Roman Law and Economics: Volume II: Exchange, Ownership, and Disputes (Oxford Studies in Roman Society & Law) 0198787219, 9780198787211

Citation preview

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OXFORD STUDIES IN ROMAN SOCIETY AND LAW General Editors   

 . . c

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OXFORD STUDIES IN ROMAN SOCIETY AND LAW The aim of this monograph series is to create an interdisciplinary forum devoted to the interaction between legal history and ancient history, in the context of the study of Roman law. Focusing on the relationship of law to society, the volumes will cover the most significant periods of Roman law (up to the death of Justinian in 565) so as to provide a balanced view of growth, decline, and resurgence. Most importantly, the series will provoke general debate over the extent to which legal rules should be examined in light of the society which produced them in order to understand their purpose and efficacy.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Roman Law and Economics Volume II Exchange, Ownership, and Disputes

Edited by GIUSEPPE DARI-MATTIACCI AND DENNIS P. KEHOE

1

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2020 The moral rights of the authors have been asserted First Edition published in 2020 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2020934649 ISBN 978–0–19–878721–1 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Preface In recent years, historians of Roman law and the Roman economy have increasingly turned toward insights from economic theory to come to a better understanding of the relationship between Roman law and institutions and the economy of the Roman world. The research of such scholars has borrowed tools mainly from the fields of law and economics and new institutional economics, and the result has been very fruitful, with an increasing number of Roman historians pursuing genuine interdisciplinary work that applies modern theoretical perspectives to understanding the development of Roman jurisprudence and legal institutions in numerous areas of the law that were important to the economy, such as contracts of sale and lease and hire, markets, credit, agency, and banking, not to mention the vast field of Roman slavery. At the same time, modern legal scholars and economists have found the ancient world, particularly the Roman Empire, to be an intriguing test case to trace the role of law and legal institutions in a pre-industrial society. The Roman Empire provides a fitting test case for theories about the relationship between law and the economy because its legal institutions are relatively well documented, while at the same time the economic performance of the Roman world has been an area of intensive study and debate among economic historians. One of the principal questions that historians of the Roman economy have been addressing in recent years concerns the performance of the economy of the Roman Empire: given the favorable conditions resulting from relative peace, political unification, and a more uniform and predictable system of courts and enforcement of contracts, to what extent did the Roman Empire experience economic growth, and, if so, how were the fruits of economic growth shared across society and to what extent did the legal institutions in place contribute and change in response to it? Roman economic historians have also focused on many other questions related to these overarching issues, such as the effects of demographic change on the economy, economic planning by Roman property owners, and the role of the state in the economy, not to mention the actual effectiveness of Roman legal institutions and the

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

vi

Preface

degree to which people in the empire relied on them as they sought to protect their own most important economic interests. The rich debate over these and many other issues provides a very favorable climate for genuine interdisciplinary research on the Roman world. This collection of essays endeavors to make an important contribution to bringing together scholars with diverse research backgrounds and thus to consolidate a new field of research, the economic analysis of Roman law (or Roman law and economics). The collection does so by drawing together scholars from a variety of fields both ancient and modern, and integrating insights from legal history, economic history, and the social sciences and, in particular, economic theory and econometrics. We hope to achieve two sets of goals. First, this collection provides a novel perspective on the function, evolution, and, possibly, rationale of Roman legal institutions. While we have no ambition to provide a systematic analysis, the various chapters offer a wide coverage of the law and institutions of ancient Rome and provide an innovative perspective of often long-studied issues. Second, this collection contributes a radically interdisciplinary methodological toolbox to the analysis of Roman legal institutions (and ancient legal institutions, more generally). Through the various chapters, the reader is exposed to a rich array of methodological approaches. Careful historical analysis of both legal and economic institutions is combined with cutting-edge theoretical and empirical examination. Next to those who, like us, are fascinated by the law and institutions of ancient Rome, we hope to interest a broader community of historians (both legal and economic), legal scholars, and social scientists. On the one hand, the set of methods that are used in this book can be applied to many more issues than we could cover in two volumes. Students of Roman legal and economic institutions will hopefully find some of these methods useful. On the other hand, we hope to have demonstrated that the information available on about one thousand years of Roman history offers an interesting natural laboratory to test the explanatory power of modern theoretical approaches. This project would have not been possible without invaluable advice, support, and encouragement we received from Barbara Abatino, Erasmo Giambona, Henry Hansmann, Elio Lo Cascio, and

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Preface

vii

Peter Temin at many crucial junctures. We are also deeply indebted to the series editors, Paul du Plessis and Thomas A. J. McGinn, and two anonymous reviewers, who guided and advised us throughout a long and intense process of revision. Georgina Leighton has been extraordinarily supportive and patient as we have worked to bring the collection into its final form. We would also like to thank Tim Beck for his energetic work in copyediting, and Chandrakala Chandrasekaran for managing production. In addition, we are grateful to Claudia Kassner and Aparna Sundaram of Columbia Law School for preparing the index. Needless to say, the greatest share of efforts has been exerted by the authors of the chapters in this book; we are extremely grateful to them all. Giuseppe Dari-Mattiacci Dennis P. Kehoe

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

List of Figures 18.1. Cesena, Italy (44°N, 12°E), showing RS demarcation persistence.

213

Map data: Google, DigitalGlobe. Image © 2013 DigitalGlobe.

18.2. Ancient Rome at its greatest extent under Trajan (c.117 ).

216

18.3. Roman system of rectangular demarcation.

220

18.4. Roman (left, a) and US (right, b) rectangular demarcation systems.

222

18.5. Depiction of a groma and mensor (surveyor).

225

18.6. Documented location of centuriae in ancient Rome.

235

Source: Museo della Centuriazione.

18.7. Roman centuriation in Italia.

236

18.8. Hypothetical Roman centuriation along a river.

238

18.9. Carthage, Tunisia (36°N, 10°E), showing RS demarcation persistence.

242

Source: Google Earth.

19.1. Choice between personal and impersonal exchange to maximize social value of productive resources.

253

Source: Adapted from Arruñada and Garoupa (2005: 717, fig. 2).

19.2. Hypothesized evolution of titling when, after distant trade becomes more common, mancipatio becomes costlier and less effective and is replaced by traditio.

272

Source: Adapted from Arruñada and Garoupa (2005: 717–18, figs 2 and 3).

22.1. Deterring wrongdoing.

358

Whilst every effort has been made to secure permissions, we may have failed in a few cases to trace the copyright holders. If contacted, the publisher will be pleased to rectify any omissions at the earliest opportunity.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

List of Tables 14.1. Manumission terms in the Digest (# of fragments).

52

14.2. Some Roman-Egyptian slave prices (in silver dr.).

62

14.3. Patron rights and the effects of manumission.

67

Adapted from Loreti-Lorini 1925.

14.4. Claims to freedman inheritance (Gai., Inst. 3.40–3.44 and 3.56; Ulp., Epit. 29.1–29.3). 17.1. Probable birthplaces of forty-one prominent Roman writers and teachers born 200 – 200.

70 177

18.1. History of ancient Rome.

217

18.2. Roman measurements with US equivalents.

219

Source: Dilke 1971; Campbell 2000.

18.3. Characteristics of selected Roman centuriae in Italia.

234

Source: see text.

24.1. Aedilician and praetorian remedies.

409

24.2. Buyers’ and seller’s valuations of the goods.

412

Whilst every effort has been made to secure permissions, we may have failed in a few cases to trace the copyright holders. If contacted, the publisher will be pleased to rectify any omissions at the earliest opportunity.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

List of Contributors Barbara Abatino, Research Fellow in the Faculty of Law at the University of Amsterdam, the Netherlands. Jean Andreau, Director of Studies at the School for Advanced Studies in the Social Sciences (EHESS), France. Benito Arruñada, Professor of Business Organization at Pompeu Fabra University, Spain. Giuseppe Dari-Mattiacci, Alfred W. Bressler Professor of Law at Columbia Law School, USA. Robert C. Ellickson, Walter E. Meyer Professor Emeritus of Property and Urban Law and Professorial Lecturer in Law at Yale Law School, USA. Richard A. Epstein, Laurence A. Tisch Professor of Law at NYU School of Law, the Peter and Kirsten Bedford Senior Fellow at the Hoover Institution, and the James Parker Hall Distinguished Service Professor Emeritus of Law at the University of Chicago Law School, USA. Iole Fargnoli, Professor of Roman Law at the University of Bern, Switzerland, and the University of Milan, Italy. Robert K. Fleck, Professor, John E. Walker Department of Economics, Clemson University, USA. Andreas Martin Fleckner, Professor of Private Law, Roman Law, and Commercial Law at Humboldt University of Berlin, Germany. David Friedman, Professor of Law at the Santa Clara University School of Law, USA. Henry Hansmann, Oscar M. Ruebhausen Professor of Law at Yale Law School, USA. F. Andrew Hanssen, Professor, John E. Walker Department of Economics, Clemson University, USA.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

xviii

List of Contributors

Ron Harris, Kalman Lubowsky Professor of Law and History at Tel Aviv University, Israel. Dennis P. Kehoe, Professor, Department of Classical Studies and Andrew W. Mellon Professor in the Humanities (2010–13) at Tulane University, USA. Egbert Koops, Professor of Legal History at Leiden University, the Netherlands. Reinier H. Kraakman, Ezra Ripley Thayer Professor of Law at Harvard University, USA. Gary D. Libecap, Distinguished Emeritus Professor of Economics at the University of California, Santa Barbara and a Research Associate at the National Bureau of Economic Research, USA. Luuk de Ligt, Professor of Ancient History at Leiden University, the Netherlands. Elio Lo Cascio, Professor Emeritus of Roman History at the Sapienza University of Rome, Italy. Dean Lueck, Director of the Program on Natural Resource Governance at the Ostrom Workshop, Professor of Economics and Affiliate Professor of Law at Indiana University Bloomington, USA. Barbara Luppi, Assistant Professor of Economics at the University of Modena and Reggio Emilia and Adjunct Professor of International Economics at the Bologna Institute for Policy Research, Italy. Thomas J. Miceli, Professor of Economics at the University of Connecticut, USA. Geoffrey Parsons Miller, Stuyvesant P. Comfort Professor of Law at NYU School of Law, USA. Francesco Parisi, Oppenheimer Wolff and Donnelly Professor of Law at University of Minnesota Law School, USA, and Professor of Economics at the University of Bologna, Italy. Daniel Pi, Visiting Professor of Law at Mitchell Hamline School of Law, USA. Eric A. Posner, Kirkland and Ellis Distinguished Service Professor of Law at the University of Chicago Law School, USA.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

List of Contributors

xix

Aldo Schiavone, Professor of Roman Law at the Università di Roma La Sapienza, Italy. Richard Squire, Alpin J. Cameron Professor of Law at Fordham University School of Law, USA. Peter Temin, Elisha Gray II Professor Emeritus of Economics at MIT, USA. Hendrik L. E. Verhagen, Professor of Private International Law, Comparative Law and Roman Law and Society at Radboud University, the Netherlands, and attorney-at-law at Clifford Chance (Financial Markets, Amsterdam).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

List of Abbreviations C. CIL C.Th. D. FIRA Gai., Inst. Inst.

Codex of Justinian Corpus Inscriptionum Latinarum Theodosian Code Digest of Justinian Fontes Iuris Romani Antejustiniani Gaius, Institutes Justinian, Institutes

Standard abbreviations are used for inscriptions, papyri, and other legal texts are used, as well as for ancient literary sources.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

12 Rome and the Economics of Ancient Law II Geoffrey Parsons Miller

The first volume of this collection enlists economic theory and methodology to explore the legal and economic system of ancient Rome—the evolution of its constitution, the organization of its markets and trading relationships, and the legal form of its business enterprises. The contributions to that volume illustrate the fruitful potential of legal-economic theory for shedding light on the institutions of the ancient world, and in particular for enhancing our understanding of the legal and economic arrangements found in the Roman Republic and Roman Empire. The Introduction to the first volume discusses the essential premises of economic analysis of ancient law, which need only be summarized here. The essential points are twofold. First, legal-economic analysis provides a useful approach to understanding economic activity in the ancient world. Such analysis starts with the premise that, while we must take into account basic social factors affecting economic choices, people in ancient societies tended to use rational means to pursue the most important things that they valued, such as status, prestige, wealth, security, and the chance for a better life for their children and loved ones. This basic assumption provides a starting point for analyzing how people in the Roman world responded to various incentives as they pursued their life goals. Second, at a sufficient level of abstraction, the problems of social organization facing jurists, law-makers, and business people in

Geoffrey Parsons Miller, Rome and the Economics of Ancient Law II In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0012

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

2

Geoffrey Parsons Miller

ancient Rome were similar to those facing their counterparts of today. The society of ancient Rome, like societies today, faced the challenge of facilitating cooperative behavior while deterring exploitation and opportunism. Roman jurists and law-makers, like jurists and lawmakers today, sought to facilitate transactions that enhance value, limit risk, and allow people the means to live a decent life under conditions of reasonable economic security. In developing norms and rules to achieve these ends, legal systems both ancient and modern create incentives for people to act in value-enhancing ways and to refrain from acting in harmful ways. This second volume carries forward the economic analysis of Roman legal and economic institutions. A fundamental institution in the Roman economy was slavery. Aldo Schiavone’s chapter, “Law, Slaves, and Markets in the Roman Imperial System,” observes that the Roman imperial economy was fundamentally dependent on slavery. “[W]ithout the presence of slaves in agriculture, in manufacturing, in service functions,” Schiavone declares, “economic and administrative life of the empire would never have reached the levels we know.” Roman law dealt comprehensively with the rights, duties, and obligations pertinent to slaves: their servitude and manumission; their purchase and sale; their obligations to their masters; their rights to own property and to make contracts; their disabilities before the law; and much else besides. The author observes that Roman jurists did not succeed in conferring legal recognition on slaves but only managed to rationalize and improve the functional efficiency of slavery as a means for enriching the slaveholding class. Manumission forms the topic of Egbert Koops’s chapter, “The Practice of Manumission through Negotiated Conditions in Imperial Rome.” Koops observes that contracts for the manumission of slaves were common in Roman law. Slaves were able in many cases to acquire sufficient wealth to purchase their freedom and negotiate contracts for manumission. Koops infers that the price of manumission was not overwhelming in many cases: a slave might be able to purchase freedom within a period of years rather than decades. Manumission in the author’s account was a transaction that advanced the economic interest of slave owners: the prospect of manumission induced diligent and trustworthy efforts by slaves who hoped for their freedom, fostered cooperation and subservience from other slaves, and promised the master continued benefits flowing from the acquisition of a freedman.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Rome and the Economics of Ancient Law II

3

Credit was a crucially important feature of the Roman economy. Jean Andreau, “Banking, Money-Lending, and Elite Financial Life in Rome,” examines the development of credit laws and markets from the last centuries of the Republic until the fourth century of the Common Era. Andreau notes that lending transactions were commonplace throughout the history of ancient Rome. Particularly enlightening is the author’s description of Roman banks, which in some respects appear to have operated in a manner not so dissimilar from banks of today. Hendrick L. E. Verhagen’s “Secured Transactions in Classical Roman Law” offers an important and controversial reinterpretation of the law applicable to secured lending transactions. The conventional view is that the Roman rules on credit were awkward and cumbersome, and that, for this reason among others, the most important source of credit enhancement was personal security such as guarantees. Drawing in part on an archive of professional Roman money-lenders, the author argues that this view is mistaken. Far from being legally disfavored, loans on real security were flexible financing arrangements that played an important role in the Roman economy. The author’s analysis is informed by a sophisticated understanding of the economic principles underlying secured lending, including concepts such as monitoring, bonding, adverse selection, and moral hazard. Robert C. Ellickson’s chapter, “Ancient Rome: Legal Foundations of the Growth of an Indispensable City,” turns to Roman urbanization, and in particular to the role of Rome as the capital of a great empire. Ellickson takes issue with the view of economic historian Moses Finley, who claimed that the enormous growth in population and prosperity of ancient Rome occurred at the expense of subject populations elsewhere. Finley’s depiction of Rome as a parasite of the Empire is fundamentally mistaken, in Ellickson’s view. Far from being parasitic, Rome exported many benefits to areas under its dominion, including government, law, literature, access to financial and capital markets, and the security and peace that came with being part of a great empire. Meanwhile, Rome’s enormous growth can be explained as a function of wise economic policy and sound institutional foundations. In particular, the government refrained from engaging in populist growth-limiting policies, but instead allowed the private market to provide housing for the population, supported by government-provided public services such as aqueducts. Ellickson’s

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

4

Geoffrey Parsons Miller

analysis of Rome’s policies towards land development suggest that cities of today might have much to learn from their ancient forebear. Economic historians Gary D. Libecap and Dean Lueck carry forward the economic analysis of land law in their chapter, “Land Demarcation in Ancient Rome.” As in the case of Ellickson’s chapter, these authors find much to admire in Roman administration of land policy. In particular, they study centuriation, a system of land demarcation with decidedly modern elements based on systematic surveying and mapping of territory. Libecap and Lueck model a decision facing Roman administrators of new territories: whether to maintain the existing demarcation systems inherited from the previous rulers, or rather to implement the new “rectangular” system. They argue that despite its higher up-front costs, the rectangular system offered superior benefits by virtue of its efficacy at reducing boundary conflicts, facilitating the development of roads and canals, fostering of new settlement, lowering the costs of transactions, and improving the identification of land parcels. Benito Arruñada’s “The Institutions of Roman Markets” highlights the role of Roman law in facilitating market transactions. Impersonal markets work best when participants are protected against idiosyncratic risks due to features of counterparties that are difficult to observe or control. On the one hand, to afford such protections, the law must dispense with the strict enforcement of rights in property (otherwise all transactions would be subject to potentially unknown claims of prior owners). On the other hand, if property rights are not well protected, people will not invest optimally in productive activities. The law must seek to find the right balance between enforcement of property rights and the facilitation of impersonal transactions. Arruñada argues, based on a subtle and nuanced analysis, that the Roman state and Roman society developed marketenabling institutions that, while not impersonal in the modern sense, at least allowed for a wider scope of transactions than would have been possible under a regime of strict enforcement of property rights. No legal system can be effectively administrated if the rules are not enforced. Richard A. Epstein’s contribution, “One Step at a Time in Roman Law: How Roman Pleading Rules Shape the Substantive Structure of Private Law,” offers a favorable view of how the Roman legal system dealt with problems of enforcement in the courts. Extending on his impressive prior work in this field, Professor Epstein cogently argues that the clear classifications of Roman pleading

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Rome and the Economics of Ancient Law II

5

provided a mechanism for improving the precision of doctrines of delict, contract, property, and restitution, and by this pathway facilitated the creation of efficient substantive rules. Epstein is alert to the implications of his analysis for contemporary public policy: modern systems of pleading, he suggests, may look to Roman models when searching for ways to improve the efficiency of adjudicatory proceedings today. David Friedman’s chapter, “Private Prosecution and Enforcement in Roman Law,” underscores the private element in Roman adjudicatory contestation. Friedman points to evidence that Roman procedures evolved out decentralized systems of privately enforced law. Drawing on the analogy to the feud system in saga-period Iceland—a subject Friedman has analyzed in prior work—the author here points to holdovers of an earlier regime of private enforcement in the rules of Roman law. These include the use of damage payments as sanctions for crimes; the delegation of the enforcement of judgments; the use of sureties to enforce legal rights; and the rule of requiring disputing parties to deposit the equivalent of bonds for the benefit of the winning party. Even in the imperial period, when the task of enforcement shifted increasingly to state actors, private citizens remained principally responsible for the prosecution of both civil and criminal offenses. Friedman’s analysis of the historical antecedents of Roman procedure is necessarily conjectural, given the paucity of historical evidence, but it does appear to be a plausible reconstruction based on the residues and traces found in later law. Modern economic analysis recognizes that the principal value of legal rules is not their efficacy in rectifying wrongs, but rather their effect at deterring people from engaging in wrongdoing in the first place. Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli’s “Deterrence of Wrongdoing in Ancient Law” addresses this issue. Like Friedman’s contribution, which traces Roman procedures to a hypothesized origin in private enforcement regimes, these authors offer a dynamic account of the growth of Roman (and other) law out of earlier antecedents. In their case, the methodology of choice is a formal economic model exploring the growth of individual liability principles out of an earlier framework of lex talionis grounded in biologically based desires for revenge. Thomas J. Miceli’s chapter, “Collective Responsibility,” investigates the topic of punishment of a group for misconduct by one of its

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

6

Geoffrey Parsons Miller

members. Collective responsibility was exemplified in the Roman context by the doctrines of quasi-delicts and noxal liability. The author models the choice between individual and collective liability as reflecting a tradeoff between the costs of failing to punish the true offender and the costs of erroneously punishing an innocent person. The general (although not total) elimination of collective responsibility from modern legal systems, the author suggests, is not so much due to the inherent superiority of individual responsibility under all circumstances, but rather to the fact that under conditions prevailing in the modern world, the benefits of collective responsibility are almost never sufficient to overcome its costs. This interesting article provides a useful corrective to conventional value judgments, and illustrates the potential of the economic analysis of ancient law to place contemporary beliefs and attitudes in a more comprehensive perspective. Barbara Abatino and Giuseppe Dari-Mattiacci’s study, “The Dual Origin of the Duty to Disclose in Roman Law,” takes on a topic with a decidedly modern resonance: the legal obligations of parties to disclose private information to their counterparties in contracts of sale. The issue is salient for contemporary commercial and business law— consider the disclosure obligations of merchants under state unfair trade practices laws; rules against trading on confidential information in stock markets; and SEC requirements for public company reporting. It is a testament to the sophistication of Roman law that disclosure obligations in commercial transactions were also regulated under that body of rules. Abatino and Dari-Mattiacci note that Roman law imposed such obligations under two institutions: the aediles, pertaining to the context of private market transactions effected through auctions; and the praetor, covering private transactions in general. The remedies differed in important respects: for example, the aedilician remedies included both restitution and enforcement of the contract plus price reductions, while the praetorian remedy required the buyer to affirm the contract but allowed offsetting claims of damages. Abatino and Dari-Mattiacci argue that both remedies were efficient in their sphere of application: the aedilician remedies maximized the value of market transactions between professional sellers and private buyers in organized markets; whereas the praetorian remedy maximized the value of transactions between private parties outside of organized markets. Like other contributions to this volume, this chapter by Abatino and Dari-Mattiacci illustrates

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Rome and the Economics of Ancient Law II

7

the potential of economic analysis to shed light on institutions of ancient law that resist analysis through traditional doctrinal approaches. This second volume of Roman Law and Economics carries forward and further validates the promise of its predecessor. In its various chapters, we observe the lived reality of an ancient legal system, sensitively analyzed and illuminated by principles of economic analysis. Taken as a whole, these studies demonstrate that economic analysis has enormous potential to enhance our understanding and appreciation of Roman legal institutions.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

13 Law, Slaves, and Markets in the Roman Imperial System Aldo Schiavone

The Roman Mediterranean Sea is the historical and natural background of the most advanced western economy before the age of capitalism—a system in which slavery is both a factor of strength and a limit to any further growth. This contradiction is also inherent in Roman law. On the one hand, Roman law anticipates some of the fundamental characteristics of the private and commercial law of modern Europe: the centrality of consent, freedom of contract, reciprocity, fairness, and good faith. On the other hand, Roman law unsuccessfully tries to overcome the classical form of chattel-slavery and ends up reinforcing its presence in Roman culture and the broader economy. The chapter supports this point of view through the study of Roman legal texts from first century BC to first century AD.

13.1. LAW, ECONOMY, AND POWER The Roman Mediterranean between the end of the second century BC and the beginning of the third century AD formed the geographic backdrop for the most vast and complex economic system ever to have existed before the Industrial Revolution and the definitive capitalistic lift-off in the second half of the eighteenth century. Notwithstanding the qualitative changes that had already taken place in the Aldo Schiavone, Law, Slaves, and Markets in the Roman Imperial System In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0013

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

12

Aldo Schiavone

fabric of the European economies since the late medieval period, it was only in seventeenth-century Holland or, possibly, about a century before, the Spanish empire of Philip II, that we find levels of production performance and wealth accumulation no longer comparable to those of the Roman world.¹ The sea was the chief protagonist of ancient commerce, and the rivers and winds along with it. Water was the only means of conveying large quantities of bulk goods over long distances. “On ships with swift sails he often traveled the great sea”: thus describes an epitaph from Brundisium one merchant’s life.² The soft technology of maritime transport (hulls, sails, rudders, knowledge of winds and currents) won out over the hard one of overland freight (wheels, carts, yokes and beasts of burden, and difficult and costly road-building). Once again, the ancient economy seems to have been more dependent on geography and anthropology than on history. Its mercantile side involved almost exclusively the coastal areas. When, beginning at the end of the second century AD, the axes of the empire in the west shifted toward the interior of Europe, the commercial networks were thrown into disarray: the fragile threads of large-scale exchanges tended to fray and break (though we must not underestimate the persistence of Mediterranean trade in late antiquity, as in the high Middle Ages), and the traditionally dominant agrarian economy began to take on overtones of feudalism. Roman law would not have been what it is if its course had not become entwined first with the development and then with the maintenance of this Mediterranean and tricontinental network of production and markets: in a certain sense, the first “world economy” in our history. Historians of law have almost always forgotten this very elementary truth, led as they are by a long tradition to reconstruct legal forms as if they developed in a world apart, far removed from society, economics, and politics; but such an obstinate concealment in no way lessens the absolute evidence of the connection.³ It is less easy, however, to identify the nature of this relationship. One should not in fact imagine any unilinear causality, in one

¹ The economic history scenarios outlined in the text presuppose my work in Schiavone (2000). ² Corpus Inscriptionum Latinarum IX 60: see Giardina (1993: 261). See also the book of Temin (2013). ³ See my work in Schiavone (2012).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

13

direction or another: from economics toward law, or from law toward economics. It is necessary instead to hypothesize a complex reciprocity of interactions, where every innovation—in the field of law or economics—was at once cause and effect of further developments occurring on both fronts. Without fear of exaggeration, we can safely say that law was an invention of the Romans: in the sense that they were the first to isolate the legal function as a specific technique of social regulation, detaching it from every other kind of cultural production or institutional apparatus—from religion, from morals, from politics itself—and identifying it instead with an entirely autonomous field, with its own principles and structure: the world of ius. At the center of this universe were the jurists (a word unknown to any ancient language, except for Latin), who, at least in the centuries of Rome’s imperial heyday, were not just the adepts and interpreters of ius, but, in large part, its creators as well: a body of experts who labored for dozens of generations, moving along a previously untrod path. Admittedly, besides the work of the jurists there was in Rome, in that period, also a significant presence of legislation—from the Twelve Tables, still in the archaic age, to the laws passed by the people’s assemblies in republican and Augustan times, through to the normative provisions decided by the emperors during the Principate. And a no less important role was played, in the same epochs, by the praetor’s edict: a text formed through an uninterrupted stratification that had involved, year after year, hundreds of magistrates responsible for civil jurisdiction (the praetors). Yet this many-stranded whole was never concretely operative outside the thick weave of prescriptions, interpretations, innovations, and amendments that would be built around it by jurisprudence. “Law cannot exist unless there is some jurist who can improve it from day to day,” wrote Pomponius in the Enchiridion (in D.1.2.2.13) around the middle of the second century, when that model had already begun to slip into crisis. What we generally think as “Roman law” is therefore first and foremost a construction of the jurists. A “living law of custom,”⁴ case based and guided by experts: much closer in its structure, in its ⁴ The expression comes from Carl Schmitt (1950, who epitomized Savigny), and it is used in reference to English law. See also Buckland and McNair (1952).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

14

Aldo Schiavone

mature configuration, to modern English and (in a number of ways) American law, than to French law after the codification under Napoleon, or to Italian law subsequent to the codification of 1865. In the epoch of the great conquests—between the second century BC and the age of Augustus—the collaboration between jurists and praetors gave rise to a kind of “Mediterranean trading law,” which immediately became a decisive instrument of Roman hegemony. The imperial economy never built a unitary structure: nothing like the capitalistic unification of contemporary economies. Rather, it involved a constellation of diverse systems, determined much more by the history and geography of the different places than by the force of the impact and presence of Rome. But the empire, through the lever of tax levies, and due to the initiatives of avid, unscrupulous, and enterprising groups of merchants, managed to create an unprecedented network of trade exchanges and wealth redistribution mechanisms, which contributed in a decisive way to stabilizing and maintaining Roman power. This network required a framework of relations that it would have been impossible to build and consolidate without the experience of Roman law, which gave that context, albeit through many mediations, the certainty of a juridical foundation capable of protecting and regulating, in a rigorous and specific manner, property expectations, single and group interests, dealings, markets. In turn, the presence of such an extensive and developed economic fabric offered an inestimable stimulus for legal thought— an almost endless supply of different cases and situations—and, above all, limitless proof of an essential lesson: that to govern so very diverse spaces and peoples, consensus was much more necessary than armies, and law, then, more than force. But the picture was in reality even more complex. A third element present on the scene—besides law and economy—should also be borne in mind: politics. In fact, the functioning of a Mediterranean economy presupposed, in turn, imperial political unification: in the ancient context, it was almost always politics that underpinned economics, and not vice versa, as we are accustomed to seeing in the modern world. And what has just been said of economics holds true for politics as well. The institutional construction of the empire would have been inconceivable if a decisive role had not been played by Roman law. The latter, in turn, would never have reached the degree of refinement and formalization that we know, and that enabled it to perform its essential function, if its development had not been

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

15

constantly accompanied by an unprecedented accumulation of political power, concentrated in the circle of the same elites that were elaborating the law—if the great imperial legal culture had not taken root in the shadow of a world power. The growth of political strength, economic successes, and the disciplinary capacity of law— power and order—were inextricable strands of the same hegemonic construction. The birth of a new paradigm within Roman legal thinking, which transformed law into a science, changing the quality of its precepts from acts of authority into acts of knowledge—an epistemological revolution destined to mark the history of the West from then on—would have been impossible if it had not occurred within this intertwined web: a picture about which historians still have a lot to discover, and a lot to recount.⁵ Roman jurisprudential law was an order with a fluid, changing surface, constantly moving in the attempt to grasp as closely as possible the social reality it faced and had to regulate. The shared core of a scientific and preceptive paradigm that had been precociously developed was constantly riven by polemics between different jurists, disagreements and divergent solutions. Partly due to this intrinsic instability, and for many other reasons besides—both political and cultural, as well as more strictly technical and practical—the Roman case law shaped by the jurists was never really the law of the whole empire. It was not the law of the whole empire in a “horizontal,” geopolitical sense, because wherever possible Roman imperial praxis envisaged the regulatory independence of the local communities, which were to govern themselves “according to their own laws and their own traditions”—suis moribus legibusque suis uti.⁶ Nor was it in what might be termed a “vertical,” social sense, because Roman law almost always remained the law if not just of the elites, then at least of the property-owning orders—in Rome itself, in Italy and in those provincial contexts where Romanization had penetrated most strongly. It never became, in other words, if not following heavy mediation, the law of the vast, outlying territories under Roman domination, from the Red Sea to the Atlantic Ocean; nor, by the same token, was it the law of all layers of the population, of the entire social body.

⁵ See Schiavone (2012: 245). ⁶ Gellius, Noctes Atticae 16.13.4, referring to Hadrian’s oratio ad Italicenses.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

16

Aldo Schiavone

The law of the Roman jurists always remained an extremely circumscribed model, in terms of its effective range of application, but carried great prestige and a potent charge of exemplarity: a bright, luminous center surrounded by a myriad of minor and local legal systems, sometimes lacking any formal recognition, which assumed Roman law as a distant and unattainable point of reference that still had to be taken into account, even if only in a rough and ready manner. The activity of Roman jurisprudence had developed entirely within the heart of this economic and political supremacy, and it would be incomprehensible now if we did not trace it back to the everyday exercising of a hegemony that it helped so greatly to construct. And the jurists themselves frequently became embroiled in the management and the problems of a world government. In the republican age, insofar as they all belonged to the nobility, they were also at the head of the Republic. During the Principate, first as advisors to the princeps, then directly as high-ranking functionaries of the imperial administration (the entire history of Roman legal thought can be drawn together around three pictures, each with the same protagonist, though with different dress: that of an archaic priest; of a republican nobleman; and finally, of a great specialist at work in the milieu of the princeps and the court). But the science of the jurists seems never to have left the slightest trace of such involvement in the governing of the empire, and they always managed to avoid becoming shackled by too close a bond between legal learning and political action. On the contrary, they were capable of defending the isolation of their knowledge, rebalancing it as a technique that was selflegitimating well beyond the underpinnings of political power, even though entirely constructed right next to it. And in turn, politics would always find itself in need of alliances and compromises with jurisprudence, in which the relative equilibrium of powers was determined by the conditions prevailing at any given time.

13.2. THE SLAVE SYSTEM The Roman imperial economy had a trait that makes it quite distinctive and totally recognizable: it was a slave economy, like the one of classical Athens between the fifth and fourth centuries BC, or that of

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

17

the Southern U.S. states before the Civil War. It is hard now to doubt the evidence of this, though often in the past “the ideologies of the moderns” (to use an expression of Finley 1980)⁷ have tried to mask this incontestable truth, and many have worked to soften history and to prevent the image of the classical world—partially invented by themselves in libraries—from being sullied by such a stain. But even just from a quantitative point of view, the slightest hesitation would be unthinkable. There are grounds to sustain that in the years of Augustus, at the end of the first century BC, perhaps over 30 percent of the entire population of Italy was made up of slaves—around three million in total, a massive figure and a very high proportion, maintained for over a century. And this is without mentioning the incidence of slave labor on the structure and performance of the production fabric: without the presence of slaves in agriculture, in manufacturing, in service functions, the economic and administrative life of the empire would never have reached the levels we know. For sure the situation outside Italy was different, and we cannot expect such figures to be matched in the peripheral areas as well. But the presence of a considerable slave population can be clearly traced in Sicily, Cisalpine Gaul, Africa, in the Spanish and Greek mining regions, in Gaul, especially in the southern areas, and almost all of the eastern provinces. This was especially true of the large coastal cities: no less than one hundred thousand slaves lived in Alexandria (around 25 percent of the local population). And in Antioch in the fourth century AD some wealthy citizens still possessed between one and two thousand slaves each. Nor have we any reason to think that the picture was very different in Carthage or Marseilles.⁸ Slavery was not just an ancient phenomenon. But modern slavery was entirely colonial in origin—in Brazil, in the Caribbean, in the American South—and it became established for particular reasons (the scarcity of labor in the New World, accompanied by the European penetration along the coasts of West Africa) only in rural environments that were relatively peripheral with respect to the increasingly manufacturing-based and industrial center of the new European and Atlantic economy. New-World slavery always had ⁷ See also Andreau and Descat (2006: especially 65ff., 107ff.). A vast and accurate bibliography (until the beginnings of the 2000s) in Bellen and Heinen (2003). ⁸ Schiavone (2000: 112–13). See also Bradley (1994: esp. 10ff., 57ff.); Turley (2000); and Thompson (2003).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

18

Aldo Schiavone

problems of compatibility in relation to a different and much more expansive and dominant mode of production—both agricultural and industrial—founded on free labor. It was, in short, a recessive figure in comparison to the heart of modernity. In the ancient civilizations, by contrast, in classical Athens as in imperial Rome, the slave system was by far the most developed and dominant economic form in terms of results and organization: the real propulsive core of the whole Mediterranean economy. And it never basically had any alternative, either theoretical or practical. From Plato to Aristotle and through to the Roman jurists of the Severan age, it was constantly viewed as a social necessity, a universal principle of the civilizing of humanity, as something that could not at any rate be questioned. Indeed in this respect Roman thinking—both philosophical and legal—represented an undoubted novelty in comparison to the Greek tradition. Rather than insisting on the truth of an ontological difference that, “by nature” (kata physin), divided the human species into those who were born to command and those to obey, as sustained in a celebrated passage from Aristotle’s Politics,⁹ Roman thought, at least from Cicero onwards,¹⁰ preferred to abandon the rigidity of this model and turn to a justification of slavery that we might call sociological and historical rather than fully naturalistic: at any rate better suited to the reality of the slave system it was faced with, distinguished not only by the annihilation of personal identity and harsh repression, but also (as we shall see) by the valuing of the abilities and talents of slaves, and the regular concessions of freedom. It involved, in other words, an adjustment dictated by imperial reality: many slaves arrived from regions, like those of the eastern Mediterranean, that were culturally more advanced than the society of the conquerors, and speaking of an ontological diversity destining them to slavery would have been embarrassing. Unlike the Greek model of slavery, the Roman one never, or almost never, had specifically racial overtones. This fully explains how jurists between the second and third century AD could affirm that men should be classified as either free or slaves (and from the legal point of view, it was as if slaves did not exist, being closer to things than to human beings), but at the same

⁹ Arist., Politics 1253b–1255a.

¹⁰ Cic., De rep. 3.25.37.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

19

time that slavery did not rest on natural law (in relation to which they even went so far as to say that “all men are equal”), but only on that “of the peoples”:¹¹ it was, that is, the result of a unanimously accepted convention, not the reflection of a natural datum. In any case, the idea of a society without servile labor formed no part of the ancient Mediterranean cultures. Nor did the great slave revolts of the second and first centuries BC (in Sicily and in Italy, including that of Spartacus) set themselves such an aim.¹² They just sought to overthrow local power arrangements, and to exact vengeance on inhuman masters, not to uproot an overall system. Neither philosophical thought, nor political or legal experience offered any points of reference. And, what is more, in all the criticism raised about the uncontrolled spread of chattel slavery and the harsh conditions it entailed, the suggested alternative looked to an archaic and idealized past, rather than to the present or to the future. It just envisaged the return to a world of domestic autarchies, with a softer form of slavery, numerically limited and “according to nature,” incorporated within the family structures of former times—visions fed solely by literary commonplaces concerning the revival of the virtuous poverty of the ancestors.¹³ Such models were unable to do anything other than create a yearning for archaizing restorations, in which the rejection of the excesses of slavery inevitably carried with it an aversion towards large-scale trade, money, and mercantile networks. Basically, towards everything that was now an inerasable feature of Roman civilization, delivering the economic performance that underpinned the whole social structure of the empire: urban life, mass consumption, military and administrative bodies, intellectual creativity. And so, after brushing up against a dangerous critical point with the bloody insurrections of the previous one hundred years, the neoaristocratic stabilization that characterized Roman society with the advent of the Augustan regime was accompanied by a no less imposing stabilization of the imperial slave system. It thus consolidated its dual and contradictory aspect as an indispensable requirement for

¹¹ D. 1.1.4 (Ulp. 1 Inst.) and D. 50.17.32 (Ulp. 43 ad Sab.); D.1.5.4.1 (Florentinus 9 Inst.); D.12.6.64 (Triphoninus 7 disp.). ¹² Schiavone (2013). ¹³ I am referring in particular to the thought of Posidonius, as inferred from Athen. 6.84.263c–d = FGrHist 87 F8 = Edelstein-Kidd F 60.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

20

Aldo Schiavone

maintaining the levels of intellectual and material life achieved by that society, and at the same time as an insuperable barrier to any further development, which would instead have presupposed a massive spread of free labor, as would occur in the modern economies. Slavery underpinned that world and yet thwarted it. The terrible coercion that relentlessly accompanied the existence of millions of slaves was never substantively eased, nor was the ferocious discipline to which they were subjected. A measure introduced at the behest of Augustus himself reaffirmed, for example, that in the event of the homicide of a master, all the slaves living “under the same roof ” as him were to be tortured and put to death, because, as a great jurist Ulpian would explain two centuries later, “no house could otherwise ever have been safe, if not by obliging slaves, on pain of death, to defend their masters both from dangers arising within his home and from without.”¹⁴ In AD 61, the intransigent application of this norm, despite the disguised opposition of the emperor himself—Nero, probably influenced by Seneca—would lead to the summary execution, in the heart of Rome, of four hundred slaves, including many women and children.¹⁵ But such ruthlessness did not prevent a great many slaves, especially in the cities—hundreds of thousands perhaps, in the course of the whole of Roman history—from becoming integrated into imperial society, and not all just at the lowest levels, and from being accepted not as an extraneous body, but as an active and vital element. Such a tendency reveals the extraordinary Roman talent—as yet unequalled in world history—for the unprejudiced assimilation of different cultures and peoples, provided they participated in some way in the great process of unifying the empire. Terror and integration, then: these were the only apparently contradictory pillars that underpinned the Roman slave system. And both were analytically regulated by law. The most significant form of integration consisted of manumissions (manumissiones). Through them, every single master, without any intervention from the political authorities, could give a slave his liberty, making him a freedman (libertus): and this guaranteed him ¹⁴ D. 29.5.1 pr. and 26–7 (Ulp. 50 ad ed.). See also Tac., Ann. 14.40.1–44.4. About the sc. Silanianum see Dalla (1994); the recent contribution of Harries (2013: 51ff.); and Miceli in this volume. ¹⁵ Schiavone (2000: 108ff.).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

21

at the same time Roman citizenship (it seems incredible, but has a precise historical explanation). In the final decades of the Republic, the practice became extremely widespread. Many masters preferred to have free citizens bound to them by legal and economic obligations, as freedmen were, rather than elderly slaves who had to be maintained. And it was Augustus, once again, who had to establish that no more than one hundred slaves could be manumitted in a will, though he left it to the master’s discretion as to how many to free while he was alive.¹⁶ In the Rome of these years, tens of thousands of citizen had servile origins. Later, in the age of Nero, there was talk in the senate, according to Tacitus, that the majority of knights, not to mention many of the senators themselves, had some blood deriving from a freedman.¹⁷ So there took shape an elastic and multiform space of cultural and social intertwining, of relations and exchanges, shared by the world of the free and by that of slaves. And it was precisely in this circuit that the most lively part of the imperial economic system developed.

13.3. LEGAL THOUGHT In Roman jurisprudential texts from the time of Servius Sulpicius Rufus, at the end of the Republic, through to the Severan authors at the beginning of the third century AD, slavery occupies a position of great prominence. This centrality appears entirely evident to us, even though two important elements have long worked together to partially obscure it. The first concerns the tradition of the available documents. We owe much of our knowledge about Roman legal thought to what has been transmitted to us from Justinian’s Digest, compiled in the heart of the Byzantine age, in the first half of the sixth century AD. In the time and context in which the compilers were laboring on this great collection—at once a code and an anthology, or rather an anthology in the form of a code—slavery had become a much less crucial phenomenon, socially and economically, than it had been in the age of the ancient jurists at the height of the Roman Empire. It is therefore ¹⁶ With the lex Fufia Caninia of 2 BC, which covered all the various cases possible with great precision (one hundred was the maximum): Gai., Inst. 1.42–3. ¹⁷ Tac., Ann. 13.27. See also Hopkins (1978: 115ff.).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

22

Aldo Schiavone

reasonable to think that the Byzantine compilers, in their choice of materials, tended to attenuate rather than emphasize the original preponderance of references to slavery: a point of view that by now appeared to them to be decontextualized and lacking in topical concern. If, notwithstanding this filter, slaves still stand out so powerfully in those ancient writings, it means that their invasive presence had in any case become an integral and insuppressible part of jurisprudential thinking: preserving essential nuclei from those works also meant rendering their slave-based orientation inerasable. The second element relates instead to our perspective as moderns. For a long time European legal culture approached Roman law with practical much more than strictly historiographical intent, the objective being to extract from that legal order—updated where absolutely necessary—rules for the age. In this massive effort of actualizing interpretation, which reached a peak in nineteenth-century bourgeois Europe, the original slave-oriented imprint no longer had any utility; indeed, as far as possible it was worth concealing and passing it off as an accidental and negligible aspect, highlighting instead what was supposed to be the supratemporal dimension of that law, its capacity to regulate very different and distant societies in the best possible way. Yet despite this dual conditioning—first in the Justinianic selection, then in modern interpretation—the reality of things came to the fore: every strand of Roman law was shot through by the presence of slavery, which indeed constituted the chief condition of existence of its most advanced component, both practically and conceptually. Slavery disciplined mercantile dealings and the circulation (and accumulation) of commercial capital. What is more, this observation had already been made at the beginning of the twentieth century, opening what can still be considered the most important book on the theme in modern historiography: The Roman Law of Slavery by W. W. Buckland, published in Cambridge in 1908 (and then reprinted in 1970). In the Preface we can read: “There is scarcely a problem which can present itself, in any branch of the [Roman] law, the solution of which may not be affected by the fact that one of the parties to the transaction is a slave, and, outside the region of procedure, there are few branches of law in which the slave does not prominently appear.”¹⁸ We would not say anything different today.

¹⁸ Buckland (1908: v). See also Watson (1987).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

23

13.4. PERSONAL DEPENDENCY AND ECONOMIC AUTONOMY Servius was the first of the great Roman jurists to cast an expert and attentive eye on the economic reality of his time. We are in the middle years of the first century BC, and imperial society now appeared to be undergoing a radical transformation. We are in the heart of the Roman slave-holding and mercantile boom, at the culmination of a wave of expansion without precedent in the ancient economies: a process of growth still tied to limits and contradictions that kept it from creating the foundations for an authentic capitalist take-off, but sufficiently powerful to achieve a level of performance that, as we have said, was long unrivaled in the West: new farming estates, with impressive standards of productivity; factories created exclusively for the production of commodities; large quantities of slave labor; a spectacular accumulation of public and private money; urban centers supported by magnificent building programs; trade networks that had formed an interdependent grid of markets extending across the entire Mediterranean; opulent levels of consumption, though concentrated among tiny elites. The slave trade, in particular, was one of the most successful commercial activities in the Roman Mediterranean, which was also, no less than the Atlantic of the slave trade, a sea of slaves. In the second and first century BC—the golden age of imperial slavery—a high level of supply was maintained by the virtually uninterrupted sequence of victorious wars of conquest (thousands upon thousands of prisoners were sold to merchants following in the wake of the troops, then resold at higher prices) and the indefatigable activities of organized bands of pirate slave traders—the same ones who, having become too powerful and a danger to navigation, were eradicated by Pompey. The port of Delos (on the island sacred to Apollo) was perhaps the largest marketplace; according to Strabo, up to ten thousand prisoners could be sold there in a single day.¹⁹ The writings of Servius—who had been quaestor in Ostia at the beginning of his political career, and was thus able to gain a very good idea of what the trading reality of the empire had become—are the mirror of this world. Unfortunately, Justinian’s compilers did not use

¹⁹ Strabo 14.5.2 (Meineke).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

24

Aldo Schiavone

any work of his in assembling their collection (perhaps they no longer even had a copy), and we cannot read anything by him through the Digest. But the latter does include seventy-six texts taken from the Digest of Alfenus Varus, the first jurist to write a work with that title, and the most important, along with Aulus Ofilius, of Servius’ pupils.²⁰ Alfenus reproduced a great deal of Servius’ thought, enabling us to regard it as a full-blown edition, with commentary, of the responses of the master himself. And even if we have some difficulty in exactly distinguishing each time between the commentary of the pupil and the thought of the master, we can consider these documents to be reliable testimony of Servius’ ideas. At any rate, in at least twenty-six of these texts,²¹ at the center of the economic issue underlying the legal problem, we find the presence of a slave. This is a very high proportion: it is as if Servius had discovered the pervasiveness and essentiality of slavery in the reality of his own world. There is no doubt that in Servius’ eyes a slave was above all else a living machine, a thing, a commodity—a “speaking tool”²²—in keeping with a tradition that dated back at least as far as Aristotle.²³ But in the eyes of the jurists, this state of pure thingness, though accepted as unquestionable, did not fully describe the condition of a Roman slave. Alongside the institutional annihilation of their humanity (Quod attinet ad ius civile, servi pro nullis habentur—“as far as the civil law is concerned, slaves are regarded as not existing”—Ulpian would still write in a passage of his Libri ad Sabinum),²⁴ for jurisprudence there was also a need, dictated by the empirical rationality of the slave system which the Romans had developed with particular attention, to valorize the abilities and talents of the subjected, exploiting in the best possible way potential and aptitudes that did not fit into the scheme of their reduction to simple things. In this way there opened up what seems to us to be a glaring contradiction, but which legal thought, with a ductility that did not fear incongruence but pursued only the optimization of results, did not regard as such, but rather as a kind of opportune coexistence of a dual slave regime: one centering on the retention of the bond of personal dependence, which was never denied; and the other founded ²⁰ Lenel, Pal. I. 38–54. ²¹ Schiavone (2012: 522 n. 22). ²² As Varro put it in Res. rust. 1.17.1. ²³ Pol. 1253b; NE 1161b. ²⁴ See n. 11.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

25

on the valorization of the labor force and the productive capacity of slaves. In this respect, they were granted such a degree of autonomy in their economic activities that they could almost be mistaken for free men: a sort of permanent state of exception, valid as long as one remained within the sphere of production and markets. It was an entirely Roman peculiarity—nothing like it had been seen in Greek slavery. In the systematic scheme of Gaius (a jurist from the second century AD, and the author of a manual destined to enjoy extraordinary posthumous success), which was probably influenced by a scholastic approach that may also have been marginal with respect to the chief lines of development of the legal thought of his time, but is in case revelatory for us of a widespread attitude and tendency, the slave appeared to lie at an ambiguous point of intersection between personae and res, between “persons” and “things”: an intersection that marked the critical point of a whole civilization.²⁵ An intersection that fixed the atypical and exceptional status of a “human thing” where two different ontologies were bound up to the point of being indistinguishable, in a sphere of reciprocal influences without equal in ancient civilizations. It was as if, in the jurists’ doctrines, slavery, with all its power to annihilate identity, appeared not only to be an advantage— according to the common notion of any slave system—that permitted the total exploitation of the labor of the subjected at a very low cost, but also ended up becoming, in a completely unexpected way, a brake as well, a kind of obstacle to be overcome in order to permit that combination of human resources and trade circuits which the classical form of chattel slavery would never have made possible. Servius was the first to conceptualize this phenomenon— subjection in terms of “status,” but autonomy with regard to commercial transactions—and to employ a formal device to account for this particular condition. In a response, published by Alfenus Varus, we read: “A person leased land to a slave of his for cultivation, and provided him with some oxen. . . . ”²⁶ If we were to take as the sole point of reference the relationship of dependency between master and slave, this sentence would make no sense. The slave was totally subject to the master’s will and so could not enter into any form of economic ²⁵ Gai., Inst. 1.8–9; 48–49; 52; 2.13. See Esposito (2014: 9ff.). ²⁶ In D.15.3.16 (Alf. 2 dig.): Quidam fundum colendum servo suo locavit et boves ei dederat.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

26

Aldo Schiavone

or legal exchange with him. And yet, both Servius and Alfenus took the existence of such a relationship for granted, and in defining it, they did not hesitate to use a technical legal verb, locare, that placed it in the juridical scheme of the res locata. This was a conceptually subversive qualification, in which slave and master figured as distinct subjects, facing off against one another on a plane of formal equality. Though restricted to well-defined limits, and only with regards to transactions in progress, the formal determinations induced by the legal models of exchange overlapped with and prevailed over those of the relationship of dependency: “a person leased land to a slave of his”—servo suo locavit. The contrast could not be expressed more effectively; in these few brief words two opposite worlds really come face to face: the discriminating one of status, and the inclusive one of the contract. Servius’s idea was to construct in an analogical fashion, as far as seemed possible, the economic relations between slave and master as if they were between free men, on every occasion in which the developments of the new economy, and the need to enhance the value of the slave’s abilities, intervened to modify the typical and traditional structure of the tie of dependency.

13.5. A DUAL REGIME Servius’ account did not only speak about leasing. In the course of the event described in the same text, we see the slave selling and buying, receiving sums and failing to pay them, and becoming insolvent.²⁷ None of this was rare in that time. In the context of a broad and detailed series of cases—carefully analyzed by jurisprudence—we find slaves commanding ships, running shops, administering estates, looking after libraries, and directing commercial enterprises just like managers.²⁸ ²⁷ “ . . . but as the oxen proved unsuitable, he told the slave to sell them and to buy replacements with the money he got for them. The slave sold the oxen and bought replacements but became insolvent without paying the person he bought them from” ( . . . cum hi boves non essent idonei, iusserat eos venire et his nummis qui recepti essent alios reparari: servus boves vendiderat, alios redemerat, nummos venditori non solverat, postea conturbaverat . . . ). ²⁸ Di Porto (1984) and Carandini (1988), besides, naturally, the previously cited work of Buckland (1908).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

27

Strata of privileged slaves thus formed in the servile condition, especially in the big cities, who lived in circumstances of often considerable well-being, free from physical coercion and with full mastery of their bodies. For many of them, the availability of a patrimony was accompanied by effective recognition of the family units that had formed, and respect for their inviolability. At the same time, a minutely detailed regulation of manumissions opened up further perspectives, and in many cases came close to the goal of a liberty that could mean the attainment, even in the space of a single generation, of a social level that would provide a shield from the most serious dangers. The dual regime that governed, at least potentially, the servile condition, personal dependency and economic autonomy, ended up modifying, as we have just seen, not just the relations between the slave and a third party, but also those between slave and master, reaching the very core of the bonds of dependency. And that was not all. There was, in fact, a habit, already fully consolidated at the end of the republican age, whereby a slave was effectively permitted to possess a patrimony (peculium) independently from the carrying out of a specific commercial activity. This might consist of money or various goods—even other slaves—that in principle belonged to the master, but was actually managed freely by the slave, so much so that it could even be used to buy his freedom. The exact determination of the peculium could prove to be important in various circumstances, and indeed Servius devoted himself to that task with great care, as we learn from Ulpian: “Everything owed by the slave to the master is deducted before the peculium is valued,” he wrote, before going on to say that, to this “definition,” Servius “adds that one must also deduct anything owed by the slave to people in the master’s power, as this too is admittedly due to the master.”²⁹ We cannot say with certainty who was the first to formulate the definition mentioned by Ulpian, to which Servius “adds” a second part. But certainly it drew almost literally on a rule already formulated by Servius in the text mentioned earlier (“it did not seem to form part of the peculium, if not the balance remaining after deducting what the

²⁹ D. 15.1.9.2–3 (Ulpian, 29 ad ed.): Peculium autem deducto quod domino debetur computandum esse. . . . Huic definitioni Servius adiecit et si quid his debetur qui sunt in eius potestate, quoniam hoc quoque domino deberi nemo ambigit.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

28

Aldo Schiavone

slave owed the master”).³⁰ Evidently, then, the jurist had worked to achieve a genuine doctrinal organization of the patrimonial relations between slave and master, designed to obtain as great a flexibility as possible, in the light of the new requirements taking shape in the Roman economy. There were, however, some insurmountable barriers. Actual circumstances, acknowledged on the fringes of law in order to permit slaves to operate in the commercial reality to the best of their abilities, could not translate into formally recognized juridical obligations of the master towards the slave. Let us look at a further response of Servius, this time transcribed by Aufidius Namusa, another of his pupils: “a master left his slave five gold pieces thus: ‘let my heir give to my slave, Stichus, whose freedom I have directed in my will, the five gold pieces which I owe him by my accounts.’ Namusa reports Servius to have been of the opinion that no legacy was due to the slave, because a master could owe nothing to his slave.”³¹ The legal recognition of economic relationships within the bond of slavery ran up against an impassable asymmetry. It is the other side of the coin to Roman flexibility that we can see operating here: that of slaveholding intransigency. The admission of a debt on the part of the master counted for nothing, even if it was certified in his accounting books: to admit its existence in terms of law would have meant pushing the analogy to the point of accepting the possibility of a legal obligation on the master’s part toward the slave. If such a step had been taken, one of the fundamental presuppositions of the whole slaveholding mechanism would have been swept aside. The fact that the binding force of dependency appeared, in the context of Servius’ reasoning, only as a limit and obstacle to the legal qualification of economic relationships completely integrated into everyday life is an indication of the degree to which the jurist’s thinking was capable of capturing (“which I owe him . . . because a master could owe nothing . . . ”) the sharpest contradiction that the growth of trade had introduced into the fabric of a society based on slavery: the contrast between the ³⁰ Again in D. 15.3.16: . . . non videri peculii quicquam esse, nisi si quid deducto eo, quod servus domino debuisset, reliquuum fieret. ³¹ This text can be read through a citation by Iavolenus Priscus, 2 ex post. Lab., in D. 35.1.40.3: Dominus servo aureos quinque eius legaverat: “heres meus Sticho servo meo, quem in testamento liberum esse iussi, aureos quinque, quos in tabulis debeo, dato.” Nihil servo legatum esse Namusa Servium respondisse scribit, quia dominus servo nihil debere potuisset.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

29

desirability of valorizing productive forces, which would have required the setting aside of personal relationships of dependency, and the presence of a generalized form, without any practicable alternatives in that world, of chattel slavery.

13.6. PRAETORS AND SLAVES Another insuperable obstacle imposed by the persistence of the relationship of dependency was the absolute impossibility for the slave to seek legal redress or to be summoned himself before the law—and little did it matter if the trial concerned his own master or a third party. But the praetors responsible for the administration of civil justice, in close collaboration with the jurists, tried to get round this difficulty as well. To deal with the problem they invented a trial device that permitted the assignment to the masters of legal responsibility for business affairs conducted by slaves, without, however, canceling the role played in them by the latter. The technical expedient consisted of indicating the name of the slave in the part of the judicial “formula” in which the event was described (it involved a simple statement of the circumstances that had led to the dispute, and mention of the slave and his activities did not entail any recognition of his fitness to plead), and the name of the master in that part of the “formula” envisaging the conviction or acquittal of the defendant—which could not regard a slave without denying his condition, but only a free man. It was a manipulation carried out at the very margins of ius, utilizing all of its ancient and shrewd formalistic wisdom, a sort of exchange within the ceremony of the trial, that in some way made it possible, albeit obliquely, to incorporate personal dependency and participation in commercial networks into the slave’s activity. In the space of a few decades the edict issued by the praetor at the beginning of his mandate and modified year by year, which established the formulas of the actions permitted to those who, with the appropriate prerequisites, had requested them, reached the point of making provision for a series of cases that ended up offering legal protection for the most significant economic enterprises of slaves. These ranged from agros colere (farm management) to pecuniam faenerare (credit lending), to mercaturas redempturasque facere

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

30

Aldo Schiavone

(mercantile activities), according to the significant list proposed by Labeo (the greatest jurist of the Augustan age) and recalled by Ulpian.³² The essential (and probably original) nucleus was made up of actions de peculio (“on patrimony”) and de in rem verso (“on enrichment”)—which envisaged a single formula with two possible variants, already the subject of reflection by Servius and Labeo³³—whereby the master was called to answer for commitments made by his slaves within the limits of size of the peculium granted to the latter, or, alternatively, the enrichment gained as a result of the activities undertaken by the slave. To these, more or less in the same period of time, were added the more specific formulas of the actions exercitoria (for maritime trade), institoria, quod iussu, and tributoria (for other cases of non-maritime or speculative trading activity).³⁴ The result led to a kind of commercial law of slavery that was unparalleled in any other slave-holding society, ancient or modern. That was, moreover, the age in which the creative force of the praetorian edicts had reached its peak, in a constant effort by the aristocratic elites holding legal knowledge to adapt to and anticipate the new needs of a world dominion. A handful of solid principles, evident but never explicitly formulated, guided their work. First, consensualism (“consensus”): that is to say, acknowledgment of the agreement between the parties, however manifested, provided it was demonstrable, and expressed within the context of a typology of transactions rigorously envisioned in the edicts. Then, reciprocity (ultro citroque obligatio in the lexis of Labeo),³⁵ according to which, in a transaction, any economic performance by one of the parties should be matched by a symmetrical performance of the other party. Next, the idea of good faith (bona fides), that is to say the commitment to a reliable and trustworthy form of behavior, based on the

³² D.14.3.5.2 (Ulp. 28 ad ed.). ³³ As far as Servius is concerned, these actions are dealt with in the final part of the already-mentioned text in D. 15.3.16 (qui boves vendiderat nummos a domino petebat actione de peculio aut quod in rem domini versum esset . . . —“that person now claimed the price of the oxen from the master by means of an action ‘on the peculium’ or ‘on enrichment’ . . . ”). For Labeo, see for example D.15.3.1.1 (Ulp. 29 ad ed.), who reports the thinking of the Augustan jurist. ³⁴ See the reconstruction in Di Porto (1984: 31ff.). On the so called actiones adiecticiae qualitatis, see de Ligt (1999: 205ff.) and, more important, Miceli (2001). ³⁵ Cited by Ulp. 11 ad ed. in D. 50.16.19.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

31

Roman tradition, as a way of limiting Mediterranean mercantile cunning. Finally, and in many ways the most important of all, fairness (aequum): a criterion whereby, in the assessment of a case for trial purposes, the literal application of the traditional law (ius civile) could be set aside in favor of a more flexible measure that was better suited to the interests and values at stake. Roman law thus took on its consolidated guise: a law made up of cases and judicial actions, constructed day by day through the very long collaboration between magistrates and jurists.³⁶ It was, in short, imperial power that turned into the effectivity of a more mature legal order, which in turn favored the accumulation of new power, in a spiral that seemed to have no end.

13.7. CONCLUSI ON In appearance, it seemed that not so much was required for the decisive step to be taken, and for slavery to dissolve, as it were, from within, destroyed by a sort of precapitalist maturing of economic networks, guided and supported by a precursory legal reflection: and on certain occasions the jurisprudence between Servius and Labeo really does give the impression of being on the brink of that threshold. But that jump was never made, and Roman slavery did not slip into a gentle transition toward dominant forms of free wage labor. It was also a question of numbers: the new developments introduced by the praetors and re-elaborated by the jurists concerned only a minority— albeit not small, as we have seen, and the focus of great attention as well—of the millions of slaves who lived in Italy between the age of the Gracchi and the time of Augustus. The outlook of the majority of them, divided between the prison dormitories of the villas, the damnation of the mines, or the grinding drudgery of the most humiliating domestic duties in the city, would never extend beyond pure survival. And, what is more, never in history has the emergence from a slave system taken place from within the system itself, without the necessity of traumas or at any rate interventions induced from outside; nor have slave revolts ever sufficed. Great systemic changes have always ³⁶ Schiavone (2012: 145–53); see also Cardilli (2014).

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

32

Aldo Schiavone

been required to get out of it: the end of the ancient world, in one case, and a bloody civil war with a formidable industrial transformation behind it, in another. Slavery has always been a total institution, causing such a profound split in the consciousness both of those suffering and those imposing it as to render impossible, from within, a collective change with the specific goal of abolishing these practices. And so, over a few generations the creative and brilliant excogitations of the Roman jurists would end up in a blind alley. In order to continue, what was needed were social, economic, and cultural foundations that it was impossible for that world to provide. Later jurisprudence would not cancel out the results that had been achieved; instead, it would perfect and systematize them. But there would be no further progress of any significance along the path to further legal recognition for the slaves. And in fact the goal of legal thought had never been to arrive at the point of questioning the existence of the slave system but only to attain its maximum functional efficiency with a view to the profits of the masters, and to pursue an economic rationality that, while not rejecting slaveholding, reconfigured its structures to achieve a valorization of the productive forces compatible with the preservation of the given structure: a result that the law made possible with its empirical flexibility, and which no doubt contained within it the seeds of a culture already in some way protocapitalist, but without there emerging a class capable of taking possession of and developing it. It is a lack that marked the whole of ancient history, defining its form and destiny.

REFERENCES Andreau, Jean, and Raymond Descat. 2006. Esclave en Grèce et à Rome. Paris: Hachette. Bellen, Heinz, and Heinz Heinen. 2003. Bibliographie zur antiken Skaverei, vols I–II. Stuttgart: Franz Steiner Verlag. Bradley, Keith. 1994. Slavery and Society at Rome. Cambridge: Cambridge University Press. Buckland, William W. 1908. The Roman Law of Slavery. Cambridge: Cambridge University Press. Buckland, William W., and Arnold D. McNair. 1952. Roman Law and Common Law, ed. Frederick H. Lawson, 2nd edn. Cambridge: Cambridge University Press.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

Law, Slaves, and Markets in the Roman Imperial System

33

Carandini, Andrea. 1988. Schiavi in Italia: gli strumenti pensanti dei Romani fra tarda repubblica e medio Impero. Rome: Nuova Italia Scientifica. Cardilli, Riccardo. 2014. “Bona fides”: tra storia e sistema, 3rd edn. Turin: Giappichelli. Dalla, Danilo. 1994. Senatus consultum Silanianum. Milan: Giuffré. de Ligt, Luuk. 1999. Legal history and economic history: the case of the actiones adiecticiae qualitatis. 67 The Legal History Review 205–26. Di Porto, Andrea. 1984. Impresa collettiva e schiavo “manager” in Roma antica (II sec.a.C–II sec.d.C.). Milan: Giuffré. Esposito, Roberto. 2014. Le persone e le cose. Turin: Einaudi. Finley, Moses I. 1980 [new edn 1998]. Ancient Slavery and Modern Ideology, ed. Brent D. Shaw. Princeton: Princeton University Press. Giardina, Andrea. 1993. “The merchant,” in Andrea Giardina, ed., The Romans, trans. Lydia G. Cochrane. Chicago: The University of Chicago Press, 245–71. Harries, Jill Diana. 2013. “The Senatus Consultum Silanianum; court decisions and judicial severity in the early Roman Empire,” in Paul du Plessis (ed.), New Frontiers: Law and Society in the Roman World. Edinburgh: Edinburgh University Press, 51–72. Hopkins, Keith. 1978. Conquerors and Slaves: Sociological Studies in Roman History. Cambridge: Cambridge University Press. Miceli, Maria. 2001. Sulla struttura formulare delle “actiones adiecticiae qualitatis.” Turin: Giappichelli. Schiavone, Aldo. 2000 [repr. 2002]. The End of the Past: Ancient Rome and Modern West. Cambridge, MA: Harvard University Press (orig. edn: La storia spezzata. Roma antica e Occidente moderno. Rome and Bari 1996, several reprints). Schiavone, Aldo. 2012. The Invention of the Law in the West. Cambridge, MA: Harvard University Press (orig. edn: Ius. L’invenzione del diritto in Occidente. Turin 2005, new ed. Turin 2017). Schiavone, Aldo. 2013. Spartacus. Cambridge, MA: Harvard University Press (orig. edn: Spartaco. Le armi e l’uomo. Turin 2011). Schmitt, Carl. 1950. Die Lage der europäischen Rechtswissenschaft. Tübingen: Internationaler Universitäts-Verlag. Temin, Peter. 2013. The Roman Market Economy. Princeton: Princeton University Press. Thompson, Frederick H. 2003. The Archaeology of Greek and Roman Slavery. London: Duckworth. Turley, David. 2000. Slavery. London: Blackwell. Watson, Alan. 1987. Roman Slave Law. Baltimore: Johns Hopkins University Press.

OUP CORRECTED PROOF – FINAL, 21/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

14 The Practice of Manumission through Negotiated Conditions in Imperial Rome Egbert Koops

14.1. INTRODUCTION Roman slaves often had to meet expressly negotiated conditions to obtain their freedom (§14.2). The use of such conditions helps to explain why the Romans freed so many slaves (§14.3). They are an expression of the economic considerations that underlie the extraction and manumission model of Roman slavery (§14.4). Agreements between masters and slaves occurred in practice and were recognized at law (§14.5). Conditions could be set among the living or by testament and could consist of settling accounts, money payments, or services in kind; some followed the slave and were actionable (§14.6). The money to pay for freedom often came from the slave’s patrimony or peculium (§14.7). Although evidence is scarce, conditions and the corresponding manumission prices seem to have been of a type that could be met within years rather than decades (§14.8). Extracting a price from slaves for their freedom lessened the future claims of patrons (§14.9). For a certain type of slave, negotiated manumission conditions may have been the norm (§14.10).

Egbert Koops, The Practice of Manumission through Negotiated Conditions in Imperial Rome In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0014

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

36

Egbert Koops 14.2. TWO CASES: CRONION AND CALENUS

In  142, the cavalry trooper Antonius Silvanus drew up his will at his legion’s winter camp at Alexandria.¹ Silvanus instituted his son as heir, made out several legacies, and provided for the manumission of a slave: “I wish my slave Cronion to be free after my death if he has handled everything properly and transfers it to my aforementioned heir or agent, and I wish the manumission tax for him to be paid from my estate.”² As a libertus orcinus, someone freed from beyond the grave, Cronion’s obligations to his patron’s son would have been quite limited. An orcinus was not required to provide labor or sustenance, or to show particular deference to his patron under classical Roman law (Loreti-Lorini 1925: 36–46; Champlin 1991: 139; Mouritsen 2011: 51), making him largely independent from his manumittor’s family circle (Garnsey 1981: 362–3).³ In this case the heir was even burdened with the 5 percent manumission tax of the slave’s market value (Liebs 2000: 126; Günther 2008: 118–24).⁴ So why was Cronion freed? Referring to the clause si recte tractaverit, “if he has handled everything properly,” Wiedemann (1985: 164), recognizing that “manumission is a just reward for faithful service,” argued that Silvanus wanted to fashion himself as a benign master. Yet the clause has a different and more legal meaning. Si recte tractaverit is a condition that should be read in conjunction with the obligation to transfer something (et tradiderit). The best explanation, it seems, is that Cronion had to account for and hand over the assets of the inheritance. ¹ FIRA III 47. The will surfaced from the Egyptian sands in 1938 and remains the completest specimen of a Roman testament to date. Liebs 2000: 114–19. ² FIRA III 47, lines 31–7: Cronionem / seruom meum pos(t) mortem meam / si omnia recte tractaverit et / tra(di)derit heredi meo s(upra) s(cripto) uel / procuratori tunc liberum uolo / esse uicesima{m}que pro eo ex / bonis meis dari uolo. ³ Justinian made extensive changes to the Roman law of patronage by his constitution of  531, found in C. 6.4.4 (partially restored from the Basilica and referred to in Inst. 3.7.3). The position of the orcinus was completely revised by C. 6.4.4.27 (preserved in the Veronese Codex). On the interpolation of other texts in consequence of this law, see Loreti-Lorini 1925: 47–60; Harada 1939a, 1939b. Further discussion in §14.9. ⁴ A similar clause appears in line 53 of the so-called testament of ‘Dasumius’ (AE 2005, 191), though the clause in lines 117–18 refers to the vice(n)sima hereditatis. For an example of manumission inter vivos with payment of the manumission tax by the master, see P.Tebt. II 407 ( 199); for manumission testamento with payment by the heir, see P.Sijp. 44 ( c.130).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

37

The legacies of 1,350 denarii total represent the value of two to six slaves,⁵ quite apart from the gift of freedom to Cronion. The inheritance will have been bigger still, totaling perhaps 4,000 denarii (Champlin 1991: 54, 184–6), and so the cavalry soldier Antonius Silvanus appears to have been a moderately wealthy man. Perhaps Cronion was a trusted slave who looked after Silvanus’ accounts; or perhaps he had to render an accounting for the business he conducted out of his slave patrimony (peculium). Even though a slave had little legal personality and could technically hold no patrimony, his master could allow the use of certain assets as if they belonged to the slave, giving him a measure of autonomy and an incentive to increase his holdings (§14.7). In any case, it is fairly common in legal sources to find slaves accounting for their administration to their masters’ heirs in fulfillment of a condition for their release (Buckland 1908: 494–6). The Digest contains roughly seventy instances of a clause stipulating that a slave shall be free if he has rendered an accounting (si rationem dederit/reddidit),⁶ many of which are drawn from actual practice (Morabito 1981: 170–2). In an example that follows the wording of Silvanus’ will quite closely, probably also drawn from an actual testament, the jurist Pomponius cites the writings of the late Republican jurist Labeo: “my accountant Calenus is to be free and to have all his goods and a legacy of a hundred, if it appears that he has kept the accounts diligently” (si rationes diligenter tractasse).⁷ Pomponius goes on to explain that a diligence is required that profits the master rather than the slave, not only in calculating accounts but also in paying over the balance (reliqua). This second part is revealing. What Pomponius and Labeo have in mind is a case in which a slave receives his freedom together with his peculium, if he calculates and pays over in good faith the remainder of the accounts in his care. The slave-accountant ⁵ Compare the slave prices mentioned in Ruffing and Drexhage 2008: 321–36. The total comes to over four years’ of legionary wages at 300 denarii a year, not including donatives (Liebs 2000: 121). Difficulties in comparing slave prices: Harper 2010: 212. Influence of exonerations and guarantees: Arzt-Grabner 2010: 30–1. Market integration: Temin 2001, 2004; Scheidel 2005b. ⁶ Similar clauses appear in lines 41 and 51 of the ‘Dasumius’ testament (AE 2005, 191). ⁷ D. 40.7.21 pr. (Pomp. 7 Plaut.; Lab., Post.): Calenus dispensator meus, si rationes diligenter tractasse videbitur, liber esto suaque omnia et centum habeto. The same case is referred to in D. 40.4.8 (Pomp. 5 Sab.). For similar cases, see D. 33.8.23.1 (Scaev. 15 dig.) and D. 40.7.40.3 (Scaev. 24 dig.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

38

Egbert Koops

Calenus had personal money as well as access to his master’s money, and inevitably the two would get mixed up. The less of reliqua there would be for the master, the more peculium there would be for Calenus: hence the need to explain that the slave’s interest should not prevail. Elsewhere, Pomponius refers to a statement by the late first-century jurists Neratius and Aristo that a slave can give security for the unclear (obscurius) part of the balance. Otherwise hardly any slave would ever become free, “owing to the uncertainty of the account and this type of affair.”⁸ Celsus, writing in the first half of the second century, allows a similar glimpse. If a slave has been freed on condition of providing an account, but the heir will not allow him to sell off his peculium to raise money to pay the balance, then the slave will be free as if the condition were fulfilled.⁹ The Roman jurists consider it near-normal that the accounts of master and slave become entangled. They also consider it acceptable that a slave supply from his personal money what went missing from his master’s accounts. And they consider this sufficient to fulfill the condition for manumission.¹⁰ Africanus, from the middle of the second century, is clear on the point: “to give an accounting means no more than to pay the balance” and as long as this happens in good faith, the slave will be free.¹¹ Set against the legal sources, Cronion’s manumission appears in a different light than Wiedemann thought. “Faithful service” is not the condition for freedom, but Cronion’s truthful accounting and handing over of the balance. Much like Calenus, Cronion had access to his master’s accounts and by implication probably held a peculium. If so, it was not explicitly granted to him as a legacy and was therefore considered withdrawn in as far as anything would be left after settling the accounts.¹² Seen in this light, his manumission appears less as an act of benevolence and more as the winding-up of a partnership. ⁸ D. 40.7.5 pr. (Pomp. 8 Sab.): incerta causa rationis et genere negotii huiusmodi. ⁹ D. 40.7.23.1 (Cels. 22 dig.). This follows from the legal principle that a condition is automatically met if someone who stands to gain from its non-fulfillment hinders or prevents its fulfillment (Kaser 1971: 257). ¹⁰ e.g. D. 40.5.41.7 and 9 (Scaev. 4 resp.); D. 40.7.40 pr. (Scaev. 24 dig.). ¹¹ D. 35.1.32 (Afr. 9 quaest.): rationes reddere nihil aliud sit quam reliqua solvere. But also see D. 35.1.82 (Call. 2 quaest.) for an additional obligation to surrender the books. Columella (RR 1.8.13) advises masters not to allow farm overseers to trade for their own profit at all, because it increases the difficulty of settling accounts. ¹² Fr. Vat. 261 (Pap. 12 resp.); D. 33.8.8.7 (Ulp. 25 Sab.); C. 7.23.1 (Diocl./Max.,  294); Inst. 2.20.20.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

39

The slave’s help is indispensable to establish the contents of the inheritance, and manumission is a strong incentive to produce a truthful account.¹³ Of course Silvanus could have revoked Cronion’s privileges and peculium at any time, or could have ordered him to render an accounting (§14.5).¹⁴ But death is a natural moment to extricate the affairs of master and trusted slave, precisely because the heirs lack sufficient information about the inheritance if the slave bookkeeper does not cooperate (Dari-Mattiacci 2013: 84–7). The legacies that are often coupled to this type of manumission, as in the case of the accountant Calenus, only reinforce this view. They put a premium on a truthful account, leading to manumission, while the penalty for a detected lie is continued slavery, thereby tipping the scales in favor of truth.

14.3. THE FREQUENCY OF MANUMISSION It would be amiss to suggest that Cronion or the dispensator Calenus were typical Roman slaves. Not every employee in a modern firm has access to the board of directors or the corporate bank account and not every Roman slave had access to his master, let alone his master’s accounts. Slaves occupied a spectrum of social positions (Finley 1985: 67–8; Alföldy 2011: 196). Even so, Cronion and Calenus represent a type that is frequently encountered in literary sources, that predominates in the epigraphy, and that receives most of the jurists’ attention. Selection bias cannot be avoided, since too little is known about the vast bulk of undocumented Roman slaves, including almost all the agricultural slaves, to say whether, when, and how they were manumitted, or in what numbers. For this reason, most of what follows applies only to a certain class of privileged or skilled slaves (Fenoaltea 1984), or to those employed in positions of trust (Dari-Mattiacci ¹³ This is not to say that the account will necessarily be truthful. The slave may take the dangerous gamble that his fraud will go undetected, netting him both freedom and (more) money. D. 27.3.1.3 (Ulp. 36 Ed.); D. 40.4.22 (Afr. 9 quaest.); D. 40.5.23 pr. (Pap. 9 resp.); D. 40.5.41.11 (Scaev. 4 resp.); D. 40.7.40 pr. (Scaev. 24 dig.). ¹⁴ D. 33.8.19 pr. (Pap. 7 resp.): “when a master wanted to manumit a slave, he ordered him to provide an inventory of his peculium, and thereby the slave received his freedom” (cum dominus servum vellet manumittere, professionem edi sibi peculii iussit atque ita servus libertatem accepit). Also see D. 18.1.7 pr. (Ulp. 28 Sab.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

40

Egbert Koops

2013: 105–6), and not to Apuleius’ slave millers or the Spanish miners mentioned by Diodorus Siculus, “who would rather be dead than alive.”¹⁵ Certain claims can be made, however, with regard to this select group of documented Roman slaves. One of these claims takes the form of a question famously posed by Keith Hopkins (1978: 115): “Why did the Romans free so many slaves?” The number of references to freed slaves is staggering indeed. Rare cases of mass manumission aside (Westermann 1955: 135–6; Hopkins 1978: 115), many active, prosperous, or high-ranking freedmen appear in literary sources (Garnsey 1981: 359–61)¹⁶ and the economy of Roman Italy has been rightfully characterized as a “freedmen economy” (Verboven 2012). Three-quarters of the funerary inscriptions in Rome concern former slaves and the proportion is high as well for other Italian cities (Mouritsen 2005: 38–9, Verboven 2012: 90–1). Of the more than 506,000 Latin inscriptions catalogued in the Clauss-Slaby online database, over 27,000 refer to liberti. Freedmen figure in more than 1,000 legal fragments and their position was subject to a constant stream of legislation.¹⁷ Unfortunately, although such figures provide the notion that freedmen were common enough to matter, they reveal little about their relative number. The interests of writers and their readership shape literary sources,¹⁸ inscriptions reflect the strong epigraphical habit of freedmen (Mouritsen 2005: 55–63), and not every legal text presents a case drawn from actual practice. Some support can be found in the importance literary sources accord the vice(n)sima manumissionis, the manumission tax, but no reliable figures can be drawn from it (Bradley 1987: 149–50; Günther 2008: 98–9), nor from the number of public grain recipients (Mouritsen 2011: 120–3). Be this as it may, few scholars would dispute that the Romans freed ‘many’ slaves. Quantifying ‘many’ remains difficult, both in terms of the number of slaves at any given time and in terms of the ratio of slaves to free persons. Two figures are generally adduced: the census returns for ¹⁵ Apul., Met. 9.12; Diod. 5.38.1: αἱρετώτερος γὰρ αὐτοῖς ὁ θάνατός ἐστι τοῦ ζῆν. ¹⁶ e.g. Cass. Dio 50.10.4–5 and 51.3.3; Cic., Balb. 24; Philo, Leg. ad Gai. 155; Tac., Ann. 13.27. ¹⁷ Lopez Barja de Quiroga (1998: 161–3) provides a long list, only covering the Junian Latins. ¹⁸ Even freedmen authors such as Epictetus, Phaedrus, and Terentius provide little information on the prospects of the slave population. On Phaedrus, see Bradley (1987: 150–3). On the anonymous Life of Aesop, see Hopkins (1993: 10–27).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

41

Roman Egypt, which show a ratio close to 11 percent (Bagnall and Frier 1994: 48), and Galen’s cryptic reference to the roughly 20–30 percent slave population of Pergamum (Harris 1999: 65).¹⁹ It is difficult to proceed from these figures to an estimate for Roman Italy, let alone the rest of the Empire. As Scheidel (2005a: 64–6) has pointed out, the prevalent notion that slaves made up a third of the Roman population is nothing more than a shaky top-down guess. By stocking the Roman elite with a number of slaves according to their status, applying the data from Roman Egypt to the Italian cities, and extrapolating from the production levels and demand for a slave workforce in agriculture, Scheidel (2005a: 66–71) arrived at a bottom-up estimate of anywhere between 500,000–1,000,000 urban slaves and 250,000–750,000 rural slaves for Roman Italy. Depending on the size of the free population, a figure which is also problematic, slaves may have made up between 13–28 percent of the Italian population if one assumes the ‘low count’ for the total population (Scheidel 2008: 19–30). Such a margin of uncertainty precludes quantitative claims about the percentage of slaves manumitted, particularly since the weight of various sources in the slave supply is still heavily debated (Harris 1980; Scheidel 1997; Harris 1999; Scheidel 2005a). Because births seem to have been a major source of new slaves, female manumission rates cannot have been overly high at child-bearing age or the slave population would plummet, which is not supported by other evidence. But the incidence of male manumission and late-life female manumission has less of a demographic effect and may have been high.²⁰ Scheidel (1997: 160–6, 2005: 76) has employed a purely hypothetical model of 10–20 percent per five-year age cohort, starting at age 25, which means roughly half to threequarters of the slave population would have been released by age 50. This hypothetical model has met with general silent acceptance (Verboven 2012: 92), which means no better guess is currently available. In a seminal study of the age at death given in over 1,800 funerary inscriptions, Alföldy (1972: 346–54) has shown that a large majority of the epigraphically commemorated ex-slaves had been released

¹⁹ Gal., Propr. an. 9.13 (Corpus medicorum Graecorum V 4.1.1). ²⁰ Manumission was also used as an incentive for childbirth. Colum., RR 1.8.19; Fr. de iure fisci 13; D. 1.5.15 (Tryph. 10 disp.); D. 40.7.3.16 (Ulp. 27 Sab.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

42

Egbert Koops

before the age of 30, or at most between 30 and 40.²¹ The age of 30 was not chosen randomly: it was the legal age for manumission leading to Roman citizenship under the Augustan lex Aelia Sentia of  4.²² Alföldy claimed that manumission shortly after 30 constituted the “Normalfall” with which every slave could reckon. This assertion met with severe criticism (Hopkins 1978: 127; Garnsey 1981: 361–3; Wiedemann 1985: 162–3). It was argued that slaves were only commemorated at an early age; that only trusted slaves were commemorated at all; and that epigraphically commemorated freedmen form an atypical group of “successful” ex-slaves by their very nature. In addition, Alföldy’s figure could not generally apply to female slaves, or the slave population would never come close to reproducing itself (Scheidel 1997: 165–6). Yet the simple fact remains that the majority of epigraphically commemorated (ex-)slaves were manumitted early, when a reasonable part of their productive lives still lay ahead. True enough, the epigraphical data cannot be used as a demographic sample (Mouritsen 2011: 132–6), but it does show that for a certain class of slaves, early manumission was the norm. Employing similar methods as Alföldy with a non-overlapping data set, Weaver (1972: 97–104) concluded that manumission between the ages of 30 and 40 was normal for imperial slaves (familia Caesaris). A different, more random sample is offered by the census returns from Roman Egypt.²³ Bagnall and Frier (1994: 71, 94, 156–8, 342–3) indicate that male slaves in Roman Egypt were generally freed between ages 30 and 40, though women were as a rule not manumitted while they could still bear children. A similar trend appears in Romano-Egyptian documents that provide the age of slaves or freedmen (Wiedemann 1985: 163). Finally, a Greek census inscription from the late fourth century lists 152 named slaves and 87 ages on a single agricultural estate on the island of Thera (Harper 2008: 106–16). Males may be under-reported for tax reasons, but even so the same pattern emerges as from the Romano-Egyptian data, that is to say a high level of probable male manumission after 30. Perhaps it ²¹ As much as 67 percent in Rome itself, though regional differences should be kept in mind. As Alföldy notes, this percentage even understates the case, as freedmen who died at a later age may have been released earlier. ²² Gai., Inst. 1.18; Ulp., Epit. 1.12. ²³ The point can always be made that the published census returns for Roman Egypt cover three centuries, only mention 118 slaves, and may reflect local circumstances such as a lesser demand for slave labor.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

43

is a localized and fairly late snapshot. But at some point the question becomes how many “special cases” it takes to establish a general case, particularly since no study has provided any evidence to the contrary.²⁴ The incidence of male and female manumission varies remarkably between sources,²⁵ but for men, at least, manumission between the ages of 30 and 40 appears to have been frequent. The jurists suggest that the requirements for formal manumission were interpreted very leniently after 30, and quite strictly beforehand.²⁶ Even so, it goes without saying that some slaves were never freed.

14.4. ECONOMIC CONSIDERATIONS FOR MANUMISSION All slaves were expensive, both measured against wheat equivalent and against daily wages (Scheidel 2005b; Ruffing and Drexhage 2008: 341–7), and outside of the elite most households will have owned only a few (Bagnall and Frier 1994: 48–9), either as a status symbol and for domestic work, or to work beside their owner and provide income in the “family economy” (Groen-Vallinga 2017: 98–9). Slaves needed sustenance²⁷ and slave children at least had little earning power (Morabito 1981: 64–5).²⁸ Some were taught trades or skills at considerable ²⁴ The over 1,200 Delphic manumission records from the second century  through the first century  do not provide firm age indications (Hopkins 1978: 133–71). The same is true of the epitaphs of the urban slaves and freedmen of the elite Roman households (Treggiari 1975a: 58, 1975b: 395–400). ²⁵ Alföldy (1972: 351) noted a higher incidence of female than male manumission before 30 in the funerary inscriptions. The census returns from Roman Egypt and the Thera inscription show quite the opposite pattern, which is probably closer to reality. The difference may be due to the practice of releasing and then marrying young slave women (Wacke 2001), who stood a much better chance of being commemorated. ²⁶ Gai., Inst. 1.20: maiores vero triginta annorum servi semper manumitti solent, “slaves over 30 are wont to be manumitted at any time.” Also see D. 40.2.7 (Gai. 1 cott.); D. 40.2.8 (Ulp. 5 Ed.). ²⁷ Juvenal and Seneca complain of the cost (Sat. 3.166–67; Tranq. 8.8). A wet nurse for an infant slave foundling cost 10 drachmae and half a liter of oil per month (BGU IV 1107, 13 ); a slave wet nurse for a free infant cost 400 drachmae for two years (P. Oxy. I 91,  187). ²⁸ e.g. a legacy of the use of a slave child only takes effect after infancy, and no value is placed on his labor until the age of five: D. 7.1.55 (Pomp. 26 Q. Muc.); D. 7.7.6.1 (Ulp. 55 Ed.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

44

Egbert Koops

cost (Verboven 2012: 94).²⁹ Even the residual value of slaves was high. Petronius has a freedman claim that he paid 1,000 denarii for his freedom after forty years of slavery.³⁰ The highest actual age that can be related to a specific price is in the manumission certificate of Antonius Hermes, aged 40, for whom 20 denarii in taxes were paid, putting his value at a minimum of 400 denarii.³¹ To release any slave meant a sacrifice both in status and earning capacity for most households, not to mention the difficulty of recouping the investment under a system of early manumission. So why were they freed in such numbers? As Hopkins noted (1978: 117), “Roman society was not marked by altruism.” Of course, some manumissions were prompted by non-pecuniary considerations. Slaves and masters lived in close proximity and relationships were doubtlessly formed. Some slaves were released in recognition of their accomplishments³² or in gratitude for exceptional service,³³ and to a large extent such practice represents the literary ideal (Wiedemann 1985: 163–5; Mouritsen 2011: 30–5).³⁴ Slave girls ²⁹ P. Oxy. IV 724 = Chr. 140 (apprenticeship to stenographer, 120 drachmae,  155); but compare P. Oxy. XIV 1647 ( c.185, paid apprenticeship to weaver, wages rising from 8 drachmae per month in the first year to 20 drachmae per month in the fourth year, with eighteen days off per year). Value added by training: Plut., Cat. 21.7; Cic., Q. Rosc. 10; Hor., Ep. 2.2.5–8; Mart. 10.62. Also see the famous case reported by Paulus, D. 19.1.43 (5 quaest.). Conversely, Habinnas counts himself lucky that he spent no money to teach what his slave could also pick up on the streets (Petr., Sat. 68.7). ³⁰ Petr., Sat. 57.6 and 57.9. ³¹ FIRA III 10bis (Egypt, second century). The true price may have been higher still because of tax evasion. If both examples are explained away as cases of self-purchase (which is not evident in the case of Antonius Hermes), the next highest age is a slave woman aged 30–39 who sold for 200–25 denarii (800–900 drachmae, P. Mich. XV 707, after  185, damaged); or a male slave aged 32 who sold for 225 denarii (900 drachmae, P. Oxy. XXXVIII 2856,  91/92). In P. Hamb. I 63 ( 125/126) two prisoners of war, one aged 38 and the other’s age not given, are sold for a combined price of 350 denarii (1,400 drachmae). ³² e.g. after offering money the grammarian Lenaeus was freed gratis by Pompey in recognition of his learning (Suet., Gramm. 15), and when Cicero freed his secretary Tiro in 53 , his brother Quintus famously wrote that Tiro would rather be their friend than their slave (Cic., Fam. 16.16.1). ³³ Fending off robbers, uncovering a plot, defending or healing their master, etc.: D. 40.2.15.1 (Paul. 1 Ael. Sent.). Denouncing the murder of their master: Sc. Silanianum, D. 29.5 rubr. (Buckland 1908: 600–2). The grant of freedom for informing against a criminal master to the benefit of the state seems to be a later development (Buckland 1908: 598–9); but see App., Bell. civ. 4.11 for special circumstances (proscriptions). ³⁴ Dion. Hal., Ant. Rom. 4.24.4.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

45

were bought and set free at an early age for the purpose of marriage, which brought material benefit to their masters under the Augustan matrimony laws (Wacke 2001). The lex Aelia Sentia contained an exception for next of kin and future wives, who could be formally manumitted and acquire Roman citizenship before the prescribed age of 30.³⁵ If a union with a slave girl proved fruitful, the resulting children might be freed as well, though the relation between a master and his natural slave children should not be sentimentalized by assuming this as a necessary outcome.³⁶ Dying slaves were sometimes manumitted. The practice had little value in the eyes of a jurist and was the butt of one of Petronius’ jokes, since the bereaved master still owed manumission taxes (Kleijwegt 2002).³⁷ Sometimes, dying slaves were simply abandoned and left to fend for themselves.³⁸ Slave abandonment may either have been a common practice or, under the emperor Claudius, an outrageous but isolated incident that sparked imperial intervention. The same is true for the manumission of actors or gladiators following popular acclaim, which was criticized or forbidden by several emperors.³⁹ Equally frowned upon was the large-scale manumission to gain clients or fill out funeral processions, which Dionysius of Halicarnassus claims to have witnessed.⁴⁰ Such cases seem rare (Champlin 1991: 141). Finally, a slave could be forced to take up a debt-ridden inheritance as a necessary heir (heres necessarius), a quaint institution that merited considerable legal attention because a necessary heir was often included in the order of succession

³⁵ Gai., Inst. 1.19; D. 40.2.11 (Ulp. 6 procons.); D. 40.2.12 (Ulp. 2 Ael. Sent.); D. 40.2.13 (Ulp. 6 procons.); D. 40.2.20.2–3 (Ulp. 2 off. cons.). The slave’s age requirement was abolished by Justinian (C. 7.15.2,  530). ³⁶ Some testaments provide for the manumission of likely slave children, which implies that the master and father had no intention of manumitting immediately: FIRA III 50 = Chr. 316 (will of C. Longinus Castor,  194); FIRA III 10 (Dameis aged 13 freed with peculium and liberated from patronage, third century). ³⁷ D. 40.4.17 pr. (Iul. 42 dig.); Petr., Sat. 65.10. For deathbed manumission, see Mart. 1.101; Pliny, Ep. 8.16; CIL VI 9317; CIL X 2381 = ILS 7842; SB XXII 15345. ³⁸ Claudius decreed that abandoned slaves need not return to their masters if they recovered, but were free with the status of Junian Latins, giving them at least some recognition: Suet., Claud. 25.2; Cass. Dio 60.29.7; D. 40.8.2 (Mod. 6 reg.); C. 7.6.1.3 (Just.,  531). ³⁹ Tiberius: Cass. Dio 57.11.6; Suet., Tib. 47; Hadrian: Cass. Dio 69.16.3; Marcus Aurelius: D. 40.9.17 pr. (Paul. 3 Ael. Sent.); C. 7.11.3 (Alex.,  223–4); Cass. Dio 72.29. ⁴⁰ Dion. Hal., Ant. Rom. 4.24.8. Also see Pers., Sat. 3.105–6; Petr., Sat. 42.6.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

46

Egbert Koops

to ensure that the inheritance was taken (Buckland 1908: 505–12).⁴¹ Again, the number of cases in which this failsafe was activated and the slave was freed may have been limited (Champlin 1991: 137). The problem with such manumission motives is that blood ties, faithful service, deathbed grants, abandonment, popular success in the arena or on stage, debt-ridden inheritances, etc. can hardly account for systematic male manumission at an early age. It is difficult to reconcile that manumission was “both very common and very selective,” as Mouritsen claims (2011: 140), who admits that it was “essentially a financial transaction” (2011: 146) while simultaneously downplaying economic considerations (2011: 159–80). Instead of the fanciful notion of “common selection,” the best explanation why the Romans freed so many slaves remains the one offered by Hopkins (1978: 118) and Alföldy (1972: 360–3): slaves paid considerable sums of money for their freedom. The Roman slave system exploited the desire for freedom to great effect, offering manumission as an incentive for the ideal type of hardworking and frugal slave who would merit freedom (Hopkins 1978: 128, 147–8). This meant that extraction of the full value of a slave’s labor was postponed for some time. Even so, giving slaves a stake in their productivity by allowing them to save part of their earnings ensured cooperation and harder work at lower supervision costs (see already Cohen 1951: 222).⁴² The risk of undisclosed funds following the slave into freedom was partially countered by giving patrons a legally protected stake in any future earnings of a freedman, through an expectancy to inherit (§14.9) as his closest agnatic kin (Kaser 1971: 697; Gardner 1993: 19–20).⁴³ Thus, slavery and manumission were marked by a stepped extraction process. First, labor was extracted for which the slave received less than full compensation; then a manumission price would be paid, enabling the master to reinvest while obtaining a freedman; and upon the freedman’s death, the patron would often take part of the inheritance.

⁴¹ Gai., Inst. 2.153; Ulp., Epit. 1.14. See line 13 of the testament of ‘Dasumius’ (AE 2005, 191). ⁴² Varro, RR 1.17.5–7 and 1.19.3. But note the quite different opinion of Colum., RR 1.8.13 and 11.1.24. Aristotle hints at the problem as well (Oecon. 1.1344b). ⁴³ Gai., Inst. 3.39 and following. Gaius dedicates a full thirty-eight paragraphs to the convoluted law of freedman inheritance.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

47

The mechanism described here was not invented by Augustus. Statuliberi, slaves freed by testament on condition of payment of a sum of money, already appear in the Law of the Twelve Tables.⁴⁴ But the Augustan manumission laws surely fostered this form of manumission to serve state interests.⁴⁵ On the one hand, a high manumission rate was desirable as it meant that enough freedmen were visible and successful for manumission to present a believable scenario, fostering obedience on the part of the slave population (Bradley 1994: 163).⁴⁶ Moreover, freed slaves were taken into the citizen body, adding new citizens in a practice that some considered an important contribution to the Roman success.⁴⁷ On the other hand, Augustus aimed to add only the “best” slaves as citizens,⁴⁸ that is to say those who deserved freedom in the eye of a discerning master by their frugality or other virtues. These conflicting aims were both addressed by the lex Iunia (Norbana) of 17 ,⁴⁹ which created a waiting room class of second-rate free persons, consisting of freedmen who had been granted their liberty through informal manumission and lacked full citizenship in consequence.⁵⁰ Such Junian Latins do not stand out in the epigraphy (Weaver 1990) and are difficult to track in other sources.⁵¹ With some justice Weaver (1997: 55) has called this status group “a black hole of large but unknown ⁴⁴ Tab. VII,12 (Bruns) = Tab. VI.2 (Flach): Ulp., Epit. 2.4; D. 40.7.25 (Mod. 9 diff.); D. 40.7.29.1 (Pomp. 18 Q. Muc.). Also see Fest., s.v. statu liber. ⁴⁵ Augustus’ program included the lex Iunia (17 ), lex Fufia Caninia (2 ), lex Aelia Sentia ( 4) and the lex Papia Poppaea ( 9). With minor modifications these laws were in place for over five hundred years, until most were abolished by Justinian. ⁴⁶ Hinted at by Dion. Hal., Ant. Rom. 4.24.7, perhaps in response to the slave revolts of the late Republic. ⁴⁷ Cf. the famous (but slightly erroneous) letter of Philip V of Macedon to the city of Larissa, SIG³ 543 = ILS 8763 (214 ). Also see Cass. Dio 56.7.6. ⁴⁸ Cass. Dio 56.33.1–3 (but perhaps mistaken, see Suet., Aug. 101); Suet., Aug. 40.3–4. ⁴⁹ The date is in dispute (the other possibility being  19, under Tiberius), but since the status group is named after the Junian law, it seems likely this law preceded the lex Aelia Sentia that also regulated their position. Buckland 1908: 534–7; Sirks 1981: 250–1; Weaver 1997: 58–60; Lopez Barja de Quiroga 1998: 137–8. ⁵⁰ That is to say not by iusta manumissio in front of a magistrate or by testament, but by the mere will of their master. Gai., Inst. 1.17 and 3.56; Fr. Dos. 4–5; Quint., Decl. min. 340 and 342. Of course proof was essential: see Mart. 9.87 and the unsavory tale in Cic., Att. 7.2.8. ⁵¹ Isolated cases: Suet., Vesp. 3.1 (later declared freeborn); Plin., Ep. 7.16.4 (offer of magistrate’s assistance); Plin., Ep. 10.5.2, 10.11.2 and 10.104 (requests for citizenship grants); Mart. 1.101 (deathbed manumission); AE 1959, 297 (probatio anniculi). See now Emmerson (2011).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

48

Egbert Koops

proportions.”⁵² It seems likely that informal manumission was frequently employed. Patrons of Junian Latins retained a claim to their entire inheritance, which would fall to them as if the deceased freedman had remained a slave.⁵³ This gave Junian Latins a strong incentive to obtain full Roman citizenship by a second, formal manumission (iteratio, Buckland 1908: 714–18; Sirks 1983). The requirement of further cooperation from the patron for formal manumission ensured continued dependence and further opportunities for extraction.⁵⁴ In any case, the overall effect of the Augustan laws was that informally manumitting slaves as Junian Latins was relatively easy, maintaining a high manumission rate, while formally turning slaves into Roman citizens was more difficult, but not impossible, maintaining the desirability of citizenship. From a systematic point of view, manumission reinforced slavery (Hopkins 1978: 118).

14.5. IMPLICATIONS OF FREEDOM PURCHASE: AGREEMENTS AND TERMS The extraction process described here has two implications. The first is that manumission was often the outcome of negotiations between master and slave. This may be a difficult proposition, considering the power disparity, but there is enough evidence to believe agreements or promises occurred frequently. In the Roman view, slaves were capable of forming relations governed by bona fides or good faith (Horsmann 1986: 317), which in turn meant that to break one’s word, even against a slave, was to act against good morals (Kaser 1971: 200). Pliny allowed slaves to bequeath part of “their” property, as long as they made it over within the household, and Columella advised to offer exemption from work to slave women who bore more than two

⁵² As a status group, Junian Latins survived the mass grant of Roman citizenship by the Constitutio Antoniniana of  212, and the status was only abolished by Justinian (C. 7.6.1.12–12a,  531). Koops 2012. ⁵³ Gnomon 19 = BGU 1210; Gai., Inst. 3.56; Inst. 3.7.4; C. 7.6.1.1b (Just.,  531). ⁵⁴ Fr. Dos. 14; Gai., Inst. 1.35; Fr. Vat. 221; Ulp., Epit. 3.4. Also see Tac., Ann. 13.27. The state tapped this source of income as well, as successive emperors granted citizenship in return for any number of expensive civic services. Gai., Inst. 1.32c–34; Ulp., Epit. 3.6.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

49

children, and freedom to those who bore more than three.⁵⁵ The novelist Achilles Tatius describes a scene in which the slave girl Lakaina negotiates with Melite for her purchase for 2,000 drachmae from an abusive master, offering to repay as soon as possible.⁵⁶ In an inscription at the Rijksmuseum of Oudheden at Leiden, a certain Ericthonius is mentioned, “who died age 28, when he ought to be manumitted at age 30.”⁵⁷ The use of the verb debere may imply a social norm, but also an agreement between a master and his slave or a third party concerning the term of servitude; in any case Ericthonius’ master was felt to be morally or legally required to free him at age 30.⁵⁸ Manumission agreements, pactiones pro libertate, are common in legal sources. As early as the first century , Alfenus discusses a case in which a slave has contracted with his master to purchase his freedom; when the master dies, the slave has not yet paid the full amount but is freed by testament.⁵⁹ Whether he can reclaim what he has already paid depends on whether the master entered the payment into his own accounts, or kept it apart as belonging to the slave’s peculium. The implication that legal obligations could arise between masters and slaves as debtors and creditors is borne out by a multitude of texts (Buckland 1908: 220–5; Morabito 1981: 105–10). As Ulpian notes, slaves were bound and could bind others naturaliter by their contracts.⁶⁰ True, a master was under no legally enforceable obligation and merely under a moral obligation to uphold agreements made with his slaves. Masters might withdraw the peculium (Johnston 2002: 11–12),⁶¹ sell the slave or assign him to a different station, or renege on their promise of manumission (Buckland 1908: 205, 640–3). ⁵⁵ Plin., Ep. 8.16.1–2 (note the inversion in Petr., Sat. 53.8, where Trimalchio is disinherited by his slaves); Colum., RR 1.8.19. In the same letter, Pliny suggests that he manumits immaturos easily, that is those under 30. ⁵⁶ Ach. Tat. 5.17. ⁵⁷ RMO Leiden (unpublished, EDCS-58700011): Dis manibus/Ericthonio animae/ sanctissimae
hic/cum deberet ann(is) XXX/manumitti
ann(os) XXIIX/decessit/ C(aius) Cilnius
Philetus/filio carissimo/fec(it) (Smyrna, second to third century). Compare CIL X 4917. ⁵⁸ Especially in cases of self-sale into slavery, such agreements will have been common. Morabito 1981: 70–4; Harris 1999: 73–4. For discussion of the clause ut manumittatur, see §14.6. ⁵⁹ D. 40.1.6 (Alf. 4 dig.). Also see D. 33.8.8.5 (Ulp. 25 Sab., referring to Labeo). ⁶⁰ D. 44.7.14 (Ulp. 7 disp.). Also see D. 12.6.13 pr. (Paul. 10 Sab.); D. 12.6.64 (Tryph. 7 disp.). ⁶¹ D. 15.1.7.6 (Ulp. 29 Ed.); D. 15.1.8 (Paul. 4 Sab.). Compare Plaut., Aul. 820–32.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

50

Egbert Koops

A rescript from Diocletian states that a mistress who agreed that her slave would be free after a certain period of servitude was not required to uphold the bargain; but even so, a different rescript from the same emperor states that a master who has received a slave’s payment should at least be urged by the provincial governor to keep his promise of manumission.⁶² The ability to renege raises the question whether some masters would freeload on a social structure in which many kept their word (Watson 1987: 65, Kleijwegt 2012: 113–15). After the murder of the urban prefect Pedanius Secundus by his slave in  61, the senate debated whether to apply the law that all slaves under the same roof were to be killed. Tacitus notes that the murderer was spurred to violence either “because he had been refused the freedom for which he had paid the price” or out of a lover’s jealousy; but later, in the speech he attributes to the jurist Gaius Cassius Longinus, it is stated that the unnamed slave was avenging wrongs “because he had bargained with paternal money or because an ancestral slave was taken away”; this suggests that Tacitus rather believed a failed manumission deal to be the cause of violence.⁶³ Roman masters were well aware of this potential threat: “as many enemies as slaves,” as the proverb went.⁶⁴ Not to keep one’s word on matters as important as manumission could be dangerous in a society with a propensity toward violence (Morabito 1981: 135). Considerable legal ingenuity was expended on ways to neutralize this social danger by making the master’s obligation to manumit enforceable at law. Examples are the use of a trusted third party or the allowance of certain claims against the master if an agreement had been confirmed by testament (§14.6). Such devices are far more common in the legal sources than simple pactiones between master and slave. The second implication of agreements to purchase freedom would appear to be that Roman slaves were manumitted more often inter vivos, during the lifetime of their masters, than by testament. This runs counter to a tradition going back to Buckland (1908: 546), that testamentary manumission was more common by far because it guaranteed the use of the slave until death (Duff 1928: 25; Garnsey ⁶² C. 7.16.36 (Diocl./Max.,  294); C. 4.6.9 (Diocl./Max.,  294). Also see C. 7.16.20 (Diocl./Max.,  293). ⁶³ Tac., Ann. 14.42: seu negata libertate, cui pretium pepigerat, sive amore exoleti incensus et dominum aemulum non tolerans; Tac., Ann. 14.43: quia de paterna pecunia transegerat aut avitum mancipium detrahebatur. ⁶⁴ Sen., Ep. 47.5; Fest. s.v. quot servi (314L); Macr., Sat. 1.11.13. Also see Curt. 7.8.28.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

51

1981: 362; Patterson 1982: 223–4).⁶⁵ Recently, scholarly opinion has swung the other way, claiming that manumission inter vivos represents the standard (Roth 2010: 98–9; Mouritsen 2011: 180–3). In truth, there is little evidence for either claim. Testamentary manumission is referenced four times more often in the Digest (Morabito 1981: 166), but that may be due to the greater complexity of testaments and legacies, giving the lawyers and Byzantine lawmakers cause for attention. In literary and epigraphic sources both are encountered in roughly equal proportion, but accidents of preservation forbid any conclusion. Little importance can be attached to the rule that the slave peculium was considered granted unless withheld for manumission inter vivos, and was considered revoked unless granted for testamentary manumission. As Roth notes (2010: 95–104), because slaves paid for freedom out of their peculia, it would have been the subject of negotiation on the one hand, while on the other hand an implicit grant was often construed, for instance in cases of testamentary manumission on condition of payment of a certain sum. Since it touches on slave expectations as well as manumission motives and numbers,⁶⁶ it would seem important whether manumission inter vivos or by testament was more common. However, once one accepts the existence of a social practice of manumission payments, this opposition loses its relevance. A master would negotiate with a certain class of slaves and set certain terms. If the terms were met during his lifetime, he would usually free the slave; but it made sense to enshrine the terms in a testament or codicil as well, leading to a legally recognized position of the slave as a statuliber in case the master died before freeing his slave. Testamentary manumission put the agreement between master and slave on a securer footing for the slave: not because he expected to be released only at his master’s death, but because he knew that the bargain would hold even if his master died. The heirs were not bound by any agreements between the deceased and his slave, but they were bound by the testator’s will. A testament such as that of Silvanus (§14.2) cannot discount the possibility that Cronion may have been manumitted before the will came into effect; but if Cronion was still a slave when Silvanus died, ⁶⁵ The self-purchase model vitiates this argument since it allows for the purchase of a replacement slave and the addition of a freedman. ⁶⁶ Because of the lex Fufia Caninia of 2 , which limited the number of slaves who could be freed by testament: Gai., Inst. 1.42; Ulp., Epit. 1.24; Paul., Sent. 4.14.4.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

52

Egbert Koops

his position as a statuliber would have been unassailable, not only by the heirs but also by any new owner to whom the heirs might sell him. Such conditions followed the slave (Buckland 1908: 286–7). Of course, slaves were powerless to prevent the testament from being modified by the testator during his lifetime,⁶⁷ which happened frequently (Champlin 1991: 64–70); but that was no different than their position regarding any other agreement with their master. Moreover, there are traces of a practice of reading wills in public, asserting the social order of the household (Champlin 1991: 23–4) by communicating what lay in wait.⁶⁸ Seen in this light, the important question is not how the Romans freed so many slaves—by testament or inter vivos—but under which terms. A testament is merely a way to enshrine a negotiated agreement and make it actionable. Over 90 percent of the fragments in the Digest dealing with manumission place a condition on attaining liberty (Morabito 1981: 174). The condition may consist of an amount to be paid, at once or over time, or of years of servitude, but it may also take the form of future service as a freedman; and it may be set by agreement or by testament, either directly or through an intermediary. The possibilities are endless (Champlin 1991: 139–40). And though their frequent occurrence should be taken with a grain of salt, since conditions give rise to greater legal problems than gratis manumission, it does show that manumission terms were at the forefront of jurists’ minds (Table 14.1). Put differently, Table 14.1. Manumission terms in the Digest (# of fragments). payment of price

190

- price set inter vivos - price set by testament render accounting fixed term or fixed age imposition of operae no condition (testament) no condition (inter vivos)

(71) (119) 66 87 101 31 14

⁶⁷ If (part of) a testament containing a manumission clause was suppressed, the slave could sue the heirs under a rescript from Marcus Aurelius and Commodus: D. 5.1.53 (Herm. 1 epit.); D. 48.10.7 (Marc. 2 Inst.). ⁶⁸ Petr., Sat. 71.1–3.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

53

such extensive discussion only makes sense if the purchase of manumission was common, and freedom had its price.

14.6. FREEDOM PURCHASE AGREEMENTS AND THEIR ENFORCEABILITY Many slave systems in antiquity developed mechanisms whereby slaves purchased their freedom (Patterson 1982: 282–4). The best example is provided by the 1,200+ inscriptions from Delphi (Hopkins 1978: 133–71), though manumission payments are encountered throughout the Hellenistic world (Zelnick-Abramovitz 2005: 207–47). Even so, a similar Roman practice is harder to substantiate outside of the legal texts. Few inscriptions mention whether a freedman paid for his liberty or received it gratis.⁶⁹ Even when family relationships make paid-for manumission unlikely, a gift is hardly ever mentioned, which warrants the assumption that the reason for manumission was simply not considered worth inscribing. Unlike Greek manumission inscriptions, which were publicized on durable temple walls, Roman deeds and state records were kept on perishable materials. Freedom purchase appears quite often in the far better preserved documentary evidence from Roman Egypt (Scholl 2001),⁷⁰ but no conclusions can be drawn about its prevalence, save that informal manumission is more common than the iusta manumissio leading to Roman citizenship. Literary sources provide more references to the purchase of freedom, starting as early as Plautus (d. 184 ). When the slaves Stichus and Sagarinus encounter the courtesan Stephanium, they exclaim that “the peculium is wafted away, it is done; liberty has fled this slave.”⁷¹ Many Plautine slaves have a peculium. It may be sizable through thrift or for comic effect, or merely limited to some livestock, but the Plautine slaves generally ⁶⁹ CIL XI 5400 = ILS 7812 (doctor, 50,000 sesterces); CIL XII 5026 ([e]t pretio [obtin]uit, quod prec[e]/non valuit); CIL VI 2211 = FIRA III 80g (free); CIL VI 20905 (free, but curse tablet). CIL II² 7.432 can be read either way. ⁷⁰ e.g. P. Oxy. IV 202 = Chr. 361 (per epistulam); FIRA III 11 = Chr. 362; P. Lips. II 151; P. Oxy. IX 1205; (inter amicos); P. Oxy. IV 722 = Chr. 358; P. Turn. 19 (in front of agoranomoi); P. Flor. I 4 = Chr. 206; P. Oxy. XLIII 3117 (registration through court). ⁷¹ Plaut., Stich. 751: [Stich.] Vapulat peculium, actum est. [Sag.] Fugit hoc libertas caput.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

54

Egbert Koops

attempt to buy their liberty or that of loved ones once they acquire some money.⁷² Other authors assume that slaves will work hard and save, or rob and swindle, even cheat their own bellies to find the necessary funds.⁷³ The promise of freedom is considered an effective incentive for better work,⁷⁴ and although literary authors can conceive of slaves who prefer sheltered servitude to insecure life on their own,⁷⁵ such cases must have been rare in practice, considering the grave disadvantages of “social death” (Patterson 1982: 86–94; Hopkins 1993: 14). Martial refers to the advantages of “mastering” oneself⁷⁶ and Trimalchio’s conlibertus Hermeros, who bought his freedom for the high price of 1,000 denarii after forty years of servitude (§14.4), also claims to have purchased the freedom of his slave partner (contubernalis) to save her from groping hands.⁷⁷ “Liberty is favored above all things,” according to Gaius,⁷⁸ and though the worth of freedom cannot be measured, its value can be priced. Skilled slaves may have cost a fantastic sum in the past, Pliny the Elder notes, but if that figure is topped in the present, it is because these slaves are purchasing their freedom.⁷⁹ And when Vespasian became emperor, his old herdsman complained that the fox had changed his fur, but not his nature, so that he still needed to pay for manumission; conversely, Virgil’s old shepherd Tityrus never had a hope of freedom and hence never a thought of saving.⁸⁰ The purchase of freedom is ubiquitous in legal sources. In the Digest alone, close to two hundred fragments deal with the complications arising from its various legal forms. Aside from simple pactiones and the more elaborate obligation to render an accounting ⁷² Plaut., Asin. 497–8 (frugality); Plaut., Asin. 539–40 (flock); Plaut., Trinum. 727–8 (talent in the bank); Plaut., Trinum. 433–4 (peculium generally assumed). Purchase of freedom is mentioned inter alia in Plaut., Asin. 650–2 and 673; Aul. 309–10; Most. 300; Pers. 34–8; Rud. 1408–10. ⁷³ Cic., Par. stoic. 5.39; Sen., Ep. 80.4; Pers., Sat. 5.73–99 and 5.174–5. Also see Ter. Phor. 1.35–46 (saving); Dion. Hal., Ant. Rom. 4.24.4 (saving, theft); Plin., NH 33.6.26 (theft of food); Apul., Met. 10.14 (theft of food); Juv., Sat. 3.188–9 (bribes). ⁷⁴ Cic., Rab. Perd. 15. Also see Sen., Ep. 80.4. ⁷⁵ Plaut., Capt. 119–20 and 270–3; Epict. 4.1.33–5; Mart. 9.92. ⁷⁶ Mart. 2.68. ⁷⁷ Petr., Sat. 57.6: “I have purchased the freedom of my slave bedmate, so that no one might wipe his hands in her hair” (contubernalem meam redemi, ne quis in illius manus tergeret). ⁷⁸ D. 50.17.122 (Gai. 5 Ed. prov.): Libertas omnibus rebus favorabilior est. ⁷⁹ Plin., NH 7.39.128–9. Also see Suet., Gramm. 13 (Staberius Eros) and 15 (Lenaeus); D. 12.4.3.5 (Ulp. 26 Ed.); Tac., Ann. 13.27 (Paris). ⁸⁰ Suet., Vesp. 16.3; Virg., Ecl. 1.26–35.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

55

(§14.5 and 2), slaves could be sold under the obligation to manumit (ut manumittatur); sold in trust against a payment from the slave’s peculium or money borrowed to the purpose (emptio suis nummis); manumitted by testament under a condition of payment (statuliber); or left under a testamentary trust (fideicommissum) under the taker’s obligation to manumit against payment. Many of these arrangements were put on firmer legal footing by the emperors,⁸¹ though it should be recalled that statuliberi already appear in the Law of the Twelve Tables. Under the Trajanic SC Rubrianum ( 103),⁸² the praetor could declare a slave free if the fiduciary taker under a legacy would not free him and did not appear in court to explain himself (Buckland 1908: 611–20; Finkenauer 2010: 26–34). If they were not freed by their fiduciary buyer, slaves who bought their freedom with their own money (suis nummis) could claim their freedom in court under a rescript of Marcus Aurelius and Lucius Verus ( 161–9, Buckland 1908: 636–40; Horsmann 1986; Finkenauer 2010: 44–50).⁸³ And by a constitution of Marcus Aurelius and Commodus ( c.178), a slave who was sold or, less often, donated to be manumitted by the new master, was automatically free if the master did not comply (Buckland 1908: 628–36; Finkenauer 2010: 34–44).⁸⁴ This construction of a transfer ut manumittatur was particularly suited to cases of self-sale into slavery, which required the use of a proxy seller. The self-seller could negotiate the conditions of servitude and future manumission (including the grant of a peculium) by having the proxy seller transfer him under the condition ut manumittatur, for instance after a certain period of servitude or when a certain sum had been paid. Such a construction gave the self-seller legal redress, though still a slave, if the conditions were met but manumission did not follow. A similar but far older rule applied to statuliberi, who became free immediately once they either fulfilled the testamentary condition of paying money, rendering an accounting, serving for a certain period, etc., or were prevented from fulfilling that condition (Buckland 1908: 487–8, 492–505). In other words, slaves could actively proceed against their ⁸¹ The causes for this development are not clear. A lessened credibility of mere moral commitments, favor libertatis, the greater importance accorded to will theories and by extension to testaments, or even the desire to raise the proceeds of manumission taxes may all have played a part. ⁸² D. 40.5.26.7 (Ulp. 5 fideic.). ⁸³ D. 40.1.4 pr. (Ulp. 6 disp.) and following. ⁸⁴ D. 40.1.20 pr. (Pap. 10 resp.); D. 40.8.3 (Call. 3 cogn.); C. 4.57.2 (Alex.,  222); C. 4.57.3 pr. (Alex.,  224).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

56

Egbert Koops

masters in court on the strength of many manumission agreements, though not on a mere pactio. This forms an important departure from the rule that slaves had no capacity to appear in civil suits (Buckland 1908: 83–4; Garnsey 1981: 363).⁸⁵ Such varied manumission arrangements imply negotiation between masters and slaves, but even more tellingly a concern for the possibility of enforcing the bargain. Two final arrangements need mentioning because they clarify the limits of this system. The Roman law of civil procedure dictated that all judgments were expressed in terms of money, which might suggest the trick of colluding with a free person to bring a suit concerning the status of a slave (Buckland 1908: 653, 665). If the master proved slave status, the suit was lost and the damages would simply be paid. Yet in opposition to the general principle of pecuniary condemnation, Papinian explicitly states that a victorious master cannot be forced to accept money if he wishes to take away the slave.⁸⁶ Negotiated manumission always depended on the will and assent of the master (Gardner 1993: 14). A second trick only strengthens the point. A slave could give someone a mandate to buy him and subsequently set him free.⁸⁷ Such a mandate has essentially no legal force, although it may give rise to obligations if the slave has indeed been sold and transferred, that is to say if the master assents (Buckland 1908: 639–40; Heinemeyer 2013: 238–80). But since a slave has no authority to sell or transfer himself,⁸⁸ and cannot irrevocably bind the master to a mandate to sell (which can always be withdrawn as long as there is no performance), it falls within the master’s power to comply with the slave’s scheme or not (Heinemeyer 2013: 270–1). For this reason, according to Ulpian, “In this case I should be no more liable on the mandate than I am forced to sell him.”⁸⁹ In the final analysis, manumission required the master’s assent. ⁸⁵ D. 5.1.53 (Herm. 1 epit.): slaves may inter alia appear in court against their masters in cases of (conditional) testamentary or fideicommissary manumission, fiduciary purchase with their own money (suis nummis), purchase with money loaned to the purpose of manumission, and to have an arbiter appointed to supervise the conditions of accounting. ⁸⁶ D. 40.12.36 (Pap. 12 resp.): Dominus qui optinuit, si velit servum suum abducere, litis aestimationem pro eo accipere non cogetur. If he accepted the pecuniary condemnation however, the slave became a Junian Latin: C. 7.6.1.8 (Just.,  531). ⁸⁷ D. 17.1.19 (Ulp. 43 Sab.); D. 17.1.54 pr. (Pap. 27 quaest.); C. 4.36 l (Diocl./Max.,  293). ⁸⁸ A slave is never part of his own peculium: D. 33.8.16.1 (Afr. 5 quaest.). ⁸⁹ D. 17.1.19 (Ulp. 43 Sab.): nec magis in hunc casum debeo mandati teneri, quam ut eum tibi venderem.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

57

14.7. THE PECULIUM AS A FREEDOM FUND The linchpin of the Roman system of self-purchase by slaves is the peculium, the slave patrimony. That certain slaves were allowed the use of property that formally belonged to their master is again not uniquely Roman. The Delphic manumission inscriptions imply that the phenomenon was also known in the Greek world (Hopkins 1978: 147) and traces of a similar practice appear in Jewish law (Cohen 1951: 150–86). What is characteristically Roman, however, is the amount of legal attention devoted to the institution. In theory it could not exist as the slave had no relevant legal capacity; but in practice the jurists treated it in over 1,000 fragments of the Digest with their eyes wide shut.⁹⁰ For the present purpose, only a few aspects matter. It has been noted that the peculium could be granted and taken away at any time, though this was undoubtedly circumscribed by social norms. Whether masters always knew what their slaves controlled is another matter entirely (Johnston 2002). Some placed their money in a bank, lent it to debtors, or sent assets abroad, and many will have had secret funds hidden away.⁹¹ The implication of rendering an accounting as a condition for manumission (§14.2), with the insistence on accounting in good faith, is that “squaring up” was more important than the provenance of the funds.⁹² As noted, some peculation was assumed, but evidently outweighed by lower supervision costs. Slave peculia could be extensive, containing “chattels, land, subslaves and their peculium, and even claims against debtors,” or very small, consisting only of clothes or some livestock.⁹³ Sizable holdings

⁹⁰ D. 40.1.4.1 (Ulp. 6 disp.): “First, it seems that it cannot be properly said that someone is bought with his own money, since a slave cannot have his own money: but with closed eyes it should be thought that he has been purchased with his own money, as long as he is not bought with the money of the one who buys him” (et primo quidem nummis suis non proprie videtur emptus dici, cum suos nummos servus habere non possit: verum coniventibus oculis credendum est suis nummis eum redemptum, cum non nummis eius, qui eum redemit, comparator). ⁹¹ D. 16.3.1.33 (Ulp. 30 Ed.); D. 40.7.40.6 (Scaev. 24 dig.); D. 44.7.14 (Ulp. 1 disp.). ⁹² D. 40.7.23.1 (Cels. 22 dig.). ⁹³ D. 15.1.7.4 (Ulp. 29 Ed.): in peculio autem res esse possunt omnes et mobiles et soli: vicarios quoque in peculium potest habere et vicariorum peculium: hoc amplius et nomina debitorum. Clothes: D. 15.1.25 (Pomp. 32 Sab.). Livestock: Varro, RR 1.17.7 and 1.19.3.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

58

Egbert Koops

are attested for slaves in the imperial household,⁹⁴ though the legal sources also contain references to private slaves owning money, wine, buildings, cattle, gold and silver, and vicarii or sub-slaves.⁹⁵ Many slaves erected costly inscriptions from their own money (de suo) for themselves or for others: the masseur Arphocras, for instance, bought space in a columbarium for 80 denarii and an ossuary for 175 denarii.⁹⁶ Slaves received some sort of allowance⁹⁷ and there are glimpses of wage-earning or tenant slaves who made payments to their absent masters,⁹⁸ presumably in exchange for greater independence. Alfenus reports that he was consulted on the case of a statuliber who paid off his duty to perform day labor in money. These money payments did not count toward fulfilling the manumission condition, Alfenus explained, just as a money payment from the slave tenant of a farm instead of a payment in kind would not count.⁹⁹ Even Cato, who is not otherwise known as a considerate master, ran a paid brothel for his slaves and lent them money to invest in other slaves and share in the profit, both of which imply the grant of a peculium.¹⁰⁰ Several authors have argued that Roman businesses were structured through the peculium to benefit from the legal advantages offered by the master’s limited liability (Di Porto 1984; Cerami and Petrucci 2010: 61–7; Abatino et al. 2011). But in spite of occasional references,¹⁰¹ there is preciously little evidence outside of the legal texts¹⁰² that economic ⁹⁴ e.g. CIL VI 5197 = ILS 1514 (a dispensator Scurranus with 16+ vicarii,  14–37); Suet., Otho 5.2; Plin., NH 33.52.145. ⁹⁵ Money: D. 19.1.38 pr. (Cels. 8 Dig.). Wine: D. 33.6.9.3 (Ulp. 23 Sab.). Buildings: D. 15.1.22 (Pomp. 7 Sab.); D. 33.8.6 pr. (Ulp. 25 Sab.). Cattle: D. 15.3.16 (Alf. 2 dig.). Gold and silver: D. 14.4.5.13 (Ulp. 29 Ed.). Vicarii figure in sixty-four fragments from the Digests (Morabito 1981: 111 n. 605). ⁹⁶ A direct link between the peculium and paying for an inscription de suo is found in several inscriptions: CIL II 6338ff; CIL II² 7.981; CIL XI 6314 = ILS 3581; AE 1903, 140. Arphocras: AE 1980, 150 ( c.50). ⁹⁷ Sen., Ep. 80.8; Hor., Ep. 1.14.40; Petr., Sat. 75.4. ⁹⁸ Wage-earning: D. 12.6.55 (Pap. 6 quaest.); D. 19.2.60.7 (Lab. 5 post.); D. 40.7.3.8 (Ulp. 27 Sab.) and possibly D. 33.7.19.1 (Paul. 13 resp.). Tenancy: D. 15.3.16 (Alf. 2 dig.); D. 18.1.40.5 (Paul. 4 epit.); D. 20.1.32 (Scaev. 5 resp.); D. 26.7.32.3 (Mod. 6 resp.); D. 33.7.12.3 (Ulp. 20 Sab.); D. 33.7.18.4 (Paul. 2 Vit.); D. 33.7.20.1 (Scaev. 3 resp.); C. 4.14.5 (Gord.,  243). ⁹⁹ D. 40.7.14 pr. (Alf. 4 dig.). ¹⁰⁰ Plut., Cato 21.2 and 21.7. ¹⁰¹ Banking: Hipp., Refut. omn. haer. 9.7, but also see D. 14.5.8 (Paul. 1 Decr.). Pasturage: Varro, RR 2.10.5. ¹⁰² The Digest contains various examples of economic activity structured through the peculium. Shops: D. 14.4.5.13 (Ulp. 29 Ed.); D. 14.4.5.16–17 (Ulp. 29 Ed.) and possibly D. 15.1.47 pr. (Paul. 4 Plaut.). Banks: D. 2.13.4.3 (Ulp. 4 Ed.). Inns and

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

59

activity was purposely structured in this manner, with the institor model of direct agency appearing dominant instead (Aubert 1994; Fleckner 2010: 235; Mouritsen 2011: 175). Peculium-run businesses may have been rather small (Abatino and Dari-Mattiacci in this volume) and more often than not an extension of the family economy. Even so, peculia undoubtedly formed an important aspect of the Roman slave household, and it does not seem far-fetched to claim that most slaves held one, especially if they were involved in tasks that were difficult to monitor (Dari-Mattiacci 2013). Rather than an organizational principle for Roman businesses, the peculium appears to have been a reward or incentive for better work, both serving to ameliorate a slave’s condition and to allow the prospect of eventual self-purchase. It was a recognized legal principle that the peculium served as a “freedom fund,” and so a statuliber became free if he was blocked from the peculium which he needed to fulfill the condition of his release (Buckland 1908: 502–3). The use of sub-slaves is of particular interest in this regard. Horace explicitly connects the term vicarius to slave speech by opposing him to the co-slave (conservus) that such a slave actually is.¹⁰³ Sub-slaves could be employed to perform their slave master’s work, netting time for other activities, or to enhance the slave master’s peculium, for instance by being hired out to others. Aside from vicarii working as farmers, sailors, captains, and shopkeepers (Morabito 1981: 112), the Digest also contains the sinister example of sub-slaves put to work as prostitutes.¹⁰⁴ Perhaps the most revealing use of vicarii is that of delivering them pro libertate, in exchange for freedom, either by the slave who is seeking freedom or by a third party.¹⁰⁵ Thus, it is tacitly assumed for such stables: D. 4.9.7.6 (Ulp. 18 Ed.). Mule driving: D. 19.2.60.7 (Lab. 5 post.). Shipping: D. 4.9.3.3 (Ulp. 14 Ed.); D. 4.9.7.6 (Ulp. 18 Ed.); D. 9.4.19.2 (Paul. 22 Ed.); D. 14.1.1.22 (Ulp. 28 Ed.); D. 14.1.6 pr. (Paul. 6 brev.); D. 47.2.42 pr. (Paul. 9 Sab.). Dye factory: D. 32.91.2 (Pap. 7 resp.). Clothing factory: D. 14.4.5.15 (Ulp. 29 Ed.); D. 15.1.27 pr. (Gai. 9 Ed. prov.). Bakery: D. 33.7.18.1 (Paul. 2 Vit.). Managing conveyances: D. 11.6.3.6 (Ulp. 24 Ed.). Prostitution: D. 3.2.4.3 (Ulp. 6 Ed.). ¹⁰³ Hor., Sat. 2.7.79–80. For the jurists, the term has a precise technical meaning: a slave belonging to the peculium of another slave. Vicarii appear in many inscriptions but the meaning is less clear there, since the term was also used as a functional description (“assistant,” “underling”) for slaves belonging to the imperial household, cities, or tax associations (Weaver 1972: 200–6). ¹⁰⁴ Hire: D. 14.3.11.8 (Ulp. 28 Ed.); D. 14.3.12 (Iul. 11 dig.). Prostitution: D. 3.2.4.3 (Ulp. 6 Ed.), though it should be noted that the master slave is considered infamis after manumission. ¹⁰⁵ Slave: D. 41.3.4.16 (Paul. 54 Ed.); D. 41.4.9 (Iul. 3 Urs. Fer.). Third party: D. 19.5.5.2 (Paul. 5 quaest.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

60

Egbert Koops

exchanges that a slave of equal market value is as good as any other slave, and the only question that concerns the jurists is what happens if title to the replacement slave is disturbed.

14.8. ESTABLISHING THE PRICE OF FREEDOM It remains an open question whether Roman slaves paid a premium, a reduced fee (Morabito 1981: 57), or market price (Garnsey 1981: 364) for their freedom. All options are theoretically possible: a slave may have valued his freedom higher than a potential buyer, but the price may also have been lower out of liberality, or because the slave would not have required guarantees. Much will have depended on specific negotiations. The jurists assume an objective market price when discussing the replacement value, although they are certainly aware of subjective considerations. If a slave is killed, emotional attachment plays no part in establishing the damages, though it might have raised the cost of purchasing this specific slave, and the same rule applies if the value of a slave’s labor is to be appraised.¹⁰⁶ Freeing a slave in return for money does not lead to enrichment, as the money completely covers the loss of the slave, but if the slave pays more than his value, presumably the market value, then the master is enriched to the amount of the surplus.¹⁰⁷ A captured slave who has been ransomed from the enemy by someone else and set free need only pay his true master his estimated market price to acquire freedom.¹⁰⁸ If parents sold a child into slavery, Constantine allowed them to reclaim the child against payment of its value or delivery of a replacement slave,¹⁰⁹ again stressing the connection between market price and freedom. For calculations concerning the Falcidian fourth, release under a fideicommissum, or manumission of someone else’s slave by mistake, the market value is again generally used. In all, the overwhelming tendency among the jurists is to value slaves at their market price for replacement purposes. ¹⁰⁶ D. 9.2.33 (Paul. 2 Plaut.); D. 7.7.6.2 (Ulp. 55 Ed.). ¹⁰⁷ D. 15.3.2 (Iav. 12 Cass.); D. 15.3.3 pr. (Ulp. 29 Ed.). By implication, market value and slave are interchangeable while the surplus constitutes a gift from the peculium to the master. ¹⁰⁸ D. 29.2.71 pr. (Ulp. 61 Ed.). ¹⁰⁹ Fr. Vat. 34 (Const.,  313); C.Th. 5.10.1 (Const.,  329).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

61

The scarcity of credible and comparable manumission prices makes it difficult to establish a relation between the market price and the price of freedom in practice.¹¹⁰ Almost all slave prices mentioned in the Digest are unusable as evidence (Buckland 1908: 8)¹¹¹ and many of the literary prices are suspect. In Delphi, Hopkins (1978: 158–63) found evidence for a rising manumission premium for full freedom due to increased prices as a result of market integration. If anything, manumission prices seem to be on the high side as well in the documents from Roman Egypt, suggesting a premium, but the evidence does not allow for any conclusions. The following prices are merely given as an indication,¹¹² but they do show that manumission and replacement at market value operate on the same plane, or put differently, that the price of freedom is not an order of magnitude higher than the market value of slaves. Circumstantial support comes from relating slave prices to prices for other commodities. Ruffing and Drexhage (2008: 340–5) found that the somewhat believable documented slave prices could be expressed in terms of 250–1,000 day wages for the first century  and 200–1,000 day wages for the second century. For Roman Egypt, where better data are always available, prices lay between 1,200–2,000 day wages in the first century, between 250–2,800 day wages in the second century, and between 400–1,125 day wages in the third century, before the massive inflation of the last quarter of that century (Table 14.2). Converting slave prices to wheat equivalent, Scheidel (2005b: 2–6) arrived at prices that represented 2.7–3.2 years of income for a rural laborer in Roman Egypt. Diocletian’s Edict on maximum prices of  301 contains a schedule of slave prices, as well as ample figures on wages and daily necessities (Lauffer 1971;

¹¹⁰ Interestingly, the Sumerian Code of Lipit-Ishtar from the nineteenth century  contains a provision (§14) that a slave shall be freed if he compensates his master twofold. ¹¹¹ Slave prices are mentioned in 133 fragments of the Digest. The price is almost always 10, with 5 being half a slave or a pledged slave, and 20 being two slaves, a skilled slave, or a case of mistake or deception. Justinianic price schedules aside, the only actual price seems to be the fixed sum of 20 aurei that was to be paid by slaves who had been set free under an invalid will yet wanted their freedom: D. 4.4.31 (Pap. 9 resp.); D. 5.2.8.17 (Ulp. 14 Ed.); D. 5.2.9 (Mod., inoff. test.); D. 40.4.47 pr. (Pap. 6 quaest.). Contra Morabito 1981: 54–9; Scheidel 2005b: 6. ¹¹² See Ruffing and Drexhage 2008: 321–36 for the completest list to date. On the sale of slaves in Roman Egypt, see Straus 2004.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

62

Egbert Koops

Table 14.2. Some Roman-Egyptian slave prices (in silver dr.). Manumission prices have been bolded.¹¹³ Year ()

Price

Age

Source

c.80–100 86 91/92 101 107–15 c.100–25 ... 212 221 221 225 234 c.235

1,161 1,125 900 1,000 500 700

23 – 32 17 0–19 4

P. Oxy. LXXV 5051 P. Oxy. I 48 + XXXVIII 2843 P. Oxy. XXXVIII 2856 P. Turn. 19 P. Strasb. VI 505 BGU XI 2111

1,600 2,200 2,200 1,600 2,200 1,600

19 34 14 9 25 –

P. Oxy. XXXVI 2777 FIRA III 11 P. Vind. Bos. 7 SB XIV 11277 PSI III 182 P. Oxy. XLIII 3117

Salway 2010).¹¹⁴ It suggests that a male unskilled slave between 16 and 40 (maximum price: 30,000 denarii) cost the equivalent of 4.8–6 years of unskilled day wages (maximum price: 20–25 denarii), at 250 working days per year. It is under debate whether slave labor and wage labor stood in direct competition, and if so for which occupations (Harper 2010: 212–14, Groen-Vallinga 2017: 280–2). But even so it is interesting to note that a slave working as a day laborer— which included limited meals—would potentially earn back his purchase price within a period of years rather than decades. The Price Edict supplies enough information to fill out this amortization period by using a ‘consumption basket’ of daily necessities. Adapting the figures supplied by Allen (2009: 327–43) and assuming no rent payments, with lodgings provided by the master, the investment in an unskilled male slave of 18 years might be recouped between 4.7 years (“bare bones diet”) and 7.7 years (“respectable diet”), assuming ¹¹³ Prices in talents and copper drachmae have been converted using the scale: 1 talent = 6,000 copper drachmae = 100 silver drachmae. For tax purposes, 300 copper dr. converted to 1 silver dr., but a conversion rate of 56:1 seems to have been used in practice (Johnson 1936: 282–3). Note that P. Oxy. XXXVIII 2856 mentions both prices, at a conversion rate closer to 69:1. ¹¹⁴ Since the Price Edict offers maximum prices, without explaining their provenance or their internal relation, the schedule should be used with care. Scheidel 1996; Harper 2010: 219–20.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

63

250 working days per year. A late legal source implies an awareness of this range: if a captive is ransomed from enemies, he has to repay the ransom price or work off the debt in five years’ time.¹¹⁵ Skilled slaves may have earned more, but their value was higher as well, since skill level is an important determinant of market price.¹¹⁶ Diocletian’s Edict expressly provides that the price for skilled slaves is at most double that of unskilled slaves, which corresponds to the differentiation between skilled (50–75 denarii per day) and unskilled labor (20–25 denarii per day), and so the range stays roughly the same. Interestingly enough, the high prices that are mentioned for certain exceptional slaves (doctors, grammarians, actors), fall in the same range when set against the comparably high figures for their wages: the ratio is roughly 0.2 to 7 years of wages.¹¹⁷ Needless to say, these figures need to be taken with a large grain of salt. Their purpose is merely to draw attention to the fact that the overall ratio is one to eight years of wages, tending to five to seven years, rather than twenty to thirty years. How long, then, would a slave have to labor to purchase his freedom at a close to market price? Unfortunately, it is almost impossible to relate manumission prices to specific terms of servitude. Not a single source refers to the amount that a master would allow slaves to keep for themselves in the form of wages. Seneca states that a slave actor received 5 denarii and 5 modii of wheat for a show, but he does not mention the master’s cut; Martial refers to a gratuity of 25 asses.¹¹⁸ Such figures are hardly enough to hazard a guess, and in any case the amount of wages will have differed from case to case. Other legal systems from antiquity such as Babylonian and Mosaic law mention relatively short terms of servitude, three to six years, although these relate mostly to debt bondage and legislation

¹¹⁵ C.Th. 5.7.2 = Const. Sirm. 16 (Hon.,  408) = C. 8.50(51).20.2 (Hon.,  409). ¹¹⁶ The jurists valued the use of a skilled slave against his market price, but the use of an unskilled slave against the daily wages earned by his usual work: D. 7.7.6 pr. (Ulp. 55 Ed.). ¹¹⁷ Doctor: 250,000 HS per year (Plin., NH 29.5.7, imperial physician), manumission 50,000 HS (CIL XI 5400 = ILS 7812). Actor: 200,000 HS per year (Cic., Q. Rosc. 23), manumission 700,000 HS (Plin., NH 7.39.128). Grammarian: 100,000 HS per year (Suet., Gramm. 17), sale 700,000 HS (Plin., NH 7.39.128). ¹¹⁸ Sen. Ep. 80.5: Ille qui in scaena latus incedit et haec resupinus dicit [ . . . ] servus est, quinque modios accipit et quinque denarios; Mart. 10.75: Sportula [ . . . ] quadrantibus arida centum [ . . . ] puero diximus esse datam.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

64

Egbert Koops

on periodic debt remission (Wiedemann 1985: 166; Chirichigno 1993).¹¹⁹ Roman law is less forthcoming. As mentioned, the manumission prices from the Digest yield no information; similarly, the terms of servitude provide little insight (Wiedemann 1985: 169–74). Testamentary provisions for terms of one, two, three, five, seven, ten, and even fifteen years are mentioned, as well as provisions for payment of certain sums during or within three, five, or ten years. But without any particulars of a slave’s age and status, such clauses mean very little. Neither provisions that a slave is to be released after a certain period of time (examples are one, two, three, five, eight, ten, or twelve years), nor clauses that he is to be released when the heir or the taker under a legacy dies or comes of age, can be related to a specific period of servitude. If a manumission clause is coupled to a slave’s age, it is almost always the 30 years of age that corresponds to the lex Aelia Sentia.¹²⁰ The one exception occurs in a rescript from Alexander Severus considering a slave girl named Firmia who was sold at age 7, to be released at age 25.¹²¹ The single explicit reference to a common term of servitude comes from Cicero, who informs the senate that it entertains some hope of freedom after enduring six years of slavery, which is longer than diligent captives and frugal slaves usually suffer.¹²² Cicero is referring to the time that has passed since Caesar crossed the Rubicon, and so the period is of limited value (Mouritsen 2011: 137). Nevertheless, the reference would fall flat if slaves only had a chance of freedom after, say, thirty productive years. Considering the social advantage of gaining freedmen while recapitalizing, relatively short terms of servitude seem likely for those who became slaves later in life, while the practice of manumitting certain slaves at an early age (§14.3) may have been the lot of those born into slavery or enslaved at a very young age. Again, it should be remembered that some slaves were never freed.

¹¹⁹ Codex Hammurabi 117 (in the fourth year); Exod. 21.2; Deut. 15.12–18; Philo, Spec. leg. 2.84 (in the seventh year). ¹²⁰ D. 10.2.39.2 (Scaev. 1 resp.); D. 34.5.29(30) (Scaev. 18 dig.); D. 40.4.46 (Pomp. 7 var. lect.); D. 40.7.13.5 (Iul. 43 dig.). ¹²¹ C. 4.57.3 pr. (Alex.,  224). ¹²² Cic., Phil. 8.32: Etenim, patres conscripti, cum in spem libertatis sexennio post sumus ingressi diutiusque servitutem perpessi, quam captivi servi frugi et diligentes solent.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

65

14.9. PATRONS AND FREEDMEN Hopkins’s question, “why did the Romans free so many slaves?,” is best answered by a model in which the benefits of allowing slaves autonomy by the undisturbed use of their peculia, while factoring in the cost of peculation, outweigh the gains of brute exploitation. It implies that value extraction and recapitalization are delayed until an agreed-upon price has been met, because a slave cum peculio will eventually outperform a slave sine peculio (Alföldy 1972: 360–2). This scenario is not so unlikely, particularly when supervision costs are taken into account (Dari-Mattiacci 2013), as successful manumission deals inspire confidence in other slaves, leading to better work and subservience. But aside from the material benefit of better work at lesser cost, and the immaterial benefit of strengthened social peace, furthered by making some manumission agreements actionable, there is another mechanism at work to ensure that masters keep their word. The loss of a slave is sweetened by the acquisition of a freedman.¹²³ Few parts of Roman law are as convoluted as the law of patronage. The rights of patrons take a number of forms, and these rights also vary considerably according to the type of manumission. The difference between slaves manumitted as citizens and as Junian Latins has already been mentioned (§14.4), but even for formally manumitted slaves, the legal relationship to their patron was contingent on the mode and conditions of manumission. Freedmen manumitted directly and among the living out of kindness or generosity remained firmly under their patron’s control and supervision. All family relations were destroyed by slavery, and so the manumittor was legally closest to a parent, to whom a duty of obedience and deference was owed (Mouritsen 2011: 36–51).¹²⁴ But as a general rule, a patron who ¹²³ The patron’s patrimony is enriched by acquiring a freedman. Thus, a manumittor in good faith of a slave that was transferred to him in error is enriched to the value of the operae and his claim to the freedman’s inheritance: D. 12.6.65.8 (Paul. 17 Plaut.). From the palingenesis, it is clear that the far too general statement in D. 50.17.126.1 (Ulp. 15 Ed.), that no one is enriched who acquires a freedman, initially only referred to the Sc. Iuventianum of  129 concerning the vindicatio caducorum. See D. 5.3.20.6c (Ulp. 15 Ed.); D. 5.3.23 pr. (Ulp. 15 Ed.). A counter argument is not provided by D. 19.5.5.5 (Paul. 5 quaest.) either, since the reason why the freedman’s value is not calculated in that particular case is that the action is limited to quanti interest mea servum habere in the first place. ¹²⁴ The duty of obsequium covers any number of rules, running from an inability to sue the patron in certain suits or raise certain defenses, to an obligation enforced by

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

66

Egbert Koops

had received compensation for manumission, be it in payment or as taker under a fideicommissum, had fewer rights against his freedman (Garnsey 1981: 363–4). By extension, a trustee owner such as the transferee ut manumittatur or the buyer in an emptio suis nummis had even fewer rights, since he was merely lending his name to the manumission and not making any personal sacrifice at all.¹²⁵ Finally, the slave who was freed by testament, the orcinus, was under little obligation at all (Loreti-Lorini 1925; Champlin 1991: 139) since his true patron was dead and the patron’s heir, being next in line, had made no sacrifice nor willed the manumission. All the more reason existed to apply this rule if a slave was freed as a statuliber by testament, for instance under the condition of rendering an accounting, paying a sum of money, or working a number of years, since the heir even gained by the process when compared to direct manumission. If and when the condition was met, the former statuliber became a freedman of the deceased testator and not of the testator’s heir. Within this continuum, two categories of rights stand out for the present purpose since they carry a direct monetary value. These are claims on day labor or services (operae) and claims on the estate of a freedman upon his death, either in the form of a right to intestate succession or in the form of a claim for possession in contravention of the will (bonorum possessio contra tabulas). An overview is provided in the following, rudimentary table (Table 14.3). The claim to labor duties may seem paramount among a patron’s rights since it ensures the continued services of a freedman after manumission. But this is not the case. Operae could only be imposed upon freedmen who had been manumitted for free, and then only if they had promised or sworn to perform such duties (Garnsey 1981: 364; Gardner 1993: 20, 29–32).¹²⁶ They could not be imposed if the master had accepted any money in return for manumission or had only served as a trustee manumittor, at least since the time of Hadrian (Morabito 1981: 168–9). In other words, the obligation to perform operae was mutually exclusive with paid-for manumission. It appears

public law to treat patrons with respect (Gardner 1993: 23–5). Similar considerations of the “parental” nature of the relationship underlie a freedwoman’s need for permission to marry, reciprocal obligations of tutelage, and reciprocal claims for maintenance (alimenta) between patron and freedman. ¹²⁵ D. 40.1.4.7 (Ulp. 6 disp.). ¹²⁶ D. 38.1.7.2 (Ulp. 28 Sab.); D. 38.1.31 (Mod. 1 reg.); D. 40.4.36 (Paul. 7 Plaut.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

67

Table 14.3. Patron rights and the effects of manumission. Adapted from LoretiLorini 1925. Manumission occurred . . .

The patron acquires a right to . . .

Inter vivos, and

Labor Intestate Overrule a (operae) succession testament

Be maintained Deference (alimenta) (obsequium)

- for free - by pactio - by transfer to be freed (ut manumittatur) - by purchase suis nummis

Yes¹²⁷ No¹³² No¹³⁷

Yes¹²⁸ Yes¹³³ Yes?¹³⁸

Yes¹²⁹ Yes¹³⁴ Yes¹³⁹

Yes¹³⁰ Yes?¹³⁵ Yes?¹⁴⁰

Yes¹³¹ Yes¹³⁶ Yes¹⁴¹

No¹⁴²

No¹⁴³

No¹⁴⁴

No¹⁴⁵

No¹⁴⁶ (cont. )

to have been the functional equivalent of borrowing the manumission price from the master and paying it off in instalments. Freedmen operae are accounted in labor days while the obligation to perform a certain number of labor days as a slave is conversely mentioned as a possible condition for manumission.¹⁴⁷ The analogy between paying before and paying after manumission through labor is strengthened by the fact that freedmen could agree to pay a sum of money instead of performing labor, though patrons could not force them to do so ¹²⁷ C. 6.3.1 (Sev./Ant.,  204). ¹²⁸ Gai., Inst. 3.41; D. 26.4.3 pr. (Ulp. 38 Sab.); D. 34.5.9(10).2 (Tryph. 21 disp.). ¹²⁹ Gai., Inst. 3.41. ¹³⁰ D. 25.3.5.18 (Ulp. 2 cons.); C. 6.3.1 (Sev./Ant.,  204). ¹³¹ D. 37.15.7.4 (Ulp. 10 Ed.); D. 38.2.1 pr. (Ulp. 42 Ed.); D. 44.4.4.16 (Ulp. 76 Ed.); D. 47.2.92(91) (Lab./Paul. 2 pith.); C. 2.41(42).2 (Just.,  531); C. 6.7.1 (Ant.,  214). ¹³² C. 6.3.3 (Sev./Ant.,  206). ¹³³ C. 6.4.1 pr. (Sev./Ant.,  210). ¹³⁴ C. 6.4.1 pr. (Sev./Ant.,  210) contra D. 38.2.3.4 (Ulp. 41 Ed.), which is generally amended. ¹³⁵ By inference: no direct source available. ¹³⁶ D. 37.15.3 (Marc., resp.); C. 6.6.3 (Alex.,  223). ¹³⁷ C. 6.3.2 (Sev./Ant.,  205); D. 38.1.13 pr. (Ulp. 38 Ed.). ¹³⁸ By inference: no direct source available. ¹³⁹ D. 38.2.3.3 (Ulp. 41 Ed.). ¹⁴⁰ By inference: no direct source available. ¹⁴¹ D. 2.4.10 pr. (Ulp. 5 Ed.). ¹⁴² C. 6.3.8 (Alex.,  224). ¹⁴³ C. 6.4.4.4 (Just.,  531). ¹⁴⁴ C. 6.4.1 pr. (Sev./Ant.,  210); C. 6.4.4.4 (Just.,  531). ¹⁴⁵ D. 25.3.5.22 (Ulp. 2 cons.). ¹⁴⁶ D. 2.4.10 pr. (Ulp. 5 Ed.); D. 37.15.3 (Marc., resp.); C. 6.3.8 (Alex.,  224). ¹⁴⁷ D. 38.1.3.1 (Pomp. 6 Sab.); D. 38.1.15.1 (Ulp. 38 Ed.); D. 40.7.4.4 (Paul. 5 Sab.). Also see D. 40.7.14.1 (Alf. 4 dig.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

68

Egbert Koops

Table 14.3. Continued Manumission occurred . . .

The patron acquires a right to . . .

Inter vivos, and

Labor Intestate Overrule a (operae) succession testament

Be maintained Deference (alimenta) (obsequium)

By testament, and

Labor Intestate Overrule a (operae) succession testament

Be maintained (alimenta)

Deference (obsequium)

- directly (orcinus) - by fideicommissum

No¹⁴⁸ No¹⁵³

No¹⁵¹ No¹⁵⁶

Some¹⁵² Some¹⁵⁷

Yes¹⁴⁹ Yes¹⁵⁴

Yes¹⁵⁰ Yes¹⁵⁵

(Morabito 1981: 91–2).¹⁵⁸ Again, the number of operae owed would have been a matter of negotiation (Verboven 2012: 97). Seen in this light, there is little economic difference between manumission among the living under the imposition of operae, and manumission by testament under the condition to perform a number of services or pay a sum of money. The legal difference, however, is that manumission is already complete under the first option, while slavery ¹⁴⁸ D. 38.1.13.1 (Ulp. 38 Ed.). Changed by Justinian so that a patron’s son could impose operae (Loreti-Lorini 1925: 53). See D. 38.2.29.1. (Marc. 9 inst.); D. 40.5.33 pr. (Paul. 3 fideic.). ¹⁴⁹ Gai., Inst. 3.41. ¹⁵⁰ Gai., Inst. 3.41. ¹⁵¹ The argument is that a slave who was to be released under a fideicommissum that the taker neglected to fulfill was declared free under the Sc. Rubrianum of  103 and given an equal position to that of an orcinus: D. 40.5.26.7 (Ulp. 5 fideic.); D. 40.5.5 (Paul. 57 Ed.). By extension, the same held true for a slave bought suis nummis whom his fiduciary master would not free: D. 5.1.67 (Ulp. 6 disp.). The true orcinus cannot have been worse off than these orcini by analogy. Therefore, there probably was no obligation to provide alimenta or act with full deference. See Loreti-Lorini 1925: 35–41. ¹⁵² See the previous footnote. ¹⁵³ D. 38.1.7.4 (Ulp. 28 Sab.); D. 38.1.13.1 (Ulp. 38 Ed.); D. 38.1.42 (Pap. 9 resp.); D. 38.1.47 (Val. 6 fideic.); D. 38.2.29 pr. (Marc. 9 inst.); C. 6.3.5 (Ant.,  212); C. 6.4.4,7 (Just.,  531). Changed by Justinian so that a patron’s son could impose operae (Loreti-Lorini 1925: 53). See D. 38.2.29.1. (Marc. 9 inst.); D. 40.5.33 pr. (Paul. 3 fideic.). ¹⁵⁴ Fr. Vat. 225 (Pap. 11 quaest.); D. 38.2.29 pr. (Marc. 9 inst.); D. 38.16.3.1 (Ulp. 14 Sab.). ¹⁵⁵ Fr. Vat. 225 (Pap. 11 quaest.); D. 38.2.29 pr. (Marc. 9 inst.); C. 6.13.1 (Gord.,  239). ¹⁵⁶ D. 25.3.5.22 (Ulp. 2 cons.); C. 6.4.4.7 (Just.,  531). ¹⁵⁷ D. 2.4.9 (Paul. 4 Ed.); C. 6.3.5 (Ant.,  212); C. 6.7.1 (Ant.,  214). ¹⁵⁸ D. 38.1.32 (Mod. 1 reg.); D. 38.1.39 pr. (Paul. 7 Plaut.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

69

continues under the second option, albeit under the unassailable position of a statuliber. The paramount right of the patron, then, lies in the claim to inherit part of the estate of a freedman (Gardner 1993: 21–3) in addition to either receiving a manumission price or imposing operae. This claim travelled along the male line,¹⁵⁹ ensuring that testamentary manumission, for example, did not overly diminish the family fortune. A freedman could only exclude his patron from inheriting by leaving the inheritance to his freeborn natural children. Even then, strengthening a patron’s right to sizable estates (over 100,000 sesterces), the Augustan lex Papia Poppaea of  9 provided that a patron was entitled to a share equal to that of the freedman’s children, while it took three surviving freeborn children to exclude the patron completely. In light of the regular age at manumission, plausibly between 30 and 40, complete exclusion was possible but perhaps not likely. This is even more true for women, who tended to be manumitted at a later age (Gardner 1993: 22; Scheidel 1997: 167–8).¹⁶⁰ In addition, it was made legally difficult to resort to subterfuge to escape the patronal claim. Donating assets to children prior to death would be of no avail, since two actions were in place by the first century to rescind alienations by a freedman that could defraud the patron, further protecting his interest (Lenel 1927: 352–3; Gardner 1993: 23).¹⁶¹ This well-protected expectation of inheritance further strengthened the predilection of masters to delay extraction from their slaves until the payment of a manumission price, when they would receive replacement value and gain a freedman, instead of settling for full exploitation of a slave’s labor (Table 14.4). Manumission did not amount to a full loss of the slave’s earning capacity. In fact, a patron might benefit if a freedman did well for himself, which in turn ensured a continuing supervision that was desirable to the state.

¹⁵⁹ The claim to a Junian Latin’s inheritance was a simple patrimonial right that did not follow the rules of agnatic succession, but rather the normal rules of inheritance. Gai., Inst. 3.58 and following. ¹⁶⁰ Freedwomen would have had to acquire the ius quattuor liberorum first (Kübler 1909), freeing them from their patron’s tutelage by giving birth to four freeborn children, before they could make their own testament and even think of bypassing their patron. Considering the pattern of female manumission past childbearing age, acquisition of this right must have been limited to exceptional cases of early manumission. An example is perhaps found in CIL VI 10246 (Septimia Dionisias). ¹⁶¹ The Fabian and Calvisian actions: D. 38.5 rubr. and C. 6.5 rubr.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

70

Egbert Koops Table 14.4. Claims to freedman inheritance (Gai., Inst. 3.40–3.44 and 3.56; Ulp., Epit. 29.1–29.3). Freedman has . . .

Patron receives . . .

Status of Junian Latin No heres suus Extraneus as heir (not a child) 1 Child as heir and under 100,000 HS 1 Child as heir and 100,000+ HS 2 Children as heir and 100,000+ HS 3 Children as heir and 100,000+ HS

All All Half Nothing Half One-third Nothing

Freedwoman has . . .

Patron receives . . .

No testament Testament, less than 4 children Testament, 4 living children Testament, 3 living children, 1 dead Testament, 4 dead children

All Patron decides One-fifth One-quarter All

This legal structure also explains why the remainder of the peculium could be handed to a freedman if anything was left after paying the freedom price. Patrons might confidently expect that after serving as an incentive and a freedom fund, the money would serve as venture capital (Garnsey 1981: 367–70; Mouritsen 2011: 176–80), to be returned to the family fortune with interest in good time. Since many skilled freedmen went on to work in the family business, capital accumulation could be monitored by the patron and was safeguarded by economic family ties (Verboven 2012: 98, Groen-Vallinga 2017: 146–52). The one modality that escapes this framework is the emptio suis nummis (Heinemeyer 2013). Recognized by Marcus Aurelius, in this legal construction the slave was purchased by a fiduciary buyer for the express purpose of manumission, with money coming either from the slave’s peculium or borrowed to the purpose. The seller was not manumitting the slave and so he did not become patron, while the buyer merely lent his name to the arrangement and made no sacrifice, leading to a near-complete loss of patron’s rights including the right to inherit. This particular arrangement would seem to defeat the entire purpose of manumission for the original owner. Why not revoke the peculium, or even contract with the slave through a

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

71

manumission pact, retaining the patron rights, instead of selling the slave to a fiduciary buyer? As with slaves freed under the condition of rendering an accounting—necessary to truthfully establish the size of an inheritance—some slaves will have had enough power and influence that the realization of replacement value from their no doubt hidden peculia required the surrender of patron rights. In this view, emptio suis nummis is not the starting point of legal development toward actionable manumission agreements, but rather the secondcentury endpoint. Instead of an unattractive proposition that was to be avoided in favor of other modalities (Gardner 1993: 37), or an anomaly (Mouritsen 2011: 173–4), its existence is testament to the bargaining power of certain well-placed slaves (Morabito 1981: 166–7). It put the enforceability of manumission arrangements made inter vivos on an equal level to that of testamentary arrangements and effectively neutralized almost all of a patron’s claims. As such, it formed a logical pinnacle to the Roman system of negotiated freedom payments.

14.10. CONCLUSION Conditional manumissions are known from the Hellenistic world and Roman Egypt (Zelnick-Abramovitz 2005). This chapter has suggested that the Roman lawyers, at least, considered the fulfillment of certain conditions the normal precursor to Roman manumission as well. Important differences exist between Greece, Egypt, and Rome, particularly with regard to the residual power over conditionally freed slaves, but the general point stands that relations between masters and slaves were formalized by negotiations and agreements concerning manumission. By enshrining an otherwise merely morally binding agreement in a testament, it became enforceable by the slave after his master’s death. Gradually, as a result of imperial intervention, most manumission agreements made inter vivos became enforceable as well. The general effect of setting conditions prior to manumission was that a patron’s hold over his freedman was lessened; or put differently, that a patron who accepted a price for freedom lost part of his bundle of rights. Aside from the emptio suis nummis, however, the most important patron right remained intact, in the form of an expectancy to inherit.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

72

Egbert Koops

Agreements between master and slave could concern the period of servitude, but more often referred to a manumission price. This price was paid from a slave’s peculium, assets which masters allowed slaves to hold as their own (Garnsey 1981: 364). Serving both as an incentive and as a freedom fund, the slave peculium and the concomitant social practice of purchasing freedom explain why the Romans freed so many slaves. Though it is difficult to relate the size of the peculium to the proceeds of a slave’s labor or the duration of slavery, its social importance can hardly be underestimated. It gave slaves an interest in their labor and the hope of advancement on the one hand, while reducing supervision costs, increasing security and raising productivity for their masters on the other. Though there was no legal guarantee of tenure, the peculium and the negotiations and agreements surrounding it show that at least a certain class of Roman slaves, that is to say those encountered in literary and epigraphic sources, rose well above the status of socially dead instrumentum vocale. The practice of frequent and early manumission in exchange for a negotiated and agreed upon price does not make the Roman slave system any less brutal or exploitative during the period of enslavement. But it does help to understand why masters, slaves, and freedmen, living in close proximity, largely succeeded in maintaining social order under the Roman empire.¹⁶²

REFERENCES Abatino, Barbara, Giuseppe Dari-Mattiacci, and Enrico C. Perotti. 2011. “Depersonalization of business in ancient Rome.” 31 Oxford Journal of Legal Studies 365–89. Alföldy, Géza. 1972. “Die Freilassung von Sklaven und die Struktur der Sklaverei in der römischen Kaiserzeit.” 2 Rivista Storica dell’Antichita 97–129; repr. with add. in Helmuth Schneider, ed., Sozial- und Wirtschaftsgeschichte der römischen Kaiserzeit, Darmstadt: WBG 1981, 336–71. Alföldy, Géza. 2011. Römische Sozialgeschichte, 4th edn. Stuttgart: Steiner.

¹⁶² Research for this chapter was conducted under grant 400-09-009 from the Netherlands Organisation for Scientific Research (NWO).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

73

Allen, Robert C. 2009. “How prosperous were the Romans? Evidence from Diocletian’s Price Edict (AD 301),” in Alan K. Bowman and Andrew Wilson, eds, Quantifying the Roman Economy: Methods and Problems. Oxford: Oxford University Press, 327–45. Arzt-Grabner, Peter. 2010. “ ‘Neither a truant nor a fugitive’: some remarks on the sale of slaves in Roman Egypt and other provinces,” in Traianos Gagos, ed., Proceedings of the 25th International Congress of Papyrology. Ann Arbor: Scholarly Publishing Office, 21–32. Aubert, Jean-Jacques. 1994. Business Managers in Ancient Rome: A Social and Economic Study of Institores, 200 ..–.. 250. Leiden/New York: Brill. Bagnall, Roger S., and Bruce W. Frier. 1994. The Demography of Roman Egypt. Cambridge: Cambridge University Press. Bradley, Keith R. 1987. Slaves and Masters in the Roman Empire: A Study in Social Control. Oxford: Oxford University Press. Bradley, Keith R. 1994. Slavery and Society at Rome. Cambridge: Cambridge University Press. Buckland, William W. 1908 [repr. 1970]. The Roman Law of Slavery. Cambridge: Cambridge University Press. Cerami, Pietro, and Aldo Petrucci. 2010. Diritto commerciale romano: profilo storico, 3rd edn. Turin: Giappichelli. Champlin, Edward. 1991. Final Judgments: Duty and Emotion in Roman Wills, 200 ..–.. 250. Berkeley: University of California Press. Chirichigno, Gregory C. 1993. Debt Slavery in Israel and the Ancient Near East. Sheffield: Sheffield Academic Press. Cohen, Boaz. 1951. “Peculium in Jewish and Roman law.” 20 Proceedings of the American Academy for Jewish Research 135–234. Dari-Mattiacci, Giuseppe. 2013. “Slavery and information.” 73 Journal of Economic History 79–116. Di Porto, Andrea. 1984. Impresa collettiva e schiavo ‘manager’ in Roma antica (II sec. a.C.–II sec. d.C.). Milan: Giuffrè. Duff, Arnold M. 1928. Freedmen in the Early Roman Empire. Oxford: Clarendon Press. Emmerson, Allison A. C. 2011. “Evidence for Junian Latins in the tombs of Pompeii?” 24 Journal of Roman Archaeology 161–90. Fenoaltea, Stefano. 1984. “Slavery and supervision in comparative perspective: a model.” 44 Journal of Economic History 635–68. Finkenauer, Thomas. 2010. Die Rechtsetzung Mark Aurels zur Sklaverei. Stuttgart: Franz Steiner. Finley, Moses I. 1985 [repr. 1999]. The Ancient Economy, 2nd edn. Berkeley: University of California Press. Fleckner, Andreas M. 2010. Antike Kapitalvereinigungen. Ein Beitrag zu den konzeptionellen und historischen Grundlagen der Aktiengesellschaft. Cologne: Böhlau.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

74

Egbert Koops

Gardner, Jane F. 1993. Being a Roman Citizen. London/New York: Routledge. Garnsey, Peter. 1981. “Independent freedmen and the economy of Roman Italy under the Principate.” 63 Klio 359–71. Groen-Vallinga, Miriam J. 2017. The Roman World of Work, (diss. Leiden). Rotterdam: Optima Grafische Communicatie. Günther, Sven. 2008. “Vectigalia nervos esse rei publicae”: Die indirekten Steuern in der Römischen Kaiserzeit von Augustus bis Diokletian. Wiesbaden: Harrassowitz. Harada, Keikichi. 1939a. “Textkritische Studien zur adsignatio liberti orcini.” 59 Zeitschrift der Savigny-Stiftung 498–508. Harada, Keikichi. 1939b. “Zwei Quellenstellen zum Patronatsrecht,” in Festschrift Paul Koschaker I. Weimar: Böhlau, 401–3. Harper, Kyle. 2008. “The Greek census inscriptions of late Antiquity.” 98 Journal of Roman Studies 83–119. Harper, Kyle. 2010. “Slave prices in late antiquity (and in the very long term).” 59 Historia 206–38. Harris, William V. 1980. “Toward a study of the Roman slave trade.” 36 Memoirs of the American Academy in Rome 117–40. Harris, William V. 1999. “Demography, geography and the sources of Roman slaves.” 89 Journal of Roman Studies 62–75. Heinemeyer, Susanne. 2013. Der Freikauf des Sklaven mit eigenem Geld— Redemptio suis nummis. Berlin: Duncker & Humblot. Hopkins, Keith. 1978. Conquerors and Slaves. Cambridge: Cambridge University Press. Hopkins, Keith. 1993. “Novel evidence for Roman slavery.” 138 Past & Present 3–27. Horsmann, Gerhard. 1986. “Die ‘divi fratres’ und die ‘redemptio servi suis nummis’ (zu den Motiven der epistula ad Urbium Maximum, Dig. 40,1,4).” 35 Historia 308–21. Johnson, Allan C. 1936. Roman Egypt to the reign of Diocletian, in Frank Tenney, ed., An Economic Survey of Ancient Rome, vol. II. Baltimore: Johns Hopkins University Press. Johnston, David. 2002. “Peculiar questions,” in Paul McKechnie, ed., Thinking Like a Lawyer: Essays on Legal History and General History for John Crook on his Eightieth Birthday. Mnemosyne suppl. 231. Brill: Leiden, 5–13. Kaser, Max. 1971. Das römische Privatrecht. Vol. I: Das altrömische, das vorklassische und klassische Recht, 2nd edn. Munich: Beck. Kleijwegt, Marc. 2002. “Cum vicensimariis magnam mantissam habet (Petronius Satyricon 65.10).” 123 American Journal of Philology 275–86. Kleijwegt, Marc. 2012. “Deciphering freedwomen in the Roman Empire,” in Sinclair Bell and Teresa Ramsby, eds, Free at Last! The Impact of Freed Slaves on the Roman Empire. London: Bloomsbury, 110–29.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

75

Koops, Egbert. 2012. “Second-rate citizens: Junian Latins and the constitutio Antoniniana.” 19 Maastricht Journal of European and Comparative Law 223–39. Kübler, Bernhard. 1909. “Über das Jus liberorum der Frauen und die Vormundschaft der Mutter. Ein Beitrag zur Geschichte der Rezeption des römischen Rechts in Ägypten.” 30 Zeitschrift der Savigny-Stiftung 154–82. Lauffer, Siegfried. 1971. Diokletians Preisedikt. Berlin: De Gruyter. Lenel, Otto. 1927 [repr. 2010]. Das edictum perpetuum. Ein Versuch zu seiner Wiederherstellung, 3rd edn. Leipzig: Tauchnitz. Liebs, Detlef. 2000. “Das testament des Antonius Silvanus, römischer Kavallerist in Alexandria bei Ägypten, aus dem Jahr 142 n. Chr.,” in Klaus Märker, ed., Festschrift für Weddig Fricke zum 70. Geburtstag. Freiburg: Alber, 113–28. Lopez Barja de Quiroga, Pedro. 1998. “Junian Latin: status and number.” 86 Athenaeum 133–64. Loreti-Lorini, Bradamante. 1925. “La condizione del liberto orcino nella compilazione giustinianea.” 34 Bullettino dell’Istituto di Diritto Romano 29–66. Morabito, Marcel. 1981. Les réalités de l’esclavage d’après le Digeste. Paris: Belles Lettres 1981. Mouritsen, Henrik. 2005. “Freedmen and decurions: epitaphs and social history in imperial Italy.” 95 Journal of Roman Studies 38–63. Mouritsen, Henrik. 2011. The Freedman in the Roman World. Cambridge: Cambridge University Press. Patterson, Orlando. 1982. Slavery and Social Death: A Comparative Study. Cambridge, MA: Harvard University Press. Roth, Ulrike. 2010. “Peculium, freedom, citizenship: golden triangle or vicious circle? An act in two parts,” in Ulrike Roth, ed., By the Sweat of your Brow: Roman Slavery in its Socio-Economic Setting (BICS Suppl. 109). London: Institute of Classical Studies, 91–120. Ruffing, Kai, and Hans-Joachim Drexhage. 2008. “Antike Sklavenpreise,” in Peter Mauritsch, ed., Antike Lebenswelten: Konstanz—Wandel— Wirkungsmacht. Festschrift für Ingomar Weiler zum 70. Geburtstag (Philippika 25). Wiesbaden: Harrassowitz, 321–51. Salway, Benet. 2010. “Mancipium rusticum sive urbanum: the slave chapter of Diocletian’s edict on maximum prices,” in Ulrike Roth, ed., By the Sweat of your Brow: Roman Slavery in its Socio-Economic Setting (BICS Suppl. 109). London: Institute of Classical Studies, 1–20. Scheidel, Walter. 1996. “Reflections on the differential valuation of slaves in Diocletian’s Price Edict and in the United States.” 15 Münstersche Beiträge zur Antiken Handelsgeschichte 67–79. Scheidel, Walter. 1997. “Quantifying the sources of slaves in the early Roman Empire.” 87 Journal of Roman Studies 156–69.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

76

Egbert Koops

Scheidel, Walter. 2005a. “Human mobility in Italy II: the slave population.” 95 Journal of Roman Studies 64–79. Scheidel, Walter. 2005b. “Real slave prices and the relative cost of slave labor in the Greco-Roman world.” 35 Ancient Society 1–17. Scheidel, Walter. 2008. “Roman population size: the logic of the debate,” in Luuk de Ligt and Simon J. Northwood, eds, People, Land, and Politics: Demographic Developments and the Transformation of Roman Italy, 300 — 14. Leiden: Brill, 17–70. Scholl, Reinhold. 2001. “ ‘Freilassung unter Freunden’ im römischen Ägypten,” in Heinz Bellen and Heinz Heinen, eds, Fünfzig Jahre Forschungen zur antiken Sklaverei an der Mainzer Akademie, 1950–2000. Stuttgart: Steiner, 159–69. Sirks, A. J. Boudewijn. 1981. “Informal manumission and the lex Junia.” 28 Revue Internationale des Droits de l’Antiquité 247–76. Sirks, A. J. Boudewijn. 1983. “The lex Iunia and the effects of informal manumission and iteration.” 30 Revue Internationale des Droits de l’Antiquité 211–92. Straus, Jean A. 2004. L’achat et la vente des esclaves dans l’Egypte romaine: contribution papyrologique à l’étude de l’esclavage dans une province orientale de l’Empire romain. Munich: Saur. Temin, Peter. 2001. “A market economy in the early Roman Empire.” 91 Journal of Roman Studies 169–81. Temin, Peter. 2004. “The labor market of the early Roman Empire.” 34 Journal of Interdisciplinary History 513–38. Treggiari, Susan. 1975a. “Jobs in the household of Livia.” 43 Papers of the British School at Rome 48–77. Treggiari, Susan. 1975b. “Family life among the staff of the Volusii.” 105 Transactions of the American Philological Association 393–401. Verboven, Koenraad. 2012. “The freedman economy of Roman Italy,” in Sinclair Bell and Teresa Ramsby, eds, Free at Last! The Impact of Freed Slaves on the Roman Empire. London: Bloomsbury, 88–109. Wacke, Andreas. 2001. “Manumissio matrimonii causa. Die Freilassung zwecks Heirat nach den Ehegesetzen des Augustus,” in Heinz Bellen and Heinz Heinen, eds, Fünfzig Jahre Forschungen zur antiken Sklaverei an der Mainzer Akademie, 1950–2000. Stuttgart: Steiner, 133–57. Watson, Alan. 1987. Roman Slave Law. Baltimore: Johns Hopkins University Press. Weaver, Paul R. C. 1972. Familia Caesaris: A Social Study of the Emperor’s Freedmen and Slaves. Cambridge: Cambridge University Press. Weaver, Paul R. C. 1990. “Where have all the Junian Latins gone? Nomenclature and status in the early Empire.” 20 Chiron 275–305.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Practice of Manumission

77

Weaver, Paul R. C. 1997. “Children of Junian Latins,” in Beryl Rawson and Paul R. C. Weaver, eds, The Roman Family in Italy: Status, Sentiment, Space. Oxford University Press: Oxford, 55–72. Westermann, William L. 1955. The Slave Systems of Greek and Roman Antiquity. Philadelphia: American Philosophical Society. Wiedemann, Thomas E. J. 1985. “The regularity of manumission at Rome.” 35 The Classical Quarterly 162–75. Zelnick-Abramovitz, Rachel. 2005. Not Wholly Free: The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World (Mnemosyne suppl. 266). Leiden: Brill.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

15 Banking, Money-Lending, and Elite Financial Life in Rome Jean Andreau

In the Roman world the practice of making loans at interest was widespread and quite ancient. “In truth, the practice of making loans at interest was a perennial problem in Rome, and a frequent cause of sedition and strife” (Sane vetus urbi faenebre malum et seditionum discordiarumque creberrima causa), in the words of Tacitus (Tac. Ann. 6.16.1). He adds that the Law of the Twelve Tables, in order to remedy this evil, stipulated that interest should be limited to a twelfth part (unciarium fenus). What is meant by a “twelfth”? The twelfth part of the loaned capital, annually? It may rather have been 1/12 of this capital, but repaid every month, as Zehnacker (1980) has suggested, I think rightly. In any case allusion to the Twelve Tables attests that among the Latins, charging interest for loans had already become widespread in the mid-fifth century . Livy also mentions social and political crises in the fifth and fourth centuries  caused by heavy indebtedness affecting large numbers of people, especially the poor (Liv. 6.11–20; 7.16.1; 7.21.3; 7.27.3–4). At the time of the Twelve Tables, in the fifth century, the city of Rome had not yet begun to strike discoid coins. Some Greek cities in southern Italy or Sicily minted disc-shaped coins in silver, but it does not appear that these coins circulated freely in Roman territory. Bronze bars were used there, both as standards of value and as a medium of exchange. Some of these bars bore no markings, and they were known as aes rude (rough or raw bronze); others were scored with marks like the branches of trees, so that they are now called “ramo secco” (Thomsen Jean Andreau, Banking, Money-Lending, and Elite Financial Life in Rome In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0015

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

82

Jean Andreau

1978; Zehnacker 1990). Charging interest was possible, then, even in the absence of disc-shaped coins. Even at that period, the activity of lending was not limited to lending food, or payment in kind. There were other kinds of loans. Repeated crises brought things to the point where in 347  interest rates were limited to 1/24th part, half the amount of the previous rate. Then in 342, through the lex Genucia, charging interest was completely prohibited in the city of Rome (Liv. 7.42.1). This is the only time in the entire history of ancient Rome that such a prohibition was enacted. Probably, the edict was never formally repealed (on the situation at the beginning of the first century , see App., Bell. civ. 1.54). But was it ever applied? We just do not know. And if it was applied, for how long? Probably not for long. In fact between 318  and 310, deposit bankers appeared in the Forum (the argentarii). At the same time, Rome began to mint silver coins called RomanoCampanian didrachmas. A century later (late third century, early second century), at the time when Plautus and Cato the Elder lived, such a prohibition was completely out of the question, and so things would remain up to the time of the end of the Western Roman Empire, in the fifth century . No such prohibition would be considered in all that time. This does not mean that public authorities, in Italy and in the provinces, would not make attempts to regulate or even limit interest rates, as we shall see in what follows. During the last three centuries of the Republic, between the end of the fourth century and Augustus’ reign, the pace of economic and commercial life increased, driven by the conquests of Rome. Such developments cannot be quantified, but there is no serious doubt about the matter. The size of the estates of members of the elite of Rome became significantly larger (on average), and the lifestyle of this group was transformed. The Roman tradition may have greatly exaggerated the supposed austerity of the lives of the Roman senators of previous centuries, but there is no doubt that at the time of Cicero, members of Italian elites (senators, equestrians, municipal aristocrats) were in general richer and lived lives of luxury in comparison with the elites of a few centuries before. The practice of charging interest continued, and it created problems as before. Indebtedness probably increased as the size of estates increased. It is not possible to measure, though, the speed with which this took place, in Italy or in conquered territories. However, a fair number of documents show that charging interest was a common practice.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

83

With regard to the first century , it can be supposed that all wealthy or well-off men and women borrowed money or lent it, at one time or another. Many of these loans were secured, for instance by pignora. The same people both borrowed and lent, depending on their needs, at different times in their lives or even at the same time, depending on the case. Women of a certain social level did not specialize in lending money as a profession, but they could find themselves managing estates and engaging in either borrowing or lending. The documentation we have concerning the poor (in cities as well as in the countryside) is much more limited than what is available for rich people; but a fair number of testimonies and indicators show that the poor frequently got into debt, such that many of them were always in debt; they also show that some poor people practiced low-level money-lending themselves. Shatzman has compiled a list of twenty-five senators known to have engaged in money-lending (1975: 76), and Nicolet made a list of the names of seventeen equestrians (1966: 372–3). In view of the amount of documentation available and the nature of that documentation, this is a very significant sample. Further, at the end of the Republic, and under the early Principate, lending is mentioned in all the general texts that concern wealth and the sources of income of the rich. In addition to owning lands, livestock, personal residences, and residential investment properties, slaves and precious objects such as gold, a rich person of note (not necessarily a senator or equestrian) is likely to be a creditor (Petr., Sat. 37.8–9; Sen., Ad Lucil. 2.6 and 4.41.7; Tac., Hist. 1.20.3; Juv., Sat. 11.39–41; Tert., Cult. fem. 1.9.3; etc.). As for this last century of the Republic, there are also more references to the way such a source of income must be handled. The wealthy man is normally helped out by his family, by friends, or by those who depended on him, such as slaves or freedmen: he could borrow from them with more convenient conditions and receive information from them. But in any case, he must find out about the character of those to whom he lends, and their solvency. He must pay attention to repayment dates and other aspects of the loan agreement, and know when it is time to sue a bad debtor, although Saller (1982: 121) insists that creditors exhausted all other possible means of settling debts before bringing their debtors into courts of law. In Rome, creditors posted complaints about deadbeats and their bad faith upon the columna Maenia; Cicero, in the Pro Cluentio, speaks of one Quintus Manilius who lost his

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

84

Jean Andreau

reputation in this manner (Cic., Pro Cluent. 13.38–9; see Andreau 1987b: 163–4). Obviously sums great and small were lent out, and financial activity of this kind was not of the same importance for everyone. Loans did not all carry the same rate of interest or the same conditions. Even among members of the same social group, for example senators or equestrians, there were differences with regard to loans made at interest. Many senators needed to borrow at various times in their political careers, or for reasons pertaining to their social lives, but such situations did not mean they might not still lend money, for example to credit intermediaries. One year, a senator might need to borrow a great deal of money for a political campaign; another year, he might be obliged to splurge on his daughter’s wedding. But at other times, they might profit from certain very lucrative political functions or from inheritances. Not all the members of the elite were trained financiers, although some of them might handle very large sums. Cicero himself is an excellent example. He borrowed often, and in large amounts, but he also engaged in money-lending. One of his letters shows that he and Atticus both did, Cicero through an intermediary, Cluvius, and Atticus with the assistance of Vestorius (Cic., Ad Att. 6.2.3; see Früchtl 1912; Shatzman 1975: 416–22). According to Rauh, the debts Cicero held in 45  amounted to an enormous sum, possibly several million sesterces (Rauh 1989: 60–9). But Cicero was not a trained financier. It was fashionable to rail against money-lending, but well-known Greeks and Latins do not appear to have tried very hard to hide the fact that they were money-lending, except when there was something illegal about it (for example if the interest was high enough to constitute usury). If charging interest, which was not illegal, had really been the subject of a moral and social taboo among members of the elite, the number of general mentions of it, and the number of loans we hear about, would not be so large. Charging interest was criticized from a moral standpoint, but it was also accepted by Roman public opinion. Who were Cluvius and Vestorius, two men who are mentioned several times in Cicero’s letters? What activity were they engaged in? They lived and worked in Puteoli (Pozzuoli). They had a relationship with the businesspeople of Puteoli, but what exactly was it? Cluvius owned many things, including real estate (commercial buildings and gardens). Vestorius must have had interests in manufacturing,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

85

because in a passage in Vitruvius and another in Pliny the Elder, he is said to have been the first in Pozzuoli to produce the dye or pigment Egyptian blue, and to have created a new variety of that blue (Vitr. 7.11.1; Plin., N.H. 33.162). From the way Cicero talks about him, he knew a great deal about accounting and financial practices. At that time, in Italy, there were professional bankers (argentarii), as well as a variety of types of financiers and money-lenders, at various levels of the social hierarchy. Were Cluvius and Vestorius, who appear not to have belonged to the equestrian order, really bankers, or something different, financiers of another sort? On the basis of the available information, I am convinced that they were not professional bankers. But this cannot be established with complete certainty; if someone wishes to think of them as bankers, I cannot prove this wrong. Both interpretations can be defended (D’Arms 1981: 49–55; Andreau 1997: 99–118). Aristocrats who borrowed and lent money at interest, according to the requirements of the moment, were concerned above all with their estates, the dowries of their wives, and perhaps with the estates of their close friends. Thus, they were obliged to know how to maintain buildings and how to manage their lands and their money. Some of them, all while leading political and social lives in accordance with their rank, showed themselves to be particularly greedy, even rapacious. Such was the case for the senator Marcus Junius Brutus, with whom Cicero had dealings in his capacity as governor of the province of Cilicia and Cyprus, in the year 51–50 . The city of Salamis in Cyprus, in the year 56 , borrowed a large sum of money (the exact amount is not known) from two Italians based in Cilicia, Marcus Scaptius and Publius Matinius. When Cicero arrived in Cilicia in 51 , the debt had not been repaid. The people of Salamis thought that the two businessmen were the actual creditors, and at first Cicero also believed this; but later he learned that the capital had come from M. Junius Brutus, and the risk in the lending operation was his problem. Brutus had taken advantage of the official mission that had taken him to Cyprus in 58, in the company of the senator who would later be known as Cato of Utica, to make the deal to loan the money. His two intermediaries and he demanded a usurious interest rate; Brutus contacted a succession of provincial governors in order to put pressure on the citizens of Salamis (Cic., Ad Att. 5.21.10–13; 6.1.3–8; 6.2.7–9; 6.3.5–6; see Andreau 1997: 109). Brutus was not a

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

86

Jean Andreau

professional financier, but he was obviously more avid for profit than Cicero. Titus Pomponius Atticus, Cicero’s friend and brother-in-law, was better acquainted with the business of finance than either Cicero or M. Junius Brutus; still, he could not be counted among the greediest members of the elite. There were varying degrees separating the most disinterested members of the elite and those who were the least knowledgeable about finance on one hand, and those who were the greediest or the most experienced on the other. Among the greedy, we find professional money-lenders (the faeneratores this word means lender at interest), a certain number of equestrians who controlled the companies of publicans—companies specializing in tax collection, among other things, and a number of Italians who had set up their businesses in different provinces. Even if the imperial regime changed the political context, and even if the importance and influence of companies of tax collectors gradually decreased, such financial activities undoubtedly continued under the Principate. But with regard to the Principate, we do not have discourses such as those of Cicero, nor any correspondence comparable to his. The available evidence is less weighty.

15.1. SOCIAL PRACTICES OF GRATUITY AND GENEROSITY Aristocrats of the first century  also made loans without charging interest or at very low rates, by reason of friendship, or in order to help their relatives or clients; they might themselves stand surety for their relatives, friends, or clients, but they also made outright gifts. These financial operations having to do with some degree of generosity have been studied in depth and in a very subtle way, in recent decades, by Ioannatou (2006: 229–307) and Verboven (1993, 2002). Members of the elite were pleased to express admiration for disinterested actions and generous gestures, and they prided themselves on acting this way themselves, as we see in treatises by Cicero (De officiis) and Seneca (De beneficiis). In truth, in most cases the practice of giving gifts or no-interest (or low-interest) loans, even if it was part of the ideology of the elites, was not completely disinterested, because

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

87

the lenders in these cases were most often hoping to motivate symbolic or political results that would benefit them. As Ioannatou wrote (2006: 244), in such actions “the social aspect was far more important than the financial aspect.” The gifts and loans made without interest (or at low interest rates) were part of social habits, and were related to a social hierarchy (even within the aristocratic milieu). They were often involved in exchanges between family members or political allies, involving a series of gifts and counter-gifts (see Verboven 2002; Ioannatou 2006). Three remarks must be made with regard to the interest-free loans. First, it appears that they were rarely completely obligatory. Most of the time, the father of an aristocratic family had some latitude in choosing, even if, after the fact, he presented his choice as an absolute duty which could not be avoided. Secondly, such gestures are primarily known to us as occurring in the first century , because of the existence of Cicero’s works. But they were already made before the first century , and continued to exist under the Principate. Thirdly, some historians might be tempted to insist on the reality of these gifts from generosity, as opposed to financial practices imbued with greed and self-interest. Others might give more credence to the inverse tendency, and emphasize the predominance of greed. But neither of these positions is satisfying, because the desire for gain and riches existed alongside altruistic social tendencies. These two ways of doing things were always present in the city of Rome and in the Roman Empire, although beginning with Augustus, the emperor often tried to limit the effects of the most extreme examples. As for the law, loans at interest and loans without interest gave rise to contracts that were only partly identical. A stipulation could be made in either case, and the mutuum as well (even if in principle, the mutuum was gratuitous), because it could be accompanied by a stipulation of interest. But certain contracts, whose gratuitous status was better respected, were better adapted to interest-free loans. This is the case with the mandatum and the commodatum. The latter was a gratuitous contract that designated movable and immovable goods to be loaned for use, and which belonged to the category of good offices and services rendered as between friends or neighbors (Michel 1962: 95–127 and 168–97). One may say with Ioannatou (2006: 230–1) that interest-free loans “fitted the legal pattern of the mutuum and the commodatum,” without forgetting that the mutuum was also used in the case of loans for which interest was charged.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

88

Jean Andreau 15.2. THE BEGINNINGS OF BANKING IN ROME AND ITS DEVELOPMENT (FOURTH TO FIRST CENTURIES )

Let’s return to the matter of loans at interest and financial affairs. Between the end of the fourth century  and the beginning of the common era, not only did the volume of financial operations in Rome and Italy certainly increase by a very large amount, (although it is difficult to prove this), but the nature of these operations became more complex, as the world of business became more diversified. On one hand, deposit banks, which had existed in the Greek world since the second half of the fifth century, began to operate in Rome at the end of the fourth century, and were soon found in other cities in Italy and the western Mediterranean. In the last decades of the Roman Republic, we know that several deposit bankers, called argentarii, operated not only in Rome itself, but in western Mediterranean cities such as Pompeii, Regium, Lepcis Magna, and Syracuse (Andreau 1987a: 313–29). Later, under the Principate, inscriptions attest the presence of some banks in a larger number of cities, but, as regards the activity of these professions, the situation that can be observed in the Ciceronian period remains more or less the same until the second part of the third century . When they appeared at Rome, these bankers undoubtedly verified the value of coins and operated as money-changers, but without more documentation it is impossible to know what other services they might have provided for their clients at that time. A century later, during the time of Plautus, we can see that they received deposits, made loans, and provided routine banking services for their clients (cashing drafts, making payments to and on behalf of their clients, allowing withdrawals, etc.). Further, during the second half of the second century , bankers in Rome and Italy began to provide a service that Greek bankers had not provided: they furnished credit for purchases made at auction. Auctions were held in Greece as well as in Rome, but in the Greek world bankers did not play the same role in auctions as in Rome (see Thielmann 1961; Andreau 1974; Andreau 1987: 583–97; Petrucci 1991, 2002; Garcia Morcillo 2005, 2008, 2014). In the Roman case, at an auction, or not long afterward (a few days), the banker, the argentarius, would pay the seller the auction price for objects, slaves,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

89

or supplies purchased at auction, loaning the money to the buyer, who would reimburse the banker several months later (not more than a year, in the examples we know) and pay interest. Several texts from Cicero attest the existence of this auction credit during the times when he lived (see for example Cic., Pro Caec. 6.16–17 and 10.27). And for the Julio-Claudian period, 153 tablets were found at Pompeii, tablets that had belonged to the argentarius or to the coactor argentarius Lucius Caecilius Jucundus, of which 137 were receipts from auctions, given by the seller to the banker, attesting that he had received from the banker the price of sale. The other sixteen tablets had to do with business transactions (rentals, tax collection) the banker had concluded, during the period  50–60, with the city of Pompeii (Andreau 1974; Jongman 1988). In the De agricultura of Cato, mention is made of a coactor, that is, a private person who handles sums of money on behalf of his clients, collecting payments and debts, without however acting as a banker (Cato, De agr. 150.2). The father of Horace was a coactor in Venosa (Hor., Sat. 1.6.86), and we know of a few other coactores, at the end of the Republic and under the Principate. Finally, two other occupations appear at the end of the second century , or during the first century : the nummularii used to verify the value of coins and change money, but they were not deposit bankers (the nummularii would not become deposit bankers until the first half of the second century ). And there were also coactores argentarii, who were at one and the same time deposit bankers and coactores. Professional bankers, argentarii and coactores argentarii, all loaned money, as is shown by the documentation on credit provided for auctions, but these people were never confused with money-lenders who charged interest yet were not bankers (Barlow 1978: 67–80; Andreau 1999: 9–49). Plautus borrows his characters who are professional bankers from Greek models, and he calls them tarpezitai or trapezitai, after their Greek names, or just argentarii, the Latin term. He never speaks as if they were the same as money-lenders who were not bankers, a status he indicates by using the Greek word daneistai or by the Latin expression faeneratores (on banking in Plautus and Terence, see Andreau 1968). Plautus calls his comedies palliatae, and in them he shows himself to be the inheritor of Greek comic traditions; he in fact claims this. His plays take place in Greek cities and feature characters with Greek names. But he also describes and satirizes Roman mores. What he

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

90

Jean Andreau

says about bankers and credit is thus part of the Greek reality of his times, but is also part of Roman reality—although he distorts all these situations in accordance with the comic effect of his language. His plays are valuable sources of information about Roman life. The distinction he established between professional bankers (argentarii) and money-lenders (faeneratores) is part of the information we can glean from his writing, and what he says in this regard is largely confirmed by other sources. That which constituted the essence of the idea of a bank (argentaria) for the ancients and particularly for legal experts (jurisconsults) had two aspects. On one hand, this is what constitutes in our eyes as well the specific function of a deposit bank—the double service of receiving deposits and extending credit. The banker, acting as such, did not loan out his own money, but rather a part of the money that had been deposited with him by his clients (Plaut., Curc. 71–9). On another hand, this was the connection between the banker and the client. Clients deposited money, and they could leave it on deposit or withdraw it, or ask the banker to make payments or send money. This connection is thus exhibited in a series of operations, and by the records of transactions that were kept. The result of such operations involved deposit accounts held by clients, and such accounts were called rationes. Transfers could be made between clients of the same bank, as is attested by two passages, one in Plautus and the other in Terence (Plaut., Asin. 436–40 and Ter., Phorm. 921–3). There was no system of institutionalized compensation between the banks of the same city, but that did not prevent transfers from taking place, as we may see in a number of papyri. To facilitate such transfers, the banker sometimes held an account in the bank of one or several of his colleagues (Bogaert 1983a: 218–21; 1983b: 34). Between banks in different cities, such an arrangement, although not impossible, was very probably rarer, and must have depended on the personal relations of the individual bankers concerned (on such categories of financial operations, see Harris 2006, 2008; Hollander 2007). Professional bankers, unlike elite financiers, were local businessmen, who specialized in operations on the spot. A cheque is a written note by which a client of a bank orders it to pay a precise sum of money to the beneficiary of the cheque. But the cheque is given to the beneficiary, whereas other payment orders are given to the banker. In antiquity, there is no trace of cheques in the

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

91

western Mediterranean. However, in the eastern Mediterranean, the use of cheques is attested in at least two regions: in Canaan, following rulings made by Mosaic law on the payment of wages (Bogaert 1968: 340), and in Egypt, where cheques have been found from the end of the Hellenistic period. Such cheques continued to be used in Egypt after the Roman conquest (Bogaert 1983a, 1987: 73–7). Yet, the existence of these non-transmissible cheques was not as important historically as that of the transmisible cheques which came into use later, in modern Europe. At the end of the Republic and during the classical period of Roman law, jurisconsults had a point of view regarding deposit banks (argentaria) that was at one and the same time very rigid and very flexible. It was rigid with regard to the definition at its center: they only thought of these banks in terms of deposit accounts, and nothing that was external to accounts had anything, strictly speaking, to do with banks. It was very flexible as concerned the details of the operations that the bank account permitted: a bank account is nothing but this series of operations. For the jurisconsult and for the client, what the banker was supposed to do was to lend out part of the money deposited. But it is also assumed that the banker will make payments in the name of the depositor, and allow him to make withdrawals at will (D. 2.13.9.2, Paul. 3 ad ed.). Before examining more closely the orientation of Roman law in relation to banking and financial activity, I would like to make a few remarks about the norms applied to non-Romans, that is, to Latins and to foreigners (who were called socii, or “Allies” in Italy during the republican period, and peregrini in the provinces).

15.3. THE NON-ROMANS AND FINANCIAL LIFE Roman citizens were subject to the norms of Roman civil law. But in territories controlled by Rome, there were people who were not Roman citizens. There were non-citizens in Italy prior to the end of the Social War (91–89 ), and there were also non-citizens in the provinces, at least up to the time of the famous Constitution of Caracalla or the Antonine Constitution (Constitutio Antoniniana) of  212. What consequences did such a situation have on banking

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

92

Jean Andreau

and credit? An episode from the history of the Roman Republic allows us to understand a little better the elements of the problem. It concerns the “crisis” of 193 , concerning which we have only a short passage from Livy (Liv. 35.7). Some have called this a banking crisis, but it would be more precise to speak of a financial crisis, since nothing indicates that the professional bankers had anything to do with it. That year, many Romans were deep in debt, even though the greed of financiers (Livy relates) had been limited by a number of laws governing interest-bearing loans—laws whose substance we do not know. How was this possible? We can see that some Roman citizens, in order to avoid the effects of that legislation, transferred the debts they owned to Italian Allies, who were not subject to the same norms because they were not Roman citizens. Livy says the transfer was fraudulent, because the money came from Romans who transferred it to Allies and then allowed the Allies to lend it to Romans (Barlow 1978: 78–80). How did the Roman creditors manage this? We just do not know. The first attempt to straighten this out involved requiring Allies to declare all the loans they had made to Roman citizens, going back to the most recent Feralia. Then the extent of the practice was discovered, Livy says. So a second measure was tried: a plebiscite proposed by the tribune M. Sempronius Tuditanus required the Latins and Allies to conform to the same norms as Romans with regard to loaning money (pecuniae creditae ius). We do not know the amount of the debts declared, and this is a pity, because this is the only time we see the Roman state inquire as to the amounts of debts contracted by Romans, or in any case a part of these debts. The episode shows the consequences to which a multiplicity of statuses of persons before the public law can give rise, and how the state managed to palliate these consequences by making all persons subject to the same legal norms. Roman jurisprudence (as for example in the manual of Gaius) distinguishes between ius civile, only applicable to Roman citizens, and ius gentium, which was applicable to Roman citizens as well as to other free inhabitants of the Empire. But Livy, in this passage, does not speak of ius gentium. As Schulz has emphasized, when legal advisors or legislators wanted to apply a norm already in force for Romans to a non-Roman, they did not seek to create a system that would apply equally to both and that would be a part of the ius gentium. They simply applied the civil law to the non-Roman (Schulz 1946: 73, 137, and 163). This is why the norms pertaining to the civil

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

93

law are very often present in contexts that involve foreigners. The ius gentium only had practical significance in international law. After the beginning of the second century , the number of citizens slowly grew. After the Social War (91–89 ), all the Italians of the Peninsula became Roman citizens, and later the residents of northern Italy were granted the same status. But at the beginning of Augustus’ reign, the Italians residing in provinces and provincial citizens who had acquired Roman citizenship represented only a tiny minority of the population of those provinces. In the second century , the situation had obviously changed. At the beginning of the third century , not long before the Constitution of Caracalla, how many foreigners were left? For some historians a very large majority of the free inhabitants of the Empire had already received this citizenship. For others, Roman citizens continued to be a minority in the eastern part of the Empire, and in North Africa, communities of foreigners were still numerous (Jacques-Scheid 1990: 281–7). Otherwise, following the establishment of the Constitution, legal norms of Roman law were not applied everywhere, nor were they applied to all populations. It is important to know in which cases judgment remained in the hands of Roman magistrates (meaning by this primarily the governors of provinces), and which cases were handled by other magistrates (usually municipal magistrates). But it is just as important to know the norms to which foreigners were subject. The first question is a little less difficult than the second, but both are delicate because the documentation for them is extremely limited. Geographically, the evidence for financial and banking-related matters is not divided up in the same way as the documentation for law and the legal system. As regards the late Republic, the cases of Sicily and of Cilicia are the least obscure, simply because of the speeches of Cicero against Verres and of his letters. In imperial times, we have from Egypt (via the papyri) a copious enough amount of documentation for legal matters and for the professional part of financial life. Rather than making general remarks, I prefer to base myself on two examples, about which we are somewhat less ignorant: Sicily at the time of Cicero, and Egypt during the Principate. In the Sicily of the time of Gaius Verres, a single category of litigation was judged by Sicilian judges in accordance with Sicilian norms: litigation opposing two peregrini from the same city (Maganzani 2007). Conversely, if the two parties engaged in litigation were Roman

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

94

Jean Andreau

citizens, the procedure followed and the norms applied would obviously be based on Roman law. When the case involved two peregrini from different cities, it was necessary to appear at a session of the conventus, where the governor of the province would be acting in a judicial capacity. Banking and money-lending in Roman Egypt have been very well studied by Bogaert (1994) and Lerouxel (2008, 2012, 2016). As regards the norms applied to non-Romans in Egypt, one has to refer to the research of Mélèze-Modrzejewski (1970, 1990) and to a recent article by Yiftach-Firanko (2009). For instance, Mélèze studied the mandatum cases described in Egyptian papyri, some of which had a financial aspect. They were gratuitous mandates of the kind known under Roman law, but had to do with Greco-Egyptian customs, and not with Roman law (Mélèze-Modrzejewski 1990: II, esp. 473). In the second century , we find allusions to a “Law of the Egyptians” that was the result of a fusion between Greek elements and properly Egyptian ones. This was a sort of manual that aimed at explaining local legal customs to Roman judges, and the norms contained in it applied to all the peregrini of Egypt, whatever their origin. Conclusion: in a number of cases, when the opposing parties were not Roman, the applied norms were more or less divergent with regard to Roman law, and we know far less about them than we know about Roman law. But we should not exaggerate the size of the unknown area of the law for foreigners, because the influence of Roman law was probably always very strong.

15.4. LAW AND J URISPRUDENCE AND THE WORLD OF FINANCE Even if Roman law seems systematic and specializes in definitions and classifications, its role can be undeniably characterized as flexible. The fact that, in the Republican period, corporate and banking law came mainly under the ius honorarium, that is under the praetor’s edict, and was not mainly a matter of statute, is a symptom and an expression of this flexibility. For, in his edict, the praetor, at least in theory, might take new social and economic situations into account, whereas it would have been more difficult to change statutory law so frequently.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

95

Under the Principate, corporate law was not mainly shaped by imperial constitutions (which were equivalent to Republican leges publicae). As the praetor’s edict was in some way fossilized, in practice, and then officially (edictum perpetuum), it did not ensure the same flexibility. But this evolution of the praetor’s edict was balanced by the leading role played by jurisconsults, to whom, from Augustus’ reign onwards, the emperor granted the ius respondendi. In the second and third centuries , the major part of the available texts about banking and financial life were written by jurisconsults. It needs to be stressed how precise and concrete these fragments, included in the Digest, were. But there were a number of jurisconsults, and they did not always agree on any particular question, so some of these questions were subjects of debate. It seems that the jurists’ opinion bound the judges only when a consensus had been reached (Magdelain 1990: 103–52). Compared to the literary evidence for financial life and banking from the Ciceronian period, the evidence from the Principate is poor. We know a fair number of funerary inscriptions concerning bankers or money-changers, but the literary texts concerning banking and financial life are rather few during the Principate. For this reason, the historian who is not a specialist in law may gain the impression that there was much less innovation in the Principate. But this certainly is not correct as concerns law and juridical thought. In financial matters, a first token of the flexibility of Roman law is that jurisconsults paid a great deal of attention to the concrete application of technical procedures. The authors of the fragments collected in the Digest, as well as Gaius in his Institutes, were much concerned with the material conditions of the various occupations and activities, and they based their legal opinions thereon. The jurists often expressed their opinions on one occupation, or one form of activity, after another. So, in writing about the actio institoria, Ulpian, not hesitating to go into detail, enumerates an impressive list of functions that a slave might perform: there are managers of buildings and monuments, bankers, money-lenders, weavers and vilici, and also baker’s apprentices, spice vendors, clothing merchants, travelling salesmen, and wholesalers in oil or grain (D. 14.3.5, Ulp. 28 ad ed.). The definition of bank-account which Labeo produced and which is handed down by Ulpian is very descriptive: it is a list of counter transactions rather than a general statement about the whole banking

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

96

Jean Andreau

practice: “Labeo says that the bank-account is composed of mutual transactions which consist in paying out, collecting payments, lending, obliging, and paying off one’s debts” (D. 2.13.6.3, Ulp. 4 ad ed.). The jurists attached much significance to the bank-account, which they called either ratio (“an account”), or ratio accepti et expensi (“an account of deposits and payments”), or ratio implicita propter accepta et data, “a complex account including both deposits and payments” (D. 2.14.47.1, Scaev. 1 dig.). The fact that this notion of bank-account could be called in different ways suggests that it arose out of professional practice and was not a creation of the jurists themselves. The jurists were concerned to keep in touch with the development of business transactions and professional life, in order to handle the juridical problems caused by financial activity as efficiently as possible. But at the same time, as Thomas remarked, the art of the law provided for human activities “a means for their formal elaboration” (Thomas 2004: 212). Thus we are concerned first of all with concrete and precise observations on the conditions of an activity; and then based on these observations, we have the beginnings of conceptualization of a kind we could characterize as economic. In the financial field, this beginning of conceptualization led the jurists to insist on the specificity of banking. One might have thought that all financial activity would be handled according to the modalities of private law applicable to every citizen; but, on the contrary, one category was isolated, that of people who had the right to open deposit accounts and who did business on the basis of such deposit accounts (D. 2.14.47.1, Scaev. 1 dig.). Only the deposit banker had the right to open an account for each of his clients. His transactions were ordered around the notion of the bank-account, which created a relationship, considered as long-lasting, between his clients and himself. Besides this empirical description and its formal elaboration, the jurists found another way of handling such problems. It consisted in appealing to juridical instruments that already existed. For instance, they resorted several times to stipulation. This formal contract was very rigid, for it was made up of a question and an answer which had to be completely identical. But, despite that rigidity, it was very flexible at the same time, since the stipulation could be applied to nearly every requirement (Johnston 1999: 77–8). So, in connection with a loan contract that was in principle interest-free, such as a mutuum, a stipulation could be added to allow the payment of

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

97

interest (Johnston 1999: 84–5). The stipulations was also used to manage the credit that the bankers granted the buyers in the context of auctions (stipulatio argentaria). Another route taken by ius honorarium and jurisprudence was the creation of new instruments, for instance new actions. Conspicuous here are the five actiones adiecticiae qualitatis and the actio tributoria, which were instituted by the praetor in the late Republican period (see for instance Aubert 1994: 40–116). Their remit consisted in protecting third parties who had professional relations with slaves or with dependent sons (filiifamilias). Through these actions, third parties were allowed to take legal action against the slave’s owner or the son’s father. For our present purposes the most significant is the actio institoria. It concerned a slave or a filiusfamilias who ran a nonmaritime business, that is a shop, a workshop, a storehouse, or a bank, on behalf of his owner or his father. The actio institoria was instrumental in enabling slave owners to take advantage of their slaves’ commercial and financial activity. These actiones adiecticiae qualitatis were one of the most striking and original features of the Roman slave system. Of course, and without a doubt, this system had all the defects and flaws of slavery and slave systems, but at least it allowed a modest number of slaves and freedmen to become technicians and small or medium-sized managers. Finally, with regard to other aspects of banking and credit, debates of long duration seem to have taken place among the jurists. The best example of such debates concerns the so-called “irregular deposit,” which nevertheless is at the center of banking activity. In the juristic texts on this contract one encounters no consensus, and possible juristic interpolations of the Justinianic period does not provide sufficient explanation. In Roman law, under the ordinary contract of deposit, the depositee was not entitled to use the deposited object (Andreau 1987a: 529–44). Even if this object was an amount of money, it had to be kept as it was, until it was given back to the depositor, and it could not bear interest. But, of course, such a deposit, which was retained without being touched by the depositee, is not the defining characteristic of banking activity. The banker generally wished to be entitled to make use of the money and to return not the precise coins deposited, but the equivalent value. In practice, the argentarii, the coactores argentarii, and, later on, the nummularii received deposits of money that they were entitled to use. But, at the end of the

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

98

Jean Andreau

Republican period and under the Principate, did the jurists acknowledge such a kind of deposits (which, later on, was called an “irregular deposit”)? Or did they consider such transactions as loans received by the banker from the customer? Three elements were singled out for attention by the jurists: did the depositee use the money he had received from the client? Did he lend it at interest? And did he pay the depositors interest? The most probable hypothesis is that, in the second and third centuries , the interest-bearing deposits, i.e. the deposits for investment, were considered as loans, whereas the others, the deposits which did not provide interest to the depositors, were considered as true deposits. These observations point up some of the ways in which Roman law succeeded in coping with the problems caused by financial life and credit: it paid much attention to the practical details of financial life; it called up juridical instruments that already existed; it created new tools, and particularly new actions; and, moreover, long debates took place among the jurisconsults, especially about the depositum which, later on, after the classical period, was called the “irregular deposit.”

15.5. PRODUCTIVE LOANS AND COMMERCIAL CREDIT In Greece and in Rome, did productive loans exist, that is loans which played a part in production and trade? If such productive loans existed, to what extent did professional bankers and businessmen who used to lend money play a role in such loans? And what can be said of the role of law and jurist in connection with them? On the first of these three questions, the evidence is very scarce, but the answer is positive, beyond all reasonable doubt. Seneca, for instance, wrote in a letter to Lucilius (Sen., Ad Lucil. 119.1): “yet, you will need a creditor; if you wish to be in business, you have to contract debts” (opus erit tamen tibi creditore; ut negotiari possis, aes alienum facias oportet). Let us note that, four centuries before, Demosthenes had written a very similar sentence on the role of loans in trade (Dem. 34.51). Plutarch, when he composed a treatise condemning debts, treated in the same manner those who borrowed

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome

99

to buy estates or productive slaves and those who borrowed because they had expensive tastes or for purposes of evergetism (Plut., Mor. 830 E). In another passage, he speaks of a loan contracted in order to buy grain, vineyards, or olive groves (Plut., Mor. 523 F). Moreover, the scraps of information provided by jurists are consistent with those derived from literary texts: several fragments of the Digest, written by the jurists Africanus and Ulpian, deal with cases in which the loan is contracted in order to buy an estate, repair a ship, feed sailors, or buy goods (D. 12.1.4 pr., Ulp. 34 ad Sab.; D. 14.1.1.8–11, Ulp. 28 ad ed.; D. 14.1.7, Afric. 8 quaest.). Such texts show that the notion of productive loan was absent in antiquity. Greeks and Romans were not accustomed to distinguish between “productive” loans and loans for consumption. However, in practice, this conceptual gap did not prevent such productive loans from being contracted. They were interest-bearing loans. From the juridical point of view, they were treated in the same way as the other kinds of loans, except the maritime loans. What proportion of all loans were productive? We do not know. As regards Athens in the fifth and the fourth century , both Millett (1991: 188) and Cohen (1992: 151), who very often disagree, are as one in holding that productive loans constituted a rather substantial minority. For the Roman world of the end of the republican period and of the Principate, such a guess would also be appropriate. To what extent did professional bankers and the rest of the financiers play a role in these loans the object of which was to invest in economic activities? This second question leads us inevitably to a consideration of the three Sulpicii. The Sulpicii were three firstcentury  businessmen, of whom we know through the tablets found in the area called Murecine (or Moragine), in the territory of Pompeii. The tablets concern transactions concluded in Puteoli under the Julio-Claudians. They have been published and studied by Camodeca, who has written two outstanding books and many papers about them (Camodeca 1992, 1999, 2000, and 2003). What was the business of the three Caii Sulpicii, Faustus, Cinnamus, and Onirus? Two different hypotheses have been proposed: they were either professional bankers, or money-lenders without being bankers. Relying mainly on the tablet TPSulp. 82 (mainly, but not exclusively), Camodeca concluded that they were professional bankers (argentarii, cf. Jones 2006). On the contrary, Verboven (2003a, 2003b) disagrees,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

100

Jean Andreau

holding that they were not bankers, but faeneratores, that is, specialized financiers and money-lenders. I share Verboven’s opinion (Andreau 1999: 71–9). At the moment none of the two assumptions may be considered as surely true. Every category of businessmen could be brought to lend money for a productive purpose, and so to have an influence on economic life, but not all of them in the same manner, nor to the same extent. Howgego (1992) has listed all the categories of persons and institutions that could lend money in Roman antiquity. In brief, let us say that short-term commercial loans were made by several categories of people, and particularly by professional bankers and wholesale merchants, who lent money at the same time as they sold goods. As for long-term credit, which was certainly less frequent, it was mainly practiced by financiers belonging to the elite and by big businessmen. I have already said that, in auctions, professional bankers of the Latin world used to advance short-term loans to the buyers. These auctions often took place in places of commerce such as ports, markets for retailing or for wholesale trade. For intance, in Rome, there were auctions in the Forum Boarium and the Macellum Magnum. Outside Rome, receivers and deposit-bankers are known in towns where fairs took place, such as Cremona, or in towns with periodic markets (nundinae). These nundinae constituted regional networks. I have shown elsewhere that in Pompeii and in Puteoli the auctions took place on the day of the nundinae (Andreau 1974, 1999: 149–50, 2000; De Ligt-De Neeve 1988; De Ligt 1993). Besides serving as country markets for peasants and small landowners of the surrounding area, the nundinae had two other functions. They were one of the places where patrimonial transactions could take place, where the landowners, even the important ones, could buy and sell land, buildings, cattle, slaves, and where they could sell some of the products of their estates. At the same time, merchants were present at the nundinae, and they played a significant role in the marketing of agricultural products, which were conveyed to the coast and then to other ports and cities, for instance to Rome. When the landowners and the farmers sold agricultural products by auction, the buyers included merchants and tradesmen. A letter of Pliny provides an example of such a situation: Pliny sells his grape-harvest by auction, and the buyers are negotiatores (Cat., De agr. 146–8; Plin., Epist. 8.2).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome 101 In other cases, argentarii and coactores argentarii were involved in auctions which were more to do with craftwork or mining; however, in such cases too the buyers might be merchants. For instance, in Southern Portugal, at Aljustrel, a Roman mining site, the bronze tables of Vipasca show that mining pits could be sold by auction, besides slaves, mules, donkeys, horses, i.e. men and animals playing a role in mining, and that argentarii customarily took part in such auctions (CIL II.5181; Domergue 1983). Whatever views one holds of the activities of the Sulpicii, the professional bankers provided a short-term commercial credit, and they were not the only ones to do it. For instance, in a fragment written by Paul, a slave entrusted by his master with money-lending at interest was active in commercial credit, on his own initiative, paying barley traders on behalf of buyers (D. 14.5.8, Paul. 1 decret.). I am convinced that neither Cicero’s Cluvius nor the Sulpicii were professional bankers; but, if they were bankers, these Murecine tablets would show that, in a port such as Puteoli, the professional bankers, besides short-term credit, also made long-term loans, and that, in that way, they created more currency than I previously believed to be the case (though, in fact, the volume of their transactions was not sufficient to create a lot of currency). If the Sulpicii were bankers, the Murecine tablets would show professional bankers of greater significance than any others known from the first and the second centuries . But, in any case, the outline of professional law which I will present in what follows would be the same. And moreover, in the available evidence—including the Murecine tablets—no argentarius concludes a maritime loan. They may play a role as intermediaries in maritime loans, but do not seem to have concluded such loans themselves. In the Murecine archive, two tablets, TPSulp. 31 and TPSulp. 78, have first been considered as regarding maritime loans; but Camodeca has shown that it is not the case (Camodeca 1999: 97–9 and 177–80). Admittedly, the available evidence is scanty. But in my view maritime loans were mainly transacted by big businessmen, faeneratores or merchants who were financiers at the same time. Among the known examples of production loans (that is, of loans contributing to production, transportation, or distribution of goods), medium-term and long-term loans are very rare. This may seem surprising, but such surprise is probably misplaced, for the following three reasons. First, merchants and manufacturers could have relied on short-term loans that were repetitively refinanced. Then, even in

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

102

Jean Andreau

eighteenth-century England, such long-term loans were not as frequent or as massive as it was previously thought. Verley (1985: 48–51) has offered a compelling picture of the ways in which capital was raised for commercial ventures in pre-industrial economies. Commercial ventures often required only modest levels of start-up capital. But when they were successful, they could grow rapidly through the massive reinvestment of profits. Thus pre-industrial commercial ventures tended to rely on internal over external funding, and a low demand for fixed as opposed to liquid capital. Such conclusions are a fortiori valid as regards Roman antiquity. Self-financing was also present in antiquity. In addition, Rome had institutions that offered financing choices other than long-term loans. The first of the institutions is what we would call a limited partnership: one of the associates (socii) of this societas provided the capital to one or a few associates who were in charge of the actual work and asset management. This was well suited to the needs of members of the elite interested in increasing their patrimony: they had the opportunity to extract profits from commercial, industrial, or even financial enterprises without having to live the life of an entrepreneur. Crassus participated in such societates, as in all probability did Vespasian (Cic., Par. Stoic. 6.46; Suet., Vesp. 4.6). Some municipal notables were involved in such limited partnerships as well. Such was the case with P. Alfenus Pollio, identified as such in Jucundus’ tablet no. 45, found at Pompeii (CIL IV, Suppl. 1.3340.45). The second institution is the peculium. To get his slave to work with third parties, the master had available two juridical procedures: the preposition (praepositio) and the peculium (Andreau 2004). It is important to make a clear distinction between them. In the preposition, the master maintains his role as the entrepreneur and the slave is a “manager,” to use a term proposed by Aubert. This is not the case with the peculium. Aubert (1994: 4) wrote: “even though a slave with peculium was legally dependent, his economic activities were in practice kept separate from his master’s. . . . A slave with peculium was not acting as business manager on behalf of a principal.” He was an entrepreneur. To summarize: Roman banking and finance did not function, first and foremost, to facilitate economic investment. Nevertheless, their economic role should not be entirely dismissed.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome 103 15.6. STANDARDS IMPOSED ON PROFESSIONAL BANKERS Did commercial law exist in the Roman world as a distinct branch of law? Did Roman jurisprudence deal with commercial transactions in a specific way, in accordance with a set of standard rules and procedures? The matter has long been considered problematic, as is indicated by the books of Goldschmidt (1891) and Fadda (1903, published again in 1987). The conclusion that has been generally accepted since the beginning of the twentieth century is that a specific and distinct commercial law did not exist in Rome, but only appeared with the ius mercatorum, in the Middle Ages (twelfth and thirteenth centuries). But, since the eighties, contrary opinions have surfaced in Italy, within the framework of the controversy between “primitivists” and “modernists.” In fact, the absence of commercial law as such has been perceived by some historians as an indication that economic life and thought were at a lower level of development in antiquity than in early modern history and even in the late Middle Ages. Serrao, leader of a “modernizing” tendency in legal history, rebelled against such a vision, and his disciples, Cerami, Petrucci, and Di Porto, have returned to this problem of the existence of a Roman commercial law. Readers may refer to a book published by Cerami and Petrucci (2010; cf. Petrucci 1991, 2002). Of course, how we define commercial law is crucial. This matter, and the general problem of commercial law, will not be treated here. Our present concern is to provide a rough outline of the professional law that Roman jurisprudence created around banking. I say professional law rather than commercial law, because it only concerned the members of the banking professions, and not all those involved in finance and banking. The constituent parts of this professional law are fourfold: production of rationes; solidarity in a society of bankers; the receptum argentarii; compensation. Did these four sets of regulations give preferential treatment to the bankers, as compared with other categories of financiers, or were they disadvantageous to them? And did they have a positive influence on economic life? In my book about the bankers (Andreau 1987a), I have stressed that some of these rules might have damaged their interests,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

104

Jean Andreau

even if this was not the intention of the authorities and jurists. I am rather less convinced of this than I was twenty-five years ago. As the authorities gave a lot of attention to securing the good working order of monetary circulation and of juridical contracts, it was not in their interest that the banks should go bankrupt! On the other hand, the financial situation of the bankers’ enterprises was certainly precarious, if one considers their size and how risky their activity was. As regards the Greek banks, Bogaert (1968) has very much stressed this fragility. If the bankers had been faced with the authorities’ hostility or with disadvantageous regulations, the fact that they carried on with their activity for several centuries would not be plausible. So, in order to reassess this conclusion, we need to look again at the relevant regulations. First of all is the “production of the bank-accounts, of the register” (editio rationum). All the operations effected on an account were entered in a register kept by the banker, the rationes, which constituted the tangible reality of his clients’ accounts. We do not know precisely how such a register of accounts was organized (Andreau 1987a: 615–26). Anyway, it had to be produced by the banker whenever a client was involved in a lawsuit, even if the banker was not personally concerned. The banker was required to produce (edere) everything relating to the particular account, for, in some manner, these records were considered to be the property of the client. This standard did not make the bankers’ activity easier, and it could have been an indirect way to control their professional behavior. However, it simplified the financial practice, for it allowed the relationship between bankers and clients to be exclusively based on accounts kept by the banker without adding other formalities to each operation. Second norm: compensation. If the client Titius had deposited 30 and owed 40 to the banker, and if Titius took the banker to court, he was not expected to calculate the compensation between all the operations of his account. The banker, however, did have to calculate this compensation, taking into consideration all the debts and deposits of the client. He could claim only the outstanding difference, 10, and moreover had to be careful to make no errors in his calculation of that difference. This rule may be regarded as harmful to the bankers; but it was favorable to the clients; a rule sensibly favorable to clients, if known by clients, would reduce transaction costs and help generate business for bankers. Moreover, in some way, like the previous rule, it may have been intended to minimize the distance

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome 105 between the banker and his client. Banking is a commercial business, and the banker runs the financial risk of his activity. But at the same time he is a credit intermediary, and these regulations made him a kind of direct financial assistant of his client. A third element of the professional law relates to the partnership of bankers. Generally speaking, in Roman law, the private partnerships of merchants or craftsmen did not have legal personality. They did not possess a patrimony distinct from the patrimonies of their socii. From that point of view, the bankers’ partnerships (argentarii socii) were no different from other private partnerships. Nevertheless, as regards their partners’ joint liabiblity, they did have a particular legal status. If one of the partners was creditor of a third party, the sum could be claimed and perceived by another partner (which is called active solidarity). In the same way, if one of the socii bankers contracted a debt, the creditor could claim for the entire sum from his partner or partners (which is called passive solidarity). These two solidarities were not applied to other commercial or manufacturing partnerships (Arangio Ruiz 1965: 78–83 and 144–5; Andreau 1987a: 626–31). In the bankers’ case, they seem to have become more substantial with the passing of time. In fact, the Rhetorica ad Herennium, at the beginning of the first century , considers it as a simple custom which was not a matter of legal standards, whereas, for the jurist Paul, at the beginning of the third century , it was a well-established juridical rule (Rhet. Her. 2.13.19; D. 2.14.25 pr., Paul. 3 ad ed.). This rule concerning solidarity (in case of a societas of bankers) gave more authority to the banking profession (compared to other businesses) and to the bank-accounts. Like the norm concerning editio rationum, it shows that banking was considered a prestigious activity, even if the bankers themselves were socially less important than some money-lenders belonging to the senate or the equestrian order. The specificity of banking was understood by the ancients, as the jurists’ texts stress. Thus, some of the texts speak of the publica causa and of the utilitas publica of the banks, or of the fides publica of the bankers. These complex notions deserve fuller discussion than is possible here. They show that the Romans regarded banking specifically as having some standing, even dignity—in connection with the notion of bank-account (ratio)—and this had consequences for the rights and duties of bankers. The last element, but not the least: the receptum argentarii. The receptum was reserved for argentarii (until the second century ,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

106

Jean Andreau

when it was extended to the nummularii). It was the undertaking that the banker gave to a third party, when he promised to pay him the money that his client owed to a third party (Lenel 1881: 62–71; Andreau 1987a: 597–602). The receptum involved three people (the banker, his client, and the third party), but it was binding upon only two of them, the banker and the third party. Legally, according to the receptum, the client did not need either to express his agreement or to be present. Clearly, however, in one form or another, an agreement also existed between the client and his banker on the payment of the debt in question. A praetorian action was instituted as regards the receptum, at the end of the second century or at the beginning of the first century , and it was used down to the third century  (after which it lapsed). The receptum is probably the most important among these professional regulations, but also the least known, because it had disappeared in the post-classical period: nearly nothing is said of it in Justinian’s Corpus iuris. It simplified the client’s role and the relations between him and the banker: the client only needed to worry about paying the debt to the banker and not about the arrangements that the banker had to make with the third party. Moreover, it is one of the indirect ways in which Roman law created forms of agency relationships. G. Dari Mattiacci draws my attention to the fact that the four standards I have just explained might reduce transaction costs among bankers, clients, and third parties, and I think it is a very useful idea, from the point of view of a neo-institutionalist analysis (on such analysis in Roman economic history, see for instance Lerouxel 2006, 2012, 2015, 2016; Terpstra 2008; Kehoe, Ratzan, and Yiftach 2015). In fact, even if all of them were not favorable to the bankers themselves, they could protect their clients’ interests, that is, those of artisans, wholesale merchants, and retailers, and play a stimulating influence on commercial and financial life. The bankers’ professions were linked with commerce and manufacture. With their presence in the fora, in markets, ports, and auctions, they were part of the fabric of commercial and manufacturing life. This is particularly relevant to Peninsular Italy, where they played a role in the foodsupply of the city of Rome, through the ports of Puteoli, Ostia, and Portus. Anyway, in the first century  and in the first century , there is no indication that banks suffered particularly because of these rules.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome 107 But the situation was transformed in the third century , because of monetary, military, and political developments, particularly during the last third of the third century, when prices soared. In that period, the banking professions virtually disappear from sight, for several generations. For financial life and the banking professions, it is a period of major crisis. The institutional role of bankers in auctions definitively disappears and will never reappear in history, in its earlier form. The professions of argentarii and coactores argentarii, which were directly linked to the provision of credit in auctions, also disappeared completely for almost a century; from the first half of the fourth century , the word argentarii designates silversmiths. It is only in the last part of the fourth century that money-changers and bankers are attested again (Andreau 1997: 133–55).¹

REFERENCES Andreau, Jean. 1968. “Banque grecque et banque romaine dans le théâtre de Plaute et de Térence.” 80 Mélanges de l’École Française de Rome 461–526. Andreau, Jean. 1974. Les Affaires de Monsieur Jucundus. Rome: Éditions de l’École Française de Rome. Andreau, Jean. 1987a. Vie financière dans le monde romain, Les métiers de manieurs d’argent (IVe siècle av. J.-C.-IIIe siècle ap. J.-C.). Rome: Éditions de l’École Française de Rome. Andreau, Jean. 1987b. “L’Espace de la vie financière à Rome,” in L’Urbs, Espace urbain et histoire (Ier siècle av. J.-C.-IIIe siècle ap. J.-C.). Rome: Éditions de l’École Française de Rome, 157–74. Andreau, Jean. 1997. Patrimoines, échanges et prêts d’argent: l’économie romaine. Rome: “L’Erma” di Bretschneider. Andreau, Jean. 1999. Banking and Business in the Roman World. Cambridge: Cambridge University Press. Andreau, Jean. 2000. “Les marchés hebdomadaires de Latium et de Campanie au 1er siècle ap. J.-C.,” in Elio Lo Cascio, ed., Mercati permanenti e mercati periodici nel mondo romano. Bari: Edipuglia, 69–91. Andreau, Jean. 2004. “Les Esclaves ‘hommes d’affaires’ et la gestion des ateliers et commerces,” in Andreau, France, and Pittia, eds, 111–26. Andreau, Jean, Jérôme France, and Sylvie Pittia, eds. 2004. Mentalités et choix économiques des Romains. Bordeaux-Pessac: Ausonius. ¹ I am grateful to Giuseppe Dari-Mattiacci and Robert Ellickson for their useful comments on a first version of this chapter.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

108

Jean Andreau

Arangio Ruiz, Vincenzo. 1965. La Società in diritto romano. Naples: Eugenio Jovene. Aubert, Jean-Jacques. 1994. Business Managers in Ancient Rome: A Social and Economic Study of Institores, 200 – 250. Leiden/NewYork/ Cologne: Brill. Barlow, Charles T. 1978. Bankers, Moneylenders and Interest Rates in the Roman Republic. Ann Arbor/London: University Microfilms International. Bogaert, Raymond. 1968. Banques et banquiers dans les cités grecques. Leiden: Sijthoff. Bogaert, Raymond. 1983a. “Note sur l’emploi du chèque en Egypte ptolémaïque.” 58 Chronique d’Egypte 212–21. Bogaert, Raymond. 1983b. “Les kollybistikai trapezai dans l’Egypte grécoromaine.” 3 Anagennesis 21–64. Bogaert, Raymond. 1987. “Recherches sur la banque en Egypte grécoromaine,” in Tony Hackens and Patrick Marchetti, eds, Histoire économique de l’Antiquité. Louvain-la-Neuve: Séminaire de Numismatique Marcel Hoc, 49–77. Bogaert, Raymond. 1994. Trapezitica Aegyptiaca: recueil de recherches sur la banque en Egypte gréco-romaine. Florence: Gonnelli. Camodeca, Giuseppe. 1992. L’Archivio puteolano dei Sulpicii, vol. I. Naples: Eugenio Jovene. Camodeca, Giuseppe. 1999. Tabulae Pompeianae Sulpiciorum: edizione critica dell’Archivio puteolano dei Sulpicii. Rome: Quasar. Camodeca, Giuseppe. 2000. “Per un primo aggiornamento all’edizione dell’archivio dei Sulpicii (TPSulp.).” 11 Cahiers du Centre Gustave Glotz 173–91. Camodeca, Giuseppe. 2003. “Il Credito negli archivi campani: il caso di Puteoli e di Herculaneum,” in Elio Lo Cascio, ed., Credito e moneta nel mondo romano (Atti degli Incontri Capresi di Storia dell’Economia antica (Capri, 12–14 ottobre 2000). Bari: Edipuglia, 69–98. Cerami, Pietro, and Petrucci, Aldo. 2010. Lezioni di diritto commerciale romano, 3rd edn. Turin: Giappichelli. Cohen, Edward E. 1992. Athenian Economy and Society: A Banking Perspective. Princeton: Princeton University Press. D’Arms, John H. 1981. Commerce and Social Standing in Ancient Rome. Cambridge, MA: Harvard University Press. de Ligt, Luuk, and Piter-Willem W. de Neeve. 1988. “Ancient periodic markets: festivals and fairs.” 66 Athenaeum 391–416. de Ligt, Luuk. 1993. Fairs and Markets in the Roman Empire. Amsterdam: J. C. Gieben. Domergue, Claude. 1983. La Mine antique d’Aljustrel (Portugal) et les tables de bronze de Vipasca. Bordeaux-Talence: Publications du Centre Pierre Paris.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome 109 Fadda, Carlo. 1903 [repr. 1987]. Istituti commerciali del diritto romano. Naples: Jovene. Früchtl, Alois. 1912. Die Geldgeschäfte bei Cicero. Erlangen: Junge & Sohn. Garcia Morcillo, Marta. 2005. Las Ventas por subasta en el mundo romano, La Esfera privada. Barcelona: Universitat de Barcelona. Garcia Morcillo, Marta. 2008. “Auctions, bankers and public finances in the Roman world,” in Verboven, Vandorpe, and Chankowski, eds, 257–75. Garcia Morcillo, Marta. 2014. “Publicidad, transparencia y legitimidad: subastas en la obra de Cicerón,” in Catherine Apicella, Marie-Laurence Haack, and François Lerouxel, eds, Les Affaires de Monsieur Andreau, Économie et société du monde romain. Bordeaux-Pessac: Ausonius, 209–22. Goldschmidt, Lewin. 1891. Universalgeschichte des Handelsrechts. Stuttgart: F. Enke. Harris, William V. 2006. “A revisionist view of Roman money.” 96 Journal of Roman Studies 1–24. Harris, William V. 2008. “The nature of Roman money,” in William V. Harris, ed., The Monetary Systems of the Greeks and Romans. Oxford: Oxford University Press, 174–207. Hollander, David B. 2007. Money in the Late Roman Republic. Leiden/ Boston: Brill. Howgego, Christopher. 1992. “The supply and use of money in the Roman world, 200  to  300.” 82 Journal of Roman Studies 1–31. Ioannatou, Marina. 2006. Affaires d’argent dans la correspondance de Cicéron, L’aristocratie sénatoriale face à ses dettes. Paris: De Boccard. Jacques, François, and Scheid, John. 1990. Rome et l’intégration de l’Empire, 44 av. J.-C.–260 ap. J.-C., vol. 1, Les Structures de l’Empire romain. Paris: Presses Universitaires de France. Jakab, Eva. 2011. “Financial transactions by women in Puteoli,” in Paul J. du Plessis, ed., New Frontiers: Law and Society in the Roman World. Edinburgh: Edinburgh University Press, 123–50. Johnston, David. 1999. Roman Law in Context. Cambridge: Cambridge University Press. Jones, David Francis. 2006. The Bankers of Puteoli: Finance, Trade and Industry in the Roman World. London: Tempus. Jongman, Wim. 1988. The Economy and Society of Pompeii. Amsterdam: J. C. Gieben. Kehoe, Dennis P., David M. Ratzan, and Uri Yiftach, eds. 2015. Law and Transaction Costs in the Roman Economy. Ann Arbor: University of Michigan Press. Lenel, Otto. 1881. “Beiträge zur Kunde des Edicts und der Editcommentare.” 2 Zeitschrift der Savigny-Stiftung für Rechtsgeschichte 14–83.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

110

Jean Andreau

Lerouxel, François. 2006. “Les femmes sur le marché du crédit en Égypte romaine (30 av. J.-C.–284 apr. J.-C.). Une approche néo-institutionnaliste.” 37 Cahiers du Centre de Recherches Historiques 121–36. Lerouxel, François. 2008. “La banque privée romaine et le marché du crédit dans les tablettes de Murecine et les papyrus d’Égypte romaine,” in Verboven, Vandorpe, and Chankowski, eds, 169–98. Lerouxel, François. 2012. “Le marché du crédit privé, la bibliothèque des acquêts et les tâches publiques en Égypte romaine.” 67 Annales, Histoire Sciences Sociales 943–76. Lerouxel, François. 2015. “The βιβλιοθήκη ἐγκτήσεων and transaction costs in the credit market of Roman Egypt (30 ... to ca. 170 ..),” in Kehoe, Ratzan, and Yiftach, eds, 162–84. Lerouxel, François. 2016. Le Marché du crédit dans le monde romain, Rome: Éditions de l’École Française de Rome. Magdelain, André. 1990. Ius imperium auctoritas: Études de droit romain. Rome: Éditions de l’École Française de Rome. Maganzani, Lauretta. 2007. “L’Editto provinciale alla luce delle Verrine: profili strutturali, criteri applicative,” in Julien Dubouloz and Sylvie Pittia, eds, La Sicile de Cicéron: Lectures des Verrines. Besançon: Presses Universitaires de Franche-Comté, 127–46. Mélèze-Modrzejewski, Joseph. 1970. “La Règle de droit dans l’Egypte romaine,” in D. H. Samuel, ed., Proceedings of the Twelfth International Congress of Papyrology. Toronto: A. M. Hakkert, 317–77. Mélèze-Modrzejewski, Joseph. 1990. Droit impérial et traditions locales dans l’Egypte romaine. Aldershot: Ashgate. Michel, Jacques-Henri. 1962. La Gratuité en droit romain. Bruxelles: Université Libre de Bruxelles. Millett, Paul. 1991. Lending and Borrowing in Ancient Athens. Cambridge: Cambridge University Press. Nicolet, Claude. 1966. L’Ordre équestre à l’époque républicaine (312–43 av. J.-C.), vol. I: Définitions juridiques et structures sociales. Paris: De Boccard. Petrucci, A. 1991. Mensam exercere: studi sull’impresa finanzaria romana: II secolo a.C–metà del III secolo d.C. Naples: Jovene. Petrucci, A. 2002. Profili giuridici delle attività e dell’organizzazione delle banche romane. Turin: Giappichelli. Rauh, Nicholas. 1989. “Finances and estate sales in republican Rome.” 63 Aevum 45–76. Saller, Richard P. 1982. Personal Patronage under the Early Empire. Cambridge: Cambridge University Press. Schulz, Fritz. 1946. History of Roman Legal Science. Oxford: Clarendon Press. Shatzman, Israel. 1975. Senatorial Wealth and Roman Politics. Bruxelles: Latomus.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Banking, Money-Lending, and Elite Financial Life in Rome 111 Terpstra, Taco. 2008. “Roman Law, Transactions Costs and the Roman Economy: Evidence from the Sulpicii Archive,” in Verboven, Vandorpe, and Chankowski, eds, 345–69. Thielmann, Georg. 1961. Die römische Privatauktion, zugleich ein Beitrag zum römischen Bankierrecht. Berlin: Duncker & Humblot. Thomas, Yan. 2004. “Travail incorporé dans une matière première, travail d’usage et travail comme marchandise. Le droit comme matrice des catégories économiques à Rome,” in Andreau, France, and Pittia, eds, 201–25. Thomsen, Rudi. 1978. “From libral aes grave to uncial aes reduction, the literary tradition and the numismatic evidence,” in Les Dévaluations à Rome: Epoque républicaine et impériale, vol. 1. Rome: Éditions de l’École Française de Rome, 9–30. Verboven, Koen. 1993. “Le système financier à la fin de la République romaine.” 24 Ancient Society 69–98. Verboven, Koen. 2002. The Economy of Friends: Economic Aspects of Amicitia and Patronage in the Late Republic. Bruxelles: Latomus. Verboven, Koen. 2003a. “The Sulpicii from Puteoli, argentarii or faeneratores?,” in Pol Defosse, ed., Hommages à Carl Deroux, vol. III, Histoire et Epigraphie, Droit. Bruxelles: Latomus, 429–45. Verboven, Koen. 2003b. “The Sulpicii from Puteoli and Usury in the Early Roman Empire.” 71 Tijdschrift voor Rechtsgechiedenis 7–24. Verboven, Koen, Katelijn Vandorpe, and Véronique Chankowski, eds, Pistoi dia tèn technèn: Bankers, Loans and Archives in the Ancient World, Studies in Honour of Raymond Bogaert. Leuven: Peters. Verley, Patrick. 1985. La révolution industrielle. Paris: Editions MA. Yiftach-Firanko, Uri. 2009. “Law in Graeco-Roman Egypt: Hellenization, fusion, Romanization,” in Roger S. Bagnall, ed., The Oxford Handbook of Papyrology. Oxford: Oxford University Press, 541–60. Zehnacker, Hubert. 1980. “Unciarium fenus (Tacite, Annales, VI, 16),” in Mélanges de littérature et d’épigraphie latines, d’histoire ancienne et d’archéologie (= Mélanges Pierre Wuilleumier). Paris: Belles Lettres, 353–62. Zehnacker, Hubert. 1990. “Rome, une société archaïque au contact de la monnaie (Vie–Ive siècles av. J.-C.),” in Crise et transformation des sociétés archaïques de l’Italie antique au Ve siècle av. J.-C. Rome: Éditions de l’École Française de Rome, 307–26.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

16 Secured Transactions in Classical Roman Law Hendrik L. E. Verhagen

“For historians and lawyers interested in the connection of law with economics the Roman law of real security will forever remain an instructive phenomenon.” (Schulz 1951: 405)

16.1. INTRODUCTION When the rules for the taking and enforcement of security are “cumbersome, inefficient and awkward” (Fleisig 2008: 90) security may lose its essential economic function of mitigating risk for lenders. This is the condition of the Roman law of real security according to the prevailing view in modern literature on Roman law. It is often added, however, that flaws in real security were not a major problem, as personal security (guarantees, suretyships) was far more important in ancient Rome than real security.¹ This may have been true for the “economy of friends,” where loans were provided within the framework of amicitia (“friendship”) relationships of members of the elite ¹ Schulz (1951: 405)—see §16.5. In a similar sense, Zimmermann (1996: 115–16); Johnston (1999: 94). More nuanced Kaser (1971: 457); Kaser, Knütel, and Lohsse (2017: 180). For positive assessments of the Roman law of real security, see Krämer (2007: 1–6); Harris (2006); Terpstra (2008: 356–60). Hendrik L. E. Verhagen, Secured Transactions in Classical Roman Law In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0016

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

114

Hendrik L. E. Verhagen

and friends were expected to enhance each other’s borrowing capacity by standing surety and to discourage one another’s threatening insolvency.² But in the archive of the Sulpicii, a collection of legal documents involving a group of professional Roman (Puteoli) money-lenders, we see that in ordinary commercial relationships personal security was used for small loans and real security for larger ones.³ In the papyri of Roman Egypt personal security is rare throughout the Principate (Lerouxel 2016: 245–91). It is also hard to believe that the Roman law of real security was relatively insignificant when so many opinions of the jurists on pignus have survived.⁴ The Roman law of security was highly sophisticated and versatile, allowing multiple charges, non-possessory security, and even floating charges. As legal systems often adapt in reaction to impulses from their economic environment, the complexity of the Roman law of real security would suggest that pignus and fiducia did play an important role in the Roman economy.⁵ In this contribution it will be argued that the prevailing view is wrong and that the Roman law of real security was capable of facilitating credit in a similar manner as its descendants in modern economies. The economic functions of real security will be reviewed (§16.2). The requirements for an effective law of security, which are to a large extent a reflection of these functions, will subsequently be applied to Roman law (§16.3). The

² Kaser (1982: 216–7); Krämer (2007: 3); Verboven (2002: 151). ³ Gröschler (2008: 303–5); Krämer (2007: 182). For the loans secured by real security no personal security seems to have been granted. In Kaser, Knütel, and Lohsse (2017: 180) it is noted that, although at the beginning of the Principate personal security was still preferred, in the second century  real security had developed into the most significant form of security. The archive of the Sulpicii would suggest that this development already took place in the first century . This is confirmed by Chemain (2015): 377–386. There are two excellent editions of the Sulpicii archive: Camodeca (1999) (TPSulp) and Wolf (2012) (TPN). ⁴ In the Digest a whole book (D. 20) and a large title (D. 13.7) are devoted to real security (pignus and hypotheca), while the Codex also contains a large number of rescripts and other imperial constitutions dealing with pignus (C. 4.24; C. 8.13–8.34). Also elsewhere in the Digest there are many fragments on pignus and hypotheca. In the sources, the term pignus (“pledge”) is often used as the generic term for possessory and non-possessory pledges, while the term hypotheca usually (but by no means always) denotes non-possessory pledges specifically. ⁵ On the co-evolution of law and economy, see in particular Roe (1996) and Luhmann (2004: 230–73). See also the contributions to Zumbansen and Callies (2011) (with many further references).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

115

(lack of) publicity in the Roman law of real security will be given separate attention (§16.4), as it is the most controversial aspect of the Roman law of real security.

16.2. ROMAN LAW AND ECONOMICS

16.2.1. Economic Analysis of Roman Security Interests In this chapter it will be examined what law and economics has to say about the role of the law of secured credit in modern economies and it will be indicated whether there is enough similarity with the role of the law of secured credit in the Roman empire in order to apply the findings of law and economics to it. The laws of real security of modern civil law jurisdictions are built on Roman foundations and are strikingly similar to them. In particular the Roman legal rules on pignus—as contained in the Corpus iuris civilis—were received in the ius commune of medieval Europe and were eventually (usually in only slightly modified form) adopted in the great codifications of modern Europe. There is therefore a continuity of congruent legal solutions, which allows the Roman law of real security to be analyzed in a similar fashion as modern laws (Wieacker 1988: 29–30). It is true that the formal rules in the codifications may have acquired different meanings in the course of history. Moreover, the functioning of the formal rules may be subject to different informal constraints. More generally, the socio-economic environment of modern legal rules may be so different from that of their Roman ancestors that a one-to-one transposition cannot be made.⁶ One of the most important historians of the Roman economy, Moses Finley, speaks of “the inapplicability to the ancient world of a market-centred analysis” (Finley 1999 (1973): 26). If Finley is right that ancient society did not have a market economy and that the ancient economy is better analyzed with sociological methods examining the status of persons and their relationships, then the economic analysis of ancient law might also be doomed to fail. ⁶ As Schiavone (2000: 177) warns us: “one should not be misled by the apparent persistence of the juridical forms that cover the economic relationships, in cases in which the modern usage of Roman law seems to be evidence of uninterrupted contact between the ancient and the modern.”

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

116

Hendrik L. E. Verhagen

Finley has been enormously influential in the Anglo-American world. The school of thought which is considered by the Oxford Handbook of Roman Studies (Scheidel) as the most influential one, still has clear “primitivist” connotations. It considers the Roman economy as “a highly localized, fragmented, and largely agrarian economy that sustained a thin veneer of coerced transfers and trade in luxuries and a network of towns that were dominated by landowning elites” (Scheidel 2010: 593). The purpose of this chapter is not to take sides in the “Bücher/Meyer controversy” between “primitivists” and “modernists.” This chapter merely seeks to demonstrate that Roman law did offer a legal framework sophisticated enough for a market economy in which finance plays an important role. If that is so, however, one wonders how likely it is that a state with a “primitive” economy, even as described by moderate “neo-primitivists” as having only a “thin veneer of coerced transfers and trade in luxuries,” would develop a highly versatile and differentiated law of secured credit?⁷ In any case, although for other areas of the law this may well be different, I would claim that the interests at stake when a grain merchant in first-century  Puteoli pledges his stocks in favor of his financier are not much different from those when his colleague in twenty-first-century Rotterdam does the same.⁸ As will be elaborated below, the economic functions that are attributed to real security in studies on secured credit in modern economies can, without difficulty, also be applied to the Roman law of real security.

16.2.2. Economic Functions of Security 16.2.2.1. Primary Function of Security: Risk Reduction In their groundbreaking Yale Law Journal article Jackson and Kronman (1979: 1143) observed that “[t]o a considerable extent, the value of a security interest depends on the degree to which it insulates the secured party from the claims of the debtor’s other creditors.” The primary economic function of security is therefore ⁷ In a similar sense, with reference to the Roman law of sale, Crook (1996: 35–6). So also for commercial law generally Johnston (1999: 110–1), quoted with approval by Frier (2000: 447). ⁸ See also Arruñada (in this volume), who assumes that “the basic problems that market participants faced in Roman times were essentially the same as they face now . . . ” (p. 256).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

117

risk reduction for the creditor (Röver 1999: 108). In case of secured loans the risk concerned is that of non-payment of principal and interest at the agreed time. The more this risk is reduced by taking security, the greater its value will be to the creditor. The risk manifests itself in particular when the debtor is insolvent. In that event creditors are likely to receive only a small portion of their claims, since the liquidation proceeds of the debtor’s assets are usually significantly less than the aggregate amount of his or her debts. Under the principle of pari passu or paritas creditorum (“equality of creditors”) all creditors receive distributions proportionate to their claims. Real security, however, creates a separate class of creditors, with preferential rights to the proceeds of specific assets (the charged assets). Thus, a secured lender who has obtained a security interest on one or more of the debtor’s assets is not affected by this pari passu principle. In other words, where the law allows a creditor to obtain effective security, this will considerably reduce the risk that he or she will suffer losses when the loan is not fully repaid at the agreed time. The corresponding benefit for debtors granting security is that they may obtain credit in a manner (amount, term) which otherwise would not be available to them or only at higher costs (interest). Also, in ancient Rome the primary reason for taking security cannot have been anything else than maximizing the lender’s prospects for recovery, and so it is likely to have facilitated the borrower’s access to credit.⁹

16.2.2.2. Monitoring, Bonding, and Moral Hazard Another economic function of real security has to do with monitoring the debtor. Where security is taken in order to secure recovery, the relatively simple process of monitoring the market value of the charged assets can replace the far more complex process of monitoring the borrower’s business as a whole. By focusing on the continued availability of the charged assets, the secured creditor can achieve a substantial reduction of his or her monitoring costs (Jackson and ⁹ Harris (2006: 6) mentions Cicero, Ad Att. 16.6.3 as an example of a borrower’s differential treatment of secured and unsecured debt. The Sulpicii archive unfortunately does not contain documents mentioning interest rates (although they must have calculated interest). In particular for Roman Egypt there is relatively much evidence for the enhancement of borrowing capacity by real security. See Lerouxel (2015), who does, however, not discuss whether the interest rate for secured loans was substantially lower than for unsecured loans.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

118

Hendrik L. E. Verhagen

Kronman 1979: 1153).¹⁰ A related function is that where the security interest is attached to the charged property, this will restrict debtor misbehavior, consisting of transferring assets to third parties (Jackson and Kronman 1979: 1153). The same purpose (i.e. bonding) could be achieved by contractual covenants in loan agreements in which the borrower undertakes not to sell certain assets without the lender’s permission. However, when the borrower is declared bankrupt, a contractual action for breach of covenant will normally only be marginally recoverable. A proprietary security interest will allow the lender to enforce the security, even when the charged assets have been transferred to third parties. Taking security may also prevent the debtor from shifting risks to the detriment of his or her creditors by adopting a more risky investment strategy after the credit has been granted (moral hazard). The advantages of such a new strategy would normally accrue for the debtor, while the disadvantages (insolvency) would belong to the creditors. Security will entail that the debtors will lose the charged assets if they default on their borrowings, which may discourage them from adopting risky investment strategies (Röver 1999: 14). One economic function of security therefore is that it can effectively restrict the debtor’s ability to transfer certain assets of stable value to fund more risky transactions (Armour 2008: 7–8). Again there is no reason why these functions should not also be performed by real security in Rome. Let us look at one example derived from the Sulpicii archive. In TPN 43 (TPSulp 51) the debtor declares that the wheat, pledged in order to secure the initial loan of HS 10,000, is stored “with me” (penes me) in the public granaries of Puteoli. When some time later an additional loan of HS 3,000 was granted, the same wheat was pledged to secure the total debt of HS 13,000. This time the debtor no longer declares that the wheat is stored “with me” (TPN 44/TPSulp 52). From another document (TPN 86/TPSulp 45) it becomes clear why: the creditor rented the space in the granaries where the pledged wheat was stored (for a nominal amount of HS 1 per month), thus effectively bringing the pledged property under his control. What must have happened here is that the pledge was originally created as a non-possessory pledge and was later converted into a possessory one (Krämer 2007: 300–38; Verhagen 2011: 28). The motivation behind this conversion is not

¹⁰ See also my discussion of impersonal exchange below, §16.2.2.3.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

119

expressed, but is likely to have been that the creditor wished to strengthen his position. Bringing the pledged wheat under his control would entail that the creditor could be sure that as long as there was sufficient wheat in the storeroom rented by him, it would be less urgent to keep an eye on the debtor (monitoring). Moreover, the borrower’s loss of control over the charged wheat may have been an important stimulus for the proper repayment of the secured loan and would prevent him from using the assets for more risky investment strategies (bonding and moral hazard).

16.2.2.3. Depersonalizing Credit Relationships: Impersonal Exchange and Adverse Selection Economic development is facilitated by legal and other social institutions that support impersonal exchange (Arruñada 2012: 17). In particular Verboven, however, has demonstrated that highly personalized relationships based on amicitia played an important role in the Roman economy. Amicitia was not merely a relationship based on mutual affection and altruism; it could also involve the exchange of money, goods, and services. In many cases, the exchange aspect was even the raison d’être of the amicitia relationship. In financial transactions amicitia entailed that one friend helped out another, by providing interest-free loans or at low(er) interest (Verboven 2002: 116). Also productive loans at higher interest (but still on more favorable terms) were preferably obtained from friends. Amicitia would provide the lender with an extra guarantee that repayment would take place, while it would protect the debtor from excessive interest rates and allow him to count on more leniency when unable to repay the loan in time (Verboven 2002: 170–4). However, if amicitia was as pervasive in the financial world as Verboven claims it to be, one wonders how the substantial evidence for real security can be explained.¹¹ Precisely real security enables a lender to provide credit outside the circle of his or her relatives, friends, and their dependents. The lender’s lack of information as to the borrower’s creditworthiness is much less a concern when the

¹¹ Terpstra (2013: 29): “if indeed amicitia and clientela were all-important in business, it is simply not visible in the Murecine documents.” The Murecine documents are those of the Sulpicii archive, discovered at the locality of Murecine near Pompeii (but originally stemming from Puteoli).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

120

Hendrik L. E. Verhagen

value of the collateral will be sufficient to cover the debt.¹² The law of real security may have provided a mechanism of bringing lenders and borrowers together who otherwise would not have engaged in credit transactions. The Roman law of real security may have facilitated impersonal exchange in financial transactions and may thus have contributed to economic growth in the Roman empire. To put it differently, the existence of a large body of fragments in the Digest and the Code on real security at least suggests that many credit relationships were depersonalized.¹³ In these depersonalized relationships Roman real security may also have had an information function, distinguishing good borrowers from bad ones (adverse selection). Adverse selection may occur in markets with asymmetrical information: information available to one party (often about himself or herself) may remain hidden for the other party. This could lead to adverse selection: parties who have most to hide are most eager to enter into certain contracts (Stiglitz and Weiss 1981: 393–410). The availability of security may, however, reduce this process of adverse selection. Professional lenders generally anticipate that only serious borrowers will be prepared (or be able) to grant security. Thus security not only has a risk reducing function but also an information function, distinguishing good borrowers from bad ones (Röver 1999: 116–17).

16.2.2.4. Zero-Sum Game or Redistribution of Wealth to Secured Creditors? From a macro-economic perspective, the most important advantage of the microeconomic function of risk reduction is that it increases available credit, which leads to higher investments, increased production, and ultimately a higher gross domestic product (Röver 1999:

¹² For Roman Egypt, see Lerouxel (2012: 960–3). According to Kay there are two ways around the problem of the lender’s lack of information as to the borrower’s creditworthiness: using a financial broker (with information) introducing the borrower to the lender and deposit banking. In the latter case the lender takes a credit risk on the banker rather than on the ultimate borrower, while the banker’s expertise is in evaluating the creditworthiness of persons. Kay (2014: 110) does not mention, however, real security as another way of remedying the lender’s lack of information. ¹³ On the other hand, rights of pledge were also created in highly personalized contexts, such as between husband and wife (e.g. D. 20.6.11, Paul. 4 resp.) and between brothers (e.g. D. 20.4.3.2, Pap. 11 resp.).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

121

109–10). A fundamental objection, however, against the efficiency of proprietary security is that secured credit involves a zero-sum game: interest rate savings for debtors granting security are wiped out by corresponding interest rate increases for debtors unable to offer security (Schwartz 1984). So-called “redistributive” theories even hold that the benefit of lower interest rates is acquired at the expense of nonsecured creditors, who will receive less in the insolvency of a debtor who has granted security rights to one or more other creditors. The growing empirical literature on the use of secured credit, however, refutes these redistributive theories and demonstrates that, as Armour (2008: 11–2) states, “secured credit is, on the whole, socially beneficial, and . . . such benefits are highly likely to outweigh the social costs of any transaction motivated by redistribution.” Armour argues that granting security, by facilitating monitoring and bonding, reduces the probability of the debtor engaging in risky wealth-reducing transactions and thus increases the value of all creditors’ claims (2008: 7–8). Some years ago it was argued that, in the more than thirty years of debate since the Jackson/Kronman article, “a comprehensive justification of secured commercial credit on efficiency grounds is unproven and perhaps not provable” (McCall 2009: 11). However, legal history seems to provide strong indications that security is efficient. As McCormack (2004: 26) has put it: “an inefficiency conclusion would go against the grain of history. Security devices are widespread and pervasive not only in the modern industrialized world but also in ancient societies, and one might ask the rhetorical question: why does secured credit persist for so long if it is inefficient?”

16.3. EFFECTIVENESS OF THE ROMAN LAW OF REAL SECURITY

16.3.1. Priority and Certainty in Ranking The law of real security will, as we have seen, only be able to perform its economic function of facilitating access to credit if the secured lender’s claims are insulated from the claims of other creditors. An effective law of security therefore requires, first and foremost, that the creditor’s claims can be satisfied with priority over the other creditors from the proceeds of the charged assets. In addition, the basic

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

122

Hendrik L. E. Verhagen

requirement of priority should be supplemented with rules creating certainty in ranking. It should—at the outset of a secured finance transaction—be perfectly clear to the lender what the ranking of his or her security right will be: whether the tax authorities will have a higher-ranking privilege, whether other secured or privileged creditors could prevail, et cetera. Roman law does to a large extent meet these requirements for an effective law of security. There could be no doubt for secured creditors that the proceeds of the charged property were exclusively available in order to discharge the secured debt. Also when insolvency proceedings were conducted in respect of the debtor, the secured creditor could take recourse against the charged property with priority over other creditors. If the creditor was in possession of the charged property, he or she could sell it to third parties and would only be required to pay the surplus to the person conducting the insolvency proceedings (in particular: the bonorum emptor). If the pledge was non-possessory, the creditor could institute the actio Serviana to recover the pledged property from the person in charge of the relevant stage of the insolvency proceedings (Kaser and Hackl 1996: 402). There could, however, sometimes be creditors (in particular: the imperial treasury and local authorities) with superior rights of recourse against the pledged property, but this was rare.¹⁴ Only from the second century onwards there was an increasing number of cases in which all the debtor’s assets would be subject to a right of pledge arising by operation of law, in particular in favor of the tax authorities. These statutory general hypothecs may not have been as devastating for secured credit as is often assumed in modern literature. They would generally also be subject to the prior tempore principle, so that they could not adversely affect existing security rights. Besides, when a creditor satisfied the tax debts of a debtor, ¹⁴ In D. 49.14.28 (Ulp. 3 disp.) the imperial treasury’s preferential right of recourse over an after-acquired asset was granted preference over an anterior general pledge created by the debtor in favor of a private individual. On this interesting text, in which Ulpian restates an opinion by Papinian, see Wagner (1974: 180–92). Sometimes the ranking was reversed by operation of law, in particular when the money provided by a later secured creditor was used for the benefit of the pledged property, such as the equipment, crew, or repair of ships or the storage of merchandise. See D. 20.4.5 (Ulp. 3 disp.) and D. 20.4.6 (Ulp. 73 ad ed.). In a rescript from 293 by Diocletian and Maximian (C. 8.17.7) the pledge over real estate granted to the lender who financed the purchase of the estate is considered to have priority over any other pledge (e.g. a general pledge created earlier).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

123

such creditor would be granted the same preferential position as the tax authorities (Wagner 1974: 191). From Diocletian onwards, however, higher tax pressures may have increased the cases where assets were previously charged with a fiscal pledge.

16.3.2. Creating Security One benefit of taking security is that it provides a lower-cost method for achieving what otherwise should be achieved by more complex contractual arrangements. Jackson and Kronman (1979: 1157) observe that if the law prohibited debtors from preferring one or more creditors over others by granting security, a similar “network of priority relationships could be expected to emerge by consensual arrangement between creditors.” However, particularly where the number of general creditors is large, such arrangements will be more complicated to achieve and involve more transaction costs. This would be true for ancient Rome as well. However, security will only have this advantage of saving on transaction costs for more complex priority relationships when the law of secured credit offers efficiency in creation. In order to reduce transaction costs, there should be as few formalities for creating security as possible. Ideally there would be a single method for creating security rights rather than a multiplicity of methods for different types of charged assets (Armour 2008: 16). Also the creation of security over after-acquired property should not be cumbersome. It seems that originally in Rome res mancipi were always provided as collateral by way of fiducia (cum creditore) and res nec mancipi by way of pignus. In case of fiducia the debtor would—through mancipatio—transfer land, slaves, horses, or cattle as collateral to a creditor. A transfer in court by way of in iure cessio would also be possible, though is likely to have been rare. Gaius explains why: “But generally, in fact more or less always, we mancipate. For there is no point and no need to do with greater difficulty in the presence of a praetor or the governor of a province what we can do ourselves in the presence of friends” (Gai., Inst. 2.25).¹⁵ Gaius can be regarded as offering secondcentury advice to save on transaction costs! The epigraphic sources all ¹⁵ Plerumque tamen et fere semper mancipationibus utimur: quod enim ipsi per nos praesentibus amicis agere possumus, hoc non interest nec necesse cum maiore difficultate apud praetorem aut apud praesidem prouinciae agere. Translation taken from Gordon and Robinson (1997: 134–5).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

124

Hendrik L. E. Verhagen

concern a fiducia cum creditore with a mancipatio (Noordraven 1999: 154–5). In the classical period it was possible that res mancipi could also be charged by way of pignus (Verhagen 2011: 8–9).¹⁶ Thus a single method for creating security over all types of asset became available, which, moreover, could also be used for property subsequently. For the creation of a right of pignus no formalities existed. Initially it was probably required that the charged property was brought under the factual control of the creditor, but at a later stage non-possessory pledges could be created “by mere agreement” (nuda conventione).

16.3.3. Comprehensiveness Another requirement for an effective law of real security is that of comprehensiveness. First, granting and taking security should be available to a wide range of debtors and creditors. For instance, granting security should not be confined to persons acting in the exercise of a trade or profession, and taking security should not be confined to licensed credit institutions. No such limitations existed in Roman law. Moreover, women could grant or accept security, and security was often granted by or to slaves acting for their masters, the latter category often in connection with a banking business.¹⁷ Security could also be granted for someone else’s debts (D. 20.1.5 pr., Marcian., ad form. hyp.), although for women this was prohibited by the senatus consultum Vellaeanum ( 46?). Furthermore, security rights should be possible with respect to a wide range of assets: not only movable and immovable property, but also receivables (the pledgor’s claims against his or her debtors). This was the case in Roman law: all types of assets (including receivables) could be encumbered with a right of pignus (Verhagen 2013a: 60–2). Moreover, it should be possible to create generic charges, for instance over a herd or a shop’s inventory, without it being necessary that each individual item be identified and that frequent updates be executed for after-acquired items. This would increase transaction costs and

¹⁶ Ultimately pignus completely supplanted fiducia (although not yet in the late classical period); Justinian’s compilers deleted fiducia from the jurists’ fragments: one of the most important examples of systematic interpolation. ¹⁷ See e.g. TPN 73 (TPSulp 90) and the Mancipatio Pompeiana, FIRA III, nr. 91 (security (fiducia) by women); TPN 43, 44, and 69 (TPSulp 55, 51, 52, and 79) (slaves).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

125

would be likely to affect the availability or cost (interest rate) of credit. Under classical Roman law generic pledges could be created over present and future assets, without the need to specify each individual item.¹⁸ From the second century  it even became possible to create a “floating charge” over all the debtor’s present and future assets (Verhagen 2013a: 63–6).¹⁹ Finally, it should be possible to create security for a wide range of debts. Roman law also complied with this condition for an effective law of security. Security could be granted not only for debts assumed at the time of its creation, but also for previously existing debts, contingent debts, and even future debts. The priority of the right of pledge could then be determined by the time at which the pledge was granted rather than by the time the secured debt arose. For example, in D. 20.4.9 pr. (Afr. 8 quaest.) the slave Eros was first pledged to a landlord, in order to secure debts under a lease that had not yet entered into effect, and then to a lender in order to secure a loan, which (as was often the case in Rome) was immediately due and payable. Although the debt owed to the lender came into being first, the lessor’s right of pledge was first-ranking, as it had been created earlier. Where, on the other hand, the contract pursuant to which the secured debt would arise had not yet been entered into at the time of creation, the pledge’s priority would be determined by the time at which the secured debt arose. Thus, where a pledge had been granted in order to secure a loan that had yet to be entered into, the priority of the pledge would be determined by the time at which the borrowed money was actually paid out to the borrower (D. 20.4.1.1, Pap. 8 quaest.).

16.3.4. Rights of Use 16.3.4.1. Hypotheca and Debtor’s Right to Dispose Empirical studies for contemporary economies demonstrate that non-possessory security leads to a greater availability of credit (Armour 2008: 19). Charged assets that are essential for the debtor’s

¹⁸ D. 20.1.13 pr. (Marcian., ad form. hyp.): herd; D. 20.1.34 pr. (Scaev. 27 dig.): shop. ¹⁹ e.g. D. 20.1.1. pr. (Pap. 11 resp.). See Wagner (1968). For a complete overview of all Digest fragments on generic and general pledges, see Mentxaka (1986: 279–350). For a comparison of the Roman general pledge with the floating charge of the common law, see Verhagen (2013b: 135–42).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

126

Hendrik L. E. Verhagen

business should still be able to generate income with which the secured debt can be discharged. By allowing the debtor to remain in possession of the charged assets, he or she will be able to continue using the assets for the exercise of a trade or profession. Also, the debtor should be able (with the creditor’s permission) to dispose of the charged assets in the ordinary course of his or her business. In all likelihood the Roman right of pignus originated in the early ius civile as a form of possessory security.²⁰ It is often held in modern writing that the non-possessory pledge was only first recognized at the time of the second-century jurist Julian, who is the author of the earliest legal opinions on this form of security.²¹ It seems, however, that pignus already could exist as a non-possessory security right much earlier than Julian. In the Digest there are several fragments from before Julian that seem to concern non-possessory security (Krämer 2007: 249–75).²² The archive of the Sulpicii provides epigraphic evidence of a transactional practice of creating nonpossessory pledges already in the first century .²³ In any case, it is clear that in the classical period pignus could be created as a nonpossessory security interest and that it was widely used. Moreover, although originally the debtor could not transfer pledged property, this later changed. With the secured creditor’s (express or implied) consent the debtor could even transfer the pledged property without the pledge continuing to attach to it.²⁴

16.3.4.2. Multiple Pledges In particular when the (expected) liquidation proceeds of a charged asset exceed the amount of the secured debt, the debtor can optimize ²⁰ The great majority of modern authors accepts this: see Krämer (2007: 121), with further references. ²¹ Krämer (2007: 37), whose own view is that the non-possessory pledge was recognized much earlier. See also Kaser (1971: 457). ²² One example is D. 13.7.30 Paul. 5 Epit. Alf. dig.), in which Paul restates the opinion of the jurist Alfenus Varus (first century ) on what may have been a nonpossessory pledge of a boat. Krämer (2007: 211–48). Another (in my opinion more problematic) example is D. 13.7.18.3 (Paul. 29 ad ed.), referring to the opinion of Cassius (first century ). ²³ See §16.2.2.2 above. Fiducia cum creditore had always been possible as a form of non-possessory security. The mancipatio did not require an actual transfer of possession. Often the debtor would rent the property from the creditor or hold it as bailee (precarium): Gai., Inst. 2.60. ²⁴ See §16.4.4.4 below.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

127

the collateral value of the asset if the law enables him or her to create several layers of security in favor of multiple creditors. In classical Roman law this was for some time problematic. For, after the first fiduciary transfer, a debtor would no longer be the owner of the property and would therefore not have the power to transfer it validly to another creditor. When the security granted to the creditors was a right of pignus the position originally was largely the same. Although there is good reason to assume that the non-possessory pledge was recognized relatively early, the recognition of its “logical sequel” (Kunkel 1973: 155)—the multiple pledge—took much longer. For a long time, the granting of the pledge would deprive the debtor of his or her power to dispose of the pledged asset. This may have been a relic from the past, when pignus—like fiducia—may also have been regarded as a form of ownership (Kaser 1982: 72–3). The functional similarity and analogous treatment of fiducia and pignus by the jurists may also have contributed to this (Krämer 2007: 370). During the course of classical law the inability to create multiple pledges over the same assets disappeared, under the influence of transactional practices (Kaser 1976: 201–5). In fact, one of the reasons why fiducia ultimately lost the struggle for life against pignus and disappeared from practice may have been that it never did enable a debtor and creditor deliberately to create lower-ranking security.

16.3.4.3. Rights of Use for Creditor: Antichresis, Rehypothecation, and Assignment of the Secured Debt Sometimes, the availability of credit may be enhanced by giving the creditor rights to use the charged assets. In Roman law such rights of use could take several forms. The creditor could use the (natural or civil) fruits of the charged property to discharge the secured debt or to cover the interest (antichresis). The creditor could also charge the collateral for his or her own debts (“rehypothecation”).²⁵ Also, when a creditor assigns a secured claim to another person, that other person (the assignee) should be able to enforce the security when the assigned debt is not properly discharged. Although the assignment of debts was problematic in Roman law, devices were used to achieve

²⁵ e.g. D. 36.4.5.21 (Ulp. 52 ad ed.): fruits; D. 44.3.14.3 (Scaev. quaest. publ. tract.): rehypothecation.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

128

Hendrik L. E. Verhagen

the same results, and in those situations the “assignee” was entitled to enforce the security.²⁶

16.3.5. Enforcing Security In many if not most legal systems, going to court is a relatively costly and time-consuming way for creditors to get what is due to them. However, precisely an effective law of real security enables a secured creditor to take enforcement measures without the need to go to court first (Jackson and Kronman 1979: 1143). Empirical studies indicate that efficient enforcement mechanisms lead to greater availability of credit, lower interest rates, and less collateral for equivalent levels of borrowing (Armour 2008: 19). Also, the opening of insolvency proceedings should not adversely affect the secured creditor’s ability to take recourse against the charged assets. At the same time, it should be born in mind that certain formalistic enforcement procedures, in particular public auctions, may have the positive effect of generating the highest possible price for the charged property. This may benefit not only the secured creditor, but also (in case of a surplus) lowerranking creditors or the debtor. Under Roman law the collateral would originally be forfeited to the creditor, who would become its owner. From that point on the creditor could treat the object as his or her own, even if its value exceeded the amount of the secured debt (Verhagen 2013a: 69–78).²⁷ Already for this early stage, we have epigraphic evidence of contracting parties expressly agreeing that the creditor would be authorized to sell the object of security in case of a payment default, in which case the creditor should turn over any surplus to the debtor, while the debtor would remain liable for any deficit.²⁸ In practice the creditor will usually have auctioned the charged property. In the first century  a well-organized auction practice existed, both for “ordinary” commercial sales and for enforcement sales (Noordraven 1999: 247; Lerouxel 2008: 189).²⁹ In the Sulpicii archive we find many ²⁶ e.g. D. 18.4.6 (Paul. 5 quaest.). ²⁷ The only difference was that in case of fiducia ownership would transfer immediately to the creditor, whereas in case of pignus this would take place upon the debtor’s default. ²⁸ e.g. TPN 69 (TPSulp 79): pignus; Formula Baetica, FIRA III, nr. 92: fiducia. ²⁹ The documents in the Sulpcii archive concerning auctions are reviewed by Noordraven (1999: 247–59). See also Romeo (2006), with further references.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

129

documents relating to the auction of assets (slaves, cloth, immovable property) that had been pledged or fiduciarily transferred.³⁰ We also know that there were many bankers (argentarii) in the Roman empire specializing (inter alia) in providing credit to bidders at auction (Andreau 1999: 38–9). The Sulpicii may have belonged to this category, which would mean that enforcing security by way of auction would be very familiar to them. For a long time, it was purely at the creditor’s discretion whether or not recourse would be taken against the charged assets by selling them (at auction or otherwise). Only at the end of the classical period had the creditor become obliged by operation of law to sell the object of security and turn over the surplus to the debtor. Court permission, however, was still not required for selling the charged assets and taking recourse against the proceeds of the collateral. Only forfeiture required the permission of the imperial chancery: impetratio dominii. This evolution from taking recourse by forfeiture to mandatory enforcement by way of sale can also be observed in other legal systems (Verhagen 2011: 11–13, 2013b: 152–9). As Moses Finley observed (1953: 266): “[i]n its earliest form, security is always substitution, a forfeit.” According to Finley the cause of this “profound economic transformation” lies outside the law of security and has everything to do with the growing importance of credit. Credit providers are not interested in obtaining ownership of the object of security as such: they want to liquidate it as soon as possible, so as to be able to take recourse against the proceeds in order to satisfy their claims. At the same time, debtors wish to lay their hands on the excess value of the collateral they provided. Thus, rules providing for a compulsory sale of collateral and the return of the superfluum (excess proceeds) to the debtor, as had developed in Rome by the end of the classical period (but have their origin in transactional practices from the early Principate), result in an equilibrium in which the interests of creditors and debtors are more or less balanced. Interestingly, forfeiture clauses, in the form of conditional sales of the charged property to the secured creditor, were still used at the end of the classical period, so apparently there was a need for them. They were legally valid, provided that the purchase price payable by the secured creditor to the debtor reflected the fair market value (iustum ³⁰ TPN 70, 71, 73, 74, 75, 76, 78, 79, and 80 (TPSulp 83, 84, 90, 91, 92, 85, 87, and 89).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

130

Hendrik L. E. Verhagen

pretium) of the property, so that when its value exceeded the amount of the secured debt the surplus would go to the debtor (D. 20.1.16.9, Marcian., ad form. hyp.; Verhagen 2013a: 73–7). Conditional sales of this type can indeed be an efficient means of taking recourse against charged assets, as they dispense with the need to auction the property soon after the debtor defaults, so that in a depressed market the creditor has the option to wait for better times. What at the beginning of the twenty-first century is regarded as “state of the art” in collateral enforcement was already possible in the Roman law of the second century !³¹

16.4. PUBLICITY

16.4.1. Publicity: Introduction The main criticism endorsed by modern Romanists on the Roman law of real security concerns its lack of publicity. Non-possessory security could be created over movable and immovable property, whether or not by way of a “floating charge,” without registration in a public register. In contrast to most developed modern countries, Rome did not have public mortgage registries for immovable property that could be consulted by prospective secured creditors in order to determine whether the debtor was the owner of the property and whether “limited” rights in rem (e.g. pledges, usufructs) had previously been created over the property. Moreover, again in contrast to modern legal systems, Roman law did not recognize a general principle allowing third parties in good faith to rely on possession of movable property.³² In other words, the adverse consequences of lack of publicity were not compensated for by rules protecting bona fide third parties. The oldest right in rem (nearly) always prevailed and could not be adversely affected by dispositions made without the consent of the oldest title holder (Verhagen 2014: 982). “Not ‘security of intercourse’ but security of the vested right was dear to the Roman heart” (Schulz 1936: 252). ³¹ See e.g. Directive 2002/47/EC of the European Parliament and of the Council of June 6, 2002, on financial collateral arrangements and revised (2006) article 2078 of the French civil code. ³² See however usucapio, discussed in §16.4.4.2 below.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

131

16.4.2. Registries Support Impersonal Exchange In a monograph on the economic and legal functions of property registries, Benito Arruñada (2012) sketches a dilemma concerning security interests that can be summarized as follows. The point of departure is that impersonal exchange expands the number of transactions that can be entered into and thus enhances economic growth (Arruñada 2012: 17). In secured lending transactions impersonal exchange means that the characteristics (reliability, solvency, etc.) of the debtor are not so much relevant as those of the charged assets. Granting the creditor a pledge, mortgage, or other form of security right in rem enables him or her to rely effectively on the charged assets. A proprietary security interest has a strong enforcement advantage, in that it allows the creditor to take recourse against the charged assets with preference over other creditors. The downside for commercial and financial intercourse of rights in rem, however, is that the other creditors run the risk of getting something of much less value than they expected. This may happen, for instance, when previously charged property is purchased by, or charged to, someone who is not aware of the charge. Should, on the other hand, the purchaser or the subsequent secured creditor be protected against a previously created right in rem, this would reduce the enforcement advantage for the original secured creditor. Property registries can solve this dilemma by providing verifiable information, so that the law can allow rights in rem (benefiting the original secured creditor) while at the same time enabling subsequent buyers or secured creditors to verify what the legal status of the property is.³³ However, in a particular (contemporary or ancient) society this will only be true when registries are technically feasible and reliable and can be used at relatively low costs.

16.4.3. Public Registers in the Roman Empire 16.4.3.1. Registration in the Roman Empire Douglass North (2010: 15) describes efficiency as “a condition in which given the state of technology and information costs, the market has the lowest production and transaction costs attainable.” ³³ For a distinct summary of the advantages of registration, see Arruñada (2012: 65).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

132

Hendrik L. E. Verhagen

According to North the “term is almost always used in relative rather than absolute terms.” Thus whereas in a digitalized world a public registration system for security rights may be efficient, it may very well be an inefficient system for the ancient Roman world. For immovable property operating a public registration system may generally be less problematic (although it may still be difficult) than for movable assets (e.g. stocks) and intangible assets (e.g. trade receivables). This may already have been the case in the ancient world: Roman Egypt did have registration systems for real estate (Taubenschlag 1955: 222–30; Wolff 1978: 184–262; Lerouxel 2012). In Egypt the Romans even introduced a new registration system for the most valuable assets (real estate and slaves): the “archive of acquisitions” (βιβλιοθήκη ἐγκτήσεων). Although the prime motivation for these changes may have been to control documentary practice with a view to taxes and liturgies, the archive clearly also had a publicity function for private parties. In particular, one concern of the Romans was to prevent making dispositions of land, houses, and slaves by persons without title (Von Woess 1924: 29). Moreover, the archive also recorded security interests and even fiscal preferences (protopraxia), so that potential purchasers or secured creditors could see whether the property had been previously charged (Von Woess 1924: 3–4).³⁴ Von Woess and more recently Lerouxel have stressed the high quality and accessibility of the archive (Von Woess 1924: 31; Lerouxel 2012: 967–8). Lerouxel (2015: 178) even considers the creation of the archive a “pivotal moment” in the history of the private credit market in Roman Egypt, causing a “tremendous growth” in the number of secured loans and the amounts loaned out by individuals. a. Registration in Roman Italy The registration of charges of real estate was not a phenomenon completely unknown to Roman Italy. When citizens entered into certain contracts with the Roman state or with local municipalities,

³⁴ A papyrus discovered in Oxyrhynchos records an ordinance dating from 89 by Mettius Rufus, prefect of Egypt. One of the provisions of this ordinance reads as follows: “ . . . Therefore I command that all owners shall register their property at the record-office within six months, and all lenders the mortgages that they hold, and other persons the claims which they possess.” FIRA I, nr. 60 (P.Oxy. 2, 237): translation taken from Bowman and Wilson (2009: 40).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

133

they might be required to grant a special security interest over real estate in favor of these bodies: the cautio praedibus praediisque. This form of security was created by a special declaration of the owner of the estate (subsignatio), which was registered in the records of the state (tabulae publicae) or of the municipality (tabulae communes). This form of (non-possessory) security was not a private law institution but rather a charge arising under public law (Kaser 1971: 459–60).³⁵ As Schulz (1951: 412–13) observed, “whenever we cross the boundary of private law we enter a new world: we find documents, registration and publicity, phenomena which were entirely foreign to classical private law.” The Roman practice of recording security interests primarily served to enforce the state’s claims rather than to publicize these interests to potential future acquirers or secured creditors. These observations also apply to another form of security granted over real estate to public bodies, recorded in the so-called tabulae alimentariae (Criniti 1991). In the second century the imperial treasury granted loans secured by a pledge over real estate, with the intention that the interest generated by these loans be for the benefit of children in need. The secured loans were recorded in a document, containing the names of the borrowers, descriptions of the real estate, the amounts of the loans, and the rates of interest, all of which was stored in public archives (Wenger 1953: 761–6). Moreover, there is evidence that already during the Principate Roman cities did have a cadaster, a registry whose purpose it was to demarcate individual plots of (often: public) land (Moatti 1993; Arruñada in this volume). For taxation purposes, registries of land in the census existed from the time of Augustus.³⁶ So it would have been technically possible for the Romans to create a registration system for immovable property, similar to the Egyptian archive of acquisitions. But apparently the Romans saw no need for it in Rome and elsewhere in the western part of the empire.

³⁵ This is undisputed for praedia granted to the Roman people, but less so for praedia granted to municipia: van Gessel (2003: 97). ³⁶ Arruñada (in this volume), with further references. That such registers could also be relevant in property disputes between private parties is demonstrated by D. 10.1.11 (Pap. 2 resp.), where it is observed that in boundary disputes, when there are no “old records” (vetera monumenta), one should follow the most recent registration by the tax authorities (census).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

134

Hendrik L. E. Verhagen

b. Why No Property Registries in Rome? Why did the Romans introduce a property registry in Egypt, but fail to do so in their home territory?³⁷ By way of conjecture I submit that the following factors could have played a role. In Egypt the Romans built upon existing registration systems set up and operated under the Ptolemies (Von Woess 1924: 28). The costs of building upon existing registration systems, including a body of professionals (notaries) experienced in executing and registering transaction documents, may be considerably lower than setting up an entirely new system.³⁸ Also the Romans’ desire to get a grip on property and commercial relationships in their new province may have played a role, in particular with a view to taxation.³⁹ Another reason is more of a psychological nature and may be that Roman citizens would consider it an unacceptable infringement upon their privacy if third parties could see whether their property had been charged for indebtedness (Von Woess 1924: 30).⁴⁰ A lesser degree of social cohesion in Roman Egypt than in Italy may also provide (part of ) the explanation. As will be discussed below, an important reason why the Roman law of property did not develop a principle protecting third parties in good faith is likely to have been that in the Republic and in the Principate social norms (amicitia, infamia) generally provided safeguards against a defrauding debtor who charged someone else’s property or failed to disclose an earlier charge. These social norms may not have been as stringently observed in Roman Egypt, where the population was largely non-Roman.⁴¹

³⁷ According to Arruñada (in this volume), the costs of archiving should not have been a serious obstacle for creating property registries. ³⁸ Lerouxel (2012: 958): “À l’époque romaine, l’Égypte est donc une grande terre de notariat et elle a atteint un haut degré de culture pratique de l’écrit.” ³⁹ It is not uncommon that institutions are introduced in colonial territories that are considered unnecessary in the dominating country. For instance, a land registration system was introduced in Ireland and the colonial possessions of the United Kingdom long before it was introduced in England and Wales: Watson (2001: 56). ⁴⁰ Arruñada (2012: 108–9) mentions that the failure of the Crown’s attempt to introduce property registers for land in England in 1581 and 1673 has been attributed to the opposition of the nobility, who did not want their debts to be known to the public. Other European states have similar experiences. ⁴¹ Lerouxel (2012: 946, 967–76) argues that the interaction between the private credit market and the finance of state expenditures may explain why the archive was created. These funding techniques were, however, not unique for Roman Egypt, as also Lerouxel notices.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

135

16.4.4. Informal Publicity through Possession 16.4.4.1. No Protection of Bona Fide Third Parties Two fundamental principles in the modern continental laws of property—which are of Roman origin—are the nemo plus-principle and the prior tempore-principle. The nemo plus-principle entails that a non-owner cannot transfer ownership or grant other rights in rem (e.g. pignus) to someone else.⁴² The prior tempore-principle entails that when two or more competing real rights exist over the same asset the oldest of these rights prevails.⁴³ In modern civil law jurisdictions the nemo plus-principle and the prior tempore-principle can—in respect of movable property—be overridden by a third principle of property law, what one could call the “legitimizing effect” of possession. A person who in good faith relied on the semblance of ownership reflected by his or her predecessor’s possession will be protected: possession legitimizes the other party’s belief that the possessor was entitled to dispose of the property. Applied to security interests, this principle provides that when a non-owning possessor creates a possessory security interest over movable assets in favor of a creditor who had no reason to believe that the debtor did not own the property, the creditor will acquire a valid and enforceable security interest.⁴⁴ Likewise, when the owner in possession grants a possessory security interest in favor of a creditor, that security interest will prevail in ranking over a previously created non-possessory security interest of which the creditor could not have been aware.⁴⁵ In Roman law all this is different. Here the nemo plus-principle and the prior temporeprinciple were not counterbalanced by a general principle allowing third parties in good faith to rely on their counterparty’s possession. There was therefore no general protection for the creditor who in good faith accepted security on property already charged in favor ⁴² The maxim nemo plus iuris ad alium transferre potest quam ipse haberet (“no one can transfer more in law to another than he himself has”) is as such articulated as a general principle of property law from the twelfth century onwards by the Glossators (see e.g. Glossa Ordinaria, gl. Nemo plus ad D. 50.17.54). For pignus, see e.g. D. 20.1.3.1 (Pap. 20 quaest.). ⁴³ Prior tempore potior iure (“earlier in time, stronger in law”). See Ant. C. 8.17.3 ( 212, on pignus of land). Another principle of property law is that rights in rem follow the property: “droit de suite.” For pignus see D. 13.7.18.2 (Paul. 29 ad ed.). ⁴⁴ e.g. §1207 German civil code (BGB); art. 3:238(1) Dutch civil code (BW). ⁴⁵ e.g. §1208 BGB; art. 3:238(2) BW. In both cases bona fide creditors will only be protected when their security interest is possessory.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

136

Hendrik L. E. Verhagen

of someone else or on property owned by someone else.⁴⁶ The oldest right in rem prevailed and could not be adversely affected by dispositions made without the consent of the oldest title holder (Verhagen 2014).

16.4.4.2. Usucapio and Possessory Interdicts Roman law did have an institution which could have evolved into such general principle protecting third parties in good faith: acquisitive prescription by way of usucapio. Already at a very early stage, however, the practically important category of stolen or embezzled property (res furtivae) was excluded from usucapio (Kaser 1971: 37). Moreover, usucapio could neither lead to the acquisition of a pledge created by a non-owner, nor could it protect a secured creditor against an earlier pledge unknown to him or her. Nevertheless, usucapio did fulfill a useful role in Roman society. Not only did it dispense with the need to transfer res mancipi by way of mancipatio, it also relieved the purchaser (or other acquirer) from the probatio diabolica (“diabolical proof”) of an unbroken chain of rightful acquisitions.⁴⁷ As such it could also benefit the creditor when he or she had to demonstrate that his or her right of pledge had been granted by a debtor owning the pledged property. To this it may be added that any possessor, even of stolen property, would have the possessory interdicts at his or her disposal, also against the owner, in order to act against threatening or actual interferences with his or her possession. The possessory interdicts were also available to the secured creditor whose factual possession of the pledged object was (or threatened to be) interfered with. The rules on possessio and usucapio thus reached a “compromise between the Roman dislike of interference with vested rights and the practical need to give some recognition to established facts” (Nicholas 1962: 124–5). ⁴⁶ Similarly a purchaser of previously charged property would not be protected. ⁴⁷ Already in preclassical law when the civilian (Quiritary) owner would transfer a res mancipi by way of traditio (informal transfer of possession) the transferee would after one (movable property) or two (immovable property) years acquire civilian ownership through usucapio. In the meantime the transferee could reclaim the property from every possessor with the actio Publiciana and ward off a rei vindicatio by the civilian owner with the exceptio rei venditae et traditae. This strong legal position is in modern literature (e.g. Jolowicz and Nicholas (1972: 99–100)) often referred to as “bonitary” ownership. See in particular Ankum and Pool (1989).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

137

16.4.4.3. Occasional Protection against Lack of Ownership In a limited number of cases Roman law did protect a creditor against the invalidity of a pledge created by a non-owner (Verhagen 2014). To this category belong, in the first place, texts dealing with a “relative” pledge. Wubbe (1960: 58–62, 92) has argued that whenever a non-owner in factual possession of the pledged asset would grant a pledge, the creditor could use the actio Serviana to recover the pledged assets from anyone possessing them, except from the full (pleno iure) owner, the bonitary owner, and their successors. A less radical position is that only a bona fide possessor, who was in the process of acquiring civilian ownership through usucapio, could create a relative right of pledge (Verhagen 2014: 986–8).⁴⁸ Secondly, a rescript by Septimius Severus and Caracalla from  205 (C. 8.15.2) holds that, although a pledge created without the permission of the owner is invalid, the situation is different when the owner knowingly allowed the pledge to be created. Fraudulent behavior in respect of an innocent creditor by the owner has the consequence that the invalidity of the pledge cannot be invoked against a pledgee in good faith (Ankum 2010: 37). Here Roman law comes close to results reached by modern rules on protecting bona fide secured creditors.

16.4.4.4. Purchasers of Pledged Property The position of the purchaser of pledged property in classical Roman law is that the secured creditor could always invoke the pledge even when the purchaser was in possession of the property and could reasonably have been unaware of the pledge. Obviously this legal rule could have seriously affected commercial intercourse in Rome. However, we have a large number of texts indicating that the secured creditor’s permission to the debtor to sell the pledged assets would enable the debtor to transfer the pledged assets unencumbered.⁴⁹ In other words, where the debtor was allowed by the secured creditor to sell goods in the ordinary course of business, commercial

⁴⁸ According to Ankum and Pool (1989: 32) it was even required that the debtor be a bonitary owner, i.e. someone to whom res mancipi had been conveyed by an owner. ⁴⁹ See in particular the fragments in D. 20.5, D. 20.6, and C. 8.25.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

138

Hendrik L. E. Verhagen

intercourse would not be disrupted. Moreover, Digest fragments on general pledges indicate that, although a practice existed of creating security over all one’s assets, such a security interest did not prejudice the interests of third parties dealing with the debtor in the course of business (Verhagen 2013a: 63–6). In other words, the general pledge may have created an equilibrium between, on the one hand, the secured creditor’s interest of having a preferential right of recourse against the assets comprised from time to time in the debtor’s estate, and, on the other hand, the debtor’s interest of being able to sell assets in the ordinary course of business and the debtor’s customers’ interest of acquiring them free of a security interest. This equilibrium is essentially the same as that reached by modern rules allowing third parties in good faith to rely on their counterparty’s possession.

16.4.5. Transactional Practices Entailing Publicity 16.4.5.1. Introduction In ancient Athens there was a practice of placing “mortgage stones” (horoi) on mortgaged real estate, indicating the name of the creditor and the amount secured by the mortgage (Fine 1951; Finley 1952, 1953: 249–68). Even this (simple) practice is not attested for Rome. But this is not to say that Roman legal practice did not have other practices informing third parties or otherwise reducing the risk that the secured creditor would be confronted with adverse claims by the real owner or another secured creditor. There were certain transactional practices which, in combination with the social and criminal sanctions discussed below, must have offered some compensation for lack of publicity and protection of good faith.

16.4.5.2. Mancipatio Under Roman law res mancipi were transferred as security to a creditor by way of fiducia cum creditore. The debtor would, through mancipatio, transfer land, slaves, or cattle as collateral to a creditor. As a consequence of the mancipatio, the creditor acquired dominium ex iure Quiritium (civilian ownership) of the object of fiducia. In the accompanying pactum fiduciae it would normally be expressly agreed that the creditor would retransfer the collateral if and when the secured debt was discharged. The mancipatio can be regarded as a

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

139

symbolic sale, even when it took place by way of security.⁵⁰ It required the presence of six Roman citizens: five witnesses and a sixth person, the libripens, who held a pair of scales. In the presence of these persons the transferee would (in case a slave was transferred) declare: “I assert that this man is mine by Quiritary right, and let him have been acquired to me with this piece of bronze and a bronze scale.”⁵¹ Subsequently, the transferee would strike the scales with the piece of bronze (aes) and give it to the transferor in payment of the “purchase price.” Although the original function of the witnesses may have been that of members of the “clan” or family expressing their consent to the transfer of property to which they were co-entitled, by the time of the Twelve Tables its main function must have been to create certainty as to the validity of the transfer of ownership of important capital goods (Kaser 1956: 120–1; Amunátegui Perelló 2012: 343). The large number of witnesses not only served to facilitate proof that the mancipatio had actually taken place, it also enhanced publicity in the preliminary stage (Kaser 1971: 42). Through the formal invitation of witnesses (rogatio) to the mancipatio, the intended transfer was made public, so that persons with adverse claims could come forward and object to the transfer, for example, by refusing to act as witnesses, by asking other members of the local community to do so, or by initiating legal proceedings (Kaser 1956: 120–2). Thus the publicity was also aimed at preventing transfers of property over which third parties exercised better rights. In other words, the presence of witnesses publicized the intended and the completed mancipatio. When the res mancipi was a slave, horse, or cattle the acquirer could put up an uncontested mancipatio as a defense against the accusation of theft (Kaser 1956: 134). Moreover, an uncontested mancipatio would make it easier for the acquirer to prove that he was the true owner of the property, especially when he could also

⁵⁰ Also in the case of fiducia cum creditore the “sale” was for a nominal purchase price of HS 1. See the Mancipatio Pompeiana, FIRA III, nr. 91 (singula sestertis nu[mmis sin]gulis) and Formula Baetica, FIRA III, nr. 92 (sestertio n(ummo) I). The transferor’s liability for eviction under the actio auctoritatis for double the purchase price was based on the mancipatio itself, so that a purchase price of HS I effectively exempted the transferor from liability. ⁵¹ Hunc ego hominem ex iure Quiritium meum esse aio, isque mihi emptus esto hoc aere aeneaque libra. Translation taken from Jolowicz and Nicholas (1972: 144).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

140

Hendrik L. E. Verhagen

prove a chain of earlier transfers by way of mancipatio (Kaser 1956: 134–6). As the mancipatio was, in contrast to the traditio, “abstract,” it was not necessary to prove that the earlier transfers had all been made pursuant to valid contacts or other legal transactions. The mancipatio may have served this publicity function well when Rome was a small agrarian community around the Palatine, but must have lost much of its effectiveness when Rome developed into a much larger community with hundreds of thousands of inhabitants.⁵² This may have been one of the reasons why—in the Principate— mancipatio was in practice often replaced by informal traditio and res mancipi came to be charged by way of the (informal) pignus rather than by way of fiducia cum creditore. Nevertheless, in the Principate, the presence of witnesses probably continued to have a publicity function, particularly when agricultural property (farms, arable land) was transferred in rural areas.⁵³ The formalities for the mancipatio itself were relaxed during the course of the classical period. According to Gaius the parties and witnesses did not actually need to be on the land when the mancipatio took place.⁵⁴ In the late classical period the oral formalities and the use of the scale were replaced with a document in which the mancipatio was recorded. There is, however, difference of opinion in modern literature as to whether the presence of witnesses was still required in the late classical period (Kaser 1975: 274). In any case, it is likely that in practice witnesses were used both in mancipatio and in an informal traditio whenever there were no other means of

⁵² Also the fact that the mancipatio could only be used by Roman citizens or citizens benefiting from commercium may have contributed to this: Sturm (1993: 349). Another factor contributing to the demise of mancipatio was that simpler legal arrangements, in particular the stipulatio duplae (undertaking to pay twice the amount of the purchase price), were available to make a seller liable in case of eviction: Sturm (1993: 354). ⁵³ Documents recording the conveyance of land usually contained the name of the plot of land, its extent, the municipality (and sometimes also: pagus and vicus) in which it was situated, the names of neighbors, natural features of the land, buildings on the land. See for instance Tabula Veleia (Criniti 1991) and Formula Baetica (FIRA III, nr. 92). ⁵⁴ Gai., Inst. 1.121. In postclassical law this changed. At the beginning of the fourth century  Emperor Constantine issued a long constitution which required that the sale and transfer of land should take place on site (reversing Gai., Inst. 1.121) and in the presence of neighbors. See Fragmenta Vaticana 35: Honoré (1989: 142–9); Voss (1982: 135–77).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

141

establishing the boundaries of the transferred land.⁵⁵ In case of provincial land (which was res nec mancipi) this was common practice and this is likely to also have been the case with land in Italy.⁵⁶

16.4.5.3. Documents The fact that in classical Roman law writing was rarely prescribed as a legal formality does not mean that in practice many transactions were not recorded in written documents and witnessed. As the archive of the Sulpicii and other epigraphic and papyrological evidence demonstrate, the creation of security (fiducia or pignus) was often evidenced in writing in the style of chirographa or testationes.⁵⁷ Writing served as evidence for the granting of a pledge (D. 20.1.4, Gai., ad form. hyp.). This was particularly relevant for non-possessory pledges (hypothecae), which were otherwise more difficult to prove than possessory pledges. Rather than using the relatively inexpensive medium of papyrus, the parties even chose to record their transactions in wooden writing tablets (tabulae ceratae), to which the seals of witnesses were attached. The sealed tablets included the date on which the pledge was created and sometimes an express warranty that the security was validly created and first-ranking.⁵⁸ Obviously the use of documents did not guarantee that the creditor could enforce his or her pledge as first-ranking creditor. There are Digest fragments discussing cases in which documents were lost (D. 44.2.30.1, Paul. 14 quaest.), not dated (D. 20.1.34.1, Scaev. 27 dig.), wrongly dated (D. 48.10.28, Mod. 4 resp.), or deliberately antedated by the parties in order to defraud other creditors (D. 22.4.3, Paul. 3 resp.; Wacke 1969: 399–400).

⁵⁵ The main function of witnesses in the classical period may been to demarcate the plot of land that was actually transferred (e.g. Jav., D. 18.1.63.1, Iav. 7 ex Cassio): Voss (1982: 158–64). ⁵⁶ D. 18.1.35.8 (Gai. 10 ad ed. prov.); D. 18.1.63.1 (Iav. 7 ex Cassio). Honoré (1989: 139) observes that the traditio of provincial land was influenced by the mancipatio. ⁵⁷ On the use of chirographa, testationes, and tabulae ceratae, see Verhagen (2018: 251–62). ⁵⁸ See TPN 40, 43, 44, and 69 (TPSulp 55, 51, 52, and 79). For an express warranty see e.g. FIRA III, nr. 91 (Mancipatio Pompeiana) and D. 20.1.34.1 (Scaev. 27 dig.).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

142

Hendrik L. E. Verhagen

16.4.5.4. Handing-Over Deed of Sale An interesting fragment from the second-century jurist Scaevola refers to a debtor’s handing over the deed of sale (pursuant to which he had acquired the real estate) to the secured creditor (D. 13.7.43 pr., Scaev. 5 dig.).⁵⁹ Several explanations have been offered for this practice. It has been argued that the practice demonstrates the difficulties in departing from the principle that only possessory pledges could be created (Wagner 1968: 59–60). This is not very likely, as by the time of Scaevola the pledge created by mere agreement (nuda conventione) was firmly established. A more convincing explanation is that handing over the deed of sale had always been customary, but had never been mentioned by the jurists as they were not particularly interested in these practical matters (Wagner 1968: 60). Scaevola’s writings, however, are an exception and show a great interest in actual transaction practices, so that may explain why the practice is first attested precisely in one of his opinions. There is also a rescript by Septimius Severus and Caracalla (C. 8.16.2,  207, when the concept of a non-possessory pledge was at least a century old) on a case in which the deed of sale of land had been handed over to the secured creditor. The rescript states that when someone has pledged the deed of sale of his land he must have intended to pledge the land itself. The creditor’s possession of the deed of sale certainly could have practical advantages. As the Severan rescript demonstrates it could serve as evidence of the pledge. It could make a prospective secured creditor or purchaser suspicious if the seller would be unable to produce the deed of sale. Moreover, possession of the deed of sale could enable the creditor to sell the property more easily when the debtor was in default. Also, as D. 19.1.48 (Scaev. 2 resp.) demonstrates, documents concerning real estate were used in order to determine which rights the seller’s predecessor had over the property sold: this could also be relevant for the creditor who took security over real estate.⁶⁰ The disadvantages of this practice are that the ⁵⁹ See also Sev./Ant. C. 8.16.2 ( 207) and D. 19.1.48 (Scaev. 2 resp.). On this practice in other times and places, see Arruñada (2012: 46–7). ⁶⁰ Titius had inherited the fundus Sempronianus from Sempronius and sold it to Septicius. The contract of sale included the following clause: “Whatever rights Sempronius had in the Sempronian farm are yours by purchase for such and such a number of coins.” The question was asked to the jurist Scaevola whether purchaser

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

143

secured creditor could (without good reason) impede a sale and that it could prevent multiple security interests being created over the same property when there was sufficient “equity” in the property to do so (Arruñada 2012: 46).

16.4.6. Reputation-Based Mechanisms and Criminal Liability As D. 20.1.15.2 (Gai., de form. hyp.) suggests (“the danger run by those who create multiple charges’”), it can reasonably be assumed that loss of credibility, civil liability, and infamia must have had a strong deterrent effect against pledging someone else’s property or the failure to disclose already existing charges.⁶¹ Particularly for Romans belonging to the higher social classes, infamia could have serious repercussions: it (inter alia) entailed the exclusion from important offices (Krämer 2007: 367–8). For a long time the combination of social norms, legal rules, and transactional practices must have worked reasonably well to remedy the lack of publicity and of protection of bona fide third parties. The Romans must have been aware—in the first century —that the registry they created for Egypt had been very beneficial for the availability of (secured) credit. Is it too speculative to conclude from this that they would surely have set up similar registries in the city of Rome if their own law of real security was, because of lack of registries, so poorly adapted to the Septicius could institute the action on purchase (actio empti) against seller Titius, in order to force him to show what rights Sempronius had on the basis of records relating to the inheritance and to point out the boundaries. Scaevola answered that decisive is what the parties are considered to have intended with the clause cited above. But if this is unclear, good faith entails that seller Titius must produce the farm’s records (instrumenta fundi) as well as point out its boundaries. Interestingly, Papinian observes in D. 10.1.11 (2 resp.) that in boundary disputes, when there are no “old records” (vetera monumenta), one should follow the most recent registration by the tax authorities (census). ⁶¹ When one reads D. 20.1.15.2 (periculum, quod solent pati qui saepius easdem res obligant) one would think that the “danger” (periculum) referred to would have been criminal liability (stellionatus). However, the prevailing view seems to be that stellionatus only became a crime at the end of the second century . Before that time the “danger” must have been that of being sued with the actio de dolo (Kaser 1976: 183). A condemnation on the basis of the actio de dolo would lead to infamia. See also D. 20.1.34.1 (Scaev. 27 dig.), in which the debtor declares that the creditor must have given credence to the debtor’s assurance “as man of honor” that the pledge was firstranking.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

144

Hendrik L. E. Verhagen

needs of the financial world? It was only by the end of the second century  that imperial rescripts stated that the fraudulent nondisclosure of earlier pledges should be criminally prosecuted.⁶² Knowingly pledging someone else’s property or property that had already been charged constituted a special crime (crimen stellionatus).⁶³ The maximum punishment was severe: forced labor for plebeians and banishment or exclusion from their order for members of the higher classes.⁶⁴ It must be assumed that this criminal liability served to enhance security of transactions and that it took place in response to abuses that threatened commercial intercourse (Wacke 1969: 400; Bürge 1979: 128). It must be doubted, however, whether these criminal sanctions were always an effective remedy to ensure a properly functioning law of secured credit.⁶⁵ The uncertain legal status of pledged assets, which despite the newly introduced criminal sanction, apparently remained a problem, is likely to have contributed in the late classical period to the increasing practice of excluding liability for eviction in contracts entered into by creditors enforcing their rights of pledge by selling the pledged assets to third parties.⁶⁶ In that case the purchaser could sue the debtor with an adapted version of the action which a purchaser would normally have against the seller (actio empti utilis; D. 20.5.12.1., Tryph. 8 disp.). A creditor liable for eviction could assign to the purchaser his or her contractual action arising against the debtor under the pledge agreement. In both scenarios the

⁶² See D. 13.7.36 pr. (Ulp. 11 ad ed.), referring to the imperial rescripts; D. 13.7.16.1 (Paul. 29 ad ed.), expressly mentioning pledging someone else’s property and pledging property already pledged to someone else; Philip. C. 9.34.4 pr. ( 244). When a valuable asset had been pledged for a small debt, granting a second pledge without disclosing the first one was not a crimen stellionatus (D. 13.7.36.1, Ulp. 11 ad ed.). ⁶³ The pledgor would not be criminally liable if he or she had disclosed the earlier pledge to the second pledgee: D. 20.1.15.2 (Gai., de form. hyp., discussed above, p. 143). ⁶⁴ Philip. C. 9.34.4 pr. ( 244); D. 13.7.36.1 (Ulp. 11 ad ed.); D. 47.20.3.2 (Ulp. 8 de off. proc.). ⁶⁵ As Arruñada (2012: 51) observes with respect to the need to rely on criminal law: “the need to rely on such harsh mechanisms for enforcement provides a glimpse of the opportunity costs caused by lack of proper institutions for conveyancing land and securing credit.” ⁶⁶ D. 21.2.68 pr. (Pap. 11 resp.); D. 17.1.59.4 (Paul. 4 resp.). As with so many transactional practices, these clauses may have evolved into a rule of law. From Alex. C. 8.45.1 ( 223) it appears that the secured creditor is not liable for eviction, unless he or she has expressly agreed otherwise.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

145

purchaser could directly sue the debtor who had pledged someone else’s property. This would, however, be of little benefit when the debtor was insolvent, and for this reason the possibility of eviction may have had a depressing effect on prices paid for pledged assets (Bürge 1979: 130–1).⁶⁷ From D. 13.7.22.4 (Ulp. 30 ad ed.) it appears that in the late classical period it became a customary practice for creditors selling pledged property to enter into a stipulatio duplae with purchasers, this promising twice the amount of the purchase price in case of eviction. This may have had a positive effect on sales prices realized by creditors enforcing their pledge. At the same time this transactional practice increased risks for secured creditors, which may have resulted in increased interest rates for borrowers.⁶⁸ Criminal liability, stipulationes duplae for eviction, granting remedies for eviction to the creditor or the purchaser of pledged property, these are all symptoms of serious disruptions of the credit markets in the late classical period resulting from the lack of publicity and of other forms of protection against pledges created over someone else’s property or previously charged property. More generally, it seems that the uncertain legal status of land was becoming more and more problematic. It is therefore not surprising that in late antiquity measures were introduced to remedy this.

16.4.7. Late Antiquity: Publicly Executed Documents and gesta municipalia In classical law many juridical acts (e.g. mancipatio, stipulatio) required that certain words were to be used by the parties executing them. These juridical acts derived their legal validity from these formalities. Already during the course of classical law this perception ⁶⁷ Liability for eviction could be a reliable alternative for the secured creditor when creditworthy third parties are liable, such as title insurers (Arruñada 2012: 51). I have not carried out extensive research on this, but, although via a stipulatio or fideiussio (or other form of personal security) such liability was technically possible, it seems unlikely that there was an established Roman practice of title insurance. ⁶⁸ In the classical period the stipulatio duplae was broadly used in the context of ordinary sale. In a number of fragments additional remedies are granted to the executing pledge-creditor who entered into a stipulatio duplae: D. 13.7.8.1 (Pomp. 35 ad Sab.) (right of retention); D. 13.7.22.4 (Ulp. 30 ad ed.) and D. 13.7.23 (Tryph. 8 disp.) (right of recourse against debtor). See also D. 13.7.24 (Ulp. 30 ad ed.): the creditor has an actio emptio utilis when evicted after impetratio dominii.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

146

Hendrik L. E. Verhagen

changed and the will of the parties came to be regarded as the cause of the legal consequences of a juridical act. The consequence of this was that the old oral formalities were replaced with new forms, which now only served to enhance legal certainty. The new forms were mainly (witnessed) documents and the cooperation of public authorities (Kaser 1975: 73–4). We have already briefly looked at the first form (written mancipatio in late classical law), and will now focus on the second form, which eventually resulted in an imperial constitution on the ranking of pledges. In the late Roman period private documents were often deposited with local archives (gesta municipalia), particularly in the western Roman Empire. A protocol was prepared describing the essentials of the transaction recorded in the deposited document or repeating the text of the registered document verbatim. The deposited document and the protocol were stored in an archive, while the parties received copies (Hirschfeld 1904: 63–4). Direct evidence of gesta records is to be found in the Ravenna papyri, dating from after the fall of the western Roman Empire.⁶⁹ An extremely interesting example, which appears to follow a procedure of (late) Roman origin (Tjäder 1955: 286), is a protocol of  489, concerning a donation by the (first) Germanic king of Italy, Odoacer. This document gives us a detailed picture of how title deeds were registered in late antiquity and the early Middle Ages. The donation by Odoacer was drawn up by a notary of the royal chancery (notarius regiae sedis) in Ravenna. The text of the donation was recited and copied into the gesta protocol executed in Ravenna and deposited with the gesta muncipalia of that city. Both the notarial donation deed and the Ravenna gesta protocol were also registered with the gesta municipalia of Syracuse (Sicily), where the land was situated and where possession of it was transferred to the donee.⁷⁰ Subsequently the name of the former owner was replaced with that of the new owner (the donee) in the tax registry (a polypthicis ⁶⁹ The authoritative edition (with translations in German) of the Ravenna papyri (P.Ital.) is Tjäder 1955 (P.Ital. 1–28) and Tjäder 1982 (P.Ital. 29–59). In one document, a letter, reference is made to a mortgage (fiducia) of land. P.Ital. 1, Tjäder (1955: 168–78). ⁷⁰ P.Ital. 10–11, Tjäder (1955: 279–93). The donation was to a Roman supporter of Odoacer, Pierius, who had acted as a military commander under the king. Pierius died a year later at the battle of Adda ( 490) in which Odoacer was defeated by Theoderic the Great. Tjäder (1955: 283–4). Another interesting example of a protocol accompanied with a “letter of transfer” (epistula traditionis), executed in Ravenna ( 540), is P.Ital. 31, Tjäder (1982: 62–71).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

147

publicis).⁷¹ Although the core function of the gesta municipalia was fiscal (i.e. to record the real estate and transactions relating thereto with a view to taxation), they were also voluntarily used by private individuals to authenticate and give some form of publicity to their transactions.⁷² In particular, registering a deed with public authorities would enhance the evidentiary force of the document.⁷³ From Constantinople we even have an enactment by Emperor Leo I from  472 (C. 8.17.11) specifically dealing with the ranking of pledges created by “publicly executed instruments” (instrumentis publice confectis). According to this constitution “ . . . if anyone seeks to claim for himself the right of pledge or hypothec under such documents, the one who uses publicly executed instruments is preferred even if there is a later date on them . . . ”⁷⁴ In other words, a publicly executed deed of pledge takes precedence over a privately executed one. This constitution is very important for our subject: for the first time in the history of Roman law the “public” execution of a document fundamentally affects the ranking of a right of pledge.

16.4.8. Comparative Perspective 16.4.8.1. Registration From a comparative perspective it is important to note that classical Roman law was not unique in not having title registers for immovable

⁷¹ According to Hirschfeld (1904: 21) the registration was in a property registry, which not only listed the actual owners of the property, but also the holders of limited rights in rem (hypothecs, usufructs, etc.). According to Tjäder (1955: 442, with further references) the registry was a tax register, registering for each plot of land the name of the owner, the extent of the property and the amount of taxes to be paid. ⁷² Brown (2013: 102). The functioning of the gesta municipalia in the late Roman period is discussed at p. 95–102. See also Hirschfeld (1904: 33–49). See also Everett (2013: 63–94), who criticizes the view that the primary purpose of the gesta was the creation and storage of authoritative documents (2013: 72), but later discusses several instances where the parties clearly had that purpose in mind (2013: 79, 80, and 82). Moreover, as far as I am aware, nobody has suggested that this was the primary purpose of Constantine and other Roman emperors for organizing gesta muncipalia. ⁷³ Kaser (1975: 80–1). This (secondary) function of the gesta muncipalia was rather that of “recordation of documents” than that of “registration of rights.” For this distinction, see Arruñada (2012: 55–67). ⁷⁴ Sin autem ius pignoris vel hypothecae ex huiusmodi instrumentis vindicare quis sibi contenderit, eum qui instrumentis publice confectis nititur praeponi, etiamsi posterior dies his contineatur. Translation Frier (2016).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

148

Hendrik L. E. Verhagen

property. In respect of land, houses, and other immovable property, registration as a general institution is a relatively recent phenomenon, even for modern countries with developed economies. The function of informing prospective secured lenders was carried out by other mechanisms, such as a network of local notarial archives in sixteenthcentury and seventeenth-century Paris (Hoffman, Postel-Vinay, and Rosenthal 2000: 20–1). French, English and other modern experiences show that developed credit markets can exist without registration systems and that setting-up and operating such registries is not without difficulties.⁷⁵ As far as movable property (e.g. stocks, inventory, machinery) and receivables are concerned, there are widely divergent views as to the desirability of having public registration systems in contemporary jurisdictions.⁷⁶ In the United States in particular, it is regarded essential to establish a public general security rights registry for recording notices about the possible existence of a security right.⁷⁷ In the great majority of civil law jurisdictions (e.g. Germany, France, Italy, Netherlands) there are no public filing systems for assets other than immovable property.⁷⁸ In any case, where in contemporary jurisdictions a registration system for movable property and receivables is at least technically feasible, in ancient Rome this must have been—given the state of technology—much more difficult to achieve. The exception was slaves, the ownership and charge of whom could be registered in the archive of acquisitions of Roman Egypt. However, slaves must generally have been easier to identify with a certain estate or household than other movable assets. All this demonstrates that it would be too simplistic to discredit the ⁷⁵ According to Arruñada (2012: 233), “governments have struggled for almost ten centuries to organise reliable registries . . . .” Even today few countries have succeeded in making property registries fully functional. ⁷⁶ Arruñada (2012: 84–6) does not appear to recommend registration of property rights on movable property. ⁷⁷ Under Article 9 of the Uniform Commercial Code a security interest can be “perfected,” so that it can be invoked against competing claimants and insolvency officials, in several ways, such as by possession or control or by public filing. ⁷⁸ The original draft for the 1992 Dutch Civil Code did provide for a nonpossessory pledge which needed to be registered in a public register. The banking and trade community opposed such publicly registered security interest. Under the present Civil Code, a non-possessory pledge can be created either by notarial (or other “authentic”) deed or by registering the deed of pledge with the Dutch tax authorities in a non-public register (article 3:237 Civil Code). In 2018 Belgium introduced the registration in a public register of non-possessory pledges over movable assets (title XVII, Book III, Belgian Civil Code). See Dirix (2013).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

149

Roman law of real security simply because it lacked registries for security interests.

16.4.8.2. Protection of Bona Fide Secured Creditors More remarkable is that in respect of movable property Roman law did not develop a general principle protecting bona fide persons against lack of title or already existing charges. However, even this feature of Roman law can be put in perspective by a comparison with modern law. In contemporary market economies contractual claims are valuable assets and therefore an important category of collateral: receivables-based financing is of immense importance for many companies.⁷⁹ Nevertheless, as far as the protection against multiple dispositions or unauthorized dispositions of claims is concerned, in many (if not most) civilian jurisdictions, the situation is exactly the same as with movable property in classical Roman law. There is neither a public filing system for receivables nor are there rules on the protection of bona fide purchasers or chargees of receivables.⁸⁰ In the credit market this is no obstacle for taking receivables as collateral, as there are sufficient non-legal incentives to prevent a debtor from creating security over claims that have already been charged (Flessner 2008: 336–49). These non-legal incentives may, at least for many centuries, even have been stronger in ancient Rome.

16.5. CONCLUDING OBSERVATIONS The Roman law of security satisfied nearly all the requirements for an effective law of secured transactions. Secured creditors would in principle be able to take recourse against the charged assets with preference over the other creditors. Roman law offered a single and ⁷⁹ “In developed economies the bulk of corporate wealth is locked up in debts,” Oditah (1991: vii). ⁸⁰ See, however, art. 1690 section 3 and art. 2075 of the Belgian Civil Code. For English law, see Dearle vs. Hall (1828) 3 Russ 1. In these jurisdictions notification fulfills a function for the protection of bona fide assignees against earlier assignments: the assignee who has first given notice of the assignment to the debtor has priority. Oditah considers it a very poor form of publicity to make the debtor of the assigned claim “a kind of public register,” Oditah (1991: 140).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

150

Hendrik L. E. Verhagen

informal method for creating security rights rather than a multiplicity of methods for different types of charged asset. Security could be granted by a wide range of debtors to a wide range of creditors and could cover a wide range of debts (including future debts). Rights of use could be granted to the debtor (non-possessory and multiple pledges) and to the creditor (antichresis, rehypothecation, and assignment), thus preventing the collateral from remaining “dead capital.” The rules on enforcement allowed creditors to take recourse against the charged assets by selling them to third parties without the need to obtain a court’s permission. The Achilles’ heel of the Roman law of security was its lack of publicity for security created over real estate and its lack of protection of bona fide third parties in respect of more liquid assets. Schulz (1936: 251) observed that “under the narrower and more easily controllable condition of life in ancient days there was no urgent need for the principle of publicity and the protection of a person acquiring a thing in good faith.” In respect of security interests in particular, Schulz writes that “hypothecary law did not, in those times, play the same part in economic life as it does today and that therefore the absence of protection for a creditor accepting security in good faith was less noticeable.” Although Schulz was usually highly perceptive of the interactions between law and economy in Rome, his view on real security may need revision. Already in the first century  the city of Rome itself had several hundred thousand inhabitants. In other words, here the “narrower and more easily controllable condition of life” no longer existed. Second, the increasing sophistication of the law of real security in the classical period of Roman law during the Principate coincided with a time of economic growth. The law seems to have responded to the demands of the economy, by allowing non-possessory pledges, pledges of claims, general pledges, multiple pledges, et cetera. Roman law thus did offer a legal framework for secured credit which would be versatile enough for a market economy in which secured finance plays an important role. If that is so, one wonders how likely it is that real security was as insignificant as Schulz (and others) profess it to have been. For a long time the Roman economy seems to have done fairly well without registries for security interests on immovable property and protection of bona fide creditors taking security over movable property. The smooth operation of the Roman (secured) credit market may only have become seriously disrupted by the end of the classical

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

151

period. It may not be a coincidence that only by the end of the classical period the fraudulent non-disclosure of earlier pledges came to be criminally prosecuted and that contractual practices emerged dealing with the creditor’s liability when selling the charged property. In other words, (only) by the end of the classical period the system of secured credit may have become considerably less effective.⁸¹

REFERENCES Amunátegui Perelló, Carlos F. 2012. “Problems concerning mancipatio.” 80 Tijdschrift voor Rechtsgeschiedenis 329–52. Andreau, Jean. 1999. Banking and Business in the Roman World. Cambridge: Cambridge University Press. Ankum, Hans, and Eric Pool. 1989. “Rem in bonis meis esse and rem in bonis meam esse: Traces of the Development of Roman Double Ownership,” in Peter Birks, ed., New Perspectives in the Roman Law of Property. Oxford: Oxford University Press, 5–41. Ankum, Hans. 2010. “Spätklassische Problemfälle bezüglich der Verpfändung einer res aliena,” in H. Altemeppen et al., eds, Festschrift für Rolf Knütel zum 70. Geburtstag. Heidelberg: Müller, 35–44. Armour, John. 2008. “The Law and Economics Debate about Secured Lending: Lessons for European Lawmaking?,” in Horst Eidenmüller and Eva-Maria Kieninger, eds, The Future of Secured Credit in Europe. European Company and Financial Law Review, Special Volume 2. Berlin: De Gruyter, 3–29. Arruñada, Benito. 2012. Institutional Foundations of Impersonal Exchange: The Theory and Policy of Contractual Registries. Chicago: University of Chicago Press. Brown, Warren C. 2013. The gesta municipalia and the public validation of documents in Frankish Europe,” in Warren C. Brown et al., eds, Documentary Culture and the Laity in the Early Middle Ages. Cambridge: Cambridge University Press, 95–151. Buckland, W. W. 1963. A Text-Book of Roman Law: From Augustus to Justinian, 3rd edn, revised by Peter Stein. Cambridge: Cambridge University Press. Bürge, Alfons. 1979. Retentio im römischen Sachen- und Obligationenrecht. Zürich: Schulthess. ⁸¹ The author thanks Benito Arruñada and Giuseppe Dari-Mattiacci for their helpful comments. Some of the descriptions of the law of pignus and fiducia in this chapter are adaptations from parts of Verhagen (2013a, 2013b, and 2014).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

152

Hendrik L. E. Verhagen

Camodeca, Giuseppe. 1999. Tabulae Pompeianae Sulpiciorum, Edizione critica dell’archivio puteolano dei Sulpicii. Rome: Quasar. Chemain, J.-F. 2015. L’argent des autres: le cautionnement dans le monde romain du IIe siècle avant J.-C. au Ier siècle après J.-C. Versailles: Via Romana. Criniti, Nicola. 1991. La tabula alimentaria di Veleia: introduzione storica, edizione critica, traduzione, indici onomastici e toponimici, bibliografia veleiate. Parma: Deputazione di Storia Patria per le Province Parmensi. Crook, J. A. 1996. “Legal history and general history.” 41 Bulletin of the Institute of Classical Studies 31–6. Deakin, Simon, and Fabio Carvalho. 2011. “System and evolution in corporate governance,” in Peer Zumbansen and Gralf-Peter Callies, eds, Law, Economics and Evolutionary Theory. Cheltenham/Northhampton: Elgar, 111–30. Dirix, Eric. 2013. De hervorming van de roerende zakelijke zekerheden. Mechelen: Kluwer. Everett, Nicholas. 2013. “Lay documents and archives in early medieval Spain and Italy, c.400–700,” in Warren C. Brown et al., eds, Documentary Culture and the Laity in the Early Middle Ages. Cambridge: Cambridge University Press, 63–94. Fine, John V. A. 1951. Horoi: Studies in Mortgage, Real Security, and Land Tenure in Ancient Athens. Athens: American School of Classical Studies at Athens. Finley, Moses I. 1952. Studies in Land and Credit in Ancient Athens, 500–200 : The Horos-Inscriptions. New Brunswick, NJ: Rutgers University Press. Finley, Moses I. 1953. “Land, debt, and the man of property in classical Athens.” 68 Political Science Quarterly 249–68. Finley, Moses I. 1999 (1973). The Ancient Economy, updated with a foreword by Ian Morris. Berkeley: University of California Press. Fleisig, Heywood. 2008. “The Economics of Collateral and of Collateral Reform,” in Frederique Dahan and John Simpson, eds, Secured Transactions Reform and Access to Credit. Northampton, MA: Edward Elgar, 81–109. Flessner, Axel. 2008. “Security interests in receivables—A European perspective,” in Horst Eidenmüller and Eva-Maria Kieninger, eds, The Future of Secured Credit in Europe, European Company and Financial Law Review, Special Volume 2. Berlin: De Gruyter, 336–49. Frier, Bruce W. 2000. “Roman law’s descent into history.” 13 Journal of Roman Archaeology 446–8. Frier. Bruce W. 2016. The Codex of Justinian: A New Annotated Translation, with Parallel Latin and Greek Text, 3 vols. Cambridge: Cambridge University Press. Gordon, W. M., and O. F. Robinson. 1997. The Institutes of Gaius, 2nd edn. London: Duckworth.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

153

Gröschler, Peter. 2008. “Die Mittel der Kreditsicherung in den tabulae ceratae,” in K. Verboven et al., eds, Pistoi dia tèn technèn: Bankers, Loans and Archives in the Ancient World, Studies in Honour of Raymond Bogaert. Louvain: Peeters, 301–19. Harris, Ron. 2003. “The encounters of economic history and legal history.” 21 Law and History Review 297–346. Harris, William V. 2006. “A revisionist view of Roman money.” 96 Journal of Roman Studies 1–24. Hirschfeld, Bruno. 1904. Die Gesta municipalia in römischer und frühgermanischer Zeit, Marburg: [no publisher] 33–49. Hoffman, Philip T., Gilles Postel-Vinay, and Jean-Laurent Rosenthal. 2000. Priceless Markets: The Political Economy of Credit in Paris, 1660–1870. Chicago: University of Chicago Press. Honoré. Tony. 1989. “Conveyances of land and professional standards in the later Empire,” in Peter Birks, ed., New Perspectives in the Roman Law of Property. Oxford: Oxford University Press, 137–52. Jackson, Thomas H., and Anthony T. Kronman. 1979. “Secured financing and priorities.” 88 Yale Law Review 1143–82. Johnston, David. 1999. Roman Law in Context. Cambridge: Cambridge University Press. Jolowicz, H. F., and Barry Nicholas. 1972. Historical Introduction to the Study of Roman Law, 3rd edn. Cambridge: Cambridge University Press. Kaser, Max. 1971. Das römische Privatrecht, Erster Abschnitt: Das altrömische, das vorklassische und klassische Recht, 2nd edn. Munich: Beck. Kaser, Max. 1975. Das römische Privatrecht, Zweiter Abschnitt: Die nachklassischen Entwicklungen, 2nd edn. Munich: Beck. Kaser, Max. 1976. Ausgewählte Schriften, vol. II. Napels: Jovene. Kaser, Max. 1982. Studien zum römischen Pfandrecht. Napels: Jovene. Kaser, Max, and Karl Hackl. 1996. Das römische Zivilprozessrecht, 2nd edn. Munich: Beck. Kaser, Max, Rolf Knütel, and Sebastian Lohsse. 2017. Römisches Privatrecht: Ein Studienbuch, 2nd edn. Munich: Beck. Kay, Philip. 2014. Rome’s Economic Revolution. Oxford: Oxford University Press. Krämer, Gerd. 2007. Das besitzlose Pfandrecht: Entwicklungen in der römischen Republik und im frühen Prinzipat. Cologne/Vienna: Böhlau. Kunkel, Wolfgang. 1973. “Hypothesen zur Geschichte des römischen Pfandrechts.” 90 Zeitschrift der Savigny-Stiftung für Rechtsgeschichte: Romanistische Abteilung 15–170. Lerouxel, François. 2012. “Le marché du crédit privé, la bibliothèque des acquêts et les tâches publiques en Égypte romaine.” 67 Annales HSS 943–76.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

154

Hendrik L. E. Verhagen

Lerouxel, François. 2015. “The βιβλιοθήκη ἐγκτήσεων and transaction costs in the credit market of Roman Egypt (30 ... to ca. 170 ..),” in Dennis P. Kehoe, David M. Ratzan, and Uri Yiftach, eds, Law and Transaction Costs in the Ancient Economy. Ann Arbor: University of Michigan Press, 162–84. Luhmann, Niklas. 2004 (1993). Law as a Social System. Oxford: Oxford University Press (orig.: Das Recht der Gesellschaft, Frankurt am Main: Suhrkamp). McCall, Brian. 2009. “It’s just secured credit: a natural law case in support of some forms of secured credit.” 43 Indiana Law Review 7–44. McCormack, Gerard. 2004. Secured Credit under English and American Law. Cambridge: Cambridge University Press. Mentxaka, Rosa. 1986. La pigneracion de colectividades en el derecho romano clasico. Bilbao: Servicio Editorial de la Universidad del País Vasco. Moatti, Claude. 1993. Archives et partage de la terre dans le monde Romain (IIe siècle avant—Ier siècle après J.-C.). Rome: École française de Rome. Noordraven, Bert. 1999. Die Fiduzia im römischen Recht. Amsterdam: Gieben. Nicholas, Barry. 1962 [repr. 2010]. An Introduction to Roman Law. Oxford: Oxford University Press. North, Douglass C. 2010. Understanding the Process of Economic Change. Princeton/Oxford: Princeton University Press. Oditah, Fidelis. 1991. Legal Aspects of Receivables Financing. London: Sweet & Maxwell. Roe, Mark J. 1996. “Chaos and evolution in law and economics.” 109 Harvard Law Review 641–68. Romeo, Stefania. 2006. “Fiducia auctionibus vendunda nelle Tabelle Pompeiane. Procedure e modalità di redazione delle testationes nelle avctiones puteolane del 61 d.C.” Polis. Studi interdisciplinari sul mondo antico 207–58. Röver, Jan-Hendrik. 1999. Vergleichende Prinzipien dinglicher Sicherheiten. Munich: Beck. Röver, Jan-Hendrik. 2007. Secured Lending in Eastern Europe: Comparative Law of Secured Transactions and the EBRD Model Law. Oxford: Oxford University Press. Schäfer, Hans-Bernd. 2008. “The law and economics debate about secured lending: lessons for European lawmaking. Commentary?,” in Horst Eidenmüller and Eva-Maria Kieninger, eds, The Future of Secured Credit in Europe. European Company and Financial Law Review, Special Volume 2. Berlin: De Gruyter, 30–5. Scheidel, Walter. 2010. “Economy and quality of life,” in Alessandro Barchiesi and Walter Scheidel, eds, The Oxford Handbook of Roman Studies. Oxford: Oxford University Press, 593–609. Schulz, Fritz. 1936. Principles of Roman Law. Oxford: Clarendon Press. Schulz, Fritz. 1951 [repr. 1969]. Classical Roman Law. Oxford: Clarendon Press.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Secured Transactions in Classical Roman Law

155

Schwartz, Alan. 1984. “The continuing puzzle of secured debt.” 37 Vanderbilt Law Review, 1051–69. Schiavone, Aldo. 2000 (1996). The End of the Past: Ancient Rome and the Modern West. Cambridge, MA/London: Harvard University Press (orig.: La storia spezzata: Rome antica e Occidente modern. Rome/Bari: Laterza). Stiglitz, Joseph E., and Andrew Weiss. 1981. “Credit rationing in markets with imperfect information.” 71 American Economic Review 393–410. Sturm, Fritz. 1993. “Das Absterben der mancipatio,” in Stephan Buchholz et al., eds, Überlieferung, Bewahrung und Gestaltung in der rechtsgeschichtlichen Forschung. Paderborn (etc.): Schöningh. Taubenschlag, Raphael. 1955. The Law of Greco-Roman Egypt in the Light of the Papyri. 332 –640 , 2nd edn. Warsaw: PNW. Terpstra, Taco. 2008. “Roman law, transaction costs and the Roman economy,” in Koenraad Verboven et al., eds, Pistoi dia tèn technèn: Bankers, Loans and Archives in the Ancient World, Studies in Honour of Raymond Bogaert. Louvain: Peeters, 345–69. Terpstra, Taco. 2013. Trading Communities in the Roman World, A MicroEconomic and Institutional Perspective. Leiden/Boston. Tjäder, Jan Olof. 1955. Die nichtliterarischen lateinischen Papyri Italiens aus der Zeit 445–700, I. Papyri 1–28. Acta Instituti Romani Regni Sueciae, Series in 4o, XIX:1. Lund: Gleerup. Tjäder, Jan Olof. 1982. Die nichtliterarischen lateinischen Papyri Italiens aus der Zeit 445–700, II. Papyri 29–59. Acta Instituti Romani Regni Sueciae, Series in 4o, XIX:2. Stockholm: Paul Åström. Van Gessel, Christian. 2003. “Praedes, praedia, cognitores: les suretés réelles et personelles de l’adjudicataire du contrat public en droit romain (textes et réflections),” in Jean-Jacques Aubert, ed., Tâches publiques et entreprise privée dans le monde romain. Actes du Diplôme d’Etudes Avancées, Universités de Neuchâtel et de Lausanne, 2000–2002. Geneve: Droz, 95–122. Verboven, Koen. 2002. The Economy of Friends: Economic Aspects of Amicitia and Patronage in the Late Republic. Brussels: Latomus. Verhagen, Hendrik L. E. 2011. “Das Verfallpfand im frühklassischen römischen Recht.” 79 Tijdschrift voor Rechtsgeschiedenis 1–46. Verhagen, Hendrik L. E. 2013a. “The evolution of pignus in classical Roman law.” 81 Tijdschrift voor rechtsgeschiedenis 51–79. Verhagen, Hendrik L. E. 2013b. “Ius honorarium, equity and real security: parallel lines of legal development,” in W. J. Zwalve and E. Koops, eds, Law and Equity: Approaches in Roman Law and Common Law. Den Haag/ Boston: Martinus Nijhoff, 129–60. Verhagen, Hendrik L. E. 2014. “The (lack of) protection of bona fide pledgees in classical Roman law,” in Rena van den Bergh et al., eds, Meditationes de iure et historia: Essays in Honour of Laurens Winkel. Fundamina. Editio specialis 2014 (20–2). Pretoria: Unisa Press, 982–92.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

156

Hendrik L. E. Verhagen

Verhagen, Hendrik L. E. 2014. “Chirographs in classical Roman law: constitutive or probative?” 65 Revue Internationale des Droits de l’Antiquité 251–306. Voss, Wulf Eckart. 1982. Recht und Rhetorik in den Kaisergesetzen der Spätantike: Eine Untersuchung zum nachklassischen Kauf- und Übereignungsrecht. Forschungen zur Byzantinischen Rechtsgeschichte 9. Frankfurt am Main: Löwenklau. Wacke, Andreas. 1969. “Prozessformel und Beweislast im Pfandrechtsprätendentenstreit.” 37 Tijdschrift voor Rechtsgeschiedenis 369–414. Wagner, Herbert. 1968. Voraussetzungen, Vorstufen und Anfänge der römischen Generalverpfändung. Marburg: Elwert. Wagner, Herbert. 1974. Die Entwicklung der Legalhypotheken am Schuldnervermögen im römischen Zeit (bis zur Zeit Diokletians). Cologne/Vienna: Böhlau. Watson, Alan. 2001. Society and Legal Change, 2nd edn. Philadelphia: Temple University Press. Wenger, Leopold. 1953. Die Quellen des römischen Rechts. Vienna: Holzhausen. Wieacker, Franz. 1988. Römische Rechtsgeschichte. Quellenkunde, Rechtsbildung, Jurisprudenz und Rechtsliteratur, Erster Abschnitt. Munich: Beck. Woess, Friedrich von. 1924. Untersuchungen über das Urkundenwesen und den Publizitätsschutz im römischen Ägypten. Munich: Beck. Wolf, Joseph Georg. 2012. Neue Rechtsurkunden aus Pompeji, Tabulae Pompeianae Novae, 2nd edn. Darmstadt: Wissenschaftlichte Buchgesellschaft. Wolff, Hans Julius. 1978. Das Recht der griechischen Papyri Ägyptens in der Zeit der Ptolemaeer und des Prinzipat, Bd. 2: Organisation und Kontrolle des privaten Rechtsverkehrs. Munich: Beck. Wubbe, Felix B. J. 1960. Res aliena pignori data: de verpanding van andermans zaak in het klassieke Romeinse recht. Leiden: Universitaire Pers. Zimmermann, Reinhard. 1996 [1990]. The Law of Obligations: Roman Foundations of the Civilian Tradition. Oxford: Oxford University Press. Zumbansen, Peer, and Gralf-Peter Callies, eds. 2011. Law, Economics and Evolutionary Theory. Cheltenham/Northhampton: Elgar.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

17 Ancient Rome Legal Foundations of the Growth of an Indispensable City Robert C. Ellickson

“The ancients themselves were firm in their view that civilized life was thinkable only in and because of cities.” M. I. Finley (1977: 305) “For the upper-class Roman, it would seem, the provincial towns did not exist: Rome’s prestige held them, as London and Paris hold similar groups today. To live well, he must dwell in Rome . . . .” Lewis Mumford (1961: 210)

In 200 , the city of Rome had a population of about 200,000 (Morley 1996: 39). Two hundred years later, the city’s population had leapt to roughly 1,000,000, the largest urban agglomeration the world had yet seen.¹ Almost two millennia would elapse before Europe would again witness equivalent urban expansions, those of London and Paris around 1800. ¹ Temin (2013: 31) summarizes the varying estimates of ancient Rome’s population. At the time of the early Empire, about one-seventh of the peninsula of Italy’s population resided in Rome (ibid.: 259). During civil wars and other periods of sociopolitical instability, the rise in the city’s population may have paused or reversed (Turchin and Scheidel 2009). In  1500, Rome’s population numbered roughly 50,000 (Jongman 1988: 73). At that time, the combined populations of the twelve largest cities of Christian Europe totaled about 1,000,000, the equivalent of ancient Rome’s population at its peak (Scheidel 2007: 78). Robert C. Ellickson, Ancient Rome: Legal Foundations of the Growth of an Indispensable City In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0017

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

160

Robert C. Ellickson

The explosive growth of Rome during and after the late Republic poses two major questions. The first is whether the city’s emergence as a major metropolis benefited other residents of the Empire, or instead, as Moses Finley (1999) and others have contended, was parasitic on the resources of outsiders. I contribute to this longstanding debate by applying the theory of cities that Edward Glaeser (1998) and other urban economists have developed. This body of theory helps identify the many sorts of agglomeration benefits that ancient Rome provided to persons residing outside the city. I include an analysis of the origins and career destinations of forty-one renowned Romans, among them poets, historians, and jurists. This small dataset strongly supports the thesis that many of Rome’s most talented inhabitants were born elsewhere and migrated to the city. They saw the immensity of Rome as advantageous because it enabled them to specialize and to exchange ideas with peers. Interactions among these talented people generated information spillovers that benefited not only themselves, but also outsiders who consumed the many services that the capital exported. Far from being parasitic, the city of Rome in its prime contributed mightily to the relative prosperity of the Mediterranean region.² My second central undertaking is to identify the institutional underpinnings that enabled the city of Rome to grow as rapidly as it did. Land was the principal form of ancient wealth (Crook 1967: 147; Finley 1999: 188). I contend that lawmakers in both the Republic and early Empire generally resolved key issues affecting urban development more sagaciously than had lawmakers of other notable regimes in the ancient world. Like most of the prominent states that predated it, Roman lawmakers authorized private property in land. The Romans, however, were exceptionally willing to permit a landowner to alienate real estate by sale, lease, or testamentary disposition.³ For example, the Twelve Tables, legislation adopted

² Support for the geographic breadth of prosperity in the Roman world can be found in Engels (1990: 125–6); Ward-Perkins (2005: 87–100); Lo Cascio (2007: 619–22); and Temin (2013: 220–61). Benefits hardly were universal. During their conquests, the Romans killed and enslaved many. ³ The phrase real estate challenges Finley’s (1999) vision that transactions in the ancient world differed in kind from those in a mature commercial nation. Ancient Rome unquestionably lacked many of the institutions present in a twenty-firstcentury market economy, including specialized real estate brokers and large-scale mortgage lenders. See n. 63. But market transactions were pervasive (Temin 2013).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 161 soon after the founding of the Roman Republic, affirmed unfettered freedom of testation, an extraordinary legal principle in a neararchaic society. The Romans’ willingness to put land tenure in motion helped unleash the energies of the host of private entrepreneurs who developed and managed the apartment blocks of the expanding city. The decisions that determined the scope of the public sector in the city of Rome also were relatively sound. Public officials provided public works such as forums and aqueducts that were essential to Rome’s growth. But these officials also limited the scope of government in ways that enhanced the city’s prosperity. They declined to provide private goods, such as housing, that the city’s entrepreneurs could more efficiently supply, and learned to resist political pressures for growth-inhibiting populist enactments, such as rent controls and edicts cancelling debts. I do not contend that the legal policies just identified were sufficient to assure the rise of the city of Rome. They surely were not. But in the absence of this cluster of enabling policies, Rome would have remained a backwater.

17.1. THE CITYSCAPE OF ANCIENT ROME Especially for the benefit of non-classicists, I begin with a brief description of Rome’s demography and landscape during its prime, defined here as the conveniently symmetric period between 200  and  200. During that stretch, Rome was, for better or worse, a city in a class by itself. Zipf ’s Law, an intriguing nugget of urban theory, predicts that the population of a nation’s largest city will be roughly twice the population of its second largest city, three times the population of its third largest, and so on (Gabaix 1999). Around  150, the populations of the cities throughout the Roman empire comported well with Zipf ’s Law. After Rome, a city with about 1,000,000 inhabitants, came Alexandria, with about 500,000, and then Carthage, with 300,000 (Wilson 2011: 181–6). But within the boundaries of present-day mainland Italy, the second most populous city in  150 A “capitalistic” economy can be defined as one in which decentralized actors routinely exchange goods and services for money. By this definition, Rome’s land and housing markets were capitalistic.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

162

Robert C. Ellickson

is thought to have been Ostia, Rome’s seaport. Ostia’s population then was a mere 40,000, a twenty-fifth of Rome’s (Wilson 2011: 182).⁴ On the peninsula of Italy itself, the city of Rome dwarfed all rivals.⁵

17.1.1. Private Buildings and the Entrepreneurs Who Built and Owned Them During the late Republic and early Empire, curule and plebeian aediles and other public officials supervised the provision of public works such as the streets and aqueducts essential to Rome’s growth. There is no indication, however, that officials of either the late Republic or early Empire ever sponsored the building of housing projects (Frier 1980: 25; Robinson 1992: 4; see also Frier 1977: 36).⁶ Private entrepreneurs, whose names are mostly lost in history, produced and managed the apartment blocks, lodging houses, and shops that made up the bulk of the city’s buildings. The best evidence about Roman housing comes from archaeological remains in Ostia and the Vesuvian cities (Pirson 1997). Apartment blocks built in Ostia during the first century , in particular, are thought to have been patterned after structures in Rome (Packer 1971: 63; Garnsey 1976: 129; Frier 1980: 3–6, 14). During the early Empire, the typical private housing development in Rome, at least for upscale tenants,⁷ was an apartment block (insula)⁸ on the order of four to six stories in height.⁹ An apartment block commonly included shops (tabernae) on the ground floor fronting on the street, and apartments (cenacula) on the floors above. ⁴ Morley (1996: 182) provides a somewhat different set of estimates of the populations of ancient cities on the Italian peninsula. ⁵ When analyzing the growth of urban giants, Ades and Glaeser (1995: 216–18) selected ancient Rome as one of their case studies. ⁶ By forfeit or escheat, however, a domus or insula might fall into government ownership (Frier 1980: 25). ⁷ On the existence and size of a Roman “middle class,” see Mayer (2012). Perhaps 6 to 12 percent of the Roman Empire’s population could be so categorized (ibid.: 21). For more discussion, see Scheidel and Friesen (2009). ⁸ The proper translation of insula is disputed (Storey 2002). I adopt the interpretation in Frier (1980). Dubouloz (2011), reviewed by Machado (2012), unconventionally defines an insula not as a type of building but as a property that was owned indirectly. ⁹ Frier (1980: 14–16) describes the Casa di Via Giulio Romano, the insula in Rome that has best survived.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 163 Over the centuries, improvements in building technology raised the average quality of Rome’s apartment blocks. Among these advances were the refinements of Roman concrete (opus caementicium) during the late Republic, and fired brick during the first two centuries of the Empire (Schneider 2007: 162–3). These materials enabled construction of taller, stronger, and more fire-resistant apartment blocks, commonly ones graced with arches and vaults.¹⁰ Within an apartment, however, rafters and other structural elements commonly were of timber (Robinson 1992: 34–5). Textual sources indicate that tenants in Rome’s apartment blocks were frequently endangered by fire and collapse (Crook 1967: 154). An apartment on an intermediate floor of an upscale insula might have an area of 150 to 300 square meters (Frier 1980: 6). This typically would have accommodated a household of six to eight occupants, counting slaves¹¹ and other hangers-on (Frier 1980: 16). Upper apartments were reached by way of staircases, commonly external (Pirson 1997: 177–8). An apartment on a lower floor might be serviced by gravity-fed piped water (Robinson 1992: 99, 119–20; see also Beard (2008: 64), on Pompeii), and perhaps a toilet seat situated over a chamber pot or pipe into a wall (Beard 2008: 107). Most of Rome’s relatively poor households—perhaps as much as 90 percent of a city’s free population (Frier 1980: 5)—found housing in one of three ways.¹² As noted, the ground floor of an insula fronting a well-traveled street included specialized shops.¹³ The artisans and traders—commonly slaves or ex-slaves—who worked in these tiny shops commonly slept in a back room or on an elevated mezzanine (Pirson 1997: 173).

¹⁰ Vitruvius estimated that a concrete tenement house in Cicero’s day would have had a lifespan of eighty years, an expected duration shorter than that of a brick structure (Garnsey 1976: 129). ¹¹ The percentage of the population enslaved likely was higher in Rome than elsewhere, and perhaps peaked at the outset of the Empire at between 25 percent and 40 percent. For estimates, see Morley (1996: 39); Scheidel (2005); and Temin (2013: 135–6). See generally Koops in this volume. ¹² A sub-middle-class household was far less likely to include slaves and ex-slaves, and might have had five to six occupants on average (Bowman and Wilson 2011b: 12). On the residences of the poor of Pompeii, see Beard (2008: 105–6). Wallace-Hadrill (1994: 75–9) reports that the distribution of dwelling sizes in Pompeii was broad and not bimodal. ¹³ Holleran (2012: 99–158) describes the various services that these shops offered.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

164

Robert C. Ellickson

Second, a landlord of an apartment block whose lower floors were designed for prosperous tenants might subdivide the topmost stories into small rooms or cubicles that poor households could afford.¹⁴ These quarters, although perhaps enjoying better light and view, lacked piped water and required a long climb. Residents of upper floors also typically bore the greatest risks of both fumes wafting and fires spreading from the apartments below. Because landlords tended to market apartments on upper floors to the less wealthy, in Rome there commonly was a vertical sorting of tenants by income.¹⁵ Households in Rome also were horizontally sorted (Yavetz 1958: 505–6).¹⁶ The poorest neighborhoods contained disproportionate concentrations of tenements and lodging houses (known as deversoria, among other terms). The owner of a lodging house offered short-term rentals of squalid quarters and perhaps also meals in a common room (Frier 1977: 31–5; Frier 1980: 27–9, 51; Scobie 1986: 402). Although archaeological evidence is sparse, some lodging houses may have been made with cheaper materials such as mud brick and wood (Pirson 1997: 95). Public latrines provided an essential service. The wealthiest Romans, less than 1 percent of the population, sought the distinction of owning and residing in a domus, a singlehousehold abode that typically was spacious, luxuriously appointed, and situated on a large lot.¹⁷ A domus household might include dozens, or even hundreds, of slaves, ex-slaves, and other hangers-on (Saller 2007: 107). During the late Republic, some members of the upper crust clustered their houses on Palatine Hill, close to the jewelers, goldsmiths, and pearl dealers on Via Sacra (Engels 1990: 243; Holleran 2012: 55). Rome’s neighborhoods likely were less stratified by social class than those of a twenty-first-century city. Because the richest households

¹⁴ An apartment builder indeed might use wooden materials, not masonry, to frame the top floor or two (Robinson 1992: 37). ¹⁵ Compare the vertical stratification characteristic of the luxurious Hausmannian apartment houses of late nineteenth-century Paris, in which tiny rooms for chambermaids were provided on the top floor. ¹⁶ Frier (1980: 35) refers to Argiletum and Aventine as “lower-class districts” of Rome. ¹⁷ Treggiari (1999) depicts Cicero’s emotional attachments to his domus. A census of Rome at the end of the third century  counted 1,790 domus structures and 44,300 insulae (Benevolo 1980: 176).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 165 included numerous slaves and ex-slaves, pedestrians on Via Sacra likely would have been socially diverse. And residents of upscale apartments commonly lived in close proximity to the ground-floor shops that housed low-status artisans and traders.¹⁸ Little is known about the entrepreneurs, designers, contractors, financiers, and construction workers who together erected Rome’s private buildings.¹⁹ Because senators tended to regard conspicuous involvement in real estate development as undignified (Finley 1999: 41–2, 60–1), most subdividers and apartment builders likely were drawn from a somewhat lower social rank, perhaps relatively wealthy freedmen. Cicero’s letters mention two real estate operators, C. Sergius Orata and Damasippus, the latter a subdivider of land into lots (Rawson 1976: 100–2).²⁰ Plutarch’s Parallel Lives includes a depiction of M. Licinius Crassus, an exceptionally rich Roman famed for his military, political, and real estate exploits. Crassus’ adulthood spanned the first half of the first century . Writing in Greece almost two centuries later, Plutarch asserted that Crassus built nothing except his own house, but eventually acquired “the largest part of Rome” from sellers panicked by fires breaking out in their apartment blocks.²¹ Plutarch surely exaggerated. Even the wealthiest Roman lacked the resources to buy a major fraction of the private real estate in the city, and had no feasible means of managing thousands of apartment blocks. There were no large business enterprises in Rome (Crook 1967: 229), partly because Roman law provided no workable mechanism for the creation of a private business entity that would have been attractive to capital investors (Fleckner in volume I; Hansmann et al. in volume I). Ownership of Rome’s many thousands of apartment blocks likely was distributed widely among individuals and small enterprises. The wealthiest residents of Rome tended to invest not only in rural property, but also in urban real estate (Frier 1980: 21–6; Parkins 1997; Rosenstein 2008). Cicero,

¹⁸ On the paucity of horizontal stratification in Pompeii, a much smaller city, see Wallace-Hadrill (1991: 261–4); Beard (2008: 62); and Mayer (2012: 51–8). ¹⁹ Kehoe (2012: 123–4) provides a brief discussion of builders and construction workers, and Hawkins (2012: 180–1), of specialty subcontractors. A well-established profession of land surveyors existed; see Campbell 2000; and Libecap and Lueck in this volume, but most of the textual evidence depicts their activities in rural areas, not cities. ²⁰ Rawson draws on Cicero, Ad Atticum 12.33. ²¹ Plut., Crass. 2.4–2.5; for discussion, see Frier (1980: 32–4).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

166

Robert C. Ellickson

whom Finley (1999: 41–2) cites as contemptuous of urban commerce, himself owned several apartment blocks (Garnsey 1976: 126–7). Many housing complexes would have been in the hands of either individuals who had inherited their ownership interests, or guardians managing inheritances for wards. Patronage networks supported by gift exchange pervaded many aspects of Roman society.²² These helped foster trust among longdistance traders, for example. The owner of an apartment block similarly could take advantage of a patron-client relationship when managing the building. The owner might rely on a former slave, for instance, to handle relations with tenants.²³ A patron-builder similarly might prefer to choose a client, rather than a more socially distant provider, as the architect or contractor. In a dense city such as Rome, however, the reputational concerns of a repeat player would have partially allayed a builder’s worries about the trustworthiness of an agent who was not a client.

17.1.2. Density During the reign of Augustus, Rome is estimated to have had about five hundred residents per hectare (Morley 1996: 38). This was roughly twice the density of Manhattan’s population in 2010. In Rome, both people and buildings were packed together. Streets were narrow. By the early Empire, most of Rome’s streets had a width of about 4.5–5.0 meters between building lines (Chevallier 1976: 67).²⁴ An apartment block typically abutted its sidewalk and covered most of its lot. In most neighborhoods, narrow streets lined with shopfronts threaded through masses of hulking apartment blocks bursting with humanity. Pedestrians had to share a street right-of-way with, among others, peddlers of goods, guiders of pack-animals, riders on horseback, and drovers of carts drawn by oxen (Finley 1999: 126). A staple of urban economic theory holds that land values tend to decline with distance from a city’s center (Alonso 1964).²⁵ As this ²² See nn. 35–7 and accompanying text. ²³ See text accompanying nn. 99–100. ²⁴ Beard (2008: 58) states that street pavements in Pompeii typically were less than 3 m wide. ²⁵ Frier (1980: 23 n. 7) marshals some of the few extant sources on land prices in ancient Rome.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 167 theory predicts, rents in central Rome seem to have been astronomic.²⁶ The price of agricultural land also was roughly ten times higher near the city than elsewhere in Italy (Engels 1990: 25–6), further evidence of the value of proximity to the capital. Rome’s epicenter was the venerable Forum, which evolved into a collection of adjoining forums. Storey’s (2002: 420–1) study of fourth-century  census data suggests that residential densities were highest near the Forum and the adjoining Palatine Hill. During the period 200  to  200, Rome had a single defensive perimeter, the Servian Wall, which dated from the fourth century . A Roman living inside that wall would have resided no more than two kilometers from the Forum, a half-hour’s walk at most. The territory within roughly a day’s walk (30–7 km) of the Forum comprised Rome’s suburban district (Goodman 2007: 20, 50). Partly because the Servian Wall, despite its sixteen gates, impeded access, gradients of both land value and population density likely dropped markedly beyond that wall. Nonetheless, the decision to build the Aurelean Wall in  270–5 at a farther-outlying location suggests that the area just beyond the Servian wall had already become heavily urbanized with apartment structures.²⁷

17.1.3. Unpleasant Aspects of Life in Rome Life in the capital city was exasperating and unhealthful in many respects. Scobie (1986) provides a particularly dystopian assessment.²⁸ Roman literary figures, most notably Juvenal in Satura III (The Sixteen Satires 1974: 87–98), stress the downsides of city life and the superiority of rustic environments. Juvenal’s mouthpiece is Umbricius, a man who had failed to prosper in Rome and who rants about negative aspects of the city. Some current-day interpreters, overlooking Juvenal’s satirical approach, have been too quick to take Umbricius’ laments literally (Braund 1989). Even allowing for Umbricius’ exaggerations, the environment in Rome was indeed challenging in many respects. For starters, it was a

²⁶ Frier (1980: 43) refers to the “enormously high rents at Rome,” an assertion consistent with Juvenal (1974: 93). ²⁷ Angel (2012: 93) provides a map that demarcates both walls. ²⁸ Laurence (1997) argues that Scobie exaggerates.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

168

Robert C. Ellickson

noisy place.²⁹ Late into the evening, from the streets echoed shouts of draymen (Juvenal 1974: 95) and the clatter of iron-clad wheels of carts on stone pavements (Beard 2008: 80). Because acoustical insulation within and between apartment blocks was mediocre, a wouldbe sleeper might have to put up with the thumps of footsteps, the cries of babies, and the barking of the ubiquitous pet dogs.³⁰ Rome also was a smelly city. Latrines and other systems for disposal of excrement and urine, both human and animal, were primitive at best (Scobie 1986: 407–22; Robinson 1992: 119–22). A pedestrian risked being splattered by a dump of household waste, perhaps from a chamber pot, from a window above (Scobie 1986: 417; Beard 2008: 55).³¹ These putrid materials, when added to the wastes of draft animals, could make a pedestrian’s walk along a street an unpleasant trek (Beard 2008: 56). The remains of Pompeii suggest that Rome’s aediles might have arranged for installation of stepping stones in streets to provide a path above the slop (Beard 2008: 54). Braziers, oil lamps, and kitchen fires, all commonplace in apartments (ibid. 92), emitted fumes and posed risks of fire. Rome suffered as many as one hundred building fires per day, twenty of them large (Robinson 1992: 108) Worst of all, Rome was a sickly environment, especially for young children. Because infectious diseases were rampant, life expectancies were shorter in Rome and other cities than in the countryside (Scheidel 2007: 39). Given all these distasteful aspects of life in Rome, how could the city’s population growth have been so explosive after 200 ? Why would the head of a free household have ever chosen to migrate to the capital?

²⁹ Martial, Epigrammata, Book 12, poem 57, proclaims the impossibility of a good night’s sleep in Rome. In Epistles 56, Seneca complains of noise from a bathing establishment. ³⁰ Even poor urban households commonly kept dogs as pets (Scobie 1986: 418–19). ³¹ A dump of wastes could give rise to legal liability. “The praetor says . . . ‘If anything should be thrown out or poured out from a building onto a place where people commonly pass and repass or stand about, I will grant an action to be brought against whoever lives there for double the damage caused or done as a result’ ” D. 9.3.1 pr. (Ulp. 23 ad ed.). See also D. 9.3.1.10 (Ulp. 23 ad ed.): “If a number of people occupy a lodging house and something is thrown down from it, action may be brought against any one of them.” (Throughout this chapter, references to the Justinian Digest are to the Watson ed. (1985) version.)

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 169 17.2. WHY TALENTED PEOPLE MOVED TO ROME, AND HOW THEIR PRESENCE THERE BENEFITED OUTSIDERS To individuals and households who were free to decide where to live—as slaves obviously were not—Rome offered countervailing benefits that no other city in the eventual western empire could match. This section identifies these benefits, and, more controversially, argues that the city of Rome also provided valuable services to people throughout the empire. In spirit, my analysis echoes both Donald Engels’s (1990) assessment of ancient Corinth’s role as a service city, and many of the essays in Bowman and Wilson (2011). But, unlike the authors of those works, I explicitly draw on the theory of cities that Glaeser and other urban economists have refined since the 1990s. This theory seeks to explain why people are willing to put up with the high costs and hassles of city life, and why, throughout history, cities have tended to be centers of innovation, as ancient Rome surely was.

17.2.1. The Once Fashionable Conception of Rome as a Consumer City In several works first published in the 1970s, the deservedly eminent classicist Moses Finley (1977, 1999) depicted Rome as a “consumer city”—a “parasite” of residents of rural areas.³² Finley’s conception of the Roman economy is commonly characterized as “primitivist,” a position contrasted with the “modernist” views of his critics. Three of the premises underlying Finley’s analysis are uncontroversial. The first is the truism that, on balance, a city inhabited by people owning a disproportionate share of their polity’s financial assets and real estate will be a net importer of foodstuffs and manufactured goods. The phrase consumer city captures this truism.³³ Second, no one disputes the fact that rich Romans disproportionately lived in Rome. The members of the senatorial class, the wealthiest of all, were centered ³² Max Weber and Karl Polanyi influenced Finley’s thinking. Finley applied his thesis to ancient cities generally, not just to Rome. ³³ Glaeser (2011: 259–60), in a book that cites neither Finley nor Weber, describes contemporary London, New York and Paris as “consumer cities”—that is, places that offer services and entertainments that attract the wealthy.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

170

Robert C. Ellickson

in the capital, and indeed the senate was required to convene there. Third, it is uncontested that the city of Rome lacked large-scale manufacturing facilities (Finley 1999: 23). The city thus could hardly have counterbalanced its massive imports of food solely with exports of manufactured wares.³⁴ Much more controversial, however, has been Finley’s (1999: 35–61) assertion of a profound difference between the motivations of ancient and modern traders. He admitted that transfers of land and houses were “relatively frequent” during the time of Cicero (Finley 1976b: 4).³⁵ But he contended that parties negotiating these transactions, especially when both were members of the social elite, would have been more concerned about attaining honor and status than material wealth. A land sale or apartment lease between two wealthy parties thus would not have been an arm’s length bargain about the division of the monetary gains from trade, but rather an episode in an ongoing process of gift exchange embedded within a larger social network.³⁶ To Finley, the ancient economy was qualitatively different, not just quantitatively different, from a modern economy (Morris 1999: introduction). Many of Finley’s critics (e.g. Wallace-Hadrill 1991; Temin 2013: 4–11) reject this vision.³⁷ Finley himself was well aware that members of Rome’s upper crust frequently acted materialistically (Jongman 1988: 34–5). Land typically was sold for cash and at arm’s length (Frederiksen 1975: 168). In all known urban leases, the rent due was stated in money (Frier 1980: 60). Conversely, gift exchange is hardly absent in a current-day market economy. On the order of one-third of US GDP currently is produced within households, much of it through non-market processes (Ellickson 2006: 231–2). And inhabitants of a

³⁴ There is some evidence of the mass manufacture of pottery at sites other than Rome (Ward-Perkins 2005: 96–100; Peña 2007: 32). Classicists agree, however, that most goods were manufactured either in households or small shops (Hawkins 2012). ³⁵ Rawson (1976: 86) cites a handful of examples of land parcels, urban and rural, that were sold four or more times within a fifty-year period. ³⁶ Kirschenbaum (1987: 171–96) provides examples of transactions among intimates, some involving friends of Cicero and Pliny the Younger. On the operation of reputational constraints within these sorts of social networks, see Temin (2013: 110–13). ³⁷ On both Finley and his critics, see Morris (1999). Erdkamp (2001), who generally supports the consumer-city thesis, has also summarized the debate. Duffy and Puzzello (2014) invoke theory and laboratory evidence to evaluate, as a general matter, the relative merits of gift exchange and monetary exchange.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 171 modern market economy remain acutely status-conscious (McAdams 1997: 355–8). Thorstein Veblen would have recognized Cicero. Even if members of Rome’s upper crust had been qualitatively different from contemporary traders—to a modernist, a doubtful proposition—they constituted no more than one percent of the city’s population (cf. Engels 1990: 123–4). Would honor and status have been central when a person of average means was making a quotidian arrangement with a barber, butcher, or fuller? Might not the parties in these instances have generally sought to buy low and sell high? And would not market forces have tended to weed out suppliers indifferent to covering their costs?

17.2.2. The Agglomeration Efficiencies That Rome Generated Even if ancient plutocrats were to have been primarily motivated by considerations of honor, the conclusion does not follow that Rome was a “complete parasite-city” that sucked resources out the rest of the Empire (Finley 1999: 130).³⁸ Imagine that after 300  the city of Rome had never grown, and also that no other major city had emerged on the Italian peninsula. How would those counterfactual conditions have affected the prosperity of a tenant farmer in central Italy?³⁹ Adam Smith foreshadows, In the Wealth of Nations (1937: 356–7), a vision contrary to Finley’s: We must not . . . imagine that the gain of the town is the loss of the country. The gains of both are mutual and reciprocal, and the division of labor in this, as in all other cases, advantageous to all the different persons employed in the various occupations into which it is subdivided.

17.2.2.1. The Economic Theory of Cities Over the course of history, cities have been engines of economic growth and technological and artistic innovation. Urban economists ³⁸ See also Finley (1999: 125, 194). ³⁹ In this thought experiment, the tenant farmer would owe rent not to a landlord residing in a city such as Rome, but to a landlord who was a member of the local gentry.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

172

Robert C. Ellickson

such as Edward Glaeser (1998), building on the ideas of Alfred Marshall and others, have developed a theory of why this is so.⁴⁰ The central point is that people in close proximity can more cheaply initiate mutually valuable economic and social interactions. The “agglomeration benefits” of cities can be grouped into three categories: (1) reductions in the transportation and information costs of arranging interactions; (2) enhanced specialization of labor and capital; and (3) spillover benefits arising out of the exchange of ideas. Although some scholars in classical studies have occasionally referred to one or more of these benefits, none appears to have made systematic use of the economic theory of cities. An example featuring a hypothetical sandal-maker will serve to illustrate each type of agglomeration benefit. The sandal-maker is an ex-slave living in a shop on a well-trafficked street in ancient Rome. Assisted by others, including several slaves of his former master, he fashions sandals out of leather, wood, and cord, and sells them out of his shop. How might this sandal-maker benefit from operating a shop in Rome, as opposed to a rural village? First, and most mundanely, the sandal-maker would profit from being in close proximity to numerous suppliers, workers, and customers. Within a human beehive such as Rome, a sandal-maker could find within easy walking distance numerous different suppliers of finished leather and other inputs necessary in the trade, including workers. A large number of potential buyers of sandals also would live in close proximity to his shop. The transportation and search costs these potential customers would incur when engaging in comparative shopping for sandals would be far lower in a big city than in a rural setting. Moreover, specialty retailers in a large metropolis commonly cluster their shops together, thereby reducing the transaction costs of both customers and suppliers. There is evidence that specialty retailers indeed did cluster their outlets in Rome (Holleran 2012: 53–60).⁴¹ A big-city location thus would likely benefit both sandal-makers and sandal-buyers by enhancing their mutual gains from footwear transactions. Second, a sandal-maker located in Rome, unlike one in a rural village, could readily specialize in crafting a particular style of sandal. Opportunities for specialization benefit both suppliers with distinct ⁴⁰ Like Glaeser, Schleicher (2010) presents the theory of agglomeration efficiencies in accessible fashion. Rosenthal and Strange (2004) summarize the technical literature. ⁴¹ Beard (2008: 60–1) describes the bunching of specialty retailers in Pompeii.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 173 talents and consumers who value product differentiation. As a result, in a big city, jobs are more varied, social opportunities richer, and entertainments more diverse. As the satirist Juvenal (1974: 95) put it, “If you can face the prospect of no more public games, purchase a freehold house in the country.” Juvenal himself appears to have been unwilling to face that prospect.⁴² Third, and commonly of greatest regional value, interactions within a great city such as Rome generate positive information spillovers. In the sandal-making context, these would have entailed the invention and diffusion of new ideas about how to make and market sandals. Specialists in a big metropolis are likely to fraternize, exchange information, and stimulate, through status competition, innovative thinking. Paradigm instances include the interactions of Impressionist painters in Paris and of computer scientists in the Silicon Valley. The high point of the Golden Age of Latin literature in ancient Rome is analogous. During Augustan times, two figures centrally involved were Maecenas and Messalla Corvinus, wealthy patrons of the arts. Their social and patronage circles came to include a broad array of Rome’s most distinguished poets, among them Horace, Ovid, and Virgil. Had Virgil remained in his birthplace, by legend the small village of Andes in northern Italy, it is a safe bet that he would never have written the Aeneid. The works of these various poets diffused throughout the Empire, benefiting outsiders who had never set foot in the capital. New ideas about the design and manufacture of sandals would have spread from Rome in similar fashion.⁴³ Although urban residents obtain benefits from city living, they also must bear a variety of congestion costs. These include the expense of big-city real estate (or, in the alternative, high commuting costs), enhanced environmental disamenities, and higher prices for most goods and services.⁴⁴ The economic theory of cities posits,

⁴² Robinson (1992: 173) asserts that Juvenal, a denizen of Rome, was “someone who clearly had no intention of living elsewhere.” ⁴³ Perhaps in implicit recognition of the agglomeration benefits that cities generate, Roman law encouraged migration to Rome. A migrating citizen had full political rights in the capital, but not, at least as a formal matter, in any alternative destination (Morley 1996: 174–6). ⁴⁴ The market price of wheat was higher in Rome than elsewhere (Temin 2013: 40–52, 254–5; Temin in volume I;). The city of Rome depended on imports of wheat, and its grain dealers would have borne relatively high expenses for both space and labor.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

174

Robert C. Ellickson

unsurprisingly, that people voluntarily move to a city when they perceive that their benefits from city life would outweigh their costs.⁴⁵

17.2.2.2. The Many Services That Rome Exported The consumer-city model is overly narrow because it focuses exclusively on flows of manufactured goods and agricultural products. Rome, on balance, was not a parasite because it exported a wide variety of valuable services to residents of its hinterland and provinces. After the founding of the Empire, Rome served as the capital city until  330, when Constantine moved the capital to Byzantium. On account of Rome’s special status, members of the city’s elite assumed central responsibilities in imperial governance. As Jane Jacobs (1969: 143) succinctly put it, “ultimately government became Rome’s chief export work.” During the first several centuries of the Empire, the primary service that the city of Rome exported was the Pax Romana (Lo Cascio 2007: 621–6; Temin 2013: 222–3). In this interlude the Romans were relatively successful at deterring internal unrest, invasions across the Empire’s frontiers, and piracies at sea. Mediterranean trade flourished. Although the legions and navies that were the instruments of Rome’s hegemonic control were commonly based far from the capital city, the imperial government in Rome coordinated these expensive military forces and designed and supervised the taxation system that paid for them.⁴⁶ During both the late Republic and early Empire, the governors and other top officials of the provinces (proconsuls, propraetors, legates, and others) were mostly drawn, as part of their regular career path, from Rome’s senatorial class. In addition, residents of the city of Rome were disproportionately involved in establishing the legal and financial institutions that promoted trade throughout Italy and the empire. The praetors, jurists, and others who refined the principles of Roman law and judicial administration were disproportionately based in Rome, and the city also appears to have been a favored venue of litigants.⁴⁷ Officials ⁴⁵ Because a decision to move may give rise to externalities, the population of a city may not be optimal (Tolley 1974). ⁴⁶ Hopkins (1980, 2002) analyzes the Roman system of taxation. ⁴⁷ D. 4.6.1.1 (Ulp. 12 ad ed.) provides that a party “absent on state business” should not suffer from any resulting delay in the initiation of an action for restitution of lost property. But D. 4.6.5.1 (Ulp. 12 ad ed.) adds that “those who serve the state in Rome

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 175 in the capital likely coordinated efforts to standardize coinages (Lo Cascio 2007: 626–30), and weights and measures (Beard 2008: 164–5, 195). From Rome came ideas, funding, and perhaps even designs for sundry economic development projects throughout the empire. These included roads between population centers, aqueducts, harbors, and the planting of cities in the various provinces. Rome also likely was an important center of private finance, such as that was. The wealthiest Romans congregated in Rome and would have sought help in investing their capital. Specialists in banking and credit emerged (Andreau 1999; Andreau in this volume; Harris 2006; Temin 2013: 157–89), and many of them would have regarded the capital city as an ideal base for operations. The Roman plutocrats who invested in rural and urban real estate would have appreciated having at hand specialists, such as lenders, surveyors, and lawyers, who were capable of assisting both in the effectuation of land transfers (Crook 1967: 78–9, 89) and the management of rental property. Guardians of the estates of wards controlled a significant fraction of wealth (Saller 2007: 99–100; Nicholas 2008: 90–7). They also would have been potential consumers of financial advice, legal services, and accounting assistance. As its role in the Golden Age of Latin literature illustrates, Rome was a center of innovation in the arts, sciences, and engineering, all potentially of benefit to outsiders. Although other cultural centers, such as Alexandria and Athens, may well have been more vibrant by some measures, Rome was unmatched in the central portion of the eventual empire. Many buildings and artistic works in Pompeii, for example, reflected architectural styles in the capital (Beard 2008: 51–2). Two libraries, one for works in Latin and the other in Greek, were completed on Palatine Hill in 28 , and many more followed (Robinson 1992: 57–8). These were a boon to writers of literary, historical, and technical works. Rome became a center of schooling, tutoring, and apprenticeships (cf. Saller 2012: 83). The father of a talented son might either send him, or accompany him, to Rome to advance his education. As youths, Cicero, Horace, and Ovid all were sent to the big city. Perhaps Rome’s most important “export” was its success in attracting to the city hordes of traders, tourists, envoys, and others who paid short-term visits. These itinerants consumed services and acquired goods while in the city, and in effect exported those goods and are not absent on state business,” implying that the city of Rome was seen as an especially convenient venue for litigation.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

176

Robert C. Ellickson

services when they returned home.⁴⁸ Rome was an inland city, over 20 kilometers up the Tiber from the port of Ostia. Fifteen roads radiated out in all directions from Rome into the surrounding hinterland (Chevallier 1976: 133). The capital offered specialized marketplaces that drew in outsiders seeking to peddle or acquire goods (Holleran 2012: 159–93). Cattle, for example, were auctioned at the Forum Boarium. But the city’s attractions were far more varied. In the mid-first century , Seneca described Rome’s appeal to both permanent and temporary migrants: They have flocked here from their townships and colonies, in short, from every part of the world: some have been drawn by ambition, some by the obligation of a public service, some by the office of envoy entrusted to them, some by luxury seeking a suitable and rich field for vice, some by desire for higher studies, some by public shows. . . . ⁴⁹

Ah, public shows. Besides hosting religious festivals and theater performances, Rome was famous for its games. These typically were open to all, including women and slaves (Robinson 1992: 170). What spectator could forget an event at the Colosseum featuring lions, elephants, and rhinoceroses (Beard 2008: 270)? Rome surely imported bread, but circuses it in effect exported to well-to-do tourists drawn to the city. Visitors attracted to a show or festival might stay in Rome for several days to shop (Holleran 2012: 191), a practice that would have enhanced revenues at the city’s many inns, bars, and houses of adult entertainment.

17.2.3. Evidence That Rome Was a Magnet for Talent The economic theory of cities posits that the ancient city of Rome would have been able to attract, and retain, a disproportionate share of the empire’s most talented individuals. Residence in a city helps enable people of genius to specialize in tasks that they perform best. They also can gain from integrating themselves into a social network of people who share their interests and talents. A specialized network of this sort provides opportunities for picking up ideas, making social connections, and motivating oneself through status competition. ⁴⁸ Finley (1999: 139) himself mentions a city’s “invisible exports of trade and tourism.” ⁴⁹ Seneca, “Consolation to Helvia” 6.2, in Seneca (2007: 167).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 177 The life histories of forty-one leading writers and teachers of the late Roman Republic and early Empire indicate that the city of Rome indeed served as a magnet for talent. The names of the fortyone appear in Table 17.1. I selected these individuals prior to researching their birthplaces and career paths. Each was eminent enough to have been included in John Hazel’s Who’s Who in the Roman World (2001).⁵⁰ The forty-one can be loosely grouped into five categories: historians (n = 9), literary figures (10), jurists and advocates (13), educators (6), and writers on technical subjects (3). With the exception of the jurist Modestinus, the lives of each

Table 17.1. Probable birthplaces of forty-one prominent Roman writers and teachers born 200 – 200. A. Born prior to 1  (n = 22) City of Rome Within one day’s walk of Rome Elsewhere in Central Italy Northern Italy Southern Italy North Africa Spain Turkey

Capito (j), Hortensius Hortalus (j), Lucretius (l), Quintus Mucius Scaevola (called “Pontifex”) (j) Cato the Elder (h), Aelius Stilo (e) Cicero (j), Verrius Flaccus (e), Labeo (j), Ovid (l), Sallust (h), Marcus Terentius Varro (h), Vitruvius (t) Gaius Valerius Catullus (l), Livy (h), Virgil (l) Ennius (l), Horace (l), Orbilius (e) Terence (l) Porcius Latro (e) Alexander Polyhistor (e)

B. Born after  1 (n = 19) In Rome or within one day’s walk of it [None] Elsewhere in Central Italy Juvenal (l) Northern Italy Celsus Publius (j), Pliny the Elder (h), Tacitus (h) Julianus (j), Papinian (j), Quintus Cervidius North Africa Scaevola (j), Suetonius (h) Spain Columella (t), Lucan (l), Martial (l), Quintilian (e) France Frontinus (t) Greece Plutarch (h) Turkey Modestinus (j) Levant Josephus (h), Ulpian (j) Unknown Gaius (j), Paul (Julius Paulus) (j) Notations: (e) denotes an educator, (h) a historian, (j) a jurist or advocate, (l) a literary figure, and (t) a writer on a technical subject.

⁵⁰ Although the data in Table 17.1 are primarily drawn from Hazel (2001), in some instances Magill (1998) and Hornblower and Spawforth (2000) also were consulted. The latter two sources contain fewer entries than Hazel does.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

178

Robert C. Ellickson

overlapped at least part of the period 200  to  200.⁵¹ Twenty-two of the forty-one were born prior to 1 . For the sake of nonclassicists, Table 17.1 employs present-day geographical designations to identify the region of a birthplace. Some caveats about the data must be noted. In many cases, classicists have been forced to conjecture about birthplaces and career paths, and they have not always agreed. The selections are contestable and forty-one is a small number. Perchance this preliminary venture will provoke others to compile a larger dataset. Despite their limitations, the data robustly support the proposition that a city attracts people of talent. Of the thirty-seven writers and teachers whose career histories are sufficiently documented, thirtyfour spent much, indeed likely most, of the primes of their careers in the city of Rome.⁵² These brainy individuals, of course, were not always in the capital. A person with talent and wealth at times might travel (perhaps as a member of the emperor’s entourage (Millar 1977)), spend time in a rural villa, and perhaps even serve out a multiyear tour of duty in a province. Nonetheless, the city of Rome appears to have been the main locus of the professional lives of thirtyfour out of the thirty-seven.⁵³ Each of the remaining three of the thirty-seven, moreover, had first-hand knowledge of the capital. Plutarch was born in Greece, but lectured in Rome and “had influential friends there” (Hazel 2001).⁵⁴ Gaius, who is thought to have received his legal education in Rome, apparently moved to Berytus (Beirut) as a young man. If so, Gaius was the only one of the thirty-seven who affirmatively spurned choosing Rome as his career base. Columella, the foremost Roman writer on agriculture, unsurprisingly was the third exception. Columella owned several farms not far from Rome and likely spent most of his prime years on them. His treatise, De re ⁵¹ Modestinus apparently was born shortly after  200. The Law of Citations of  426 gave special weight to his opinions and those of four other jurists. Modestinus was added to the group of forty-one so that it would include all five of these notables. On the role of the jurists in the Roman legal system, see Nicholas (2008: 28–38). ⁵² For four—Lucretius, Alexander Polyhistor, Verrius Flaccus, and Vitruvius— there is insufficient information about the venue of the prime career. ⁵³ Two close cases are included in the thirty-four: Virgil, who appears to have split the prime of his career between Rome and Naples, and Orbilius, who moved to Rome at age 50. ⁵⁴ Plutarch had predecessors. Ando (2003: 359) identifies numerous Greek writers who visited Rome during the late Republic and then circulated back to cities in greater Greece.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 179 rustica, written late in life, includes an extensive bibliography. To compile this bibliography, Columella would have needed a good library, and it is easy to guess to what city he went to find one. Roman leaders, like their counterparts in many other ancient societies, used banishment as a sanction for perceived misconduct. During the late Republic and early Empire, banishment invariably, and tellingly, included a prohibition against residing both in the city of Rome (Braginton 1944: 392) and in a much broader territory around it. Three individuals in the group of forty-one—Cicero, Ovid, and the jurist Paul (Julius Paulus)—unquestionably were banished, and Juvenal likely was as well. Of these four, only Ovid failed to receive eventual permission to return to Rome. Exiled to a site in present-day Romania, Ovid vainly begged for dispensation in letters to Augustus and, later, Tiberius (Braginton 1944: 404). The distribution of the birthplaces of these forty-one writers and teachers powerfully demonstrates Rome’s widespread appeal. The data in Table 17.1 indicate that no more than four were born in the city of Rome itself, and only an additional two within a one-day walk of the city. At least thirty-three of these forty-one talented contributors to Roman culture were born beyond the immediate hinterland of the city. This small data-set suggests that, as the Empire matured, the talented individuals migrating to Rome increasingly came from distant parts of the greater Mediterranean. In the dataset, three (14 percent) of the twenty-two talents born prior to 1  were born beyond the boundaries of modern Italy. By contrast, of the seventeen who were born after  1 and whose birthplaces are known, thirteen (76 percent) were born outside Italy, including four in North Africa and four in Spain.⁵⁵ Birthplaces have been identified for only three of the five jurists whose eminence was recognized in the Law of Citations. The three are Modestinus, born in Turkey; Papinian, born in North Africa;⁵⁶ and Ulpian, born in Lebanon. Each migrated to Rome and spent the prime of his juridical career there. Remarkably, none of the nineteen talented individuals born after  1 is known to have been a native of either the city of Rome or its ⁵⁵ Grounded conjectures about the birthplaces of jurists Gaius and Paul could not be found. ⁵⁶ Hypotheses about Papinian’s birthplace differ. Tony Honoré’s guess, reported in Hornblower and Spawforth (2000), is North Africa. Others have surmised that Papinian was from Syria.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

180

Robert C. Ellickson

immediate hinterland. This suggests, but, because the numbers are small, hardly proves, that the city of Rome for some reason came to lack capacity to produce home-grown talent. The relative shortness of life expectancies in Rome may have been part of the story (Scheidel 1999: 280).⁵⁷ The life of Josephus, a leading Jewish historian of the first century , illustrates the varied sources of Rome’s appeal. Josephus was born in Jerusalem in  37. After participating in the Jewish war against the Romans of  66, he surrendered to the Romans and soon came to Rome. The emperor Titus eventually granted Roman citizenship to Josephus, and provided him a pension and house. Josephus appears to have spent most of the following three decades in Rome, at work on his major historical works. The data in Table 17.1, limited as they are, counter Finley’s conception that Rome was a consumer city parasitic on outsiders. From Italy and beyond, Rome attracted exceptionally talented people. By clustering in the capital, these individuals spurred one another to compose histories, poems, legal treatises, and technical works. These cultural achievements, like the Pax Romana, were services that the city of Rome exported empire-wide.

17.3. LEGAL RULES REVEALING ROME’S COMMITMENT TO “COMMERCIALIZATION OF THE SOIL” The Romans adopted two general policies that were necessary, but not sufficient, conditions for the phenomenal growth of the ancient city of Rome. This section depicts Rome’s relative willingness, compared to that of other ancient civilizations, to make land alienable by sale and lease. The next section stresses the relative sagacity of Rome’s leaders in avoiding the twin perils of too little, and too much, government. In the absence of these two foundational policies, Rome’s entrepreneurs would not have been able to build and maintain the apartment blocks that accommodated the 800,000 people added to

⁵⁷ The theory that lead-poisoning is the explanation has fallen out of fashion (Scheidel 1999: 274).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 181 the city’s population between 200  and  50. Rome doubtless benefited from other advantages. These arguably included, among many others, a favorable location, other supportive cultural and legal traditions, and luck. But Rome’s relatively sound real-estate institutions played an essential role.

17.3.1. Norms and Laws Fostering the Alienability of Land The bedrock institution of private land ownership promotes prosperity by incentivizing smallholders to conserve and improve their holdings with scant supervision from on high (Aristotle, 1946: 49–51; Demsetz 1965; Ellickson 1993). Evidence from early Mesopotamia, Egypt, and China indicates that, in each location, individuals and families were allowed to own, in perpetuity, much of the land best allocated to agricultural and residential use.⁵⁸ Roman law, for its part, firmly protected the rights of private owners of both land and movables, so much so that nineteenth-century German critics chafed at the Romans’ “property absolutism” (Whitman 1996).⁵⁹ The placement of land into commerce is a hallmark of an archaic society’s transition to a market economy. Karl Polanyi (1944: 179) refers to this event as “commercialization of the soil.” Polanyi contended that land was not bought and sold until after the demise of feudalism in the late Middle Ages. Polanyi was wrong (Ellickson 1993: 1375–8). In Mesopotamia, evidence of land sales and land leases dates ⁵⁸ On practices in the Near East, where the solidest evidence is from Mesopotamia, see Ellickson and Thorland (1995) and Westbrook (2003a). On China, see Scogin (1990). An ancient regime, while permitting private ownership in land, at times might grant some lands in feudal tenure, and manage others through a palace, temple, or other hierarchical organization. ⁵⁹ For the regulation of land in Roman law, see Kehoe (2016). According to both Mommsen and Weber, just after Rome’s founding, arable land in Latium, with the exception of small plots assigned to families (heredia), was owned communally, purportedly following traditional clan or tribal practices (Love 1991: 20). Buckland (1963: 239) concurs: “There was no separate property in land in early Rome except for the heredium, or houseplace, which was not alienable.” Historically, communal ownership has been common for pastures, but has been exceedingly rare for arable lands during the growing season (Ellickson 1993: 1331–2, 1346–8, 1399). The notion that the founding Romans owned their arable land communally thus warrants skepticism. If they did, there can be no doubt that their descendants emphatically rejected that form of land tenure.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

182

Robert C. Ellickson

at least as far back as 2000  (Westbrook 2003b), and in Egypt, to the New Kingdom (Jasnow 2003). In China, policies favoring the alienability of private land were evident in the fourth century  or earlier (Scogin 1990: 1338–46, 1354). Legal authority to use land to secure a loan requires acceptance of the possibility that the land ultimately will end up being owned by the lender. Ancient Mesopotamian regimes were willing to accept that possibility (Ellickson and Thorland 1995: 393–9), and, as early as 500 , Athenians were placing stones to signify that a land parcel had been pledged to secure a debt (Finley 1951). In most ancient societies, loyalties to family and tribe inhibited unconstrained land transfers, as feudal entails later would in England. The alienability of land can have both beneficial and harmful consequences. On the positive side, a land transfer is likely to generate gains from trade, potentially benefiting both the buyer and seller. Other affected parties also may gain. In Rome, the voluntary sale of an apartment block, on average, would have shifted its stewardship to an owner relatively more competent in arranging for the building’s management—a possible boon for tenants. More generally, land alienability enhances the mobility of workers, households, and capital, fueling the dynamism of an economy. On the negative side, unconstrained land transfers can disrupt grounded social networks. If a community insider were to sell land to an outsider, the buyer likely would have fewer loyalties to local residents than the seller had had. An influx of strangers thus may vitiate a community’s endeavors, for example, to defend and police itself and provide its members with informal social insurance. The norms and laws of archaic civilizations therefore tended to deter an owner of ancestral land from alienating it to persons other than the members of the owner’s extended-family or tribe (Ellickson and Thorland 1995: 354–8, 387–93; Westbrook 2003a: 56–62; cf. Finley 1976b: 4–5).⁶⁰ Although the inhabitants of many other ancient civilizations permitted transactions in land, the Romans were particularly eager to do so.⁶¹ The city of Rome became, by the standards of ancient

⁶⁰ Romans symbolically recognized their precommercial roots by granting “tribal” assemblies (comitia tributa) a central role in republican governance. In practice, these groups were defined by geographic boundaries, not bloodlines. ⁶¹ On the commitment of Roman authorities, from the late Republic to late antiquity, to the workings of free markets, see Lo Cascio in volume I.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 183 times, a real estate bazaar (Rawson 1976: 86–9; Temin 2013: 139–48).⁶² Ferdinand Tönnies, a prominent nineteenth-century German critic of Roman society, regretted this commercialization of the soil. Tönnies (1887) admired institutional arrangements that foster the intimate and enduring local social networks characteristic of a precommercial society (Gemeinschaft).⁶³ Like Otto von Gierke and other fellow German critics, Tönnies thought that Rome’s commercialism had led to a loss of trust and devotion to social duty (Whitman 1996: 1845–9). Although the soundness of the German critics’ normative critique is debatable, they unquestionably were right to conclude that the residents of Latium, Rome’s region, had been uncommonly willing, at an early date, to substitute commercial relationships for kinship relationships. The Romans’ openness to change would later be reflected in their eagerness to borrow from other cultures, particularly Greek ones, as opposed to hewing to their own traditions. The Twelve Tables, a collection of tenets of private law codified in 451–450 , is a primary source of evidence on the predilections of the Roman elite during the early Republic.⁶⁴ The Twelve Tables staunchly embraced principles of both private property and freedom of contract. Most pertinently, this ancient source included two truly exceptional legal policies that expedited land alienation.

17.3.1.1. Broad Freedom of Testation The Twelve Tables articulated a principle of complete freedom of testation (Crook 1967: 122; Watson 1995: 174–5).⁶⁵ This declaration, ⁶² Nonetheless, compared to a present-day real estate market, Rome’s was sluggish. There were relatively fewer transfers by sale, and more by either testamentary disposition or conferral of a dowry (Temin 2013: 144). No evidence has been found of brokers of either land sales or leases (Finley 1999: 117–18). Markets for loans secured by land appear to have been thin, in part because the wealthy preferred to borrow from family members or friends (Crook 1967: 170–8; Temin 2013: 168–71). ⁶³ Sociologists debate whether life in a large city is as socially alienating as Tönnies conceived. Wirth’s analysis (1938), supportive of Tönnies’s view, has been challenged by Fischer (1982). ⁶⁴ On the Twelve Tables, only fragments of which survive, see Westbrook (2009a). Livy asserted that, prior to the enactment of the Twelve Tables, the Roman Republic sent a mission to Athens to learn about that city’s legal traditions. Many scholars are skeptical of Livy’s account (Watson 1995: 111–12). ⁶⁵ On Roman inheritance law, see generally Crook (1967: 98–138).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

184

Robert C. Ellickson

presumably not at violence with the norms of the early republicans, overtly spurned the archaic impulse to discourage the transfer of property outside the family. On its face, the Twelve Tables authorized a paterfamilias to transfer patrimonial land by testamentary devise to a patron, a client, an influential politician, a mistress, or whomever. As a matter of comparative law, complete freedom of testation is an extreme legal outlier. In the contemporary United States, virtually all states confer on a surviving spouse an elective share (Kelly 2013: 1128 n. 9), typically at least one-third of the decedent’s estate. Louisiana, influenced by French practice, entitles surviving children to shares of their parents’ estates as a matter of right (ibid.: 1181). In preindustrial England, the norms of the aristocracy favored the voluntary use of entails to keep patrimonial real estate within the family. Roman law never permitted that option (Buckland 1963: 189, 276; Crook 1967: 121–2; but cf. Champlin 1991: 177–80). Champlin (1991) has cogently analyzed actual testamentary practices after 200 , when textual evidence first becomes available. A wealthy Roman, seeking to avoid the “horror of intestacy,” typically chose to execute a will (ibid.: 41–63). A large majority of testators, however, honored informal duties of family support and included both children and spouses among the named heirs (ibid.: 103–30). Nonetheless, many also made some use of their broad freedom of testation. They commonly chose to confer smaller shares on daughters than on sons (ibid.: 112–20), and not uncommonly provided legacies to servants and friends (ibid.: 131–54).⁶⁶ Crook (1976: 83) asserts that, in practice, testators “frequently” favored persons outside their families, “another spur to the brisk movement of land.”⁶⁷ During the late Republic, formal Roman inheritance law was amended to soften the Twelve Tables’ unbridled endorsement of freedom of testation. After these amendments, for example, a testator had to expressly disinherit a son to bar his claim (Champlin 1991: 14). And beginning no later than 52 ,⁶⁸ Roman law provided a procedure, querela inofficiosi testamenti, through which a descendent or other claimant could assert that a testator had violated moral duties by completely disinheriting the claimant (Crook 1967: 120–1; Champlin 1991: 15). This remedy seems to have been relatively frequently ⁶⁶ Because heirs were responsible for the debts of the decedent, a testator might have reason to select ones whose assets were meager. ⁶⁷ See also Rawson (1976: 85–9). ⁶⁸ Val. Max. 7.7.2.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 185 invoked, despite the burdens of litigation it placed on a challenger. But Roman law continued to be relatively protective of freedom of testation. The lex Falcidia of 40 , for example, limited a claimant who had successfully invoked querela inofficiosi testamenti to no more than one-quarter of the testator’s estate. This law also restricted the portion of the estate that the testator could give away in legacies, requiring that the heirs receive at least one-fourth of the estate.

17.3.1.2. Legal Practices That Provided Title Assurance to Land Buyers A robust real estate market depends on an effective system of title assurance. In its absence, risk-averse land buyers either may get cold feet, or have to bear the additional costs of acquiring title information.⁶⁹ Legal systems far more ancient than Rome’s struggled to provide procedures that would help assure a land buyer of the quality of a seller’s title (Ellickson and Thorland 1995: 382–7).⁷⁰ Some developed a method of publicizing an upcoming sale. Others provided a buyer a means of examining documents evidencing prior sales. The city of Rome, however, appears to have never developed a system of land records on which land buyers could confidently rely (Arruñada 2012: 12–16, and his chapter in this volume; Nicholas 2008: 104).⁷¹ From an early date, however, Roman law did provide several procedures that helped reassure wary land buyers.⁷² Two central ⁶⁹ Frier (1980: 35–6, 122–3) cites instances of the fraudulent leasing of residential property by a person purporting to be its owner. ⁷⁰ On the general problem, see Baird and Jackson (1984: 304–6) and Arruñada in this volume. ⁷¹ Rome’s officials may have used records from census surveys for a variety of purposes, such as determining individuals’ eligibility for the grain dole and vulnerability to the military draft. But the jurists seem not to have regarded evidence in land records to be pertinent to the resolution of title disputes between private parties. Land records appear to have been better developed beyond the city of Rome. The Romans’ system of centuriation in the colonies gave rise to both land maps and a version of land records (Duncan-Jones 1976: 12–20; Libecap and Lueck in this volume). In Roman Egypt a land sale was supposed to be registered as soon as it took place (Crook 1967: 147–8). By  478, the Eastern Empire may have instituted a formal system of land records (Macnair 1999: 557–8). ⁷² After their refinement by jurists, Roman law conceptions of property were significantly more sophisticated than those of prior legal systems in the Near East. Among Roman law’s achievements were the distinctions between immovable

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

186

Robert C. Ellickson

ones were usucapio and mancipatio, institutions that dated back at least to the early Republic. Specialists in Roman law have devoted great attention to both. a. Usucapio The Twelve Tables included a second provision stunningly supportive of land alienation. Virtually all legal systems entitle a long-time land possessor who has satisfied specified requirements to obtain title to another’s land without having to pay for it. In Anglo-American law, the pertinent doctrine is adverse possession. The closest Roman law analogue was usucapio (see generally Buckland 1963: 241–9; Nicholas 2008: 122–5; Arruñada in this volume). To usucapt successfully, an adverse possessor had to have acquired the property on a lawful basis and in good faith—that is, to have thought himself the rightful owner—and to have remained in overt and continuous possession for a sufficiently long period of time.⁷³ How long? The Twelve Tables provided that two years would suffice. Compared to other legal systems, a two-year period is extraordinarily short. England during its Industrial Revolution required twenty years for successful adverse possession. A solid majority of US states currently require a period of ten years or more, and none require as few as two (Foster and McKinney 2011: 201). Remarkably, the two-year period for usucapio announced in the Twelve Tables remained a rule of Roman law for at least the ensuing seven hundred years (Honoré 1989: 141). The shortness of this statute of limitations, on balance, promoted the alienability of land by limiting how long a good-faith purchaser had to worry about the quality of a seller’s title.⁷⁴

property (real estate) and movable property, and between rights in personam (against single individuals) and in rem (against the world) (Nicholas 2008: 99–100). By clarifying landowners’ legal rights and remedies, these conceptual advances might have contributed to Rome’s growth (cf. ibid.: 1; Westbrook 2009a: 69). I myself suspect that the size of the economic boost from these abstractions would have been trivial at most. ⁷³ Watson (1971: 64) thinks that the requirement of good faith likely was added after 200 . ⁷⁴ Offsetting some of this gain would be a buyer’s increased risk of later losing purchased land to a third-party claimant invoking usucapio.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 187 b. Mancipatio Even prior to the Twelve Tables, Roman citizens regarded mancipatio as the favored procedure for an inter vivos transfer of Italic land (Buckland 1963: 235–41). A mancipatio ceremony required the presence of not only the transferor and transferee, but also (in effect) six witnesses, each required to be an adult male citizen. The Digest and other textual sources reveal nothing about the procedures governing the selection of these witnesses,⁷⁵ and commentators have seldom focused on how they were chosen.⁷⁶ If negotiating in good faith, both a prospective seller and a prospective buyer of land would have gained from the presence at a mancipatio ceremony of witnesses likely to be knowledgeable about prior transactions involving the same property. Fitting this description might be a prior owner of the land being transferred, or an owner or long-time occupant of adjacent property. A custom of choosing knowledgeable witnesses would have made it more difficult for an unscrupulous seller to sell the same property twice. A buyer might also have worried about a hidden encumbrance, such as an undisclosed land-security interest⁷⁷ or servitude.⁷⁸ If able to reduce the ⁷⁵ In Berger (1953), an unusually exhaustive source, neither the entry under Mancipatio nor that under Testis indicates the party or parties responsible for selecting these witnesses. ⁷⁶ Arruñada in this volume cogently stresses how a mancipatio ceremony could serve to publicize a transaction; see also Nicholas (2008: 142, 152). But the usually insightful Raymond Westbrook (1989: 209) failed to discern mancipatio’s potential title assurance function. Cf. Champlin (1991: 75–81), presenting evidence on the selection of witnesses for the execution of a will. ⁷⁷ During the Republic, fiducia denoted a loan secured by property that passed into the ownership of the creditor. Roman law later recognized a number of alternative arrangements (pignus, hypotheca, and antichresis) (Buckland 1963: 431–3, 473–81; Crook 1967: 243–9). Because parties typically used mancipatio to execute a fiducia transaction (Wigmore 1941: 384–5; Watson 1971: 84–8), a subsequent buyer’s adroit selection of mancipatio witnesses potentially could have reduced his risks of undisclosed security interests (Buckland 1963: 231). On land security generally, see Verhagen in this volume. ⁷⁸ An urban servitude was not res mancipi, and thus not creatable by a mancipatio ceremony, but rather only by an in iure cessio proceeding before a magistrate (Watson 1971: 82). On account of this difference in mode of creation, neighbors, assumed here to be a buyer’s favored mancipatio witnesses, might be less likely to know about servitudes than about other title defects. But Roman law did provide a land buyer several protections against a hidden servitude. A mancipatio ceremony created a warranty of title that a buyer could enforce against a seller by means of an actio auctoritatis action. This warranty appears to have included protection against an undisclosed servitude, especially if the mancipation had included the locution optimus

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

188

Robert C. Ellickson

risks of title flaws through the adroit selection of mancipatio witnesses, a buyer would have been willing to pay a higher purchase price. If so, the seller and buyer would have had more surplus to share. If the parties involved in a land sale had wished the witnesses at a mancipatio ceremony to be locals, they likely would have preferred to hold the ceremony at or near the land being transferred. But, in his Institutes published in  161, Gaius declares that “Lands, however, are usually mancipated at a distance” (Gai., Inst. 1.121). The customs for siting these ceremonies doubtless varied across the many centuries and territories under Roman influence. An off-site location, the routine choice according to Gaius, would have been sensible if the buyer and seller were to have been trusting intimates. Suppose, for instance, that a rich patron residing in the city of Rome had wished to sell a rural estate in Northern Italy, farmed by tenants, to a client who resided close to the estate. These trusting parties might have decided to conduct the mancipatio event in Rome to save the patron the burden of making the long journey to the rural site. By contrast, suppose the seller and buyer were to have been strangers, and the property in question, an apartment block in Rome itself. Under those circumstances, the parties likely would have preferred to hold the mancipatio ceremony on-site, where it would have been easier to assemble witnesses knowledgeable about prior real estate transactions in the immediate vicinity.⁷⁹ In a society where land titles were uncertain, the mancipatio ritual thus was structured to serve as a potential source of buyer comfort.

17.3.2. Legal Practices That Fostered the Leasing of Houses and Shops Ancient Rome was a city of renters. A third-century census found that apartment blocks outnumbered houses by more than maximus (Crook 1976: 72–3). Roman law also refused to enforce a servitude that either imposed an affirmative duty or whose beneficiaries did not own appurtenant land (Nicholas 2008: 142–3). These policies would have lessened a land buyer’s risks of surprise. ⁷⁹ Constantine the Great, emperor from 306 to 337 , eventually decreed that a land-sale ceremony had to be conducted on site and in the presence of neighbors (Macnair 1999: 556–7).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 189 twenty-to-one.⁸⁰ The condominium form, an ownership arrangement that might have enhanced the value of some apartment blocks, was not recognized (Crook 1967: 143–4; Natelson 1987). These facts imply that 90 percent of imperial Rome’s residents, if not more, lived in rental units.⁸¹ For a relatively prosperous household that could not afford the luxury of a house, the best option was to lease an apartment in an upscale apartment block for a multiyear term. Members of a poorer household either resided in the shops where they worked, were short-term tenants in a squalid rooming house, or rented a cubicle apartment on an upper floor of a tall apartment building.⁸² Frier (1980) provides a definitive analysis of landlord-tenant relations in Roman cities during the early Empire.⁸³ He persuasively argues that, in a city such as Rome, the formal landlord-tenant law depicted in the Digest would have governed at most the upscale portion of the rental housing market (Frier 1980: 39–52). In the downscale sector, a legal proceeding typically would have been far too costly for either a landlord or tenant to consider. And even in the upscale sector, in most residential-lease disputes, the stakes would have been small, and the landlord and tenant both would have been anticipating continuing relations. Under these circumstances, the disputants commonly would have decided to forgo the expense and hassle of legal proceedings. They instead would have tended to resolve their disputes informally, applying social norms that might not have corresponded to the strictures of formal law (Ellickson 1991; cf. Kehoe 2007: 13–15, 100–5). If Roman landlord-tenant law were to have severely impeded the ready consummation of long-term and short-term residential leases, the city could never have grown as it did. Dozens of passages in Justinian’s Digest refer to urban landlord-tenant disputes. Many of

⁸⁰ See n. 17. ⁸¹ Although the members of a household dwelling in a domus typically were far more numerous than those of a household in an insula, an insula might accommodate several dozen households, while a domus accommodated only one. ⁸² See nn. 11–16 and accompanying text. ⁸³ The practice of leasing real estate long predated the rise of Rome. In Mesopotamia, the leasing of both houses and fields was common by the middle of the third millennium  (Ellickson and Thorland 1995: 369). But, as in other areas of law, the Romans fleshed out, and in this instance helped urbanize, a legal institution that others had pioneered.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

190

Robert C. Ellickson

the classicists who have analyzed these entries have assessed the various doctrines separately. This section offers a holistic, if stripped-down, account of applicable Roman law and custom. I discuss the interplay among three norms that were central in Roman urban leasing practice. Each of the three one-sidedly favored either the landlord or tenant, and each is entirely at odds with contemporary US landlord-tenant practice. Frier has concluded that Roman residential landlord-tenant law, in the limited domain where it mattered, achieved a reasonable “balance” between the interests of landlords and tenants, thereby enabling the flourishing of an active market for urban apartments (Frier 1980: 192–3, 209).⁸⁴ Because the three one-sided norms crudely offset each other, on balance, I agree with Frier’s assessment. First, Roman law protected a lessee’s entitlements against a landlord solely by means of liability rules, as opposed to property rules (Calabresi and Melamed 1972; Frier 1980: 64). In the eyes of Roman jurists, a landlord retained dominium (Buckland 1963: 187–9). The upshot, in their view, was that a landlord could oust a tenant from leased premises at any time, albeit at the risk of being liable to the tenant for damages incurred.⁸⁵ These legal consequences of dominium were regarded as immutable, that is, not modifiable by the parties. Most US states, and doubtless many other jurisdictions, rightly reject this rule and instead, as a default, provide propertyrule protection to a tenant not in breach. A residential tenant who has robust security of tenure has fewer worries about suddenly losing a home, and is more likely to invest in both improving the premises and developing local social connections. Roman law included a second rule that also was decidedly prolandlord. If a lease so provided, and perhaps even if it did not, a landlord was entitled to unilaterally seize most of the tenant’s furnishings to satisfy the tenant’s arrears on rent payments, and possibly other debts as well (Frier 1980: 105–35).⁸⁶ In most contexts, a legal system should disfavor this remedy, known as distraint in Anglo⁸⁴ Kehoe (2007: 193–7) similarly concludes that Roman jurists sought to balance the interests of landlords and tenants when developing the imperial law that governed the leasing of farms. ⁸⁵ By extension, the buyer of a landlord’s interest was immutably entitled to oust a sitting tenant. If ousted in this fashion, the ex-tenant’s sole possible remedy would have been an in personam action against the former landlord (Crook 1976: 73–4). ⁸⁶ Beard (2008: 89–93) describes furnishings in Pompeii apartments.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 191 American law. A tenant may be provoked to resist violently a landlord’s agents who have peremptorily entered the tenant’s apartment and begun to empty it. Most US states prohibit this self-help action (Rabin 1984: 538), and Roman jurists were aware of its risks.⁸⁷ But there was a third, and countervailing, practice. During the period of Rome’s rapid growth, the Roman law of apartment leases embodied an overarching, although hardly blanket, commitment to freedom of contract (Frier 1980: 61–3).⁸⁸ Many residential tenants appear to have succeeded in wielding their contracting power to restore a rough balance of power in their relationships with landlords. At least during the early Empire, most urban leases either explicitly or implicitly provided that a tenant’s rent did not become due until after a period of occupancy (Frier 1977: 29; 1980: 37).⁸⁹ This remarkable practice, coupled with the related custom that a tenant did not have to provide a security deposit in advance (Frier 1980: 61 n. 14), vastly increased the leverage of urban tenants. The term of a residential lease in Rome typically extended for several years, began on the standard moving day, July 1, and ended on June 30. The multiyear term was divided into payment periods, seldom less than six months each. In the middle of a payment period, a landlord therefore would have been apprehensive that the tenant would fail to remit the significant sum due at the period’s end. The tenant’s control over this contingency would have deterred the landlord from unexpectedly ousting or otherwise abusing the tenant in mid-period. An ousted tenant likely would have become enraged, and less likely to perform. Twenty-first-century landlord-tenant law rejects each of these three features of Roman landlord-tenant practice. Considered in isolation, each is suspect on the merits. In most instances, legal rules should protect, perhaps immutably, both a tenant’s security of tenure and immunity from a landlord’s unilateral seizure of ⁸⁷ Kehoe (2007: 155–7) reviews Roman law’s limits on the use of distraint in a rural context. ⁸⁸ See e.g. D. 19.2.15.1 (Ulp. 32 ad ed.): “ . . . [O]r if they agree on something else in a clause of the hire and this duty is not carried out, there will be an action on hire.” In some contexts, bargaining to modify a default rule resulted in a pro-landlord clause (Frier 1980: 61–3, 141–2). See e.g. D. 19.2.11.1 (Ulp. 32 ad ed.), in which a tenant is deemed liable for damages from a fire that the lease forbade. Conversely, a tenant could bargain for a waiver of the landlord’s remedy of distraint (Frier 1980: 115). ⁸⁹ D. 19.2.19.6 (Ulp. 32 ad ed.) mentions a rent that was payable in advance. Frier (1980: 37) asserts that this provision’s detailed discussion implies the rarity of advance payments, an interpretation contrary to Crook’s (1967: 154).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

192

Robert C. Ellickson

furnishings. And a landlord has good reasons for insisting that a tenant both pay rent in advance and provide a security deposit. The Romans’ trio of one-sided landlord-tenant practices, however, did crudely offset, resulting in the “balance” that Frier has identified.

17.4. THE ACTIVE, BUT CABINED, ROLE OF GOVERNMENT During Rome’s heyday, the praetors, aediles, emperors, and others who had a hand in the capital’s governance generally made sensible, if hardly flawless, decisions about the scope of governmental undertakings. Either overly passive governance, or overly intrusive governance, would have held the city back (Morris et al. 2007: 11–12). Despite their occasional confiscations of landed property, Rome’s officials largely respected Roman law’s staunch commitments to protection of property rights and freedom of contract. They also were cleverly selective in identifying additional roles that the city government might affirmatively play. This prudence was another necessary, although not sufficient, condition for the city’s emergence as the titan of the Mediterranean.

17.4.1. Policies Addressing Negative Land Use Externalities As noted, most of Rome’s neighborhoods were noisy, smelly, and unsanitary. The city’s governors nonetheless applied a light hand to problems of negative land use externalities. Although Pirson (1997: 175) refers to work by German scholars on “building codes,” there is no evidence that a builder needed a public permit in advance (Robinson 1992: 33).⁹⁰ Nor is there any evidence of zoning policies.⁹¹ Augustus’ decree limiting building heights along Rome’s public streets to 70 feet, if actually enforced, seldom would have been ⁹⁰ Scobie (1986: 405) cites a regulation limiting the width of a party wall to 1.5 inches, a puzzling requirement. ⁹¹ Holleran (2012: 58–9) refers to clusters of tanners in Rome, and Goodman (2007: 47) to a concentration of industry on the right bank of the Tiber. A cluster of enterprises of course may arise in the absence of a government directive.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 193 both a binding and burdensome constraint on the developer of an apartment block.⁹² In some periods, city authorities did require a landowner who wished to demolish a structure both to obtain permission in advance and immediately build a replacement structure (Crook 1967: 261–2; Garnsey 1976: 133–6). This policy, like the limit on heights, may have been motivated in part by aesthetics. It also helped assure lateral support of adjoining buildings. To achieve land-use coordination at the block level, Rome’s public authorities wisely tended to rely on decentralized mechanisms more than on inflexible command-and-control regulations.⁹³ Roman law included the rudiments of a law of nuisance.⁹⁴ A neighbor damaged by another’s fire or smoke, for example, might be entitled to pursue a legal action for damages.⁹⁵ More important seems to have been the well-developed Roman law of servitudes (Buckland 1963: 259–68; Crook 1967: 149–52; Rodger 1972; Nicholas 2008: 140–8; Bannon 2009).⁹⁶ Neighbors could resolve a lurking externality problem, such as a potential blockage of light or view from an apartment block, by negotiating a servitude that bound and benefited not only themselves but also their successors in ownership (Buckland 1963: 264). Perhaps most important, Roman customs of real estate development produced a set of agents who were well-positioned to negotiate servitudes and other decentralized solutions to externality problems. Coase (1961) famously observed that, regardless of the state of the law, neighboring landowners are incentivized to resolve potential ⁹² The best-preserved insula in Rome, mentioned in n. 9, slightly exceeds Augustus’ height limit (Scobie 1986: 406). Trajan may later have shortened the limit to 60 feet (Robinson 1992: 35–6). ⁹³ On land use issues, see generally Crook (1967: 151, 165–7); Watson (1971: 75–83); and Rodger (1972). ⁹⁴ Anglo-American law students learn that nuisance law derives from the principle sic utere tuo ut alienum non laedas (use your property in such a way that you do not damage another’s). This Latin phrase appears to be a post-Roman invention. In English reports, the first sighting is Edwards vs Halinder, 74 Eng. Rep. 385 (1594). Lord Coke later quoted the maxim in his famous opinion in William Aldred’s Case, 77 Eng. Rep. 816, 821 (K.B. 1611). ⁹⁵ See e.g. D. 9.2.27.9 (Ulp. 18 ad ed.) (applying negligence standard to determine liability for spread of a fire); D. 8.5.8.5 (Ulp. 17 ad ed.) (a much-discussed passage authorizing relief against smoke from a ground-level cheese shop); but cf. D. 8.5.8.6 (Ulp. 17 ad ed.) (a moderate amount of smoke from a hearth would not be actionable). ⁹⁶ Servitudes predate Roman law. They were recognized in both ancient Mesopotamia (Westbrook 2003b) and New Kingdom Egypt (Jasnow 2003).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

194

Robert C. Ellickson

nuisance conflicts in mutually advantageous fashion. He also stressed that the transaction costs of reaching an agreement might foil their bargaining efforts. In ancient Rome, Coasean bargaining to resolve a potential landuse conflict typically would have involved relatively few agents, a condition conducive to decentralized coordination. Most houses and apartment blocks were situated on spacious lots. City blocks in ancient Rome, as in modern Rome, generally were small in area. Because city officials apparently did not attempt to regulate the size and shape of lots, entrepreneurs could consolidate and subdivide lots within a city block at will. Most apartment blocks in Ostia, thought to be similar to those in Rome, sat on lots that had varying shapes and commonly included an area that exceeded 1,000 square meters (Packer 1971: 93–110).⁹⁷ A small city block in Rome thus might have included only handful of apartment blocks, each teeming with tenants. A high-status owner of an apartment block or lodging house in Rome had several means of avoiding the hassle, and potential social stigma, of involvement in the building’s day-to-day management. An owner commonly entrusted supervision of the building either to a middleman who had taken out a master lease on the entire premises (Frier 1980: 30–1, 36; du Plessis 2006), or to a caretaker (insularius) (Frier 1980: 24, 29–30; Wallace-Hadrill 1994: 132; Frier and Kehoe 2007: 130–4). Given the limitations of Roman agency law, promising candidates for the caretaker role would have included the owner’s son, slave, or ex-slave.⁹⁸ Roman law recognized a particular business form, the peculium, that was especially adapted to a slaverun enterprise (Kirschenbaum 1987: 31–47; Hansmann et al. 2006: 1360–4; Abatino and Dari-Mattiacci in volume I). A hard-working slave had a good prospect of eventually buying freedom from a master.⁹⁹ In a much-quoted advertisement painted on a wall of an

⁹⁷ Compare Frier (1980: 9), describing as “typical” an Ostian apartment block that was situated on a slightly rhomboid lot whose sides averaged 21 meters. ⁹⁸ Roman law generally did not permit an agent other than a son, slave, or ex-slave to bind a principal (Kirschenbaum 1987: 1–7, 13–14). ⁹⁹ Love (1991: 136) estimates “that probably over three-fifths of slaves were freed before they were thirty.” This is probably incorrect. Koops in this volume implies that manumission after age 30 might have been more typical. On slavery in the Roman Empire, including the incentives of slaves to work, see Scheidel (2012b).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 195 insula in Pompeii, prospective tenants were advised to “consult Primus, slave of Cnaius Alleius Nigidius Maius” (Beard 2008: 109). On each lot, in short, there typically would have been an individual with presumptive authority to bargain over smoke, noise, and other externality issues. In the case of an apartment block, the key figure likely would have been either the caretaker whom the owner had selected, or the middleman to whom the owner had leased the entire complex. Roman law incentivized an owner of an apartment block to prompt the owner’s agent to consider the welfare of current tenants. The partial blockage of a tenant’s light by a neighbor, for example, could give rise to a tenant claim against his own landlord (Frier 1980: 102). For a domus, the agent for Coasean bargaining normally would have been the paterfamilias of the household (Saller 2007; Nicholas 2008: 65–9).¹⁰⁰ The paterfamilias would have been well-positioned not only to expeditiously resolve land-use conflicts among members of his often-large household, but also to represent the household in negotiations with neighbors over block-level environmental issues. During the late Republic and early Empire, Rome attracted continuing waves of immigrants. Most of these newcomers must have been aware of the noise, fires, diseases, and other risks of life in the big city. Narrowly targeted building regulations possibly would have made Rome an even more attractive place of residence. But crude building regulations, such as a draconian limit on the heights of the city’s buildings, would have stifled the supply of private housing. The Romans likely were wise to relegate the resolution of most nuisance and building-quality issues to blocklevel actors.

17.4.2. The Supply and Financing of Public Goods Rulers in ancient Mesopotamia, Egypt, and China attained legitimacy in part by providing dams, canals, and other water projects that would have been beyond the capabilities of non-state actors ¹⁰⁰ A Roman household did not invariably include a paterfamilias. Probably over 20 percent of land was owned by women, and almost as large a fraction by orphans (a male orphan would also be a paterfamilias) whose affairs were being managed by a guardian (Saller 2007: 97, 100). But in these situations as well, there likely would have been an agent authorized to speak for the household.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

196

Robert C. Ellickson

(Wittfogel 1957). The Romans similarly developed a public sector that bore responsibility for providing aqueducts and other public goods. Roman jurists made a sharp distinction between public property (res publica) and private property, and favored the staunch protection of both (Buckland 1963: 183). In the absence of government improvements and services, Rome could never have grown as it did. The theory of public finance, pioneered by Paul Samuelson (1954) and Richard Musgrave (1959), seeks to identify the goods and services that a government can provide more capably than can the decentralized actors in the private sector. In brief, a public good is one that either is nonrivalrously consumed, such as national defense, or one, such as a street, from which a provider cannot practically exclude entrants. Rome’s officials, most prominently the curule and plebeian aediles, had no overarching theory of the proper scope of government, and determined its ambit by muddling through. During the period of the city’s ascendance, the decisions of these officials, compared to those of many other ancient polities, were relatively sound. Ancient travelers commonly regarded Rome’s public works to be among its greatest achievements (Robinson 1992: 59, 99). Management of the city streets was a key public undertaking (Robinson 1992: 59–82). Archaeological findings indicate that the city of Rome evolved from a settlement that dated back to the early second millennium . On account of path dependence, the locations and widths of the settlement’s earliest street rights-of-way would have tended to endure (Ellickson 2013). Rome’s officials began to pave its streets in 238  (Chevallier 1976: 71) with polygon-shaped stones, likely of basalt. Street congestion eventually became a serious problem. In 45 , Julius Caesar famously banned, with some exceptions, traffic by carts between sunrise and late afternoon (Beard 2008: 80). Evidence from Pompeii (Beard 2008: 65–70) implies that Rome’s officials may have experimented with one-way streets and pedestrian-only streets. City officials also made efforts to beautify streetscapes. By the early Empire, there were an estimated 500,000 statues in Rome (Beard 2008: 77–8), many of them placed in streets, forums, and other public spaces. Rome’s public sector also saw to the provision of public buildings, many of world-class design. These included temples to various gods, public baths, theaters, libraries, and the famous venues for games (Robinson 1992: 47–58). Censors, the officials mainly responsible for

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 197 public contracts during the Republic, typically hired specialized contractors (publicani) to execute these building projects (du Plessis 2004; Hansmann et al. 2006: 1360–4). City officials provided essential utility services as well. By 200 , Rome already had two aqueducts. The fast-growing city added seven more by  52, quadrupling Rome’s public water supply (Martini 1976: 563). The aqueducts supplied the city’s public fountains, where most households drew their water. Partly to help clear this incoming water, Rome also became a pioneer in sewer technology (Robinson 1992: 117–19). Besides attending to these physical improvements, public officials provided services that arguably can be characterized as natural monopolies. These included the coordination of street cleaning (Beard 2008: 56) and, beginning in the early Empire, the provision of mobile crews of fire-fighters.¹⁰¹ Although theorists of public finance might quibble over the wisdom of some of Rome’s public undertakings, most of them likely would give the city’s high marks overall. As important as what Rome’s governors did, however, was what they declined to do. During the period of the city’s rapid growth, the production and management of houses, apartment blocks, shops, and many other goods and services were relegated to the private sector. Most theorists of public finance similarly would endorse the decision to leave this portion of the economy privatized. A private individual who supplies a good or service personally bears the cost of not making a profit, and therefore is likely to be more cost-conscious than a public supplier. More generally, economic forces, in a competitive market, tend to be far better than political forces at weeding out inefficient suppliers. Consistent with this theory, the governments that initiated major efforts to socialize housing supply during the twentieth century typically performed poorly.¹⁰² The ancient Romans avoided this stumble.

¹⁰¹ See Robinson (1992: 105–8), describing the roles of the hundreds of vigiles who patrolled Rome nightly. ¹⁰² During the mid-twentieth century, the governments of France, Netherlands, the United Kingdom, and many other nations assumed responsibility for providing a majority of their nation’s rental housing stock. These policies, like the public housing program in the United States, have generally been more wasteful than government housing assistance programs that would have relied more on the private providers of rental housing (Ellickson 2010: 986, 995–1003).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

198

Robert C. Ellickson

In an economy growing on the order of 0.1 percent per year, only a few basic policy mistakes would have been sufficient to push Rome off the path toward prosperity.¹⁰³ The Romans also were relatively adroit at choosing methods of financing their various public goods. Although localities collected general taxes in certain periods (Lo Cascio 2007: 625), these seldom were a major source of revenue. Rome defrayed more costs through the imposition of user fees on benefited landowners. An owner of land abutting a street, for example, might be charged with responsibility for maintaining that stretch of street, and, on failure to perform, billed for the costs of a contractor hired to do the job (Crook 1967: 257–60; Robinson 1992: 59–73). Some consumers of water also were charged user fees (Robinson 1992: 104). Far more important in the system of public finance was the custom of euergetism, a refinement of a tradition previously developed in some Greek city-states (Engels 1990: 128; Finley 1999: 150–4). Rome’s richest residents were under great pressure from both government officials and social peers to finance voluntarily public improvements and services.¹⁰⁴ During the early Empire, for example, the emperor might set a good example by “giving” a public building or group of gladiatorial games to the people of Rome (Millar 1977: 195, 365). Victorious generals and ambitious plutocrats contributed in similar fashion (Robinson 1992: 160–72). Supplementing the system of euergetism were informal requirements that men appointed to powerful positions make out-of-pocket donations (summae) to finance public endeavors. After Augustus appointed his friend M. Agrippa curator aquarum, Agrippa paid for a new aqueduct out of personal funds, a precedent soon followed by the emperor Claudius (Rodgers 2004: 15–16).¹⁰⁵ Rome’s sumptuary laws (Harris 2007: 530) may have been partly designed to redirect the expenditures of the wealthy away from conspicuous consumption (a zero-sum game of

¹⁰³ The 0.1 percent figure is Richard Saller’s rough estimate of the average annual rate of economic growth in the western Roman empire during the three centuries following 200  (Morris et al. 2007: 5). Temin (2013: 197) stresses the roughness of this figure. ¹⁰⁴ In Pompeii, local notables similarly financed the erection of statues in public places (Beard 2008: 197). ¹⁰⁵ Champlin (1991: 158–9) describes bequests to fund public buildings and services.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 199 individual status enhancement), and toward the financing of public goods (potentially a positive-sum game).

17.4.3. Rome’s Relative Resistance to Populist Policies Ancient regimes in the Near East and elsewhere commonly adopted measures that were likely to generate immediate popular acclaim, but impair long-term economic growth.¹⁰⁶ Three examples of these populist policies were edicts cancelling debts, rent control measures, and conferrals on land sellers of immutable rights to repurchase (redeem) transferred land. Rome’s officials certainly succumbed at times to populist political pressures (Lévy 1967: 55–8, 68–9).¹⁰⁷ But, particularly after the ascension of Augustus, Roman officials were relatively resistant to them.¹⁰⁸ On assuming office, a king of an ancient Near Eastern empire commonly proclaimed a partial or complete cancelation of debts owed to private creditors (Ellickson and Thorland 1995: 401–4; Westbrook 2009b: 151–60). Although applauded by debtors, a debtcancelation decree (misharum) typically would have counterproductive effects on financial markets in the long term.¹⁰⁹ A potential lender aware of the risk of cancelation might either refuse to lend, or charge a higher interest rate. At minimum, when negotiating loan terms, a lender and borrower uncertain about the timing of the next debt cancelation would incur higher transaction costs. In 86 , in response to shortages of coinage, the Roman Republic approved a law cancelling 75 percent of outstanding debts, and, in 49 , another law cancelling roughly 25 percent (Harris 2007: 519). Julius Caesar ordered a partial remission of rents in 48  (Frier 1977: 34), prompting Finley (1999: 143) to speculate that Caesar’s opponents worried that he would cancel yet more debts. In 41 , three ¹⁰⁶ On the prevalence of these policies in even more ancient regimes, see Ellickson and Thorland (1995: 400–8). ¹⁰⁷ In the late Republic there were efforts to limit interest rates through usury laws (Harris 2007: 520). And Tiberius and Nero made sporadic efforts to control the price of grain (Temin 2013: 33–5). ¹⁰⁸ In 58 , the tribune Clodius, the late Republic’s most notable populist, helped institute what became a long-lived, but increasingly restricted, policy of doles of free wheat to adult male citizens resident in Rome (Lo Cascio 2007: 639–41). ¹⁰⁹ A debt-cancelation measure that cooled popular wrath conceivably might have aided an ancient economy by preventing an even more destructive event, such as a bloody revolution or civil war (cf. Roe 1998).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

200

Robert C. Ellickson

years after Caesar’s assassination, the Second Triumvirate ordered yet another rent remission (Frier 1977: 36). After Augustus assumed the emperorship, however, for several centuries there appear to have been no debt-cancelation decrees. By enabling developers and buyers of apartment blocks to obtain loans on more favorable terms, this abstention would have contributed to the growth of the city of Rome. Rent controls are embraced by many twenty-first-century states. The details of these measures vary, and so do, accordingly, the inefficiencies that they cause. Crude controls on urban rents severely crimp both the construction of new private housing and the maintenance of existing housing. Rent ceilings also can reduce household mobility and produce mismatches between households and dwellings (Glaeser and Luttmer 2003). Other than the aforementioned remissions of rents in 48 and 41 , the city of Rome appears to have been free of legislated rent controls (Frier 1977: 36).¹¹⁰ The notion of a “just price” was not formally recognized in Roman law until the late Empire (Gordley 1981: 1601–2). During the period of Rome’s growth, market prices generally were regarded as fair prices (Lo Cascio in volume I). A judge therefore would have rebuffed a tenant’s complaint that a negotiated rent had been pegged too high. Prior to 200 , states in China and the Near East commonly provided a land seller, and his heirs, an unwaivable entitlement to repurchase sold land at the original purchase price (Westbrook 2009b: 148–51; Ellickson 2012).¹¹¹ This redemption right would have helped keep ancestral land, perhaps mortgaged during a period of financial distress, within a family. In some contexts, these rights might have generated net economic benefits by preserving local social networks and providing informal social insurance (Ellickson 2012: 294). In a large city such as Rome, however, redemption rights on balance would have fouled land markets. They would have made, for example, a potential buyer of a lot suited for a new apartment block more wary of purchasing, and after purchase, of building.

¹¹⁰ Centuries later, emperors Constantine and Justinian attempted to peg rents of agricultural lands to customary levels (Kehoe 2007: 134–5). ¹¹¹ Although Roman law did not create implicit redemption rights, it permitted parties to create a repurchase right by express agreement. See D. 18.1.75 (Hermogenian 2 iuris epit.); D. 19.1.21.5 (Paul. 33 ad ed.); D. 19.5.12 (Proculus 11 epist.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 201 Roman law, in contrast to the law of ancient China, Israel, and several Near Eastern states, appears never to have conferred immutable redemption rights.¹¹² Yet another populist bullet thus was dodged.¹¹³

17.5. CONCLUSI ON The tools of law and economics generate insights into the causes and effects of the unprecedented growth of the ancient city of Rome. Cities historically have been centers of creativity, and urban economics helps reveal why this has been so. Rome was the indispensable nerve center not only of the empire’s predations, but also of its accomplishments. The economic perspective helps elucidate the many services that Rome exported, and why talented individuals from around the Mediterranean flocked to it. The theory of public goods similarly illuminates the city of Rome’s division of responsibilities between its public and private sectors. A law-and-economics perspective provides a useful prism, although hardly the only one, on the merits of specific institutions. By present-day standards, many of Rome’s core policies were deeply flawed. The city’s governance system was clumsy. Legal barriers kept women from fully developing their talents. Slavery, and the many other formal status distinctions that pervaded Roman culture, tended to impair social and economic mobility.¹¹⁴ The institutions governing real estate transactions—the focus of much of this chapter—also had

¹¹² Westbrook (1989: 208–9) asserts that mancipatio was a procedure designed to extinguish redemption rights, an entitlement that had “entirely disappeared” from Roman law by classical times. ¹¹³ Populist policies were more in vogue during the late Empire. For example, in  301, Diocletian’s Edict on Maximum Prices sought to regulate the prices of over a thousand commodities. Bartlett (1994) asserts that unbridled statism of this stripe contributed significantly to the decline and fall of the Western Empire. ¹¹⁴ As a general matter, Roman slavery was less harsh than the subsequent slavery systems of the post-conquest Americas (Temin 2013: 114–38). In exceptional instances, some freeborn individuals in Rome voluntarily entered slavery to obtain managerial positions that otherwise would have been unavailable (Dari-Mattiacci 2013: 98).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

202

Robert C. Ellickson

shortcomings. Rome lacked a reliable system of land records and a condominium form of ownership. The state’s power of compulsory purchase (eminent domain) was ambiguous.¹¹⁵ And Rome’s officials at times baldly confiscated real estate, such as Cicero’s house on the Palatine (Hales 2000: 45–6).¹¹⁶ A growing city requires the construction of buildings, and confiscations hardly foster their production. Measured against the practices of other ancient civilizations, however, Rome’s land institutions were relatively growth-promoting. Provisions of the Twelve Tables signaled that the people of the early Republic, despite their roots in a tribe-based culture, were uncommonly committed to putting land into commerce. Embracing the foundations of a market economy, the Romans recognized private property in land, labor, and capital, and also conferred on an owner broad discretion to transfer an asset by sale, lease, or will. In addition, the Romans came to be relatively resistant to populist enactments. Their embrace of this cluster of policies was a necessary, although hardly a sufficient, condition for the transformation of a polity in the backwater of Latium into a city that coordinated the conquest of the entire Mediterranean. An intriguing question is why the descendants of the members of tribes residing along the left bank of the Tiber had initially opted for this relatively radical set of policies. On that issue, classicists steeped in Roman culture will have ideas far better than mine.¹¹⁷

¹¹⁵ In 1625 in De jure belli et pacis, Hugo Grotius coined the phrase dominium eminens (supreme lordship) to describe this power. Roman lawyers appear never to have crisply conceptualized the issue. Crook (1967: 262–4) asserts that public authorities only “sparingly” forced land transfers, and, when they did, typically provided at least partial compensation; see also Reynolds (2010). Taylor (2000: 93–127), by contrast, argues that Roman governments lacked the power of compulsory purchase, but were willing to confiscate an owner’s assets to punish a perceived misdeed. ¹¹⁶ Land could be confiscated piecemeal by court order or imperial fiat, or en masse as a result of military action. In the latter case, the land became ager publicus, a spoil available for grant or subsidized lease to soldiers or others (Finley 1999: 119). On land confiscations in societies more ancient than Rome’s, see Ellickson and Thorland (1995: 345–6). ¹¹⁷ For comments and other help, I thank Clifford Ando, Jean Andreau, Emmanuelle Chevreau, Cyril Courrier, Giuseppe Dari-Mattiacci, Adriaan Lanni, Joseph Manning, David Schleicher, Steven Shavell, Pierre Vesperini, James Whitman, and the two reviewers. Responsibility for errors is entirely mine. Zachary Herz provided essential research assistance, and Sarah Kraus was indispensable in locating sources.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 203 REFERENCES Ades, Alberto F., and Edward L. Glaeser. 1995. “Trade and circuses: explaining urban giants.” 110 Quarterly Journal of Economics 195–227. Alonso, William. 1964. Location and Land Use: Toward a General Theory of Land Rent. Cambridge, MA: Harvard University Press. Ando, Clifford. 2003. “Review of Being Greek under Rome by Simon Goldhill.” 57 Phoenix 355–60. Andreau, Jean. 1999. Banking and Business in the Roman World. Cambridge: Cambridge University Press. Angel, Shlomo. 2012. Planet of Cities. Cambridge, MA: Lincoln Institute of Land Policy. Aristotle. 1946. Politics, trans. Ernest Barker. Oxford: Clarendon Press. Arruñada, Benito. 2012. Institutional Foundations of Impersonal Exchange: Theory and Practice of Contractual Registries. Chicago and London: University of Chicago Press. Baird, Douglas, and Thomas Jackson. 1984. “Information, uncertainty, and the transfer of property.” 13 Journal of Legal Studies 299–320. Bannon, Cynthia. 2009. Gardens and Neighbors: Private Water Rights in Roman Italy. Ann Arbor: University of Michigan Press. Bartlett, Bruce. 1994. “How excessive government killed ancient Rome.” 14 Cato Journal 287–303. Beard, Mary. 2008. The Fires of Vesuvius: Pompeii Lost and Found. Cambridge, MA: Harvard University Press. Benevolo, Leonardo. 1980. The History of the City, trans. Geoffrey Culverwell. Cambridge, MA: MIT Press. Berger, Adolf. 1953. “Encyclopedic dictionary of Roman law.” 43 (New Series) Transactions of the American Philosophical Society 333–809. Bowman, Alan, and Andrew Wilson, eds. 2011a. Settlement, Urbanization, and Population. Oxford: Oxford University Press. Bowman, Alan, and Andrew Wilson, eds. 2011b. “Introduction,” in Alan Bowman and Andrew Wilson eds, Settlement, Urbanization, and Population. Oxford: Oxford University Press, 1–14. Braginton, Mary V. 1944. “Exile under the Roman emperor.” 39 Classical Journal 391–407. Braund, Susan H. 1989. “City and Country in Roman Satire,” in Susan H. Braund, ed., Satire and Society in Ancient Rome. Exeter: Exeter University Publications, 23–47. Buckland, William Warwick. 1963. A Text-Book of Roman Law, ed. Peter Stein, 3rd rev. edn. Cambridge: Cambridge University Press. Calabresi, Guido, and A. Douglas Melamed. 1972. “Property rules, liability rules, and inalienability: one view of the cathedral.” 85 Harvard Law Review 1089.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

204

Robert C. Ellickson

Campbell, Brian. 2000. The Writings of the Roman Land Surveyors. London: Society for the Promotion of Roman Studies. Champlin, Edward. 1991. Final Judgments: Duty and Emotion in Roman Wills, 200 – 250. Berkeley: University of California Press. Chevallier, Raymond. 1976. Roman Roads, trans. N. H. Field. London: B. T. Batsford. Coase, Ronald H. 1960. “The problem of social cost.” 3 Journal of Law and Economics 1–44. Crook, John A. 1967. Law and Life in Rome. Ithaca: Cornell University Press. Crook, John A. 1976. “Classical Roman law and the sale of land,” in Finley, ed., 71–83. Dari-Mattiacci, Giuseppe. 2013. “Slavery and information.” 73 Journal of Economic History 69–116. Dubouloz, Julien. 2011. La propriété immoblière à Rome et en Italie (1er–Ve siècles). Rome: École Française de Rome. Duffy, John, and Daniela Puzzello. 2014. “Gift exchange versus monetary exchange: theory and evidence.” 104 American Economic Review 1735–76. Duncan-Jones, Richard P. 1976. “Some configurations of landholding in the Roman Empire,” in Finley, ed., 7–33. du Plessis, Paul. 2004. “The protection of the contractor in public works contracts in the Roman Republic and early Empire.” 25 Journal of Legal History 287–314. du Plessis, Paul. 2006. “Janus in the Roman law of urban lease.” 55 Historia 48–63. Ellickson, Robert C. 1991. Order without Law: How Neighbors Settle Disputes. Cambridge, MA: Harvard University Press. Ellickson, Robert C. 1993. “Property in land.” 102 Yale Law Journal 1315–1400. Ellickson, Robert C. 2006. “Unpacking the household: informal property rights around the hearth.” 116 Yale Law Journal 226–328. Ellickson, Robert C. 2010. “The false promise of the mixed-income housing project.” 57 UCLA Law Review 983–1021. Ellickson, Robert C. 2012. “The costs of complex land titles: two examples from China,” 1 Brigham-Kanner Property Rights Conference Journal, 281–302. Ellickson, Robert C. 2013. “The law and economics of street layouts: how a grid pattern benefits a downtown.” 64 Alabama Law Review 463–510. Ellickson, Robert C., and Charles Dia. Thorland. 1995. “Ancient land law: Mesopotamia, Egypt, Israel.” 71 Chicago-Kent Law Review 321–411. Engels, Donald. 1990. Roman Corinth: An Alternative Model for the Classical City. Chicago: University of Chicago Press. Erdkamp, Paul P. M. 2001. “Beyond the limits of the ‘consumer city’: a model of the urban and rural economy in the Roman world.” 50 Historia 332–56.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 205 Finley, Moses I. 1951. Studies in Land and Credit in Ancient Athens, 500–200 : The Horos Inscriptions. New Brunswick, NJ: Rutgers University Press. Finley, Moses I., ed. 1976a. Studies in Roman Property. Cambridge: Cambridge University Press. Finley, Moses I. 1976b. “Introduction,” in Finley, ed., 1–6. Finley, Moses I. 1977. “The ancient city: from Fustel de Coulanges to Max Weber and beyond.” 19 Comparative Studies in Society and History 305–27. Finley, Moses I. 1999. The Ancient Economy, ed. Ian Morris. Berkeley: University of California Press. Fischer, Claude S. 1982. To Dwell among Friends: Personal Networks in Town and City. Chicago: University of Chicago Press. Foster, Lynn, and J. Cliff McKinney, Jr. 2011. “Adverse possession and boundary by acquiescence in Arkansas.” 33 University of Arkansas at Little Rock Law Review 199–262. Frier, Bruce W. 1977. “The rental market in early imperial Rome.” 67 Journal of Roman Studies 27–37. Frier, Bruce W. 1980. Landlords and Tenants in Imperial Rome. Princeton: Princeton University Press. Frier, Bruce W., and Dennis P. Kehoe. 2007. “Law and economic institutions,” in Scheidel, Morris, and Saller, eds, 113–43. Frederiksen, Martin W. 1975. “Theory, evidence and the ancient economy.” 65 Journal of Roman Studies 164–71. Gabaix, Xavier. 1999. “Zipf ’s law for cities: an explanation.” 114 Quarterly Journal of Economics 739–67. Garnsey, Peter. 1976. “Urban property investment,” in Finley, ed., 123–36. Glaeser, Edward. 1998. “Are cities dying?” 12 Journal of Economic Perspectives 139–60. Glaeser, Edward. 2011. Triumph of the City. New York: Penguin Press. Glaeser, Edward L., and Erzo F. P. Luttmer. 2003. “The misallocation of housing under rent control.” 93 American Economic Review 1027–46. Goodman, Penelope J. 2007. The Roman City and its Periphery. London: Routledge. Gordley, James. 1981. “Equality in exchange.” 69 California Law Review 1587–1656. Hales, Shelley. 2000. “At home with Cicero.” 47 Greece & Rome (2nd Series) 44–55. Hansmann, Henry B., Reinier Kraakman, and Richard Squire. 2006. “Law and the rise of the firm.” 119 Harvard Law Review 1333–1403. Harris, William V. 2006. “A revisionist view of Roman money.” 96 Journal of Roman Studies 1–24. Harris, William V. 2007. “The late Republic,” in Scheidel, Morris, and Saller, eds, 511–39.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

206

Robert C. Ellickson

Hawkins, Cameron. 2012. “Manufacturing,” in Scheidel, ed., 175–94. Hazel, John. 2001. Who’s Who in the Roman World. London: Routledge. Holleran, Claire. 2012. Shopping in Ancient Rome. Oxford: Oxford University Press. Honoré, Tony. 1989. “Conveyances of land and professional standards in the later empire,” in Peter Birks, ed., New Perspectives in the Roman Law of Property. Oxford: Clarendon Press, 137–52. Hopkins, Keith. 1980. “Taxes and trade in the Roman empire (200 ..–.. 400).” 70 Journal of Roman Studies 101–25. Hopkins, Keith. 2002. “Rome, taxes, rents and trade,” in Walter Scheidel and Sitta von Reden, eds, The Roman Economy. New York: Routledge, 190–230. Hornblower, Simon, and Tony Spawforth, eds. 2000. Who’s Who in the Classical World. Oxford: Oxford University Press. Jacobs, Jane. 1969. The Economy of Cities. New York: Random House. Jasnow, Richard. 2003. “Egypt: New Kingdom,” in Raymond Westbrook, ed., A History of Ancient Near Eastern Law, vol. 1. Leiden: Brill, 289–359. Jongman, Willem. 1988. The Economy and Society of Pompeii. Amsterdam: J. C. Gieben. Juvenal. 1974. The Sixteen Satires, trans. Peter Green, rev. edn. Harmondsworth: Penguin Books. Kehoe, Dennis P. 2007. Law and Rural Economy in the Roman Empire. Ann Arbor: University of Michigan Press. Kehoe, Dennis P. 2012. “Contract labor,” in Scheidel, ed., 114–30. Kehoe, Dennis P. 2016. “Tenure of land and agricultural regulation,” in P. J. du Plessis, C. Ando, and K. Tuori, eds, Oxford Handbook of Roman Law and Society. Oxford: Oxford University Press, 646–59. Kelly, Daniel B. 2013. “Restricting testamentary freedom: ex ante versus ex post justifications.” 82 Fordham Law Review 1125–85. Kirschenbaum, Aaron. 1987. Sons, Slaves and Freedmen in Roman Commerce. Jerusalem: Magnes Press. Laurence, Ray. 1997. “Writing the Roman metropolis,” in Helen M. Parkins, ed., Roman Urbanism: Beyond the Consumer City. London: Routledge, 1–20. Lévy, Jean-Philippe. 1967. The Economic Life of the Ancient World, trans. John G. Biram. Chicago: University of Chicago Press. Lo Cascio, Elio. 2007. “The early Roman empire: the state and the economy,” in Scheidel, Morris, and Saller, eds, 619–47. Love, John R. 1991. Antiquity and Capitalism: Max Weber and the Sociological Foundations of Roman Civilization. London: Routledge. Machado, Carlos. 2012. “Review of La propriété immobilière à Rome et en Italie (Ier–Ve siècles) by Julien Dubouloz.” Bryn Mawr Classical Review 2012.07.36.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 207 Macnair, Mike. 1999. “Vicinage and the antecedents of the jury.” 17 Law and History Review 537–90. Magill, Frank N., ed. 1998. Dictionary of World Biography, vol. 1: The Ancient World. Chicago: Fitzroy Dearborn Publishers. Martini, Pierluigi. 1976. “A review of the history of the aqueducts of Rome.” 68 Journal of the American Water Works Association 560–7. Mayer, Ernst Emanuel. 2012. The Ancient Middle Classes: Urban Life and Aesthetics in the Roman Empire, 100 –250 . Cambridge, MA: Harvard University Press. McAdams, Richard H. 1997. “The origin, development, and regulation of norms.” 96 Michigan Law Review 338–433. Millar, Fergus. 1977 [repr. 1991]. The Emperor in the Roman World (31 – 337). London: Duckworth. Morley, Neville. 1996. Metropolis and Hinterland: The City of Rome and the Italian Economy 200 – 200. Cambridge: Cambridge University Press. Morris, Ian. 1999. “Foreword to updated edition,” in Moses I. Finley, The Ancient Economy, updated edn. Berkeley: University of California Press, ix–xxxvi. Morris, Ian, Richard P. Saller, and Walter Scheidel. 2007. “Introduction,” in Scheidel, Morris, and Saller, eds, 1–12. Mumford, Lewis. 1961. The City in History. New York: Harcourt, Brace & World. Musgrave, Richard A. 1959. Theory of Public Finance: A Study in Public Economy. New York: McGraw-Hill. Natelson, Robert G. 1987. “Comments on the historiography of condominium: the myth of Roman origin.” 12 Oklahoma City Law Review 17–58. Nicholas, Barry. 2008. An Introduction to Roman Law. Oxford: Oxford University Press. Packer, James E. 1971. The Insulae of Imperial Ostia. Memoirs of the American Academy in Rome, vol. 31. Ann Arbor: University of Michigan Press. Parkins, Helen. 1997. “The ‘consumer city’ domesticated? The Roman city in élite economic strategies,” in Helen M. Parkins, ed., Roman Urbanism: Beyond the Consumer City. London: Routledge, 83–111. Peña, J. Theodore. 2007. Roman Pottery in the Archaeological Record. Cambridge: Cambridge University Press. Pirson, Felix. 1997. “Rented accommodation at Pompeii,” in Ray Laurence and Andrew Wallace-Hadrill, eds, Domestic Space in the Roman World: Pompeii and Beyond. Journal of Roman Archaeology Suppl. 22. Portsmouth: Journal of Roman Archaeology, 165–81. Polanyi, Karl. 1944. The Great Transformation. New York: Rinehart. Rabin, Edward H. 1984. “The revolution in residential landlord-tenant law: causes and consequences.” 69 Cornell Law Review 517–84.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

208

Robert C. Ellickson

Rawson, Elizabeth. 1976. “The Ciceronian aristocracy and its properties,” in Finley, ed., 85–102. Reynolds, Susan. 2010. Before Eminent Domain: Toward a History of Expropriation of Land for the Common Good. Chapel Hill: University of North Carolina Press. Robinson, Olivia F. 1992. Ancient Rome: City Planning and Administration. London: Routledge. Rodger, Alan. 1972. Owners and Neighbors in Roman Law. Oxford: Clarendon Press. Rodgers, Robert H., ed., 2004. Frontinus, De Aquaeductu Urbis Romae. Cambridge: Cambridge University Press. Roe, Mark J. 1998. “Backlash.” 98 Columbia Law Review 217–41. Rosenstein, Nathan 2008. “Agriculture and aristocrats in the middle and late Republic.” 98 Journal of Roman Studies 1–26. Rosenthal, Stuart S., and William C. Strange. 2004. “Evidence on the nature and sources of agglomeration economies,” in J. Vernon Henderson and Jacques-François Thisse, eds, Handbook of Regional and Urban Economics, vol. 4. Amsterdam: Elsevier, 2119–71. Saller, Richard. 2012. “Human capital and economic growth,” in Scheidel, ed., 71–86. Saller, Richard P. 2007. “Household and gender,” in Scheidel, Morris, and Saller, eds, 87–112. Samuelson, Paul A. 1954. “The pure theory of public expenditure.” 36 Review of Economics and Statistics 387–9. Scheidel, Walter. 1999. “Emperors, aristocrats, and the grim reaper: towards a demographic profile of the Roman elite.” 49 Classical Quarterly 254–81. Scheidel, Walter. 2005. “Human mobility in Roman Italy, II: the slave population.” 95 Journal of Roman Studies 64–79. Scheidel, Walter. 2007. “Demography,” in Scheidel, Morris, and Saller, eds, 38–86. Scheidel, Walter, ed. 2012a. The Cambridge Companion to the Roman Economy. Cambridge: Cambridge University Press. Scheidel, Walter. 2012b. “Slavery,” in Scheidel, ed., 89–113. Scheidel, Walter, and Steven J. Friesen. 2009. “The size of the economy and the distribution of income in the Roman empire.” 99 Journal of Roman Studies 61–91. Scheidel, Walter, Ian Morris, and Richard Saller, eds. 2007. The Cambridge Economic History of the Greco-Roman World. Cambridge: Cambridge University Press. Schleicher, David. 2010. “The city as a law and economic subject.” 2010 University of Illinois Law Review 1507–63. Schneider, Helmut. 2007. “Technology,” in Scheidel, Morris, and Saller, eds, 144–71.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Legal Foundations of the Growth of an Indispensable City 209 Scobie, Alex. 1986. “Slums, sanitation and mortality in the Roman world.” 68 Klio 399–433. Scogin, Hugh T., Jr. 1990. “Between heaven and man: contract and state in Han Dynasty China.” 63 Southern California Law Review 1325–1404. Seneca. 2007. Dialogues and Essays, trans. John Davie. Oxford: Oxford University Press. Smith, Adam. 1937. The Wealth of Nations. New York: Modern Library. Storey, Glenn R. 2002. “Regionaries-type insulae 2: architectural/residential units at Rome.” 106 American Journal of Archaeology 411–34. Taylor, Rabun. 2000. Public Needs and Private Pleasures: Water Distribution, the Tiber River, and the Urban Development of Ancient Rome. Rome: “L’Erma” di Bretschneider. Temin, Peter. 2013. The Roman Market Economy. Princeton: Princeton University Press. Tolley, George S. 1974. “The welfare economics of city bigness.” 1 Journal of Urban Economics 324–45. Treggiari, Susan. 1999. “The upper-class house as symbol and focus of emotion in Cicero.” 12 Journal of Roman Archaeology 33–56. Turchin, Peter, and Walter Scheidel. 2009. “Coin hoards speak of population declines in ancient Rome.” 106 PNAS 17276–9. Wallace-Hadrill, Andrew. 1991. “Elites and trade in the Roman town,” in John Rich and Andrew Wallace-Hadrill, eds, City and Country in the Ancient World. London: Routledge, 241–72. Wallace-Hadrill, Andrew. 1994. Houses and Society in Pompeii and Herculaneum. Princeton: Princeton University Press. Ward-Perkins, Bryan. 2005. The Fall of Rome and the End of Civilization. New York: Oxford University Press. Watson, Alan. 1971. Roman Private Law around 200 . Edinburgh: Edinburgh University Press. Watson, Alan, ed. 1985. The Digest of Justinian. Philadelphia: University of Pennsylvania Press (available for download at Project Muse, http://muse. jhu.edu/). Watson, Alan, ed. 1995. The Spirit of Roman Law. Athens: University of Georgia Press. Westbrook, Raymond. 1989. “Restrictions on alienation of property in early Roman law,” in Peter Birks, ed., New Perspectives in the Roman Law of Property. Oxford: Clarendon Press 207–13. Westbrook, Raymond. 2003a. “Introduction: the character of ancient NearEastern law,” in Raymond Westbrook, ed., A History of Ancient Near Eastern Law, vol. 1. Leiden: Brill, 1–90. Westbrook, Raymond. 2003b. “Mesopotamia: Old Babylonian period,” in Raymond Westbrook, ed., A History of Ancient Near Eastern Law, vol. 1. Leiden: Brill, 361–430.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

210

Robert C. Ellickson

Westbrook, Raymond. 2009a. “The nature and origins of the Twelve Tables,” in Bruce Wells and Rachel Magdalene, eds, Law from the Tigris to the Tiber: The Writings of Raymond Westbrook, vol. 1. Winona Lake: Eisenbrauns, 21–71. Westbrook, Raymond. 2009b. “Social justice in the ancient Near East,” in Bruce Wells and Rachel Magdalene, eds, Law from the Tigris to the Tiber: The Writings of Raymond Westbrook, vol. 1. Winona Lake: Eisenbrauns 143–60. Whitman, James Q. 1996. “The moral menace of Roman law and the making of commerce: some Dutch evidence.” 105 Yale Law Journal 1841–89. Wigmore, John H. 1941. “The pledge-mortgage idea in Roman law: a revolutionary interpretation.” 36 Illinois Law Review 371–93. Wilson, Andrew. 2011. “City sizes and urbanization in the Roman empire,” in Bowman and Wilson, eds, 161–95. Wirth, Louis. 1938. “Urbanism as a way of life.” 44 American Journal of Sociology 1–24. Wittfogel, Karl. 1957. Oriental Despotism: A Comparative Study of Total Power. New Haven: Yale University Press. Yavetz, Zvi. 1958. “The living conditions of the urban plebs in republican Rome.” 17 Latomus 500–17.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

18 Land Demarcation in Ancient Rome Gary D. Libecap and Dean Lueck

18.1. INTRODUCTION Ancient Rome was a vast social network—lasting nearly one thousand years—that created wealth by facilitating specialization and trade at the level not seen in the world before or for many centuries afterward.¹ At its height the Empire comprised over 5 million square kilometers, or roughly three-fourths of the land area of the continental US and half of land area of the European continent. For centuries, the population of the city of Rome was about 1,000,000.² After the (Western) Empire fell in the fifth century CE Rome’s population plummeted to 20,000 and did not again reach 1,000,000 until the twentieth century.³ The networks that facilitated wealth generation included common languages and a common currency, a large standing army (the legions), a large navy, and a system of law (and public administration). The Roman Empire also included a system of roads (and

¹ Temin (2006, 2013) provides evidence of wealth and economic growth during the Empire period and also provides evidence of a wide market for wheat across the Empire. Arruñada (2012) discusses the importance of contractual registries in fostering networks of impersonal trade, although the evidence for Roman registries is not clear. ² Alexandria and Carthage were the second and third largest cities in the Roman world and had populations perhaps as high as 500,000 and 300,000, respectively. ³ We follow the modern scholarly convention of using CE for “common era” rather than AD, and BCE for “before common era” for BC. Gary D. Libecap and Dean Lueck, Land Demarcation in Ancient Rome In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0018

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

212

Gary D. Libecap and Dean Lueck

shipping routes) and property rights in land that promoted a vast market in agricultural products that not only sustained a wealthy capital city but also a population that, on a conservative estimate, reached a maximum of about 65 million in the second century CE, roughly 15 percent of the entire world population (Temin 2013). Indeed much of the Empire’s agricultural production came not from the area near Rome, but from North Africa, the Po Valley of northern Italy, and from Spain and France. A key component of Rome’s agricultural productivity was its sophisticated system of land demarcation known as centuriation. Centuriation was the demarcation of large tracts of land into a collection of uniform squares (rectangles in some cases) with precise borders surveyed by highly trained professionals. Vast tracts were demarcated by Roman surveyors (called agrimensores, literally meaning land measurers) in North Africa and the Po Valley. This demarcation typically followed the conquest or annexation of new territory. Once the land was demarcated it was often colonized by Roman settlers, who might also serve as soldiers, and put to agricultural use. Land demarcation has been dominated by indiscriminate or unsystematic patterns of irregularly shaped plots, such as metes and bounds (Thrower 1966; McEntyre 1978; Linklater 2002; Libecap and Lueck 2011a, 2011b; Libecap, Lueck, and O’Grady 2011). This type of demarcation includes nearly all of Europe and the eastern United States (Thrower 1966; Estopinal 1998; Linklater 2002). Metes and bounds (MB) systems demarcate boundaries in terms of natural features of the land (e.g. rivers, cliffs, beaches) and even some relatively permanent human structures (e.g. bridges, walls). The most famous deviation from metes and bounds is the American rectangular system (RS) established in 1785, now governing the vast majority of land in the U.S. (Estopinal 1998; Libecap and Lueck 2011a, 2011b; Linklater 2002). This was part of a general effort to implement the RS in the British Empire, the largest empire to follow that of Rome (Libecap, Lueck, and O’Grady 2011). The Roman system of land demarcation was a purposeful, highly centralized and coordinated effort to standardize land into welldefined and enforced parcels. It was the first widespread system of rectangular demarcation. Under centuriation land was typically demarcated land into square units called centuriae quadratae, with sides of 710 meters (Bradford 1957). Each of these squares was comprised of 100 square heredia (Johnson 1976). Typically, at the

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

213

center of the centuria a north–south axis intersected an east–west line, making four quarters.⁴ Despite having been established two millennia ago centuriation still persists in the landscape throughout the Roman world. It is found, for example, in northern Italy, especially in the Po River Valley, at Braga in Portugal, Chester in England, Tarragona and Merida in Spain, Cologne and Trier in Germany (Stanislawski 1946), and Carthage in Tunisia. Figure 18.1 shows a satellite image of the rectangular centuriation system still present and dominant near the town of Cesena, Italy (northeast Italy). The image is dominated by squares but the left-hand side of the figure and the far upper-right corner also shows land demarcated under a nonrectangular system. This chapter examines how the centuriation system of land demarcation was used and implemented as part of the great Roman

Figure 18.1. Cesena, Italy (44 N, 12 E), showing RS demarcation persistence. Map data: Google, DigitalGlobe. Image © 2013 DigitalGlobe.

⁴ Unlike the modern US system, however, the Roman system did not cover contiguous stretches of land, but was established at various new cross-points and thus varied somewhat with natural land features. And as we note below there was considerable variation over time and across space as well.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

214

Gary D. Libecap and Dean Lueck

network. We merge the economics of property rights with the economics of networks to examine the establishment and evolution of the Roman centuriation.⁵ Our approach predicts that centuriation will be adopted in those areas where the potential value from settled agriculture and colonization will be largest. This suggests that Romans established the centuriation system in areas where potential land values were high and where the terrain was relatively flat so that the costs of implementing the system were relatively low. These are also the areas where the gains from coordinated demarcation to promote tillage, drainage, irrigation, land trading, and addressing for taxation and other purposes were high. We also examine how changes in political and bureaucratic institutions and markets for land and agricultural commodities, as well as improvements in survey techniques influenced Roman land demarcation and its effects. Until now historians and archaeologists have been the primary scholars of Roman centuriation and have by far the most welldeveloped theories to explain the practice. For example, Larsen (1938) argued that the practice was derived from political goals to increase tax revenue and establish colonies. Alternatively Stambaugh (1988) discussed squares as religious artifacts. Our approach is purely economic, although consideration of political forces will be important in the empirical analysis because of their impact on implementation costs. In this regard we differ from Roman legal scholars in that we do not rely on verification of motivation from primary sources, such as the writings of Roman jurists. Instead, we draw functional parallels between ancient and modern solutions to similar problems, and we use modern economic theory to inform the evidence on land demarcation that comes from a variety of sources. The chapter is organized as follows. We begin in §18.2 with a brief history of ancient Rome and its land demarcation systems. In §18.3 we develop an economic framework for analyzing the demarcation of land under the Roman system. §18.4 is our empirical analysis of the adoption of the system of centuriation in ancient Rome. In §18.5 we summarize our findings and discuss their implications.

⁵ The works of Acemoglu and Robinson (2012) and others in the literature on economic institutions and economic development are related though our approach is distinct.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

215

18.2. HISTORY OF LAND DEMARCATION IN ANCIENT ROME Classical scholars typically divide the history into three periods that coincide with three distinct political regimes: the Kingdom or Monarchy (753–509 BCE), the Republic (509–27 BCE), and the Empire (27 BCE–476 CE).⁶ The Republic was marked by a political system in which authority was given to elected magistrates and unelected senators. During this early period territorial expansion was moderate with Roman territory consisting of Italy, Spain, Sicily, and parts of Gaul, North Africa, and the Near East (Cornell and Matthews 1982). The start of the Imperial period marked an era of more aggressive territorial expansion with the Empire ultimately spanning about 6.5 million square kilometers (2.5 million square miles) at its peak under Emperor Trajan. In the later Empire the focus shifted from expansion to defense against invading enemies (Adkins and Adkins 2004). Diocletian first divided the Empire into an eastern and western half in 293 CE and it was permanently divided in 395 with the establishment of the Byzantine Empire and the Western Roman Empire. In 476 CE, Germanic invasions caused the fall of the Western Roman Empire,⁷ but the Byzantine Empire did not fall until 1453 when Constantinople was conquered by the Ottoman Turks. Figure 18.2 shows a map of the Roman Empire in the first century CE when the Empire was at it largest extent. The map reveals the division of the Empire into provinces such as Africa and the Hispaniae (Spain), in addition to Italia (Italy). Table 18.1 contains a summary of the history of Rome relevant for our study. For each century it shows the system of government, land area and colonization, population, and major events.

⁶ We rely on various general sources including Adkins and Adkins (2004). Temin (2013) also provides a concise economic history of ancient Rome. There is of course debate over the precise ending of the Empire and this date is specific to the Western Empire and does not include the Eastern Roman (Byzantine) Empire. ⁷ In the fall of 476 CE the German military office general Odoacer deposed the emperor Romulus Augustus and became the king of Italy.

Figure 18.2. Ancient Rome at its greatest extent under Trajan (c.117 CE).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

1,950,000 sq km (752,899 sq mi)

2,750,000 sq km (1,061,781 sq mi)

Republic

Republic and Empire

Empire

Empire

Empire

Empire

Empire

Empire

2 BCE

1 BCE

0

1 CE

2 CE

3 CE

4 CE

5 CE

NA

90,000–150,000

700,000–800,000

800,000

56,000,000# NA

NA

1,000,000

NA

1,000,000

NA

NA

6,944,000

4,063,000***

394,336

214,000 **

150,000

30,000

NA

NA

Population of city of Rome

Vandals invade Gaul Fall of (Western) Roman Empire

Theodosius the Great declares Christianity to be the sole religion of the empire Split in West and East Empire

Abandonment of Dacia

Conquest of Dacia

Roman invasion of Britain

Octavian named Augustus, first emperor of Rome, annexation of Syria

Third Punic War, Conquest of Macedonia, Africa, and Greece

Annexation of Corsica and Sardinia, First Punic War between Rome and Carthage, Second Punic War

Rome colonizes and conquers Italy

Foundation of Rome

Important Events

* These figures do not include the inhabitants of the Latin colonies nor of the allied states. ** The falling off from the number of the preceding census of 220 BCE was a result of the Hannibalic war. *** These figures and those of the enumerations for 8 BCE and 13 CE are from the Monumentum Ancyranum. The increased number are given by the census of 70 BCE over that of 115 BCE registers the result of the admission to the city of the Italians at the end of the Social war. Available on website http://www.tulane.edu.

4,400,000 sq km (1,698,849 sq mi)

5,000,000 sq km (1,930,511 sq mi)

4,200,000 sq km (1,621,629 sq mi)

1,200,000 sq km (463,323 sq mi)

Republic

3 BCE 262,322

165,000*

Republic Republic

10,000,000 sq km (3,861,000 sq mi) 360,000,000 sq km (138,997,000 sq mi) 800,000,000 sq km (308,882,000 sq mi)

NA

Kingdom

20,000

6 BCE

2,750,000 sq km (1,061,781 sq mi)

Kingdom

7 BCE

Total Population

5 BCE 4 BCE

Size of area

Government

Century

Table 18.1. History of ancient Rome.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

218

Gary D. Libecap and Dean Lueck

18.2.1. The Roman System of Centuriation The main source of understanding for Roman demarcation comes from the Corpus Agrimensorum Romanorum (Corpus), a collection of Roman surveyor manuals copied later into manuscripts by monks.⁸ It is the most important source of information on centuriation.⁹ The manuscripts that make up the Corpus were written by different authors through several historical periods. The most important manuscripts date from the sixth and ninth centuries CE (Dilke 1971: 17). The earliest technical writer of the Corpus is Sextus Julius Frontinus, governor of Britain from 74–8 CE.¹⁰ Below we discuss the key features of centuriation and Roman land demarcation.

18.2.1.1. The Basic System The process of centuriation is a system of land division that is built around two wider central lines or avenues called limites, the cardo maximus and decumanus maximus, and subsequent narrower limites, creating a grid pattern.¹¹ There is a sophisticated process of measuring and extensive boundary marking, with a system of measurements used for land and other assets. The basic unit of measurement for length was the Roman foot and the main unit of measurement for area was the actus, with a length of 120 Roman feet on each side. The length that constituted a Roman foot, however, changed over the course of Roman civilization.¹² Because the actus was based on the foot, inconsistencies are observed ⁸ We rely on Bradford (1957), Campbell (2000), and Dilke (1971) for our primary understanding of centuriation and Roman land institutions. ⁹ However, monks who copied the manuals had little understanding of the technical Latin used to describe Roman surveying procedures (Dilke 1971: 17). As a result, the contents of the Corpus are not always consistent. ¹⁰ Some of the manuscripts are accompanied by well-preserved and colored miniatures. These illustrations served as a tool for teaching Roman surveying students. Within the manuals, land disputes are addressed, as well as the technicalities and proper methods associated with measuring, marking, mapping, and allocating land. ¹¹ In Latin, cardo generally referred to a hinge (as in a door) or a pivot (as with a clamshell), so the application to demarcation and mapping as a north–south or main axis is intuitive. Among English writers cardo is sometimes spelled “kardo” (e.g. Dilke 1971). ¹² There were three different standards that defined a Roman foot: The early foot, the normal foot, and the late Roman foot, which came into use in the third century CE (Dilke 1971: 84). Each new foot was shorter than the previous one.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

219

in the size of the actus at different locations that were centuriated in different time periods (Dilke 1971: 85). The length of a standard 20 × 20 actus centuria measures 709.68 meters on a side using the normal foot. The average length of 20 actus in Italy has been calculated to be about 707 meters, with a range from 704 to 714 meters (ibid.).¹³ Table 18.2 defines the units used in the Roman measurement system. The actus is the basic unit of land area for the Roman land system, measuring 14,440 square Roman feet or approximately 0.312 acres. Two actus made up one iugerum and two iugera made up a heredium, which was commonly thought of as a reasonable parcel size for an individual. As Table 18.2 shows a heredium is equivalent to 1.246 acres. The 100 heredia grouped together gave the centuria its name. A standard centuria consisted of 20 × 20 actus, or 200 iugera. Sides of centuriae were always multiples of actus, which helps determine if traces of squares are a part of Roman centuriation; however, there were non-standard centuriae indicated from the Corpus, and from observation in the field. Figure 18.3 shows an actus and iugera within a heredium. Table 18.2. Roman measurements with US equivalents. Roman Measurement

Definition

US Equivalent

Pes

Early Roman foot Normal Roman foot Late Roman foot

11.70 in/29.73 cm 11.64 in/29.57 cm 11.58 in/29.42 cm

Passus

Paces, 5 Roman feet

58 in/147.85 cm

Mille passus Actus

The mile, 100 paces 14,400 sq Roman feet

4833.33 ft 13,920 sq ft/0.312 acres

Iugerum

2 sq actus, 28,800 sq Roman feet

27,840 sq ft/0.624 acres

Heredium

2 iugera

55,680 sq ft/1.246 acres

Centuria*

200 iugera, 100 heredia

124.6 acres

Saltus

Area of woods or pasture, or a continuous tract of land, 2  2 centuria or 5  5 centuria

500 acres or 3,115 acres

Source: Dilke 1971; Campbell 2000.

¹³ In Tunisia the average is about 708 m, but some areas have been measured as low as 703 m. The low measurements in Africa reflect the use of the later Roman foot. Some of the variations seen in Italy, however, may simply be due to errors in measuring (Dilke 1971: 85).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

220

Gary D. Libecap and Dean Lueck

ACTUS

HEREDIUM IUGERUM (2 sq. actus)

Panel A: Illustration of an actus

Panel B: Depiction of two iugeria within a heredium

Figure 18.3. Roman system of rectangular demarcation.

18.2.1.2. Orientation and Division The main lines of survey, the decumanus maximus (DM) and the cardo maximus (CM), were typically north–south and east–west, respectively, as calculated by a sundial (Dilke 1971: 86). In the minority of cases, centuriation was laid out on an alignment other than the standard N–S orientation. In such cases, alignment was oriented around existing main roads or geographical features. In a minority of cases, orientation of demarcated land did not follow these rules. Some surveyors were influenced by the length of the land, picking the direction where it was the longest to mark the DM.¹⁴ In the division of the measured land, there were limites (boundary lines) of different sizes. The DM was the widest, with the CM slightly narrower. At first, there were no set widths for the limites until Augustus standardized the system and established that the DM would be 40 feet in width and the CM 20 feet. Outer limites would be sighted at a distance of 2,400 Roman feet, or the length of 20 actus, in each direction from the main intersection. The inner limites can be referred to as limites intercisivi, and intersect a centuria into multiple plots. Every fifth main limes was made slightly wider so to ensure that it would become a usable road. These limites were called quintarii. Each of these limites was sighted at a right angle and cross-checked to ensure accuracy.

¹⁴ There is evidence that the origins of such orientations take their root from Etruscan practices (Dilke 1971: 87, 89), including the origin of the word cardo and many surveying practices.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

221

18.2.1.3. Allocation, Numbering, and Mapping Allocating centuriated land was one of the surveyor’s most important responsibilities. They were required to direct the settlers to their plots of land to avoid confusion and mistakes. To record the details of the allocation, surveyors were responsible for creating maps. Some maps were inscribed on wood tablets, other were inscribed on parchment or bronze. The government kept the maps of all divided and allocated lands. If there was ever a dispute, the sanctuarium (where the maps were kept) could be referenced. Figure 18.4 shows the centuriation system and also shows the US rectangular system developed more than two millennia later.¹⁵ In order to keep track of the allocations, it was important to have a numbering system. The system uses the following abbreviations: (1) DD, to the right of decumanus; (2) SD, to the left of decumanus; (3) UC, beyond cardo; and (4) CC, on the near side of cardo. Following each of the inscriptions would be a Roman numeral indicating the distance from either the CM or DM, depending on what boundary it refers to. As Figure 18.4 shows, intersection of the decumanus and cardo maximus (called the umbilicus) is not at the center of the settlement. In the Corpus, Hyginus Gromaticus prescribes that in centuriated colonies, the main intersection of the DM and the CM should be in the center. In practice, however, this layout was not common, except in North Africa (Dilke 1971: 88). If an existing road ran through the planned settlement, then this would often become the place of intersection (ibid.). In cases of rough terrain, or if ¹⁵ The American RS also begins with the establishment of an Initial Point with a precise latitude and longitude. Next, a Principal Meridian (a true north–south line) and a Baseline (an east–west line perpendicular to the meridian) are run through the Initial Point. On each side of the Principal Meridian, land is divided into square (six miles by six miles) units called townships. A tier of townships running north and south is called a “range.” Each township is divided into thirty-six sections; each section is one mile square and contains 640 acres. These sections are numbered 1 to 36 beginning in the northeast corner of the township. Each section can be subdivided into halves and quarters (or aliquot parts). Each quarter section (160 acres) is identified by a compass direction (NE, SE, SW, NW). Each township is identified by its location relative to the Principal Meridian and Baseline. For example, the Seventh Township north of the baseline and Third Township west of the First Principal Meridian would be T7N, R3W, First Principal Meridian. In this manner, properties are positioned relative to one another in a standardized way. Dilke (1971) discusses the likelihood that Thomas Jefferson, who was the prime architect of the US system, based it on the Roman system.

DD I CK II

SD I CK II

DD II CK II

DD II CK I

DD II VK I

DD II VK II

K AR D O

quintarius

MA XI MU S

Direction of Sighting

quintarius

8 17 20 29 32

7 18 19 30 31

5

6

Figure 18.4. Roman (left, a) and US (right, b) rectangular demarcation systems.

DD I CK I

DD I VK I

DD I VK II

SD I CK I

SD I VK I

SD I VK II

M A X I M U S

D E U C M A N U S 33

28

21

16

9

4

34

27

22

15

10

3

35

26

23

14

11

2

EXAMPLE 3

36

25

24

13

12

1

TOWNSHIP 1 NORTH. RANGE 1 WEST. OILA & SALT RIVER BASELINE & MCHIDIAN

SECTION DETAIL

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

223

centuriation was being implemented on an existing settlement, the surveyor would often prefer to make a fresh start at a point not far outside the existing or planned settlement (ibid. 88–9). In cases where mountains, rivers or some other major obstruction impeded centuriation, the intersection could be at a distant point (ibid. 89). Roman survey techniques and the administrative bureaucracy that made such sophisticated demarcation possible also gradually improved as the empire expanded. These allowed for more accurate measurement and extension of the RS. Roman survey and administration were unprecedented in their precision and scope in the Empire. With the fall of Rome many of these practices were lost until redeveloped and augmented in the late eighteenth and nineteenth centuries in Britain and the United States (Libecap, Lueck, O’Grady 2011).

18.2.1.4. Agrimensores and Surveying In the first century BCE, a surveyor was also referred to as metator or a mensor, both of which mean measurer (Dilke 1971: 37). Colonization under Julius Caesar led to early growth of surveying. The need for more land for citizens and veteran soldiers, especially in Augustus’ time, encouraged the natural growth and development of centuriation. It was under Augustus’ rule that the first steps leading to the creation of a large bureaucracy were taken (ibid.). Conquered land was resurveyed and measurements standardized; 40 feet for DM; 20 feet for CM; for other decumani and cardines 12 feet; and for subsidiary roads 8 feet. Augustus’ successors continued to settle veteran soldiers on newly acquired lands. After the Empire’s expansion halted less new land was added and fewer new soldiers required settlement. Accordingly, the need for land reallocation and the role of surveyors and administrators changed. With the development of centuriation and distribution procedures, there were also advancements in the surveying. At first, agrimensores were not hired, but instead offered their services and expertise as a favor. Any compensation that they received was optional. Later, as the quantity of new land grew, Roman influence spread, and the number of soldiers and settlers seeking parcels for settlement increased, the demand for surveyors rose. The Roman state began to view surveyors as professionals with certain expertise that could be held accountable. If a surveyor made an incorrect measurement, one could take legal action against him.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

224

Gary D. Libecap and Dean Lueck

During the imperial period, the establishment of new colonies was less common, so that the quantity of land needing to be surveyed began to decrease while the frequency of legal disputes increased. The rise in land disputes corresponds with population increase and growing land scarcity. As a result, the role of surveyors shifted to serving as judges or arbitrators in land law. When there was a land dispute, it was the provincial governor’s responsibility to send out surveyors, or they could be employed privately. Surveyors were expected to know the extensive details of land law, to analyze land disputes, and to give advice to judges. The two instruments commonly used and improved over time by surveyors were the groma and the perticae. The groma was mainly constructed out of wood, so there are limited remains of this tool that have survived to modern times. The Corpus does not give a picture of the groma, consequently, scholars have had limited resources to aid them in the creation of a model. The only example of what might be a true groma is one whose metal parts were located in Pompeii in 1912 (ibid. 15). The groma was a staff, around two meters long with a cross mounted at the top and, likely, a tripod or sharp point at its base (ibid. 69–70). The arms of the cross were wooden with metal sheeting. To prevent inaccuracy due to the wearing of wood, the arms were reinforced near the center of the cross by bronze angle brackets. A plumb line hung through holes at the end of each arm. Sighting was done by lining up opposite plumb bobs. Sighting using the plumb lines required that the cross be off-centered from the staff to avoid obstruction. To achieve this task, the cross was placed on a bracket which projected out from the top of the staff. The distance from the center of the cross to the staff was 23.5 cm (9.25 in). Once plumb bobs were aligned, sights were set on a second groma, positioned a specified distance away (such as one actus). Perticae were measuring rods that allowed surveyors to measure standardized distances (ibid. 73). Sights were set on a third groma the same distance away, making a right angle with the other two gromae, creating a square. The groma allowed for the measurement of straight lines, squares and rectangles as required by Roman surveying for RS. A surveyor with a groma, plumb-line level, sundial, perticae, and writing and drawing materials could trace out rectangular plots for land demarcation. Figure 18.5 shows how Roman scholars believe the groma was designed and used.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

225

Figure 18.5. Depiction of a groma and mensor (surveyor).

A necessary skill for the agrimensor was that of sighting and leveling through use of the groma. There were multiple cross-checks in order to ensure accuracy. Cross-checks were scheduled at every fifth intersection, or quintarius. If there was a defect in the groma, or the sighting, a wrong observation would be recognized and could be corrected by resurveying. Uneven land and rivers, however, made measuring rectangles more complex and the RS more costly.

18.2.1.5. Boundaries Boundaries were an important part of centuriation and were marked by stones. Land boundaries (termini) had religious connections for the Romans (Dilke 1971: 98). They were named for the god of boundary stones, Terminus (Adkins and Adkins 2004: 304). To move a boundary stone without permission was considered not only a civil offense but also a religious one. As noted above when boundary disputes occurred, the surveyor acted as a judge or arbitrator.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

226

Gary D. Libecap and Dean Lueck 18.3. ECONOMICS OF LAND DEMARCATION SYSTEMS

The discussion of the history of Roman land demarcation suggests a number of questions, including what determines when centuriation will be chosen and implemented; what determines the characteristics of a centuria; and what are the effects of centuriation on land use and land markets. We describe an economic framework to examine these and other related questions and derive implications for empirical analysis. Our approach is to distinguish the Roman centuriation from non-systematic or metes and bounds demarcation. We use RS to represent a centuriated system (RS for Roman system or rectangular system) and MB for non-Roman demarcation (or metes and bounds). In the context of Roman land, MB refers generally to a pre-existing, often uncoordinated local demarcation practice.¹⁶ We briefly examine the choice between RS and MB as well as the optimal configuration of land parcels, land disputes, land markets, land values, and long-term land value and use under both MB and RS.¹⁷

18.3.1. Demarcation Basics: Parcel Shapes, Sizes, Alignment, and Topography We begin by examining a simple decentralized demarcation problem. In our framework, the individual parcel demarcation decision will depend on both the expected value of plot productivity, demarcation costs, and the demarcation regime. The innate characteristics of the land will partly determine its productivity and the cost of demarcating each parcel. In particular, topography or ruggedness plays a critical role. More rugged terrain will have lower productivity and higher surveying and policing costs. The demarcation regime will constrain the size, shape, and alignment (i.e. spatial or directional orientation) of the parcels.¹⁸

¹⁶ See Dilke (1971) for a discussion of pre-Roman demarcation systems. ¹⁷ We draw on the analysis in Libecap and Lueck (2011b) and Libecap, Lueck, and O’Grady (2011). We later consider the choice of leaving a tract of land un-demarcated. ¹⁸ It is also likely the RS affected irrigation systems by lowering the costs of moving water and coordinating landowners, though we do not collect evidence on irrigation.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

227

We can thus define the net value of a parcel of land as a function of area, shape, perimeter, topographic features, and the demarcation regime. Similarly, demarcation costs for land parcels are a function of area, parcel perimeter, topography, and demarcation regime (RS or MB). Topography and the demarcation system will affect parcel size and perimeter choices, and parcel alignment. They will also directly affect both value and demarcation costs. Under each regime initial landholders pick a parcel size and shape to maximize land value. Under MB land claimants make these selections independently, based on their own net value calculations, while under RS parcel demarcation is predetermined and generally rectangular parcels are presented to claimants. Accordingly, the question facing Roman administrators was which demarcation system provided for the highest value of the land for claimants, and hence, highest productive value for the Empire. Once chosen, the demarcation system was implanted on lands acquired by the expanding Empire. The next section examines the implications of these demarcation choices as well as the implications for land markets and values generated by the different demarcation regimes under varying topography. As we describe, a Roman administrator may choose to impose RS or utilize the existing MB arrangement.

18.3.2. Initial Parcel Demarcation To consider the comparative implications of the two land demarcation systems, we make the following assumptions. First, we assume that under each exogenously imposed (by Rome) demarcation regime there is a large tract of land whose external boundary is enforced collectively or otherwise by the sovereign. Second, within this large tract, there is a group of non-cooperative agents who claim and enforce separate plots in order to maximize the value of their land, net of demarcation and enforcement costs. Each claimant can only choose and demarcate a single parcel. Within the external borders, there is no coordination or contracting among claimants.¹⁹

¹⁹ We assume simultaneous claiming and we ignore the optimal time to claim under first possession rules (Lueck 1995).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

228

Gary D. Libecap and Dean Lueck

18.3.2.1. Demarcation under Non-Roman Systems (MB) Under non-Roman local demarcation, or MB, each claimant chooses parcel size and the length of parcel boundaries to enforce in order maximize land value less demarcation costs. All parcels fill the new area acquired by the Empire. Under MB there are few central constraints on the size and shape of individual claims, and no mechanism to coordinate alignment of individual parcels. The solution to the parcel demarcation problem facing each claimant will depend on the structure of value and cost functions, and importantly, on how topography affects them. For example, rugged terrain will affect agricultural potential and boundary marking and observation costs. In the simple case in which claimants have the same productivity and the same enforcement costs, and value does not depend on topography or shape, the claimant’s problem is simply to minimize border demarcation and enforcement costs, constrained by the productivity of the land. Because we assume a total land constraint and that all land is to be included in tangent parcels, and we further note that rectilinear plots have important production advantages for agriculture and urban use (Lee and Sallee 1974; Amiama et al. 2008), we predict square parcels in flat areas for MB. Squares fill the interstitial space or gaps between parcels.²⁰ They have relatively low perimeter-to-area ratios and thus have low demarcation and enforcement costs. Survey and fencing (enclosure) costs are lower for plots with fewer angles and longer straight boundaries (Johnson 1976). Deviating from flat topography, however, will alter the net value function by affecting both land value and demarcation costs. If demarcation and enforcement costs (surveying-fencing-policing) and land value depend on terrain, borders will roughly follow topography under non-Roman demarcation (e.g. follow streams and ridge tops rather than cross them). We expect a pattern of parcel sizes and shapes that depends on the character of the land and of the potential of claimants (farming productivity, violence and monitoring

²⁰ Regular polygons maximize the area enclosed by a given perimeter and thus have the lowest p/a ratio for any n-sided polygon but only three regular polygons— triangles, rectangles (squares), and hexagons—can create patterns with a common vertex and have no interstitial space. Eliminating interstitial space means less open access waste and fewer conflicts in the future.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

229

productivity). Non-Roman or MB demarcation is flexible and allows parcels to be adapted to terrain in order to maximize land values. In addition, because MB does not coordinate or otherwise constrain the spatial structure of individual parcels, alignment will be solely the choice of the individual claimant. Under MB, the lack of coordination among claimants means that individual parcels will not have a common alignment. Even in the case where the land is flat and the optimal shapes are squares, there is no reason to expect a grid of squares. The coordinated grid is a public good that will not be provided individually, unless there is an authority coordinating the claimants’ actions, perhaps in the form of a developer or, in the case of Rome, a sovereign.

18.3.2.2. Demarcation under the Roman System (RS) Initial demarcation under RS is distinct from that under MB. The land is surveyed by the sovereign into squares prior to original allocation. Unlike MB, claimants under RS do not explicitly choose area and perimeter, but are constrained by the system so that the choice under RS is to pick a (square) parcel that maximizes difference between land value and acquisition costs. Demarcation costs are born initially by the sovereign and hence are zero for the claimant. Because, however, RS demarcation does not allow for deviation from squares, the value of parcels will decrease and demarcation costs will rise with more rugged topography.

18.3.3. Aggregation: Total Value and Land Markets The previous analysis considered initial demarcation of individual parcels under MB and RS. In this section we examine the net value of aggregated land under both systems and also examine how each system affects the market for land after initial demarcation.

18.3.3.1. Total Value of the Land in a System As above, we assume the region governed by either a pre-Roman MB or RS is a fixed area, totally partitioned into parcels. Some areas acquired by Rome might have had zero net values after consideration of agricultural productivity and demarcation costs, and hence

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

230

Gary D. Libecap and Dean Lueck

remained unclaimed and not demarcated. We incorporate a temporal dimension to account for difference in demarcation setup and continuation costs for the two regimes. Under MB the net value of the land is the sum of individual values and costs, less the continuing costs associated with individual parcel adjustments resulting from the lack of coordination. These include individual enforcement efforts, responding to parcel border disputes, search, and other costs associated with misaligned plots. Under MB there is a one-time demarcation cost born by each claimant. In addition, land demarcation and production begin immediately, and the continuing costs associated with MB are assumed to be increasing in proportion the size of the region, parcel size, topography, and to rise over time as these problems accumulate. The total value under RS has a different structure from MB because it is a coordinated, imposed regime. We assume the network effects of RS are such that any person’s or group’s use of the system also benefits others and that it further increases the incentive to participate (Baird, Gertner, and Picker 1994; Farrell and Klemperer 2007). The network benefits are the public goods of common addresses, survey coordination, and standardized, aligned and fixed parcel boundaries that reduce border disputes and expand the land market because parcels are of uniform size and shape and can be more easily located than with MB. These network and coordination benefits come, however, at the cost of a necessarily extensive system.²¹ Under RS there are upfront costs of design, survey, and controlling land access until demarcation is completed. Under these assumptions the total present value of the land in the RS region is a function of parcel square shapes, sizes, and perimeters; topography; and the discount rate. The RS system setup costs occur prior to claiming and use. Network effects are incorporated into the parcel value function, which is increasing in the number of parcels governed by the RS. RS system costs are increasing in proportion to the size of the area to be surveyed and demarcated, but at a decreasing rate, revealing network economies. RS costs are increasing because of topography and become greater than MB because of the squareparcel constraint. It is possible to extend the network effects by ²¹ Our MB–RS cost distinction is similar to Dixit’s (2003) distinction between local (informal) and large (formal-legal) trading systems, where the latter have greater setup costs like RS.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

231

assuming that system costs fall over time as RS is implemented in many other areas, while there are also improvements in survey technique and system administration. Under RS surveying occurs prior to parcel selection, so the time horizon for generating value from the land begins after land has been allocated.

18.3.3.2. Net Value of the Roman System This framework implies that it is efficient to implement RS when it provides greater economic value to Rome than does the existing local (MB) system. This will be the case in some settings but not in others. Here we focus on the determinants of the potential net gains of the RS relative to MB in order to derive testable implications for analysis. Total value of land under each system in a particular area is the sum of the values of the individual parcels. If the land is flat, then centralized RS parcel shapes and sizes are the same as would be chosen under decentralized MB, although alignment would differ and there would be no coordinated addressing under MB. The net gain of RS over MB for a particular region is determined by four factors: (1) the increase in land value under RS compared to MB from the time that the RS is implemented, arising from the productivity advantages of squares, addressing, alignment, and reduced parcel border disputes; (2) the foregone land rent during the setup period and setup costs during that period assumed by the sovereign; (3) the one time MB demarcation cost assumed by individuals; and (4) the continuing MB costs as described earlier. These four factors constitute the tradeoffs between RS and MB.²² The first focuses on the network gains from RS over MB; the second on the gains from MB that would be sacrificed during the period the RS is being implemented, in terms of output under MB and RS setup costs; the third is the foregone individual demarcation costs under MB not required under RS; and the fourth is the avoided continued costs of MB over the time horizon of Roman land control. Comparatives statics emerge: the net value of RS will increase in the size of the governed land area; increase in the expected time horizon for capturing net rents; decrease in the time of RS implementation facilitated by flat land; and increase with improvements in survey technology and ²² A similar analysis can be undertaken for the pre-Roman choice to demarcate using MB or to leave the land un-demarcated.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

232

Gary D. Libecap and Dean Lueck

administration that occurred over time. A number of predictions follow: Prediction 1: A rectangular system is more likely to be adopted when (a) Rome controls large tracts of land for RS implementation; (b) when the time horizon is longer; (c) when implementation can be rapid; and (d) when network effects are larger. Considering forces likely to change the model parameters can illuminate these predictions. For instance, more rugged topography would reduce net gains from RS by increasing the costs and time of RS implementation and perhaps even by reducing the losses of suboptimal parcel shape. Similarly, we expect that implementing RS in a region where no incumbent demarcation system existed would lower costs. There would be no resistance from incumbent land holders to the demarcation change and no need to move or replace such demarcation-specific investments as road, fences, buildings, and canals. Where the topography dictates benefits from coordinated drainage of land (e.g. a large flat swampy area), then RS is more likely to be implemented. Finally, political authority and stability will increase the expected time horizon and make RS adoption more likely.

18.3.3.3. Land Markets: Roman (RS) versus Non-Roman Demarcation (MB) RS has economic implications for land markets once it has been adopted. Because parcel boundaries are standardized and aligned, there are fewer overlapping borders and unclaimed gaps outside property descriptions. These factors imply another prediction: Prediction 2: There will be fewer legal disputes (and litigation) over boundaries under the RS than under non-Roman demarcation. Centralized RS demarcation with uniformly defined parcels and common addressing lowers the cost of using the market and allows for reorganization of plots as conditions change (Barzel 1982). This should be observed as a greater number of market transactions under RS than MB and an increase in the value of land.²³ The following predictions are implied by this analysis. ²³ Over time the Romans used different size distributions. Subsequent market activity, however, would lead to subdivision or aggregation of rectilinear plots to

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

233

Prediction 3: There will be higher land values under RS than MB.²⁴ Prediction 4: There will be more land transactions under RS than under MB and parcel size will have a lower variance under RS and MB. Prediction 5: There will be higher portion of land improved under RS than MB. The coordinated clarity of RS is also expected to have an impact on public infrastructure, such as roads, drainage ditches, and irrigation canals, that require long rights-of-way and coordination among owners. Contiguous linear borders should lower the cost of assembling rights of way and coordination along parcel boundaries.²⁵ This implies another prediction: Prediction 6: There will be more roads and canals for drainage and irrigation per unit of land under the rectangular system than under metes and bounds.

18.4. EMPIRICAL ANALYSIS In this section we use historical information and data on specific centuriation systems to test some of the predictions about the use and characteristics of centuriae and their effects on economic activity in ancient Rome. As we noted in the introduction, we interpret the evidence in light of our economic framework rather than look for verification in primary Roman legal sources.

18.4.1. Detailed Data on Centuria Table 18.3 shows currently available data from Roman Italia (Italy). The information includes the name and location of the centuriae. It achieve optimal production sizes and we argue that the RS would smooth these market transactions. ²⁴ As indicate above, we also predict that under RS land values will be decreasing in the ruggedness of the topography of the land and that there would be a break-even point beyond which the gains of the RS would be offset by the costs of rigidness. ²⁵ The value of straight roads and their benefits for surveyors and civil engineers is discussed by Johnson (1976: 167).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

234

Gary D. Libecap and Dean Lueck

Table 18.3. Characteristics of selected Roman centuriae in Italia. Roman name

date of Roman foundation

Pre-Roman Size of centuria settlement

Alignment

Aecae Alba Fucens Allifae Aquileia Aquinum Arimimium (Ariminum) Asculum Aug. Bagiennorum Augusta Praetoria Salassorum Augusta Taurinorum Belunum Beneventum Bononia Brixia Caesena Capua Cosa Cremona Fanum Ferentinum Florentia Formiae Grumentum lulia Concordia Lazio Luca Luceria Luni Minturnae Norba Ostia

214 BC 303 BC 241 BC 181 BC 100 BC 268 BC

yes yes

Via Traiana mountains

yes yes

269 BC 1 century BC 25 BC

yes

20  16 actus

yes yes

21  20 actus 20  20 actus 25  16 actus 20  20 actus

Parma Patavium Pisarum Pompeii Praeneste Puteoli Sena Gallica Surrentum Tarracina Venaf rum Venetia Verona Vicetia Source: see text.

28 BC 181 BC 268 BC 189 BC 1 century BC 3 century BC 312 BC 273 BC 218 BC 1 century BC 2 century BC 59 BC 3 century BC 3 century BC 42 BC 326 BC 180 BC 321 BC 177 BC 295 BC 2 century BC 338 BC 183 BC 45 BC 184 BC 4 century BC 90 BC 194 BC 283 BC 1 century BC 509 BC 2 century BC

yes yes yes

12  12 actus 21  20 actus

20  20 actus 20  20 actus 16  32 actus 21  20 actus

road

Via Aemilia North-South mountains coast line

20  20 actus

North-South

20  18 actus 80  16 actus yes

yes

193.94  125.70 meters 20  20 actus 20  20 actus

yes

18  23 actus

North-South river mountains river Via Aemilia North-South coast line

yes yes 16  16 actus

Via Appia Via Postumia

75–50 BC 135 BC

20  20 actus

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

235

Figure 18.6. Documented location of centuriae in ancient Rome. Source: Museo della Centuriazione.

also includes information about the size of the centuriae, the date of establishment, prior settlement, and the type of centuriation (town, colony, military). As noted above, prediction 3 states that large landowners or sovereigns are more likely to adopt a centralized rectangular system because it provides the public goods of systematic location of properties, coordinated survey, reduced title conflict, and greater infrastructure investment. Figure 18.6 shows documented centuriae based on information from Museo della Centuriazione Romana (Museum of Roman Centuriation) in Borgoricco, Italy.²⁶ Figure 18.7 shows the locations of centuriae in the Roman Italy. As Table 18.3 and Figures 18.6 and 18.7 suggest, the locations of centuriae can be determined and linked to data as described above.

²⁶ See http://www.euromuse.net/en/museums/museum/view-m/museo-dellacenturiazione-romana/ accessed February 5, 2014.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

236

Gary D. Libecap and Dean Lueck

Figure 18.7. Roman centuriation in Italia.

Flatter land with higher quality soils is more likely to be centuriated than other lands. A simple examination of the location of the centuriae in Figure 18.7 provides evidence consistent with this prediction as the highest density of centuriae is in the flat fertile Po River Valley. An absence of centuriae in the Alps and Apennines is evident as well.

18.4.2. Historical Information on Centuriation Many of the implications from the model can be tested against information from the scholarly literature on ancient Rome and scholars’ understanding of centuriation and how it was linked to the market for agricultural products. As Temin (2006, 2013) notes, there is considerable information on ancient Rome but relatively little that can be used in standard econometric analysis. In this section we provide details of centuriation that are consistent with the prediction.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

237

18.4.2.1. Agrimensor Responsibilities and Skills For centuriation to create an efficient network the system of surveying must have been well designed and consistently applied. Surveyors measured the land, decided boundaries, distributed individual allocations, directed settlers to their parcels, made maps, recorded land transactions, and dealt with disputes. As mentioned before, they were at the center of the most important activities in the Roman Empire. In order to fulfill these responsibilities significant education was required. The skills that surveyors needed to master included: orientation, sighting and plotting lines, accurate measurement, boundary marking and recognition, and legal knowledge. Many of these skills were best learned by observing other skilled surveyors, so in many cases those who wanted to become surveyors contacted an established surveyor in the community to become fully trained in geometry, measurements, and tax calculations, all of which a surveyor needed on a regular basis (Dilke 1971: 51–6, 87). A surveyor was required to be able to determine orientation using the equipment described above. Usually in centuriation, sundials and compass points were employed to establish north–south orientation. This practice was most likely to be followed for military sites and remote settlements. In some cases, orientations of centuriae were fixed by reference to existing roads or geographical features. Two examples include Dalmatia (modern day Croatia), where most centuriation conforms to the coastline, and the Po valley, where the decumanus was the Via Aemilia, the major Roman trunk road in that region. As predicted, over time the operations of the Empire became increasingly bureaucratic, including those of the agrimensor. Agrimensores were given the responsibility of handling property disputes requiring considerable legal knowledge. Surveyors needed to be familiar with both the body of law governing the classification of land and the body of law concerned with boundaries and boundary disputes. Even though legal definitions were important in property lawsuits, surveyors had to apply reason and make allowances for custom. These surveying and bureaucratic improvements suggest that RS would be adopted more over time.

18.4.2.2. Optimal Alignment Parcel alignment was an important feature of the RS. In the American RS all land was demarcated and all land was aligned along a north–south

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

238

Gary D. Libecap and Dean Lueck

axis. The Romans did not implement a contiguous RS over their entire empire, presumably because it was not efficient to have a uniform alignment rule. Instead each centuriation system was an independent demarcated region surrounded by lands that were either un-demarcated or demarcated in a local MB system. As noted above, the default alignment was north–south, but depending on the topography a north–south alignment might be more costly than an alternative orientation. In particular, if there was an existing transportation corridor (road or river), it was cheaper to align the centuriae perpendicularly to the corridor rather than north–south. Such an alignment preserves optimal square parcels along the river or the road system (or along a coast). Because such a centuriation system is generally isolated, it will not create alignment conflicts with other lands. Figure 18.8 shows possible alignment outcomes and illustrates how alignment can depend on the topography of the land. Although we have no predictions about the alignment, the issue is a good example of how important coordination was for the RS and how land demarcation followed a maximization pattern. The historical record offers evidence consistent with the hypothesis that alignment depended on the topography of the land. Dilke (1971: 87) lists

N

A

D B

A: oriented N/S

C

B: oriented along river parallel C: oriented N/S D: oriented along river parallel

Figure 18.8. Hypothetical Roman centuriation along a river.

River

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

239

settlements that conformed to a non-standard orientation:²⁷ “at Tarracina (now Terracina) the decumanus was the Via Appia; in much of the Po valley it was the Via Aemilia; at Acelum (Asolo) it was the Via Postumaia. In the area round Florence the topography, particularly as regards rivers, dictated the alignment of the centuriation. . . . In some coastal areas the alignment of the coast affected the centuriation; thus much of the Dalmatian coast runs roughly NW–SE, and at Iader (Zara) the centuriation was made to conform to this coastline, even spanning the coastal indentations.”

18.4.2.3. Types of Land There are different categories of land in ancient Rome, which allows us to examine the determinants of these categories that were governed by different legal rules and by different demarcation systems. The discussion here that follows Campbell (2000) indicates that more valuable land tends to be governed by the RS. In particular Campbell (ibid.) discusses several types of land: Land that was divided and allocated. There are two categories, land contained within limites and land allocated by means of straight-line boundaries, rigores. Land that received limites is centuriated land, which Frontinus calls land within decumani and cardines (ibid. 3). Land that was divided by rigores lengthwise is said to be separated by strigae, while land divided breadthwise is called by scamna. The choice of demarcating land by strigae or scamna was based on longestablished custom in the way in which public arable land in Italy was cultivated. RS was usual in these settings via centuriation. Land that was contained in a survey throughout its extent. This land was allocated to a community and in some provinces a tax was imposed as a lump sum on the community as a whole (ibid. 319), which explains the importance of accuracy and thoroughness in demarcating this type of land. The land of private individuals was also surveyed on the same principle (ibid. 3). It was common for a surveyor to enter land to map it according to how it had been divided and allocated, even though only its boundaries had been surveyed. RS was often used through centuriation. ²⁷ The Via Appia, Via Aemilia, and Via Postumaia were major trunk roads in the Roman system of roads.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

240

Gary D. Libecap and Dean Lueck

Land of uncertain boundary not contained in any survey (arcifinius). The term ager arcifinius (also called as ager occupatorius, ibid. 3, 472) refers to land that was not surveyed but was bound by rivers, ditches, roads, or mountains. This land and these borders were given boundary markers over time because of the frequency of legal disputes in the places where the territory ended. Ager arcifinius was often land that had been newly conquered by the Romans, but which had not been properly annexed and allocated. This may have resulted from a lack of resources to do so, or because the land was very remote as well as potentially of low value (ibid. 472). The first settlers of these remote areas were not organized by the central government in Rome, but, instead, were individuals or small groups of Roman farmers who migrated to the region. These early settlers were frontiersmen who were constantly faced with the possibility of violent confrontation from the recently expelled occupants. To mitigate the risk of attack the early settlers often demarcated their land holdings along natural defenses such as streams and hills. Even so, there was little formal demarcation because of the lack of central survey. MB was common when there was no official survey or presence of the legion to enforce RS. When an area was surveyed and the land within it divided by limites into rectilinear plots, there was often land left over that was not divided into centuriae. This type of land was classified under the general term subsecivum. The word subsecivum is derived from the word subseco, meaning the line that cuts it away. Campbell (ibid.) notes two types of subseciva that occurred in centuriated settlements. The first type was the area on the outer boundaries of lands allocated for centuriation, but where square centuriae could not be completed. The second type was land that was in the middle of a centuriated area but that is not allocated and was marked off by a line. In noncenturiated settlements, land that lay between the outer boundaries of a settlement and rectilinear divisions was also called subsecivum (ibid. 3). Land designated as subsecivum was at the disposal of the founder of the colony (ibid. 321). The surveyor also marked some land as public areas such as woods and pastures (Dilke 1971: 107). These lands were not available for private ownership, but rather were left for firewood for the public bathhouses or cemeteries for the poor (ibid.). As a result, in these areas there was a mix of MB and RS as well as RS centuriation that did not involve rectangular parcels.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

241

18.4.2.4. Long-Term Effects and Path Dependence One of the most striking features of Roman centuriation is its astonishing persistence. Perhaps no other institution shows the path dependence of land demarcation. Governments come and go. Cities rise and fall. Twenty centuries after the Roman agrimensores created a grid of square parcels, the land is still farmed in rectilinear fields and the roads run straight with right angles as they did under Roman rule. As John Bradford (1957) describes the modern traces of ancient systems of centuriation: The forceful imprint of the elaborate gridded road-systems which betoken it can be still traced across some thousands of square miles on both sides of the Mediterranean. In origin, most of these systems were carved out of territories raw from conquest, and even now, in retrospect, their appearance deeply stirs the imagination,—so boldly artificial was the conception and drastic the creation as compared with any earlier man-made landscape in this region. Such a purposeful regular dissection of land for cultivation was not again matched in formal precision until comparatively recent times: e.g. in the settlement of North America or in the systematization of the Great Alfold after the Ottoman tide had receded from Hungary in the 18th century. Bradford (1957: 145, 154)

The satellite photograph of the area around Cesena, Italy (Figure 18.1) is an example. In fact, in the modern era, the vast extent of centuriation was first noticed by pilots during World War II (Bradford 1957) and ultimately led to the field of aerial archaeology.²⁸ Another example is Carthage, Tunisia. Figure 18.9 shows a modern satellite photograph of the city and surrounding countryside and its unmistakable Roman demarcation. Carthage, according to tradition, was founded in 814 BCE and was the capital of an extensive empire during the early and middle years of the Roman Republic. The regime was often in conflict with Rome and the city was ultimately destroyed by Romans in 146 BCE after the Third Punic War. Not only did the Romans burn the city and destroy surrounding farmland, they also re-demarcated the land using the centuriation system.²⁹ Soon ²⁸ Libecap and Lueck (2011b) find dramatic persistence of demarcation in Ohio (USA) and in the differential value of the land under RS and MB systems. ²⁹ The Romans also extensively centuriated farmland in North Africa and this area was an important source of wheat over the centuries.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

242

Gary D. Libecap and Dean Lueck

Figure 18.9. Carthage, Tunisia (36 N, 10 E), showing RS demarcation persistence. Source: Google Earth.

thereafter Carthage became one of the largest cities in the Roman world, with a population of hundreds of thousands. As Figure 18.9 shows, the Roman grid system is still intact over two thousand years later. It goes without saying that since the Romans implemented their centuriation system much has changed in Carthage, but not the demarcation of the land.

18.5. CONCLUDING REMARKS We study the demarcation of land in Ancient Rome in order to examine the choice between centralized and decentralized coordinating property institutions.³⁰ This chapter is the first systematic economic analysis of land demarcation outside the United States, examining the two dominant systems in ancient Rome. Our findings increase our understanding of a fundamental feature of property institutions. In particular, we find ³⁰ As we have stressed, we do not claim that Roman jurists conceptualized such problems in the same terms as we do.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

243

that the implementation of the rectangular system of land demarcation known as centuriation is consistent with a wealth-maximizing model; and that such demarcation institutions have a remarkable persistence and impact on land use and land markets. Both findings suggest the importance of initial institutional arrangements. Still this chapter leaves many questions unanswered and available for further study, including the role of political institutions on the use and structure of centuriation, the impact of centuriation on the development of the Roman economy, and the effects of the decline in the Empire on the use of centuriation in peripheral regions. This study of land demarcation is connected to a growing literature on the nature of property rights in contributing to different patterns of economic growth across countries (de Soto 1989; North 1990; Acemoglu and Robinson 2012). It is also linked to an empirical literature on property rights with application to natural resources (Bohn and Deacon 2000; Libecap and Smith 2002), American Indian reservation agriculture (Anderson and Lueck 1992), and urban residential land development (Miceli et al. 2002). It thus informs the larger issues of institutions and economic development.³¹

REFERENCES Acemoglu, Daron, and James A. Robinson. 2012. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. New York: Crown Publishers. Adkins, Lesley, and Roy A. Adkins. 2004. Handbook to Life in Ancient Rome, updated edn. New York: Facts on File. Amiama, C., J. Bueno, and C. J. Alvarez. 2008. “Influence of the physical parameters of fields and of crop yield on the effective field capacity of a self-propelled forage harvester.” 100 Biosystems Engineering 198–205. Anderson, Terry L., and Dean Lueck. 1992. “Land tenure and agricultural productivity on Indian reservations.” 35 The Journal of Law and Economics 427–54. Arruñada, Benito. 2012. Institutional Foundations of Impersonal Exchange: The Theory and Policy of Contractual Registries. Chicago: University of Chicago Press. ³¹ We received helpful comments from the volume editors Paul du Plessis, Joe Manning, Julio Ramos, and two anonymous reviewers. We have also benefitted from the assistance of Bruno Pegorin, Francesco Mazzucato, and Director Silvia Cipriano at the Museum of Roman Centuriation in Borgoricco (Padova) Italy.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

244

Gary D. Libecap and Dean Lueck

Baird, Douglas C., Robert H. Gertner, and Randal C. Picker. 1994. Game Theory and the Law. Cambridge: Harvard University Press. Barzel, Yoram. 1982. “Measurement cost and the organization of markets.” 25 Journal of Law and Economics 27–48. Bohn, Henning, and Robert T. Deacon. 2000. “Ownership risk, investment, and the use of natural resources.” 90.3 American Economic Review 526–49. Bradford, John. 1957. Ancient Landscapes: Studies in Field Archaeology. London: G. Bell and Sons. Campbell, Brian. 2000. The Writings of the Roman Land Surveyors: Introduction, Text, Translation and Commentary. London: Society for the Promotion of Roman Studies, Journal of Roman Studies Monograph No. 9. Cornell, Tim, and John Matthews. 1982. Atlas of the Roman World. Oxford: Phaidon. De Soto, Hernando. 1989. The Other Path. New York: Harper and Row. Dilke, Oswald A. W. 1971. The Roman Land Surveyors: An Introduction to the Agrimensores. New York: Newton Abbot. Dilke, Oswald A. W. 1985. Greek and Roman Maps. Ithaca: Cornell University Press. Dixit, Avinash. 2003. “Trade expansion and contract enforcement.” 111 Journal of Political Economy 1293–1317. Estopinal, Stephen V. 1998. A Guide to Understanding Land Surveys, 2nd edn. New York: John Wiley and Sons. Farrell, Joseph, and Paul Klemperer. 2007. “Coordination and lock-in: competition with switching costs and network effects,” in Handbook of Industrial Organization. Amsterdam: Elsevier, 3, Chapter 31. Johnson, Hildegard Binder. 1976. Order Upon the Land: The US Rectangular Land Survey and the Upper Mississippi Country. New York: Oxford University Press. Larsen, J. A. O. 1938. “Roman Greece,” in A. H. M. Jones, ed., An Economic Survey of Ancient Rome. Baltimore: The Johns Hopkins Press; London: Milford. Lee, D. R., and G. T. Sallee. 1974. “Theoretical patterns of farm shape and central place location.” 14 Journal of Regional Science 423–30. Libecap, Gary D., and Dean Lueck. 2011a. “Land demarcation systems,” in Kenneth Ayotte and Henry E. Smith, eds, Research Handbook on the Economics of Property Law. Research Handbooks in Law and Economics. Northampton, MA: Edward Elgar. Libecap, Gary D., and Dean Lueck. 2011b. “The demarcation of land and the role of coordinating property institutions.” 119 Journal of Political Economy 426–67. Libecap, Gary D., Dean Lueck, and Trevor O’Grady. 2011. “Large scale institutional changes: land demarcation within the British Empire.” 54 Journal of Law and Economics S295–S327.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Land Demarcation in Ancient Rome

245

Libecap, Gary D., and James L. Smith. 2002. “The economic evolution of petroleum property rights in the United States.” 31:S2 Journal of Legal Studies S589–S608. Linklater, Andro. 2002. Measuring America. London: Harper Collins. Linklater, Andro. 2013. Owning the Earth. London: Harper Collins. Lueck, Dean. 1995. “The rule of first possession and the design of the law.” 38 The Journal of Law and Economics 393–436. McEntyre, John G. 1978. Land Survey Systems. New York: John Wiley and Sons. Miceli, Thomas J., Henry J. Munneke, C. F. Sirmans, and Geoffrey K. Turnbull. 2002. “Title systems and land values.” 45 The Journal of Law and Economics 565–82. North, Douglass C. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press. Stambaugh, John. 1988. The Ancient Roman City. Baltimore: Johns Hopkins University Press. Stanislawski, Dan. 1946. “The origin and spread of the grid-pattern town.” 36 Geographical Review 105–20. Temin, Peter. 2006. “The economy of the early Roman Empire.” 20 Journal of Economic Perspectives 133–51. Temin, Peter. 2013. The Roman Market Economy. Princeton: Princeton University Press. Thrower, Norman W. 1966. Original Survey and Land Subdivision. Monograph Series of the American Association of Geographers. Chicago: Rand McNally and Company.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

19 The Institutions of Roman Markets Benito Arruñada

19.1. INTRODUCTION: THE INSTITUTIONS OF IMPERSONAL EXCHANGE Opportunities for economic growth are greater when trade, instead of being limited to known people, is impersonal.¹ This is so especially when “impersonal” means that contractual performance is totally independent of hard-to-observe parties’ characteristics, including not only their reputation and wealth but also their legal authority to commit strangers to the contract.² However, such fully impersonal trade requires contractual enforcement to be based on assets (i.e. to convey in rem rights or, in general, priority to acquirers) and this poses a conflict between those holding and those acquiring property rights. This conflict between owners and buyers is present in both property and business contexts in which owners explicitly act through contractual agents. This chapter analyzes the sophisticated institutions used in classical Rome to overcome such conflict, reducing information asymmetry without endangering property rights. Given that property rights are the foundation of economic incentives and prosperity, one might think that, in case of conflict with acquirers, goods should always be returned to their owners unless ¹ This section summarizes the argument in Arruñada (2012: 15–42) and Arruñada, Zanarone, and Garoupa (2019). ² Other concepts of impersonal exchange use less stringent requirements, such as, mainly, trade in the shadow of an independent court (North 1990: 34–5), but also trade with strangers, equal treatment of market participants, presence of assurance intermediaries, or posted prices available to any buyers (Arruñada 2012: 15–18). Benito Arruñada, The Institutions of Roman Markets In: Roman Law and Economics: Exchange, Ownership, and Disputes Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0019

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

248

Benito Arruñada

they had granted their consent. However, such strict enforcement of property rights as rights in rem would increase transaction costs by worsening the information asymmetry suffered by acquirers of all sorts of rights, who would always have to gather the consent of the previous owners without even knowing who they are. Strictly enforcing property rights would therefore endanger trade. Moreover, it would also endanger specialization, because specialization is often based on having agents acting as owners’ representatives, and, with universal strict enforcement, acquirers would have reasons to doubt the legal authority of sellers. Economic growth therefore requires this conflict between property enforcement and transaction costs to be minimized, so that both current owners and acquirers are efficiently protected. Protecting owners’ property rights encourages investment, and reducing the transaction costs faced by acquirers encourages them to trade impersonally and thus improves the allocation and specialization of resources. Owners’ consent must be preserved but enforced in a way that makes trade possible. And current owners have an interest in tackling the conflict, not only because they are potential sellers but also because, being acquirers with respect to the previous owners, they could eventually lose their title. The nature of the problem can be clarified by considering that most economic transactions are interrelated sequentially, as most transactions legally interact with previous transactions. In the simplest sequence, with only two transactions, one or several “economic principals”—such as owners, employers, shareholders, creditors, and the like—first voluntarily contract with one or several economic “agents”— possessors, employees, company directors, and managers—in an “originative” transaction. Second, the agent then contracts “subsequent” transactions with third parties.³ Understandably, it is necessary to optimize the total costs of transacting, considering both originative and subsequent transactions. Sequential exchange is necessary to specialize the tasks of principals and agents—between landowners and farmers, employers and employees, shareholders and managers, and so on in the originative contract. But it also gives rise to substantial transaction costs, because, ³ This use of “agency” language for describing property cases may puzzle readers familiar with the legal concept of agency. However, it is convenient for generalizing the argument, encompassing all types of transactions.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

249

when third parties contract with the agent in the subsequent contract, they suffer information asymmetry regarding not only the material quality of the goods or services being transacted but also the legal effects of the previous originative contract. In particular, third parties are often unaware if they are dealing with a principal or an agent, or if the agent has sufficient title or legal power to commit the principal. Moreover, principals face a commitment problem when trying to avoid this asymmetry because their incentives change after the third party has entered the subsequent contract. In an agency setup, before contracting, principals have an interest in third parties being convinced that agents have proper authority. However, if the business turns out badly and there are no further incentives in place, principals will be inclined to deny such authority.⁴ The typical dispute triggered by the sequential nature of transactions is one in which the principal tries to elude obligations assumed by the agent in the principal’s name, whether the agent had legal authority or not. In principle, judges may adjudicate in such disputes in favor of the principal or the third party. I will refer to favoring the third party as enforcing “contract” rules, as opposed to the seemingly more natural “property” rules that favor the principal.⁵ The effects of these rules are ⁴ This is easy to grasp in a legal agency setup as e.g. when contemplating the relationship between a shareholder (principal), a corporate representative (agent), and a corporate lender (third party). A typical property conflict is that between a first buyer (principal), a seller (agent), and a second buyer (third party). In such a case, the commitment refers to the seller’s promise not to sell twice. Ex ante, she is interested in committing herself, in order to encourage the buyer to buy; but her incentives change after the sale. The text could therefore read: before the originative sale contract, sellers (i.e. agents) have an interest in buyers (i.e. principals) being convinced that they will not cheat through a subsequent transaction (e.g. a second sale), but their incentives change after the first sale. ⁵ These labels parallel the legal origin of the dichotomy and should prevent confusion with related but drastically different concepts. In particular, the rules are similar but distinct from the “property” and “liability” rules defined in an influential work by Calabresi and Melamed (1972) because, instead of a taking that affects only two parties, here the rules are defined in the context of a three-party sequence of two transactions. Moreover, my analysis focuses on the role played by the parties in each transaction, disregarding that current third parties will often act as principals in a future sequence of transactions. Consequently, when good-faith third parties win a dispute over their acquisitive transaction (i.e. when they are given a property right), they do not win by virtue of a property rule, as this—by definition—would have given the good to the original owner. In such a case, the third party does not pay any monetary damages to the original owner, as in Calabresi and Melamed’s liability rule. A final difference is that Calabresi and Melamed’s property rule is weaker, referring only to the ability to force a would-be taker to bargain for a consensual transfer similar

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

250

Benito Arruñada

clear. Take the simple case in which an agent exceeds his legal powers when selling a good to an innocent third party (i.e. a good-faith party who is uninformed about the matter in question). When applying the “property rule” that no one can transfer what he does not have, judges have the sold good returned to the previous owner, who is therefore granted a right in rem, and give the innocent third party a mere claim in personam against the agent. Conversely, judges may apply an indemnity or contract rule so that the sold good stays with the acquiring third party and the owner-principal only wins a personal claim against the agent. The difference in the value of these legal remedies—real or personal— is substantial because the enforceability of these two types of right is often markedly different. While rights in personam are only valid against specific persons, inter partes, rights in rem are valid against all individuals, erga omnes. The latter, therefore, provide the strongest possible enforcement: without the consent of the rightholder, rights in rem remain unaffected. When the thing is a parcel of land, the difference in value often ranges from full value for the party being adjudicated the land, to zero value for the party being given a claim to be indemnified by an insolvent person. (Similar differences arise in business and corporate contexts, where a parallel distinction is often made, but framed in more general terms: not in terms of rights but in terms of legal priority.) If judges apply a property rule, maximizing property enforcement, owners will feel secure with respect to future acquisitions but all potential acquirers will suffer greater information asymmetry with respect to legal title, endangering trade. Conversely, if judges apply a contract rule, minimizing information asymmetry for potential acquirers, they weaken property enforcement, making owners feel insecure, endangering investment and specialization. The choice of rule therefore involves a conflict between property enforcement and transaction costs—more generally, a conflict between the transaction costs of originative and subsequent contracts. To overcome the conflict, expanding the set of viable contractual opportunities with minimal damage to property rights, different solutions will be appropriate, depending on the circumstances of each type of right and transaction. The legal system may directly to specific performance, which thus arguably has little to do with a right in rem (Merrill and Smith 2001).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

251

choose to enforce some rights in a certain way or, more generally, give freedom to rightholders about which rule to apply. However, when rightholders are free to choose, the legal system must guarantee their commitment, which in some cases is automatic as a byproduct of market interactions but in others requires costly organizations such as public registers. In general, for judges to apply property rules, which favor owners, owners must have publicized their claims or their rights, which should protect acquirers. That is, owners can opt for a property rule to make their rights stronger, but, thanks to publicity, acquirers suffer little information asymmetry. Conversely, for judges to apply contract rules, which favor acquirers, owners must have granted their consent, which should protect them. That is, when it is in the owners’ interest to reduce transaction costs, they choose a contract rule, so that acquirers’ rights are stronger, whereas owners’ rights are weaker. This weakening of property is safeguarded by the fact that it is owners—in general terms, principals—who choose e.g. the agent to whom they entrust possession or appoint as their representative, while implicitly, and simultaneously, opting for a contract rule. However, commitment to their choice is also necessary. Smooth operation of this conditional application of rules poses varying degrees of difficulty for different transactions. The difficulty is relatively minor when the originative transaction produces verifiable facts, such as the physical possession of movable goods by a merchant or a house by a lessee, or the ordinary activity of an employee. For these cases, judges can base their decisions on this public information, which is produced without any explicit formal intervention. What judges or legislatures have to do is only to define clearly the rules to be applied. The difficulty is greater when the originative transaction produces fewer verifiable facts, making private (i.e. contractual, as opposed to institutional) solutions harder to apply. Such private solutions may even be impossible if all the information on the transaction remains hidden and its consequences are not verifiable. Consider, for example, the difficulties of using land as collateral when hidden mortgages are enforced on the basis of contractual documents (Arruñada 2003). Given the possibility of antedating mortgage deeds, judges would be basing their decisions on unreliable information, and lenders would be reluctant to contract for fear of previous mortgages emerging. In such a context, rules alone are not enough because applying them

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

252

Benito Arruñada

requires verifiable information on the titling-relevant elements of originative contracts. To produce such information, it is necessary to enforce only those mortgages that have been made public, usually by entering them into a public register. Something similar happens when establishing by purely private contract the existence of a corporation, distinguishing the corporation’s assets from the personal assets of its shareholders (Arruñada 2010). This analytical perspective, which focuses on contract interaction and sequential exchange, in which there are at least two transactions and three relevant parties, is necessary for improving our understanding of transaction costs in impersonal markets, and the role that institutions play to contain them. In particular, it prevents contractual problems from being considered as referring to only two parties, even for situations in which the dominant conflict may well be multilateral. For instance, when considering the institutions of Roman markets, a prominent survey (Frier and Kehoe 2007: 134–7) discusses the tragedy of the commons and the bilateral conflict between the owner and the potential trespasser or usurper, but sidelines the multilateral conflict, which is central for any property market, between alternative claimants and acquirers (e.g. between owner, seller, and acquirer or between an owner in debt and successive creditors).

19.2. INSTITUTIONAL CHOICE Environmental circumstances, such as the cost and benefits of registry services and their palliatives, as well as the distribution of trading opportunities in the economy, will also affect the relative value of personal and impersonal solutions. To structure the analysis, and taking land transactions as the baseline, Figure 19.1 represents the hypothetical social decision on whether or not to build the institutions to support impersonal exchange. In essence, it assumes that more impersonal exchange requires a fixed cost F for putting in place the necessary institution, such as a land registry.⁶ If such an institution is ⁶ Using discrete categories of personal and impersonal exchange is a simplification. Moreover, personal exchange also often requires institutional support, which incurs fixed and variable costs, as the historical examples below will make clear. However, for the sake of simplicity, I will omit them, assuming that they are lower than those for

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets Social value of resources subject to titling conflict

253

45° Public titling

G

Private titling (i.e., “privacy”)

L L

Indifference point for individual decisions

Value of resources assuming no titling conflict

Figure 19.1. Choice between personal and impersonal exchange to maximize social value of productive resources. Note: The figure represents how the social value of resources such as e.g. land (vertical axis) is a function of their theoretical value assuming no conflicting claims (horizontal axis) and the availability of institutions (e.g. land registries) for public titling. Social choice of institutions is driven by these costs and the statistical distribution of the value of resources in the economy along the horizontal axis. Area G represents the potential gain from using institutions suitable for impersonal exchange for high-value resources, and area L the potential loss from using them for low-value resources (by e.g. overtitling low-value land). Source: Adapted from Arruñada and Garoupa (2005: 717, fig. 2).

created, individuals can then choose to trade personally or impersonally. To trade impersonally, they incur an additional unit cost but their assets become more valuable. Following Arruñada and Garoupa (2005), the social value of an asset, such as e.g. a parcel of land (represented on the vertical axis), is assumed to depend on the

impersonal exchange, so that the costs considered in the analysis must be understood as the cost differences between institutions enabling relatively more or less impersonal exchange.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

254

Benito Arruñada

probability that claimants with better legal rights may appear and is thus a fraction of its value in an ideal world with no conflicting claims, represented on the horizontal axis. This fraction depends on the availability of institutions and the individual decisions to use them or not, and it is lower if the asset is traded in a more personal way—in land, it remains under private titling, or “privacy,” for short. To simplify matters, the figure represents all the assets available in the area served by a certain institution, such as a registry, that costs an amount F to put in place. Publicly titling a particular parcel of land means that, by incurring an additional unit cost (given by the intercept of the public titling line), the probability of a conflicting claim on that land being successful falls with respect to privacy. In other words, by publicly titling, the value of the asset increases by a certain percentage: with respect to the privacy line, the public titling line has a negative intercept but is steeper. The social choice of institutions is therefore driven by the fixed and variable costs of the different institutions and the value of resources in the economy, represented in the figure by their distribution along the horizontal axis. When public titling is available, owners incur a unit per-parcel cost for publicly titling their land, given by the intercept of the public titling line. They will choose not to publicly title the less valuable parcels: those below the indifference point in the figure. Area G therefore represents the potential gain from using public titling institutions, which are only used for higher-value resources.⁷ In a given geographical or historical context, the optimality of introducing public titling hinges on the difference between the titling gains G and the fixed costs, if any, of establishing the titling system, F; and the main determinants of these gains and costs are: (1) the statistical distribution of the resource in the economy, represented by its distribution along the horizontal axis; (2) the costs and effectiveness of public titling, given respectively by the differential fixed cost of establishing the titling system (F), the differential variable or unit costs represented by the intercept of the public titling function and its slope (which for land is equal to one, minus the probability of ⁷ For a more detailed analysis, see Arruñada and Garoupa (2005: 713–25). Area L is the potential loss when public titling is made mandatory, so that even low-value resources are publicly titled, resulting in overtitling of low-value land. This possibility is more theoretical than real, because, in practice, the effectiveness of mandatory titling is limited to initial allocation of titles, as shown by the recurrent failure to register second transactions after land titling efforts (Arruñada 2012: 147).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

255

eviction); and (3) the cost and effectiveness of privacy, given by the variable costs represented by its intercept (assumed in the figure to be zero) and the slope of the privacy function. The situation in the classical period of Roman law (including the last two centuries of the Republic and up to the second half of the third century CE) can be hypothesized as: (1) offering considerable gains as a consequence of relatively extensive markets, so that a substantial proportion of resources lies to the right of the horizontal axis, capturing the effect of greater demand and opportunities for impersonal exchange; (2) maybe suffering high fixed and variable costs of creating (i.e. cost F) and operating (the line’s intercept) what would be relatively ineffective (i.e. with a relatively flat line) registries; and (3) enjoying refined institutions for personal exchange (privacy), which made palliatives relatively effective by enacting rules and maximizing the use of available evidence (i.e. obtaining a steeper slope for the personal exchange line). The rest of the chapter will address these three issues in the Roman context. §19.3 explores to what extent Roman markets would have benefited from institutions making impersonal exchange easier. §19.4 examines the difficulties that the Romans of the West may have faced for archiving information, making registries more costly or less effective. Then, the bulk of the chapter studies the main sets of palliative solutions used for channeling different types of exchange in the areas of property (§19.5), business (§19.6) and the enforcement of personal obligations (§19.7), making personal exchange (privacy) regimes more effective. Sections 19.7.3 and 19.7.4 explore the factors behind the institutional choice while sections 19.7.5 to 19.7.7 review their content. §19.8 concludes by examining the overall role of the Roman state with respect to the market. Before moving ahead, some caveats are in order to clarify the boundaries of the analysis. First, the above conjectures assume that privacy was the predominant solution for the conveyancing of land during the classical period. Even though some public titling solutions were also available—in particular, public conveyances through ceremonies (mancipatio) and judicial intervention (in iure cessio)—these were not based on registries and therefore presented different costs and effectiveness (see, for a complementary discussion, Arruñada 2015).⁸ ⁸ Especially in the pre-classical period, the Romans used a form of conveyance known as mancipatio that I will analyze in section. It was public but did not rely on

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

256

Benito Arruñada

Second, when the analysis is applied not to land but to other types of goods, it must be kept in mind that for originative transactions that produce some verifiable consequences, such as the possession of an asset or the activity of an employee, these consequences can be relied upon to ground judicial decisions on subsequent contracts with third parties. In terms of figure 19.1, this means that the personal-exchange line will be—ceteris paribus—steeper than for those other transactions which lack verifiable consequences, such as, for instance, mortgages. Third, from this perspective, solutions for personal and impersonal exchange are seen as substitutes: if personal exchange is made easier, impersonal exchange becomes more expensive.⁹ However, the state also builds institutions that make personal exchange more easily enforceable: e.g. if the state opens jails for putting debtors in prison, these will be a complement to personal exchange. The same holds for civil registries and other systems that facilitate the identification of individuals and therefore the enforcement of their personal obligations. Fourth, there has been much discussion about to what extent it makes sense to use modern economic concepts when analyzing ancient economies. The above analysis assumes, and the next sections aim to show, that the basic problems that market participants faced in Roman times were essentially the same as those they face now: the need to overcome information asymmetries efficiently with respect to title and legal authority in order to enforce property rights and reach specialization advantages in contracting. Institutions of course adapt to the specific technological and economic circumstances at any time, but this adaptation is always governed by the same basic forces. The analysis may be new but the problems are not. The relative importance of the problems is probably different, and, in the Roman case, this is applicable, in particular, to the relative prevalence and the consequent interactions between personal and impersonal exchange. registries, therefore did not incur a fixed cost F, but its effectiveness was likely limited to the local market. Roman Egypt did rely on a sophisticated system of registers, but it predated Roman conquest. ⁹ Note that my argument is about substitution between personal and impersonal exchange, both happening in the market, not between market and non-market forms of exchange, as seemingly discussed by e.g. Temin when arguing that the Roman “economy of friends” was a complement of the formal market (Temin 2013: 110–11, 148); i.e. to the extent that reputation, friendship, and other social ties acted as safeguards of market exchange, they were complements of the market, but making personal exchange possible, which was a substitute for impersonal exchange.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

257

However, one should not ascribe personal exchange to nonmarket relations. Even today, most market exchange is personal to the extent it relies on information about the parties’ attributes. In particular, exchange can be personal without being based on domination or power, an aspect that might be relevant for the ongoing discussion between “primitivist and/or substantivist” and “modernist and/or formalist” positions (using the labels in Scheidel 2012: 7–10). Lastly, I will not analyze factors making all types of market exchange—both personal and impersonal—and, in particular, longdistance trade viable, such as, mainly, the effectiveness of the Roman state in providing: (1) a stable currency; (2) a tolerable tax burden and little government intervention; (3) security of property in a wide geographical area, with e.g. little piracy in the Mediterranean; and (4) a hybrid but comprehensive judicial system which, without destroying local legal systems, covered the whole of the Empire through potential appeals to Rome. My objective is to analyze the main features of a few salient institutions from the perspective of personal-versus-impersonal exchange, in the hope of suggesting some alternative interpretations. With this in mind, I will examine some classic controversies, aiming to clarify some of their main elements.

19.3. INSTITUTIONAL DEMAND: OPPORTUNITIES FOR IMPERSONAL EXCHANGE IN THE ROMAN ECONOMY With the expansion and consolidation of the Roman state during the Republic, the Roman economy developed in the direction of providing greater opportunities for impersonal exchange. In terms of figure 19.1, it is therefore likely that the distribution of resources moved to the right along the horizontal axis. It is debatable, however, to what extent such movements and opportunities were limited to a few activities or even geographical areas. Let us now review and discuss some empirical indicators in this regard.

19.3.1. An Extensive Market Economy Several indicators point out that the economy of the early Roman Empire was an exchange economy to an unprecedented level, albeit

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

258

Benito Arruñada

with some redistribution by the state and the presence of customary systems in rural areas. It has been estimated that 7.4–19.5 percent of Italy was urbanized at that time (Wilson 2011: 191). If these numbers are taken as an index of economic performance, the Roman economy performed similarly to those of the most advanced European economies in 1600–1750. This was achieved “through the combined operation of moderately stable political conditions and markets for goods, labor and capital, which allowed specialization and efficiency” (Temin 2006: 134). It has even been argued that the Roman economy was not merely a collection of independent markets but an integrated market economy: “markets for goods, labor and capital were relatively well-developed in ancient Rome, which in turn encouraged specialization and efficiency. These markets were able to work well in the environment created by public authorities who provided local public services in cities and a functional rule of law across most of the Empire” (Temin 2006: 137). In particular, “the market for land in the Roman Empire worked approximately like the land market today” (Temin 2013: 139) because land was typically held in fee simple with few impediments to land usage and sale. The market in urban real estate enjoyed not only a substantial provision of public goods but, compared with other ancient societies, much less interference from public authorities in terms of debt moratoria, rent control, mandatory redemption, and tenant protectionism (Ellickson in this volume). In addition, prices varied substantially and transactions were sophisticated, including mortgages and subdivisions. Moreover, the Roman economy was not only well developed in terms of size and specialization but in the predominant role of market exchange in the allocation of all types of resources. In addition to a market for slaves that had little in common with “closed” slavery (Scheidel 2008; Temin 2013: 114–38), there was an active market for the rights to sue or to inherit (Riggsby 2010: 127), and monetary compensation was acceptable even for criminal acts (Nicholas 1962: 209). The environment for market exchange only ceased to be favorable in the later centuries of the Empire, when trade collapsed at least in some periods and areas as a consequence of political instability. The market economy was seriously damaged in some regions in the third century CE, with the collapse of frontiers and other events: the

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

259

doubling in size and cost of the army; the increase in taxation; the emergence of a new more militaristic and local ruling class; and the rise of Christianity (Brown 1971). Both the market and the common culture retreated earlier in the West: e.g. market institutions remained in place in the Eastern empire in the seventh century (Brown 1971: 145).

19.3.2. Discussion However, the extent, integration and nature of market exchange in ancient economies, and in Rome in particular, remain controversial among historians.¹⁰ For our purposes, even if the Roman economy was a market economy, it is debatable to what extent it opened up opportunities for impersonal exchange: in terms of figure 19.1, even if productive resources moved to the right along the horizontal axis, it is unclear to what extent or in which areas they did move. Certainly, in addition to being extensive, the Roman market economy also showed signs that it might have benefited from less personal exchange, such as substantial productive specialization and longdistance trade. There was specialization mainly in agricultural production and processing industries, with technological change and diffusion (Greene 2000). Since the second century BCE, provinces had specialized in producing different types of goods: e.g. Italy exported wine and oil; Egypt, grain; North Africa, oil and grain; Spain, wine, oil, and minerals, etc. Long-distance trade in bulk commodities was prevalent (Temin 2006), and the degree of specialization achieved for bulk commodities suggests that the difficulties for impersonal exchange were overcome at least in those markets. Documentary evidence shows that at least the large estates were organized with the aim of reaching economies of scale and producing marketable surpluses (Rathbone 1991). Moreover, the city of Rome, far from ¹⁰ Most ancient historians would subject claims about the integration of markets in the Roman Empire to numerous caveats. See e.g. Erdkamp (2005) and Bransbourg (2012). Moreover, judgments on ancient economies must always be taken cautiously, given the lack of adequate evidence. See, for instance, Scheidel (2012: 2–5), for a summary of the limitations for unambiguously interpreting the evidence on the Roman economy, and Scheidel and Friesen (2009) for a comparative assessment of works obtaining substantially different estimates of the mere size of the Roman economy.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

260

Benito Arruñada

being an extractive parasite, as Finley (1999) contended, specialized in producing and exporting all sorts of services (Ellickson in this volume). After the First Punic War (264–241 BCE), the increase in foreign population in Rome made it necessary to create a specialized magistracy, the praetor peregrinus, for dealing with non-Roman citizens by applying the ius gentium to them, with a focus on conflicts related to interregional trade (Nicholas 1962: 7; Brennan 2001: 87–9; Winkel 2013: 3553). Financial intermediaries and transactions were sophisticated, and bankers provided such an array of services (Andreau 1999) that it has even been suggested that “financial institutions in the early Roman Empire were better than those of eighteenth-century France and Holland” (Temin 2013: 189). Moreover, much borrowing was for commercial purposes,¹¹ and financial markets in different regions of the Empire were linked, most likely through financial intermediaries (Temin 2013: 178–9). Credit was also abundant, so much so that indebtedness was often considered a problem (Crook 1967: 171–2). Lastly, if interest rates, “perhaps the most evident quantitative dimension of the efficiency of the institutional framework” (North 1990: 69), were indeed not much higher than those of today (Crook 1967: 211), they would suggest low default risk and therefore effective guarantees. Despite usury regulations, it seems that market participants were able to adjust the interest rate to the risk of the loan.¹² However, opportunities for impersonal exchange might have been limited to some markets and transactions. With respect to trade, it has been debated to what extent most production did remain local, with most long-distance trade taking place only by water, given the

¹¹ Or even most borrowing, according to some authors (e.g. Johnston 1999: 84). ¹² Rates were often regulated via usury ceilings, but they were often evaded by e.g. lending to peregrini (Livy 35.7; Andreau in this volume), not stating the rate in the contract, and possibly granting hidden discounts on the principals (Verboven 2003). In fact, “repeated statutes were made, for obviating all elusions, which by whatever frequent expedients repressed, were yet through wonderful devices still springing up afresh” (Tacitus, Ann. 6.16.2). In the light of usury-avoidance practices (Koyama 2010), the prevalence in the ancient evidence of loan contracts stating the regulated and legally enforceable maximum rate of 12 percent or no rate at all (e.g. Lerouxel 2012b), irrespective of the borrower’s default risk, suggests that the effective implicit rate—easily disguised by e.g. discounting the loaned amount below the stated principal—was probably higher. Similarly, in the opposite direction, the Sulpicii archive also provides some basis for believing that variable interest rates below the legal limit were charged (Bransbourg 2014).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

261

higher costs of transporting goods over land.¹³ Similarly, there are doubts on the nature of the exchanges. For instance, for rural land in Roman Egypt, Rowlandson (1981) considers that many of the transactions she analyzes took place between relatives. Certainly, for a land market to exist it is not necessary for most land to be transferred by purchase, as Temin points out (2013: 144); however, transactions between relatives are more likely to be personal. Similarly, in the area of sureties, the effectiveness of guarantees raises the question about to what extent exchange was personal or impersonal. Some authors even give somehow contradictory accounts on this. For example, Crook claims that the abundant supply of credit to the dominant class was structured as a web of loans and personal sureties, in a circular fashion, relying on officium, the “mutual serviceableness between status-equals” (Crook 1967: 94) but based, deep down, on the supposedly secure market for land (ibid. 170–2). This description suggests that personal credit was grounded on secured and therefore impersonal credit.

19.4. INSTITUTIONAL SUPPLY: PUBLIC REGISTRIES IN ROME In a context of substantial opportunities for impersonal exchange, it may come as a surprise that land registration “was quite unknown to Roman law” (Nicholas 1962: 104), especially considering that Rome revitalized the land registers of Egypt and built numerous cadastres. In accordance with the analysis in §19.1, there are two possible reasons: the costs of registries and the effectiveness of hybrid palliatives and purely personal-exchange solutions. First, as argued above, the supply of institutions is costly and, therefore, for a given demand, it might be inefficient to provide them even if they make some additional exchange opportunities possible.¹⁴ Second, the efficiency ¹³ It is estimated that “the envisaged price ratio for moving a given unit of cargo over a given unit of distance is 1 (sea) to 5 (downriver)/10 (upriver) to 52 (wagon).” (Scheidel 2013: 4). ¹⁴ This complements the views in some of the chapters in volume I. For example, Abatino and Dari-Mattiacci argue that lack of proper agency institutions constrained growth; while de Ligt seems to hold that institutions followed market demands, and Fleckner suggests that there was not enough demand for more stable business

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

262

Benito Arruñada

of registries and, in general, of all solutions for impersonal exchange depends on those for personal exchange. This section will explore some elements of the cost hypothesis while the following sections will examine Roman palliatives and, in particular, §19.7 will focus on the personal-exchange hypothesis.

19.4.1. Constraints on Property Registries in the Roman West The cost of archiving and, especially, using written documents as judicial evidence is difficult to ascertain but should not have been a serious impediment to developing property registries. It is true that Romans lacked paper and printing, and relied for recordkeeping on perishable media, such as whitewashed boards and wax tablets (Rhodes 2013). However, their reliance on wax tablets for all sorts of important documents and archiving seems to have been a choice (Meyer 2004), as they also made frequent use of papyrus imported from Egypt, especially in the East (Bagnall 2010), and more generally for the administration of the Empire (Innis 1950). The use of papyrus could have reduced the costs of registries, but sealed wooden tablets seemingly offered greater protection against fraud by tampering (Meyer 2004: 2), so were harder to refute in court than other documents (Meyer 2004: 219, 226). Moreover, a basic facilitating factor was probably a level of literacy that is believed to have been high not only among the upper class (W. Harris 1989) but also among lowerclass participants in economic transactions (Haeussler 2013). Roman archiving laws are often considered weak (Riggsby 2010: 89). Similarly, “the files of the government (except perhaps the archives for census material) seem often to have been in disarray” (Eich 2013). It is even claimed that Romans suffered archiving difficulties for legislation, as the imperial office was sometimes unable to find its own legislation, which “may surprise those accustomed to think of the Romans as an administratively efficient people” (Crook 1967: 32). Moreover, forgery of public documents was prevalent at least in certain periods (Moreau 1994). Even authors inclined to believe that the Roman bureaucracy was well-developed recognize associations. Other authors, however, hold a middle ground view (Hansmann et al. in volume I.).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

263

that the archives were at least sometimes in a state of abandonment and negligence (Nicolet 1988: 179). However, different levels of the Roman government did organize extensive public archives for administrative purposes, where they kept copies of different types of documents such as laws, contracts for public works, international treaties, documents related to public lands, minutes of political bodies, imperial rescripts, and mortgage loans to provide subsidized assistance to children (Sánchez-Moreno 2013: 662‒4). Closer to property registries, they also operated public registries on individuals, mainly the census, which was updated every five years (Crook 1967: 46), and a birth registry, established by Augustus in 4 CE, which served to provide evidence of status as well as family or household size. Moreover, Romans kept records of transactions in both the private and public spheres. Private recordkeeping was effective. For example, some banks preserved contracts of third parties for which they had facilitated payment, for them or for the benefit of their clients (Temin 2013: 184), and, more importantly, lenders’ books were used as evidence in courts. The books of bankers were seen as “unimpeachable legal evidence . . . . [And, with some constraints to protect sensitive financial data,] the praetor required bankers to disclose their entries as evidence on behalf of anyone to whose case they were relevant” (Crook 1967: 233). This made their records a sort of partial register, acting as notaries or specialized witnesses. Similarly, the state also kept a record of transactions in the grain trade (Temin 2013: 107). Therefore, registries seem to have suffered neither insurmountable nor specific technological difficulties. Nor did the Romans lack knowledge about property registries: their conquest of Egypt (30 BCE) put them in touch with a long tradition of property registration. In both Ptolemaic and Roman Egypt, public registers played an effective role in enforcement and trade of property rights in land.¹⁵ Furthermore, the Roman provincial

¹⁵ See e.g. Manning (2003); Monson (2012); and Muhs (2015). Very different institutions for land contracting (deeds, notaries, eviction guarantees, witnesses, checks on parties’ identity, formal consent of potentially affected rightholders, repeated formal protestations of title, recordation of contractual transcripts in public registries) were in place in different stages of Egyptian history and changed substantially in line with increases in the volume and value of transactions (Muhs 2015). The effectiveness of the register to protect good-faith purchasers against fraudulent vendors (e.g. Daumas 1987: 162) and the other legal functions analyzed by Manning (2003) and Muhs (2015) point towards a legal (i.e. transaction-cost-reducing) function of

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

264

Benito Arruñada

administration of Egypt created between 68 and 72 CE a sophisticated “registry of acquisitions,” the bibliotheke enkteseon, for property rights in land and slaves. It was not merely a passive recording of deeds but an active register of rights with certification of title: deeds were void unless the grantor, to prove his title, presented a certification issued by the bibliotheke to the scribe writing the deed, and the scribe was then required to send the deed to the bibliotheke. Lerouxel (2012a: 648‒51) shows that the introduction of the bibliotheke coincided with an increase in the volume of transactions. Moreover, it was created with the objective of protecting property rights and reducing private transaction costs.¹⁶ This objective is also clear from the fact that, from a merely fiscal perspective, tax authorities had an easier alternative: to define land taxes and obligations as rights in rem, so that they enjoy priority over all other rightholders, including innocent third parties who thought they were acquiring free of charges. In fact, this was the common solution in later centuries, at least in the West (Nicholas 1962: 153; Crook 1967: 246).

19.4.2. Discussion The situation is far from clear, however, and several aspects remain in doubt. First, the registration of land in the census, begun with Augustus,¹⁷ has sometimes been seen as “a system of land registration” (Crook 1967: 147). However, this part of the census was probably a mere administrative cadastre, instead of a property register,¹⁸ as it was based on voluntary declaration, as cadastres usually are, given

Egyptian registries, as does the claim that they did not contain maps (Kain and Baigent 1992: 1). Applying Demsetz’s (1967) argument, several factors could have made land registries more useful in Egypt: the possibly higher value per unit of land and recurrent annual flooding. ¹⁶ Lerouxel (2016: 145–92), and mainly Jördens (2010) and Yiftach-Firanko (2010), discussed by Lerouxel (2012a: 652). ¹⁷ See, however, Nicolet (1988: 226‒32) for possible antecedents of cadastral documents. ¹⁸ The confusion between cadastres and registries is a version of a wider confusion in historiography between administrative (mainly tax) and contractual or “economic” archives (Sánchez-Moreno 2013: 660). The necessarily public element in rights in rem exacerbates the confusion by adding such a public element to what are functionally private contractual registries (Arruñada 2017, 2018).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

265

that they only create obligations, not rights (Arruñada 2012: 205). The summary of the forma censualis reproduced in the Digest from Ulpian, a jurist writing in the third century CE, makes clear that both the properties and their “owners” were identified in the census, but its purpose was fiscal, as most of the recorded information (on surface areas, uses, products, and values) was irrelevant for titling purposes. Consequently, the information could be based merely on what the “owner” declared, without any reference to an independent review of his title claim. Moreover, the Ulpian fragment makes clear that failing to declare an ownership claim in the census did not damage the claim: “a man who claimed ownership did not declare it, it is right that his claim should remain” (D. 50.15.4.4; Ulp. 2 de cens.). This declarative nature is a common feature of cadastres everywhere because tax authorities are happy to record a claimant as owner to ensure tax collection, even if his claims are not legal. It also explains why the census was not used as proof of title but only as one more source of evidence: “the census-list, being based on individual declarations, was only evidentiary, not automatically proof” (Crook 1967: 149). Of course, as in many other jurisdictions, ancient and new, after volunteering to record their claim in the cadastre, Roman claimants probably tried to use their position as taxpayers to perfect their title.¹⁹ But, given the lack of independent reviewing and purging of the information contained in the cadastres, judges did not rely on them (Ellickson in this volume, n. 65),²⁰ so property acquirers had no reason to rely on them either. In particular, even if the vast majority of the claimants recorded as owners in the cadastre were the true legal owners, acquirers could not rely on such records to reduce the information asymmetry that they suffer with respect to sellers. Therefore,

¹⁹ Indeed, payment of the land tax seems to have been considered proof of ownership later, in the Byzantine empire (Laiou and Morrisson 2007: 50). This is reminiscent of current practice in some jurisdictions: “In Illinois, for example, an adverse possessor may establish his claim merely by paying taxes on the property, at least against an owner who is familiar with real estate practice and records” (Rose 1985: 80). A mention in the Digest also refers to relying on the census for settling boundary disputes in the absence of evidence on alterations (D. 10.1.11; Papin. 2 resp.). ²⁰ Considering the reluctance of judges to grant conclusive effects to evidence produced by property registries, which has been observed in different countries and historical circumstances (Arruñada 2012: 241, n. 39), the prevalence of fraud in public Roman documents (Moreau 1994) may have added another difficulty.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

266

Benito Arruñada

it might well be the case that a variety of public registries did indeed exist for all the purposes mentioned e.g. by Nicolet (1988: 213‒40), and that the authorities did have lists that allowed them to identify “owners” (in fact, claimants) to the level of legal certainty required by administrative law in matters such as tax collection or street maintenance. However, it is equally likely that such lists provided little valuable information for ascertaining legal title to property. Second, Rome also did extensive land surveying and recording during its complex and repeated processes of land redistribution, which relied on lists of beneficiaries and land parcels at different levels (e.g. Moatti 1994). However, these were not prepared for facilitating private transactions; therefore, labeling them as land registers may be misleading (as in e.g. Duncan-Jones 1976). Even if a double set of records with different information may have been used, relying on bronze for the maps and more perishable media for more detailed information, including the beneficiaries’ names and the surface area assigned to each (Nicolet 1988: 219‒21), the information was not updated for property transfers, so that “to clearly establish property rights, it was necessary to perform all of a series of searches to find their origin” (Nicolet 1988: 221). In addition to mapping and administrative functions, the forma played a legal attestation function “with the reserve, however, that not containing the transfers, it must be, in case of title conflict, completed by other means” (Moatti 1994: 109). Third, securities granted to public bodies over private land were recorded by different means (Verhagen in this volume). However, the purpose of this archiving might well have been merely that of enforcing the state’s claims, instead of publicizing them to potential third parties. From a land titling perspective, this archiving by the state would therefore have served the same purpose as the safekeeping of contracts and property deeds by landowners and lenders, and, even if the rightholder was the state, it was in essence of a private nature. When land charges were inscribed in stone or bronze, as in the alimenta schemes of the Veleia and Ligures Baebiani tablets, by which the state granted loans to landowners, the returns from which were then used to maintain poor children, a publicity purpose is more likely, but the function of such publicity is also unclear. In addition to that of cautioning potential land acquirers that the land was subject to the alimenta charge, publicity might also have had euergetic purposes—i.e. to show that the emperor was doing

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

267

good, not only caring for children but also providing a stimulus to agriculture (Lo Cascio 2000: 223–64). Fourth, there seems to be growing consensus that the practices of Egypt were more representative of the rest of the Roman Empire than it was thought in the past and that available evidence may be seriously biased by the type of media used (e.g. Ando 2010: 179). However, this bias does not explain the lack of references to land registers in both the preserved records of contracts documenting private land transactions,²¹ and in the legal texts (until the last centuries of the Empire).²² Both absences suggest that at least in the West and during the classical period of Roman law, land registration did not play a significant role in private contracting. Considering that jurists often minimize the role of modern registries in property transactions (Arruñada 2012: 114‒18), therefore exaggerating their own role, which is diminished in the presence of effective registries, one may wonder if Roman jurists simply omitted mentioning the titling role played by property registries in Roman times. It seems unlikely, however, that such bias would have left no trace of registries. Modern treatises teach most property law as if private conveyancing (i.e. without registries) were still the dominant paradigm. However, all of them need to cover conveyancing “in registered land,” to use the

²¹ For instance Adams (2013); Camodeca (2013); and Rotman (2013). This is in contrast to those found in Egyptian contracts (Yiftach-Firanko 2013). Private safekeeping of documents might have been prevalent even for publicly issued documents, given that “it does appear that the safest way to preserve a decision that concerned you was to take a copy of it to yourself when it was promulgated; and this is in fact how our surviving documents have mostly come down to us” (Crook 1967: 33). This private safekeeping of public documents led to forgeries (Moreau 1994). ²² The involvement of public authorities in private legal transactions increased in the last centuries of the Empire, with some authorities recording private documents and issuing copies of their records that enjoyed privileges as proof (Harter-Uibopuu 2013). In 313, Constantine enacted publicity for conveyances, with a clear fiscal purpose (Honoré 1989: 142–6), and in 323 he made the public registration of donations mandatory. A century later, Valentinian III (425‒55) required the registration of conveyances, allegedly with little visible effect with the exception of the papyri of Ravenna, whose registration is attributed to this city being an important capital and administrative center (Everett 2000: 73‒6). However, it is unclear how prevalent these solutions were both before or even after Constantine. Other contractual evidence on property, dated at the end of the fifth century, during the period of Vandal rule in Africa, does not mention registers (Corcoran 2013). Several factors indicate that these institutional changes had little, if any, impact on reducing transaction costs: their fiscal objective, their coincidence with the decline in market activity, and their apparent failure.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

268

Benito Arruñada

English terminology. They tend to present a distorted view of the two modes of conveyancing but for obvious reasons they need to present both. Similarly, considering practices in medieval European cities, one may think that local courts kept written records of in iure cessio transactions, which might then have been used as a sort of land recordation system. But even this possibility remains unlikely because such transactions were rare (Gai., Inst. 2.25) and reliance on such records should have left some traces in the legal sources. Nor do the municipal registration practices observed in later centuries seem to be relevant with respect to classical Roman law, because the earlier available evidence on municipal land recordation in the West, the papyri of Ravenna, dates transactions starting in 445 CE (Everett 2013). Lastly, suggestions that lack of land registration in Rome was due to an overriding preference for privacy seem unsupported. For instance, Temin argues that “there was no registration of property ownership on behalf of the state, as in Egypt and probably other provinces too. Roman law and administration operated in a quite different tradition of civil honesty compared to the prying bureaucratic registers of Hellenistic kingdoms” (2013: 184). However, reluctance to publicize transactions seems to be absent in the legal texts and commentaries (while it abounds e.g. in modern European discussions), and, perhaps most importantly, traditional Roman rules on conveyancing did require publicity, as examined in the next section. Therefore, the reason probably lay in other factors. As argued below, more effective palliatives and stricter enforcement of personal obligations possibly made registration less necessary in Italy and the West than in Egypt and the Hellenistic kingdoms.

19.5. PALLIATIVE SOLUTIONS IN PROPERTY The Roman law of property provided a variety of solutions for contracting rights in rem, enabling the type of asset-based impersonal exchange which is a main focus of this work. Allegedly, these solutions changed substantially over time in parallel with changes observed in the economy and, in particular, with the expansion of long-distance trade.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

269

19.5.1. Ceremonial Publicity: mancipatio In principle, to transfer ownership of the most valuable goods, such as land, slaves and cattle,²³ Roman law required a formal and public conveyance, either through a collusive proceeding before the court (in iure cessio) or through a public ceremony known as mancipatio. In the latter, the seller and the acquirer appeared upon the property before a person holding a scale and five witnesses. The acquirer then declared the property to be his, struck a token coin against the scales and handed it to the seller, who said nothing. This ceremony produced a public conveyance which was separate from the private contract (the two-step contracting process characteristic of property, as analyzed in Arruñada 2003). In this case, the public step served at least two major purposes. First, it triggered the start of a prescription period that can also be understood as a purging procedure. Contradictory in rem rights were downgraded to in personam status unless rightholders opposed the intended transaction either during the ceremony or within the prescription period. Second, it provided the basis for future legal procedures, as the law granted substantial procedural advantages to possessors against other claimants. The ceremony therefore had a titling function as it publicized the conveyance, but also served to gather the consents of affected rightholders and purge conflicting claims either at once or after a certain period. In contrast to the creation of rights in personam by formal stipulatio, which did not in principle require any witnesses (Nicholas 1962: 104), mancipatio required numerous witnesses, and this “was probably needed not so much to prove that the act had taken place as to give it publicity so that any defect of title (which in a small society would be likely to be known to the witness) could be investigated immediately” (Nicholas 1962: 256).²⁴ Indeed, in such a context, witnesses often performed several functions, including, in addition ²³ The separation made economic and legal sense: given their value and durability, res mancipi were assets suitable for enforcing multiple rights in rem. For this reason, it made sense for the legal system to require publicity when transferring them. Views considering that the separation was driven by the legal requirement probably reverse the direction of causation. The inclusion of some types of movables in res mancipi should not be seen as a “useless but ubiquitous complication” (Crook 1967: 141), because those movables were high-value capital goods suitable for holding multiple rights on them. ²⁴ It has even been argued, in the context of documents formalized in tabulae, that “these testes or superstites, as they were called, were not witnesses in our sense,

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

270

Benito Arruñada

to merely testifying, those of pointing out possible title clouds and granting their consent. Acquirers achieved some title protection by choosing the right witnesses: those with a good reputation (when adjudicating a case, judges considered the reliability of witnesses)²⁵ and specially those who might know or even directly hold claims on the land (including e.g. relatives and neighbors). These potential claimants, by acting as witnesses, are implicitly consenting to the transaction.²⁶ This also holds for boundary disputes, for which, in Rome, “though not a legal requirement, it was standard practice to call the neighbors to witness a conveyance by delivery. This allowed the transferee to verify that the transferor and his neighbors agreed on the boundaries, or, if they did not, to pinpoint the area/s of dispute. If the dispute was not settled at the time, the conveyance could be confined to the agreed area, or if necessary postponed” (Honoré 1989: 139). Lastly, even if mancipatio did not in principle require delivery of possession (Buckland 1912: 95–6; Gai., Inst. 4.131), the transferor had a duty to deliver it (Honoré 1989: 139). In most cases, if the transferee did not obtain possession publicly, it was not protected against dispossessors by possessory interdicts, this being a speedy procedure by which the judge would adjudicate on possession without looking at title, forcing possible owners to rely on a more cumbersome action, that of vindicatio, in which the possessor would be the defendant and would not therefore bear the burden of proof (Nicholas 1962: 108–9).²⁷

expected later to testify to what they had seen or read, but judges, expected to stop an act at the time of its making if the performance were flawed” (Meyer 2004: 118). ²⁵ Interestingly, witnesses sealing the Campanian tablets were ordered hierarchically, starting with those of higher status (Meyer 2004: 118). ²⁶ Referring to conveyances in Ptolemaic Egypt, it is said that “the seller identifies those most likely to challenge the title of the new owner as his children and siblings” and “children were present and consented to the transaction, or ceded their claim to the property” (Muhs 2015). ²⁷ It must be kept in mind that, in Rome, possession was narrowly understood to mean “not simply the holding of a thing but rather the holding of a thing in the manner of an owner, the exclusive holding of a thing” (Nicholas 1962: 111). Those holding without possessing had “detention”: they were physically occupying the land but lacked relation to it (ibid. 112). This was mere possessio naturalis (Buckland 1912: 77).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

271

19.5.2. Evolution towards Privacy Conveyance by mancipatio therefore served for providing some degree of publicity, gathering consents and purging property rights. However, public ceremonial conveyances become less efficient with economic development.²⁸ First, their cost increases if parties move more and need to purchase from a distance. Second, they are more effective in local markets, where transactions take place between neighbors, as is common in rural societies, than in wider markets: e.g. “the six participants in the mancipatio, though adequate enough for a small community, could constitute no hindrance to secrecy in so vast a society as imperial Rome” (Nicholas 1962: 104). Indeed, for neighbors, it is easy to notice announcements and public deals, especially for the kinds of rights common in rural societies, many of which are linked to family and household matters. This is probably less so for non-neighbors, as indicated by the longer prescription periods usually granted to those absent, apparently balancing costlier knowledge with longer time. Figure 19.2 represents what might have been the historical evolution in Rome towards more private conveyancing. At the starting point, most land is conveyed by mancipatio, with some public titling for the highest-value land through the judicial procedure known as in iure cessio. I hypothesize that the development of a wider market for land had two effects: first, it increased the cost of holding public ceremonies, which drove the mancipatio line downwards; and, second, and probably more important, it made such ceremonies less effective, reducing the line’s slope. With this move of the mancipatio line, a more private titling procedure, using increasingly private delivery and known as traditio (or even, in the late Empire, merely granting a written deed),²⁹ a procedure which previously was

²⁸ Moreover, two other factors may be present, at least in the short term. On the one hand, there is a permanent demand for privacy by economic agents, because of reputational concerns and tax avoidance, a demand that could be uncorrelated to economic growth. As argued by Nicholas, “there was a similar struggle to achieve secret conveyancing in English law, culminating early in the seventeenth century in the recognition of the device of a bargain and sale for a term followed by a release” (1962: 104 n. 3). On the other hand, conveyancers—mainly lawyers—may have an interest in privacy solutions to the extent that they increase the demand for their professional services. ²⁹ Procedural law is also important here. In post-classical times, the more bureaucratic system of judicial procedure known as cognitio extra ordinem ends up

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

272 Value of property right with titling conflict

Benito Arruñada 45°

Judicial titling

Mancipatio (1) Traditio

Mancipatio (2)

Value of property right without titling conflict Indifference points between judicial titling and mancipatio (1), right, and traditio, left

Figure 19.2. Hypothesized evolution of titling when, after distant trade becomes more common, mancipatio becomes costlier and less effective and is replaced by traditio. Note: The figure aims to represent the historical evolution caused by the lesser efficiency of mancipatio with the development of markets, which led to the predominance of conveyancing through traditio, characteristic of classical Roman law. Source: Adapted from Arruñada and Garoupa (2005: 717–18, figs 2 and 3).

suboptimal, became individually optimal for land parcels whose value lay to the left of the new indifference point.³⁰ A testable prediction of the analysis is that the abandonment of mancipatio should lead to an

superseding the old formulary system. After Diocletian (284–305), documents issued by a scribe started to constitute a proof superior to witness testimony (Thür 2013). However, the role of written documents—mainly, on wooden tablets—may have been underestimated in legal texts and scholarship (Meyer 2004: 112–20). ³⁰ Note that, in this context, relying on traditio could be individually optimal even if mancipatio had remained more effective than deed titling in terms of title assurance. (Imagine that in the figure the traditio line had simply moved downwards—and, therefore, become more costly—while remaining parallel to the old one.)

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

273

increase in the number of judicial procedures, which would then also be used for land values between the two indifference points.

19.5.3. Exercise of Possession as Access to Ownership: usucapio To the extent that the exercise of possession has observable consequences, it is capable of providing the basis for titling, by having it upgraded to ownership with the lapse of time—by acquisitive prescription.³¹ In Rome, initially, undisturbed possession of land and other capital goods for only two years (one year for other movables) led to ownership by usucapio if acquired in good faith and based on proper cause, even if conveyed by simple traditio (i.e. delivery of possession) without mancipatio, and with the only exception of goods stolen or taken by force. Furthermore, during the late Republic the praetor enhanced the protection enjoyed by the good faith possessor on the way to usucapio by granting him an action that ensured he succeeded even against the formal owner (if the possessor’s title was formally defective: a recipient of land by informal delivery or traditio instead of formal mancipatio) or against everyone but the owner (if the title was substantively defective because e.g. he had bought from a non-owner). As a consequence, “the recipient of a res mancipi by traditio was for nearly all practical purposes in the position of an owner” (Nicholas 1962: 127). It is believed that, thereafter, informal delivery by traditio replaced formal mancipatio.³² The key element for usucapio to work effectively as a titling procedure is that the exercise of possession provides claimants with the information they need to litigate and protect any property rights infringed by the intended transaction (e.g. Rose 1985: 78–81). It therefore implicitly provides the basis to start a selective purge and ³¹ For a more extensive treatment of the titling role of possession, see Arruñada (2015). ³² In fact, it has been argued that mancipatio was retained longer for land, at least until the fifth century, because it did not require the presence of the parties on the spot (Buckland 1912: 73). However, it is likely that the mancipatio term appearing in later Roman documents really means only conveyance (Nicholas 1962: 117). Meyer argues that formal conservatism (syntax, style, rhythmic prose) in Roman documents conveyed authority (Meyer 2004: 59). Her argument might help explain why the word mancipatio was still used even when the title was transferred by granting a written deed.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

274

Benito Arruñada

acquirers need to check only if the seller has effectively been in possession for the prescription period. In this context, doubts about the information provided by possession may arise, especially if the delivery of possession is unobservable. Indeed, traditio was increasingly made without any visible transfer of the good. This, together with the decreasing use and limited efficacy of mancipatio in a vast economy, leads some authors to think that Roman publicity was ineffective (e.g. Nicholas 1962: 104). However, this critique may be unwarranted because, for existing rights to be protected under usucapio, possession only needs to inform affected rightholders with sufficient time for them to take action. And owners can always make their possession public if, foreseeing a possible sale, they want to convince future acquirers that they will be protected by the usucapio of the seller. It is consistent with this titling view of possession that the time for usucapio increased when the information provided by publicity might have taken longer to process. Standard usucapio was part of the civil law and therefore applied only to Roman ownership: i.e. Italic land held by Roman citizens. For provincial land, longer periods were required: for longi temporis praescriptio of land, ten or twenty years when parties were or were not in the same district (Nicholas 1962: 128). This makes sense because of the greater distance and time probably needed by adverse claimants to get notice. The time for usucapio also became longer in the last centuries of the Empire. At the time of Justinian, three years were required for usucapio of movables, longi temporis praescriptio became the only usucapio for land, and a longissimi temporis praescriptio of thirty years was introduced for good faith acquisitions even for things that had been stolen (ibid.). This lengthening of usucapio periods at the time of Justinian seems consistent with greater civil unrest and consequent information and purging difficulties.

19.5.4. Discussion The public ceremonies of mancipatio, initially, and the sophisticated use of possession and usucapio seemingly provided an effective palliative for transfers of ownership. The limitations of such a palliative are most visible in the inability of Roman law to implement efficient real securities, which is understandable due to their abstract nature

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

275

and the consequent difficulties of making them publicly known by such means. Given the lack of registers,³³ to provide lenders with real security, mortgages were instrumented on the basis of the fiducia.³⁴ This therefore merely reduced the default risk of the lender by creating a dispossession risk for the borrower, which helps explain why they most often resorted to personal sureties.³⁵ Alternatively, Romans could, of course, easily add real securities to their personal obligations.³⁶ However, due to the possibility of hidden charges, the ³³ Efforts to make hypothecs public in the West came quite late: not until 472 CE was priority granted to hypothecs created before a public authority or before three witnesses (Nicholas 1962: 153), therefore with an element of publicity. ³⁴ A fiducia, “like the late medieval English mortgage, was a conveyance subject to a covenant for reconveyance on payment of the debt” (Nicholas 1962: 151). In fiducia, the owner gave ownership to the creditor by mancipatio or iure in cessio, usually including pacts for reconveyance on payment of the debt, authorizing the creditor to sell if the debt was not paid and allocating the possible surplus to the borrower (e.g. Crook 1967: 244–5). There is also evidence in the Digest that at some point real securities might have been reinforced, as in old Mesopotamia and modern England (Arruñada 2012: 46–7), by pledging the deed or the whole chain of deeds with the creditor (Scaevola D. 13.7.43 pr.; Scaevola 5 dig.). This pledging of deeds also subjects the debtor to the creditor’s moral hazard, as he may fail to produce the deed when the debtor needs it or may even fraudulently sell, and also makes it harder to subrogate the debt and contract second mortgages. The Digest case makes this point clear, as the lender fails to produce the deed to the borrower, who is therefore unable to prove the extent of his land and cannot fully develop it. ³⁵ Personal security was very common for any substantial transaction on credit (Nicholas 1962: 151). It has even been argued that Romans preferred personal rather than real securities because of status: “a wealthy Roman’s word was his bond, and security as potent as any pledge” (Johnston 1999: 88). Understandably, posting a real security might even have been detrimental to personal reputation, as it might signal that the debtor’s word was not valued by potential creditors and, worse still, that he had no friends willing to back him. From this perspective, reluctance to accept publicity would have been limited to the posting of real securities, not to mere indebtedness; and such reluctance would be hard to reconcile with the abovementioned prevalence of circular credit networks among friends. ³⁶ Compare Verhagen (in this volume), for whom “the Roman law of real security was capable of facilitating credit in a similar manner as its descendants in modern economies” and, in particular, offered “efficiency in creation” because parties could create a variety of securities very easily. However, ease of creation is, at most, a partial and doubtful advantage. Partial, because such ease ex ante comes at the price of, first, information asymmetry at the time of contracting (given the possibility of hidden charges and title clouds, lenders know less than borrowers about the value of the security); and, second, ineffectiveness of the guarantees ex post (if a prior hidden charge appears or a title risk materializes) and costly enforcement to clear title before repossession (even if hidden charges or title clouds do not surface). It is also a doubtful advantage when considering that total transaction costs may well be lower when the law reduces the variety of real rights by means of a numerus clausus of such rights (Heller 1999; Merrill and Smith 2000; and Hansmann and Kraakman 2002).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

276

Benito Arruñada

enforcement of such securities was in fact most likely based on the personal sureties offered by the borrower and his guarantors, not on the collateral value of the borrower’s land.³⁷ Therefore, the presence and sophistication of real securities in surviving contracts, or their treatment in legal texts tells us little about their economic relevance.³⁸ Similarly, and whatever its effectiveness, the legal decision to impose harsher or criminal sanctions on those who hid such charges can also be seen as a sign that secrecy and fraud were hindering these transactions. Other controversial issues are also illuminated by the titling role played by possession. For a start, analyses that consider that granting ownership by prescription based on possession and the lapse of time “makes possible the proof of ownership by depriving it of its ordinary meaning” (Nicholas 1962: 156) disregard that the public nature of possession allows claimants to vindicate ownership during the prescription period. It therefore implicitly provides the basis for a judicial purge of ownership claims: ownership does not exist as a titling- (i.e. judicially) independent reality. Similarly, seeing possession as a main element of the titling process denies claims that the importance Roman law attached to possession rather than ownership “indicates that Roman law was seriously concerned with preservation of the status quo and keeping the peace, and less so with questions of formal entitlement” (Johnston 1999: 60). Possession was a tool to ascertain which claimant would be the owner. Ownership claims should not be confused with ownership rights. Lastly, when considering the role of possession in the titling process, views of a two-part land tenure system in Roman land law also seem unwarranted. It is said, for example, that “jurists and legal commentators described a complex two-part law of land tenure. There were two sets of rules,

³⁷ Interestingly, the alimenta schemes mentioned above provide some indication that (at least at the time for which most evidence on these schemes is available—early second century CE) the collateral value of land was low, as owners had to provide a security which has been estimated at an average of 12.5 times the amount of the loan (Fernández 2013: 315). Low collateral value is consistent with the conjecture that real securities were of little effect in reducing default risk. ³⁸ Verhagen (in this volume) bases its positive assessment of Roman real securities on two facts: the Sulpicii bankers used real securities for large loans, and the extensive treatment of real securities by jurists, as well as in the Digest and the Codex. However, the Sulpicii relied on pignus—transferring possession to the creditor—an old form of real security. And extensive legal treatment of a given topic may well be an indication of conflict rather than low transaction costs, as shown, for instance, by the examples on chirographa taken from the Digest by Verhagen in this volume.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

277

one for what has been labeled ownership and one on possession. The line between these categories is not clear, and one puzzle for modern observers is to know which set of rules was applicable in any particular case” (Temin 2013: 147). From a titling perspective, this duality makes sense not as a two-part tenure system for different land but as a two-step titling process for the same land: possession was simply the publicizing element in the second step within the titling process ending with usucapio. Possession was the path to ownership, not its alternative.

19.6. PALLIATIVE SOLUTIONS IN BUSINESS Sequential exchange is prevalent in business, because using contractual agents capable of committing their principals often provides substantial specialization advantages, expanding the scope and coverage of business operations. This explains why in many situations the strict enforcement of property rights is superseded by “contract” rules granting third parties rights in rem or giving them priority access to the principals’ assets—i.e. making the principal liable for the agent’s contractual obligations.

19.6.1. The Palliatives of Contractual Agency However, early Roman law lacked a proper law of agency, as “the strictly personal character of the Roman obligation . . . made agency impossible [in the sense of creating a direct relationship between principal and third party], and the mandatarius alone was both liable and entitled on any contract which he made” (Nicholas 1962: 189). Contracts therefore could only negatively affect their parties: what in English law is called the “privity” of contract was rigidly applied (ibid. 199).³⁹ This strict enforcement of what I have labeled the “property” rule made sense in a personal market with few trade opportunities but ³⁹ Initially, the only exception were rights acquired by slaves and filiusfamilias, which vested in the paterfamilias (Nicholas 1962: 199). This asymmetry in the rule (principals not liable but able to acquire rights through their sons’ and slaves’ actions) made sense because the third parties granting a right to the slave owner were themselves owners; therefore they had ultimate control of the situation and could

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

278

Benito Arruñada

not after trade opportunities substantially increased in the second century BCE.⁴⁰ Consistent with this trade-expansion hypothesis, praetorian edicts introduced important changes diluting such strictness, by making principals liable and therefore protecting innocent third parties. The key innovation was to enhance the role of slaves and sons as economic agents, in two opposite directions: on the one hand, making masters and fathers liable for the acts of their slaves and sons; and, on the other hand, by limiting the masters’ and fathers’ liability to the assets assigned to the sons or slaves. I will argue that these solutions were close to modern solutions both in their function and their operation, as both were based on appearance and publicity. First, consistent with the analysis in §19.1, the principal was made liable when it was clear he had granted his consent and therefore authorized the agent. Since the second century BCE, by the remedies later referred to as actiones adiecticiae qualitatis, praetorian law made the paterfamilias liable for obligations entered by his sons or slaves, with his liability increasing with the authority he had granted or with his involvement in the businesses run by his sons and slaves (Nicholas 1962: 202; Abatino et al. 2011: 374–5). In particular, he was liable in full when he had authorized the agent, either explicitly, implicitly or by allowing similar dealings: not only when the master ordered the transaction, but also when he put the agent in charge of a business or a ship. In such cases, the scope of liability was limited to contracts made in connection with the business, and could be narrowed by giving proper notice (Crook 1967: 190; Johnston 1999: 103).⁴¹ Moreover, similar principles were extended outside the household when a freeman was appointed as a manager or institor (Nicholas 1962: 203–4). And at least the actio institoria made it possible to commit take good care of their interests. With little potential for conflict, this solution must have facilitated some degree of exchange and specialization, as slave owners needed to be present to commit themselves on any obligation but did not need to be present to acquire rights. Indeed, we find similar asymmetries in today’s economy: banks are often allowed to modify the land registry, but only against their own interest (e.g. by filing a mortgage cancellation). ⁴⁰ According e.g. to archaeological evidence on the number of shipwrecks (Parker 1992). ⁴¹ “The area of competence of the agent was defined in a charter (lex praepositionis) or by custom or common sense; it could be restricted by posting at the workplace a written sign (proscriptio) in any language, or extended to specific activities by a special invitation (iussus/-um) addressed by the principal to third contracting parties” (Aubert 2013: 3466).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

279

the principal on the basis of appearance (Petrucci 2003), so it played a role similar to that of possession in property. The principal was protected only by an explicit prohibition and the rules seems to have evolved in such a way that the master was liable, and, therefore, third parties were protected, when the agent exceeded his powers through actions not explicitly forbidden and made adequately public by the master (Petrucci 2003: 607–10, discussing D. 14.5.8; Paul. 1 decret.). Roman law was therefore considering the same basic elements of consent and publicity that we find today in business agency (Arruñada 2010). Second, the principal could limit his liability to the part of his wealth managed by the agent (the peculium),⁴² therefore enjoying some degree of “asset partitioning,”⁴³ but with a complex set of legal actions that arguably resembled contemporary remedies aimed to tackle standard conflicts of interests between shareholders and company creditors, such as—in current terms (Johnson et al. 2000; Djankov et al. 2008)—the abnormal distribution of dividends and selfdealing,⁴⁴ or the “tunneling” of corporate resources to insiders.⁴⁵ In general, the rules seemingly tried to link the liability of the principal to the benefit he was obtaining from the business (Crook 1967: 189). In terms of §19.1, this normative linkage to benefits avoided the typical attempt by principals to get rid ex post of their own ex ante commitments (i.e. to have a property rule applied), once the business entered by their agent went against the principal’s interests.

⁴² The peculium was a separable set of resources which, even if by law it pertained to the master, was controlled by the slave and was first in line in terms of satisfying the contractual commitments made by the slave to third parties. Even if slaves’ peculium was more important in economic terms, “it is overwhelmingly probable that married sons living independently had such a fund” (Crook 1967: 110). ⁴³ See Hansmann and Kraakman (2000); and Hansmann et al. (2006). Having a slave operating a business provided masters with the possibility of delegating the management of business activities and benefiting from direct agency, limited the master’s liability, provided some “entity shielding” against the master’s creditors, and helped to ensure the continuity of business after changes in ownership and management. See Abatino et al. (2011), who find evidence of entity shielding between different businesses of the same owner. ⁴⁴ e.g. this could be behind the rule constraining the discretion of the master to pay himself to the detriment of the slave’s creditors, according to Gaius (Inst. 4.72). ⁴⁵ e.g. the rule making the master liable for payments made to him from the peculium (Gai., Inst. 4.73).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

280

Benito Arruñada

19.6.2. Discussion Many aspects of the Roman law of business agency are subject to alternative interpretations, however. A major question is why sons and slaves played such a prominent role as contractual agents, which suggests that they enjoyed a comparative advantage over individuals sui iuris, those who were “in their own power” and not, like sons and slaves, under the patria potestas of a paterfamilias. This advantage seems to have lasted at least until the early third century (Johnston 1999: 106) and is supported by the occasional practice of free-born individuals becoming slaves to act as chief manager or accountant of a big household (Crook 1967: 60; Ramin and Veyne 1981). It has often been argued that the advantage was linked to failures in formal law—i.e. to the incompleteness of agency-independent labor (Nicholas 1962: 203; Johnston 1999: 105–6)—but the survival of this formal constraint seems unlikely, considering the adaptability shown by praetorian law. Alternatively, from the perspective of §19.1, this comparative advantage may have been rooted in lower transaction costs in either their originative and/or their subsequent contracts: by easier contracting of originative agency and/or subsequent trade—i.e. by better incentives for agents and/or lower information asymmetry for third parties. First, the incentives provided by family (i.e. household) and slave law lowered agency costs, not only by making harsher punishments available, which may be suitable in conditions of information asymmetry (Dari-Mattiacci 2013), but by judicial forbearance that implies a radical asymmetry of decision rights, as I will argue in §19.7.2.1. Second, using sons and slaves as contractual agents also reduced information asymmetry in subsequent transactions if they were easier to identify as such by third parties and, in particular, their status was not easy to modify ex post, thereby preventing principals from freeing themselves from their undesirable commitments. As argued in §19.7.2.2, this explains the structure of the formalities which dealt with household membership. The advantage seems more nuanced in terms of subsequent transactions with third parties: third parties may have been more reluctant to contract with filiusfamilias and slaves to the extent that in some circumstances they would have had access only to their peculium, which made recovery harder (e.g. Gai., Inst. 4.74). However, for the

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

281

same reason, such contracting was safer for the paterfamilias, suggesting the presence of a tradeoff which was arguably easy to optimize by modifying the public involvement of the principal in the business and, consequently, his liability. There also seems to be substantial discrepancy on the prevalence of corporate entities in the Roman Empire. On the one hand, through the contract of societas (e.g. Riggsby 2010: 13), someone selling to the partnership was selling to the individual partner with whom he was dealing. If he was not paid, he had no recourse against other partners. In general, “legal personality was not available for private businesses” (Abatino et al. 2011: 368). On the other hand, big tax farming companies created by publicani enjoyed special rules: e.g. death of a partner or litigation between partners did not dissolve the company (Crook 1967: 234). It has even been claimed that there were proper companies, at least for tax farming and shipping, and that they were functionally similar to the European joint-stock companies of the sixteenth and seventeenth centuries, with their own legal identity and continuity after the death of their investors, with separation of ownership and control and with shares being traded at variable prices (Malmendier 2009). Given that at least partnerships for tax collection and building projects were governed by special rules that included incorporation and entity shielding, the legal and judicial expertise for contracting them was available: the basic “technology” for corporate contracting had already been developed. However, in connection with the argument in §19.7, corporate contracting is exceedingly costly when, in the absence of company registries, it is based purely on informal publicity (Arruñada 2010: 556–8), as shown by the experience of English unincorporated companies during the Industrial Revolution (R. Harris 2000). Even if the company form was available to the Romans, it must have been ineffective when used without some form of public chartering. The historical record is consistent with the interpretation that, as in the modern English case and given the lack of company registries, large corporations were viable when based on the public chartering process leading to the concession grant but were hardly viable on a purely contractual basis.⁴⁶ However, in ⁴⁶ Compare Malmendier (2009), who omits any consideration of the difficulties faced by companies created on a purely contractual basis and emphasizes political considerations in the demise of the large societas publicanorum in the Roman Empire.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

282

Benito Arruñada

contrast with the English case, in which we observe for decades unsatisfied demand for proper incorporation, leading to the solutions analyzed by R. Harris (2000), there are no indications of such unsatisfied demand in Roman times. The analysis in §19.2 therefore provides a more systematic explanation as to why corporations were not used for other activities and industries. The most basic reason might have been that there was little need to use them for such activities. There was not enough demand for more stable business associations: available technology did not require the large amounts of capital and continuity that, for instance, infrastructure projects and railroads were to require at the time of the Industrial Revolution.⁴⁷ The situation would have corresponded to the one represented by interpreting impersonal exchange as the corporate form of business organization and personal exchange as non-corporate organization. In the Roman context, few business opportunities lie to the right of the horizontal axis, making—even under voluntary registration, which makes the cost L irrelevant—the fixed cost of creating a company register (F) bigger than the benefits (G). Therefore, only a few large companies incorporated, relying on the publicity provided by the public calls and auctions linked to tax farming and government subcontracting.

They lost importance and ultimately disappeared because the state changed the way it collected taxes and produced public services, moving away from pure tax farming into more sophisticated forms of subcontracting, with greater control of subcontractors, sharing of revenues and development of the state’s bureaucracy (Gibbs 2013). It is not clear that these changes in public finance worsened the environment for the market economy, which in fact flourished during at least a couple of centuries after them; but, whatever their causes and merits, the most pressing question is why the corporate form used by tax farming did not prosper in other activities and, in particular, why this form was not used for organizing other private economic activities either before or after such changes in fiscal policy. (The existence of corporate entities in shipping, alleged by e.g. Malmendier (2009: 1089) and Temin (2013: 103) on the basis of a statement attributed to Cato by Plutarch, who wrote almost two hundred years later and might have been prone to exaggeration (see, for another likely example, Ellickson in this volume), seems to be grounded on a weak, temporary, basis (Verboven 2002: 285), even if shipping is an activity that allegedly poses relatively discontinuous and therefore simpler contractual problems.) ⁴⁷ Compare to the firm size constraint argument in Abatino and Dari-Mattiacci (in volume I).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

283

19.7. EFFECTIVE ENFORCEMENT OF PERSONAL OBLIGATIONS Both economic transactions and legal systems substitute between impersonal and personal exchange. When impersonal exchange is impossible, transactions are either made relying on personal safeguards or are not made at all and become lost opportunities. In particular, the harder it is to enforce rights in rem, more rights will be enforced in personam. For example, when real securities for credit are weak, credit will tend to be more personal and transactors will rely more on personal sureties. Informal norms and formal rules might also be adapted to facilitate such substitution. If a society relies more on personal exchange, one could expect individuals’ reputation to play a greater role and personal sanctions to be harsher (Arruñada 2012: 50–2). This was seemingly the case in ancient Rome, where informal social norms motivated people to fulfill their personal obligations. Interestingly, both formal rules and organizationally supported institutions were also used to reinforce such informal social norms.

19.7.1. Informal Norms and Formal Rules Roman society placed great importance on personal honor and reputation, which was the basis for strong horizontal and vertical bonds that provided safeguards for all sorts of exchanges. As a consequence, fulfilling contractual obligations was a matter of personal honor (Temin 2013: 12). Horizontal bonds meant that substantial mutual help was customarily provided by people equal in status, mainly under the informal institutions of amicitia and officium: respectively, the informal bond and the obligation of friendship. “Friendship (amicitia) gave rise to serious and substantial duties. Roman friends made claims on each other which would cause a modern ‘friend’ to break off the relationship without delay” (Schulz 1951: 555).⁴⁸

⁴⁸ There are signs that informal bonds became less effective over time, as indicated by Seneca’s complaint in the first century CE that fides was no longer enough and debt now needed to be formalized by means of written and sealed documents, with the seal giving increased physical protection (Meyer 2004: 156).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

284

Benito Arruñada

Vertical bonds amounted to several layers of patronage by people of higher status, who acted as patrons for their lower-status “clients,” the clientela. In the early Republic the situation was close to serfdom, but survived later mainly for the relation between patron and freedman (Crook 1967: 93). Moreover, informal personal ties were also important among Roman merchants for gathering information relevant for contractual enforcement. Traders often belonged to the same social groups (Temin 2006: 139), were able to exchange information (Kessler and Temin 2007) and, to reduce information asymmetries, relied not only on personal ties, household members, and peer monitoring but also on formal guilds and some state institutions (Temin 2013: 100). These personal bonds were based on individuals’ informal reputation but were also supported in formal reputational mechanisms. Thus, losing reputation was subject to informal sanctions, but these soon acquired the formal version of ignominia and infamia, with a judicially imposed exclusion from the legal protections enjoyed by a Roman citizen (Crook 1967: 83–5). Interestingly, reputation itself was also formally protected: defaming another person was itself harshly punished with infamia. The role of reputation was reinforced with the census, which established a formal classification of individuals defining their rights and duties and including a negative mark for those who offended public morality. In this way, it provided a formal register of reputation. Lastly, the formal legal system also relied on and reinforced the informal system of social norms to the extent that the value that Roman judges granted to evidence depended to some extent on the social standing of those providing or witnessing it (Meyer 2004). Moreover, informal social norms were reinforced by a whole array of formal rules ensuring enforcement of personal obligations by several means, which included, to mention only a few, a self-help version of debtors’ prison, strict punishments, and sophisticated allocations of liability. Personal enforcement or self-help initially consisted of execution of the wrongdoer by the creditor, including victims of crimes, who were entitled to a payment from the wrongdoer. Later, creditors could only privately imprison defaulting debtors but were responsible with respect to other creditors (Nicholas 1962: 209–10). Judicial enforcement was also based on personal execution: the plaintiff was authorized to imprison the judgment debtor and sell his whole property (Crook 1967: 82–3; Johnston 1999: 108–10). After

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

285

326 BCE, a borrower could not pledge himself or a son as collateral for a loan (nexum), to become the lender’s slave in case of default, but debt bondage could still result from the debtor’s default (Crook 1967: 173). Lastly, the law can be understood to have tried to make personal obligations effective by increasing punishments in line with asymmetric information. In particular, an early legal rule imposed double damages for eviction of res mancipi (Nicholas 1962: 161 n. 1), which motivated sellers not only to disclose but also to discover and purge title clouds.⁴⁹

19.7.2. Institutions Empowering Private Ordering At a more general level, social norms protecting personal exchange were also reinforced by two sets of arrangements enabling the extended familia to act as a legal entity: first, those allocating most decision rights to its head, the paterfamilias; and, second, those defining the familia’s legal boundaries and, therefore, who could inherit and, crucially in the context of impersonal trade, act as his contractual agent, committing the familia’s assets to meeting his obligations.

19.7.2.1. The Role of Judicial Forbearance in Enabling Self-Enforcement First, concentration of property and decision rights in the paterfamilias (including those on slaves and filiusfamilias) defined an ambit of “forbearance” in which judges would not enter (using the term coined by Williamson (1991) to designate the refusal of courts to hear disputes internal to modern firms, for example, between their divisions). The refusal of judges to enter into intra-familia disputes allowed an asymmetric allocation to the paterfamilias of property

⁴⁹ It was also made a crime to mislead a creditor about prior charges (Johnston 1999: 93). Verhagen (in this volume) emphasizes that this happened only by virtue of imperial rescript practice as from the end of the second century CE. This relatively late criminalization could have been a response to greater incidence of fraud caused by the slow erosion of personal bonds. This weakening of personal bonds might also have been behind the earlier growing reliance on written and sealed legal documents and the increasing physical protection of such seals, as claimed by contemporaries (Meyer 2004: 156).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

286

Benito Arruñada

rights on other familia members.⁵⁰ This should have made it possible for familias to fully rely on self-enforcement in their internal transactions and to act as quasi-firms with respect to third parties.⁵¹ Several factors increased the importance of this system of selfenforcement. To begin with, it encompassed not only the familia stricto sensu but the wider clientela: “it would have been impious and unlawful for patrons and clients to litigate or testify against each other or to support each other’s enemies” (Verboven 2013: 1578). Second, judicial forbearance was enhanced by the paucity of mandatory rules about key aspects of family law, including, most prominently, substantial freedom of testation (e.g. Crook 1967: 118–27), which had obvious consequences in terms of stronger incentives for sons and other possible heirs or beneficiaries, including slaves. Third, the aforementioned reliance on self-help can also be understood as a form of judicial forbearance, to the extent that it was parties themselves who were in charge of enforcement. Fourth, this concentration of rights in the paterfamilias had consequences both inside and outside the familia. Inside, it facilitated enforcement. Outside, it helped achieving some of the “modularity” provided now by business firms (Smith 2006, 2009). For instance, reputation was linked not only to individuals but to the whole familia, alleviating possible horizon problems. Lastly, it makes sense that Roman law carefully protected the familia against possible mistakes when the young age of the paterfamilias could lead him to exercise these decision rights poorly. It did so not only with the institution of guardianship, applied to children younger than 14 and in a different form to adult women, but also with a “caretakership” which was common for minors aged between 14 and 25, probably to avoid the difficulties they would have faced for contracting, given that, otherwise, they could easily have the transaction nullified later on by alleging fraud (Crook 1967: 113–18).

⁵⁰ About agency, it is said that “these were relations that never reached the inside of a courtroom. Their entire tone precludes contract and suit, action and liability; yet they were most effective in fulfilling the roles and needs lawyers associate with agency” (Kirschenbaum 1987: 180, cited by Temin 2013: 111). Indeed, they precluded contract and suit, but only inside the familia, in a similar fashion to modern firms. ⁵¹ Asymmetric allocation of decision rights and judicial forbearance play an important role in today’s economy (e.g. for franchising, Arruñada, Garicano, and Vázquez 2001).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

287

19.7.2.2. The Role of Formalities with Respect to Household Membership The importance of the familia as a legal entity also explains the sophisticated formalities and cautions used by Roman law to define its boundaries. All contractually relevant entries and exits of the familia were subject to ceremonial formalities and, therefore, to publicity: adoption, which was commonly used to achieve succession by adopting adults who entered the patria potestas of their adoptive fathers; emancipation, which led to dissolution of potestas, and was often used as a sanction; as well as the purchase and manumission of slaves. The fact that the dominant form of marriage under the Empire was consensual and had no ceremonial requirement ties in with the argument: public notice was unnecessary because the legal status of the wife remained unchanged—that is, she remained under the patria potestas of her father if she was alieni iuris and she remained independent and retained her property if she was sui iuris.⁵² However, publicity was required for manus marriage, in which the wife was in the position of a daughter to her husband, as well as for any mancipatio, adoption, emancipation, and manumission. All of them were ceremonial and required the presence of witnesses or some other form of publicity (Nicholas 1962: 71–83; Crook 1967: 103–13; McGinn 2013; Scafuro 2013). Moreover, such witnesses often played a purging function, as explained above. Some adoptions were even subject to judicial supervision and popular vote: adrogatio, being the adoption of a person not under the patria potestas of another man, often produced the merger of two familiae, making public approval advisable, not only because it extinguished a familia (Nicholas 1962: 77) but possibly because it may have had serious consequences for third parties, mainly creditors. Lastly, the very prevalence of adoption in Rome, by indicating the extraordinary importance of succession, indirectly supports the argument as to the legal-entity nature of the familia. These requirements of publicity are understandable considering that these legal acts could potentially affect third parties: in the framework of §19.1, they are originative transactions which may ⁵² Even if form-free (Nicholas 1962: 80–1), marriage required a clearly manifested intention, customarily evidenced by two-step contracting (a more private engagement and a more public wedding) and elaborate public ceremonies (Crook 1967: 102–3; Hersch 2010; Arruñada 2012: 238–9 n. 14).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

288

Benito Arruñada

have serious consequences for subsequent transactions. The originative transactions were those establishing or modifying individuals’ status, such as birth,⁵³ emancipation, adoption, marriage, manumission, etc. The affected subsequent transactions were those in which the status of the individuals or their children was a key condition for their legal effects. For example, marriage determined the status of children, and adoption affected the right to inherit and the power of the adopted to commit the paterfamilias. The consequences of slavery and manumission were similarly important, as slavery also made a form of contractual agency viable and manumission gave the right to have born-free children: freedmen and freedwomen could marry (Crook 1967: 52) and their children were freeborn (see a case in ibid. 48–9).

19.7.3. Discussion Based on this analysis of the enforcement of personal obligations, it is worth considering a possible connection between the strength of personal obligations and the convoluted development of the Roman law of agency. When enforcement of all types (by first, second, or third parties) is stricter, it becomes more necessary to ensure that proper consent has been given when entering into a commitment, which may be seen as an application of the principle in principalagent theory that establishes some degree of complementarity between the intensity of both incentives and monitoring. From this perspective, the strict enforcement of personal obligations, characteristic of Roman law, thus ties in with, first, its initial reluctance to commit an individual by the actions of others; and, second, its ulterior development of a palliative of contractual agency. This palliative was largely based on the role as agents of sons and slaves, who provided an additional safeguard by being under the patria potestas of their father ⁵³ The birth registry was voluntary, but (as often happens with public registries) people were motivated to record to enjoy legal effects (Rowlandson 2013). The evidence can be interpreted as confirming the argument: recordation in the kalendarium was declarative within thirty days of birth but in principle illegitimate children (with fewer or no rights) could not be included in the album, which suggests some form of purge, especially considering that the album was publicly displayed, with both archives working respectively as a mere recording diary and as a proper register. It is consistent with this interpretation that the copies used as proof of status were produced from the album and were authenticated by up to seven witnesses.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

289

and master, a situation that was carefully identified and often made public by institutional solutions.⁵⁴ For a similar reason, protecting the strict enforcement of personal obligations might also have been behind the underdevelopment of the law of companies.

19.8. CONCLUSI ON For decades, ancient history has been divided between divergent “modernist” or “formalist” views, which commonly emphasize economic growth and market exchange, and “primitivist” or “substantivist” views, which tend to highlight constraints, power, and domination.⁵⁵ My analysis of Roman institutions from the perspective of personal vs impersonal exchange aims to transcend two aspects of this recurrent discussion, relating to the nature and extent of the market economy in Roman times, and the market-facilitating role of the Roman state. With respect to the nature and extent of the market, by focusing on rights in rem, the paradigm of truly impersonal exchange, the chapter sets a benchmark for the impersonality of market transactions, one of the dimensions of transaction costs and, therefore, the “quality” of market exchange. Using this metric, one may be tempted to frame the discussion in ancient history in terms of personal versus impersonal exchange, considering, from a primitivist standpoint, that only impersonal markets deserve to be considered as markets. This, however, would be mistaken on both empirical and theoretical grounds. Empirically, because even today most markets and transactions remain personal or at least retain personal dimensions, even if other dimensions or attributes have been depersonalized.⁵⁶ More ⁵⁴ The fact that the ability to commit the principal was limited to dependents under his potestas also refutes the argument by Verboven (2002: 260–3), according to which the strong informal ties between Roman friends made agency unnecessary. In my view, friends were less effective than sons and slaves as agents because, whatever the strength of their ties with the principal and in contrast to sons and children, their status as friends was totally informal and therefore more easily subject to misunderstanding and information asymmetry on the part of and with respect to third parties. ⁵⁵ See e.g. for a summary account Scheidel (2012: 7–10). ⁵⁶ In reality as opposed to abstract models, impersonality is a more or less continuous attribute of transactions (Arruñada 2012: 15–18) and even specific dimensions of transactions may be situated at different levels of that continuum.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

290

Benito Arruñada

fundamentally, efficiency judgments must consider the fixed and variable costs of introducing and operating alternative institutions; and, even within the narrow framework of §19.2, impersonal exchange is not intrinsically superior to personal exchange. Therefore, the lack of land and company registries, and the consequent reliance on palliative solutions ensuring strict enforcement of personal obligations should not necessarily be considered as a negative indicator of the extent of Roman markets or of the support given to the market by the Roman state. With respect to this support, it has been hotly disputed if the Roman state favored the development of the market or merely pursued extractive objectives that were conducive, as an unintentional byproduct, to the development of market exchange. In particular, it has been argued that Roman markets were not supported by the state and only worked in a limited manner within the informal networks of merchants and households (e.g. Bang 2008). In essence, I proposed in §19.2 a simple model of the state’s institutional decisions and then confirmed in the following sections that the empirical evidence on Roman institutions is broadly consistent with the model’s predictions.⁵⁷ I did not discuss how the Roman state effectively provided conventional public goods which were essential for market exchange, such as security of property, a stable currency, standard weights and measures, and an interconnected judicial system, but did analyze key areas of the law in which the Roman state helped expand the market by developing institutions that enabled a wider scope of transactions. They were, if not impersonal, at least made possible by the effective enforcement of personal obligations. These market-enabling institutions included more or less centralized solutions. Relatively decentralized solutions were based on superseding allegedly inefficient legal rules by means of judicial decisions, in a process which was to some extent similar to how modern common law is often assumed to work. The most salient case is the evolution of praetorian remedies, which filled gaps in traditional law. More centralized solutions, closer to the modern civil law tradition, were rarer and are here represented by the bibliotheke enkteseon, the new land and ⁵⁷ The model assumes a benevolent decision maker, which is unreal but, under proper constraints, individual maximization should be compatible with social welfare, and judging the effectiveness of such constraints is not easier than judging overall consequences.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

291

mortgage register created by the provincial administration of Egypt. This creation suggests that Roman administration was not inimical to this type of institution when it was well suited but, more widely, was interested in facilitating market transactions and limiting the marketdebasing effect of the in rem enforcement of tax duties through tacit charges.⁵⁸ Lastly, the state contributed by supporting personal exchange within and between familiae. In this area, the state helped the market not only by enforcing strict rules on personal obligations and not interfering with private contracting within familiae (forbearance), but also by providing organized support for personal transactions. Roman personal exchange was not mere private ordering, as the state enacted publicity rules (e.g. on emancipation and manumission) and implemented public personal registries (e.g. the birth registry and the census) that reduced transaction costs and enhanced the enforcement of personal obligations. From this perspective and to the extent that informal social norms enforcing personal obligations were supported by such formal rules and registries, the separation between formal and informal institutions, which seems to play a major role in current discussions in ancient history (Lerouxel 2012a: 642), becomes fuzzier, while the market-enabling role of the Roman state becomes greater and more nuanced. The picture that emerges when considering the whole set of solutions used in the classical period of Roman law is one in which, broadly speaking and relying on the framework developed in §19.1, transactions are based on, first, automatic verifiability of originative contracts (e.g. publicity of possession for property and public appearance for business transactions); and, second, substantial public support for the effective enforcement of personal obligations. In particular, institutional support provided practically no organized verifiability (i.e. registries of originative contracts) but did provide rules (e.g. penalties, family law), general organizational principles (judicial forbearance), and even organizations (census) that facilitated the enforcement of personal obligations.⁵⁹ ⁵⁸ See Lerouxel (2012a: 656–7). Of course, this emphasis on enabling the market was not permanent. Especially in later centuries, the state simply gave privileged status to the treasury (Nicholas 1962: 153; Crook 1967: 246) without paying attention to the increase in transaction costs caused by these tacit charges in the absence of registries. ⁵⁹ This work has benefited from advice and comments received from Maria Brosius, Yun-chien Chang, Giuseppe Dari-Matiacci, Damián Fernández, Stephen

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

292

Benito Arruñada R E F E RE NC E S

Abatino, Barbara, Giuseppe Dari-Mattiacci, and Enrico C. Perotti. 2011. “Depersonalization of business in ancient Rome.” 31 Oxford Journal of Legal Studies 365–89. Adams, Colin E. P. 2013. “Abinnaeus archive,” in Bagnall et al., eds, 5–6. Ando, Clifford. 2010. “The administration of the provinces,” in David S. Potter, ed., A Companion to the Roman Empire. Oxford: WileyBlackwell, 177–92. Andreau, Jean. 1999. Bankers and Business in the Roman World. Cambridge: Cambridge University Press. Arruñada, Benito, and Nuno Garoupa. 2005. “The choice of titling system in land.” 48 Journal of Law and Economics 709–27. Arruñada, Benito, Giorgio Zanarone, and Nuno Garoupa. 2019. “Property rights in sequential exchange.” 35 Journal of Law, Economics, and Organization 127–53. Arruñada, Benito, Luis Garicano, and Luis Vázquez. 2001. “Contractual allocation of decision rights and incentives: the case of automobile distribution.” 17 Journal of Law, Economics, and Organization 256–83. Arruñada, Benito. 2003. “Property enforcement as organized consent.” 19 Journal of Law, Economics, and Organization 401–4. Arruñada, Benito. 2010. “Institutional support of the firm: a theory of business registries.” 2 The Journal of Legal Analysis 525–76. Arruñada, Benito. 2012. Institutional Foundations of Impersonal Exchange: Theory and Policy of Contractual Registries. Chicago and London: University of Chicago Press. Arruñada, Benito. 2015. “The titling role of possession,” in Yun-chien Chang, ed., The Law and Economics of Possession. Cambridge: Cambridge University Press, 207–33. Arruñada, Benito. 2016. “How Rome enabled impersonal markets.” 61 Explorations in Economic History 68–84.

Hansen, François Lerouxel, Joe Manning, Tom McGinn, Fernando P. Méndez González, Pedro Pernas, Paul du Plessis, Dominic Rathbone, José Remesal Rodríguez, Jane Rowlandson, Carlos Sánchez-Moreno Ellart, Peter Temin, Rick Verhagen, an anonymous reviewer, and participants at the Economic History Seminar of Carlos III University, the Bozen workshop on “The Institutions of Property Rights,” the Italian Society of Law and Economics, the Spanish Association of Law and Economics, and the 18th annual conference of the International Society for New Institutional Economics (Duke University). Usual disclaimers apply. It received support from the Spanish Ministry of the Economy and Competitiveness through grants ECO201457131-R and ECO2017-85763-R, and the Severo Ochoa Program for Centers of Excellence in R&D (SEV-2015-0563). This work is a considerably expanded and revised version of Arruñada (2016).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

293

Arruñada, Benito. 2017. “Property as sequential exchange: the forgotten limits of private contract.” 13 Journal of Institutional Economics 753–83. Arruñada, Benito. 2018. “Evolving practice in land demarcation.” 77 Land Use Policy 661–75. Aubert, Jean-Jacques. 2013. “Institor,” in Bagnall et al., eds, 3466–7. Bagnall, Roger S. 2010. Everyday Writing in the Graeco-Roman East. Sather Classical Lectures. Vol. 69. Berkeley: University of California Press. Bagnall, Roger S., Kai Brodersen, Craige B. Champion, Andrew Erskine, and Sabine R. Huebner, eds. 2013. The Encyclopedia of Ancient History, 1st edn. Malden, MA/Oxford: Wiley-Blackwell. Bang, Peter F. 2008. The Roman Bazaar: A Comparative Study of Trade and Markets in a Tributary Empire. Cambridge: Cambridge University Press. Bransbourg, Gilles. 2012. “Rome and the economic integration of Empire.” Institute for the Study of the Ancient World Papers 3. http://dlib.nyu.edu/ awdl/isaw/isaw-papers/3/ (visited July 30, 2015). Bransbourg, Gilles. 2014. “Les taux d’intérêt flottants des Sulpicii,” in Catherine Apicella, Marie-Laurence Haack, and François Lerouxel, eds, Les affaires de Monsieur Andreau: économie et société du monde romain. Bordeaux: Ausonius, 123–35. Brennan, T. Corey. 2001. The Praetorship in the Roman Republic, vol. 1: Origins to 122 BC. Oxford: Oxford University Press. Brown, Peter. 1971 [repr. 2006]. The World of Late Antiquity AD 150–750. London: Thames & Hudson. Buckland, William Warwick. 1912. Elementary Principles of the Roman Private Law. Cambridge: Cambridge University Press. Calabresi, Guido, and A. Douglas Melamed. 1972. “Property rules, liability rules, and inalienability: one view of the cathedral.” 85 Harvard Law Review 1089–1128. Camodeca, Giuseppe. 2013. “Tabulae Herculanenses,” in Bagnall et al., eds, 6507–9. Corcoran, Simon. 2013. “Tablettes Albertini,” in Bagnall et al., eds, 6499–500. Crook, John A. 1967. Law and Life of Rome. Ithaca: Cornell University Press. Dari-Mattiacci, Giuseppe. 2013. “Slavery and information.” 73 The Journal of Economic History 79–116. Daumas, François. 1987. La civilisation de l’Égypte pharaonique. Paris: Les Editions Arthaud (Spanish translation used: Barcelona, Óptima, 2000). Demsetz, Harold. 1967. “Toward a theory of property rights.” 57 American Economic Review 347–59. Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer. 2008. “The law and economics of self-dealing.” 88 Journal of Financial Economics 430–65. Duncan-Jones, Richard P. 1976. “Some configurations of landholding in the Roman Empire,” in Moses I. Finley, ed., Studies in Roman Property by the

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

294

Benito Arruñada

Cambridge University Research Seminar in Ancient History. Cambridge: Cambridge University Press, 7–33. Eich, Peter. 2013. “Administration, Roman,” in Bagnall et al., eds, 90–5. Erdkamp, Paul. 2005. The Grain Market in the Roman Empire: A Social, Political and Economic Study. Cambridge: Cambridge University Press. Everett, Nicholas C. 2000. “Scribes and charters in Lombard Italy.” 41 Studi Medievali 39–84. Everett, Nicholas C. 2013. “Ravenna Papyri,” in Bagnall et al., eds, 5739–41. Fernández, Damián. 2013. “Alimenta schemes,” in Bagnall et al., eds, 315–16. Finley, Moses I. 1999. The Ancient Economy. Berkeley: University of California Press. Frier, Bruce W., and Dennis P. Kehoe. 2007. “Law and economic institutions,” in Walter Scheidel, Ian Morris, and Richard P. Saller, eds, The Cambridge Economic History of the Greco-Roman World. Cambridge: Cambridge University Press, 113–43. Gibbs, Matt. 2013. “Tax farming,” in Bagnall et al., eds, 6553–5. Greene, Kevin. 2000. “Technical innovation and economic progress in the ancient world: M. I. Finley re-considered.” 53 Economic History Review 29–59. Haeussler, Ralph. 2013. “Literacy, Rome and provinces,” in Bagnall et al., eds, 4104–8. Hansmann, Henry, and Reinier Kraakman. 2000. “The essential role of organizational law.” 110 Yale Law Journal 387–440. Hansmann, Henry, and Reinier Kraakman. 2002. “Property, contract, and verification: the numerus clausus problem and the divisibility of rights.” 31 Journal of Legal Studies S373–S420. Hansmann, Henry, Reinier Kraakman, and Richard Squire. 2006. “Law and the rise of the firm.” 119 Harvard Law Review 1333–1403. Harris, Ron. 2000. Industrializing English Law: Entrepreneurship and Business Organization, 1720–1844. Cambridge: Cambridge University Press. Harris, William V. 1989. Ancient Literacy. Cambridge: Harvard University Press. Harter-Uibopuu, Kaja. 2013. “Gesta,” in Bagnall et al., eds, 2908–9. Heller, Michael A. 1999. “The boundaries of private property.” 108 Yale Law Journal 1163–1223. Hersch, Karen K. 2010. The Roman Wedding: Ritual and Meaning in Antiquity. Cambridge: Cambridge University Press. Honoré, Anthony Maurice. 1989. “Conveyances of land and professional standards in the later Empire,” in Peter Goodwin Birks, ed., New Perspectives in the Roman Law of Property: Essays for Barry Nicholas. Oxford: Clarendon Press, 137–52. Innis, Harold A. 1950 [repr. 2007]. Empire and Communications. Toronto: Dundurn Press.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

295

Johnson, Simon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer. 2000. “Tunneling.” 90 American Economic Review Papers and Proceedings 22–7. Johnston, David. 1999. Roman Law in Context. Cambridge: Cambridge University Press. Jördens, Andrea. 2010. “Nochmals zur Bibliotheke Enkteseon,” in Gerhard Thür, ed., Symposion 2009: Vorträge zur griechischen und hellenistischen Rechtsgeschichte (Seggau, 25–30 August 2009). Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 277–90. Kain, Roger J. P., and Elizabeth Baigent. 1992. The Cadastral Map in the Service of the State: A History of Property Mapping. Chicago: University of Chicago Press. Kessler, David, and Peter Temin. 2007. “The organization of the grain trade in the early Roman Empire.” 60 Economic History Review 313–32. Kirschenbaum, Aaron. 1987. Sons, Slaves and Freedmen in Roman Commerce. Jerusalem: The Hebrew University Magnes Press. Koyama, Mark. 2010. “Evading the ‘taint of usury’: the usury prohibition as a barrier to entry.” 47 Explorations in Economic History 420–42. Laiou, Angeliki E., and Cécile Morrisson. 2007. The Byzantine Economy. Cambridge: Cambridge University Press. Lerouxel, François. 2012a. “The private credit market, the Bibliotheke Enkteseon, and public services in Roman Egypt.” 67 Annales. Histoire, Sciences Sociales 629–59. Lerouxel, François. 2012b. “Des prêts sans intérêt? Le taux d’intérêt dans le nome oxyrhynchite avant 79.” 181 Zeitschrift für Papyrologie und Epigraphik 161–72. Lerouxel, François 2016. Le marché du crédit dans le monde romain (Égypte et Campanie), Bibliothèque des Écoles françaises d’Athènes et de Rome 374. Rome: École française de Rome. Lo Cascio, Elio. 2000. “Gli alimenta, l’agricoltura italica e l’approvvigionamento di Roma,” in Elio Lo Cascio, ed., Il princeps e il suo impero: studi di storia amministrativa e finanziaria romana. Bari: Edipuglia, 223–64. Malmendier, Ulrike. 2009. “Law and finance ‘at the origin.’ ” 47 Journal of Economic Literature 1076–1108. Manning, Joseph G. 2003. Land and Power in Ptolemaic Egypt: The Structure of Land Tenure. Cambridge: Cambridge University Press. McGinn, Thomas A. J. 2013. “Manus,” in Bagnall et al., eds, 4270–1. Merrill, Thomas W., and Henry E. Smith. 2000. “Optimal standardization in the law of property: the numerus clausus principle.” 110 Yale Law Journal 1–70. Merrill, Thomas W., and Henry E. Smith. 2001. “What happened to property in law and economics?” 111 Yale Law Journal 357–98.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

296

Benito Arruñada

Meyer, Elizabeth A. 2004. Legitimacy and Law in the Roman World: Tabulae in Roman Belief and Practice. Cambridge: Cambridge University Press. Moatti, Claude. 1994. “Les archives des terres publiques à Rome (1er siècle av.–1er siècle ap. J.-C.): le cas des assignations,” in Ségolène Demougin, ed., La mémoire perdue: à la recherche des archives oubliées, publiques et privées, de la Rome antique. Paris: Publications de la Sorbonne, 103–19. Monson, Andrew. 2012. From the Ptolemies to the Romans: Political and Economic Change in Egypt. Cambridge: Cambridge University Press. Moreau, Philippe. 1994. “La mémoire fragile: falsification et destruction des documents publics au 1er siècle av. J.C.,” in Ségolène Demougin, ed., La mémoire perdue: à la recherche des archives oubliées, publiques et privées, de la Rome antique. Paris: Publications de la Sorbonne, 121–47. Muhs, Brian. 2015. “Transaction costs and institutional change in Egypt in the early first millennium B.C.E.,” in Dennis P. Kehoe, David M. Ratzan, and Uri Yiftach, eds, Transaction Costs in the Ancient Economy. Ann Arbor: Michigan University Press, 80–98. Nicholas, Barry. [1962] 2010. An Introduction to Roman Law. Oxford: Oxford University Press. Nicolet, Claude. 1988. L’inventaire du monde: géographie et politique aux origines de l’Empire romain. Paris: Hachette. North, Douglass C. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press. Parker, Anthony John. 1992. Ancient Shipwrecks of the Mediterranean and the Roman Provinces. British Archaeological Reports International Series 580. Oxford: Tempus Reparatum. Petrucci, Aldo. 2003. “Orígenes romanísticos de la buena fe objetiva en la contratación mercantil.” 15 Anuario Mexicano de Historia del Derecho 601–15. Ramin, Jacques, and Paul Veyne. 1981. “Droit romain et société: les hommes libres qui passent pour esclaves et l’esclavage volontaire.” 30 Historia 472–97. Rathbone, Dominic. 1991. Economic Rationalism and Rural Society in ThirdCentury A. D. Egypt: The Heroninos Archive and the Appianus Estate. Cambridge: Cambridge University Press. Rhodes, P. J. 2013. “Documents and archives,” in Bagnall et al., eds, 2181‒4. Riggsby, Andrew M. 2010. Roman Law and the Legal World of the Romans. Cambridge: Cambridge University Press. Rose, Carol M. 1985. “Possession as the origin of property.” 52 University of Chicago Law Review 73–88. Rotman, Youval. 2013. “Babatha,” in Bagnall et al., eds, 1004–5. Rowlandson, Jane. 1981. “Sales of land in their social context,” in Roger S. Bagnall, Gerald Browne, Ann E. Hanson, and Ludwig Koenen, eds,

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Institutions of Roman Markets

297

Proceedings of the Sixteenth International Congress of Papyrology. New York: Chico Scholars Press, 371–8. Rowlandson, Jane. 2013. “Birth, registrations of,” in Bagnall et al., eds, 1128–31. Sánchez-Moreno Ellart, Carlos. 2013. “Archives,” in Bagnall et al., eds, 660–5. Scafuro, Adele C. 2013. “Witnesses, Greek and Roman,” in Bagnall et al., eds, 7114–16. Scheidel, Walter, and Steven J. Friesen. 2009. “The size of the economy and the distribution of income in the Roman Empire.” 99 Journal of Roman Studies 61–91. Scheidel, Walter. 2008. “The comparative economics of slavery in the GrecoRoman world,” in Enrico Dal Lago and Constantina Katsari, eds, Slave Systems, Ancient and Modern. Cambridge: Cambridge University Press, 105–26. Scheidel, Walter. 2012. “Approaching the Roman economy,” in Walter Scheidel, ed., The Cambridge Companion to the Roman Economy. Cambridge: Cambridge University Press, 1–21. Scheidel, Walter. 2013. “The shape of the Roman world.” Princeton/Stanford Working Papers in Classics. Version 1.0. April. Schulz, Fritz. 1951. Classical Roman Law. Oxford: Clarendon Press. Smith, Henry E. 2006. “Modularity in contracts: boilerplate and information flow.” 104 University of Michigan Law Review 1175–1222. Smith, Henry E. 2009. “Modularity in property, intellectual property, and organizations.” Harvard Law School Working Paper, September 17. Temin, Peter. 2006. “The economy of the early Roman Empire.” 20 Journal of Economic Perspectives 133–51. Temin, Peter. 2013. The Roman Market Economy. Princeton: Princeton University Press. Thür, Gerhard. 2013. “Proof, legal, Greek and Roman,” in Bagnall et al., eds, 5572–4. Verboven, Koenraad. 2002. The Economy of Friends: Economic Aspects of Amicitia and Patronage in the Late Republic. Brussels: Latomus. Verboven, Koenraad. 2003. “The Sulpicii from Puteoli and usury in the early Roman Empire.” 71 Tijdschrift voor Rechtsgeschiedenis 7–28. Verboven, Koenraad. 2013. “Clientela, Roman Republic,” in Bagnall et al., eds, 1577–82. Williamson, Oliver E. 1991. “Comparative economic organization: the analysis of discrete structural alternatives.” 36 Administrative Sciences Quarterly 269–96. Wilson, Andrew I. 2011. “City sizes and urbanization in the Roman Empire,” in Alan K. Bowman and Andrew I. Wilson, eds, Settlement, Urbanization, and Population. Oxford: Oxford University Press, 161–95.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

298

Benito Arruñada

Winkel, Laurens. 2013. “Ius gentium,” in Bagnall et al., eds, 3553–4. Yiftach-Firanko, Uri. 2010. “Comments on Andrea Jördens,” in Gerhard Thür, ed., Symposion 2009. Vorträge zur griechischen und hellenistischen Rechtsgeschichte (Seggau, 25–30 August 2009). Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 291–9. Yiftach-Firanko, Uri. 2013. “Law, Greco-Roman Egypt,” in Bagnall et al., eds, 3941–45. Ancient Sources Digest. The Digest of Justinian, rev. Eng. edn Alan Watson. Philadelphia: University of Pennsylvania Press. Gaius. The Institutes of Gaius, trans. Edward Poste, rev. edn E. A. Whittuck. Oxford: Clarendon Press, 1904. Livius. The History of Rome, Books 27 to 36, trans. Cyrus R. Edmonds. http:// www.gutenberg.org/ebooks/12582 (visited December 1, 2013). Tacitus. The Reign of Tiberius, Out of the First Six Annals of Tacitus; With his Account of Germany, and the Life of Agricola, trans. Thomas Gordon, ed. Arthur Galton. http://www.gutenberg.org/ebooks/7959 (visited December 1, 2013).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

20 One Step at a Time in Roman Law How Roman Pleading Rules Shape the Substantive Structure of Private Law Richard A. Epstein

One broad measure of the success of any legal system is its durability. All legal systems must survive a world that is dominated by scarcity and individual self-interest. Set the wrong set of rules and the selfinterest of the regulated parties will destroy the system. Set the right rules and societies will, within the limits of their resource endowments, thrive relative to their competitors. By this test, the extensive body of private Roman law has done very well indeed. Its basic precepts and organizational structure not only served as the foundation for a great empire, but have migrated, largely intact, across a wide range of different cultures, both in time and in space. For the Romans, pleading rules played an integral part in their procedural system (Metzger 2013: 13–28). The early works of Justinian and Gaius contain an accurate exposition of Roman pleading rules. Indeed, it has been observed more than once that many of the pleading rules of English law are direct descendants of the Roman pleading principles, down to the details of nomenclature, for instance where the term replication is carried over from the Roman law into the English to refer to the next responsive plea after the defendant has

Richard A. Epstein, One Step at a Time in Roman Law: How Roman Pleading Rules Shape the Substantive Structure of Private Law In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0020

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

302

Richard A. Epstein

put forward a valid, but defeasible, defense.¹ Yet at the same time that these rules have sparked admiration, little has been said about how the mature Roman pleading rules established a procedural template that lends greater precision to the private law doctrines of property, contracts, and tort that work as well today as in Roman times. The purpose of this paper is not to explain the historical development of Roman procedural law, but to show why the Roman practices work as well today as they did historically and why they tend to lead to the creation of efficient legal rules. In order to achieve this goal, I shall proceed as follows. §20.1 explains the operation of the Roman pleading system in order to show how it facilitated the orderly development of substantive law. §20.2 then offers concrete examples of how the basic distinctions in Roman law facilitate its success in ways that carry over to modern legal analysis applicable in a quite different historical period. For these purposes, two bodies of law play a critical part. The first is the distinction between actions in rem and actions in personam. The second, whose use is set up by the first, is the Roman system of staged pleadings, which tended toward efficient legal solutions by using a set of successive approximations, each of which brought the legal system closer to its ideal formulation. In §20.2, I shall then give a number of concrete examples drawn from historical materials which help explain the efficient results that the Romans achieved in a number of difficult substantive areas of the law.

20.1. THE ROMAN PROCEDURAL SYSTEM The great strength of the Roman procedural system was its clear classifications, aided by the deliberate way in which the mature Roman pleading system converted broad cases into precise issues ready for an authoritative decision. On its face, this technical body of rules appears to derive little from economic theory, given that its ¹ For the early invocation of replication in the Roman law, see The Institutes of Gaius, Book 4, pt I, §§ tl 5–29 (ed. Francis de Zulueta, 1946) (noting the use of the replication as a way to admit new matter that overcomes what would otherwise be a valid exception to the prima case). The replication made its way into English law as well (Sutton 1929: 78, 190, 198).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 303 basic framework stresses the organization of procedural elements of a wide array of real and personal actions, where the substantive legal framework emerges through the gradual application of the procedural system. But a closer look at the rules does reveal something of the implicit economic principles operating at every level. The emergence of these categories was not just a fluke but rested on the happy historical circumstance that the Romans institutionally separated the formulation of their basic legal rules from the application of those rules to particular disputes. The basic legal rules and principles, applicable across the board, were carried over from each praetor, or some other magistrate, in a continuous arc of legal development (Gai., Inst. 4.111; Lenel 1927). A separate fact finder then applied those principles to the given case (Metzger 2013: 8). This strong degree of separation meant that the standard legal formulas could be articulated with a rare precision uncluttered by the factual disputes that encouraged the parties to make strategic decisions in presenting their cases. In the course of litigation, triers of fact are constrained by a host of practical and formal restraints. They must dispose of litigation within a reasonable time. They also must make sure that each party receives due process, so that he has a right to be heard—audi alteram partem—before a neutral judge— nemo judex in causa sua—and fair notice of the charges against him (Danner 1912: 27, 33). Those requirements all speak to the simple point that each party should have his day in court to present his case. The institutional framework that the Romans used to formulate their pleadings severed doctrinal developments from the short-term necessities of practical litigation. The central importance of this separation lay in the way in which the Romans classified their causes of action and pleaded their cases to joinder of issue. The want of any litigant pressure meant that the Romans could develop rational classifications of causes of action. The division between rights in rem and rights in personam set the stage for the simple but critical line of actions ex contractu and ex delicto. This two-stage design may appear intuitive, but even with the Roman precedents, the English ad hoc system of substantive rights insecurely organized within the forms of actions created an unfortunate dissonance between the form of action and the underlying cause of action, which was evident in dealing with both actions for debt (which covered consensual and nonconsensual claims) and actions on the case (which covered both tort and contract actions; Maitland 1936).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

304

Richard A. Epstein

Once their basic classifications were established, the Romans were able to develop their substantive rules by adding in additional elements of the case through exceptions, replications, duplications, and so on (Nicholas 1962: 24–5). The modern framework always provides on grounds of natural law (or the American due process of law) both litigants the right to present their cases before the court (Schauer 1976: 47). This twopart division may be justified on grounds of judicial economy insofar as it lets each side speak exactly once and then uses some pretrial meeting to set the stage for the trial.² But conceptually, this procedure sets out the wrong template for pleading any case to joinder of issue. The great advance in Roman pleading law was that it did not confine the development of the underlying legal doctrine to two stages of a claim and then a defense, but rather it allowed the parties’ back and forth to continue so long as either party wanted to add some new matter to the case that incorporated all allegations from the proceeding stages of the complaint.³ At each stage, the party then opposing the pleadings (in the second stage, this would be the defendant, to whom the plaintiff would respond in stage three, and to whom the defendant would subsequently respond in stage four, and so on) has three options. The first of these is to deny the allegations. The second is to admit that they are true, and then to insist that they do not state a sufficient prima facie case or subsequent plea (in effect, to demur). The third is to admit that the new material is a sufficient plea and then to insist that some additional factor, by way of replication or duplication, is consistent with a sense of justice and therefore should be considered an exception to the previous plea. Gaius does not use the English common-law classifications in which these three options are termed: to deny, to demur (or object), or to add new material through a plea of confession and avoidance (Sutton 1929; Epstein

² Federal Rule of Civil Procedure Rule 16(a) provides: “In any action, the court may order the attorneys and any unrepresented parties to appear for one or more pretrial conferences for such purposes as: (1) expediting disposition of the action; (2) establishing early and continuing control so that the case will not be protracted because of lack of management; (3) discouraging wasteful pretrial activities; (4) improving the quality of the trial through more thorough preparation; and (5) facilitating settlement.” ³ These rules were set out with great clarity in Book 4 of Gaius’ Institutes (§§102–29); Just., Inst. 4.6, 12. The development of this theme has been a career obsession (Epstein 1997, 2005).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 305 1973a). Yet his analysis tracks these terms precisely, so I shall use them here for purposes of clarity. Importantly, the demurrer introduced new material into the case, and its legal significance (at that stage) was determined by whether it was then accepted or rejected. It is through this dialogue of pleadings that the Roman law narrowed and refined the precise legal issues in dispute. These procedural niceties are vital for institutional reasons, but I shall for the moment put them aside to concentrate on the substantive defenses, some good and some bad, that can be raised in these Roman cases, such as assumption of risk, contributory negligence, infancy, insanity, plaintiff ’s entry, and the like. Once the defendant presents these substantive defenses, the correct procedure is for the plaintiff at stage three of the case (the reply or replication) to have the same set of choices that is made available to the defendant at stage two. The denial again raises no important principle of fact. The demurrer to the exception asks the same question about its sufficiency that a demurrer to the complaint asks about the basic cause of action: is the matter pleaded sufficient to switch the balance back in favor of the defendant? Yet it is at this point that the third alternative, the introduction of new material by a plea of confession and avoidance, shows why the common law’s two-stage procedures, which may be administratively convenient, in fact distort the overall development of substantive law. As Gaius notes: “Sometimes an exception, which in the absence of counter allegations seems prima facie to be just to the defendant, is unjust to the plaintiff, and then, to protect the plaintiff, the praetor adds to the instructions a clause called Replication, because it is an undoing and counteraction of the force of the exception” (Gai., Inst. 4.26). Once three stages are admitted into the argument, there is no way in which it is possible to impose any artificial limits on the process whereby new information is added. Allowing the replicatio thus paves the way for allowing the duplicatio or rejoinder, after which the cycle can continue indefinitely until a particular party finds that it has nothing more that it would like to say (Gai., Inst. 4.127–9). A few observations are in order about the formation of this system. The first of these is that the system proceeds by private initiative followed by government approval. The state does not tell individuals what issues they ought to raise or when they ought to raise them. That part is left in private hands, subject of course to the oversight of the court as to which of these pleas is allowed, and when. The choice of

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

306

Richard A. Epstein

this method has the sometimes unappreciated virtue of removing any external constraints on the operation of the system as a whole. The kinds of questions that can be raised are not those which arise solely from the abstract reflections of bureaucrats, but from concrete disputes as they percolate from the ground up, assuring that the doctrinal issues that matter will receive careful attention while those that do not will be passed by in relative silence. The second point is that, in dealing with these issues, Gaius does not make the mistake of H. L. A. Hart in his famous lecture on “Ascription of Rights and Responsibilities,” which sought to turn defeasibility into a question of the meaning of key words like “contract” (Hart 1949: 171–94).⁴ But in fact that is not the case. A perfectly adequate definition of contract refers to an agreement between two or more people calling for the performance or forbearance of various actions by either or both parties. That definition makes it clear what kinds of activities are governed but leaves open whether a particular agreement is the kind that should be enforced, or whether, even if it falls within the class of enforceable agreements and the defendant has not performed, the promise still should not be enforced because of some act of the plaintiff or extrinsic circumstances. None of these matters are questions of the definition of contract. All are questions of substantive judgment. The modern law and economics scholar might raise his or her eyebrows about Gaius’ constant assertions that certain forms of conduct are either “just” or “unjust,” which reflects the modern inclination to distrust any pronouncement about natural law. But I think that this charge is overstated. A close look at the individual cases in which Gaius pronounces judgment shows a virtual unerring sense for the right result that would please a thoughtful law and economics scholar who subscribes to the overtly consequentialist arguments that dominate modern economic thought. Thus, to give Gaius’ example in Inst. 4.26, in which the prima facie case is that the defendant has not paid a debt to the plaintiff, the defendant can introduce as an exception that the plaintiff had agreed not to sue on the debt. But then the debt could be reinstated if the plaintiff, by way of the replication (in the next stage), could demonstrate that he had renewed the debt by a fresh promise: what Gaius ⁴ For criticism of Hart’s position, see Geach (1960: 221–5); Pitcher (1960: 226–35). Hart eventually abandoned that position (Hart 1983). However, none of these three authors hit on the Roman formulation.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 307 called a “contrary agreement.” Clearly, a duplication is in order if that fresh promise was in turn induced by fraud, and so on. Indeed, as I have argued elsewhere, the phrase “naturalis ratio”—literally the natural reason—rests on implicit use of a modern economic argument that could not be articulated at the time because no one then thought in terms of the economic principles on which that approach rested (Epstein 1989: 713–51; 2003). The task of law and economics is to explain the durability of these rules, and not to dismiss them as some relic of outmoded thought, which decidedly they are not.

20.2. THE USES OF THE PLEADING SYSTEM It is one thing to note the organization of any legal system. It is quite another to show its economic utility. In this regard, the bottom-up nature of the pleading system is congenial with the economic approach that wants to gain a holistic look at the overall set of rules. Modern law and economics scholars tend to develop formal models that identify the optimal distribution of rights and duties. In so doing they tend to rely on overbroad generalizations suggesting that, for example, a negligence rule with a contributory negligence defense will provide for the optimal allocation of resources (Brown 1973: 323–49). The claim is often made in both a positive and normative guise, and is wrong in both those contexts. In this regard, the constant Roman tendency is to subdivide causes of action into discrete units, each with its own outcomes. That approach often leads descriptively and normatively to the use of a wide range of potential liability rules in different contexts. In stranger cases, the basic position sets the prima facie case for all law suits as strict liability, so as to flesh out that system. But in harms that arise out of consensual situations, no uniform theory of liability can deal with all of the actions and omissions that cause harm, so that a wider range of solutions may prove necessary. The advantage of the pleading system, rightly construed, is that it resists the kinds of overgeneralizations to which modern law and economics often falls prey. To see how this system functions, it is important to show in some detail how the pleading rules work, when rigorously and consistently applied. There is a tendency in Roman sources to use these rules precisely in dealing with such matters as pleas in abatement and res

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

308

Richard A. Epstein

judicata. But by the same token, as noted long ago by Danner (1912), there was less willingness to plead separately all of the substantive elements of a given claim. Instead it was common to allow matters that technically might be part of an exceptio to be introduced at trial under a general denial of the plaintiff ’s case (Danner 1912: 36–7). But the method loses much of its power by this relaxation of its formal strictures. In order to ensure that the doctrine is properly aligned, it is necessary to make sure that there is a unity across the system, which can only be obtained by making a couple of critical observations about the notion of “sufficiency” of a claim. First, in a negative sense, the term sufficient in law never means what it does in formal logic—if A, then B follows. Rather what it means for legal pleadings is that if A’s claim is sufficient, then B cannot file a demurrer but must either deny or plead over by confession and avoidance. Second, in order for this process to remain harmonious, the prima facie case should only require the minimum set of allegations that is necessary to shift the burden of explanation, whether by excuse or justification, to the other side, leaving all other potentially favorable facts for use, if needed, at some later stage in their argument, should some valid plea by the defendant then give rise to new opportunities for further elaboration. The point here is intended to impose a major level of self-restraint on the plaintiff who might wish to plead more. The choice of action starts with the identification of the plaintiff ’s interest that has been harmed, which is of course established, as will become clear, in accordance with the rights embedded in the in rem/in personam distinction. The next part of the prima facie case identifies the act or omission that is said presumptively to interfere with either of the two issues just identified. Here we can examine some examples from common law torts and contracts. By way of introduction, “tort,” in contrast to the Roman “delict” (Buckland, McNair, and Lawson 1965: 352–66), is drawn from English law. Delict has long been understood to be “the act of a person, by fault or maliciously, which harms or otherwise causes damage to another” (Pothier 1761 [1806]). It should be apparent that this definition excludes strict-liability offenses in their entirety, which is consistent with the Roman view that did not distinguish sharply between public enforcement of criminal law and private enforcement of personal wrongs. The law of furtum clearly has this intentional element, but the Lex Aquilia reads differently. It starts out

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 309 in strict-liability fashion—if anyone kills—so that the addition of negligence and malice is through gloss only. I regard that approach as a serious intellectual error in both Roman and modern times, and for the same reason. This gloss ignores the role that strict liability (with defenses) has and should have in any system of private law. In using the term “tort” to refer to private wrongs under both the common-law and the Roman-law system, I am talking only about civil enforcement and intend to stress the close parallels between the two legal systems within this limited domain. In actions for bodily harm of property damage, plaintiffs will often choose to include in their pleadings that the defendant’s act intended to cause harm, or that the defendant failed to take reasonable care in dealing with matters that hurt the plaintiff. That view is in tension with the explicit language of the Lex Aquilia, where the word occidere (killing) is not coupled with either intention (dolus) or some form of fault (culpa). Instead the word used is iniuria, which literally means “unlawfully,” without linking it to the denial of negligence and intention (Nicholas 1962: 222), as is commonly done.⁵ That position is also taken explicitly in Gaius, where it is said baldly, “Now a person is understood to kill wrongfully (iniuria) when this death occurs through his willful intent (dolus) or negligence (culpa)” (Inst. 3.211). Note that this approach precludes the possibility of all strict liability cases, prior to any examination of the substantive arguments in favor of that position. Under strict liability the unadorned claim that the defendant killed the plaintiff is dismissed just as if the complaint had said that the plaintiff had been killed by an act of God, which is on both moral and philosophical grounds indefensible.⁶ Nor does this ostensible definition of iniuria leave the field open for the full range of affirmative defenses based on plaintiff ’s conduct, let alone instruct on how to treat those cases in which some excuse or justification is offered, when it is very unlikely that the identical set of defenses applies alike to negligence and intentional harm cases. As becomes clear later on, iniuria is best understood not ⁵ For a discussion of the common law of torts, see Williams (1951). ⁶ See Holmes (1881: 96): “Unless my act is of a nature to threaten others, unless under the circumstances a prudent man would have foreseen the possibility of harm, it is no more justifiable to make me indemnify my neighbor against the consequences, than to make me do the same thing if I had fallen upon him in a fit, or to compel me to insure him against lightning.” But what of the want of foresight by the plaintiff of the defendant’s application of force?

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

310

Richard A. Epstein

as a description of any mental state or extrinsic circumstance, but as a conclusion of law that could never be substituted for a particular allegation, so that just as the earlier definition of iniuria was too limited, so too is its common alternative, which has, of course, wide acceptance in modern tort law as well (Holmes 1881: 88–96; Brown vs Kendall 1850). A similar issue could arise with a breach of promise, where, however, it is rarely held that no promise is actionable unless the breach is either willful or negligent. And once again, allowing the strict liability case at the first level still permits the introduction of other mental states further down the road. It is, to be sure, a wise pleading tactic on the part of plaintiffs in an unregulated system to allege any negligence or harmful intention in the prima facie case if only to quiet any moral qualms about the strength of the claim, so that a trier of fact will be strongly inclined to rule in the plaintiff ’s favor. Nonetheless, that approach must be strictly forbidden as a systematic matter. The point here is not that mental states and care levels are irrelevant to the law of tort; rather, they have played a central role in fleshing out the system from the beginning. The only question on this score is where to slot them into the overall system, not whether they have a part to play in it. The insistence that each case be only presumptive also structures the terms that are permissible to introduce at any stage of a case. Thus, the plaintiff cannot plead that the defendant trespassed on his land, for that creates the conclusion that the defendant had committed a wrong, which may or may not have been the case. The correct way of putting the point therefore is to say that the defendant entered the plaintiff ’s land, which creates a prima facie case of trespass, which in turn might be excused because of a mistake in the location of the boundary or justified by showing that the plaintiff permitted the defendant to enter the land. Likewise, the plaintiff cannot say that the defendant killed the plaintiff ’s relative. The prima facie case is one of (deliberate) killing, which is a prima facie case of murder, but not necessarily murder given the possibility of affirmative defenses (how intention to harm is introduced will be discussed later). Similarly, on the contract side, the prima face case in the pleading can never say that the defendant breached his contract. The correct form of argument is that the defendant did not keep his promise, which leaves open the possibility that there was some excuse or justification for that case of nonperformance, as the case may be. In effect the English language (and I dare say all others) has these pairs—nonperformance,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 311 breach; entry, trespass; deliberate killing, murder—where the first of the pair sets the presumption which becomes a legal wrong only when the various responsive pleas are inapplicable in the case. On this score, therefore, systematic reliance on pleadings offers a useful decision tree that allows the analyst to see where each discrete allegation fits into the overall picture. The insistence that there be only the minimum elements stated at a given point in the analysis means that there is then an opportunity for additional elements to be worked into the issue later on. Indeed, to have two or more alternative theories of liability that start with the prima facie case leads to the obvious question of how the two of them can fit simultaneously into the overall picture, and how, and why, to parcel out different causes of action between them, which is the difficulty of the standard account of iniuria. That problem is avoided by the system of staged pleadings, which makes it possible in time to work all relevant elements into the overall system. The same result is reached when the issue is looked at from a more explicitly economic view. The tapestry of potentially relevant facts is always very large, and the correct approach for ordering them is to ask of the plaintiff to select from the full range of facts those that advance his case, that is, to give reason why the law will create a social improvement if it imposes prima facie form of liability in the plaintiff ’s case. Once that is done, the plaintiff can rest as the defendant comes forward with an argument to explain why first appearances are, as Gaius said, deceiving and that further information may well show why it is now unwise to allow that cause of action, subject to further pleadings on the matter. It should be apparent whenever this system is used that the vast bulk of the cases are governed in practice by those issues that come up early in the pleading sequence. Most cases of force used against strangers, for example, are not used in selfdefense, so that much is gained if the collection of accidental injuries is handled before tackling any difficult questions arising from intentional harms. The same kind of arguments can be made about various issues in contract law, where there is again the same alternation. The plaintiff has to plead the prima facie case in minimalist terms, after which the defendant is allowed to excuse or justify his nonperformance subject to further sequential pleadings by both sides until the issues are finally joined. The success of the pleading system does not depend solely on the fact of any one particular alternation. It is also critical to see that the

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

312

Richard A. Epstein

right substantive boxes are filled to make sure that the system does not go wrong. In order to do this, the legal system has to solve two key questions. The first is to determine the initial baseline that establishes the relative position among all individuals, dispersed as they are across the globe. The second is to develop a set of legal responses for persons who make moves that are inconsistent with these critical assignments of rights. The first task is often discussed in political theory under the rubric of “state of nature” theory or, more recently, under the more abstract Rawlsian formulation of the “original position,” which is intended to make the question look less historical and more analytical. The second task is to develop the rules that go into place once the basic norms are established. The pleading issues are key to this second development, but they can only be understood by starting with how the key Roman distinction between actions in rem and actions in personam sets the table for the subsequent pleading analysis, by indicating the scope of the plaintiff ’s interest as protected by law. Thereafter the pleadings set the roadmap for dealing with cases in which individual defendants do not follow, as it were, the rules of the road established in the initial legal position.

20.2.1. Actions in personam and in rem That basic distinction between actions in rem and in personam is part of every legal system, even if the Latin words “against the thing” or “against the person” do not translate easily into English (Epstein 2011). In a procedural sense, an action is said to be in rem if it is directed toward the thing, which the plaintiff claims to be his own (Nicholas 1962: 99–103). That vindicatio rei (vindication of title in the thing) does not name a defendant as such, but clearly anyone who wishes to contest the claim of ownership is entitled to appear on his own behalf. The hard procedural questions are the extent to which notice has to be given, so that others with a claim to the property have a chance to be heard before their ownership rights are blocked, and that notice could be provided in principle either by personal service to known rivals or publication to others. One way in which this is done in English law, with chattels, is for the plaintiff to make a demand for the return of the object, which when refused creates a prima facie case of the tort of conversion against this particular defendant, leaving all other parties (none of whom are in possession) out of

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 313 the picture. But the formal differences by which these suits begin should not conceal their common object, which is to restore possession of the disputed thing to the plaintiff, who then has an interest in the property that is good against the entire world. In that sense the vindicatio rei is different from any action brought against a particular person for some wrong that he has done to the plaintiff insofar as it protects rights in rem (i.e. claims for forbearance against the rest of the world) and not rights in personam, by which more exacting obligations were imposed on a particular person or persons by agreement. Conceptually, this distinction is as relevant as to modern law as it is to Roman law (Chang and Smith 2012). We are now in position to see the close connection between rights in rem and actions in tort, and rights in personam and rights in contract. In both cases, the basic method comports first with the basic logic of ordinary language and second with a rational economic approach that allows the overall system to be developed by a series of successive approximations, each of which, if decided correctly, takes us closer to some social optimum, much the way a mathematician finds the area under a curve by approximating it with ever thinner rectangular slices, which in the limit define the area under a curve.

20.2.2. Actions ex delicto In the tort context, if mutual forbearance from entry is the requirement of an in rem action, the action that constitutes the prima facie tort is a physical invasion that violates the plaintiff ’s space. The bare entry is enough of course to allow the plaintiff to remove the defendant. Similarly, the taking of a chattel offers a prima facie case for its recovery. Actions for damages in the event of partial or complete destruction of the property follow on closely, because no defendant should be able to escape liability by the use of this special stratagem of timing the sequence of taking or destroying. It would be grotesque to allow a defendant to first destroy the property without legal consequences, and then be responsible only for the return of the land or object that he has destroyed. The constitutional law of the United States treats takings and tort damages as a pair, precisely because the private law has made that same fundamental structural move. This narrow definition of the prima facie case prevents the disintegration of the “harm principle” in the law of tort so that it covers

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

314

Richard A. Epstein

any externality, i.e. an action that has a negative impact on other individuals. It is more than coincidental that the violation of these in rem rights involves in the first instance some physical invasion of the property of another—the boundary-crossings that Robert Nozick made famous in Anarchy, State, and Utopia, without any knowledge of its Roman antecedents or economic logic (Nozick 1974). Once that invasion has taken place, the party responsible has singled himself out of the mass of other individuals by upsetting the balance of mutual forbearances found in the in rem world. At this point, traditional notions of corrective justice, drawn in large measure from the work of Aristotle, kick in and shape the remedy that should be awarded (Nicomachean Ethics, Book 5). In those cases where the land or chattel is taken, the remedy is its restoration to the true owner. That restoration cannot, of course, restore the lost (rental) value of interim use or compensate for any damage to the property while in the hands of other individuals. Accordingly, the law supplies damages to fill in the gap with the object of returning the system to the status quo ex ante to the extent that the law is able. It is critical to stress that corrective justice, so conceived, is thus not diametrically opposed to some economic approach to the subject, even if it is often treated in that fashion. To the contrary, a sound understanding of corrective justice provides a platform that allows for the achievement of particular economic objectives. In those instances where the chattel or an improvement on land has been destroyed, the restoration remedy is no longer feasible, so the system turns, for want of a better solution to damages as a second-best substitute for unwinding the transaction. Those damages are in general awarded on the same ground as in the case of damage or delay. But as they occupy the entire landscape instead of some small fraction of it, the weaknesses in that system become more acute. In dealing with slaves, for example, the Romans did not normally award pain and suffering. In the analogous action allowed to a free person, some estimate of those damages should normally be allowed notwithstanding the want of any market standard to determine the harm in question. The basis of liability under the principles of minimum sufficiency outlined above should reflect (as the text of the Lex Aquilia suggests in ch. 1) a strict liability system for anyone who kills (occidere = killing by cutting) a slave of either sex or a herd animal (pecus). By the same token the parallel passages in chapter three dealing with “burning, breaking, or destroying” are subject to the same basic reading, Here,

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 315 as earlier, it is key to distinguish between a denial that the defendant has acted and an effort by the defendant to excuse or justify his actions. Conceptually, it is one thing for a defendant to be pushed, and quite another for him to act under duress in order to escape some greater harm.⁷ Thus it becomes important to ask what happens whenever B enters A’s land because he, B, has been driven there under threats of force by C. At this point, B cannot plead the general denial, because acting under compulsion is not the same as not acting at all. In principle, one could allow this narrow defense on substantive grounds, by insisting that the only proper action should be brought against C, who did not enter the land but who had maximum control over the situation. The result is that if all persons were present and solvent in a single proceeding, B would be exonerated even though the direct cause of the harm, and C would be liable in full although he did not enter the land. Yet even this simple scenario poses serious procedural difficulties under the Lex Aquilia, which does not supply an action of indemnity by B against C—you made me kill him. Any direct action by A against C forces these substantive ambiguities to the surface, because it is now necessary to unpack longer causal chains with multiple defendants. The correct set of answers requires not a yes/no or on/off answer on causation, but instead demands an explicit causal hierarchy among all relevant parties, such that A can recover from either B or C, and B can recover from C, and C can recover from no one at all (Epstein 1973b). The English common-law cases raise exactly those same ambiguities. Thus, in Weaver vs Ward (1616), the judges write: no man shall be excused of a trespass (for this is the nature of an excuse, and not of a justification, prout ei bene licuit [as it well appeared to him]) except it may be judged utterly without his fault. As if a man by force take my hand and strike you, or if here the defendant had said, that the plaintiff ran cross his piece when it was discharging, or had set forth the case with the circumstances, so as it had appeared to the Court that it had been inevitable, and that the defendant had committed no negligence to give occasion to the hurt.

⁷ For the variations, see for the Roman version (D. 9.2.7.3; Ulp. 18 ad ed.), and for the English law the contrast between Smith vs Stone (1647) and Gilbert vs Stone (1647).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

316

Richard A. Epstein

Note that the evocative phrase “utterly without fault” conceals the difference between the denial of voluntary action—“as if a man take my hand by force and strike another”—and an affirmative defense based on the plaintiff ’s conduct (running “cross his piece”). The full passage closes the triad with the affirmative defense of inevitable accident, which is overcome in the third stage by the reply that the defendant must lose if he had done some prior act to “occasion” the harm. The use of the weaker causal word “occasion” strikes the correct balance, because the use of force has been pleaded at stage one, and it cannot therefore be raised anew in the replication. Since, moreover, no justification could be offered for an accidental killing—in this case an early instance of friendly fire—the court rightly said that only an excuse could be raised, without seeing, as becomes clear later on, the distinctions between excuses that enter the pleadings at stage two and stage six. This discussion of both the Roman and English cases shows how easy it is to jumble up different elements in a complex fact pattern when the pleading rules are not scrupulously observed. The narrow definition of causation involving the use of force, however, also sets the stage for invoking a broader sense of causation which in Roman law covers the case of “furnishing a cause of death” (causam mortis praestare), which is set in opposition to kill (occidere), which gives rise to an action in factum, or what the English call an action on the case. The Digest makes this leap from direct force so that the protection of the plaintiff ’s individual autonomy cannot be circumvented by taking simple evasive steps, as in the case where the defendant literally sets poison in front of the woman who then drinks it herself (D. 9.2.9 pr.–1; Ulp. 18 ad ed.). Yet this expansion of liability is not open-ended because it will not cover cases in which the woman drinks the poison knowing of its danger when not subject to any external threat of force. As economic theory predicts, the key cases are those of asymmetrical information when the defendant knows, as the plaintiff does not, that the drink contains poison. The situation takes on an added dimension because ordinary Latin (and ordinary English) both speak of possible excuses and justifications for given courses of conduct. No one has to justify actions that do not harm other people, or even those that harm them accidentally: only excuses based on the plaintiff ’s conduct come into play in the latter case. Blocking the plaintiff ’s right of way (which presupposes a system of property rights, lest someone claim that the plaintiff ’s face

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 317 got in the way of the defendant’s fist) provides a good causal defense against strict liability. People can debate whether that defense should be total or partial, so long as they understand that proof of antecedent negligence by the plaintiff (except in the odd sense of negligence per se, a move back to strict liability in disguise), is technically speaking not needed to raise the defense (Epstein 2012). A parallel affirmative defense is that the plaintiff was injured while on the defendant’s land. That entrance establishes plaintiff ’s prima facie trespass—which can be answered in two ways: one for strangers, and one for licensees and invitees, who enter consensually. By the former, the plaintiff concedes the trespass but rightly insists that the defendant deliberately struck the plaintiff who had crossed a field where javelins were being thrown at “an inopportune time” (D. 9.2.9.4; Ulp. 18 ad ed.). Lawson mistakenly treats deliberate harm as a species of negligence, even though the word culpa means culpability, i.e. a generalized form of blameworthiness, and not negligence. Clearly the word “noxious” lodges a more powerful charge than negligence, which generally will not suffice, because the plaintiff ’s trespass does not impose a duty of care (Lawson 1950: 222–3). Nonetheless, the exact content of deliberate harm still needs further elaboration, especially in cases where the defendant acts in reckless disregard of consequences. But what now becomes crystal clear is that deliberate harms only make their way into the original strict liability case at the third stage of the argument. Compressing the pleadings into two stages thus conceals that the defendant’s mental element is relevant if, and only if, it is necessary to overcome an initial excuse based on plaintiff ’s conduct. Negligence can also become operative in cases of consensual entry. Thus, if the plaintiff alleges a want of reasonable care, the defendant responds at stage four that the defendant as a social guest assumed that risk as well, for example, of latent defects not known to the defendant. But that defense will fail against a business guest with respect to latent dangers discoverable by reasonable inspection.⁸ Similarly, when a defendant surgeon cuts the plaintiff patient, assumption of risk is a defense against a strict liability case, but ⁸ For the old categories of duties owed to various kinds of entrants, see Robert Addie & Sons (Collieries), Ltd. vs. Dumbeck (1929); for their rejection, see the California Supreme Court’s decision Rowland vs. Christian (1968). For commentary, see Epstein 1999: 329–31.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

318

Richard A. Epstein

ordinarily not against negligence. The Lex Aquilia picks up that point independently of the pleadings: “Proculus says, if a doctor operates unskillfully on a slave, an action lies either ex locato (for hire of services) or under the Lex Aquilia” (D. 9.2.7.8; Ulp. 18 ad ed.). These contractual variations are great whenever two parties who have singled out each other for special dealings. Accordingly, this concurrence of actions only applies for “operations.” The defendant who fails to treat a patient when there is a duty to do so is easily held liable on the contract of hire, but not under the Lex, which has great difficulty in dealing with cases of omissions. In principle, consensual parties may allocate risk as they choose, for they are not bound by the simple allocation rules that govern cases between strangers. To return to these stranger cases, justifications for deliberate harms only enter the case at the fourth stage. The first type, relevant at stage two, involve excuses for accidental harms based on plaintiff ’s conduct. The second type of excuse enters at stage six as follows. A justification for deliberate harm, e.g. self-defense, enters at stage four, after an allegation of deliberate harm. At stage five, the plaintiff argues that the defendant exceeds the scope of that justification, say by the use of excessive force, or even disproportionate force (a much tougher nut to crack). In both cases, the measure of damage is reduced, to cover only the excessive force in the first case and the disproportionate force in the second. Next at stage six the defendant points to the weaknesses in his mental or physical ability that prompted the use of excessive force. At this point an extended version of the defense of “you take your victim as you find him” is incorporated in both Roman and English law (D. 9.2.7.5; Ulp. 18 ad ed.).⁹ The same logic applies to the excuse of non-consent to deliberate harm. At stage five, the plaintiff could claim that he was induced to accept the operation by fraud, or perhaps by the failure to disclose known risks that a fiduciary physician should make clear. The injured pugilist may allege that the defendant did not use proper equipment, so that injuries inflicted went beyond the scope of the consent that was given. Assumption of risk for accidental harms, at the second

⁹ “But if one gives a slight blow to a sick slave and he dies, Labeo rightly says that he is liable under the Lex Aquilia, because different things are fatal to different people.” The text is not quite right because it does not distinguish between the treatment case, which is governed by the rules of medical malpractice, and the case involving the stranger, which is not.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 319 stage is thus distinguished from consent to deliberate ones at the fourth. By contrast any plea of insanity (as noted in Weaver vs Ward) is quite inappropriate at the second stage. The tort system with its strong views of individual autonomy does not allow any insane person a direct claim of assistance from B, a stranger. If A cannot force the obligation of care on B, a stranger, he cannot achieve that result by raising the defense of insanity when his own actions have injured or killed another. The same argument can be made about infancy or any other forms of legally-recognized capacity impairments. The infant, the incompetent, or the novice cannot be asked, as the victim of an attack, to perform beyond his capabilities. Nonetheless, the law cannot allow any infant or insane person to hide behind his status to use unnecessary force. To counter that risk, the usual way of framing the stage six excuse asks whether the defendant’s use of either excessive or disproportionate force was done reasonably and in good faith. This balancing act helps address the irreducible uncertainties that surround defensive actions under stress. The good faith element allows a certain subjective inquiry; the reasonableness element tries to cabin it within an objective standard. Once the initial balance has been upset, the clear rules that work at stages one and two no longer apply. But so long as the hardest cases arise in only the lowest probability situations, the system will work even if perfection is unattainable.

20.2.2.1 Summing up the tort side These disparate pleading paths, then, explain how peaceful coexistence among the three basic theories of liability should be achieved. Strict liability sets the uniform prima facie case. Negligence comes in as an offset to the assumption of risk cases in the various special relationships of which physician and lifeguard, babysitter, landlord and occupier are some common examples. These cases will be subject to contractual variations, both express and implied, which explains why divergence in pleas at stages three and four forms a necessary and proper part of the overall system. By the same token, the use of deliberate force at stage three raises squarely issues of consent and self-defense, which show how closely entangled justifications, and the limitations on those justifications, apply as well. None of these

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

320

Richard A. Epstein

arguments are particularly dependent on the fine points of Roman culture. Indeed, the rules that were developed first with slaves cover all sorts of situations involving personal injury and property damage in a more or less timeless fashion. The move from Rome AD 160 to New York AD 2018 is relatively seamless. That is indeed as it should be, lest every change in social circumstance require a recalibration of the fundamental rules of tort law. By the same token, the variations that arise in cases of assumption of risk and consent allow for degrees of individuation that are as valuable across different people working in the same general social milieu, as it does for social changes across milieus and millennia. The care expected of a babysitter is likely to vary less significantly over generations than the care expected of an experienced surgeon, and these rules accommodate all the relevant distinctions. Start with the pleadings and the system can run its course.

20.2.3. Actions ex contractu The extended account of the role of pleadings in tort theory makes it easier to see how the basic approach carries over to contract disputes. The initial statement of the complaint should be bare bones, leaving the complications for later pleas when they become relevant. It is commonly said on moral grounds that people ought prima facie to keep their promises.¹⁰ In this regard, the law deviates evidently from the Kantian premise that “there are certain duties of perfect obligation, such as those of fulfilling promises, of paying debts, of telling the truth, which admit of no exception whatsoever” (Ross 1930: 18). But that rigid position goes so much against the accumulated weight of every legal system that it is idle to take it as the final resting point. What is needed is a way to systematize these insights as presumptions instead of per se rules, which is what the pleading metric allows. In this context, we start with the proposition that it is prima facie wrong not to keep one’s promises. The point here is that it is only through promising that people can gain the benefits of cooperation ¹⁰ See, for example, the discussion of prima facie obligations, in Ross 1930: 20–5. The notion of the prima facie case “is an idea that simply had not occurred to critics of deontological theories (Mill, Moore, Sedgwick, Rashdall) and therefore represented at the time a major advance in the dialectic existing between utilitarians and nonutilitarians” (Stanford Encyclopedia of Philosophy 2012).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 321 over time. People make promises that they are willing to keep in order to elicit a like response from others, so that both sides can mutually improve their position. To allow the demurrer to this basic prima facie case would upset every known fact of social organization. To keep the plea sufficiently general, the basic case starts with a general strict liability standard, so that presumptively contracts create warranties of result, not promises for effort. Yet this presumption can be overcome in a variety of ways. In most cases, the moral force of a promise is not explicitly tied to the use of legal remedies. It should therefore be a defense against the legal enforcement of any particular promise for the defendant to demonstrate that the parties had no intention to create legal relations (Balfour vs Balfour 1919), which could be proved either explicitly or implicitly from context. In addition, it could also be urged that the promise in question was not supported by consideration so that it was a gift promise, which should not ordinarily be enforced.¹¹ To that position, it is then possible to create exceptions with respect to certain forms of gift promises, including those for gifts to charities or on the occasion of marriage, which might be made enforceable as a substantive matter. It is also possible to argue that even though these promises may not be enforceable when fully executory, they are enforceable after the plaintiff delivered some object to the defendant, for which under the law of bailments the transferee has an obligation to return. Or in some cases, the promise could simply be that the defendant agreed to take care of a chattel that he borrowed from the defendant, so that the issue of executory enforcement is nowhere to be found. It is also possible in many cases to make the further argument that the defendant who is prima facie obligated to repay a debt could plead any release as binding, even if it is not enforced by consideration, which is the Roman rule but not the English (Gai., Inst. 3.169–76). The possible defenses to the enforcement of a promise are not solely related to the issues of intent and consideration. In other cases, an affirmative defense could be used in context-specific ways to undercut the strictness of the basic contractual obligation. That adjustment will in general not be needed for the provision of certain kinds of standardized products. No one thinks that the implied warranty of merchantable products like milk or canned goods is

¹¹ See, for example, Congregation Kadimah Toras-Moshe vs DeLeo (1989).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

322

Richard A. Epstein

met by the seller’s good faith effort to deliver a safe product. But when the analysis turns for contracts dealing with services, including those for medical malpractice, the assumption of risk defense becomes powerful given the intrinsic uncertainty in medical outcomes, so that liability for negligence (usually based on customary practice) becomes the norm. In rare cases that can be displaced, as when the defendant is anxious to conduct experimental treatments of the plaintiff for his own collateral benefit, at which point the standard of care remains strict (Hawkins vs McGee 1929). It is through this lens of assumption of risk, which is always highly context-specific, that the multiple standards used, for example, in the bailment cases are developed, each depending in some measure on the distribution of benefits between the two parties (Epstein 2009; Coggs vs Bernard 1704). The greater the benefit to the bailee, the higher the standard of care should be, and the similarly so in reverse. It is only in a limited subclass of cases that the standards commonly chosen are the reasonable-care standard associated with the formula Judge Learned Hand developed in United States vs Carroll Towing (1947). The same element of pleading is exceedingly important in connection with the law of conditions, which deals with issues related to the sequence of performance of various promissory obligations in a contract. Thus, the simplest agreement is one in which A promises to sell goods for $10, with nothing said about the order of performance. The modern form of pleading generally says that the plaintiff should plead the performance or the occurrence of all conditions in order to maintain the action, where the former term refers to conditions within the control of the defendant and occurrences refers to natural events (Federal Rules of Civil Procedure 9(c)).¹² But that general pronouncement has no heft, for all the work comes when the defendant is required to deny that performance or occurrence “specifically and with particularity.” The denial then puts the condition into play and sets up the possibility that the plaintiff may be able to explain in the third stage why the nonperformance or

¹² “In pleading conditions precedent, it suffices to allege generally that all conditions precedent have occurred or been performed. But when denying that a condition precedent has occurred or been performed, a party must do so with particularity.”

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 323 nonoccurrence of the condition should not be a defense. Thus, it could be said that the defendant need not pay the money because he has yet to receive the goods. But to that there is a further reply that the plaintiff has tendered the goods so that this defense is not available, or the defendant waived the condition so that replication undoes the force of the defense. At this point, the system introduces excuses and justifications for the nonperformance of conditions. There are still other implied conditions, based on impossibility of performance or fundamental mistake that are accepted in some case but not in others, including the difficult Roman cases where the plea against the enforcement of a promise is error in substantia, so that the thing delivered is not the thing promised. The same approach that is congenial to the law of tort carries over well here. Other conditions are based on the sequence of performance that is either stated in the agreement or which can be teased out from its other terms or from a background set of default norms, the chief purpose of which is to minimize the exposure that each side has over the life of a complex contract, which happens when one party gets too far ahead of the performance of the other side. The underlying rationale is strongly economic. “The policy of the law, here as in the tendency to construct concurrent conditions, is to minimize credit risks” (Patterson 1942: 920). The simple rule that may demand either payment or performance unless a party is prepared to tender his obligations is one example of that basic position. Timing progress payments on staged performance in construction contracts is yet another way to minimize the risk that one party will get too far out in front of the other. There is, in the context of a short article on pleading, no possibility of going through all these variations, each of which can be teased out of the basic position of Roman and Anglo-American sales law. All that can be done in this instance is to note that the well-organized system of pleading tracks the usual strategy of a complete contingent state contract, whereby, at each point in the sequence of performance, each party knows what he must do to keep the contract alive, and what consequences—damages, release of the other side, delay in performance—will attend each type of nonperformance. Once again, the right system of pleading does not make the substantive judgments, but it does help set up the problem so as to allow doctrine to be organized, and cases to be distinguished, with greater precision that would otherwise be the case.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

324

Richard A. Epstein 20.3. CONCLUSI ON

The strategy of this article has been two fold. It first develops the key distinctions of Roman procedural law that helped facilitate the development of the substantive law of property, contracts and torts. It then applies those rules to a number of critical choices in the organization of any legal system. To be sure, the historical development of these rules reveals little or no conscious interest on the part of classical Roman writers—or for that matter their English and American peers—to find some explicit economic logic that guides the law of pleading. But initial appearances can deceive, especially when it is possible to examine these various doctrines with some degree of closeness to demonstrate how they facilitate the articulation of a substantive set of jurist-made or judge-made decisions that in large measure have just that effect. The Roman system proceeded in a bottom-up incremental fashion, by breaking up complex cases and dividing them into their constituent parts that are then carefully reassembled into a large whole. This effort to strip out one issue after another is quite different from the modern approach in all substantive law, which invokes some overarching notion of reasonableness very early in the inquiry. There is no doubt that every legal system ultimately involves an interestbalancing inquiry that is implicit in these modern tests of reasonableness. But the key point in this case is not whether these judgments should be made, but how they should be made. I have indicated—and this is the second main point of this article—how various maneuvers dealing with the definition of terms such as iniuria can lead to an oversimplification of tort and contract doctrine that conceal a set of distinctions that, once laid bare, make the system more coherent than if reasonableness were the starting point of analysis. The idea of prima facie obligations has been around for a long time. The great achievement of the Roman doctrine of pleadings is that it shows with great force and clarity the way in which it can be extended from a vague set of moral intuitions into a systematic body of law that functions as well in modern times as it did in ancient times.¹³ ¹³ My thanks to Mateo Aceves, University of Chicago Law School, Class of 2015, and to Bijan Aboutorabi, Thomas Malloy, and John Tienken University of Chicago Law School class of 2018—four accomplished Roman law students—for their invaluable research assistance on earlier drafts of this article.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

How Roman Pleading Rules Shape the Substantive Structure 325 REFERENCES Brown, John Prather. 1973. “Toward an economic theory of liability,” 2 Journal of Legal Studies 323–49. Buckland, William Warwick, Arnold Duncan McNair, and F. H. Lawson. 1965. Roman Law and Common Law: A Comparison in Outline. Cambridge: Cambridge University Press. Chang, Yun-Chien, and Henry E. Smith. 2012. “An economic analysis of civil versus common law property.” 88 Notre Dame Law Review 1–55. Danner, Henry R. 1912. Roman Law Pleading: An Outline of its Historical Growth and General Principles. Indianapolis: H. R. Danner. Epstein, Richard A. 1973a. “Pleadings and presumptions.” 40 University of Chicago Law Review 556–82. Epstein, Richard A. 1973b. “A theory of strict liability.” 2 Journal of Legal Studies 151–204. Epstein, Richard A. 1989. “The utilitarian foundations of natural law.” 12 Harvard Journal of Law and Public Policy 713–51. Epstein, Richard A. 1997. “The modern uses of ancient law.” 48 South Carolina Law Review 243. Epstein, Richard A. 1999. Torts. Gaithersburg: Aspen Law & Business. Epstein, Richard A. 2003. Skepticism and Freedom: A Modern Case for Classical Liberalism. Chicago: University of Chicago Press. Epstein, Richard A. 2005. “The not so minimum content of natural law.” 25 Oxford Journal of Legal Studies 219–55. Epstein, Richard A. 2009. “The many faces of fault in contract law: or how to do economics right, without really trying.” 107 Michigan Law Review 1461–77. Epstein, Richard A. 2011. Design for Liberty: Private Property, Public Administration, and the Rule of Law. Cambridge, MA: Harvard University Press. Epstein, Richard A. 2012. “The irrelevance of the Hand formula: how institutional arrangements structure tort liability,” in Jef De Mot, ed., Liber amicorum Boudewijn Bouckaert. Brugge: Die Keure, 65–76. Geach, Peter T. 1960. “Ascriptivism.” 69 Philosophical Review 221–5. Hart, Herbert L. A. 1949. ‘The ascription of responsibility and rights in XLIX.” 49 Proceedings of the Aristotelian Society (New Series) 171–94. Hart, Herbert L. A. 1983. Essays in Jurisprudence and Philosophy. Oxford: Clarendon Press. Holmes, Oliver Wendell. 1881. The Common Law. Boston: Little, Brown. Lawson, Frederick H. 1950. Negligence in the Civil Law. Oxford: Clarendon Press. Lenel, Otto. 1927. Das Edictum Perpetuum: ein Versuch zu seiner Wiederherstellung. Leipzig: Tauchnitz.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

326

Richard A. Epstein

Maitland, Frederick W. 1936. The Forms of Action at Common Law: A Course of Lectures, ed. Alfred H. Chaytor and William J. Whittaker. Cambridge: Cambridge University Press. Metzger, Ernest. 2013. “An outline on Roman civil procedure.” 9 Roman Legal Tradition 1–30. http://romanlegaltradition.org/contents/2013/RLT9METZGER.PDF. Nicholas, Barry. 1962. An Introduction to Roman Law. Oxford: Clarendon Press. Nozick, Robert. 1974. Anarchy, State, and Utopia. Basic Books. Patterson, Edwin W. 1942. “Constructive conditions in contracts.” 42 Columbia Law Review 903–54. Pitcher, George. 1960. “Hart on action and responsibility.” 69 Philosophical Review 226–35. Poithier, Robert Joseph 1761 [1806]. Traité du contrat de louage et traité des cheptels, revised edition by M Hutteau. Paris: Letellier. Ross, William David. 1930. The Right and the Good. Oxford: Clarendon Press. Stanford Encyclopedia of Philosophy, “William David Ross,” revised 19 June 2012, available at http://plato.stanford.edu/entries/william-david-ross/. Schauer, Frederick F. 1976. “English natural justice and American due process: an analytical comparison.” 18 William and Mary Law Review 47–72. Sutton, Ralph. 1929. Personal Actions at Common Law. London: Butterworth & Co. Ltd. Williams, Glanville. 1951. “The aims of the law of tort.” 4 Current Legal Probs. 137–76.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

21 Private Prosecution and Enforcement in Roman Law David Friedman

21.1. INTRODUCTION Modern legal systems use two mechanisms to impose costs on those who injure others. Under tort law, the victim or his agents initiate and prosecute the case; if the verdict is a monetary penalty, it is paid to the victim. Under criminal law, the victim is replaced by the state. The offense is treated as an offense against the state; it is the state that initiates, controls, and pays for prosecution and, if the verdict is a fine, collects it. Anglo-American common law developed out of a legal system, Anglo-Saxon law, in which offenses were for the most part treated as torts; a killing created a claim for compensation by the kin of the victim.¹ The state had a claim only if and to the extent that the offense was also seen as a breach of the king’s peace. In a purer form of such a system, exemplified by the legal system of saga-period Iceland, not only the prosecution of the case but also the enforcement of the verdict was entirely the responsibility of the victim.² The state’s role was limited to authorizing the use of private force through its system of laws and courts. Even that role is missing in a still purer

¹ For an extensive survey of wergeld in Anglo-Saxon and other early societies, see Seebohm (1902). ² Friedman (1979); Friedman et al. (2019: 147–69). David Friedman, Private Prosecution and Enforcement in Roman Law In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0021

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

328

David Friedman

feud system such as the traditional Somali legal system,³ based on customary law and the private arbitration of disputes. The logic of a feud system is simple. If someone wrongs me, I demand compensation and threaten to harm him if I do not get it. If someone kills me, someone else, typically my heir, inherits my claim and the duty to enforce it. In order for this to work as a way of discouraging wrongs, the threat must be more believable when I have been wronged than when I have not—there must be some mechanism such that right makes might. In both the Icelandic and Somali systems, the relevant mechanism is a court trial, in the former case in a pre-existing court system, in the latter in a court created for the case under traditional procedures. In the feud system currently existing among the Romanichal gypsies in Britain, the mechanism is more informal. When I accuse you of wronging me and threaten to beat you up if I do not receive suitable compensation, both of us know that if you did wrong me according to the norms of our community, my friends will back me and your friends will not back you.⁴ The role of the avenger of blood, the heir of the victim of a killing in Jewish law, suggests that that legal system developed out of a preexisting system of decentralized, privately enforced law. If the court finds the killing capital, it is the avenger of blood who is supposed to execute the sentence. If the court instead sentences the killer to temporary exile,⁵ the avenger of blood has the right to kill him if he can catch him on his way to the city of refuge.⁶ The law forbids the ³ A good account of the Somali system can be found in Lewis (1961). See also Friedman et al. (2019: 170–82). ⁴ Weyrauch (2001: ch. 3). ⁵ Exile is in one of the cities of refuge and lasts until the death of the high priest. See Neusner (1988: The Order of Damages, Makkot, 2:4–2:7, 614–15); Klein (1954: V.1.195). ⁶ Another piece of evidence that Jewish law was built on top of a system of privately enforced law is the treatment of situations where the amount of a debt is ambiguous— where, for instance, a carelessly drafted document refers to the sum as 50 zuz in one place and 100 zuz in another. The court can only compel the payment of 50 zuz, since it does not know that the larger sum is owed. But if the creditor has seized 100 zuz from the debtor, the court cannot compel him to return part of it, since it does not know that 100 zuz was not owed. “Therefore, in the case of every instrument which contains a recital with two different meanings, one referring to a greater sum and the other referring to a lesser sum, the holder of the instrument collects only the lesser sum. But if he seized the greater sum it may not be reclaimed from him without clear proof ” (Rabinowitz 1949: 187). A similar pattern occurs with tort damages: “If an animal stamps on the ground on the premises of the plaintiff and as a result pebbles spring up and cause damage there, the defendant must pay for (only) one quarter of

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

329

avenger of blood from agreeing to a money payment instead,⁷ which suggests that the equivalent of the Icelandic wergeld had at some previous point been a common practice. In Jinayat, the part of Muslim law that applies to one person killing or injuring another, the offense is treated as an offense against the victim or, if he is killed, his kin. The court determines guilt and sets a penalty, either a payment in camels or retaliation against the offender. The victim or kin can either enforce the penalty or agree to waive it, gratis or on condition of some lesser penalty.⁸ Can a similar pattern can be found in Roman law? Is there evidence that it too developed out of a system in which prosecution of offenses and the enforcement of verdicts was primarily the responsibility of the victim?

21.2. THE HISTORY The traditional date for the foundation of Rome is 753 BC. The political system was converted from monarchy to republic, again according to the traditional dating, in 509 BC, and to an empire in 27 BC. The first written text of the law which can be substantially, although not completely, reconstructed⁹ is the Twelve Tables, composed according to the Roman accounts in 451–450 BC. While there was extensive writing on the law over the next thousand years, the next complete official account of the law was Justinian’s codification project of AD 529–34.¹⁰ the damage . . . . But if the plaintiff seizes one-half of the damage, we may not take it away from him” (Klein 1954: XI.1.2, part 6, p. 8). That implies a system where such claims were at least sometimes privately enforced. ⁷ “and Scripture says, Moreover ye shall take no ransom for the life of a murderer (Num. 35:31)” (Klein 1954: V.1.194). ⁸ Hallaq (2009: 310, 320–2); Friedman et al. (2019: 98–9). There is an additional penalty due in the form of penance—freeing a Muslim slave or fasting during daylight for two consecutive months. ⁹ Reconstruction is from multiple sources, not all consistent with each other. For a discussion, see Crawford (1996). ¹⁰ Consisting of the Institutes, Digest, and Codex. There is, however, an almost complete legal textbook by Gaius from the second century, Gordon and Robinson (1988), which is a major source of information about the legal system at and before his time.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

330

David Friedman

Over that period of time there were major changes in the law, the result of legislation by the assemblies of the people and edicts of the magistrates, along with interpretation, first by the pontiffs and then by jurists. By the late Republic, the ius honorarium, the law laid down by the magistrates, had substantially transformed the existing legal system. The process continued under the empire, supplemented by imperial constitutions. The pattern of development in civil procedure was one of an increasing role for the magistrate as judge, a reduced role for the parties.¹¹ Roman law as codified under Justinian, the form in which it was passed down to influence later legal systems, was, like modern law, a system in which state judgment and enforcement played a central role.

21.3. PRIVATE LAW: STRONG SENSE Originally, the term ius (plural, iura) denoted that which is due in human relations—the rightful power of a community member to act in a certain manner vis-à-vis his fellow citizens. It referred to a course of conduct that the community would take for granted and, in that sense, endorse. Thus, a person who appropriated an object, entered upon land, ejected or imprisoned another individual may in so doing be exercising ius. The community had a general awareness of the circumstances when such acts would be construed as iura and these were established by custom. . . . At this stage, the exercise of ius had no connection with state organization and thus ius was defined as any instance of approved selfhelp. (Mousourakis 2007: 19–20) To the legislators of the Twelve Tables, however, an interpretation of criminal law in which the state played a part was still entirely foreign. For them the natural and only result of a crime was the victim’s right of vengeance, and they were merely concerned to restrict the right of physical vengeance to the more serious crimes, to keep this right under judicial control, to isolate the wrongdoer who had been found guilty, and thus to protect the commonwealth from the effects of devastating family vendettas. (Kunkel 1973: 29)

As these passages by two scholars of Roman law suggest, a central characteristic of the legal process of the early Republic was its reliance ¹¹ du Plessis (2015: 79–83).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

331

on private action by the plaintiff.¹² It was his job to bring the defendant into court, if necessary by force.¹³ A debtor could bind himself by pledging his body as security; if he defaulted, his creditor could seize him without a prior court verdict for enslavement or execution.¹⁴ If the court verdict created a debt by the defendant to the plaintiff and the defendant failed to pay in thirty days, the plaintiff was entitled to seize him and, if the debt was not paid, eventually to sell him into slavery or execute him.¹⁵ As in the Icelandic system, the court authorized the use of force but it was, in most or all cases,

¹² Mousourakis (2007: 64) describes litigation as essentially private arbitration under the approval of the state, functioning as a substitute for self-help. For detailed discussion of the early history of Roman courts, see Kaser and Hackl (1996: 25–34). ¹³ Twelve Tables, Table I. “1. If the plaintiff summons the defendant to court the defendant shall go. If the defendant does not go the plaintiff shall call a witness thereto. Only then the plaintiff shall seize the defendant. 2. If the defendant attempts evasion or takes flight the plaintiff shall lay hand on him. . . . 3. If one of the parties does not appear the magistrate shall adjudge the case, after noon, in favor of the one present” (Johnson et. al. 1961). ¹⁴ “The Twelve tables contain a harsh form of debt-contract, in which the borrower, through the receipt of the money weighed out to him before witnesses, literally gave himself into the power of the creditor (whence the transaction’s name, nexum, ‘binding’). If he could not free himself punctually by repayment he fell into a condition of debt-bondage without the necessity of the judgement of a court” Kunkel (1973: 26). These options were abolished by the Lex Poetelia (326 BC). ¹⁵ Twelve Tables, Table III. Execution of Judgment: “1. Thirty days shall be allowed by law for payment of confessed debt and for settlement of matters adjudged in court. 2. After this time the creditor shall have the right of laying hand on the debtor. The creditor shall hale the debtor into court. 3. Unless the debtor discharges the debt adjudged or unless someone offers surety for him in court the creditor shall take the debtor with him. He shall bind him either with a thong or with fetters of not less than fifteen pounds in weight, or if he wishes he shall bind him with fetters of more than this weight. 4. If the debtor wishes he shall live on his own means. If he does not live on his own means the creditor who holds him in bonds shall give him a pound of grits daily. If he wishes he shall give him more. 5. . . . Meanwhile they shall have the right to compromise, and unless they make a compromise the debtors shall be held in bonds for sixty days. During these days they shall be brought to the praetor (possibly anachronistic for consul) into the meeting place on three successive market days, and the amount for which they have been judged liable shall be declared publicly. Moreover, on the third market day they shall suffer capital punishment or shall be delivered for sale abroad across the Tiber River” (Johnson et. al. 1961).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

332

David Friedman

private force. And, as in that system, there was a schedule of payments owed for injuries to the victim.¹⁶ All of these conclusions must be qualified by the limitations of the available evidence. The earliest written text of the law described in Roman tradition, the laws of the kings, the so-called ius Papirianum, ascribed to a Papirius, a putative contemporary of Tarquin the Proud, has not survived; some scholars doubt that it ever existed, although purported fragments exist in later literary sources.¹⁷ While the law of the Twelve Tables certainly existed and was for a thousand years regarded by the Romans as the foundation of their legal system, no complete text survives, leaving us with only incomplete reconstructions based on passages quoted in later texts.¹⁸ The information on the law of the Twelve Tables can, however, be supplemented by descriptions of legal process as it existed in the early Republic and by later enactments that, by abolishing rights of private action, demonstrated that they once existed. Thus, for instance, as Paul du Plessis (2015: 346) points out: “There had to be a dishonest and forcible taking of property; if the taker acted in good faith, he was not liable. This rule appears to have exonerated those who violently enforced genuinely held claims. Not surprisingly, Marcus Aurelius decreed that bona fide claims should be forfeited if enforced in a violent manner.”¹⁹ And a decree in AD 389 (C. 8.4.7) provided that an owner who forcibly seized his property should forfeit both ownership and possession, suggesting that self-help was still common enough to be seen as an issue at that late a date.²⁰

¹⁶ Gai., Inst. 3.223: “Under the Twelve Tables the penalty for this delict was, for a damaged limb, retaliation (talio); for a broken or bruised bone, on the other hand it was 300 ‘asses’ if a free man’s bone had been broken but 150 if it was a slave’s; for all other contempts (inuriae), on the other hand, the penalty established was twenty-five ‘asses’ ” (trans. Gordon and Robinson 1988). ¹⁷ Mousourakis (2007: 23). “The collection attributed to Papirius is probably an apocryphal publication of the end of the Republic, or of the time of Augustus” (Girard 1906: 32). On the so-called leges regiae, see Johnson et al. (1961: 3–6); Crawford et al. (1996: 561–3); and du Plessis (2015: 28–9). ¹⁸ Girard (1906: 51–2). Watson (1992: 14–20) provides an account of the sources for the reconstruction and an analysis of its reliability. ¹⁹ The context was the action for violent theft (actio vi bonorum raptorum). ²⁰ du Plessis (2015: 182). Gaius (Inst. 4:155) describes a similar result at an earlier date. “Sometimes, however, even if the person whom I have forcibly dispossessed was possessing from me by force, stealth or licence, I am compelled to restore possession to him, for instance, if I expelled him by force of arms. For by such gross wrongdoing

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

333

Litigation in early law was “a private arbitration established under the approval of the State, as a substitute for self-help, the business of the State officials being only to see that this arbitration is conducted in proper form” (Buckland and McNair 1952: 400). Mousourakis suggests that the early procedure for trial (legis actio) “probably derived from the practice established by custom where contested claims were voluntarily submitted to arbitration, and must have been in habitual use before its formal adoption.” In the form in which it was recorded, it consisted of two stages, the first before a consul or pontiff, the second “before a citizen appointed as the judge (iudex) by the magistrate and the parties concerned” (Mousourakis 2007: 32). In the earliest and most important of the legal actions (legis actio sacramento), each party deposited a sum of money as a wager that it was in the right. The prevailing party recovered its deposit, that of the losing party was forfeited to the state. The outcome of the trial demonstrated to other parties which claimant was in the right, providing the mechanism by which right made might in that society. At least one scholar interprets the procedure of each party swearing to the truth of his claim as a way of getting the priests, who were in charge of legal procedures, to provide a verdict, since a false oath was a religious offense.²¹ The law, as it later developed, distinguished between ius and lex. The former corresponded roughly to the underlying principles defining the rights of individuals towards each other.²² The later was legislation, specific rules created by an assembly to enforce ius. That I must in all circumstances submit to the action restoring possession to him” (trans. Gordon and Robinson 1988). ²¹ “The two parties took an oath of the truth of their pretensions, in such a way that there was necessarily on the one side or the other perjury,—a sin and, consequently, a delict, in this epoch when religion and law were not separated, and when, in order to know who had incurred the penalty, it was necessary for the king, head of religion as of criminal justice, to inquire who was right” Girard (1906: 28). And, on the same point in the context of the Twelve Tables, Girard (1906: 59) writes: “But we see by what roundabout means legal process is here grafted onto procedure under which one takes justice for oneself without legal process.” ²² “In this broad normative sense ius is not the same as morality nor as positive law; rather, it is right law, or positive law as it ought to exist in light of what morality and justice ordain. Ius, as defined above, was distinguished from lex (plural leges). The latter term signified a law created by a competent legislative organ of the state in conformance with a prescribed procedure. During the Republic the term lex was used to denote a statute enacted by a popular assembly and created in a form directed to all citizens” Mousourakis (2007: 20).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

334

David Friedman

suggests that ius may have represented the system of privately enforced customary law that preceded explicit legislation. Early Roman law also made a distinction between ius, private rights, and fas, divine rights. Where state intervention in English law was initially justified on the grounds that some acts that violated individual rights also violated the king’s peace, state enforcement in Roman law seems to have originated with enforcement of religious rules.²³ Exactly where the line was drawn between prosecution by magistrates and private action is unclear. “The distinction was not between capital and pecuniary cases: private actions, e.g. against a thief or a debtor, might give the prosecutor the right to kill or enslave a citizen, while on the other hand tribunes often prosecute criminally for a fine” (Jones 1972: 38). It sounds from Jones’s account as though public prosecution was primarily for offenses actually against the state—treason, military incompetence, peculation. The legal status of murder in the early law is unclear. By the time of the Twelve Tables, it was treated as a public rather than private offense; the penalty was not a monetary payment to the heirs of the victim, as in Icelandic and Somali law, but execution. There is, however, a passage sometimes ascribed to the Twelve Tables that suggests that at some point blood revenge was a legitimate response to a killing.²⁴ The rules in the Twelve Tables, at least as we have them, are not entirely consistent. One passage forbids putting anyone to death without a trial while another specifies that it is legal to kill someone who steals by night, provided that the killer raises a clamor to draw

²³ “Priestly officials were entrusted to enforce the religious duties as prescribed by the norms of fas. These officials independently discharged their task by devising and administering coercive practices” (Mousourakis 1007: 19). Note also that the earliest form of action involved an exchange of oaths and a wager upon them placed in the hands of the pontiffs. ²⁴ Table VIII, 24a (VIII 13 Crawford). “If a weapon has sped accidentally from one’s hand, rather than if one has aimed and hurled it, to atone for the deed a ram is substituted as a peace offering to prevent blood revenge.” Crawford et al. (1996: 693) discuss a tradition that this rule originated with the laws of Numa. So it may reflect a rule from a point before the punishment of killing shifted from privately to publicly enforced. Alternatively, the phrase “to prevent blood revenge,” present in the reconstruction by Johnson et al., absent in some others, might be a mistake and the ram could be interpreted as an offering to the gods. Jolowicz (1932: 178) suggests that the ram was “given to the agnates—a relic of the time when the agnates of the slain man were entitled to take vengeance on the slayer.”

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

335

attention to the fact.²⁵ Similarly, one passage permits retaliation in kind for a broken limb if compensation is not agreed on, while the next specifies a fixed money payment.²⁶ This suggests that, at the time of the composition of the Twelve Tables, the law was in the process of moving from a system of private vengeance with the possibility of private compensation to one of court authorized force with predetermined monetary payments. That is consistent with Mousourakis’s view that self-help was established in some royal decrees prior to the republic, with retaliation (talio) for some personal injuries.²⁷

21.4. PRIVATE ENFORCEMENT OF CONTRACTS AND JUDGMENTS: SURETIES, DISTRAINT, AND HONOR Our modern view of the surety and his functions is almost entirely based on the fideiussor of classical Roman law. But he is at once the product of a lengthy evolution and the symbol of a relatively advanced social and economic organization. Behind him, it is true, loom the dim figures of the uindex, the uas and the praes, but these survive in the works of the great jurisconsults only as the ghosts of a vanished order . . . . (Binchy 1970: 356)

Societies that lack a state apparatus willing and able to enforce contracts and legal judgments develop other mechanisms for the purpose, one of which is the institution of suretyship. Early Irish law provides examples which, Binchy argues, may represent the survival of institutions common to other early Indo-European legal systems, including the Roman. To what extent is that conjecture supported by what we know of suretyship in Roman law?

²⁵ IX 6. “For anyone whomsoever to be put to death without a trial and convicted . . . is forbidden.” VIII 12. “If a thief commits a theft by night, if the owner kills the thief, the thief shall be killed lawfully” (Johnson et al. 1961). One might interpret this as distinguishing between execution by a magistrate and legal killing by a private party. ²⁶ VIII 2: “If anyone has broken another’s limb there shall be retaliation in kind unless he compounds for compensation with him.” VIII 3: “. . . If a person breaks a bone of a freeman with hand or by club, he shall undergo a penalty of 300 asses; or of 150 asses, if of a slave . . .” (Johnson et al. 1961). ²⁷ Cf. du Plessis (2015: 29).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

336

David Friedman

The Irish system recognized three sorts of sureties created in the process of forming a contract.²⁸ The naidm, or enforcing surety, had the job of guaranteeing that the party for whom he was a surety would fulfill his obligations and was entitled to use force to make him do so. The rath, or paying surety, was obligated to pay what his party owed if the party failed to do so, and he was also entitled to collect from his party what he had paid plus an additional sum. The aitire, or hostage surety, guaranteed his party’s obligations not with his wealth but with his body. If the party defaulted, the surety was obliged to surrender himself to the claimant, eventually ransoming himself back—and, again, obtaining a claim against his party for considerably more than the sum originally owed. The fideiussor in Roman law was to some degree the equivalent of the Irish rath. He was obligated to pay what the party for whom he was a surety owed if the party himself failed to do so. There might be multiple sureties for a single contract. In the initial version of the institution, the creditor could claim the entire debt from any of them. A rule created under Hadrian allowed a surety to require a creditor to divide his claim against any surety by proportion in the total number of sureties. Redress of sureties among themselves was only firmly fixed by Justinian in AD 535.²⁹ The praes seems to have been an earlier form of the fideiussor, combining the roles of rath and aitire, guaranteeing a debt with both his estate and his person.³⁰ Another early form of surety was the vas, who pledged to pay a fixed amount if the third party he was guaranteeing failed to fulfill an obligation such as paying a debt or appearing before a court.³¹ ²⁸ For details, see Kelly (2009); Binchy (1970). ²⁹ For earlier redress among sponsores and fidepromissores (but not fideiussores) see Gai., Inst. 3.122 (against other sureties) and 3.127 (against the creditor). ³⁰ “To guarantee performance of a duty by nominating praedes was the original form the Romans had used in archaic times. It meant that if the duty was not duly fulfilled by the contractor himself, the surety was bound in a personal way: he was obligated through nexum or mancipium to the creditor. For the later period that concerns us, this form of surety had been abandoned in private law and was replaced by the principle that the surety was a second debtor who guaranteed the debt with his means (i.e. his entire estate) but not with his person. The surety was no longer called praes but was referred to as sponsor or fideipromissor or fideiussor, in accordance with the formal legal act that established this special relationship between the creditor and the person willing to guarantee for a debtor” (Rainier 2013: 178). ³¹ “In the oldest times the ‘vadimonium’ served the purposes of a suretyship. Vadimonium was a solemn promise to pay a specified penalty, if a certain third party failed to meet a particular obligation (e.g. to pay a debt or answer a summons

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

337

The vindex in early Roman law plays a different role from any of the Irish sureties. The vindex could intervene at the point when a debtor was about to be seized and led off in bonds by his creditor, taking the obligation on himself and apparently freeing the debtor from it. It was then up to him to show that the debt was not owed. If he failed, he was liable for twice the original amount.³² While the Roman forms of suretyship differ from the Irish, their existence is consistent with Binchy’s suggestion that institutions for enforcement in a decentralized system that survived in Irish law late enough to be recorded in some detail may date back to a much earlier point in the history of Indo-European institutions. Another institution of which the same may be true is distraint. Irish law provided elaborate mechanisms permitting a claimant to seize cattle in satisfaction of his claim by a series of stages, taken in presence of witnesses, which the defendant could interrupt by agreeing to arbitration of the claim.³³ The Roman equivalent, predating the Twelve Tables, was the action-at-law of pignoris capio.³⁴ It is unclear what the plaintiff could do with the property seized, other than holding it in order to force the defendant to perform his obligation. Nor is it clear in what sorts of cases it was permitted. But it is a self-help remedy of the sort needed in a legal system where state enforcement is absent or weak, one that may, as Binchy has suggested, have been more similar to the Irish distraint in the period before the Twelve Tables.³⁵ before court). A ‘vas’ is not a surety in our sense of the word, because the liability he undertakes is not the same as that of the principal, but is a new liability with different contents . . .” Sohm (1892: 8 n. 2). ³² “In the process described by Gaius and Gellius the intervention of the vindex apparently set the debtor free, and all liability was assumed by the vindex. This is clearly a later stage in an institution whose origin has been the subject of considerable speculation” (Schiller 1910: 206; Gai., Inst. 4.21, 25). ³³ Kelly (2009: 177–82). ³⁴ du Plessis (2015: 71–2). ³⁵ “Further, it seems to me reasonably certain that the legis actio in the older Roman law called pignoris capio is also a vestigial relic of this archaic right of selfhelp, even though by the time of Gaius it had become a rigidly circumscribed and conventionalized exception. Indeed, this was the view held by most (if not all) previous legal historians, though some modern Romanists reject it. But have not these scholars lost sight of the fact that in the earliest stages of Roman society, well before the Twelve Tables, when Rome was still, to use Ludwig Wenger’s word, a Stammstaat, a remedy of this kind must have existed and that the old term for it was retained in classical Roman law (as happened to many such terms in other legal systems) for a much more specialized procedure?” (Binchy 1973: 24).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

338

David Friedman

In Irish law every individual had an honor price determined by his status, affecting both the amount owed for offenses against him and the upper limit of the amount he could commit himself to in a contract or as a surety. One consequence of some offenses was the loss of honor price. The nearest Roman equivalent was the reduction in status (but not in limits on suretyship) that made someone an infamis or intestabilis or both as a result of some sort of discreditable conduct. As du Plessis (2015: 107–9) describes, infames could not “hold offices or positions of honor, could not vote (at least in early law), or bring criminal accusations, or appear as advocates, or act as representatives (or be represented) in litigation.” Intestabiles “could not act as witnesses, either in litigation or in formal transactions (such as the making of wills or the conveyance of property).” One further feature of early Roman legal procedure is how formalistic it was—a plaintiff could lose his case for putting it in the wrong words.³⁶ The same seems to have been true of Icelandic law. In one passage in Njal’s Saga, a step in a legal procedure requires a formula to be recited three times. Two of the three recitations may be imprecise, but one must be verbatim.³⁷ Such rigid rules may make sense in a decentralized system of privately enforced law; it is important that third parties can tell who is legally in the right, and in the Icelandic context there is no superior authority to decide what variations in the rules are or are not legitimate. Given the limitations in our source materials, it is hard to be certain of the nature of Roman law in the monarchy and early Republic, but such evidence as we have is consistent with the view that, like other early forms of law, it originated as a privately enforced feud system.

³⁶ Mousourakis suggests that the formality of the law reflected the religious origin and character of many legal rules. An example is a plaintiff who charged the defendant with cutting his vines when the legal formalism required him to describe them as trees in order to fit the wording of the legal rule in the Twelve Tables (Mousourakis 2007: 29). Gai., Inst. 4.11: “This is why the opinion was given that a man who raised an action over the cutting down of vines in a way that used the word ‘vines’ in the action had lost his case. He ought to have used the word ‘trees,’ because the Twelve Tables, under which the action for cutting down vines was available, spoke in general terms about cutting down trees” (trans. Gordon and Robinson 1988). ³⁷ Magnusson and Palsson (1960: 76–8).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

339

21.5. PRIVATE LAW: WEAK SENSE The system of procedure, or lack of a system, constituted a major disincentive to involvement in litigation, especially in Republican Rome, since the State’s involvement in civil procedure was minimal (in both the legis actio and formulary procedure). This was because of the notion that litigation was essentially “a private arbitration established under the approval of the State, as a substitute for self-help, the business of the State officials being only to see that this arbitration is conducted in proper form.” (Buckland and McNair 1952: 400)³⁸

So far I have been offering evidence that Roman law developed from a legal system that was private in the strong sense of a feud system, privately prosecuted and enforced. To what extent, especially in the later and better-documented period, was there private law in the weaker sense, law that, like modern tort and contract law, is privately prosecuted but publicly enforced? The answer is that, through the republican period and into the early Empire, Roman civil law was private in a sense intermediate between the two. Under both the earlier legis actio and the later formulary system, both getting the defendant to court and executing the judgment were the responsibility of the plaintiff, not the state. Under the earlier system, if the defendant, once summoned, failed to either come to court with the plaintiff or find a guarantor (vindex) for his later appearance, the plaintiff was entitled to call witnesses and then drag the defendant to court by force. If he was unable to do so and the defendant failed to appear, the defendant was considered indefensus, “a status that entitled the plaintiff to seize his property on the authorization of the praetor.”³⁹ Under the later formulary system, the defendant “could make a promise (vadimonium) to appear on a particular day, in which case he would promise to provide security and to pay a penalty for failure to appear at the agreed time in the proximity of the court. . . . Sanctions were imposed on an indefensus— a defendant—who tried to avoid being summoned (e.g. by hiding) or who refused to obey the summons or to provide a vindex. “For example, the praetor could allow the plaintiff the possession of the defendant’s estate, with a possible right of sale. The right to seize a

³⁸ du Plessis (2015: 64).

³⁹ du Plessis (2015: 66).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

340

David Friedman

reluctant defendant and drag him to court still survived but was probably used only as a last resort” (du Plessis 2015: 72–3). In modern legal systems, we take for granted that judgments will be effectively enforced (“executed”) with the full authority and backing of the State. But for much of Rome’s history, the state took only an indirect role in the execution of judgments. It authorized the successful plaintiff to pressurize the defendant into complying with the judgment—a form of regulated self-help, the onus being firmly on the plaintiff to obtain satisfaction. (du Plessis 2015: 70)

Under the legis actio, if the case was in rem, for possession of a thing, and the winner was already in possession he remained so. If the loser was in possession but had granted temporary security at an earlier stage in order to remain so, the security was forfeited to the plaintiff. “If the security was unforthcoming or insufficient, further proceedings could be brought to assess the compensation payable . . . . When the amount had been assessed, execution of the judgment proceeded along the lines of a case in personam . . .” (du Plessis 2015: 70). If the case was in personam, for enforcing an obligation owed by the defendant to the plaintiff, the successful plaintiff in the early period could enforce the verdict either by manus iniectio, physical seizure of the defendant, or by pignoris capio, seizure of his property, the latter option being available only for certain categories of cases. In the former case, the defendant had thirty days in which to pay the judgment, after which the judgment creditor could seize him, take him before a magistrate, and if he neither paid nor produced a vindex to take his place imprison him for sixty days. If nobody came forward to pay the debt, at the end of that time the judgment debtor could be sold into foreign slavery or put to death. A rule instituted in 326 BC eliminated those alternatives. “Instead, the creditor could keep the debtor for as long beyond the 60 days’ limit as it took for the latter to work off the debt” (du Plessis 2015: 71). Under the later formulary system, the creditor was eventually given the additional option of seizing and selling property of the debtor in satisfaction of the debt.⁴⁰ The formulary system was introduced in the third century BC, initially for cases involving foreigners, and made available more generally in the next century. The earlier legis actiones system “fell largely into disuse in the late Republic and was formally abolished by ⁴⁰ du Plessis (2015: 77–8).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

341

Augustus in 17 BC . . . ” (du Plessis 2015: 72). It was in turn supplemented in the imperial period by the cognition system and formally abolished in AD 342. That change reflected the conversion of civil law from a system privately prosecuted and to a large extent privately enforced to something more like modern civil law, with the responsibility for enforcement shifting from the plaintiff to the state. Under the type of procedure that depended on an “investigation” (cognitio) by a single magistrate, the plaintiff lodged a statement of his claim with the magistrate and the magistrate ordered the defendant to undertake to appear in court and provide security that he would do so. If not, he could be arrested by the magistrate. If the defendant lost the case and failed to pay up, his property could be seized by court bailiffs and sold. Roman law took its final and most complete form under Justinian in the Digest, a lengthy account of the law in the form of selected passages from the writings of earlier authorities, the Codex, a compilation of imperial enactments, and the Institutes, a condensed version intended for students. One of the first things that strikes a modern reader is how much of it is privately prosecuted. Theft, for example, is prosecuted by the victim, either the owner of the stolen property or someone else injured by its theft. The convicted thief owes damages of twice or four times the value of what has been stolen, the larger amount if caught in the act.⁴¹ The law distinguishes what we would describe as compensatory damages from the additional punitive sum. The former is an obligation that is inherited by the thief ’s heirs, the latter is not. The distinction between private and public prosecutions, in Roman law as in Athenian law,⁴² is that public prosecutions “are called public because as a general rule any citizen may come forward as prosecutor in them” (Just., Inst. 4.18.1). Finally, it should be observed that a person who has been outraged always has his option between the civil remedy and a criminal indictment. If he prefers the former, the penalty which is imposed depends, as we have said, on the plaintiff’s own estimate of the wrong he has suffered; if the latter, it is the judge’s duty to inflict an extraordinary penalty on the offender. (Just., Inst. 4.4.10) ⁴¹ Just., Inst. 4.1; Gai., Inst. 3.189–90. The thief is also liable to return what was stolen by an action in rem or its value by an action in personam. ⁴² MacDowell (1978: 57–8).

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

342

David Friedman

The Digest distinguishes between civil and criminal law, but most criminal law, as in eighteenth-century England,⁴³ appears to have been privately prosecuted.⁴⁴ That raises, for a modern reader, the question of what distinguishes the two.⁴⁵ The answer is never made explicit in the texts. At least part appears to be that criminal conviction results in punishment of the criminal but not in a damage payment or transfer of property to the private prosecutor. Another difference is that in a criminal case the accuser was not free to drop the case, as legislated in the Senatus Consultum Turpillianum of AD 61.⁴⁶ Conceptually, from the beginning of the legal system, criminal law had to do with offenses against the gods or the polity, civil law with disputes among individuals. As in a modern legal system, the same act might sometimes be prosecuted as either a civil or a criminal wrong, as shown by the passage quoted above. According to a constitution of the emperors Valens, Gratian, and Valentinian, “whenever both a private law action and a criminal proceeding lie concerning property, both may be brought, without reference to which one is brought first . . . ” (C. 9.31.1 pr.; Trier, AD 378). 21.6. WHY? Over a period of about a thousand years, Roman law developed from a legal system in which most claims were privately prosecuted and ⁴³ Friedman (1995); Friedman et al. (2019: 223–58). ⁴⁴ This raises the interesting question of what the incentive was for a private party to accuse another of a criminal offense. In Periclean Athens, the successful private prosecutor typically received a sizable share of the fine paid by the convicted defendant, but there seems to have been no equivalent mechanism in Roman law except perhaps the delatores who notify the imperial Fiscus of certain events (D. 49.14.1 pr.; Callistr. 1 de iure fisci) in exchange for a reward (lex Papia Poppaea, see Tacitus, Annals 3.25 and 3.28). Where the accuser was a victim of the crime, the obvious incentive would be revenge—or, to put it in something closer to the language of economics, a commitment strategy of prosecuting those who committed crimes against one in order to produce deterrence as a private good. Another incentive, as in the Athenian case, might be to injure an enemy. And one cannot dismiss the possibility that at least some Roman citizens would regard enforcement of the criminal law as a civic obligation. ⁴⁵ du Plessis (2015: 325–6) discusses the civil/criminal distinction and the intermediate status of the law of delicts. ⁴⁶ See D. 48.16 and C. 9.45. In eighteenth-century English law, it was an offense for a private prosecutor to compound a felony—accept a payment to drop charges.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

343

privately enforced to something more like a modern legal system, although with a larger role for the victim as prosecutor than in modern criminal law. A similar pattern can be observed in English common law.⁴⁷ That raises the question of why, in both cases, the change occurred. Broadly speaking, there are three answers. One is that people discovered that state enforcement of law was superior to private enforcement—the Whig theory of history broadly defined.⁴⁸ One is that circumstances changed so that, while private enforcement might have been the best viable option in the early period, by the later period it no longer was, and the legal system changed accordingly. Both of those explanation depend on some mechanism to push legal institutions towards optimality. The second also requires an explanation of what changed in the environment of the legal system and why those changes implied a change in how that system ought to function. An example of such an explanation in the English case would be the shift from eighteenth-century private prosecution of crime to nineteenthcentury public prosecution, arguably due to an increasing urbanization that reduced the victim’s reputational incentive to prosecute. The third answer is that the change occurred not because it was an improvement but because making it was in the interest of those controlling the legal system. The clearest example of such a change for which that is at least a plausible conjecture is the shift from private to public enforcement, from what was in effect a tort system to a criminal system, in early English law. Under Anglo-Saxon law, as under the contemporary and better recorded Icelandic system, the chief legal effect of what we would consider a crime was to establish a claim for damages by the victim or his heirs against the offender. Over time, under late Anglo-Saxon and then Norman law, that was converted to a claim by the crown for damages, forfeiture of property, in capital cases a fine paid in exchange for a pardon. The result was that, ⁴⁷ Actually, it can be observed twice, once in the transition from early Anglo-Saxon to medieval Norman, a second time in the transition from privately prosecuted (but publicly enforced) criminal law in the eighteenth century to publicly prosecuted law by the latter half of the nineteenth century. ⁴⁸ Narrowly defined, the Whig theory described in Butterfield 1930 is the view of English history propounded by nineteenth-century historians, such as Macauley and Stubbs, that interpreted past historical events in terms of their role in producing the desirable institutions of the present. Broadly defined, it is the view of history as trending towards a more nearly optimal set of institutions.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

344

David Friedman

in medieval England, the right to run courts and collect fines was viewed as a valuable part of the bundle of rights held by a feudal lord, sometimes purchased from the Crown.⁴⁹ I do not know enough about the workings of the Roman legal system as it developed to say if something similar is a plausible explanation in that case. Perhaps someone who does can.

21.7. CONCLUSI ON Roman law probably originated in the context of a decentralized system of self-help, a feud system where enforcement was private and the court system provided a mechanism for providing an authoritative verdict on private disputes. In the earliest form of which we have a good account, the private role both in bringing the defendant to court and in enforcing the verdict made Roman civil law more private than modern tort law. As the system developed, under first the Republic and then the Empire, the state took an increasingly active role. But even in its final form as codified by Justinian, prosecution of both civil and criminal offenses was primarily the responsibility of private citizens.⁵⁰

REFERENCES Binchy, D. A. 1970. “Celtic suretyship: a fossilized Indo-European institution?,” in George Cardona, Henry M. Hoenigswald, and Alfred Senn, eds, Indo European and Indo-Europeans: Papers Presented at the Third IndoEuropean Conference at the University of Pennsylvania. Philadelphia: University of Pennsylvania Press, 355–67. Binchy, D. A. 1973. “Distraint in Irish law.” 10 Celtica 22–71. Buckland, William Warwick. 1957. A Manual of Roman Private Law, 2nd edn. Cambridge: Cambridge University Press. ⁴⁹ I discuss some of these issues in Friedman (1996). For examples of revenue from law enforcement under the Angevins, see Warren (1978): 177. ⁵⁰ I would like to thank Egbert Koops for extensive advice and the correction of some of my errors on Roman law and Dennis Kehoe for updating my references. Also Giuseppe Dari-Mattiacci for giving me a reason to learn some, if not enough, about the subject.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

Private Prosecution and Enforcement in Roman Law

345

Buckland, William Warwick, and McNair, Arnold D. 1952. Roman Law & Common Law A Comparison in Outline, 2nd edn. Cambridge: Cambridge University Press. Crawford, Michael H. 1996. Roman Statutes, vol. II. London: Institute of Classical Studies. du Plessis, Paul. 2015. Borkowski’s Textbook on Roman Law, 5th edn. Oxford: Oxford University Press. Friedman, David. 1979. “Private creation and enforcement of law: a historical case.” 8 Journal of Legal Studies 399–415. Friedman, David. 1995. “Making sense of English law enforcement in the eighteenth century.” 2 The University of Chicago Law School Roundtable 477–85. Friedman, David. 1996. “Beyond the tort/crime distinction.” 76 Boston University Law Review 103–12. Friedman, David, Peter Leeson, and David Skarbek. 2019. Legal Systems Very Different from Ours. Self-published. Girard, Paul Frédéric. 1906. A Short History of Roman Law, trans. Augustus Henry Frazer Lefroy and John Home Cameron. Toronto: Canada Law Book, Co. Hallaq, Wael B. 2009. Sharia: Theory, Practice, Transformations. Cambridge: Cambridge University Press. Johnson, Allan Chester, Paul Robinson Coleman-Norton, and Frank Card Bourne. 1961. Ancient Roman Statutes: A Translation, Commentary, Glossary, and Index. Austin: University of Texas Press. Jolowicz, H. F. 1932. Historical Introduction to the Study of Roman Law. Cambridge: Cambridge University Press. Jones, A. H. M. 1972. The Criminal Courts of the Roman Republic and Principate. Totowa, NJ: Rowman and Littlefield. Kaser, Max, and Karl Hackl. 1996. Das römische Zivilproxeßrecht, 2nd edn. Munich: Beck/Hackl. Kelly, Fergus. 2009. A Guide to Early Irish Law. Dublin: Institute for Advanced Studies. Klein, Hyman, trans. 1954. The Book of Torts, Book Eleven of the Code of Maimonides (Mishnah Torah). New Haven and London: Yale University Press. Kunkel, Wolfgang. 1973. An Introduction to Roman Legal and Constitutional History, 2nd edn. based on the 6th German edn. of Romische Rechtsgeschichte, trans. J. M. Kelly. Oxford: Clarendon Press. Lewis, I. M. 1961. A Pastoral Democracy: A Study of Pastoralism and Politics among the Northern Somali of the Horn of Africa. Oxford: Oxford University Press for the International Africa Institute. MacDowell, Douglas M. 1978. The Law in Classical Athens. Ithaca: Cornell University Press.

OUP CORRECTED PROOF – FINAL, 22/4/2020, SPi

346

David Friedman

Magnusson, Magnus, and Hermann Palsson, trans. 1960. Njal’s Saga. London: Penguin. Mousourakis, George. 2007. A Legal History of Rome. London and New York: Routledge. Neusner, Jacob. 1988. The Mishnah: A New Translation. New Haven and London: Yale University Press. Pound, Roscoe. 1906. Readings in Roman Law. Lincoln: Jacob North & Co. Prichard, A. M. 1961. League’s Roman Private Law, 3rd edn. London: Macmillan. Rabinowitz, Jacob J., trans. 1949. The Code of Maimonides Book Thirteen: The Book of Civil Laws. New Haven: Yale University Press. Rainer, J. Michael. 2013. “Public building contracts in the Roman Republic,” in Thomas A. J. McGinn, ed., Obligations in Roman Law: Past, Present, and Future. Ann Arbor: University of Michigan of Michigan Press, 174–88. Schiller, A. Arthur. 1978. Roman Law: Mechanisms of Development. Berlin: Mouton De Gruyter. Seebohm, Frederic. 1902. Tribal Custom in Anglo-Saxon Law. London: Longmans, Green, and Co. Sohm, Rudolf. 1892. The Institutes of Roman Law. Oxford: Clarendon Press. Warren, W. L. 1978. King John. Berkeley: University of California Press. Watson, Alan. 1992. The State, Law and Religion: Pagan Rome. Athens and London: University of Georgia Press. Weyrauch, Walter O., ed. 2001. Gypsy Law: Romani Legal Traditions and Culture. Berkeley: University of California Press. Ancient References Moyle, J. B., trans. 1913. The Institutes of Justinian. Oxford: Clarendon Press. http://gutenberg.readingroo.ms/5/9/8/5983/5983-h/5983-h.htm. Frier, Bruce W., ed. 2016. The Codex of Justinian: A New Annotated Translation with Parallel Latin and Greek Text, 3 vols. Cambridge: Cambridge University Press. Gordon, W. M., and O. F. Robinson. 1988. The Institutes of Gaius. Ithaca: Cornell University Press. Johnson, Allan Chester, Paul Robinson Coleman-Norton, and Frank Card Bourne. 1961. Ancient Roman Statutes: Translation, with Introduction, Commentary, Glossary, and Index, gen. ed. Clyde Pharr. Austin: University of Texas Press. http://avalon.law.yale.edu/ancient/twelve_tables.asp.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

22 Deterrence of Wrongdoing in Ancient Law Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli

Deterrence is among the principal objectives of modern legal systems in a variety of disparate subject areas. For example, the criminal law seeks to deter individuals from undertaking certain “wrongs”; tort law seeks to deter individuals from exercising less than due care when engaging in risky activities; environmental law seeks to deter pollution of the environment, and contract law seeks to deter individuals from reneging on promises. Indeed, broadly interpreted, the deterrence objective may be transposed to nearly every realm of human activity in which the law manipulates incentives to influence behavior. So fundamental is the concept of deterrence to the law, it may easily be imagined that the very first legal institutions were created with this precise purpose in mind. Yet framing the function of law in terms of deterrence has not historically been regarded as intuitively obvious. Even in criminal law, arguably the most natural application of the deterrence rationale, legal scholars are still today in disagreement about whether deterrence matters,¹ or indeed whether

¹ See e.g. Farrington et al. (1994); Langan and Farrington (1998); Hirsch et al. (1999); and Gendreau et al. (1999) (finding that increasing prison sentences actually correlated with a 3 percent increase in recidivism). See generally Tonry (2008) and Webster and Doob (2012) (surveying the empirical research contradicting deterrence theories’ predictions). Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli, Deterrence of Wrongdoing in Ancient Law In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0022

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

348 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli deterrence should matter.² Thus, it is unlikely that protean legal institutions consciously embraced deterrence as a policy objective. Nevertheless, we find that in many cases modern legal systems have evolved not only to effect deterrence, but indeed to incentivize efficient levels of deterrence; this raises the question how the law evolved to so precisely achieve the deterrence objective in such a wide variety of contexts without necessarily recognizing it explicitly. In this chapter, we explore the evolution of deterrence in ancient law. We shall offer a combination of historical, comparative, and economic analyses in identifying the salient forces at work in the emergence, development, and refinement of the deterrence objective in early legal systems. We begin in §22.1 with a description of the evolutionary and cultural origins of “wrongdoing” and the desire for revenge in economic terms. We trace the historical development from these inherited biological traits to the intertribal norm of lex talionis. In §22.2, we analyze the deterrent effects of the principles of lex talionis. In §22.3, we describe the economic pressures that effected a shift from communal liability to individual liability. In §22.4, we describe the evolution of compensatory damages from the principle of lex talionis, which would later develop into the modern law of torts. §22.5 traces a parallel branch of legal evolution, describing the emergence of criminal law from lex talionis. We conclude with some general remarks on the role of the deterrence objective in the formation and maturation of law.

22.1. “WRONGDOING” AND RETRIBUTION Let us postpone defining what “wrongs” are and provisionally assume the naïve view that wrongs are offenses under some exogenously determined standard. Instead, let us focus our attention on the effect that wronged individuals tend to desire vengeance. We begin our analysis by asking why victims of wrongs seek to “repay” the disutility they have suffered by causing disutility to their injurers. One plausible hypothesis is that victims of wrongs experience positive utility from vengeance, and that “revenge” is simply the product of private utility maximization.³ The utility they get from vengeance outweighs the ² See e.g. Lewis (1953); Kant (1965); Duff (1986); and Hirsch (1993). ³ Posner (1980: 27) seems to suggest something like this.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

349

cost of undertaking revenge activities. Recent psychological research seems to indicate that this explanation is dubious, finding that individuals tend not to experience greater happiness after exacting revenge.⁴ However, the hypothesis may be partially recovered by distinguishing between the pursuit of revenge and its result: we may plausibly suppose that the act of exacting revenge itself provides utility, though the end state of having achieved revenge may not. The fact that individuals tend to seek revenge seems to reveal that they perceive the pursuit of revenge at least to be a desirable activity. To explain this preference, we find the argument from evolutionary psychology persuasive: that individuals have a biologically inherited predisposition toward retaliatory behavior, manifest by the sense that offending parties deserve retaliation for perceived offenses. Naturally, knowing that victims will have a predisposition for revenge creates an expected cost for would-be injurers, which deters “wrongful” conduct, thereby reducing the incidence of wrongs and improving the welfare of would-be victims.⁵ Parisi and Fon (2005) show that this behavior can lead to equilibria with low rates of “wrongful” conduct.⁶ When an individual from one tribe harms an individual from another tribe, either because the act of harm itself creates utility for the injurer, or because injury results as the byproduct of some other activity that creates utility for the injurer, an externality arises. If members of the victim’s tribe desire revenge as a consequence of the injury, then their utility-maximizing choice will often be to retaliate, causing harm to the injurer, reducing the benefit extracted from the activity. We may describe the utility of the injurer by the function: UI ðwÞ ¼ BðwÞ  C  πρðκÞ;

ð1Þ

where B(w) is the benefit of the harmful activity, w is the activity level, C is the opportunity cost, π is the probability of revenge being ⁴ See e.g. Carlsmith et al. (2008), finding that successful revenge tends to generate disutility for the avenged. ⁵ It is relatively straightforward to understand how such a phenotype would be selected. See e.g. Jacoby (1983); Chagnon (1988). ⁶ We think the intuition underlying this analysis to be reasonably straightforward. Readers interested in the details of the formal analysis should consult Parisi and Fon (2005). See also Posner (1988: 27).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

350 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli triggered, and ρðκÞ is the magnitude of revenge, given the revenger’s investment in revenge κ. Clearly, when πρ > B  C, the rational prospective injurer will choose to engage in some other activity. That is, he will be deterred from the harmful activity. We may describe the utility of the victim by the function: UV ðκÞ ¼ HðwÞ  κ;

ð2Þ

where H(w) is the magnitude of harm, and κ is the cost of exacting revenge. Let us assume that revenge is costly, but that it creates a @ρ greater cost for the injurer than the revenger. That is, @κ >1. However, assuming that the game is played sequentially, such that injurers move first and victims respond, it seems that the rational choice for injurers is not to seek revenge. That is, maxκ UV ¼ 0. This result is true for one-shot interactions, but does not hold for repeat interactions, where an investment in κ >> 0 may leads to a reduction in equilibrium of w over multiple rounds of play, resulting in better long-run payoffs than a strategy of κ ¼ 0. It is plausible that our prehistoric ancestors were incapable of making such conclusions. The absence of abstract language in early humans or limited cognitive faculties in pre-human hominids may have rendered the information cost of rational deliberation in repeated games prohibitive. Yet if we modified equation 2 so that @r the act of exacting revenge generated utility rðκÞ, such that @κ >1 for an interval ð0; nÞ, then some level of revenge-seeking would maximize the function:⁷ UIR ðκÞ ¼ HðwÞ  κ þ rðκÞ:

ð3Þ

Such behavior would improve the long-term prospects of tribes, facing repeat interaction with neighboring tribes. It is therefore plausible that such a revenge-preferring phenotype would tend to succeed through a process of natural selection.⁸

⁷ The literature on “negative reciprocity” is very well developed. For more detailed discussions of this topic, see e.g. Güth et al. (1982); Camerer and Thaler (1995); Roth (1995); Falk and Fischbacher (2006); and Falk et al. (2008). ⁸ Parens (2014: 170).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

351

We may observe this protean form of “retributive justice” manifested in a variety of early societies.⁹ For example, the informal intertribal law of Native American communities consisted, in many cases, exclusively of the “law of revenge.”¹⁰ Biblical accounts also describe revenge-motivated “justice” in early Judeo-Christian society.¹¹ The honor value of “blood revenge” also figured prominently in norms of ancient Athens,¹² Arab society,¹³ as well as in premodern Japanese society.¹⁴ Indeed, nearly all legal systems seem to have sprouted from the seed of satisfying the pre-legal desire for vengeance.¹⁵ Nearly the same happened in Roman law, which later developed a set of rules that has served as a model source of law in several civil law countries for many centuries after the fall of the Roman Empire, until the age of the modern codifications.¹⁶ At the beginning of the Roman legal experience the victim had the right to revenge, although retaliatory justice was administered under the control of the community.¹⁷ This begs the inquiry: how exactly did the desire for vengeance mature into a social norm and later into a basis for law? In addition to the innate desire for revenge identified by evolutionary psychology,¹⁸ we observe notions of desert and retribution taking shape in early societies, reinforced intra-tribally through “duty” and “honor” norms, often bolstered by religious rules.

⁹ Though for present purposes, we may describe the desire for revenge as retributive, it bears distinguishing revenge from retribution. Retributivists in the philosophy of punishment nearly unanimously distinguish retribution from revenge. See Moore (1997: 89). We do not wish to err in conflating revenge and retribution. Therefore, we intend special emphasis to be understood in our use of the qualifier “protean” here. Though retribution is not conceptually identical with revenge-seeking, we think notions of retributivism in moral philosophy arose from the pre-philosophical desire for revenge as a historical matter. ¹⁰ See e.g. Reid (1970); Hudson (1976); Blackburn (1980); Chagnon (1988). ¹¹ e.g. Genesis 4:10–24; Proverbs 6:13; II Samuel 12:13–18. ¹² Allen (2001) (“Anger was thus assumed to be not only the source of particular punishments but also at the root of law itself. The Athenians accordingly felt relatively little uncertainty or unease about why (that is, in response to what causes) they punished: they acted in response to anger”). ¹³ Ginat (1997). ¹⁴ Mills (1976). ¹⁵ An interesting exception may be observed in Indian culture, where incapacitative objectives seem to have been primary in the development of criminal and tort law. See Das Gupta (1930); Doongaji (1986); Olivelle (2011). ¹⁶ See Manthe (2000: 55–6). ¹⁷ Jhering (1906: 118). ¹⁸ See Joyce (2006).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

352 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli Though the evolutionary basis for revenge-seeking is clear from the game-theoretic analysis, this alone is not sufficient to explain why communities embraced it as a desirable practice. It does not follow that activities which are explainable through biological evolution are necessarily reified as social norms. To understand a community’s motivation for adopting a social norm of revenge-seeking, we note that wrongs were not regarded, insofar as “justice” was concerned, as perpetrated by or against individuals. Instead, the family, clan, or tribe was regarded as being collectively responsible for wrongs committed by members and for the retaliation of wrongs committed by other tribes.¹⁹ Tribe members who failed to exact vengeance for a wrong were regarded by their community as dishonorable and faced intra-tribal social sanctions. Suppose that a member of a revenge-seeking tribe had the utility function described in equation 2. That individual could free-ride on the revenge efforts of his fellows, enjoying the benefit of the deterrence that they generated by investing in revenge, while expending less or no κ himself. This situation is analogous to a cartel game. For any given individual, the benefit of defection will tend to generate greater benefits (i.e. the effort not dissipated by participating in communal revenge) than the loss due to decreased deterrence (i.e. the reduction in protection from deterrence). By not participating in revenge-seeking, such an individual could improve his utility at a cost to the tribe. Thus, the social norm of revenge-seeking reinforces the utility derived from revenge (for individuals who possessed idiosyncratically weak revenge-seeking preferences), while reducing the utility of free-riding (by imposing intra-tribal social sanctions). This process of social reinforcement begins the transformation from biologically motivated preference into social norm. Yet the road to legal rule is only just begun. Historically, the bare retaliation norm tended to be a short-lived stage of development.²⁰ Bare revenge, though it solves the repeated game problem, is a potentially ¹⁹ See Posner (1980) for a detailed description of the communal unit in archaic societies and an economic explanation for its development. For the role of communities in specific archaic societies, see also Reid (1970); Hudson (1976); Mills (1976); Blackburn (1980); Chagnon (1988). ²⁰ At this stage of development, it is dubious whether retaliation is even a social norm—we think it something intermediate between an evolutionary instinct and a social norm.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

353

catastrophic norm. When a multiplier is attached to retaliation (i.e. when ρ ¼ aH, where a > 1), a revenge act may be regarded as a fresh offense by the original injurer, leading to spirals of escalating revenge. Initially, a “kind-for-kind” restriction arose to constrain the type of retaliation warranted by the harm suffered. The kind-for-kind constraint restricts the scope of available retaliatory options to like harms. For instance, theft warranted revenge-theft, and killing warranted revenge-killing. The kind-for-kind constraint is later reinforced by the notions of “desert” and “talis” (equal in kind) captured by the etymology of the word “talio” (retaliation) in ancient Roman law. These social norms helped to tame the useful but potentially volatile revenge impulse. Yet, though early norms constrained retaliation as kind-for-kind, limitations on magnitude arose only later.²¹ Legal scholars describe this initial stage of development, when the norm of kind-for-kind retaliation became established, as “discretionary retaliation.” If the member of one tribe harmed the member of another tribe, the offended tribe was obliged to seek like vengeance (of variable magnitude) upon the offending tribe. The retaliatory conduct at this stage was still often subject to an arbitrary multiplier. For instance, Genesis 4:14 states, “[W]hosoever slayeth Cain, vengeance shall be taken on him sevenfold.” Similarly, II Samuel 12:13–18 prescribes a four-fold measure, while Genesis 9:5 prescribes a two-fold measure.²² Though it constrained the type of revenge-harm, discretionary retaliation could still lead to an escalating infliction of reciprocal harms. For example, one group, regarding the retaliation of another group as unwarranted or excessive, might counter-retaliate with vengeance of greater magnitude, inciting the other group to counter-counterretaliate with vengeance of yet greater magnitude.²³ Such cycles of increasing violence would clearly be undesirable for the participants in the long run, and discretionary retaliation norms tended to evolve limitations on the permissible magnitude of retaliation.²⁴

²¹ Blau (1916: 7). ²² Interestingly, Sulzberger (1915: 33) analyzes the two-fold measure as arising from the doctrine of “double blood-guilt.” The retaliation is given twice: once for the individual injurer, and again for the injurer’s tribe for failing to prevent or punish the crime internally. ²³ See Parisi (2001). ²⁴ The escalation of blood feuds is commonly observed in societies with discretionary retaliation norms. See e.g. Richter (1992) on escalating “mourning wars” between Iroquois tribes. Parisi (2001) suggests several elements that may lead to

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

354 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli The second stage of development is characterized by the emergence of the lex talionis norm—that retaliatory harms should be equal in magnitude to the harms suffered, popularly captured by the phrase, “an eye for an eye.”²⁵ This development had the obvious benefit of preempting the mechanisms that led to devastating spirals of reciprocal violence. It is at this stage of social development that moral conceptions of desert and retribution began to take shape. The impulse to strike back transforms into a sense that harms are “moral wrongs,” for which an equivalent retaliatory harm is “owed.” In economic terms, the inchoate moral calculus implicit in this new norm acted as an upper bound on ρ, such that ρ  H. It is easy to see why the upper bound of lex talionis is an improvement over unbounded ρ. First, it reduces excessive expenditure on κ on the victim’s side, which generates over-deterrence when ρ > H. Second, it sets a benchmark for proportional retaliation, against which injurers are neither inclined nor obliged to counter-retaliate. Third, it eliminates the mechanism for spiraling retaliations of increasing magnitude. Talionic law was an important development in the formalization of early legal institutions. In addition to specifying the magnitude of liability (the distinction between criminal liability and tort liability had not yet developed), procedural clarifications facilitated the orderly execution of talionic retaliation. In the biblical tradition, the institution of Go’el (nearest of blood) determined the right of the victim and his extended family to exact vengeance.²⁶ This effectively reduced coordination problems that arose in the discretionary retaliation paradigm, when it was often unclear upon whom the duty to exact revenge fell, possibly leading to duplicative efforts/harms. The exceptional case of ancient Indian law follows a similar history of formalization. Though Indian law was anomalously based not on retributive, but rather incapacitative grounds, its early development included a comparable systemization of consequences, specifying upper bounds for punishments. For example, the harshest punishment for theft was dismemberment of the hands. Though this may seem draconian by modern standards, it is important to recognize that the function of this formal rule was not to expand, but rather to escalating retaliatory actions, including variations in the perceptions of blameworthiness of a wrongdoer, and overestimation of the gravity of harm suffered. ²⁵ See Miller (2007). ²⁶ Good (1967).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

355

delimit the magnitude of available remedies. The king (ideally an impartial official) was also specified as the legal executor of punishments, with procedural constraints imposed by the Brahmins, further reducing the risks of overinvestment in punishment.²⁷ Although the second stage of development is characterized by the emergence of the principle of lex talionis, several other important changes in the proto-legal institutions of early societies may be observed. First are the codification and increasing formalization of the law, which not only specifies the measure of retaliation due, but also delimits the duties and rights of parties after the occurrence of a harm. Second, and more abstractly, it should be observed that the rationale underlying retaliation transformed from the pre-legal, primordial desire for revenge into a moral prescription for desert. It is also important to notice that in economic terms, the principle of lex talionis crudely assigns a measure of retaliation equal to the harm suffered, ρ  H, thus forcing potential injurers to internalize the cost of their “wrongdoing.” It thus approximates, however crudely, something approaching an efficient measure of deterrence. In its earliest period, Roman law applied talionic law for cases of iniuria, including the case of intentional injury to another with the breaking of a limb or a bone or insult.²⁸ But in a different case Romans also admitted a non-proportional retaliation. The Twelve Tables,²⁹ an archaic law promulgated in the fifth century BC, prescribed that the victim could lawfully kill the thief, if he had been caught in the act and was operating at night.³⁰ The murder was the lawful reaction for theft. It is also possible to ascribe to a legalization of vengeance the first personal compulsory settlement that the Romans ordered in the law of Twelve Tables, the manus iniectio (laying of a hand upon a person).³¹ This procedure authorized a private citizen to formally ²⁷ See Das Gupta (1930); Doongaji (1986); Jaishankar and Haldar (2004); Olivelle (2011). ²⁸ Law of the Twelve Tables, Table 8.2: Si membrum rup[s]it, ni cum eo pacit, talio esto (“If a victim is maimed, then in the absence of compensation, retaliation is in kind”). ²⁹ See Humbert (2005). ³⁰ Law of the Twelve Tables, Table 8.12: Si nox furtum faxsit, si in occisit, iure caesus esto (“If the theft has been done at night, if [the owner] kills the thief, [the thief] shall be considered to have been lawfully killed”). ³¹ Law of Twelve Tables, Table 1.1: Si in ius vocat,. Ni it, antestamin[o]: igitur em capito (“When [the plaintiff] brings a case against [the defendant] before the court,

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

356 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli arrest the debtor, if—without giving a guarantor (vindex)—he did not present himself at the trial as defendant or, independently from the trial, he did not fulfill his debt.³² The creditor could sell the debtor as addictus (slave until the discharge of his debt), but if there was no buyer found at the market, the creditor could lawfully kill him.³³ The cruel eventuality to cut the body of the debtor into pieces was probably related to the case in which there was more than one creditor.³⁴ These conceptual transformations were important ones. Recall that the predisposition to seek revenge seemed best explained in evolutionary terms as a means of effecting deterrence. This biologically successful phenotype was tamed by social practice into a retributive ground for retaliation. Yet this transformation brought us full circle back to what appears to be a crude attempt at effecting optimal deterrence, revealing the mechanism likely at work in effecting the institutional changes. This evolution of the grounds for punishment suggests a closer relationship between retributive and deterrence theories than many scholars have hitherto supposed.³⁵ Rather than being two competing objectives of criminal law (and to a lesser extent, tort law), “retribution” may thus be regarded as a refinement of the deterrence objective achieved through social norms.

22.2. “WRONGDOING” AND DETERRENCE Many features of lex talionis suggest that it was a far more dynamic system than often supposed, designed to take into account more complex elements, such as general deterrent value, enforcement [the defendant] should appear. If [the defendant] doesn’t appear for trial, [the plaintiff] should call the witnesses. Afterwards he should catch him [the defendant]”). ³² Kaser (1971: 132). ³³ Law of Twelve Tables, Table 3.6: Tertiis nundinis partis secanto. si plus minusve secuerunt, se fraude esto (“On the third market day they ought to cut off pieces of the corpse. Whether they have cut off too much or too little, should be left unpunished”). ³⁴ Talamanca (1990: 293). ³⁵ Though a substantial number of “mixed” theories of punishment have recognized the close relationship between retribution and deterrence rationales. Duff (1986); Hirsch and Ashworth (2005); Moore (2007).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

357

errors, and specific enforcement problems. As we remarked earlier, the ancient body of laws dealing with “wrongdoing” treated both tortious and criminal behavior as one, often without distinguishing between intentional, culpable, or accidental wrongdoing (Cairns and Robinson 2001). Modern legal systems, with their separate treatment of torts and crimes, tailor remedies according to the intentional or non-intentional nature of the act, generally pursuing goals of absolute deterrence through criminal remedies (teaching prospective wrongdoers that “crime does not pay”), and goals of relative deterrence through tort remedies (teaching prospective tortfeasors that “precautions pay”). Specifically, modern tort law forces wrongdoers to compensate the victim for the harm suffered (H). By contrast, modern criminal law creates deterrence by imposing sanctions S based on the benefit enjoyed by the injurer B and the probability of detection, ρ, oftentimes irrespective of H. In modern criminal law the sanction is tailored to the incentives of the criminal. For example, thefts of sheep for slaughter or for work are both punished as thefts; that is, the punitive element of the sanction is the same. Deterrence is achieved when the benefit to the wrongdoer is less than the probability of enforcement multiplied by the sanction, i.e. B < pS, where S is independent of H. This difference between criminal law and tort law influences the effects of the two remedial regimes: modern tort law pursues goals of relative deterrence (“precautions pay”) by forcing the internalization of the loss H, while criminal law pursues goals of absolute deterrence (“crime does not pay”) taking into account the criminal’s temptation to commit the crime, B. Talionic regimes do not bifurcate remedial instruments and tackle both objectives of criminal and tort remedies at once. By replicating the harm suffered by the victim onto the wrongdoer, and giving the parties an option to settle through the payment of blood-money (kofer), lex talionis provided a hybrid remedy that led parties to take into account both the harm to the victim and the benefit to the wrongdoer. As will be discussed below, the effectiveness of the talionic regime hinges upon the ratio of the benefit to the offender to the harm that he caused. Talionic law looks at the ratio B/H to achieve punitive and compensatory goals. Since the sanction is tied to the loss to the victim, and because it is the victim’s (or his tribe’s) right to impose this penalty, both the victim’s loss and the benefit to the wrongdoer are

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

358 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli Unified Law of Wrongs (B/H)

Tort Law (B)

Criminal Law (H)

Figure 22.1. Deterring wrongdoing.

ultimately taken into account, balancing objectives of absolute and relative deterrence in view of the ratio between B=H. Figure 22.1 illustrates how the talionic system relates to its tort and criminal law alternatives. When different instruments are available to pursue relative deterrence of unintentional wrongdoing through tort remedies and absolute deterrence of intentional wrongdoing through criminal remedies, problems of deterrence spillovers across different categories of wrongdoing can be reduced and enforcement levels can be tailored to the respective policy objectives (Parisi 2001: Dari-Mattiacci and Parisi 2004). When a single set of remedies is adopted, the choice of optimal enforcement levels becomes essential for the pursuit of effective deterrence. Prior to the lex talionis regimes, where kind-for-kind was the only requirement for punishment, multipliers were used for retaliation: the punishment imposed on the wrongdoer was more severe than the harm originally suffered by the victim (Sulzberger 1915). When the lex talionis imposed a measure-for-measure limit to the punitive remedy, possible situations of underdeterrence could emerge.³⁶ With enforcement errors, only a fraction of wrongdoers would be caught and punished, and prospective wrongdoers could expect to face retaliatory punishment only a fraction of the time. The relationship between retaliatory punishment and deterrence is particularly interesting when considering enforcement levels.³⁷ With retaliatory punishment, a lower level of enforcement is needed to deter “inefficient crimes.” Retaliatory punishment instead becomes less effective in deterring “efficient crimes” (i.e. those crimes for which the benefit to the injurer is equal or greater than the harm to ³⁶ Blau (1916: 7). ³⁷ See Czabanski (2008: ch. 4) for a discussion of the relationship between retaliation and deterrence in ancient law. See also Stein (1999).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

359

the victim). To understand the “deterrence paradox” of retaliatory punishment with enforcement errors, it is necessary to take into account the disparity between the benefit achieved by the wrongdoer and the expected cost of punishment. With retaliatory punishment, deterrence is preserved for inefficient crimes, but underdeterrence may arise for relatively efficient crimes. Retaliatory punishment thus links the harm to the offender to the harm suffered by the victim. When the multiplier α is equal to one and with a less than 100 percent enforcement rate, deterrence cannot be achieved for efficient crimes. Unlike the internalization of the harm in tort law, absolute deterrence of crime requires offsetting the offender’s benefit with a threat of punishment. Deterrence should teach the wrongdoer that crime does not pay. A rational criminal should be adequately deterred if the benefit from the crime, B, is less than the expected sanction, equal to the probability of enforcement, p, times the sanction S. B < pS Under lex talionis regimes, the sanction S is set equal to the victim’s loss H. Hence, enforcement should occur with a probability p, such that p > B=H Crimes are generally inefficient. This is because the benefit that the wrongdoer captures from committing the wrong is generally less than the cost he imposes on the victim, B < H. As a result, a wrongdoer would not generally agree to suffer retaliatory punishment in exchange for the right to impose harm on his victim. The degree of inefficiency of a crime can vary greatly, 0  B=H < 1. Lower values of B=H indicate “more inefficient” crimes. Despite the mechanical proportionality of retaliatory punishment (i.e. more malicious crimes lead to more malicious sanctions), S ¼ H, it is interesting to observe that, as H increases relative to B, the enforcement level necessary to preserve adequate deterrence actually decreases. When enforcement is imperfect, it becomes rational for some wrongdoers to carry out some wrongful activities, and the wrongs that will be rationally perpetrated will be those that are generally less inefficient (i.e. those characterized by higher values of B=H).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

360 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli The deterrence paradox is easily found in early legal rules.³⁸ The crimes mentioned in the early codes that impose measure-formeasure retaliatory punishment (murder, mutilation, wounding, etc.) are crimes that generally take a lot from the victim and give a smaller benefit to the perpetrator. For example, the crimes covered by the lex talionis are characterized by lower values of B=H. The optimal enforcement necessary to ensure deterrence ends up being lower for the most inefficient crimes. For these crimes, enforcement level could be kept low, without compromising effective deterrence. Policing and enforcement of brutal crimes could be kept low, while policing of less violent crimes would paradoxically need to be boosted. Contrary to a retributivist view of punishment, given that policing and enforcement are costly, when retaliatory penalties are utilized, more serious crimes should be enforced less rigorously. As the deadweight loss of the crime becomes smaller (i.e. when B=H goes up), society should instead strengthen policing and enforcement measures. Paradoxically, this means society should devote more resources to deter efficient (i.e. less inefficient) crimes, or else adopt a different standard of punishment. The foregoing analysis further explains why early legal systems did not apply the measure-for-measure cap on punishment to crimes that entail a larger prospective benefit for the wrongdoer (Parisi 2001). The optimal enforcement necessary to ensure deterrence in such cases would be higher. One could imagine crimes such as theft where the benefit to the victim approaches the sanction in a kindfor-kind regime such as lex talionis. To promote adequate deterrence, early legal systems adopted higher α > 1 in the case of theft, to account for the lack of deterrence inherent to theft.³⁹ This ensured that the ratio B/S would still be below the probability of enforcement. A measure-for-measure requirement for theft would have been futile here, while for violent crimes it would have been effective. Within the confines of theft, further adjustments were made to correct for possible detection problems. These deterrent aspects of lex talionis are further implicated by the higher penalties for animals marked for slaughter than animals for work. For example, animals intended for ³⁸ See Kavka (1978) for a discussion on the emergence of deterrence paradox. See also Fagan and Meares (2008) focusing on the deterrence paradox in minority communities. ³⁹ e.g. Maine (1861: 378).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

361

slaughter had a higher multiplier than those marked for work. Sheep had a multiplier of 5:1, while oxen had a multiplier of 4:1.⁴⁰ This is due to the specific detection problems with respect to animals intended for slaughter. Animals for slaughter may likely be stolen and perhaps eaten shortly after their theft. The perpetrator could literally swallow the evidence. Animals for work, however, may be recognized by a unique brand or other individual characteristics of the animal. Since the probability of enforcement was necessarily lower for slaughter animals, the lex talionis regime compensated by increasing the ratio of B to H, making enforcement less necessary for adequate deterrence, once again suggesting an understanding of the role of B/H in ancient times. Yet even within the category of work animals, special exceptions were made when they considered two animals for work, both of which need to be retrained.⁴¹ Here the animals are differentiated by loyalty. One only takes orders from the owner; thus, instructions from strangers will not be executed. Other animals would listen to the new owner immediately. A corollary today would be the difference between a dog, loyal to its owner and a horse, generally more likely to obey whoever is riding it. Considering this problem in terms of social efficiency, we can see that the theft of an animal that one would have to retrain is less efficient. Intuitively, one would expect that theft to be more socially undesirable and it would have to be deterred more severely. However, that may still encourage crimes of animals that more easily adapt to a new owner. In fact, to discourage theft, lex talionis punished thefts that required no retraining with greater multipliers. Again the B/S ratio does not look at loss to the victim, but it rather looks at the benefit to the perpetrator relative to the sanction imposed on the same perpetrator. If the prospective perpetrator can capture the full value of an animal, that is a very temping theft, or a high B value. The animals that needed to be retrained had a lower B value. To compensate for the higher B value of non-loyal animals, a higher S value was implemented as well. Again, deterrence played the primary role in norms dictating the penalties of the time.

⁴⁰ Exodus 22:3–4.

⁴¹ The Code of Hammurabi, Par. 8.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

362 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli 22.3. THE EVOLUTION OF INDIVIDUAL LIABILITY The lex talionis principle marks a stage in the evolution of ancient law, after which civil and criminal penalties begin to diverge. Before exploring these two branches of development, we pause to consider an orthogonal development: the evolution of individual liability. In nearly all early societies, there existed a clear division between intertribal and intra-tribal norms. The revenge laws governing the interaction between tribes did not necessary apply to the internal interactions of members within the same tribe. The intertribal revenge norms gradually evolved into modern legal systems. Meanwhile, intratribal norms grew less relevant as the influence of tribal units deteriorated. The mechanisms and relationships between these transformations are not immediately obvious and require some explication.⁴² It is important to observe that the tribal unit was essential to the development of intertribal norms, inasmuch as its dissolution was essential to the further development of individual liability. Posner (1980) discusses the dynamics of intra-tribal interactions in great detail, identifying important characteristics of early societies that distinguish them from modern societies. First, life in early societies was characterized by an extreme scarcity of resources. Second, information costs were prohibitive in many situations. And third, government was either weak or nonexistent. These factors entailed important consequences. Due to the scarcity of resources and high information costs, tribes served an important and desirable insurance function. By pooling resources, individuals were better equipped to weather risks. To counteract the lack of centralized government and high information costs, tribes lived in close-knit communities. By living in crowded conditions lacking privacy, where each individual knew what other members of the tribe were doing (while also being monitored by other members of the tribe), information costs were effectively reduced, leading to an informal, decentralized reciprocal surveillance.⁴³ ⁴² For a detailed formal discussion of this topic, see Dari-Mattiacci and Parisi (2004). ⁴³ However, due to the strongly aligned interests of individuals and their tribes, it is doubtful whether internally committed wrongs were punished. See Faris (1915), pointing out that the patriarch of a tribe has little incentive to punish members of his own tribe.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

363

It is the dual functions of insurance and self-policing that allowed intertribal revenge norms to develop. Assigning liability not upon individual wrongdoers, but instead upon the wrongdoers’ tribes allowed ancient societies to overcome the detection and solvency problems that would have attended individual liability. Moreover, the structure of early societies amplified the deterrence effect of intertribal retaliation, because tribes were incentivized to exercise collective “self-control” over their members to avoid the shared burden of retaliation. Historically, the constraints of lex talionis tended to be pursued with an absurd degree of fidelity to the rule. For instance, in the Native American “law of revenge,” it was the right of the wronged tribe to inflict retaliation—identical in kind and magnitude—for harms suffered. If this retaliation were executed by anyone except a member of the injured tribe, however, that act would not be recognized as retaliatory, but regarded as a fresh wrong requiring its own retribution.⁴⁴ Biblical and Babylonian implementations of lex talionis required an even more extreme formal symmetry in the acceptable retaliatory act: a man who causes the death of another man’s child is to lose his own (Daube 1947: 169);⁴⁵ the wife of a rapist is to be raped by the victim’s family (Middle Assyrian Laws, Paragraph 55);⁴⁶ and if a building collapses, killing the owner’s son, then the builder’s son is to be put to death (Code of Hammurabi, Paragraph 230). However, as the environmental constraints motivating communal liability, identified by Posner (1980), diminished over time, so too did the effectiveness of intra-tribal punishments in deterring harms. For example, Dari-Mattiacci and Parisi (2004) formally deduced that as a tribe’s wealth and population increased, common pool problems (the tendency of one person to use commonly owned property without regard for the interests of other owners), moral hazard, and free-rider problems (when people benefit by the efforts or expenditures undertaken by others without contributing themselves) tended to increase as well, diminishing the deterrence effect of communal liability. ⁴⁴ Thus, the law of revenge was intransitive. For example, if Tribe A injures Tribe B, and Tribe B injures Tribe C, and Tribe C injures Tribe A, then the owed retaliations do not cancel out. Rather, Tribe B must retaliate against Tribe A Tribe C must retaliate against Tribe B and Tribe A must retaliate against Tribe C for “justice” to be satisfied. ⁴⁵ Similarly, see Code of Hammurabi, Par. 210 (requiring that the daughter of a man who strikes a pregnant woman and causes a miscarriage be put to death). ⁴⁶ See Roth (1995: 174–5).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

364 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli This conclusion is easy to grasp intuitively. While we may suppose that the utility that an individual gains from harming (or risking harm to) individuals of another tribe is indifferent to changes in wealth or population, the decreasing share of communal liability for each individual as the population increases in size reduces that individual’s expected costs from committing the wrong. To illustrate the point by use of an extreme case, contrast the impact of retaliation against a tribe consisting of a dozen individuals who all provide essential functions to the community, versus retaliation against a wealthy tribe, consisting of many hundreds of individuals, some of whom are only distantly acquainted with the wouldbe injurer. It should be intuitively obvious that the cost of communal liability in the latter case is substantially less than in the former. Thus, as a consequence of increased wealth and population, the individuals’ interests will tend to diverge from the social optimum, as incentives become misaligned. Unsurprisingly, we do in fact observe communal liability collapsing into individual liability in our historical specimens. For example, by the sixth century BCE, communal liability was renounced in the biblical tradition: “The fathers shall not be put to death for the children, neither shall the children be put to death for the fathers” (Deuteronomy 24:16).⁴⁷

22.4. THE COMMODIFICATION OF CIVIL LIABILITY Up to this point, ancient legal systems treated liability monolithically, failing to distinguish between intentional and accidental harms. The division of tort liability and criminal liability begins to emerge with the commodification of retaliation. Readers familiar with Coase (1960) will not be surprised, given the retaliatory right under lex talionis regimes, to learn that the kind-forkind constraint eventually gave way to pecuniary compensation. In Coasean terms, the initial allocation of the retaliatory right was not

⁴⁷ See generally Parisi (1992, 1997).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

365

assigned to the highest-valuing party. Thus, we expect the injurer and victim would bargain for a transfer of the right. Several very plausible assumptions are implicit in this claim. First, we assume that injurers value the retaliatory right more than victims. Though individuals are indubitably predisposed to seek retribution for harms suffered (recall the discussion from §22.1), it seems likely that the desire of injurers to avoid retaliation will tend to be greater than victims’ desire to exact vengeance. For example, if A harms B, resulting in the loss of a limb, then by lex talionis, B obtains the right to dismemberment of A. Yet it seems quite likely that although B’s desire to retaliate may be great (call the magnitude of his desire x), A’s desire to keep his body intact will be greater still (call it y½> x). Thus, there will be a price ρ in the range p ∈½x; y, such that A can simply pay B not to exercise the retaliatory right, whereupon A enjoys the surplus y  p, and B enjoys the surplus p  x, assuming zero (or sufficiently low) transaction costs,⁴⁸ leading to a Pareto superior outcome: B gets more than the revenge was worth to him, and A gets to keep his limb. In fact, we may observe such bargaining in all lex talionis regimes throughout history. For example, in times of peace, ancient Athenians tended to exercise compensation alternatives;⁴⁹ blood revenge was rarely carried out in practice (McHardy 2008: 9). Interestingly, however, blood revenge remained common during times of war, when vengeance was exacted on the battlefield.⁵⁰ We may speculate that this was due to two factors. First, transaction costs during times of war may have been prohibitive. Second, since the legal right to kill enemy combatants during war was redundant with the right to retaliation, paying off the victim in exchange for one right did not prevent the victim from exercising the wartime right to kill the injurer anyway. At any rate, the Athenians seem clearly to have practiced Coasean bargaining of the right to retaliation. Likewise, later Talmudic interpretations of lex talionis understood “an eye for an eye” in compensatory terms as “the value of an eye for ⁴⁸ Given the stakes in the hypothetical, it is plausible that transaction costs will not exhaust the bargaining space in most situations. ⁴⁹ Demosthenes, Orationes 43.57. ⁵⁰ It is also interesting that Athenian women were less willing to accept compensation for harm, insisting on “specific performance” of the blood revenge remedy. McHardy (2008: 9). We speculate that this may be due to qualitative differences in the crimes committed against women in ancient Greece.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

366 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli an eye.” The rationale evolved from an apparent paradox in the strict application of the kind-for-kind constraint. The rabbinic scholars point out that Leviticus 24:22 requires, “Ye shall have one manner of law,” and yet the requirement, “an eye for an eye,” interpreted literally would be impracticable in many cases: “What then will you say where a blind man put out the eye of another man, or where a cripple cut off the hand of another, or where a lame person broke the leg of another?” (Talmud, Baba Kamma 84a). Thus it was recognized that “an eye for an eye” must, on pain of self-contradiction, be interpreted to mean “the value of an eye for an eye.” Pecuniary compensation was eventually established as the exclusive form of retaliatory right, rather than a mere alternative to retaliation, in Jewish law.⁵¹ We observe similar commodification of the lex talionis principle universally in the development of other ancient legal systems, including the Roman law of the Twelve Tables, the Visigothic Code,⁵² the Salic Code,⁵³ and informally in the Native American law of revenge.⁵⁴ Thanks to the numerous sources that survived throughout the centuries, we can observe a similar—and very well-defined—pattern of evolution in the history of Roman law. The Roman law of the Twelve Tables includes a provision that an agreement between the parties can obviate the literal application of the lex talionis. The option to buy out the right to talionic revenge through the payment of a sum of money in cases of iniuria was an important step in this evolution,⁵⁵ which is often regarded as the genesis of the concept of delictual obligations.⁵⁶ The right to retaliation in kind outlives this early period throughout classical Roman law. An example is found in the case of manifest theft (the so called furtum manifestum, when the person is caught in the act; a theft was manifest if the thief was caught on the day of the theft with the stolen property before reaching the place where he intended to take it). This became a delict that created an obligation for the ⁵¹ Pasachoff and Littman (2005). ⁵² Visigothic Code at 7.3.3 (“Should they wish to do so, they may exact from the kidnapper, the legal compensation for homicide; that is to say, three hundred solidi . . . ”). See Scott (1910: 248). ⁵³ Henderson (1910: 176–89): “Title LXII. If any one’s father have been killed, the sons shall have half the compounding money (wergeld); and the other half the nearest relatives, as well on the mother’s as on the father’s side, shall divide among themselves.” See also Hessels and Kern (1880). ⁵⁴ See Reid (1970); Hudson (1976); and Richter (1992). ⁵⁵ Law of the Twelve Tables, Table 8.2. ⁵⁶ Talamanca (1979: 2–4); Scherillo and Gnoli (1994: 127).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

367

wrongdoer to pay a sum of money, which was a multiple of value of property stolen. The thief could no longer be lawfully killed, but he was required to pay back the double of the value of the stolen property (or four times that value, if he was caught in flagrante). An important turning point in the Roman history of tort was the Lex Aquilia, a statute from the time of the Republic. This statute for the first time created the opportunity to receive damages (that is to receive money as a compensation for damage that one has suffered), laying the foundations for modern tort liability.⁵⁷ A distinction arises at this point between more serious crimes, such as murder or high treason, that were publicly sanctioned,⁵⁸ and other delicts, like theft, assault, or wrongfully inflicted loss, that were privately prosecuted by the victim, creating an obligation for the offender to pay the victim a certain sum of money.⁵⁹ From the compensatory scheme that evolved from the lex talionis principle, we begin to see the outlines of modern tort law taking shape. In economic terms, the compensatory modification approximates efficient injurer incentives. We now analyze the efficiency of the “blood money” rule that gradually replaced lex talionis in criminal and tort cases. We will examine the efficiency property of blood money, distinguishing between intentional and negligent crimes. A general result is that bargaining creates an inefficient compensation measure and induces underdeterrence of criminal activity. In the case of intentional crimes, bargaining between the offender and the victim (or victim’s family) will take place only in the case of inefficient crimes, i.e. when the criminal’s benefit from committing the crime B is lower than the harm H imposed on the victim. In the absence of any enforcement error, a kind-for-kind lex talionis regime will provide efficient deterrence, since it imposes on the criminal a sanction equal to the harm suffered by the victim, and the offender is not willing to pay to avoid the sanction. However, with inefficient crimes the offender is willing to avoid the kind-for-kind sanction since it would impose on him a greater harm than the benefit gained from the crime.

⁵⁷ See Lübtow (1971: 7); Frier (1989: 3); Hausmaninger (1990: 3). ⁵⁸ Santalucia (1989: 42–3); Kunkel and Schermaier (2005: 41–2). ⁵⁹ Kaser (1971: 609–10).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

368 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli Let γ denote the victim’s bargaining power. The offender is willing to pay up to B þ γðH  BÞ in order to avoid the application of the kind-for-kind sanction required by the lex talionis. This bargaining cannot take place for efficient crimes, since the criminal will still have a positive payoff, equal to the net benefit, B–H, after the kind-for-kind sanction. Bargaining between the offender and the victim takes place in case of negligent crimes, providing an insufficient compensatory measure and hence underdeterrence. In both cases, the criminal is not willing to pay more than γH, escaping the full kind-for-kind sanction, equal to H. This creates room for underdeterrence of delinquent behavior, and inefficient incentives to take optimal precautions. In the presence of enforcement errors, the underdeterrence generated by an inefficient compensatory damage measure is reinforced. Underdeterrence emerges at equilibrium and the incentive to undertake efficient precautions is diluted in the case of negligent crimes. While underdeterrence is the only possible equilibrium for negligent crimes, optimal deterrence of crime could still be achieved in case of intentional crimes. This is solely the case for inefficient crimes, when the enforcement error e is lower than the ratio of benefit to harm, i.e. B e < BþγðHBÞ . Note that blood money is less frequently observed in the presence of enforcement error since the range of values that support efficient deterrence is reduced after the introduction of bargaining. For efficient intentional crimes, however, we will not observe any bargaining and the only possible way to achieve higher efficiency in the deterrence of crime is to adopt more severe sanctions, replacing a kind-for-kind sanction with higher multipliers α in order to correct for the enforcement error e. Higher multipliers α should be optimally implemented for most efficient crimes in order to counteract underdeterrence. Several conceptual steps remain in the evolution of tort liability, separating the compensatory modification of lex talionis from modern tort law. These include: (1) recognition of the distinction between intentional and accidental wrongs; (2) elimination of the bargaining element via the introduction of fixed remedies; (3) identification of compensation with harm; and (4) recognition of victim precautions as a fundamental element in the calculus of tort liability. However, these sophisticated changes occurred in relatively developed legal systems, which fall outside the scope of our present inquiry.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

369

Yet it bears mentioning the durability of the blood money stage of legal development. Unlike bare vengeance, discretionary retaliation, and lex talionis, the mechanisms forcing further evolution from blood money are subtler and less forceful. In regimes where legal change is difficult, the relative efficiency of blood money represents a possible stopping point in the refinement of tort liability rules. Indeed, though the context of our discussion is ancient law, it is worth pointing out the persistence of blood money in modern day Sharia in the concept of Qis.ās..⁶⁰ Likewise, in the customary law of Somalia, blood money compensation continues to be practiced. And informally, blood money compensation remains a common extralegal social practice in both Japan (mimaikin) and Korea (hapuigeum). For a more detailed analysis of the evolution of tort liability in ancient law, see Parisi (2001).

22.5. THE CENTRALIZATION OF CRIMINAL LIABILITY The literature distinguishing criminal law from tort law is vast.⁶¹ In the philosophy of punishment, retributivists (and mixed theorists) identify a putatively moral dimension in criminal conduct as the distinctive element separating crimes from torts.⁶² Scholars in this field are interested in the distinction for normative reasons, seeking to determine what acts the state ought to punish, and what justifies punishment from an ethical standpoint. However, careless scholars may try to transform the claim into a descriptive one. If moral considerations are given as a descriptive distinction between the criminal law and the law of torts, it is difficult for such claims to avoid endogeneity problems. Even assuming that crimes necessarily punish moral wrongs, and that torts do not (already a disputable point), it is not clear whether such activities are regarded as criminal because they are immoral, or whether in at least some cases they are seen as immoral because they are criminalized. ⁶⁰ El-Awa (1981). ⁶¹ See e.g. Morris (1933); Posner (1980); Coffee (1991); Epstein (1997); Weinstein (2001); and Simons (2008). ⁶² See e.g. Hirsch and Ashworth (2005); Moore (2007); Husak (2008).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

370 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli These questions raise meta-ethical problems about the existence of moral facts, which are tangential to our present inquiry.⁶³ We think that the historical dimension of the crime/tort distinction may be more usefully seen in economic terms. Posner (1980) takes the strictly instrumental view that the crime/tort distinction is a function of wealth, government, and information costs. As we discussed earlier, lex talionis norms taking the tribe as the bearer of liability developed due to the high costs of detection, lack of state authority, and the lack of wealth needed to compensate victims.⁶⁴ Given these constraints, the legal instruments of modern criminal law and tort law would have been impracticable. As wealth increased in ancient societies, the possibility of compensatory payments emerged, and lex talionis was modified to allow compensation as an alternative to reciprocal retaliation. As discussed in the previous section, this innovation evolved into tort law. However, state authority tended to become established concomitantly with increases in wealth, allowing state enforcement and sanctions to emerge from the undifferentiated liability of lex talionis. This process led to the development of the criminal law. The relationships between economic development, population growth, and the evolution of criminal law are complex. Increasing wealth and population allowed for the establishment of more efficient deterrence mechanisms, such as the compensatory scheme discussed in the previous section. Yet these more efficient rules contributed to greater predictability, stability, and wealth for those communities, which in turn allowed for yet more efficient rules. However, the effects of increasing wealth and population on the law were not all positive. As discussed in §22.3, the move from tribal liability to individual liability was occasioned by increases in population and wealth to avoid free-rider and common pool problems with respect to intertribal offenses. However, it is not difficult to see how the misalignment of interests resulting from population growth would also lead to an increased need for intra-tribal mechanisms for deterring inefficient conduct. ⁶³ The denial of moral facts is central to a large number of skeptical theories. A particularly influential and clear account of one variant—moral error theory—may be found in Mackie (1977), further developed in Joyce (2001, 2006). ⁶⁴ Posner (1980: 43).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

371

In addition, although the increasing collective wealth of communities may have allowed the compensatory modification of lex talionis to replace retaliatory violence prior to the dissolution of communal liability, this process would not apply to intra-tribal remedies. Though mostly tangential to our discussion, it is here worth observing that during this phase of legal development, individual property rights would also be emerging, leading to heterogeneity in the personal wealth of community members. Thus, though relatively wealthy communities might be capable of mustering pecuniary compensation for victims of harm caused by its members, there would still be relatively poor individuals within the communities incapable of mustering compensation privately. Thus, as private incentives diverged from the social optimum, these individuals would not be deterred by damages-based deterrence measures. Thus, as population and wealth increase, we expect that early societies will tend to develop a system of criminal law, responding to two pressures: first, the increased need for deterrence in cases of intra-tribal harms; and second, the move from communal to individual liability in cases of intertribal harms. In both cases, the presence of insolvent individuals will undermine the deterrent effect of the compensation mechanism. But recall that the change from retaliatory lex talionis to compensatory lex talionis was a Pareto improving move. Compensation-based deterrence, by forcing injurers to internalize the harms they cause, will tend to create efficient incentives (for wealthy injurers, at least). Communities should therefore be reluctant to abandon (or forgo) compensation-based deterrence, simply because it may be ineffective in a subset of cases. And indeed, ancient civilizations did not do this, but instead invented different mechanisms to effect a different kind of deterrence where compensation-based deterrence fails: the criminal law. Additional factors may have contributed to the motivation for development of a criminal law. For activities where victims are unlikely or unable to demand compensation, such as incest or pollution, or where the harm is suffered by the public rather than a particular individual, such as “witchcraft” or treason, the compensation system might not provide adequate (or indeed any) deterrence. In yet other cases, the harm may be so great that the compensation remedy is insufficient to force individuals to internalize the harm. Finally, where injuries to others result in a reduction in tax revenue (e.g when the injured party is less productive as a result of his injury),

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

372 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli then this represents an additional social harm (above the private harm to the victim), and in such cases, criminal penalties may force potential injurers to internalize this additional cost.⁶⁵ Criminal law as an instrument may also provide additional benefits that civil liability is less capable of achieving. For example, assuming that moral beliefs are at least partly shaped by the law, criminalization may perform an important (albeit difficult to quantify) expressive function—transforming the preferences of community members, and generating a more efficient mix of norms. The expressive effect of the criminal law may also generate focal points, resolving coordination problems for various activities.⁶⁶ To summarize: the principle of lex talionis created a general deterrence against undifferentiated “wrongs.” As early societies developed, more efficient compensatory mechanisms arose, which later developed into the modern law of torts. However, the deterrent effect of civil liability, while more efficient than that of lex talionis, is identical neither in magnitude nor scope. Thus, we may regard the criminal law as having arisen to fulfill those functions for which compensation-based deterrence was either insufficient or inapplicable.

22.6. CONCLUSI ON Historical overviews such as this chapter are not unlike theoretical models. Models pick out features of their explananda as salient, while turning a blind eye to others, to lay bare their essential elements and relations. By forcing reality into a quantifiable framework, models filter an undifferentiated mass of facts into something meaningful— something accessible to analysis. Likewise, no telling of history can be exhaustive; instead, we must choose which aspects of the past to highlight, and which to disregard. In selecting certain features of the past, while ignoring others, we begin to construct a story, abstracted from reality to shine light upon some insight. In this chapter, we have surveyed the evolution of deterrence in ancient law. We began with the observation that individuals often ⁶⁵ See Posner (1980: 51–2). ⁶⁶ For an impressively general theory of law as a mechanism for generating focal points to solve coordination problems, see McAdams (2000).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

373

seem to desire revenge for perceived wrongs committed against them. We offered an explanation for this apparently innate preference in terms of evolutionary psychology: that a disposition for retaliatory behavior acted as a deterrent against other tribes causing harm. Though investment in revenge-seeking was undoubtedly costly, when framed in game theoretic terms, we described how it could represent an efficient equilibrium in repeated-games. We then described the process through which this preference for revengeseeking transformed into a social norm, constrained by kind-forkind limitations of discretionary magnitude. Later, an additional measure-for-measure constraint developed to limit spirals of escalating blood feuds. Concomitant with these changes, liability shifted from tribes to individuals. Finally, more specialized tools for effecting deterrence were developed, and the monolithic liability of lex talionis branched out on two separate paths: tort liability and criminal liability. At each stage, we observed that early societies tended to adopt norms (later laws) that were efficiency-improving. Indeed, each stage of development seems to have been characterized by the adoption of the ever more efficient rules, constrained by severely limiting circumstances. As ancient societies grew in wealth and population, and as state power grew, more efficient rules became available. Thus, when the broad sweep of legal history is viewed through an economic lens, we are able to identify the deterrence objective undergirding many of the major “paradigm shifts” in the law.⁶⁷ From its antediluvian pre-legal origins through to what would eventually become the modern law of torts and criminal law, we have seen the deterrence objective effect an alignment of social and private incentives through a gradual process of accretion. The resilience of the deterrence objective is evident in the remarkable variety of its disparate manifestations in human interaction—from base instinctual response to social norm to legal rule. That the evolution of these

⁶⁷ We are sensitive to possible criticism that applying Kuhn’s (1962) concept of “paradigm” to a social practice, such as the law, is a misuse of the term. The mechanisms effecting large-scale transformation in the sciences, identified by Kuhn, are not directly transplantable to the law. Our use of the term should be understood as a loose analogy, upon which nothing essential depends. Certainly, we do not intend to imply the epistemic relativism that many of Kuhn’s followers (though not Kuhn himself ) endorsed.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

374 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli apparently distinct realms should share a common cause, to solve and refine a common problem, is a testament to the robustness of the economic forces driving efficiency.

REFERENCES Allen, Danielle S. 2004. “Angry bees, wasps, and jurors: the symbolic politics of ὀργή in Athens,” in Susanna Braund and Glenn W. Most, eds, Ancient Anger: Perspectives from Homer to Galen. Yale Classical Studies (no. 32). Cambridge: Cambridge University Press, 76–98. Blackburn, Bob L. 1980. “From blood revenge to the Lighthorsemen: evolution of law enforcement institutions among the five civilized tribes to 1861.” 8.1 American Indian Law Review 49–63. Blau, Joel. 1916. “Lex talionis,” in Isaac Rabbi, ed., 26 Yearbook of the Central Conference of American Rabbis. Philadelphia: CCAR, 336–66. Cairns, John W., and Robinson, Olivia F. 2001. Critical Studies in Ancient Law, Comparative Law and Legal History. Oxford: Hart Publishing. Camerer, Colin F., and Richard H. Thaler. 1995. “Anomalies: Ultimatums, Dictators and Manners.” 9.2 Journal of Economic Perspectives 209–19. Carlsmith, Kevin M., Timothy D. Wilson, and Daniel T. Gilbert. 2008. “The paradoxical consequences of revenge.” 95 Journal of Personality and Social Psychology 1316–24. Chagnon, Napoleon. 1988. “Life histories, blood revenge and warfare in a tribal population.” 239 Science 985–92. Coase, Ronald H. 1960. “The problem of social cost.” 3 Journal of Law and Economics 1–44. Coffee, John C., Jr. 1991. “Does ‘unlawful’ mean ‘criminal?’: reflections on the disappearing tort/crime distinction in American law.” 71 Boston University Law Review 201–52. Czabanski, Jacek. 2008. Estimates of Cost of Crime: History, Methodologies, and Implications. Warszawa: Springer. Dari-Mattiacci, Giuseppe, and Francesco Parisi. 2004. “The rise and fall of communal liability in ancient law.” 24 International Review of Law and Economics 489–506. Das Gupta, Ramaprasad. 1930. Crime and Punishment in Ancient India. New Delhi: Bharitiya Publishing House. Daube, David. 1947. Studies in Biblical Law. Cambridge: Cambridge University Press. Demosthenes. 2002. Orationes. Oxford: Oxford University Press. Doongaji, Damayanti. 1986. Crime and Punishment in Ancient Hindu Society. Delhi: Ajanta Publications, Distributors, Ajanta Books International.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

375

Duff, Antony. 1986. Trials and Punishments. Cambridge: Cambridge University Press. Falk, Armin, and Urs Fischbacher. 2006. “A theory of reciprocity.” 54.2 Games and Economic Behavior 293–315. Frier, Bruce W. 1989. A Casebook on the Roman Law of Delict. Atlanta: Scholars Press. El-Awa, Mohamed. 1981. Punishment in Islamic Law. Indianapolis: American Trust Publications. Epstein, Richard A. 1997. “The modern uses of ancient law.” 48 South Carolina Law Review 243–65. Fagan, Jeffrey, and Tracey L. Meares. 2008. “Punishment, deterrence and social control: the paradox of punishment in minority communities.” 6 Ohio State Journal of Criminal Law 173–229. Falk, Armin, Ernst Fehr, and Urs Fischbacher. 2008. “Testing theories of fairness—intentions matter.” 62.1 Games and Economic Behavior 287–303. Faris, Ellsworth. 1915. “The origin of punishment,” in Albert Kocourek and John H. Wigmore, eds, Primitive and Ancient Legal Institutions. Boston: Little, Brown & Co. 151–61. Farrington, David P., Patrick A. Langanand, and Per-Olof H. Wikström. 1994. “Changes in crime and punishment in America, England and Sweden between the 1980s and the 1990s.” 3 Studies on Crime and Crime Prevention 104–31. Gendreau, Paul, Claire Goggin, and Francis T. Cullen. 1999. The Effects of Prison Sentences on Recidivism. A Report to the Corrections Research and Development and Aboriginal Policy Branch. Solicitor General of Canada. Ottawa: Public Works & Government Services Canada. Ginat, Joseph. 1997. Blood Revenge: Family Honor, Mediation and Outcasting. Brighton: Sussex Academic Press. Good, Edwin M. 1967. “Capital punishment and its alternatives in ancient Near Eastern law.” 19 Stanford Law Review 947–77. Güth, Werner, Rolf Schmittberger, and Bernd Schwarze. 1982. “An experimental analysis of ultimatum bargaining.” 3.4 Journal of Economic Behavior and Organization 367–88. Hausmaninger, Herbert. 1990. Der Schadensersatz der lex Aquilia. Vienna: Manz. Henderson, Ernest F. 1910. Select Historical Documents of the Middle Ages. London: George Bell and Sons. Hessels, Jan H. 1880. Lex Salica: The Ten Texts with the Glosses and the Lex Emendata. London: John Murray. Hirsch, Andrew von. 1993. Censure and Sanctions. Oxford: Clarendon Press. Hirsch, Andrew von, Anthony E. Bottoms, Elizabeth Burney, and Per-Olof H. Wikström. 1999. Criminal Deterrence and Sentence Severity: An Analysis of Recent Research. Oxford: Hart Publishing.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

376 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli Hirsch, Andrew von, and Andrew Ashworth. 2005. Proportionate Sentencing. Oxford: Oxford University Press. Hudson, Charles M. 1976. The Southeastern Indians. Knoxville: University of Tennessee Press. Humbert, Michel, ed. 2005. Le Dodici Tavole—dai Decemviri agli Umanisti. Pavia: IUSS PRESS. Husak, Douglas. 2008. Overcriminalization: The Limits of the Criminal Law. Oxford: Oxford University Press. Jacoby, Susan. 1983. Wild Justice. New York: Harper & Row. Jaishankar, Karuppannan, and Debarati Haldar. 2004. “Manusmriti: a critique of the criminal justice tenets in the ancient Indian Hindu code.” 1.3 ERCES Online Quarterly Review 1–15. Jhering, Rudolf von. 1906. Geist des römischen Rechts, I./5th–6th edn. Leipzig: Breitkopf und Härtel. Joyce, Richard. 2001. The Myth of Morality. Cambridge: Cambridge University Press. Joyce, Richard. 2006. The Evolution of Morality. Boston: MIT Press. Kant, Immanuel. 1965 [1787]. Critique of Pure Reason, trans. Norman Kemp Smith. New York: St. Martin’s Press. Kaser, Max. 1971. Das römische Privatrecht. Erster Abschnitt: Das altrömische, das vorklassische und klassische Recht. Munich: C. H. Beck. Kavka, Gregory S. 1978. “Some paradoxes of deterrence.” 75.6 Journal of Philosophy 285–302. Kuhn, Thomas S. 1962. The Structure of Scientific Revolutions, 1st edn. Chicago: University of Chicago Press. Kunkel, Wolfgang, and Martin Schermaier. 2005. Römische Rechtsgeschichte. Cologne/Weimar/Vienna: Böhlau. Langan, Patrick A., and David P. Farrington. 1998. Crime and Justice in the United States and in England and Wales, 1981–96. Washington, DC: Bureau of Justice Statistics. Lewis, Clive S. 1953. “The humanitarian theory of punishment.” 6 Res Judicatae 224–30. Lübtow, Ulrich von. 1971. Untersuchungen zur lex Aquilia de damno iniuria dato. Berlin: Duncker & Humblot. Mackie, John L. 1977. Ethics: Inventing Right and Wrong. New York: Penguin Books (repr. in Penguin Books in 1990). Maine, Henry J. S. 1861. Ancient Law: Its Connection with the Early History of Society, and its Relation to Modern Ideas. London: John Murray. Manthe, Ulrich. 2000. Geschichte des römischen Rechts. Munich: C. H. Beck. McAdams, Richard H. 2000. “A focal point theory of expressive law.” 86 Virginia Law Review 1649–1729. McHardy, Fiona. 2008. Revenge in Athenian Culture. London: Duckworth.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Deterrence of Wrongdoing in Ancient Law

377

Miller, William. 2007. An Eye for an Eye. Cambridge: University Press. Mills, Douglas E. 1976. “Kataki-uchi: the practice of blood-revenge in pre-modern Japan.” 10.4 Modern Asian Studies 525–42. Moore, Michael. 2007. Placing Blame. Oxford: Clarendon Press. Morris, Raphael C. 1933. Law and the Social Order. New York: Harcourt Brace & Company. Olivelle, Patrick. 2011. “Penance and punishment: marking the body in criminal law and social ideology of ancient Indi.” 4 Journal of Hindu Studies 1–19. Parens, Erik. 2014. “Afterword: living with the ancient puzzle.” 45.2 Hastings Center Report 50–2. Parisi, Francesco. 1992. Liability for Negligence and Judicial Discretion, 2nd edn. Berkeley: University of California Press. Parisi, Francesco. 1997. “Liability without Fault and Communal Responsibility.” 97–08 Ancient Legal Systems. GMU Law and Economics Working Paper Series. Parisi, Francesco. 2001. “The genesis of liability in ancient law.” 3 American Law and Economics Review 82–124. Parisi, Francesco, and Vincy Fon. 2005. “The behavioral foundations of retaliatory justice.” 7 Journal of Bioeconomics 45–72. Pasachoff, Naomi E., and Robert J. Littman. 2005. A Concise History of the Jewish People. Lanham: Rowman and Littlefield Publisher Inc. Posner, Richard A. 1980. “A theory of primitive society, with special reference to law.” 23.1 Journal of Law and Economics 1–53. Posner, Richard A. 1988. Law and Literature: A Misunderstood Relation. Cambridge, MA: Harvard University Press. Reid, John P. 1970. A Law of Blood: The Primitive Law of The Cherokee Nation. New York: New York University Press. Richter, Daniel. K. 1992. The Ordeal of the Longhouse: The Peoples of the Iroquois League in the Era of European Colonization. Chapel Hill: University of North Carolina Press. Roth, Martha T., trans. 1995. Law Collections from Mesopotamia and Asia Minor. Atlanta: Scholars Press. Santalucia, Bernardo. 1989. Diritto e processo penale nell’antica Roma. Milan: Giuffrè Editore. Scherillo, Gaetano, and Franco Gnoli. 1994. Lezioni sulle obbligazioni: corso di diritto romano. Milan: Monduzzi Editore. Scott, Samuel P., trans. and ed. 1910. The Visigothic Code. Boston: Boston Book Company. Shalom, Paul. 1970. Studies in the Book of the Covenant in the Light of Cuneiform and Biblical Law. Leiden: Brill. Simons, Kenneth W. 2007–8. The crime/tort distinction: legal doctrine and normative perspectives. 17 Widener L. J. 719–32.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

378 Francesco Parisi, Daniel Pi, Barbara Luppi, and Iole Fargnoli Stein, Peter. 1999. Roman Law in European History. Cambridge: Cambridge University Press. Sulzberger, Mayer. 1915. The Ancient Hebrew Law of Homicide. Philadelphia: Julius H. Greenstone. Talamanca, Mario. 1979. “Obbligazioni (dir. rom.),” in Enciclopedia del diritto, vol. XXIX. Milan: Giuffrè Editore, 2–78. Talamanca, Mario. 1990. Istituzioni di diritto romano. Milan: Giuffrè Editore. Tonry, Michael. 2008. Crime and Justice: An Annual Review of Research. Chicago: University of Chicago Press. Webster, Cheryl M., and Anthony N. Doob. 2012. “Searching for Sasquatch: deterrence of crime through sentence severity,” in Joan Petersilia and Kevin Reitz, eds, Oxford Handbook on Sentencing and Corrections. New York: Oxford University Press, 173–95. Weinstein, Mark I. 2001. “Limited liability in California: 1928–1932.” Paper No. 00–17 USC Olin Research.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

23 Collective Responsibility Thomas J. Miceli

Collective guilt, whatever form or justification it takes, has to be one of the most evil notions the human race has concocted, most likely the cause of more suffering of innocents than any other vile belief in history. (Simic 2013: 22) The principle of collective responsibility—so abhorrent to modern sensibilities—may be efficient in the conditions of primitive society. (Posner 1983: 194)

23.1. INTRODUCTION Economic models of law enforcement beginning with Becker (1968) have focused on the goal of identifying and apprehending the perpetrator of a crime based on a principle of “individual responsibility.” Throughout much of history, however, society operated under a much different system that attributed responsibility for wrongdoing to an entire community or group of which a wrongdoer was known to be a member. Indeed, examples of “collective responsibility,” or group punishment, are pervasive in ancient societies as well as biblical stories and mythologies. And surprisingly, the practice persists in various forms in modern society, such as when an entire class is punished for the actions of a prankster, a corporation is held liable Thomas J. Miceli, Collective Responsibility In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0023

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

380

Thomas J. Miceli

for the actions of its employees, or a country is sanctioned for terrorist acts by one of its citizens. This chapter provides an economic analysis of the choice between individual and group punishment and derives the conditions under which the latter is an optimal strategy.¹ According to the theory, the chief gains from group punishment are the certainty of punishing the actual offender plus the saved detection cost, while the primary drawback is aversion to wrongfully punishing innocent members of the target group. The use of group punishment has therefore varied over time and circumstance based on differing perspectives on the attribution of guilt and the perceived cost of wrongful punishment. Several important examples of collective responsibility are found in Roman law, as described in §23.2. §23.3 offers other examples from the ancient world, including the Bible, and §23.4 provides some modern uses of the practice. §23.5 then develops the theoretical analysis and derives the principal conclusions, the formal details of which are provided in the Appendix. Finally, §23.6 concludes.

23.2. EXAMPLES FROM ROMAN LAW Probably the best example of collective responsibility in Roman law is the Senatus Consultum Silanianum, a resolution passed by the Roman senate around 10 CE concerning the treatment of the slaves of a murdered master. The resolution provided that any slaves under the roof of the master at the time of his murder, or accompanying him if he was murdered elsewhere, were implicated in his death, with the punishment being torture, or, if it was determined that the slaves had not done enough to prevent the murder, death. However, a slave who discovered his master’s murderer was given his freedom. The early third-century jurist Ulpian explains that masters could not otherwise be safe unless slaves were compelled to provide protection at the risk of their own lives against threats from within the household and from ¹ Previous economic models of individual versus group punishment are found in Heckathorn (1990); Parisi and Dari-Mattiacci (2004); and Miceli and Segerson (2007). Informal treatments of this question from a legal and economic perspective include Posner (1983); Levmore (1995a, 1995b); and Levinson (2003). For a philosophical perspective on collective responsibility, see the various readings in May and Hoffman (1991).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

381

without (D. 29.5.1 pr.; Ulp. 50 ad ed.). Likewise, Modestinus specifically imposed on slaves the duty of protecting a master with weapons, their hands, shouting, and throwing their own bodies in the way; if they failed in this duty, capital punishment would rightly be exacted from them (D. 29.5.15; 8 pandect.). In addition, the law provided that the will of a slain person could not be opened until the murder had been investigated; heirs who did not avenge (through a criminal case) a murder would lose their inheritance, which would be confiscated by the imperial treasury. Much of the juristic discussion as it pertains to slaves concerns defining when they were in a position to come to the aid of their master, what type of aid certain slaves, such as women, children, or the infirm, would be expected to provide, and under what circumstances they might be exempt from questioning under torture and punishment. Punishments for slaves who failed to protect a master apparently were first established in a law of Sulla under the Republic. The senate in 57 CE revised the SC Silanianum by subjecting slaves who had been freed in the master’s will to its sanctions (Tac., Ann. 13.32). The sanctions of the SC Silanianum applied no matter who killed the slaves’ master, but the notorious killing by his own slave of the city prefect Pedanius Secundus in 61 CE touched off a riot when the senate debated the execution of the numerous slaves in his household (Tac., Ann. 14.42–5). Tacitus depicts a forceful speech by the jurist C. Cassius Longinus, who claimed that the four-hundred-odd slaves in his household had failed to protect Pedanius, and that without the implementation of the death penalty, no senator would be safe. His view prevailed in the senate, although there was significant opposition.² Pliny the Younger likewise reports about a vigorous debate in the senate on the fate of the freedmen of the senator Afranius Dexter, who had died under suspicious circumstances, possibly by his own hand (Ep. 8.14). The senate was divided on whether the freedmen should be killed, relegated (a form of exile), or freed; although Pliny does not explicitly state this, the penalty was likely to have been relegation.³

² For the details of the laws and the ancient sources, including D. 29.5 and C. 6.35, see Kaser (1971: 283 and n. 3, 293 and n. 3), as well as Sherwin-White (1985: 463). For discussion, see Gardner (2011: 430–1); Harries (2013). ³ Harries (2013: 66).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

382

Thomas J. Miceli

Another form of collective responsibility in Roman law is related to delicts, which represented the principal form of wrongfully imposed harm. The delict is most similar to a tort in modern law in the sense that enforcement was by a private action, and the usual remedy was payment of damages by the responsible party.⁴ The tort analogy is not exact, however, as in some cases the assessment of damages had more of the character of a criminal fine (or punitive damages) than of mere indemnification for the victim’s losses (du Plessis 2015: 325). Indeed, Buckland et al. (1952: 344–5) note that the “[d]elict is imbued with the idea of vengeance,” prompting the authors to argue that “[t]he Roman law of delict has far more affinity to the criminal law than to the law of tort . . .”; du Plessis (2015: 325) makes a similar argument. Roman law did include a distinct concept of crime (crimen), which gave rise to a criminal (public) rather than a civil action, but to modern eyes “the conceptual difference between some delicts and public offenses would have been hard to establish” (Harries 2007: 58), with the two concepts apparently lying on a continuum that depended on the seriousness of the act (Watson 1991: 212). Many offenses that are nowadays considered crimes would have been classified as delicts in Roman law, as is apparent in the cases of theft (furtum), robbery and violence (rapina), as well as in some instances of damage to property (damnum iniuria datum), and personal injury or insulting behavior (iniuria).⁵ In early Roman law, criminal offenses were generally limited to truly public harms like treason and sacral wrongs. Only later were murder and other violent acts included as crimes, but even in genuine criminal cases, private prosecution remained the rule (Zimmermann 1996: 918). The fuzzy line between crimes and torts in Roman law is a common feature of ancient legal systems (du Plessis 2015: 325). The actual perpetrator of harm was responsible under the delict, but there are elements of collective responsibility in a related category of obligation in Roman law, namely the so-called quasi-delict. The term quasi-delict refers to remedies that were created to address situations in which harm arose “as if from a delict,” but the remedies available in the actions on delicts did not apply because of a lack of direct fault on the part of the responsible party (Watson 1991: 75; Cursi 2002). Examples included liability of owners of dwellings from ⁴ See Friedman in this volume. ⁵ See, generally, Watson (1991); du Plessis (2015), ch. 10.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

383

which something was thrown or poured so as to cause damage to passersby; liability of ship’s captains, innkeepers, or stable owners whose employees caused harm or committed fraud; and liability of judges who, through breach of their official duty, caused damage to others.⁶ In all of these cases, the person being held liable need not have committed or authorized the harmful act himself, or even have known about it, to be found responsible for the damages. One interpretation of quasi-delicts is that they represent a form of strict liability on the part of individuals who are “in control of a potential source of danger to other peoples’ lives, health and property” (Zimmermann 1996: 17). In this sense, quasi-delicts share elements with the modern concept of vicarious liability (du Plessis 2015: 359). The related concept of noxal liability provided a way of establishing liability for delicts committed by children in the power of a paterfamilias (head of the household) or his slaves. Originally, the paterfamilias could discharge any liability for such wrongs by surrendering his dependent (noxal surrender), but in classical law the paterfamilias could avoid this step by compensating the injured party. The surrender option thus effectively limited the amount of the paterfamilias’s liability, though if the paterfamilias himself had been involved in the wrong, he usually forfeited his right of surrender.⁷ Parisi (2001: 112–13) argues that noxal surrender represented a kind of transitional phase from a retaliatory, or talionic, law enforcement regime (“an eye for an eye”) to one based on compensation.⁸ Generally, all cases of quasi-delict and noxal liability reflect a separation of responsibility and fault, which is a distinguishing feature of group punishment and vicarious liability. As will be discussed in detail below, there are often good reasons for this, including the “practical inability to get evidence of fault,” and “the need to get a better defendant, for the workman in charge of dangerous things is not likely to be able to compensate for the damage he does” (Buckland et al. 1952: 397). A final example of group responsibility in Roman times is found in the practice of decimation, which involved the random execution of one in ten soldiers as a means of maintaining military discipline.⁹ The ⁶ For example, a judge “become[s] (emotionally) so entangled in the case that he lacks the necessary impartiality . . .” (Zimmermann 1996: 16 n. 87). ⁷ Watson (1991: 75); Zimmermann (1996: 916–17). See also Kaser (1971: 631–2). ⁸ Also see the chapter by Parisi et al. in this volume. ⁹ Lintott (1999: 41–2).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

384

Thomas J. Miceli

Italian army employed a similar procedure in World War I to punish companies that broke and ran during combat, a practice described by Ernest Hemingway in his autobiographical novel about World War I, A Farewell to Arms. The example of decimation differs from vicarious liability, however, in that there is a sense in which all members of the targeted group are equally responsible, but punishing each one of them would obviously be impractical in a military context. As will be seen below, however, random individual punishment is theoretically equivalent to a scaled-down version of group punishment, and therefore has many of the same attributes (both positive and negative).

23.3. OTHER ANCIENT EXAMPLES As several scholars have pointed out, a defining feature of ancient law enforcement was its reliance on collective or group responsibility (Posner 1983:193–5; Parisi and Dari-Mattiacci 2004). If a person committed a harmful act, the victim’s kinship group often retaliated against the injurer or his kinship group. Even states as civilized as ancient Greece believed that a murderer “polluted” his city and family, thereby creating a form of guilt that fell upon all citizens of the city as well as descendants of the family (Posner 1983: 217). Of course, the most famous example of this is the pollution of Thebes by Oedipus’ murder of his father, as depicted in Sophocles’ play Oedipus Rex. Two more modern examples of the same concept are found in Samuel Taylor Coleridge’s poem, “The Rime of the Ancient Mariner,” which, according to Kitson (1989), symbolizes collective guilt for the unrealized promise of the French Revolution; and the work of Nathaniel Hawthorne, which reflects “the theme that the sins of the father are visited upon the children” (Buel 1986: 360). (Hawthorne, of course, believed that he carried the guilt of his forebear, who was a judge at the Salem witch trials.) The Hammurabi Code contained a provision that allowed a victim of robbery, if unable to identify the guilty party, to seek compensation from the city in which the robbery occurred (Levmore 1995a: 117). Likewise, the kinsmen of a murder victim could seek compensation from the city in which the murder took place. Old English law contained a similar provision, as chronicled by Blackstone (1766, Book 3: 161) in his Commentaries, that obliged residents of a village

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

385

in which a man was robbed to “make hue and cry after the felon,” and if they failed to identify him, the residents were collectively liable for “damages equivalent to his loss.”¹⁰ The Bible includes numerous allegories centering on group punishment. The stories of Noah’s flood (Genesis 7) and Sodom and Gomorrah (Genesis 18–19) dramatically portray the idea that the sins of a few cast guilt upon an entire community for which all citizens, the innocent and guilty alike, must pay. (Although, as Levmore 1995a: 95 points out, if all citizens were truly guilty, then these are not really examples of collective responsibility.) The Old Testament also includes an example of random punishment in the story of Jonah, who boards a ship to escape God’s command to go to the wicked city of Nineveh to serve as a missionary. As punishment for Jonah’s defiance, God creates a storm that threatens to destroy the ship and its crew. The mariners then draw lots to “detect” the person whose transgression was responsible for the storm, and when Jonah is miraculously identified as the guilty party, he is thrown overboard, where he is swallowed by a whale (Jonah 1:1–17). The fact that Jonah was correctly identified as the wrongdoer by the lottery is clearly meant to convey to readers the idea that the truly guilty can never escape the ever-present surveillance of God, which, in a world where detection and punishment of criminals is difficult, provides an important source of social control.¹¹ The pervasiveness of group punishment in ancient society reflects two characteristics of law enforcement in that era: the idea of punishment as revenge (Holmes 1881 [1963]: 6; Posner 1983: ch. 8),¹² and the absence of an effective centralized enforcement mechanism. The former requirement meant that someone had to pay in order to satisfy and/or compensate victims and their sympathizers

¹⁰ Also see Feinberg (1991: 65–6). ¹¹ On the role of religion in monitoring behavior, see Wade (2009); Johnson (2016). A modern version of this same idea is found in Adam Smith’s construct of an “impartial spectator,” which he argued was the basis for moral behavior by individuals (Smith 1759 [1976]). ¹² As Parisi and Dari-Mattiacci (2004: 497) note, “Several ancient rules of talionic justice require that the penalty be imposed with uncompromising symmetry, replicating the harm suffered by the victim and his or her family.” For discussion of the circumstances in which collective sanctions could play a positive role in an ancient polity, see Lanni (2017).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

386

Thomas J. Miceli

(a retributive motive), while the latter reflected the impracticality of identifying and exacting punishment or compensation from the actual offender.

23.4. EXAMPLES FROM MODERN LAW Although collective responsibility is commonly associated with ancient law enforcement, vestiges of it remain in contemporary doctrines and practices.¹³ The principal objection to collective responsibility is the idea that innocent parties are punished. Indeed, Parisi and Dari-Mattiacci (2004: 502–3) link the evolution from group to individual responsibility in the Old Testament to the emergence of the idea that “the collective guilt referred to in the earlier sources was no longer the basis of the communal responsibility for wrongs. An individual’s sin did not place any blame on the clan or class to which the wrongdoer belonged.” Modern examples of group responsibility thus tend to involve situations in which “innocent” members of the target group are nevertheless seen as being at least partially culpable for the offender’s action. In other words, the cost of punishing them is not seen as being high. In that case, the savings in detection costs under group punishment make it an attractive strategy. A timely example of this view in modern times is the imposition of sanctions, or the waging of warfare, against nations known to harbor terrorists or other wrongdoers. Historical examples of this same strategy include the attack on Troy by the Greeks in retaliation for the kidnapping of Helen (as chronicled in the Iliad);¹⁴ the policy of total destruction wreaked by Union general William T. Sherman in his infamous “march to the sea” during the American Civil War; the declaration of war against Germany and Japan by the United States following the bombing of Pearl Harbor (which included the bombing of cities); and the resulting internment of Japanese Americans during that war. As Levmore (1995a: 99) notes, “We are accustomed to the fact that war claims the lives of innocent people,” and “one nation ¹³ See, for example, the surveys in Levmore (1995a, 1995b); and Levinson (2003). ¹⁴ Archaeologists have concluded that the mythological story of the attack of Troy was likely based on an actual historical event (Wood 1998).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

387

may be justified in warring against another when the other nation is unable to control the actions of some of its inhabitants.” In these cases, citizens of the target country are viewed as somehow collaborating (or at least sympathizing) with the wrongdoer and are therefore partially culpable. Most modern examples of group responsibility, however, are found in the context of tort law, and can be justified in various ways. One is as a form of insurance for liability, which spreads the cost of accidental harms across an injurer’s risk group (Levinson 2003: 371). In this case, individuals pay premiums in proportion to their expected contribution to the risk of harm, and, in this sense, are “punished” appropriately in relation to their individual “guilt.” Thus, there is no perceived cost of erroneous punishment, and, as long as premiums are calibrated to individual risk (i.e. are actuarially fair), there is no loss of deterrence. However, if such calibration is costly or impossible, deterrence may be sacrificed owing to the usual moral hazard problem (though collective governance within the group, or by the insurer, can help to restore incentives) (Levinson 2003: 372). Group monitoring of individual members is undoubtedly the principal justification for other recognizable forms of group punishment in modern law. May and Hoffman (1991: 1) call this the “conspiracy model” of collective responsibility. Examples include various forms of vicarious liability like the doctrine of respondeat superior, under which an employer can be held responsible for torts committed by his employees; the related criminal law rule that a corporation can be held criminally liable for crimes committed by its employees;¹⁵ and joint and several liability, under which any one of a group of injurers can be held responsible for the victim’s entire loss.¹⁶ These rules help to promote deterrence in those circumstances where offenders are out of reach of victims or law enforcers, but are under the surveillance of other members of the group, who, it is hoped, will either impose their own form of discipline or turn the offender over to plaintiffs (Varian 1990). The Roman example of the Senatus Consultum Silanianum is a good example of this, as those accompanying a murder victim are

¹⁵ As Garoupa (2000: 244) notes, “Within the context of corporate liability, shareholders become quasi-enforcers.” ¹⁶ Menell (1991: 109) notes that under joint and several liability for environmental harms, “disposal of a thimbleful of hazardous waste at a large disposal site exposes an entity to enormous potential liability.”

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

388

Thomas J. Miceli

likely in a good position to identify, and perhaps thwart, the offender. In addition, as noted above, vicarious liability rules allow victims (or prosecutors) to pursue “better defendants” (i.e. deeper pockets) in the sense of having sufficient resources to fully compensate victims. In discussing the information revelation function of group punishment, Levmore (1995b) describes the well-known case of Ybarra vs Spangard,¹⁷ in which a court allowed a plaintiff to jointly collect from several health care professionals for harm suffered during an operation. This case reflects a general class of torts in which courts apportion damages in the presence of uncertainty over causation (Shavell 1985), but it is unique in the fact that the members of the group held liable presumably knew which of them actually caused the injury. The court’s reasoning in holding them all liable was that the absence of proof that any one person’s negligence caused the accident, which ordinarily would have resulted in a finding of no liability, should not be grounds for denying any compensation to the victim, especially given that the absence of evidence was apparently due to the unwillingness of the parties to testify against one another. The threat of group punishment in such a setting may therefore be an effective way to induce one or more of the members to reveal the true offender’s identity. At the same time, failure of any to do so, thus triggering group responsibility, would not be seen as imposing wrongful punishment because, as in the terrorism example, they are all then seen as “harboring the guilty party,” which is itself a form of guilt. Levmore (1995b) further speculates on why the court did not impose liability on each member of the group in excess of the harm suffered by the victim so that even the guilty party would have had an incentive to confess. For example, if the total harm was $500, then threatening to impose damages of, say, $600 on each person in the operating room would presumably induce the responsible party to confess so that he or she would only have to pay $500. In other words, the guilty party would have saved $100 by confessing. Such an “overextraction” rule would, in addition, give other members of the group an enhanced incentive to turn over the offender. The obvious disadvantage of such a strategy is that it potentially imposes high error costs if the scheme fails to reveal the offender. Further, as Levmore

¹⁷ 154 P.2d 687 (Cal. 1944).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

389

(1995b: 1595) observes, the offender may anticipate that society would not tolerate the actual imposition of such disproportionate penalties on innocent offenders, making the scheme non-credible. Indeed, it is this intolerance of wrongful or disproportionate punishments that makes group punishment objectionable in all but a few settings nowadays, as described above.

23.5. THEORETICAL ANALYSIS This section describes the results of a simple theoretical model devoted to answering the question of when group punishment is preferred, on economic grounds, to individual punishment. (The formal derivations of most of the results are contained in the Appendix.) The analysis is based on the following scenario. A harmful act has been committed by a single individual who is known to be a member of a group of size n. In other words, n is defined to be the smallest group of which the offender is known to be a member with certainty. The enforcement authority—whether the victim, the victim’s kinship group, or the state—considers two punishment strategies: (1) seek out and punish a single member from the group (individual punishment), or (2) impose a uniform sanction on the whole group (group punishment). The enforcer’s objective in choosing between these two strategies is taken to be maximization of the net benefits from punishment, which consist of the gains from correctly punishing the true offender, less the costs of punishment (including the cost of wrongful punishment, if any) and detection. The gains from punishment may come from various sources, including optimal deterrence, retribution, and/or compensation of victims.¹⁸ Whereas optimal deterrence is the primary motivation in economic models of crime (Becker 1968; Polinsky and Shavell 2000), ancient societies seem also to have valued proportionality of punishments to crimes (based on the talionic dictum of “an eye for an eye”), which originally had its origins in notions of vengeance (Posner 1983; Parisi 2001). But even in modern times, there seems to be a strong inclination toward proportional penalties, based primarily on notions ¹⁸ Parisi et al. (in this volume) specifically discusses the goal of deterrence in ancient law.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

390

Thomas J. Miceli

of fairness, regardless of whether or not they further the goal of deterrence.¹⁹ The basic conclusions of the following analysis will not, however, depend on the specific motivation for punishment.²⁰ In comparing the two punishment strategies, the primary trade-off is between certainty of punishing the true offender and the costs of erroneously punishing the innocent. Thus, modern examples of group punishment seem to be found in contexts where wrongful punishment is not a large concern, either because the sanction is not really deemed to be punishment (as in the case of the insurance examples), or because all members of the target group are seen as at least somewhat culpable for the underlying offense (as in the cases of military actions). To the contrary, in cases of true criminal punishment, especially when imprisonment is involved, the predominant reliance in modern society on individual punishment reflects an overriding concern for avoiding wrongful punishment. Indeed, the significant safeguards of criminal defendants’ rights in the Anglo-American and other modern criminal justice systems are a manifestation of this social value. The other factor that has contributed to (or, more properly, allowed) the rise of individual criminal punishment, of course, is the emergence of centralized enforcement authorities (principally governments) and the development of technologies (like fingerprinting and DNA) that allow the (reasonably) accurate detection of offenders. Prior to the emergence of a “science of detection,” individual punishment, if employed, often amounted to little more than random punishment, as in the Roman practice of decimation (which, as noted, continued to be used by the Italian army into the twentieth century). As shown in the Appendix, however, if there are constant or increasing returns to scale in punishment costs (including error costs), then random individual punishment—that is, punishing one

¹⁹ For example, one of the key prescriptions of Becker’s model, which is based solely on deterrence, is that punishments should be raised to the maximum extent possible, and apprehension probabilities scaled back correspondingly, so as to save on overall enforcement costs. Actual punishment schemes, however, rarely if ever implement such a policy. On the inclusion of fairness in Becker’s model, see Miceli (1991). ²⁰ See Miceli and Segerson (2007) for a formal analysis that explicitly distinguishes between deterrence and retribution as motives and shows how they affect the choice between individual and group punishment. The principal conclusion is that when retribution is the primary goal, either punishment strategy may dominate, but when deterrence is the goal, individual punishment dominates.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

391

randomly selected individual from the group—is dominated by group punishment. The reason is that, if punishing one person in the group is socially advantageous (i.e. if it yields a net social benefit), then punishing all members must yield at least n times that benefit. This logic is most easily seen in the case where the enforcer incurs no punishment costs. When that is true, it makes no sense to punish only one member of the group since by punishing all of them, the true offender is punished with certainty, thereby achieving the desired level of deterrence, revenge, or incapacitation. Adding punishment costs does not alter this conclusion as long as the gain from punishment exceeds the costs on a per-person basis as the scale of punishment is increased. Random punishment does, however, possibly dominate group punishment if punishment costs are increasing at the margin. Thus, in cases where detection is either very difficult, or where all group members are seen as culpable but there are decreasing returns to scale in punishment costs, random individual punishment may make sense. The case of decimation is a good example of this scenario since the imposition of some punishment no doubt provided a strong deterrent against mutiny or flight in battle, but execution or incapacitation of all members of an offending company would have been impractical for obvious manpower reasons. The other situation where group punishment possibly offers strong advantages, as noted above, is when the members of the group can thereby be induced to reveal the identity of the true offender, or to impose their own form of punishment. In this case, the threat of group punishment effectively enlists the members of the group as surrogate enforcers, thus minimizing the cost of enforcement to the central authority. Economic theory provides further insight into this particular motivation for group punishment. As shown in the Appendix, the size of the group likely has an important impact on the efficacy of group punishment for inducing the group to impose internal sanctions, but the effects are cross-cutting. On one hand, it would not seem to be effective if the offender is a member of a large group because then it would be fairly easy for the latter to remain anonymous to other members, and free-rider problems would preclude them from investing much effort in identifying him. Based on this logic, Parisi and Dari-Mattiacci (2004: 504) argue that growth in the size of communities over time “might explain the historical evolution towards systems of individual responsibility and in general towards

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

392

Thomas J. Miceli

the limitation of the size of the group that was to be held collectively responsible for the wrongs of its members.” On the other hand, small groups would likely have significant group solidarity, and therefore would be reluctant to turn over one of their own members. Thus, for example, sanctioning of fraternities for campus pranks would not likely be an effective way to discover the offender’s identity (as exemplified by the movie Animal House).²¹ Another example, suggested by Akerlof and Yellen (1994) in the context of gang behavior, emphasizes the importance of a community’s willingness to cooperate with police in identifying offenders. In this case, the use of group sanctions may significantly inhibit such cooperation by fostering an “us vs them” attitude among gang members. (Although in this case, members of the group in some sense inherit the offender’s guilt by their refusal to “out him.”) In the same vein, some contend that harsh sanctions against countries harboring terrorists might actually incite more terrorist acts. Taken together, these arguments suggest that the information-revealing function of group punishment might best be implemented for moderately sized groups and those with memberships that are not long-lasting so as to avoid the emergence of strong intra-group solidarity.²² The use of vicarious liability in Roman law and the modern manifestations of this practice represent a closely related form of group responsibility. Under this strategy, a principal (an employer or head of a household) is sanctioned in the hopes that he or she will pass the sanction on to an offending agent who is either in the principal’s employ or is otherwise under his or her control (as in the case of noxal liability). As Levinson (2003: 349) notes, “Vicarious liability might be understood as a special type of collective sanction regime, characterized by a target group consisting of (only) two individuals bound together in a contractual relationship.”²³ In

²¹ But see Feinberg (1991: 61), who argues that collective responsibility would be most effective in groups with a “high degree” of solidarity because such groups would be more likely to share common interests and thus to suffer sanctions in common. Although this might promote mutual restraint before an act is committed, it will also likely prevent the group from turning over a guilty party after the fact. ²² See Heckathorn (1990) for a formal analysis of intra-group control norms in the presence of group sanctions. ²³ For a formal model of this in an environmental context, see Segerson and Tietenberg (1992). For more general treatments, see Kornhauser (1982); Sykes (1984); Feinberg (1991); and Dari-Mattiacci and Parisi (2004).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

393

addition to providing incentives for offenders to refrain from harmful acts, vicarious liability can also increase the chances of obtaining compensation for victims (when that is one of the social goals) by correcting for judgment-proof problems, such as when employees, minor family members, animals, or slaves lack the personal resources to cover any damages that they impose (Parisi 2001: 111).²⁴ An offsetting effect, noted by Arlen (1994) in the context of corporate liability, is that when corporations face the threat of vicarious liability they might actually refrain from policing their employees for fear that doing so will increase the chances that crimes will be uncovered, thereby potentially raising the company’s expected liability.²⁵ Notwithstanding the foregoing benefits from group responsibility, it is undoubtedly the case that the primary objection in modern eyes to this form of punishment in most situations is the attribution of guilt to all members of the group. In addition to the stigma that this creates, especially when the punishment is criminal in nature, there is a risk that it might justify other forms of punishment or discrimination against the group. For example, the practice of statistical discrimination (or profiling) based on race, gender, or ethnic background is a form of group punishment that may be justified on purely efficiency grounds due to imperfect information, but is in many (though not all)²⁶ cases viewed as offensive and possibly unconstitutional (Posner 2003: 689–90). This has not always been the case, however, as exemplified by the example, noted above, of the internment of Japanese Americans during World War II, an action that was accepted as a necessary security measure at the time, but which would be unthinkable today, only two generations later. What these examples reveal is that the acceptability of group responsibility as an enforcement strategy depends on factors that vary with both time and circumstance, and reflect a balancing of social values about guilt and punishment. As Feinberg (1991: 67) observes, the general demise of collective responsibility has not occurred because “individual responsibility is an eternal law of

²⁴ On the judgment-proof problem in a tort context, see Shavell (1986); Miceli and Segerson (2003). ²⁵ For further discussion of collective responsibility in the context of corporations, see French (1991); Velasquez (1991). ²⁶ The setting of different auto insurance rates based on age and gender, for example, apparently remains acceptable.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

394

Thomas J. Miceli

reason,” but because the conditions that made it reasonable or necessary in ancient societies are rarely present in modern times.

23.6. CONCLUSI ON Collective responsibility for wrongdoing is a law enforcement practice that is often associated with ancient society, or at least with notions of vengeance, as exemplified by the infamous Hatfield–McCoy feud (Alther 2012). A careful examination of the practice, however, reveals that several forms of group punishment remained in use in highly civilized societies like ancient Greece and Rome, and continue to be used in various ways in modern society, both informally and doctrinally. The persistence of collective responsibility, or group punishment, suggests that the strategy must serve a useful law enforcement function in at least some circumstances. Prompted by this insight, this chapter has examined the economic logic underlying the use of group punishment by focusing on the costs and benefits of the practice in relation to individual punishment. The conclusions reveal that collective responsibility is often a rational social response to prevailing circumstances, such as when individual punishment is inefficient or impractical, or when members of the group are harboring the offending party or otherwise share in his or her responsibility for the wrongful act.

A P P E N DI X : A S I MPL E M O DE L O F I N DI V I DU A L VS GROUP PUNISHMENT This appendix derives the formal results that are referred to in the body of the text. The analysis will make use of the following notation: B = social benefit of punishing the true offender, reflecting either deterrence, retribution, or the desire for compensation of victims; k = social cost incurred for each person wrongfully punished; n = smallest group of which the offender is known to be a member, n > 1; p = probability of correctly apprehending the true offender; c(p,n) = cost of detection, where cp > 0, cpp > 0, cn > 0, and cpn > 0.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

395

The cost of detection function reflects the best available technology for identifying the true offender under a strategy of individual punishment. It seems plausible to assume that both the overall cost, and the marginal cost of increasing p (i.e. of achieving greater certainty), are increasing in the group size. Given this notation, we can write the expected social benefits from individual (I) and group (G) punishment as: WI ¼ pB  ð1  pÞk  cðp; nÞ;

ð1Þ

WG ¼ B  ðn  1Þk:

ð2Þ

For sake of comparison, we assume equal punishments under the two regimes, given that both B and k will depend on the nature and magnitude of the sanction.²⁷ In the case of individual punishment, the enforcer will ordinarily be able to choose p to maximize WI, which yields the first order condition: B þ k ¼ cp :

ð3Þ

According to (3), the optimal probability of apprehension, denoted p*, equates the marginal benefit of correct punishment, consisting of the direct benefit B plus the savings in error costs k, to the marginal cost. Totally differentiating (3) and using the assumption that cpn > 0 shows that p* is decreasing in n and increasing in k and B. Thus, under individual punishment, the enforcer devotes less effort to identifying the true offender as the group size increases, and more effort as the cost of erroneous punishment and/or the benefit of conviction increases. Using this model, we can prove several of the claims asserted in the text. Result 1: When the cost of erroneous punishment is zero, group punishment is preferred to individual punishment for any choice of p. This follows immediately from a comparison of (1) and (2) when k = 0. Intuitively, group punishment both ensures that the true offender is punished and saves enforcement costs compared to individual punishment. This establishes that the primary justification for individual punishment is the aversion to wrongful punishment.

²⁷ See Miceli and Segerson (2007) for a model that allows an endogenous choice of the punishment under each regime, and also formally distinguishes between deterrence and retribution as social goals of punishment.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

396

Thomas J. Miceli

One step on the road to individual punishment is punishment of a randomly chosen member of the group known to harbor the true offender. Result 2: Random individual punishment achieves the same level of social welfare, on a per-person basis, as group punishment. To prove this, note first that under random individual punishment, p = 1/n and c(1/n,n) = 0 (given that no enforcement costs are expended). Thus, (1) reduces to: WI R ¼ B=n  ð1  1=nÞk ¼ ½B  ðn  1Þk=n:

ð4Þ

Comparing (4) to (2) shows that WI R ¼ WG =n. Further, assuming that B  ðn  1Þk > 0; WG > WI R provided n > 1. It therefore follows that: Result 3: Group punishment is preferred to random individual punishment, provided that punishment is desirable at all. Thus, if random individual punishment is desirable in the sense that it yields a net social gain, then punishing all members of the group must yield proportionately greater social benefits, given constant returns to scale. The preceding conclusions imply that the benefits of individual punishment must come from optimal detection efforts. Indeed, when p is chosen optimally we have: Result 4: As the cost of erroneous punishment rises, individual punishment will eventually be preferred to group punishment. The effects of an increase in k on welfare under the two punishment regimes are given by: @WG ¼ ðn  1Þ < 0 @k

ð5Þ

@WI ¼ ð1  p Þ < 0; @k

ð6Þ

where the Envelope Theorem was invoked to obtain (6). From Result 1 above, we know that WG > WI when k = 0, but both expressions are decreasing in k. However, WG declines linearly whereas WI is convex given that p* rises with k. Thus, for large enough k, WI will cross WG from below. One claimed advantage of group punishment is that it may induce members of the group to identify the offender.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

397

Result 5: Social welfare from group punishment is increasing in the probability that the group will turn over the true offender. Let q be the probability that the group will identify the true offender when faced with group punishment. In that case, expected social welfare under group punishment becomes: WG ¼ B  ð1  qÞðn  1Þk:

ð7Þ

This expression is clearly increasing in q, reflecting the savings in the expected costs of erroneous punishment when the group may turn over the offender. Generally, we would expect q to be a function of the group size, but the sign of q0 (n) may be positive or negative. On one hand, q0 < 0 because an offender can avoid detection more easily in a larger group, but on the other, q0 > 0 if smaller groups have greater solidarity and hence are less willing to turn over one of their own. Thus, we have: Result 6: If q0 < 0, welfare under group punishment is decreasing in group size, but if q0 > 0, welfare under group punishment may be increasing or decreasing in group size. Differentiating (7) yields: @WG ¼ ð1  qpÞk þ ðn  1Þkq0 : @n

ð8Þ

The first term is the direct increase in error costs from an increase in n, while the second reflects the effect of n on q. If q0 < 0, this term reinforces the first term and the entire expression is negative, but if q0 > 0, this term offsets the first term and the entire expression is ambiguous in sign. It follows from these results that the most effective use of group punishment may be for moderately sized groups. One may also infer that group punishment would be most effective for groups that endure only temporarily so that members don’t have an opportunity to establish significant intra-group solidarity (in which case q would be high). Finally, suppose that punishment is costly to impose (apart from error costs), and let s be the social cost per person punished. Now (1) and (2) become: WI ¼ pB  ð1  pÞk  s  cðp; nÞ;

ð9Þ

WG ¼ B  ðn  1Þk  ns:

ð10Þ

Clearly, the higher cost of punishment under group punishment is now a significant disadvantage compared to individual punishment, all else equal.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

398

Thomas J. Miceli

This will be especially true if s is itself increasing in n; that is, if marginal punishment costs are increasing. In this case, random individual punishment is no longer simply equal to WG =n due to decreasing returns to scale in punishment costs. Rather, WI R > WG =n, and it is no longer necessarily true that group punishment dominates random individual punishment. Beyond these factors, the other results derived above for the case where s = 0 remain largely unaffected.

REFERENCES Akerlof, George, and Janet Yellen. 1994. “Gang behavior, law enforcement, and community values,” in Henry Aaron, Thomas E. Mann, and Timothy T. Taylor, eds, Values and Public Policy. Washington, DC: Brookings, 173–209. Alther, Lisa. 2012. Blood Feud: The Hatfields and the McCoys: The Epic Story of Murder and Vengeance. Guilford: Lyons Press. Arlen, Jennifer. 1994. “The potentially perverse effects of corporate criminal liability.” 23 Journal of Legal Studies 833–67. Becker, Gary. 1968. “Crime and punishment: an economic approach.” 76 Journal of Political Economy 169–217. Blackstone, William. 1766. Commentaries on the Laws of England, 1st edn. Oxford: Clarendon Press. Buckland, William W., Arnold McNair, and Frederick H. Lawson. 1952. Roman Law and Common Law: A Comparison in Outline. Cambridge: Cambridge University Press. Buell, Lawrence. 1986. New England Literary Culture. Cambridge: Cambridge University Press. Cursi, M. Floriana. 2002. Iniuria cum damno: Antigiuridictia e Colpevolezza Nella Storia del Danno Aquiliano. Milan: Giuffrè. Dari-Mattiacci, Giuseppe, and Francesco Parisi. 2004. “The cost of delegated control: vicarious liability, secondary liability and mandatory insurance.” 23 International Review of Law and Economics 453–75. du Plessis, Paul. 2015. Borkowski’s Textbook on Roman Law, 5th edn. Oxford: Oxford University Press. Feinberg, Joel. 1991. “Collective responsibility,” in May and Hoffman, eds, 53–76. French, Peter. 1991. “The corporation as a moral person,” in May and Hoffman, eds, 133–49. Gardner, Jane. 2011. “Slavery and Roman law,” in Keith Bradley and Paul Cartledge, eds, The Cambridge World History of Slavery. Cambridge: Cambridge University Press, 414–37. Garoupa, Nuno. 2000. “Corporate criminal law and organization incentives: a managerial perspective.” 21 Managerial and Decision Economics 243–52. Harries, Jill. 2007. Law and Crime in the Roman World. Cambridge: Cambridge University Press.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Collective Responsibility

399

Harries, Jill. 2013. “The Senatus Consultum Silanianum: court decisions and judicial severity in the early Roman Empire,” in Paul du Plessis, ed., New Frontiers: Law and Society in the Roman World. Edinburgh: Edinburgh University Press, 51–70. Heckathorn, Douglas. 1990. “Collective sanctions and compliance norms: a formal theory of group mediated social control.” 55 American Sociological Review 366–84. Hemingway, Ernest. 1929. A Farewell to Arms. New York: Scribner. Holmes, Oliver Wendell. 1881 [repr. 1963]. The Common Law. Boston: Little-Brown. Johnson, Dominic. 2016. God Is Watching You: How the Fear of God Makes Us Human. New York: Oxford University Press. Kaser, Max. 1971. Das römische Privatrecht. Erster Abschnitt: Das altrömische, das vorklassische und klassische Recht, 2nd edn. Munich: Beck. Kitson, Peter. 1989. “Coleridge, the French Revolution, and the ‘Ancient Mariner’: collective guilt and individual salvation.” 19 Yearbook of English Studies 197–207. Kornhauser, Lewis. 1982. “An economic analysis of the choice between enterprise and personal liability.” 70 California Law Review 1345–92. Lanni, Adriaan. 2017. “Collective sanctions in classical Athens,” in Dennis P. Kehoe and Thomas A. J. McGinn, eds, Ancient Law, Ancient Society. Ann Arbor: The University of Michigan Press, 9–31. Levinson, Daryl. 2003. “Collective sanctions.” 56 Stanford Law Review 345–428. Levmore, Saul. 1995a. “Rethinking group responsibility and strategic threats in biblical texts and modern law.” 71 Chicago-Kent Law Review 85–121. Levmore, Saul. 1995b. “Gomorrah to Ybarra and More: overextraction and the puzzle of immoderate group liability.” 81 Virginia Law Review 1561–1604. Lintott, Andrew. 1999. Violence in Republican Rome, 2nd edn. Oxford: Oxford University Press. May, Larry, and Stacey Hoffman, eds. 1991. Collective Responsibility: Five Decades of Debate in Theoretical and Applied Ethics. Savage: Rowan and Littlefield. Menell, Peter. 1991. “The limitations of legal institutions for addressing environmental risks.” 5 Journal of Economic Perspectives 93–113. Miceli, Thomas. 1991. “Optimal criminal procedure: fairness and deterrence.” 11 International Review of Law and Economics 3–10. Miceli, Thomas, and Kathleen Segerson. 2003. “A note on optimal care by wealth-constrained injurers.” 23 International Review of Law and Economics 273–84. Miceli, Thomas, and Kathleen Segerson. 2007. “Punishing the innocent along with the guilty: the economics of individual versus group punishment.” 36 Journal of Legal Studies 81–106.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

400

Thomas J. Miceli

Parisi, Francesco. 2001. “The genesis of ancient law.” 3 American Law and Economics Review 82–124. Parisi, Francesco, and Giuseppe Dari-Mattiacci. 2004. “The rise and fall of communal liability in ancient law.” 24 International Review of Law and Economics 489–505. Polinsky, A. Mitchell, and Steven Shavell. 2000. “The economic theory of public enforcement of law.” 38 Journal of Economic Literature 45–76. Posner, Richard. 2003. Economic Analysis of Law, 6th edn. New York: Aspen Publishers. Posner, Richard. 1983. The Economics of Justice. Cambridge, MA: Harvard University Press. Segerson, Kathleen, and Thomas Tietenberg. 1992. “The structure of penalties in environmental enforcement: an economic analysis.” 23 Journal of Environmental Economics and Management 179–200. Shavell, Steven. 1985. “Uncertainty over causation and the determination of civil liability.” 28 Journal of Law and Economics 587–609. Shavell, Steven. 1986. “The judgment proof problem.” 6 International Review of Law and Economics 45–58. Sherwin-White, Adrian N. 1985 [1966]. The Letters of Pliny: A Social and Historical Commentary. Oxford: Oxford University Press. Simic, Charles. 2013. “Oh, what a lovely war.” 40.15 New York Review of Books, 21–3. Smith, Adam. 1759 [1976]. The Theory of Moral Sentiments. Indianapolis: Liberty Classics. Sykes, Alan. 1984. “The economics of vicarious liability.” 93 Yale Law Journal 1231–80. Varian, Hal. 1990. “Monitoring agents with other agents.” 46 Journal of Institutional and Theoretical Economics 153–74. Velasquez, Manuel. 1991. “Why corporations are not morally responsible for anything they do,” in May and Hoffman, eds, 111–31. Wade, Nicholas. 2009. The Faith Instinct: How Religion Evolved and Why it Endures. New York: Penguin Books. Watson, Alan. 1991. Roman Law and Comparative Law. Athens: University of Georgia Press. Wood, Michael. 1998. In Search of the Trojan War. Berkeley: University of California Press. Zimmermann, Reinhard. 1996. The Law of Obligations: Roman Foundations of the Civilian Tradition. Oxford: Oxford University Press (first publ. Cape Town: Juta & Co., Ltd., 1990).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

24 The Dual Origin of the Duty to Disclose in Roman Law Barbara Abatino and Giuseppe Dari-Mattiacci

24.1. INTRODUCTION To guarantee the smooth functioning of markets, it is crucial that buyers have the opportunity to invoke legal protection if the goods they have purchased turn out to lack essential characteristics. Two different sets of officials, the aediles curules (having jurisdiction over regulated cattle and slave markets) and the praetor (having general civilian jurisdiction over contracts), developed remedies for nonconformity in sale contracts in Roman law. The aedilician remedies gave buyers a choice between an off-the-contract remedy, determining the restitution of both the good and the price paid (actio redhibitoria), and an on-the-contract remedy, affirming the contract while reducing the price (actio quanti minoris).¹ In contrast, the praetor initially provided and developed a single on-the-contract remedy (actio ex empto), affirming the contact but allowing the buyer to claim damages (damnum emergens, close to the modern reliance damages), which could possibly go beyond the restitution of the price paid.²

¹ For an economic analysis of modern on- and off-contract remedies, see Brooks and Stremitzer (2011, 2012). ² See Arangio-Ruiz (1956: 237–9 and 242–3); Talamanca (1990: 657–8). An actio empti ad redhibendum (or actio empti ad resolvendam emptionem, rescinding the contract) is first attested in the imperial period. We will provide details on this remedy below in the text. Barbara Abatino and Giuseppe Dari-Mattiacci, The Dual Origin of the Duty to Disclose in Roman Law In: Roman Law and Economics: Exchange, Ownership, and Disputes, Volume II. Edited by: Giuseppe Dari-Mattiacci and Dennis P. Kehoe, Oxford University Press (2020). © Oxford University Press. DOI: 10.1093/oso/9780198787211.003.0024

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

402

Barbara Abatino and Giuseppe Dari-Mattiacci

Under pre-Justinianic Roman law, aedilician and praetorian remedies were strikingly different along three additional dimensions: first, the aedilician remedies could be claimed only for a limited set of clearly enumerated defects, while the praetorian remedies were openended; second, the aedilician remedies applied irrespective of the fraudulent conduct used by the seller to deceive the purchaser (dolus), while the praetorian remedies required proof of dolus; third, the aedilician remedies were subject to short statutes of limitations— six months for the actio redhibitoria and one year for the actio quanti minoris³—while the praetorian remedies were not subject to prescription. In this chapter, after offering some details on the lawmaking institutions and the remedies (§§24.3 and 24.4), we will show that the differences in the remedies were designed to maximize the value of transactions. The aedilician remedies (actio redhibitoria and actio quanti minoris) maximized the value of a typical “market transaction” concluded between a professional seller and a private buyer in the markets that were under the jurisdiction of the aediles. In contrast, the praetorian remedy (actio ex empto) maximized the value of the typical “private transaction” concluded between two private parties outside a regulated market. In addition, the differences between these two sets of remedies allowed buyers to optimally choose between private and market transactions, thereby optimally opting into the two different sets of remedies and hence further enhancing their efficiency (§24.5). The availability of efficient boilerplate prescriptions reduces the need for ad hoc contractual arrangements and hence reduces transactions costs. We focus our attention on the city of Rome.⁴ The remedies we study developed at a time when the jurisdictions of the two magistrates were clearly distinct. The curule aediles were magistrates in charge of public order in markets where cattle and slaves were sold through auctions. As part of the cura annonae—concerning the smooth functioning of the markets and the distribution of foodstuffs—they had the power of iurisdictio and adjudication only concerning

³ These actiones could also be used to obtain a warranty against eviction from the seller (through stipulatio); in this case the terms were shorter. See Arangio Ruiz (1956: 367–8). ⁴ But see TPSulp. 42, l. l–2, and TPSulp. l. 3–4 with references to the aedilician edict, showing the validity of the remedies within the Italian territory.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

403

controversies arising in the regulated markets. Most likely, they developed the actio redhibitoria first (plausibly, in the mid-second century BC) and later the actio quanti minoris (plausibly before the first century BC).⁵ The praetor, by contrast, had general jurisdiction over contracts, not limited to specific types of transactions. The actio ex empto was possibly introduced as early as the third century BC.⁶ Later, in the Augustan period, the praetor also introduced an off-the-contract remedy, the actio ex empto ad redhibendum, including the possibility of obtaining restitution. By the time of Justinianic Corpus Iuris Civilis, both the aediles and the praetor gradually lost their initial independence in the administration of justice and the aedilician remedies were extended to all kinds of sales. As a result, the aedilician and praetorian remedies converged and blurred together. The timing of emergence and evolution of the different remedies, and especially the fact that in the aedilician remedies the on-the-contract remedy followed the off-the-contract remedy, will be shown to be consistent with the view that the remedies were designed to maximize the value of transactions. We will conclude with some considerations on why both the aediles and the praetor so effectively catered to the interests of the contracting parties to whom their remedies were addressed (§24.6). The core of our argument is as follows. Remedies for defects in sale contracts address a fundamental economic problem: the seller is usually better informed than the buyer about the characteristics of the goods for sale and might have incentives to conceal defects or misrepresent other characteristics. Both the aedilician and the praetorian remedies addressed this problem but did so in different ways. The aediles allowed buyers a remedy only if the seller had not provided the correct information about the quality of the goods. All buyers value goods of higher quality (for instance, a trustworthy slave) more than goods of lower quality (for instance, a slave prone to theft). Other idiosyncratic characteristics of the goods (for instance, specific slave skills, such as baking) that might be important for some individual buyers but not for others were left out of the reach of the aedilician remedies. This allowed for quick transactions (only information about quality was exchanged) and fast dispute resolution (only some predefined quality markers were subject to scrutiny ex post). In addition, the buyer did not need to prove the ⁵ See Donadio (2004: 40–54, esp. 45). ⁶ See Donadio (2004: 37–8 and 2007: 510–11). Cf. Watson (1987: 167–75).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

404

Barbara Abatino and Giuseppe Dari-Mattiacci

dolus (fraud) of the seller, which further simplified and accelerated dispute resolution. Finally, quality was relatively easy to verify after the purchase, so that the prescription for the aedilician remedies could be short. In turn, a short prescription reduced the cost of having some goods returned to innocent sellers—those who had misled the buyer in good faith. The downside of the aedilician remedies was that they did not allow for an effective exchange of information about idiosyncratic characteristics of the good prior to the contract, which had a negative effect on the value of the transaction in all those cases in which a buyer valued such characteristics. In contrast, the remedies applied by the praetor allowed for the verification of both quality and idiosyncratic characteristics but limited the liability of the seller to cases of dolus. This made trials potentially longer and more expensive but had two positive effects. On the one hand, sellers had incentives to reveal more information than under the aedilician remedies, which increased the value of the transaction for those buyers who valued idiosyncratic characteristics besides quality. On the other hand, sellers were not liable for innocent misrepresentation, which had a dampening effect on prices as the contract did not include an implicit insurance against such an eventuality. The praetorian remedies were not subject to prescription. This was likely due to the fact that some characteristics might become evident after a long time—hence the buyer needed to be given more time in order to have effective protection. Moreover, since goods were not returned to the seller under the actio ex empto, the absence of a prescription period did not expose the seller to resale risk. The measure of compensation could not be established in relation to a market price due to the private nature of the transaction; therefore, damages were calculated according to the buyer’s negative interest, that is, his situation before the contract. Typically, buyers involved in market transactions were most likely those with little interest in idiosyncratic characteristics, while those looking for specific slaves or cattle would probably self-select into private transactions. As a result, the transactions that fell under the jurisdiction of the aediles were those for which the aedilician remedies were most effective—and, hence, buyers had, even if given a choice, no incentives to opt for a praetorian remedy. Likewise, the transactions that fell under the jurisdiction of the praetor were those involving idiosyncratic buyers who valued the extended protection given by the praetorian law. In the following, we will elaborate upon and demonstrate these claims.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

405

24.2. INSTITUTIONS AND REMEDIES

24.2.1. Institutions The praetorship was created in 367 BC to take over some of the duties of the consuls. In exchange for opening the consulship to plebeians, the patricians obtained that the praetor would be a patrician. Yet, the first plebeian praetor was elected as early as 337 BC, so the praetorship cannot be defined as a patrician institution. The praetor had the power of iurisdictio, that is, to resolve disputes between litigants⁷ and of ius edicendi, that is, the power to issue an edict listing the remedies available to litigants.⁸ The aediles were lower-ranking plebeian magistrates created in 449 BC⁹ and elected by the plebeian council (concilia plebis) to collaborate with the main plebeian magistrates, the tribunes, in overseeing and policing markets. Two additional patrician aediles elected by the comitia tributa were added probably after 367 BC. Like the praetor, the aediles also held office for one year and their functions included the cura urbis (the management of the city roads, baths, and buildings), the cura ludorum (the management of the public games), and the cura annonae (the control and policing of the city markets).¹⁰ The aediles had limited coercive powers, including the power to issue fines and, within their prerogatives concerning markets, iurisdictio and ius edicendi, the power to resolve disputes and issue an annual edict listing remedies available to litigants. However, as opposed to the praetor’s edict, the edict of the aediles was limited to the remedies necessary to resolve the disputes arising in the markets under their jurisdiction.¹¹ Therefore, the aedilician remedies and, in particular the actio redhibitoria, were created as specific solutions for cases in which the seller had failed to disclose a particular defect (vitium) to the buyer ⁷ The praetor urbanus had jurisdiction over disputes between Roman citizens. Proceedings before the praetor were in iure: the praetor would typically give the parties a formula, stating the parties’ claims and instructing the iudex to verify the facts and award a remedy accordingly. From the third century BC, a praetor peregrinus with jurisdiction over disputes involving foreigners was added to the praetor urbanus. ⁸ See de Ligt in volume I. ⁹ See the Lex Valeria Horatia de tribunicia potestate. ¹⁰ On permanent and periodic markets, see de Ligt and de Neeve (1988: 391–416); de Ligt (1993). Cf. also Lo Cascio (2000). ¹¹ On the aedilician edict as portio iuris honorarii, see Guarino (1955, 1956). Cf. Volterra (1955, 1956). Recently see Chevreau (2012: 225–9).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

406

Barbara Abatino and Giuseppe Dari-Mattiacci

prior to a sale effected in markets where cattle and slaves were traded.¹² Sales in these markets were typically executed through an open ascending price auction (now known as an English auction): goods were offered for sale by an auctioneer (praeco) for a base price and assigned to the highest bidder.¹³ The auctions were announced in advance (TPSulp. 83–6, 88, 90–3)¹⁴ and, at least at Pompeii and Puteoli in the early imperial period, organized by a banker (argentarius or coactor)¹⁵ who could also lend money to the buyer.¹⁶ If the banker was hired by the seller—normally through a contract of hire, locatio conductio¹⁷—he could also guarantee (promissio auctionatoris; Donadio 2011) the payment of the price after having deducted the expenses for the organization of the auction (TPSulp. 81; D. 46.3.88.5). As with the praetor, the edict of the aediles also gradually lost importance during the imperial period.

24.2.2. Remedies It is not known when exactly the aedilician remedies were introduced. However, it seems certain that the aediles introduced first the actio redhibitoria (most likely before the second century BC; Donadio 2004: 45) and later the actio quanti minoris (most likely before the first century BC; Cic., De off. 3.17.71 and 3.23.91; Jakab 1997: 126–7). Both remedies imposed on the seller a duty to disclose whether the good for sale was affected by any of the defects indicated in the aedilician edict. Failure to do so gave rise to the buyers’ right to ask for a remedy. Note that not all defects were relevant but only those listed, ¹² See Serrao (2000); cf. Donadio, (2007: 464 and n. 11) with additional literature references. On the economic analysis of the aedilician remedies, see Kupisch (2002: 21–54, esp. 41 n. 74). ¹³ On modern auction theory, see McAfee and McMillan (1987); Krishna (2002). On the function of auctions in ancient Rome, see Malmendier (2002: 101–5); cf. Ankum (1972: 377–93). ¹⁴ TPSulp. 81 describes a case in which the intermediary lends money to the buyer to pay the seller. TPSulp. 90–3 documents instead cash payments by the buyer. For epigraphic evidence, see Gröschler (1997: 35–8); Camodeca (1999: 185–206). Cf. Bove (1975: 322–31); Wolf (2010: 105–21). ¹⁵ See Andreau (1974: 77–81). ¹⁶ See the apochae Iucundianae (e.g. CIL IV 3340, 1–137) and TPSulp. 82. For more details, see Camodeca (1999: 185); Lerouxel (2016: 204–9, 281–7). ¹⁷ On locatio conductio, see Thielmann and Bettermann (1961); cf. Thomas (1966) and Bove (1975: 327).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

407

although later the jurists tried to give an expansive interpretation of such list. Legal and epigraphic sources provide examples of slaves who are considered defective (wandering or runaway slaves) or diseased (dumb, shortsighted, or dim-sighted: D. 21.1.9; Ulp. 44 ad Sab.; D. 21.1.10.3 and D. 21.1.10.4; Ulp. 1 ad ed. aed. cur.). A well-known text by the jurist Ulpian (D. 21.1.1.1; Ulp. 1 ad ed. aed. cur.) states that “[t]hose who sell slaves are to apprise purchasers of any disease or defect in their wares and whether a given slave is runaway, a loiterer on errands, or still subject to noxal liability. . . . Again vendors must declare at the time of sale all that follows: any capital offense committed by the slave; any attempt which he has made upon his own life; and whether he has been sent into the arena to fight wild animals.”¹⁸ With the actio redhibitoria¹⁹ the buyer returned the good and asked for the restitution of the price paid, obtaining a result that tended to restore the status quo ante (in integrum restitutio; D. 21.1.23.7; D. 21.1.60). The responsibility of the seller was independent of his actual knowledge of the defect and hence was not based on dolus.²⁰ Yet, Ulpian adds that “[i]f a defect in or disease of the slave be perceptible (and defects reveal themselves generally through symptoms), it may be said that the edict has no place; its concern is simply to ensure that a purchaser is not deceived” (D. 21.1.1.6; Ulp. 1 ad ed. aed. cur.). The ordinary prescription period was six months from the day on which the contract of sale was concluded. Instead, with the actio quanti minoris,²¹ which was introduced with respect to the sale of cattle and then extended to the sale of slave, the buyer affirmed the contract but asked for the reduction of the price. A iudex would then calculate the new price with reference to the going market price for a similar (but defective) good. In this case as well, the responsibility of the seller was not based on dolus. The

¹⁸ For this and other quotations we have adapted the English translation of the Digest by Watson (2009). ¹⁹ See Guarino (1955: 295–6; 1956: 352–7); Volterra (1955: 3–7); Arangio-Ruiz (1956: 353–8); Watson (1965: 86ff.); Burdese (1975: 594–600); Memmer (1990: 1–45); Manna (1994: 137–46); Jakab (1997: 127–9); Garofalo (1998: 57–8); Donadio (2004: 173–7). ²⁰ See Hallebeek (2009: 10–179). ²¹ Giffard (1931: 682ff.); Arangio Ruiz (1990: 384, 391); Donadio (2007: 518–22).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

408

Barbara Abatino and Giuseppe Dari-Mattiacci

ordinary prescription period was one year from the day the contract was concluded.²² The praetorian remedy, the actio ex empto, was older than the aedilician remedies (third century BC) and gave the buyer the possibility to ask for damages, while affirming the contract. Understandably, since there was no market price to refer to, damages had to be calculated according to the buyer’s negative interest (damnum emergens) and were meant to make the buyer whole with respect to his position prior to the contract. Damages did not include the buyer’s lost profits (lucrum cessans) but only his reliance expenditures and the difference between the price paid and his valuation of the good, which approximated the price that the buyer would have paid rather than the objective market price. Differently from the aedilician remedies, the actio ex empto required proof of the seller’s dolus but could be used to claim any kind of defect and was not subject to prescription.²³ As a result, the praetor could consider the idiosyncratic interests of the buyer with respect to the good purchased and not only the quality indicators listed by the aediles. The actio ex empto affirmed the contract. During the Augustan period there are traces of an actio empti ad redhibendum,²⁴ giving the buyer the possibility of asking for restitution, an innovation first advocated by the jurist Labeo (D. 19.1.11.3; Ulp. 32 ad ed.). This was a later innovation, anticipating a gradual convergence of the remedies as the functions of both aediles and praetor were absorbed by the imperial administration of justice. Table 24.1 summarizes the differences between the three original remedies. Although there has been a long discussion as to the relationship among these remedies,²⁵ it seems clear that the aedilician remedies could be given only by the aediles within their jurisdiction and, likewise, the praetorian remedy could only be awarded by the praetor. Thus, buyers seeking a specific form of protection had to bring an action in the corresponding jurisdiction. A buyer in a private sale could only bring an action against the seller under the praetor’s jurisdiction using the actio ex empto, since the aediles had no jurisdiction outside the regulated markets. A buyer in a market sale could undertake a lawsuit with the aediles and choose between restitution ²² See Talamanca (1990: 591). ²³ Donadio (2007: 458 n. 7 and 510–18). ²⁴ Serrao (1959: 267–71); Vacca (1994: 60ff.). ²⁵ See Donadio (2004) for a detailed review of this debate.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

409

Table 24.1. Aedilician and praetorian remedies. Aedilician remedies

Effect

Praetorian remedies

Actio redhibitoria

Actio quanti minoris

Actio ex empto

Contract undone

Contract affirmed

Contract affirmed

Seller’s liability Restitution of the price Reduction of the price Reliance damages Seller’s dolus

Not relevant

Not relevant

Relevant

Type of defect Only listed defects

Only listed defects

Any defect

Prescription

One year

No prescription

Six months

(actio redhibitoria) and price reduction (actio quanti minoris). In addition, given the general jurisdiction of the praetor it seems plausible that a buyer in a market sale could also initiate a lawsuit with the praetor using the actio ex empto (Donadio 2004: 34). Therefore, a buyer who felt that the aedilician remedies did not fully protect his interests could choose to pursue legal action with the praetor. Yet, as we will argue, this latter option was probably not in the interest of buyers and hence, although there are no data to support this claim, was probably used infrequently.

24.3. ASYMMETRIC INFORMATION AND SYMMETRIC IGNORANCE IN SALE CONTRACTS In a typical sale contract, sellers have more information than buyers about the quality of goods for sale, a problem known as asymmetric information.²⁶ A feature of “quality” is that it has a common ranking for buyers. This is not to say that all buyers attach the same value to goods—valuations do indeed typically differ across individuals—but rather that all buyers value high-quality goods more than low-quality goods. As a result, sellers can exploit their informational advantage to the detriment of buyers, by selling low-quality goods as if they were ²⁶ On the historical development of the legal relevance of information disparities, see Decock and Hallebeek (2010: 89–133).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

410

Barbara Abatino and Giuseppe Dari-Mattiacci

high-quality goods. Uncertainty about quality makes buyers wary and unwilling to pay high prices. This in turn makes it unattractive for sellers to sells products of high quality and generates a downward spiral that drives high-quality products out of the market (adverse selection: Akerlof 1970; cf. Frier and Kehoe 2007: 119–22). This wellknown problem can be solved by imposing a legal duty to disclose information about quality on sellers, which in turn can be sustained by penalties for failure to do so (Kronman 1978). We will show that both the aedilician remedies and the praetorian remedies addressed this problem. Yet, reality is often more complex than that. Goods for sale might have other “idiosyncratic characteristics,” which, in contrast to quality, buyers rank differently. To clarify, a buyer might be willing to pay more for a slave baker rather than for a slave actor, while another buyer might have the reverse preferences. Buyers’ preferences are typically private information. Thus, as long as a seller does not know a specific buyer’s idiosyncratic preferences, he cannot exploit his informational advantage about the good’s idiosyncratic characteristics. This state of “symmetric ignorance” (Barzel et al. 2006) might actually facilitate trade if compared to asymmetric information: since both parties are uninformed, they could trade all goods for their expected value, without fear of being exploited. Ex ante, neither party has an incentive to strategically withdraw from trade and, hence, the adverse selection problem does not arise.²⁷ In contrast to asymmetric information, there are two ways to deal with symmetric ignorance. One solution requires inducing buyers to reveal their type to sellers and, subsequently, imposing a duty to disclose upon sellers. As we will demonstrate, this was the effect of the praetorian remedy. Another solution consists of preventing the parties from exchanging information, thereby avoiding that a situation of symmetric ignorance might degenerate into one of asymmetric information. Yet, in this case, a buyer might end up with a good having undesired characteristics. We will show that this was the effect of the aedilician remedies. It is worth nothing that both sets of

²⁷ It is worth noting that, if the buyers learned about the good’s characteristics they would be better informed than sellers and an inverse-adverse-selection problem would arise, since buyers could exploit their informational advantage to the detriment of sellers (Barzel et al. 2006).

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

411

remedies were designed to assure that trading parties had the same information (or lack thereof).

24.3.1. A Simple Model In this section, we sketch a very simple model of sales. The model is meant to capture the essential features of the transactions that generated the cases brought before the Roman magistrates and hence disregards many important details of these transactions. Yet, the stylized transactions that we describe in the following provide the basic ingredients of our argument. We consider a typical transaction between two parties, a seller and a buyer, for the sale of a good. Sellers are homogeneous, while buyers are of two types: one half of the buyers are of type A (for instance, they run a business in which they employ actors) while the other half are of type B (for instance, they run a business in which they employ bakers). Goods for sale can be distinguished along two dimensions: quality and idiosyncratic characteristics. Quality can be either high (H) or low (L), so that one half of the goods have high quality (for instance, a healthy slave) and one half of the goods have low quality (an unhealthy slave). The idiosyncratic characteristic of the good can be either A (for instance, the slave is an actor) or B (the slave is a baker). The crucial difference between quality and idiosyncratic characteristics is that all buyers value high quality more than low quality, while they value the idiosyncratic characteristic in a idiosyncratic way: buyer A prefers good A while buyer B prefers good B. To operationalize these ideas, let us specify the buyer’s valuations as in Table 24.2: high quality is valued at q by both buyers; further, each buyer attaches a value c to his preferred characteristic. The valuation of a good by a certain buyer is simply assumed to be the sum of these two values. Ideally, an efficient market should allocate goods with characteristic A to buyers of type A and goods with characteristic B to buyers of type B. Finally, the seller values high-quality goods at h and low-quality goods at l, irrespective of their characteristics. In an ideal world with complete information, all transactions would be efficient. However, in the real world, inefficiencies can arise because of asymmetric information. Typically, sellers know more about the goods for sale than buyers; hence, we assume that

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

412

Barbara Abatino and Giuseppe Dari-Mattiacci

Table 24.2. Buyers’ and seller’s valuations of the goods.

Idiosyncratic characteristic

L c

A

B 0

I.1.1.3 Buyer A

Quality

Quality

A

H q+c q

Idiosyncratic characteristic

B

H q

q+c

L 0

c

I.1.1.6 Buyer B

Quality

Idiosyncratic characteristic

A

B

H h

h

L l

l I.1.1.9 Seller

sellers know the quality level and the idiosyncratic characteristics of the goods, while the buyers only know their probabilities. This is a plausible assumption because learning about, say, a slave takes time and sellers might have acquired this information prior to the sale. Conversely, buyers know their own preferences. This bilateral asymmetric information can lead to inefficiencies. To illustrate, consider an example. If goods were sold without regard to asymmetric information, the price for a representative good would be somewhere between the buyer’s expected valuations, which is equal to (q + c)/2,²⁸ and the seller’s average valuation, which is equal to (h + l)/2. If, for instance, q = 120, c = 60, h = 100, and l = 20, all goods could in theory be sold irrespective of quality and idiosyncratic characteristics, because the expected buyer’s valuation, (120 + 60)/2 = 90, is higher than the average seller’s valuation for the same goods, (100 + 20)/2 = 60, and hence there is a price in between these values (say, 75) that would make both of them better off. However, the well-known problem of adverse selection (Akerlof 1970) might impair some or all of these transactions. To see why, consider that the seller knows the quality of the good and he values high-quality goods at 100. Hence, although on average the seller makes a gain by selling all goods at, say, 75, he makes an even bigger gain if the good is of low quality; this gain is partially offset by a loss ²⁸ The buyer does not know the level of quality and the idiosyncratic characteristic of the good but knows that half of the goods have high quality and that half of the goods have his preferred characteristic. Therefore, the expected value of a good is q/2 + c/2.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

413

associated with goods of high quality, which are sold below value. Thus, if the price is less than 100, the seller, having information on quality, will withdraw high-quality goods from the pool of goods for sale. Since the maximum price that a buyer is willing to pay for a good of unknown quality is 90, the only goods that the seller will put on the market are low-quality goods, those goods that the seller values at less than 90. If buyers anticipate the seller’s behavior, as they should rationally do, they will expect to be sold only low-quality goods and hence their willingness to pay will drop to c/2 = 30.²⁹ In turn, since the seller values low-quality goods at 20, the price will be between 20 and 30 (say, 25). The result is that the market shrinks, with only lowquality goods being sold due to asymmetric information and the ensuing adverse selection problem—which, in turn, drives highquality goods out of the market and pushes prices down. Note that if the seller could communicate the information he possesses about quality, high- and low-quality goods could be sold at different prices. Buyers are willing to pay 120 + 60/2 = 150 for a high-quality good with uncertain idiosyncratic characteristics, which the seller is willing to sell for at least 100. The problem with this is that information exchange needs to be supported by penalties for reticent or fraudulent behavior. Absent such penalties, the seller’s claims about the quality of the goods for sale are not credible and hence information cannot be effectively exchanged. Both the praetor and the aediles developed remedies to address this problem and, more specifically, to allow information exchange about quality. Note further than exchanging information about quality is not enough to ensure perfect matching, since A-goods could still end up in the hands of B-buyers and B-goods in the hands of A-buyers. If buyers knew the idiosyncratic characteristics and sellers knew the buyers’ types, high-quality goods could be traded for a price between 120 + 60 = 180 (the buyer’s valuation of his preferred good) and 100 (the seller’s valuation), while low-quality goods could be traded for a price between 60 and 20. With full information, A-buyers would buy A-goods and B-buyers would buy B-goods. By ensuring perfect matching ex post, full information would enhance the value of the transaction for both parties. The remedies introduced by the praetor and the aediles addressed this problem in diametrically different ²⁹ Recall that the buyer still buys without knowing the idiosyncratic characteristic in this basic setup.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

414

Barbara Abatino and Giuseppe Dari-Mattiacci

ways. While the praetorian remedies applied by the praetor stimulated exchange of information about idiosyncratic characteristics— striving to achieve symmetric information on both dimensions—the aedilician remedies impaired it—leaving the parties symmetrically ignorant about idiosyncratic characteristics but symmetrically informed about quality.

24.3.2. Transactions before the aediles The typical transaction within the jurisdiction of the aediles involved a professional seller selling goods through an auction. Since an auction does not involve negotiations, the seller did not learn the buyer’s type. Similarly, the buyer remained uninformed about the goods’ idiosyncratic characteristics and quality, since transactions were closed rapidly and the buyer normally did not have enough time to thoroughly inspect the good. Therefore, asymmetric information persisted both with respect to quality and with respect to idiosyncratic characteristics. The extent to which parties could credibly exchange information depended on the penalties that the aedilician remedies placed on false or reticent information. The liability regime of the aediles was as follows: 1. Scope: a seller was liable only for defects pertaining to quality, which were listed by the aediles. 2. Subjective knowledge: the seller’s subjective knowledge was irrelevant (no inquiry about dolus). 3. Remedy: the buyer could choose between restitution (actio redhibitoria) and reduction of the price (actio quanti minoris). 4. Prescription: short prescription terms. The aedilician remedies supported information exchange only with respect to quality, so that sellers had incentives to reveal information about quality and buyers could rely on it. This solved the adverse selection problem examined above. With respect to the idiosyncratic characteristics, both parties lacked some piece of information and hence neither party was in a position to exploit his superior knowledge. Since the buyer was ignorant of the good’s idiosyncratic characteristic and the seller was unaware of the buyer’s type, the parties were in fact in a position of symmetric ignorance. Therefore, goods were mismatched with some probability. In our simple model, half of

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

415

the goods are sold to the “wrong” buyer: an A-buyer could end up with a B-good or vice versa. The aedilician remedies induced transactions that were partially inefficient. In our simple model, we can measure this inefficiency precisely: it is c/2, because in half of the cases the good is sold to the buyer who values it the least and hence a value c is lost.³⁰ Yet, this allocative disadvantage was balanced by a reduction in the overall costs of the transactions, due to the other features of the aedilician remedies: • Market expediency: transactions were fast and did not involve lengthy negotiations precisely because information exchange was limited to important quality features and could be standardized through, for instance, the use of placards at auctions, the tituli. • Market effectiveness: if quality turned out to be low, buyers could retain the good for a lower price, if they wished to do so; this reduced the need for additional transactions ex post since some buyers might value the good even if the good was of low quality and hence they would have bought it (for the lower price) even if they had been informed. In such cases, restitution is inefficient and price reduction is a better alternative. • Low litigation costs: if problems arose, litigation was rapid and probably not very costly, since it only involved the verification of predetermined quality markers and did not require an inquiry into the seller’s subjective knowledge. • Convenient allocation of risk: the seller bore the risk of unnoticed quality failures because he was liable even if he did not know that quality was low. This was optimal, given that a professional seller was better placed to resell goods than a buyer and because this risk was reduced by the short statutes of limitation, which allowed for the quick reselling of the goods that were returned to the seller through the actio redhibitoria.

24.3.3. Transactions before the praetor The typical transaction before the jurisdiction of the praetor involved a seller who sold a good directly to a buyer. During the ³⁰ We implicitly assume that all goods offered by sellers would be traded under perfect information. This assumption is inessential for the qualitative results of the analysis.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

416

Barbara Abatino and Giuseppe Dari-Mattiacci

negotiations prior to the contract, the seller could learn the buyer’s type. For instance, the buyer searched for a good with specific characteristics (say, a slave baker) and thereby revealed his type. In this case, the seller became asymmetrically better informed along both dimensions and hence had an incentive to convey incorrect information both about quality and about idiosyncratic characteristics. As a result, the praetorian remedies were designed to induce information exchange by sellers along both dimensions: 1. Scope: a seller was liable for all defects, not only those pertaining to quality. 2. Subjective knowledge: the seller’s subjective knowledge was relevant (the buyer needed to prove dolus). 3. Remedy: the actio ex empto allowed the buyer to receive damages equal to the difference between the price paid and the value of the good to him (damnum emergens). 4. Prescription: no prescription term. The praetorian remedy induced information exchange along both dimensions and hence obtained a more efficient matching between goods and buyers, thereby removing the efficiency loss typical of the aedilician remedies. This was done at the price of inferior market expediency, because negotiations took probably longer and were more involved, and of higher litigation costs, because the inquiry also concerned the idiosyncratic characteristics of the good and the subjective knowledge of the seller. Since the remedy of the actio ex empto, at least until the introduction of an actio empti ad redhibendum, did not allow for the restitution of the good but only for the payment of damages, ex post market effectiveness was also lower. Some goods that the buyer would not have bought under perfect information remained nevertheless in his possession—at least until they were resold. Moreover, the praetorian remedies induced a different allocation of risk than the aedilician remedies. Since the seller was only liable for dolus, the buyer bore the risk for quality failures or of unsatisfactory characteristics that the seller innocently ignored. Assigning more risk to the buyer is consistent with the fact that sellers were not necessarily professionals and hence it was not necessarily optimal to shift liability to them.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

417

24.3.4. The Buyers’ Choice of Market vs Private Transactions In the model, buyers with a large appreciation of idiosyncratic characteristics compared to quality (large c relative to q) have an incentive to self-select into personal transactions. They care about obtaining the good of their preferred type and are willing to bear the higher quality risks (as liability is only for dolus) and the larger negotiation and litigation costs associated with operating under the praetorian law. In contrast, buyers characterized by a low appreciation of idiosyncratic characteristics relative to quality (small c relative to q) have an incentive to self-select into market transactions and hence accept the risk of not obtaining the good of their favorite type and, in exchange, face less risk about quality and smaller negotiation and litigation costs. The buyers’ behavior arguably enhanced the efficiency of the remedies. The aedilician remedies suffered from inefficient matching, but this problem would have been less severe if buyers who chose market transactions were, as we postulate, characterized by small appreciation for idiosyncratic characteristics. Hence, the cost-saving properties of the aedilician remedies could be advanced without much loss in terms of allocative efficiency. This argument also suggests that buyers who initially chose market transactions were unlikely to be willing to opt for the praetorian remedies and their higher resolution costs. In turn, buyers who chose personal transactions were characterized by a large appreciation for idiosyncratic characteristics, which made the costs associated to the praetorian remedies worth bearing.

24.4. ANALYSIS

24.4.1. The Aedilician Remedies The aedilician remedies possibly applied to contracting parties who interacted in the cattle and slave markets, in which sales were effected through auctions. Since sellers did not generally have the opportunity to interact with buyers prior to the sale, they were not in a position to learn about a buyer’s preferences and hence could not easily exploit their private information about the idiosyncratic characteristics of the

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

418

Barbara Abatino and Giuseppe Dari-Mattiacci

goods for sale. The aedilician remedies left buyers’ interests in idiosyncratic characteristics unprotected: the buyer could only claim the remedies if the seller had withheld information about the quality characteristics listed in the aediles’ edict, which did not allow the aediles to consider a buyer’s idiosyncratic interests. Therefore, sellers could, if they wanted, give misleading information about characteristics of the goods other than quality. This apparently inefficient solution in fact allowed for quick transactions. Sellers would expose the goods for sale (cattle or slaves) together with a placard (titulus) indicating possible defects, the slave’s nationality, and other standard quality markers.³¹ Since the only meaningful exchange of information pertained to quality, sellers and buyers would not engage in lengthy negotiations but rather would trade on the basis of standardized procedures. The aedilician remedies, however, gave sellers incentives to reveal information about quality truthfully and, in turn, induced buyers to rely on such information. If a buyer discovered that the good he purchased had one of the defects listed in the edict, which the seller had not disclosed, he could choose between the actio redhibitoria (restitution) and the actio quanti minoris (price reduction). This choice assured that buyers never paid more than what they were willing to pay, irrespective of the quality of the goods purchased. To clarify, the market generated, say, two prices: a high price for highquality goods and a low price for low-quality goods. The actio quanti minoris adjusted downwards the price of goods with undisclosed defects, so that the seller had to pay back to the buyer the difference between the high price and the low price. Since buyers’ valuations vary, a buyer who had purchased a good for a high price under the assumption that the good was of high quality might not want to retain the good once he discovered that the good was of low quality even if the price was reduced. In this case, the buyer could opt for the actio

³¹ About tituli Gellius (Noctes Atticae 4.2.1) reports that Titulus servorum singulorum scriptus sit curato ita, ut intellegi recte possit quid morbi vitiive cuique sit, quis fugitivus errove sit noxave solutus non sit (“See to it that the sale ticket of each slave be so written that it can be known exactly what disease or defect each one has, which one is a runway or a vagabond, or is still under condemnation for some offence,” trans. Rolfe 1927.) The text suggests that, by reading the titulus, purchasers obtained information about diseases, defects, noxal liability, propensity for stealing, or loitering, which might reduce the value of the slave.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

419

redhibitoria. If instead the buyer was willing to buy a low-quality good, he could choose the actio quanti minoris. These considerations also explain why the actio redhibitoria emerged earlier than the actio quanti minoris. Restitution is a safer remedy from the perspective of buyers as it brings the buyer back to the status quo ante. In contrast, price reduction could make some buyers worse-off with respect to the status quo, because it might impose a loss on buyers who have a very low valuation for lowquality goods, and hence are unwilling to buy a low-quality good for a low price. Assuming plausibly that buyers faced high transaction costs when reselling goods, this outcome would be inefficient ex post as would not optimally allocate goods to buyers. Yet, in combination with the actio redhibitoria, the actio quanti minoris enhanced the overall efficiency of the aedilician administration of justice. In fact, it allowed buyers who were willing to retain low-quality goods to do so, thereby saving on the transaction costs generated by returning the goods and offering them for sale at the next market. On the seller’s side, the buyer’s choice was relatively inconspicuous since most sellers were professionals who sold both high-quality and low-quality goods for the market price. Therefore, if the buyer decided to retain the good, we can assume that the seller was still willing to sell for the low price. Both remedies had short prescription periods. This was possible since only certain basic qualities needed to be verified by buyers. In addition, this allowed returned goods to be resold relatively quickly and hence reduced the resale costs borne by sellers. This was an important feature of the aedilician remedies. Since both remedies applied irrespective of dolus, sellers were in fact obliged to offer an implicit insurance against the eventuality that an unnoticed defect might materialize. It was probably efficient to allocate the risk of unnoticed defects to sellers due to the professional nature of most of them, and also because this risk was reduced by applying relatively short prescription periods. Summing up, the aedilician remedies allowed for quick transactions—because they reduced the amount of information that the parties exchanged—and for quick, relatively inexpensive trials— since they only allowed litigation on quality and did not examine dolus. Moreover, by offering buyers a choice between undoing and affirming the contract, they optimized the ex post allocation of goods to buyers with respect to quality. Short prescription periods reduced

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

420

Barbara Abatino and Giuseppe Dari-Mattiacci

the risk to sellers. Yet, the downside was a loss of efficiency ex post: since sellers could not credibly reveal information about idiosyncratic characteristics, buyers might end up with goods having undesired characteristics. The magnitude of this loss depended on how much buyers valued idiosyncratic characteristics. This problem was addressed by the praetorian remedy.

24.4.2. The Praetorian Remedies The praetor initially offered only one remedy, the actio ex empto, affirming the contract and awarding damages to the buyer. This remedy principally concerned private transactions—where buyers and sellers negotiated and exchanged information more intensely than in an auction—and only secondarily pertained to transactions at a public market. In private transactions, a buyer looking for a specific good (for instance, a slave actor) would implicitly reveal his type to the seller and hence put the seller in a position to exploit his informational advantage both about the good’s quality (for instance, whether the slave was prone to theft) and about its idiosyncratic characteristics (for instance, whether the slave was a good actor). Therefore, the buyer needed protection along both dimensions. Accordingly, the actio ex empto could be used to claim damages for undisclosed information about any characteristic of the good. Since both parties in a private transaction were likely to be private individuals rather than professional sellers, it is not obvious that the seller was in the best position to bear the risk of unnoticed negative characteristics. The actio ex empto awarded damages only in case of dolus, thereby allocating this risk to the buyer. Since unnoticed characteristics could also be of positive value, the buyer bore in fact offsetting risks of both negative and positive discoveries. This also explains why there was no prescription period: a buyer needed time to verify the characteristics of the good and to collect information about the seller’s prior knowledge. The natural decay of evidence over time implicitly constrained the date within which a successful suit could be brought. Moreover, since he was dealing with private transactions, the praetor could not easily refer to the market price of a good having different characteristics from those claimed by the seller. The very fact that the parties engaged in a private transaction suggests that they

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

421

may have intended to cover for idiosyncratic needs and hence that their exchange was not readily comparable to others occurring in the market. Hence, the compensation paid to the buyer was in the form of negative-interest damages, designed to bring the buyer back to his position before the contract and including damages beyond the price paid. This also shows that the actio ex empto did not have the same problems observed for the actio quanti minoris because it took into account each buyer’s private valuation for the good. This system made negotiations and trials more complex and possibly more expensive, but it allowed a more effective exchange of information and hence supported private transactions that allocated goods ex post to buyers more efficiently than the aedilician markets did, especially with respect to those buyers who valued idiosyncratic characteristics.

24.5. FORUM SHOPPING Contractual parties had an opportunity to choose between praetorian and aedilician remedies at two points in time. First, buyers and sellers could decide whether to trade in a market or through private transactions. If they opted for a private transaction, they would then fall under the exclusive jurisdiction of the praetor. If instead they opted for a market transaction, the buyer could possibly have a choice between filing a case with the aediles or with the praetor. We have shown that the aedilician remedies allowed for less expensive negotiations and litigation and involved less quality risk for buyers than the praetorian remedies. Yet, they generated larger allocative losses ex post due to the fact that idiosyncratic buyers’ interests were left unprotected. Therefore, buyers who valued a good’s idiosyncratic characteristics (for instance, a buyer looking for a slave with very specific skills) probably preferred to trade through private transactions. In contrast, buyers looking for a standard good and only interested in quality were most likely better off when purchasing in a market. The self-selection of different types of buyers into different types of transactions made the typical transaction that fell under the jurisdiction of the praetor essentially different from the typical transaction

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

422

Barbara Abatino and Giuseppe Dari-Mattiacci

falling under the aediles and reinforced the overall efficiency of the system. Since the buyers operating in markets were not too interested in idiosyncratic characteristics, the actual ex post efficiency losses were likely to be small. Moreover, these buyers were unlikely ex post to opt for the praetorian remedy, as it involved higher litigation costs and more quality risk and only a (small, given the types of buyers) advantage due to the better protection of idiosyncratic interests. Conversely, buyers interested in idiosyncratic characteristics were more likely to opt for private transactions and thereby the higher negotiation and litigation costs balanced large gains due to the more efficient allocation of goods ex post.

24.6. CONCLUSIONS In this chapter, we have shown how the aedilician and the praetorian remedies for defects in sale contracts addressed the contractors’ needs to efficiently address the problems created by asymmetric information. While the aedilician remedies catered to buyers purely interested in quality and induced quick, standardized transactions and inexpensive litigation, the remedies applied by the praetor catered to buyers interested in idiosyncratic features of the goods for sale and effectively protected those interested at the price of lengthier negotiations and more expensive litigation. These remedies slowly converged. An interesting question is how and why the praetor and the aediles fostered the economic interests of the parties over whom they had jurisdiction. The remedies they applied accumulated into an edict that, for the most part, was passed over from one magistrate to the next. In both cases, magistrates holding these functions were ascending the cursus honorum (the customary career progression) and aspired to higher functions. Since both the praetor and the aediles were in charge for only one year and were assisted by a council of jurists, they had little opportunity to entrench themselves in office and had instead clear incentives to do a good job in order to be subsequently elected to a higher office. The edict was announced at the beginning of the year and there was little room for subsequent adjustments. As a result, the lawmaking effort by both magistrates was directed to the general interests of contracting parties rather than to special interests

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

423

of individual litigants. This might have well been the engine that moved lawmaking in ancient Roman and produced such remarkable legal solutions to omnipresent problems.³²

REFERENCES Akerlof, George A. 1970. “The market for ‘lemons’: quality uncertainty and the market mechanism.” 84 Quarterly Journal of Economics 488–500. Andreau, Jean 1974. Les affaires de Monsieur Jucundus. Paris/Rome: Collection de l’Ecole française de Rome. Ankum, Hans 1972. “Quelques problèmes concernent les ventes aux enchères en droit romain classique,” in Studi in onore di Gaetano Scherillo. Milan: Istituto editoriale cisalpino—La Goliardica, 377–93. Arangio-Ruiz, Vincenzo 1956 [repr. 1990]. La compravendita in diritto romano, vols I–II. Naples: Jovene. Barzel, Yoram, Michael A. Habib, and D. Bruce Johnsen 2006. “Prevention is better than cure: the role of IPO syndicates in precluding information acquisition.” 79 Journal of Business 2911–23. Bove, Lucio 1975. “Rapporti tra dominus auctionis, coactor e emptor in Tab. Pomp. 27.” 21 Labeo 322–31. Brooks, Richard R. W., and Alexander Stremitzer. 2011. “Remedies on and off contract.” 120 Yale Law Journal 690–727. Brooks, Richard R. W., and Alexander Stremitzer. 2012. “On and off contract remedies inducing cooperative investments.” 14 American Law and Economics Review 488–516. Burdese, Alberto. 1975. Vendita (diritto romano), in 20 Novissimo Digesto Italiano. Turin: UTET, 594–600. Camodeca, Giuseppe. 1999. Tabulae Pompeianae Sulpiciorum. Edizione critica dell’archivio puteolano dei Sulpici. Rome: Quasar. Chevreau, Emanuelle. 2012. “L’édit des édiles curules: un droit des marchés avant la lettre?,” in Laurent Capdetrey et Claire Hasenohr, eds, Agoranomes et édiles: institutions des marchés antiques. Bordeaux: Ausonius, 223–33. de Ligt, Luuk. 1993. Fairs and Markets in the Roman Empire: Economic and Social Aspects of Periodic Trade in Pre-Industrial Society. Amsterdam: Gieben.

³² The authors would like to thank two anonymous referees, the participants in the Forum Romanum seminar at the University of Amsterdam, and the 2013 conferences of the International Society for New Institutional Economics in Florence and the Italian Society for Law and Economics in Lugano for helpful comments.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

424

Barbara Abatino and Giuseppe Dari-Mattiacci

de Ligt, Luuk, and Pieter W. de Neeve. 1988. “Ancient periodic markets: festivals and fairs.” 66 Athenaeum 391–416. Decock, Wim, and Jan Hallebeek. 2010. “Precontractual duties to inform in early modern Scholasticism.” 78 Tijdschrift voor Rechtsgeschiedenis 89–133. Donadio, Nunzia 2004. La tutela del compratore tra actiones aediliciae e actio empti. Milan: Giuffrè. Donadio, Nunzia. 2007. “Azioni edilizie e interdipendenza delle obbligazioni nell’emptio venditio. Il problema di un giusto equilibrio tra le prestazioni delle parti,” in Luigi Garofalo, ed., La compravendita e l’interdipendenza delle obbligazioni nel diritto romano, vol. 2. Padua: Cedam, 457–537. Donadio, Nunzia. 2011. “Promissio auctionatoris.” 39 Index 524–57. Frier, Bruce W., and Dennis P. Kehoe. 2007. “Law and economic institutions,” in Walter Scheidel, Ian Morris, and Richard Saller, eds, The Cambridge Economic History of the Greco-Roman World. Cambridge: Cambridge University Press, 113–43. Hallebeek, Jan. 2009. “The ignorant seller’s liability for latent defects: one regulator various sets of rules?,” in John W. Cairns and Paul J. du Plessis, eds, The Creation of the Ius Commune from Casus to Regula. Edinburgh: Edinburgh University Press, 175–218. Garofalo, L. 1998. “Redhibitoria actio duplicem habet condemnatio (a proposito di Gai. 1 ad ed. aed. cur. D. 21.1.45),” in Atti del II Convegno sulla Problematica contrattuale in Diritto Romano (Milano 11–12 maggio 1995). In onore di Aldo dell’Oro. Milan: LED, 57–86. Giffard, André E.V. 1931. “L’action édilicienne quanti minoris (D. 21.1.38 pr.; 13 et 14).” 12 Revue historique de droit français et étranger 682–87. Gröschler, Peter. 1997. Die tabellae-Urkunden aus den pompejanischen und herkulanensischen Urkundenfunden. Berlin: Duncker & Humblot. Guarino, Antonio. 1955. “L’editto edilizio e il diritto onorario.” 1 Labeo 295–307. Guarino, Antonio. 1956. “Ancora sull’editto edilizio.” 2 Labeo 352–7. Guarino, Antonio. 1995. “Tagliacarte.” 41 Labeo 304–5. Impallomeni, Giovanni B. 1955. L’editto degli edili curuli. Padua: Cedam. Jakab, Éva. 1997. Praedicere und cavere beim Marktkauf. Munich: Beck. Krishna, Vijay. 2002. Auction Theory. San Diego: Academic Press. Kronman, Anthony T. 1978. “Mistake, disclosure, information, and the law of contracts.” 7 Journal of Legal Studies 1–34. Kupisch, Bertold. 2002. “Römische Sachmängelhaftung: ein Beispiel für die ökonomische Analyse des Rechts.” 70 Tijdschrift voor Rechtsgeschiedenis 21–54. Lerouxel, François. 2016. Le marché du crédit dans le monde romain (Égypte et Campanie). Bibliothèque des Écoles françaises d’Athènes et de Rome 374. Rome: École française de Rome. Lo Cascio, Elio, ed. 2000. Mercati permanenti e mercati periodici nel mondo romano. Bari: Edipuglia.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

The Dual Origin of the Duty to Disclose in Roman Law

425

Malmendier, Ulrike M. 2002. Societas publicanorum. Cologne: Böhlau. Manna, Lorena. 1994. Actio redhibitoria e responsabilità per vizi della cosa nell’editto de mancipiis vendundis. Milan: Giuffrè. McAfee, R. Preston, and John McMillan. 1987. “Auctions and bidding.” 25 Journal of Economic Literature 699–738. Memmer, Michael. 1990. “Der schöne Kauf des guten Sklaven. Zum Sachmängelrecht im Syrisch-römischen Rechtsbuch.” 107 Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Rom. Abt. 1–45. Monier, Raimond. 1930. La garantie contre les vices cachés dans la vente romaine. Paris: Recueil Sirey. Rolfe, John C., ed. 1927. Gellius. Attic Nights. Loeb Classical Library 195, 3 vols. Cambridge, MA: Harvard University Press. Serrao, Feliciano. 1959. “Vacanza dell’edilità e competenza dei pretori nel I sec. a.C.” 62 Bullettino dell’Istituto di diritto romano Vittorio Scialoja 267–71. Serrao, Feliciano. 2000. “Impresa, mercato, diritto. Riflessioni minime,” in Elio Lo Cascio, ed., Mercati permanenti e mercati periodici nel mondo romano. Bari: Edipuglia, 31–67. Talamanca, Mario. 1990. Istituzioni di diritto romano. Milan: Giuffrè. Thielmann, Georg, and Bettermann Karl A. 1961. Die römische Privatauktion: zugleich ein Beitrag zum römischen Bankierrecht, ed. Ulrich von Lübtow. Berlin: Duncker & Humblot. Thomas, J. A. C. 1966. “Le ‘auctiones’ private.” 12 Labeo 395–9. Vacca, Letizia. 1994. “Ancora sull’estensione dell’ambito di applicazione dell’actio empti in età classica.” 45 Iura 35–73. Volterra, Edoardo. 1955. “Intorno all’editto,” in Scritti in onore di U. Borsi. Padua: Milani. Volterra, Edoardo. 1956. “Ancora sull’editto degli edili curuli.” 7 Iura 141–7. Watson, Alan. 1965. The Law of Obligations in the Later Roman Republic. Oxford: Clarendon Press. Watson, Alan. 1987. “Sellers’ liability for defects: aedilician edict and praetorian law.” 37 Iura 167–75. Watson, Alan, ed. 2009. The Digest of Justinian. Philadelphia: University of Pennsylvania Press. Wolf, Joseph Georg. 2010. Neue Rechtsurkunden aus Pompeji. Tabulae Pompeianae novae. Darmstadt: Wissenschaftliche Buchgesellschaft.

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Index accounting 28, 35–9, 49, 52, 54–6, 85, 175, 280 (see also rationes) actio (legal remedy) adiecticiae qualitatis 97, 278 empti utilis 144 ex contractu 303 ex delicto 303 ex empto 401, 403, 404, 408, 416, 420, 421 institoria 95, 278 legis 333, 339, 340 quanti minoris 401–3, 406, 407, 409, 414, 418, 419, 421 redhibitoria 401–3, 405–7, 409, 414, 415, 418, 419 Serviana 122, 137 tributoria 97 actus (land measurement) 218, 219 adrogatio 287 adverse selection 3, 120, 410, 412, 413–4 aedile 192, 402–6, 408, 410, 414, 415, 417, 419, 421, 422 edict of 418, 422 curule 162, 401, 402 plebeian 162 aequum 31 Africanus 38 ager acrifinius 240 agriculture 12, 17–18, 41, 116, 140, 167, 174, 178, 181, 212, 236, 243, 259, 267 agrimensor 223, 225, 237, 241 Agrippa, Marcus 198 American rectangular system 212, 221, 223–7, 229–33, 237–9, 243 amicitia 113, 119, 134, 283 antichresis 127, 150 Antonine period 91 arbitration 224–5, 328, 333, 337, 339 argentaria 90, 91, 97 argentarii 82, 85, 88–91, 97, 101, 105, 107, 129, 406 receptum 105, 106 Aristotle 18, 24

assets 36–7, 57, 69, 72, 102, 117–19, 122, 124–5, 127, 129, 132, 135, 137–8, 144, 148–50, 169, 202, 252–4, 256, 268, 277–9 charged asset 117–18, 121–31, 145, 149–51 pledged assets 136, 137, 144–5 Athens (classical) 16, 18 auction 88–9, 97, 100–1, 106–7, 128–30, 402, 406, 414, 417, 420 Augustan age 13, 14, 17, 19, 31, 41, 45, 69, 82, 93, 133, 179 Augustus 20, 21, 47, 87, 192, 198, 200, 223 bank/banking 57, 90–5, 97, 102–3, 105, 107, 124, 175, 263 bankruptcy 104, 118 bibliotheke enkteseon (property registry) 132, 264, 290 bona fides 30, 48, 332 bonding 3, 5, 119, 121, 285, 337 borrowing 55, 70, 83–5, 98–9, 114, 117–20, 125, 128, 145, 260, 275–6, 285 bundle of rights 71, 344 Byzantine period 21 Caesar, Julius 196, 199, 223 capital 22, 70, 85, 102, 150, 165, 172, 175, 182, 282 capital punishment 331, 334, 340, 343, 381, 407 capitalism 11, 14, 23, 32 Caracalla 91, 93, 142 cardo maximus 218, 220, 221, 223 Cato 58, 82, 89 cautio praedibus praediisque 133 cenacula 162 censor 196 census 133, 188, 262, 264, 265, 291 centuriation 212, 219, 226, 233–8, 240 Cicero 18, 64, 82, 84–6, 89, 93, 101, 165, 171, 179, 202 claim priority 122–3, 250, 264, 277

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

428

Index

Claudius 45, 198 clientela 284, 285 coactor 89, 406 argentarius 89, 97, 101 cognitio 341 collateral 120, 127–30, 138, 149–50, 251, 276, 285, 322 collective responsibility 227, 352, 379–80, 394 comitia tributa (tribal assembly) 405 commodatum 87 Commodus 55 common law 290, 304–5, 308–9, 315, 327, 343 common pool problem 363, 370 compensation 46, 90, 103, 104, 138, 258, 327, 335, 365–6, 371, 383–4, 386, 388–9, 393, 404 concilia plebis (plebeian assembly) 405 consensus 30 contracts, generally 5, 26, 49, 70, 87, 92, 96, 98–9, 104, 118, 120, 123, 125, 128, 132, 140, 144, 151, 183, 197, 227, 247–52, 256, 263, 266–8, 276–8, 280–1, 285–6, 288, 291, 306, 310, 318–9, 321–2, 335, 338, 392, 401–4, 406, 408, 419–21 breach 118, 190, 310–11, 383 contract law 302–3, 308, 311, 320, 324, 339, 347 contract rights 2, 49, 268, 313 contracts of sale 6, 322, 401, 403, 407, 409, 422 freedom of contract 11, 183, 191–2 performance 30, 56, 66–7, 191, 198, 306, 310–11, 322, 323 contubernalis 54 conventus 94 coordination 2, 56, 193–4, 214, 227–30, 232, 233, 235, 238, 320, 354, 372, 392 corporation 252, 281–2, 379, 387, 393 business enterprise 1, 37, 58–9, 86, 88–9, 98, 104–5, 117, 126, 137–8, 165, 277, 279–82, 286 costs, generally 4, 46, 54, 60, 62, 123–5, 128, 131, 134, 169, 171–2, 185, 189, 194, 199, 225, 229–30, 250–2, 254–6, 261, 265, 271, 280, 289–91, 365, 372–3, 386, 415, 419 demarcation cost 225–6, 228–9 (see also transaction costs)

fixed cost 252, 254–5, 282, 290 implementation cost 214, 231, 254–5, 261 monitoring cost 46, 57, 59, 62, 117, 121, 226, 288 variable cost 254–5, 290 Crassus 165 credit 3, 48, 88–90, 92, 97–8, 100–1, 105, 107, 114–16, 119, 121, 124–5, 129, 175, 261, 283, 356 access to credit 117, 120–1, 125, 127, 128, 143 secured credit 115–16, 121, 123, 128–33, 136, 142–5, 148–51 creditor 49, 83, 85, 92, 98, 105, 116–24, 126–31, 135, 137–8, 141, 144–5, 149–51, 199, 248, 252, 279, 284, 287, 331, 336–7, 340 crimen 144, 382 criminal law, generally 5, 308, 327, 330, 334, 342–3, 347–8, 356–8, 369–70, 372–3, 380, 382, 387, 390 criminal liability 145, 354, 364, 373 criminal sanction 5, 276, 329, 341, 357, 359, 361, 367–8, 370, 372, 380, 382, 390, 393 murder 311, 328, 334 cura annonae 402, 405 cura ludorum 405 cura urbis 405 cursus honorum 422 custom 13, 94, 142, 145, 187–8, 190–1, 193, 198, 237, 239, 258, 322, 328, 330, 333–4, 369 damages 6, 56, 60, 190, 193, 285, 313, 341–3, 348, 382–3, 385, 388, 393, 401, 408, 416, 420 punitive 341, 382 reliance 401, 408, 418 restitution 5, 6, 401, 403, 407, 415–16, 418–19 debt 28, 45, 49, 63–4, 82–5, 92, 96, 98, 104, 106, 120, 122, 124–8, 130, 134, 138, 150, 161, 182, 190, 199–200, 252, 258, 260, 284–5, 306, 320–1, 336, 340, 356 (see also reliqua) debtor 49, 57, 116–19, 121–31, 134–5, 137–8, 142, 144–5, 149–50, 199, 256, 284, 331, 337, 356 default 128, 142, 260, 284, 331, 336 decimation 283

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Index decumanus maximus 218, 220, 221, 223, 237, 239 deed 53, 142, 146–7, 251, 264, 266 delict 5, 308, 382 Delos 23 demurrer 304, 308, 321 deposit accounts 90–1, 95–6, 104, 105–6 (see also receptum argentarii) depositum 98 deterrence 5, 143, 347–50, 354, 356–61, 363, 367–8, 370–3, 387, 389–91 Diocletian 50, 63, 123 disclosure, generally 46, 143–4, 406, 410 failure to disclose 134, 144, 151, 318, 405–6, 410, 420 dolus 309, 402, 404, 407, 408, 414, 416, 417, 419 dominium 190 dominum ex iure Quiritium 138 domus 162, 195 due process 303–4 duty of care 317, 322, 381, 383 economic growth 11, 23, 119–20, 150, 171, 175, 198–9, 243, 247–8, 271, 289 editio rationum 104, 105 efficiency 2, 4, 5, 121, 123, 131–2, 197, 200, 237, 258, 260–2, 271, 274, 290, 302, 348, 355, 360–1, 367–71, 373–4, 379, 391, 393–4, 402, 411–12, 415, 417, 419–22 Egypt Ptolemaic 134, 263 Roman 41, 42, 53, 61, 71, 91, 93, 94, 114, 132–4, 143, 148, 259, 261, 263, 264, 267, 268, 291 elite 15, 23, 30, 41, 43, 82, 84, 86, 90, 100, 102, 113, 116, 170–1, 174, 183 emperor 87, 192, 198 enforceability 6, 50, 56, 71, 127–9, 131, 141, 145, 150, 227, 248–50, 255, 276, 284, 286, 288, 290–1, 306, 309, 321, 323, 327, 329, 335, 340 enforcement by law 344, 379, 383–7, 391, 394 private enforcement 5, 308, 334, 338–9, 341–4, 382 public enforcement 227, 308, 330, 334, 337, 339–41, 343, 370, 382 strict enforcement 4, 248, 277, 289, 291

429

equestrians 82, 83, 84, 85, 86, 105 estates landed 23, 26, 42, 82–3, 85, 99–100, 122, 148, 188, 259 inherited 36, 66, 69, 138, 175, 184–5, 336, 339 euergetism 198 exchange 26, 81, 404, 413–14, 416, 418–21 externalities 192–3, 195, 314, 349 faeneratores 86, 89, 100, 101 familia 285–7, 291 family law 286, 291 fault 308–9, 316, 382–3 feud system 328, 338–9, 344 fideiussor 335, 336 (see also surety) fiducia 114, 123, 126, 138, 140, 141, 275 finance 3, 46, 81, 84–105, 116, 119–22, 131, 149–50, 175, 196–9, 260 forfeiture 129, 340, 383 forma censualis 265, 266 formal rules 115, 189, 280, 283–4, 291, 338, 354 Forum, Roman 167 fraud 69, 92, 134, 137, 141, 144, 151, 262, 276, 286, 307, 318, 383, 402, 404, 413 free–rider problem 50, 352, 363, 370, 391 freedman (libertus) 20, 40, 47, 54, 64, 65, 69, 83, 165, 381 orcinus 36, 66 operae 66–69 furtum 136, 366, 382 (see also theft) Gaius (jurist) 25, 95, 123, 178, 188, 302, 304–6, 311 game theory 349–50, 352, 373 behavior 413, 417 repeat game 166, 350, 352, 373 gesta municipalia 146, 147 gift 53, 86–7, 166, 170, 321 good faith 30, 37–8, 48, 57, 130, 134–8, 150, 186–7, 250, 273–4, 319, 322, 332, 404 goods 12, 27, 87–8, 99–100, 119, 139, 161, 169, 173–6, 196–8, 247, 249–51, 256–9, 261, 269, 274, 323, 401, 410, 412, 414–21 nonconforming goods 401, 403, 405, 407, 416, 418–19, 422

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

430

Index

Gracchi 31 groma 224, 225 guarantee 3, 60, 113, 119, 261, 276, 336, 339, 356 Hammurabi Code 361, 363, 364 Hellenistic period 71, 91, 268 heredium 219 heres necessarius 45 hypothecae 141 idiosyncrasy 4, 352, 403–4, 408, 410–14, 416–18, 420–2 impetratio domii 129 in iure cessio 255, 269, 271 incentives 37, 54, 59, 72, 190–1, 195, 230, 247, 249, 280, 286, 288, 339, 343, 347–8, 362, 364, 367, 373, 387–8, 393, 403–4, 410, 414, 416–18, 422 indemnification 250, 315, 382 Industrial Revolution 11, 186, 281, 282 infamia 134, 143, 284, 338 information 64, 120, 173, 235, 265–6, 273, 284, 403–4, 411–14, 418, 420–1 information asymmetry 247–51, 256, 265, 274, 280, 284–5, 316, 403, 409, 410–14, 416, 422 information costs 119, 131, 172, 185, 350, 362, 370 private information 6, 410, 417 iniuria 309–11, 324, 382 innovation 13, 169, 175, 278, 370, 408 insula 162, 163 insurance 182, 200, 362–3, 387, 390, 404, 419 interest 81–2, 84–6, 88–9, 92, 96–9, 101, 116–17, 119, 135, 408, 421 interest rate 82, 84–7, 89, 119–20, 125, 128, 133, 145, 199, 260, 263 investment 44, 58, 62, 98–9, 102, 118–9, 165, 175, 190, 232, 235, 248, 250, 281, 350, 352, 373 iurisdictio 402, 405 ius civile 31, 92, 115, 126, 403 ius commune 115 ius edicendi 405 ius gentium 92, 93, 260 ius honorarium 94, 330 ius mercatorum 103 ius Papirianum 332

ius respondendi 95 iustum pretium 130, 200 judicial forbearance 280, 285–6, 291, 306, 314 Julio–Claudian period 99 jurist 13–16, 18, 20–1, 23–5, 28, 31–2, 37–9, 45, 52, 57, 60, 94–9, 104, 142, 177, 179, 190–1, 196, 200, 214, 224–5, 249–51, 265, 267, 284, 303, 324, 330, 333, 380, 383, 407, 422 Justinian 21–3, 106, 301, 329, 330, 336, 341, 403 Juvenal 167, 173, 179 kind–for–kind constraint 353, 360, 367–8, 373 Kingdom, Roman 217, 338 labor 17–8, 20, 25, 31, 36, 42, 43, 46, 58, 60–3, 66–9, 72, 172, 258, 280 (see also operae) land, generally 25–6, 57, 85, 100, 132, 138, 140–2, 145–6, 148, 160, 165, 167, 170, 180–2, 184–8, 192, 198–9, 212, 239–40, 254, 256, 261, 263–4, 266, 269–71, 273–4, 276–7, 290, 310, 313–15 land demarcation (see centuriation) land records 53, 202, 237, 266 land tenure 161, 190–1, 276–7 land value 166–7, 214, 226–32, 240, 250, 271–3 landlord 125, 164, 189–192, 195 landowner 100, 116, 160, 181, 187, 193, 198–9, 235, 248, 266 lease 25–6, 125, 160, 170, 180–1, 189–91, 194, 195, 202 legal claims, generally 50, 57, 66, 69, 71, 97, 149, 195, 250, 265, 304, 328, 337, 343, 401 cause of action 52, 65–6, 106, 118, 144, 193, 223, 270, 273, 274, 279, 287, 303, 305, 307–38, 311, 337, 408–9, 421 claim elements 303, 308, 322 legal defenses, generally 305, 309–10, 316–7, 321 excuse 308–11, 315–16, 318, 323 justification 308–11, 315–16, 318–19, 323

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Index legal obligation 21, 28–9, 36, 49–50, 54–6, 66, 129, 265, 277–8, 283, 321, 336–7, 341 legal sanction 283–4, 291, 339, 367, 380, 386–7, 389–92, 395, 410, 413 legal status 44, 56, 68, 105, 131, 144–5, 281, 287–8, 334, 338–9 legal systems 2, 16, 63, 93, 114, 128–9, 185–6, 190, 211, 250–1, 257, 283–4, 301–2, 309, 324, 327, 330, 337, 339–40, 342–4, 347, 351, 357, 364, 368 leges publicae 95 legislation 13, 63, 92, 160, 200, 251, 262, 330, 333–4 lending, generally 3, 29, 57, 82–9, 91–2, 96, 98, 100, 120, 122, 125, 131, 182, 199, 251, 263, 275, 285, 406 loans 81–92, 96, 98–102, 114, 118–19, 125, 133, 182, 199–200, 260–1, 263, 266, 275, 285 loan agreements 83–4, 118, 199–200 secured loans 3, 83, 117, 119, 132–3 Lex Aelia Sentia 42, 45, 64 Lex Aquilia 308, 309, 314, 315, 318, 367 Lex Genucia 82 lex talionis 342, 348, 354–72 liability, generally 5–6, 56, 58, 105, 143, 145, 151, 190, 225, 277–9, 281, 284, 307, 309–11, 313–19, 321–2, 332, 348, 354, 362–4, 369–70, 373, 379, 380, 382–90, 392–8, 404, 407, 414, 416–17 civil liability 5, 143, 225, 354, 372, 382 collective liability 6, 348, 352–3, 379–99 individual liability 5, 6, 348, 362, 364, 370, 373, 380, 384, 386, 389–90, 394–5, 397–8 joint and several liability 105, 387 liability rules 190, 307, 369 limited liability 58, 278–9, 383, 404 noxal liability 6, 383, 392, 407 strict liability 307, 309–10, 314, 317, 319, 321, 383 vicarious liability 383–4, 387–8, 392–3 limites 218, 220, 239 litigation 29, 31, 56, 83, 93, 104, 145, 185, 237, 273, 281, 286, 303–7, 308, 328, 333, 338–9, 356, 404, 408–9, 421 Livy 81, 92

431

mancipatio 123, 124, 136, 138–40, 145, 146, 186–8, 255, 269–74, 287 mandatarius 277 mandatum 87 manufacturing 17, 84, 101, 106, 169–170, 173–4 manumission 20, 21, 35–72, 194, 291 by fideicommissum 55, 60, 66, 68 emptio suis nummis 55, 66, 69, 71, inter vivos 50–2, 67, 68, 71 iteratio 48 iusta manumissio 53 pactiones pro libertate 49, 50, 54, 56 tax 40 Marcus Aurelius 55, 332 markets 1, 12, 23, 25, 100, 106, 131, 176, 181, 200, 251–2, 255, 260, 277, 401–3, 406, 408, 413, 419–22 housing market 189–90, 200, 214, 226–7, 229–30, 232, 243, 258, 261, 271 market price 61, 63, 200, 404, 407–8, 419, 420 market value 36, 60, 63, 117, 129 Martial 63 measure–for–measure punishment 360, 373 metator 223 metes and bounds 212, 227–31, 238, 240 monitoring 3, 70, 119, 121, 228, 284, 362, 387 moral hazard 3, 118–19, 363, 387 mortgage 138, 200, 251–2, 256, 258, 263, 291 mutuum 87, 96 negatiatores 100 negligence 305, 307, 309–10, 315, 317–19, 322, 367–8, 388 negotiations 35, 48, 51–2, 55–6, 60, 68, 71–2, 170, 187, 193, 199–200, 414, 416–18, 420–2 nemo plus principle 135 neo–institutionalism 106 Nero 20 network effects 230, 232 norms 13, 42, 49, 57, 84, 89, 91–4, 105, 134, 143, 182–4, 189–90, 279, 283–5, 291, 307, 312, 322–3, 348, 351–4, 356, 361–3, 369, 372–3 notice 278, 287, 303, 312

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

432

Index

nuisance law 193–4 nummularii 97, 106 nundinae 100 obligation (contractual/delict) 6, 30, 84, 85, 87, 96, 129, 143, 176, 249, 264, 265, 277, 278, 313, 319–24, 336–7, 340–2, 366–7, 382 involving slaves and freedmen 2, 21, 28, 36, 38, 49, 50, 54–5, 56, 65–8 personal obligations, 255–6, 268, 275, 278, 283–5, 288–91 officium 261, 283 omission 307–8, 318 operae 52, 65–96 (see also labor) optimality 24, 126, 272, 281, 356, 359, 368, 395, 402 ownership 127, 129–30, 133, 135, 137, 139, 146, 148, 165, 166, 181, 186, 193–5, 202, 240, 269–70, 273–4, 276–7, 312, 332, 361 partnership 38, 102, 105, 281 (see also societas) paterfamilias 184, 195, 278, 280, 281, 283, 285, 286, 288 patronage 35, 65, 69, 70, 71, 166, 184, 284, 286 Pax Romana 174, 179 peculium 27, 35, 37–9, 49, 51, 53, 55, 57, 58, 65, 69, 72, 102, 194, 279, 280 pecuniae creditae ius 92 peregrinus 91, 93, 94, 260 perticae 224 pignoris capio 337, 340 pignus 83, 114, 115, 123, 124, 126, 127, 140, 141 (see also pledge) Plautus 53, 82, 89, 90 pleading rules 301–2, 307, 311, 317 pledge 118, 122–3, 126–7, 131, 136–8, 141, 143–7, 150–1, 336 (see also pignus) possessory pledge 118, 122, 123, 127, 141–2, 150 pledge rights 122, 125, 144, 147 Plutarch 98, 165 Pompey 23 possession 130, 135, 136–8, 142, 146, 186, 269–70, 273–4, 276–7, 279, 312–13, 332, 339–40 praepositio 102

praetor 174, 192, 260, 290, 303, 401–6, 408–10, 415–17, 420–2 praetor’s edict 13, 29, 30, 55, 94, 95, 97, 278 preferences 132, 349, 352, 372–3, 410, 412, 417, 420 premium 60–61, 387 prescription period 186, 269, 271, 273–4, 402, 404, 407–8, 414–16, 419–20 principal–agent theory, generally 194, 277, 288 agency law 194, 277 agent 36, 166, 193–5, 227, 247–50, 277–80, 285, 288, 327 principal 102, 117, 248–9, 277–9, 281, 392 Principate 11, 16, 18, 83, 86–88, 93, 95, 98, 99, 106, 114, 150, 162, 174, 177, 179, 181, 195, 217 prior tempore principle 135 private law 11, 96, 133, 183, 302, 309, 313, 339 procedural law 5, 22, 56, 95, 102, 187, 269, 271, 273, 301–3, 330, 333, 338–9, 355 production 12, 17, 18, 25, 32, 41, 98, 120, 131, 228, 259–60 productivity 23, 46, 72, 212, 226, 228–9, 231, 253, 259 promise 306–7, 310, 320–1, 347 property, generally 5, 14, 15, 48, 123–4, 126–7, 130, 132, 134–6, 139–40, 143–5, 148–9, 160, 165, 184, 187–8, 192, 232, 235, 242, 247, 251, 255, 257, 265–6, 268–9, 279, 284, 290–1, 309, 332, 337–8, 340–1, 343, 363, 366–7, 382 alienation of property 69, 180, 182–3, 186 liability rules on 190, 307, 369 private property 160, 181, 183, 196, 202 property law 135, 237, 267–8, 302, 320, 324 property ownership 15, 268, 276 property rules 190, 249–51, 277, 279 property registration system 131–4, 143, 146–50, 185, 202, 252, 255–6, 261–4, 266–8, 275, 281–2, 291 property transfer (see transfer) public property 133, 196

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Index prosecution 327, 329, 334, 339, 341–2, 344, 367 protopraxia 132 provinces 15, 85, 259, 268 provincial governor 94 public goods 196–9, 201, 229–30, 235, 258, 290 publica causa 105 publicani 197, 281 Punic Wars, 241, 260 quaestor 23 querela inofficiosi testamenti 184 rapina 382 rationality 24, 32 rationes 37–8, 90, 96, 104, 105 (see also accounting) real estate 84, 122, 132–3, 142, 147, 150, 160, 165, 169, 173, 175, 181, 183, 185–6, 188–9, 193, 201–2, 258, 265 reciprocity 11, 13, 25, 30, 171, 350, 353–4, 362, 370 registration systems, generally 131–4, 143, 146–50, 185, 202, 252, 255–6, 261–4, 266–8, 275, 281–2, 291 of land 53, 202, 237, 266 public registration 130, 132–3, 251–2, 291 regulation 15, 20, 22, 82, 103–5, 195, 260 reliqua 37, 38 (see also debt) remedies 6, 144–5, 184, 250, 278, 280, 314, 321, 337, 355, 357–8, 368, 371, 382, 401–6, 408, 410–1, 413–14, 416–22 rent 62, 161, 164–5, 167, 170, 188–92, 199–200, 231, 258 replication 301, 304–6, 323 Republic, Roman early 161, 186, 202, 217, 330, 338 late 11, 14, 21, 23, 82, 83, 86, 88–9, 92, 95, 97, 98, 106, 150, 160, 162–4, 174, 177, 179, 181, 195, 199, 217, 255, 278, 312 res locata 26 res mancipi 123, 124, 136, 138, 140, 285 res nec mancipi 123, 141 res publica (see public property) respondeat superior 387–8

433

resources 160, 165, 171, 224, 240, 243, 253–5, 257–9, 279, 301, 362, 388, 393 restitution 5, 6, 401, 403, 407, 415–16, 418–19 retaliation 329, 335, 349, 351, 353–5, 359–60, 362, 364–6, 369–71, 373, 385 retributivism 351, 354, 360, 363, 369, 386, 389 rights, generally 2, 4, 66, 130–1, 135–6, 139, 142, 192, 212, 214, 243, 247–8, 250, 256, 264, 269, 271, 273, 277, 285, 303, 308, 313, 316, 334, 371 contract rights 2, 49, 268, 313 in personam rights 250, 269, 283, 303, 308, 313 in rem rights 130–1, 135–6, 248, 250, 264, 268–9, 277, 283, 289, 291, 303, 308, 313–4 pledge rights 122, 125, 144, 147 property rights 2, 4, 139, 142, 192, 212, 214, 243, 247–8, 256, 264, 271, 273, 277, 285, 316, 371 security rights 122–4, 131–132, 148, 150 vengeance right 330, 335, 351 risk 2, 4, 85, 104–5, 113, 117–21, 138, 145, 164, 168, 185, 188, 191, 195, 240, 260, 275, 318, 323, 347, 355, 362, 364, 380, 387, 404, 415–7, 420–2 allocation of risk 118, 318, 415–16, 419 assumption of risk 305, 317–20, 322 mitigation of risk 2, 113, 117, 120, 240, 323 security, generally 38, 72, 113, 116–17, 120–30, 138–9, 141, 149, 182, 191–2, 257, 266, 276, 290, 331, 339–41 non–possessory security 114, 125, 130, 133 real security 3, 113–21, 124, 128, 143, 149–50, 274–6, 283 security law 114, 120–2, 124–5, 128–30, 143–4, 149–50 security interest 116–18, 126, 131–3, 135, 138, 143, 149, 187 security rights 122–4, 131–2, 148, 150 seizure of security 190–2, 331–2, 337, 339–41

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

434

Index

self–help 191, 284, 286, 330, 332–3, 335, 337, 339, 344 senate 170, 381 senators 82–4 Senatus Consultum Silanianum 287, 380, 381 Senatus Consultum Turpillianum 342 Seneca (Elder) 63, 86, 98, 176 Septimius Severus 142 services 17, 35, 52, 66, 68, 87–8, 119, 160, 169, 173–6, 180, 196–8, 249 Servius (Sulpicius Rufus) 21, 23–9, 31 Severan period 18, 21 shareholders 248, 252, 279 slavery 11, 16–32, 35–72, 83, 95, 97, 102, 148, 163, 165, 194, 202, 258, 278, 280, 285, 288, 314, 356, 380, 381, 401, 403, 404, 407 actiones involving 29–31 in the familia Caesaris 42 revolts 19 slave–run businesses 59, 194, 278, 280 statuliberi 47, 51, 52, 55, 66, 69 vicarii, 58, 59 Social War 91, 93 societas 102, 105, 281 (see also partnership) socii 102, 105 socii (non–Roman) 91, 92 solvency 83, 114, 117–18, 121–2, 128, 131, 145, 250 specialization 160, 172–3, 176, 211, 248, 250, 256, 258–60, 277 stipulatio 145, 269 Sulla 381 summae honorariae 198 sumptuary laws 198 supply 23, 161, 171–2, 195, 197, 261 surety 5, 86, 113–14, 261, 275–6, 283, 335–8 (see also fideiussor) surplus 60, 122, 128–30, 188, 259 tabernae 162 tabulae alimentariae 133 tabulae communes 133 tabulae publicae 133 Tacitus 81 takings 313, 332 tax 14, 36, 40, 42, 44–5, 86, 122–3, 132–4, 146–7, 174, 198–9, 214, 237, 239, 257, 259, 265–6, 281–2, 291, 371

technology 12, 131, 148, 163, 171, 197, 231, 256, 281–2, 390 tenants 162–4, 189–92, 194–5, 200, 258 termini (land boundaries) 225 testament 37, 49–52, 55, 64, 66, 68–9, 71, 160–1, 184–5 (see also will) theft 54, 139, 332, 335, 341, 353–5, 357, 360–1, 366–7, 382, 403, 420 (see also furtum) Tiberius 179 title, generally 130, 136, 146, 149, 185–6, 188, 235, 249–50, 253–4, 256, 265–6, 269–71, 273–4, 276–7, 285, 287 title defect 269, 270, 273, 285 title purge 269, 273–4, 276, 285, 287 tort law 302–3, 308–10, 312–3, 317, 319–20, 323–4, 327, 339, 343–4, 348, 356–9, 364, 367–70, 372–3, 382, 387–8 trade 14, 19, 23, 100, 116, 124, 126, 132, 163, 166, 170, 174–5, 182, 211, 247–8, 250, 253, 258, 260, 263, 268, 277–8, 280, 284, 289–90, 410, 413, 421 export 170, 174–5, 180, 260 long–distance trade 166, 257, 259–60, 268, 272 traditio 271–4 tragedy of the commons 252 Trajan 55 transaction, generally 2, 3, 6, 22, 25, 26, 30, 46, 89–90, 95–6, 99–102, 118, 120–2, 127, 131, 138, 140–7, 170, 182, 187–8, 233, 237, 248, 252, 256, 258, 260–1, 263–4, 266–9, 271, 273, 276, 278, 280, 283, 286–8, 290–1, 338, 402–4, 411, 413–15, 418–19, 421–2 financial transaction 46, 119–20, 122 impersonal transaction 4, 119–20, 131, 252–3, 255–7, 259–62, 268, 282–3, 289–90 market transaction 4, 6, 232, 257, 289, 291, 402, 417 personal transaction 6, 253, 255–7, 261–2, 282–3, 285, 289–91, 402, 404, 417, 420–2 transaction costs 4, 104–6, 123–4, 131, 172, 194, 199, 248, 250–2, 264, 280, 289, 291, 365, 415, 419

OUP CORRECTED PROOF – FINAL, 23/4/2020, SPi

Index transfer obligations of transferee 321 of obligations 90, 92, 118, 249–50, 266–7, 269, 274, 364–65 of property rights 116, 123, 126, 127, 128, 129, 135–41, 146, 170, 175, 182–4, 187–8, 202–3, 250, 261, 266, 269–70, 273, 274, 276, 342 Triumvirate, Second 200 Twelve Tables, Law of 11, 47, 55, 81, 160, 183, 184, 186, 187, 202, 329, 330, 332, 334, 335, 337, 355, 366 Ulpian 24, 27, 30, 49, 56, 95, 99, 179, 265, 380, 407 usucapio 136, 137, 186, 273, 274, 277 usury 84–5, 260

435

utilitas publica 105 utility maximization 348–9, 389 Valentinian period 342 vindex 337, 339, 340, 356 waiver 323, 329 warranty 141, 321 wealth 1, 2, 12, 14, 23, 70, 82–3, 87, 121, 160, 164–5, 169–70, 173, 175, 178, 184, 198, 211, 243, 247, 258, 279, 336, 363–4, 370–1, 373 will 21, 37, 51–2, 66, 184, 338, 381 (see also testament) witnesses 139–41, 146, 187–8, 263, 269–70, 284, 337, 339 zero–sum game 121, 198 zoning policies 192