Exceptions to the Rule: The Politics of Filibuster Limitations in the U.S. Senate 9780815729969

Special rules enable the Senate to act despite the filibuster. Sometimes. Most people believe that, in today's par

247 23 5MB

English Pages [290] Year 2017

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Front Cover
Title Page
Copyright Information
Table of Contents
Introduction
Limiting the Unlimited: Debate in the U.S. Senate
Obscuring the Causal Chain: Majoritarian Exceptions as a Blame Avoidance Tool
Employing the Exceptions: The Case of Budget Reconciliation
The Policy Consequences of Procedural Choice: Programmatic Change Using Budget Reconciliation
Facilitating Gains and Blocking Pain: Creating Executive Branch Oversight Exceptions
Conclusion
Appendix Tables
Notes
Works Cited
Acknowledgments
Index
Back Cover
Recommend Papers

Exceptions to the Rule: The Politics of Filibuster Limitations in the U.S. Senate
 9780815729969

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

EXCEPTIONS TO THE RULE

EXCEPTIONS TO THE RULE THE POLITICS OF FILIBUSTER LIMITATIONS IN THE U.S. SENATE

MOLLY E. REYNOLDS

Brookings Institution Press Washington, D.C.

Copyright © 2017 THE BROOKINGS INSTITUTION 1775 Mas­sa­chu­setts Ave­nue, N.W., Washington, D.C. 20036 www​.­brookings​.­edu All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means without permission in writing from the Brookings Institution Press. The Brookings Institution is a private nonprofit organ­ization devoted to research, education, and publication on impor­tant issues of domestic and foreign policy. Its principal purpose is to bring the highest quality in­de­pen­dent research and analy­sis to bear on current and emerging policy prob­lems. Interpretations or conclusions in Brookings publications should be understood to be solely ­those of the authors. Library of Congress Cataloging-­in-­Publication data Names: Reynolds, Molly E. Title: Exceptions to the rule : the politics of filibuster limitations in the U.S. Senate / Molly E. Reynolds. Description: Washington, D.C. : Brookings Institution Press, 2017. Identifiers: LCCN 2017018782 (print) | LCCN 2017010290 (ebook) | ISBN 9780815729969 (paperback) | ISBN 9780815729976 Subjects: LCSH: United States. Congress. Senate—Freedom of debate. | Filibusters (Political science)—United States. | BISAC: POLITICAL SCIENCE / Government / Legislative Branch. | POLITICAL SCIENCE / Political Ideologies / Democracy. | POLITICAL SCIENCE / Political Process / Political Parties. | POLITICAL SCIENCE / Government / National. Classification: LCC JK1161 .R45 2017 (ebook) | LCC JK1161 (print) | DDC 328.73/0741—dc23 LC record available at https://lccn.loc.gov/2017018782 9 8 7 6 5 4 3 2 1 Typeset in Jenson Composition by Westchester Publishing Ser­vices

CONTENTS

One INTRODUCTION 1

Two

LIMITING THE UNLIMITED Debate in the U.S. Senate



OBSCURING THE CAUSAL CHAIN Majoritarian Exceptions as a Blame Avoidance Tool

9

Three 39

Four

EMPLOYING THE EXCEPTIONS The Case of Bud­get Reconciliation

79

Five

THE POLICY CONSEQUENCES OF PROCEDURAL CHOICE Programmatic Change Using Bud­get Reconciliation 125

Six

FACILITATING GAINS AND BLOCKING PAIN Creating Executive Branch Oversight Exceptions

147

Seven CONCLUSION 185

vi

Contents Appendix ­Tables

203

Notes 209

Works Cited

245



Acknowl­edgments

271

Index 275

One INTRODUCTION But ­every schoolboy in the United States knows that [the Senate] is practically the only parliamentary body in the world where the majority cannot transact the public business, and where the minority instead of the majority transacts the business of the country. SENATOR WILLIAM E. MASON (R-­I LL .), APRIL 21, 1897

The Senate is not a majoritarian body. SENATOR CHUCK SCHUMER (D -­N .Y.), MAY 10, 2005

I

n many other ways—­including the issues on its agenda and the demographic composition of its membership—­the U.S. Senate at the beginning of the twenty-­first ­century would be unrecognizable to a member of the body at the end of the nineteenth. The notion that s­ imple majorities do not rule, however, is a rare point of consensus across both time and party. Our understanding of the Senate as a slower-­moving, more deliberate body than the House of Representatives dates to the Constitutional Convention, where James Madison characterized the chamber as proceeding with “more coolness . . . ​[and] more system.”1 The chamber lost its ability to end debate with ­simple majority vote in 1806, and it took nearly a ­century of increasing obstruction before the cloture rule provided a supermajority solution, in 1917.2 Over the course of the twentieth ­century, the filibuster became a routine procedural tool that is often blamed for the gridlock and dysfunction that characterizes our con­temporary po­liti­cal system. 1

2

EXCEPTIONS TO THE RULE

This ability of a minority to obstruct legislative pro­g ress in the Senate permeates the understanding of deliberation and activity in Congress for academics, journalists, and ordinary citizens alike. From a scholarly perspective, the filibuster sits at the center of Krehbiel’s well-­known account of lawmaking in the separation-­of-­powers system.3 Work on gridlock and legislative productivity4 and executive nominations5 has similarly embraced the notion that the filibuster dictates what the Senate, and by extension, the House and the president, can achieve. Coverage of the Senate in the popu­lar press now also takes for granted the notion that virtually all legislative action requires the support of three-­fifths of the Senate. Discussions of specific bills are often framed as needing sixty votes for passage; in reference to the 2008 auto bailout, for example, the New York Times’s David Herszenhorn wrote that “passing any legislation to aid the auto companies would require 60 votes in the Senate.” 6 So ingrained is the effect of the filibuster rule that journalists regularly describe mea­sures that obtain more than fifty but fewer than sixty votes as failing, without additional discussion of why and how something that has majority support does not pass the chamber.7 ­There exists, however, a set of procedures in the Senate that complicates this account, which is so prevalent among Congress watchers of all stripes. Over the past nearly fifty years, Congress has repeatedly included in statutory law provisions that I call “majoritarian exceptions.” 8 By reallocating power within the chamber in three dif­fer­ent ways, ­these special procedures empower ­simple majorities and make operations of the Senate more majoritarian. Some prior work on ­these procedures explores them only in the context of broader arguments and not as an in­de­pen­dent object of interest.9 In other instances, the procedures are explored in depth but only as specific, substantive case studies10 or as a way of explaining a par­t ic­u ­lar set of legislative outcomes.11 In this book, I unify the narrow and the broad by analyzing systematically the creation, use, and policy consequences of ­these special procedures in the Senate. In chapter 2, I explore at some length what constitutes a majoritarian exception, that is, a provision included in statutory law that prevents some ­future piece of legislation from being filibustered on the floor of the Senate. A careful review of the historical rec­ord has identified 161 such provisions a­ dopted between the 91st and 113th Congresses (1969–2014). They cover a wide range of policy areas, including trade (such as the multiple provisions providing the

Introduction

3

president with fast-­track trade authority); foreign policy (including rules for the imposition or waiver of international sanctions); defense ­matters (such as procedures for closing military bases); the federal bud­get (including the pro­cess for developing and passing the congressional bud­get resolution); and health care (such as the provisions governing the adoption of proposed cuts in Medicare spending). The logic as to why majoritarian exceptions m ­ atter is s­ imple: by eliminating the possibility of a filibuster, they ease the pro­cess of building a co­a li­tion in f­avor of a par­tic­u­lar piece of legislation. B ­ ecause majoritarian exceptions apply only in specific circumstances, however, even close observers of Congress tend to think of each set of procedures in isolation, frequently describing even the best known examples, such as bud­get reconciliation, as “arcane.”12 By providing a systematic look at the exceptions together, as a single class of procedures, however, I am able to demonstrate how majoritarian exceptions represent an impor­tant procedural dynamic in the Senate in their own right. Indeed, one need only look at both the 2016 election campaign and its immediate aftermath to see the central role that majoritarian exceptions can play in public policymaking in the United States. While the presidential race was notoriously light on policy issues, one area on which then candidate Donald Trump focused heavi­ly was trade; debates over w ­ hether to ratify the Trans-­Pacific Partnership also featured prominently in the Demo­cratic primary between Hillary Clinton and Bernie Sanders. As shown in chapter 3, majoritarian exceptions ­under which trade deals come to the floor of the Senate protected from amendment and a filibuster have been central to the conduct of trade policy in the United States since the early 1970s. Chapters 4 and 5 explore a par­t ic­u ­lar majoritarian exception, known as bud­get reconciliation, that appeared in the headlines beginning the day a­ fter the election as a pos­si­ble mechanism for accomplishing some of the new unified Republican Congress’s biggest legislative goals, including repeal of the Affordable Care Act and tax reform.13 ­Those procedures, which protect certain bud­getary legislation from a filibuster and some amendments in the Senate, have been used to accomplish a range of significant policy changes since the 1980s. As demonstrated throughout the book, t­ hese kinds of high-­profile examples of the role of majoritarian exceptions are joined by many less noticeable, but still consequential, instances of policymaking that involve decisions to institute s­ imple majority thresholds for par­tic­u­lar legislation.

4

EXCEPTIONS TO THE RULE

THE ARGUMENT AND ITS BUILDING BLOCKS

Throughout this book, I document one way in which s­ imple majorities can be influential in Senate policymaking; to borrow a phrase from Krehbiel, I argue that the Senate is sometimes actively, rather than remotely, majoritarian.14 In par­tic­u­lar, the chapters to follow unite the theory and data with two principal arguments about the ends produced by this par­tic­u­lar form of majority rule. First, majoritarian exceptions ease the passage of the bills to which they apply. ­There are three potential components of a majoritarian exception—­protection from a filibuster on the floor, a prohibition on amendments, and a preclusion of committee obstruction—­and each reduces the hurdles that the mea­sure must clear on its way to passage. The committee-­related provisions, for example, reduce the chance that a ­simple majority of a committee can prevent a mea­sure from coming to the floor. When, thanks to a majoritarian exception, a bill is automatically reported out of a committee, it is impossible for the legislation to get stuck at that stage of the pro­cess; the same is true when the special procedures send a mea­sure directly to the floor.15 Decisions about when exceptions should be created and used, then, should be s­ haped by the fact that they w ­ ill make it more difficult to engage in ­future obstruction on the legislation to be considered ­under the special procedures. The second goal of majoritarian exceptions is to help the Senate’s majority ­ ill be party maintain its control of the chamber. As a result, the procedures w both created and used when d ­ oing so helps the majority party remain as such. In making this argument, I join a growing chorus of scholars who view the Senate as the home of influential parties rather than an individualistic body. Conceptually, much of this lit­er­a­ture portrays the chamber as the home of two competing partisan teams that work together to achieve shared goals at the expense of their partisan opponents.16 The majority team attempts to pass legislation it f­avors, while the minority team works to obstruct t­ hose initiatives. The majority party has a wide range of tools, both formal17 and informal,18 at its disposal as it attempts to enact its preferred policy agenda and maintain its majority status. In the context of majority maintenance, ­there are three basic ingredients for a successful defense of its status by the Senate majority party. First, majority party members must collaborate to change policies in ways preferred by constituents and generate a rec­ord of legislative accomplishment.19 Second, they

Introduction

5

must generate opportunities for individual members to produce accomplishments for which they can claim credit.20 Third, they must avoid blame for negative events.21 Expedited procedures can serve as a valuable tool as the majority party works to put each of ­these ele­ments together into a winning formula. To build this argument, I rely on several familiar assumptions. First, the individual members who constitute the majority caucus desire to be reelected.22 Second, rank-­and-­file members of the majority party delegate some of their power to the leaders of their party, who, in turn, assume responsibility for acting in the party’s interest—­that is, the party acts as procedural cartel.23 In this context, leaders must satisfy this obligation both when new procedures are created and when they are used. Fi­nally, the proximate shared goal of individual majority party senators is to enable their party to maintain its majority status;24 the benefits to a party’s members of having their party hold majority status are well documented empirically.25 As I make ­these arguments, I answer impor­tant substantive questions about how the Senate operates. Much recent work on the consequences of procedural reform in the Senate has leveraged the 2013 decision to make nominations to the executive branch and courts below the Supreme Court subject to simply majority cloture.26 In the first year ­after the change to the procedures, judicial nominees ­were confirmed more frequently and more quickly but ­were not significantly more liberal.27 Meanwhile, for other, nonjudicial appointments, confirmation rates increased in the first year, but nominations took longer to receive attention.28 ­There is ­little systematic knowledge, however, about the creation and use of expedited procedures for other legislation, and documenting ­these patterns is particularly impor­tant, given their extensive policy implications. Chapter  5 explores at length how the use of one par­tic­u ­lar exception—­the bud­get reconciliation process—­has had wide-­ranging consequences for mandatory spending programs such as Medicare, Medicaid, food stamps, and farm price supports. The examples discussed in chapters 3, 4, and 6 also include proposed and enacted exceptions involving the conduct of the war in Iraq, the sale of weapons to other countries, the negotiation of international trade agreements, the closing of military bases, and the review by Congress of regulations promulgated by executive branch agencies. The wide reach of this final exception alone suggests the breadth of the procedures’ potential policy consequences. In addition to addressing this substantive gap, this account also contributes to our understanding of from where institutions come. On one hand, a

6

EXCEPTIONS TO THE RULE

number of impor­tant accounts of Senate policymaking assume that institutions such as the filibuster are exogenous and immutable;29 both my theoretical and empirical investigations ­here demonstrate the limitations of that assumption. Instead, I join ­others who have illustrated how new procedures are created to achieve proximate po­liti­cal goals; indeed, notable works on the evolution of the filibuster have argued that Senate rules are changed in response to short-­ term po­liti­cal forces rather than to principled commitment to supermajoritarianism. One prominent account of the creation of the cloture rule (Rule 22) in 1917, for example, examines the po­liti­cal circumstances surrounding the mea­sure, whose passage was facilitated by the existence of new procedures for ending debate. Senate (majority) Demo­crats and President Wilson framed that bill, which permitted the arming of merchant ships during World War I, as a national security mea­sure, portraying the procedural question as a ­matter of policy. The new rules, they argued, w ­ ere needed if the Senate was g­ oing to enact a popu­lar and salient policy change.30 A similar dynamic holds for majoritarian exceptions. Fi­nally, by marshaling data on a range of examples in numerous policy areas, ­ ere provides useful context for the account of procedural change presented h arguments about ­whether, and ­under what circumstances, we should expect broader filibuster reform in the U.S. Senate. Exceptions to the filibuster rule, according to the evidence presented ­here, reflect the electoral priorities of the Senate’s majority party, even when adopting them requires the support of some minority party members. This explanation, when combined with the existing po­liti­cal science lit­er­a­ture, suggests that f­ uture changes to the rules are likely to be produced by po­liti­cal realities and not by senators’ principled positions on the role of unlimited debate in the chamber.

PLAN OF THE BOOK

The discussion above provides some context for our exploration of majoritarian exceptions. But a more thorough explanation of how ­these procedures fit in the broader landscape of unlimited debate in the Senate, how exactly they limit debate, and how I identify them in the historical rec­ord provides useful groundwork on which to build. In general, majoritarian exceptions can be divided into two general categories, largely based on the content of the under­lying legislative proposal that they shepherd to and through floor consideration, as described in chapter 2.

Introduction

7

One category involves efforts by the Senate to delegate some of its power to one or more actors, ­either within or outside the chamber. The actor or actors to whom this power is delegated are tasked with drafting a new policy, a­ fter which the proposal is sent to the floor of the Senate u­ nder expedited legislative procedures. The pro­cess for closing military bases is a well-­known case of this kind of exception. An in­de­pen­dent Base Realignment and Closure Commission (BRAC) is authorized by Congress to select bases for closure, and the legislation approving t­ hose se­lections cannot be filibustered or amended. ­These del­e­ga­tion procedures are explored in chapter 3, which shows how they can benefit the majority party by helping it solve internal collective action prob­lems. Chapters  4 and 5 focus on one par­tic­u ­lar majoritarian exception: the bud­get reconciliation procedure. Created in 1974, the reconciliation pro­ cess allows for changes to mandatory federal programs and revenue-­raising instruments to be made through a filibuster-­proof pro­cess that also restricts amendments. The history and development of the procedures are described in chapter 4, followed by a theoretical account that highlights how ­these par­tic­u­ lar features of the procedures can be leveraged to produce policy outcomes that reflect the majority’s preferences, making the caucus appear competent and enhancing its reputation in the eyes of voters. An empirical test and series of brief case studies illustrate how ­these dynamics have played out over the past thirty years. Chapter 5 explores ­whether the reconciliation procedures are actually used in a way that is likely to help the majority party achieve its goal of maintaining control of the chamber. I argue that the reconciliation pro­cess generates opportunities for majority party members to claim credit and avoid blame. ­Because the majority party’s ability to maintain its status involves defending dif­fer­ent sets of seats in dif­fer­ent electoral cycles, we should expect the programmatic changes made through the pro­cess to reflect ­these varying strategic concerns. I test this hypothesis using new data on programmatic reforms made using the procedures. Chapter 6 takes up the second category of exceptions, which seek to limit the president’s power to take unilateral actions in the face of a range of disincentives to do so. Depending on the degree to which Congress and the president prefer the same policy outcomes, the legislative branch may disapprove of a unilateral action taken by the president, e­ ither through an executive order, signing statement, or other method. By creating an executive branch oversight

8

EXCEPTIONS TO THE RULE

exception, Congress can make a specifically delineated unilateral action by the president subject to legislative approval. ­Because the mea­sure acceding to the president’s action is privileged for consideration and cannot be filibustered, Congress is guaranteeing, through a legislative check, that it has increased input in a par­tic­u­lar policy area. Take, for example, the provision of the International Security Assistance and Arms Export Control Act of 1976, which allows Congress to disapprove of presidentially proposed sales of major defense equipment. The resolution rejecting such a deal can be compelled out of committee by a highly privileged resolution a­ fter ten days and is limited to ten hours of debate on the floor of the Senate. Before ­these provisions ­were enacted, arms sales could be handled entirely within the executive branch, provided the president certified that the sale would “strengthen the security of the United States and promote world peace”—­a determination that was made for all proposed transactions by both Presidents Lyndon Johnson and Richard Nixon.31 Congress certainly had the power to respond to this presidential act through its regular legislative procedures before creation of the procedural exception. By changing its internal procedures for this par­tic­u­lar policy choice, however, Congress made it easier for itself to exert power in the policy area by creating opportunities for majority rule. Fi­nally, chapter 7 summarizes my findings and offers several implications of this work for the prospects of further procedural change in Congress. In par­tic­ u­lar, I discuss how existing exceptions have been used in the con­temporary Congress by senators to send messages to impor­tant constituencies outside the chamber and describe several dynamics that have been at play in the successful creation of new majoritarian exceptions in recent years. In the policymaking world, where the upper chamber is so often understood to be the sixty-­vote Senate, majoritarian exceptions are just that—­exceptions to the overall, prevailing dynamic that shapes co­ali­tion building in the chamber. Certainly, in most cases, the presence of Rule 22 can shape deliberation. Indeed, in the con­temporary era, in many instances, it does influence what the Senate, and by extension the House and the president, can accomplish. But as the pages to come show, as with so many t­ hings in the Senate, the story is not that ­simple. Let us begin.

Two LIMITING THE UNLIMITED DEBATE IN THE U.S. SENATE

A

s the United States began to recover from the ­Great Recession in late 2009, President Obama and congressional Demo­crats began to call attention to the large bud­get deficit, produced in part by the fiscal stimulus enacted in 2009 to prop up the then faltering economy.1 Obama even went as far as to call for a three-­year freeze on discretionary, non-­national-­security federal spending in his 2010 State of the Union address.2 Taking up the president’s N.D.) and Gregg (R-­ N.H.) introduced an charge, Senators Conrad (D-­ amendment to a resolution raising the federal debt limit that would create a Bipartisan Commission for Responsible Fiscal Action. This panel would be made up of two executive branch appointees (the secretary of the trea­sury and a second person selected by the president) and sixteen members of Congress, with the Senate majority leader, Senate minority leader, Speaker of the House, and House minority leader each selecting four members. The task force would be responsible for developing recommendations that “­will significantly improve the long-­term fiscal imbalance of the Federal Government.”3 While the amendment did not specify how this goal would be achieved, much of the floor debate on the m ­ atter focused on the possibility of cuts to Social Security and Medicare and of increases in taxes.4 The commission, then, would have the potential to make major changes to several of the most impor­ tant ways that individual voters interact with the federal government. Once 9

10

EXCEPTIONS TO THE RULE

submitted to Congress the proposal would be considered ­under expedited procedures in the Senate, including a protection from amendments. Although it was favored by the president, whose party held sixty seats in the chamber at the time of the vote, the Conrad-­Gregg amendment failed ­after only fifty-­three senators supported cloture.5 Among the twenty-­three Demo­crats who opposed the mea­sure was Senator Baucus (D-­Mont.), who claimed that creating the commission would limit the fundamental tasks of senators: “Two ­things most define a Senator. Senators can amend legislation, even with dif­fer­ent subjects. And Senators can debate legislation, sometimes at length. The Conrad-­Gregg proposal curtails both of t­ hose defining powers. The Conrad-­Gregg proposal completely eliminates the ability to amend. And the Conrad-­Gregg proposal sharply limits the ability to debate.” 6 As is demonstrated over the course of this book, proposals like the Conrad-­ Gregg amendment are a common feature of the con­temporary Senate. Like this 2010 proposal, they are often unsuccessful, but when Congress does choose to adopt them, they can have significant consequences for the policymaking pro­cess. Given Baucus’s characterization of unlimited amendment and debate as be­hav­iors that “most define a senator,” it is useful to consider how majoritarian exceptions fit into the broader context of the Senate’s procedural environment. To do so, I begin by exploring why limitations on debate are an exception to the chamber’s normal operating procedures. From ­there, we examine the earliest examples of restrictions on how long a mea­sure could be debated before laying out, in detail, the definition of a majoritarian exception. Next, to provide readers—­especially t­ hose from outside po­liti­cal science—­with context for the contribution of this account, I discuss how majoritarian exceptions differ from other congressional procedures. Fi­nally, to lay the groundwork for the chapters to come, I take up three impor­tant data-­related questions. First, how do we identify majoritarian exceptions in the historical rec­ord? Second, is ­there reason to believe that they are actually relevant soon ­after they are created (an assumption necessary for our argument about adoption meeting proximate po­liti­cal needs)? Third, given that ­these limitations on debate apply in specific issue areas, how do we mea­sure their policy content?

THE ORIGINS OF UNLIMITED DEBATE IN THE SENATE

To understand why limitations on debate in the Senate ­matter, it is impor­tant to understand why debate in the chamber is generally unlimited. Operation-



Limiting the Unlimited

11

ally, the possibility of unlimited debate in the Senate is built on three pillars.7 The first of ­these is the right of recognition. The presiding officer must recognize any senator who wishes to speak; Senate Rule 19, which governs floor debate, does not permit the presiding officer to simply decline to recognize a senator. If multiple senators request time to speak on a debatable question (such as a motion or bill), all must be recognized to speak before the chamber can vote on it. In general, senators are acknowledged to speak in the order they sought recognition, with the impor­tant exception that the leaders of the majority and minority parties enjoy preference in recognition, getting to jump the line to maintain orderly operations on the floor. Rule 19 does not place any limits on how long a senator may speak once recognized.8 The second pillar of unlimited debate is the absence of a germaneness requirement for debate. With the exception of the first three hours of consideration of legislative business on each calendar day (when debate must be related to the under­lying question), once recognized to speak, a senator may do so on any subject whatsoever. It is this lack of germaneness that helped produce the kind of filibuster famous in the popu­lar mind, such as when Senator D’Amato (R-­N.Y.) sang and read the same newspaper article repeatedly during a fifteen-­ hour floor speech in 1992.9 Certainly, senators engaging in extended debate often do use their floor time to speak on the subject at hand, but the absence of a requirement to do so helps ease the task of ­doing so. The third and final pillar of unlimited debate is the absence of a previous question motion in the Senate. Many other legislative bodies—­including the House of Representatives—­allow for such a motion, which lets a ­simple majority force a vote on what­ever under­lying question is currently being debated. In the Senate, however, no such motion exists—­though it has not always been absent from the chamber’s rules. Indeed, the Senate’s original rules, a­ dopted in 1789, allowed for such a motion. It was used infrequently during the Senate’s early years, however. The historical rec­ord suggests that when it was deployed, it was used to postpone consideration of a mea­sure rather than to force an end to debate and proceed to a final vote.10 Seeing the motion as duplicative of the motion to postpone (which persists to this day as part of Rule 22) and acting at the suggestion of recently indicted Vice President Burr, the Senate removed the motion from its rules in 1806.11 How unlimited debate ­shaped the chamber between 1806 and the adoption of Rule 22 (allowing two-­thirds of the Senate pres­ent and voting to invoke cloture and end debate) has been explored in depth elsewhere.12 But a brief review of the period provides useful context for this exploration of when

12

EXCEPTIONS TO THE RULE

the Senate chooses to limit debate (and other­wise make deliberation more majoritarian) for specific mea­sures. Obstruction, and effort to change the rules to contain it, was certainly pres­ent at times in the Senate in the earlier part of the nineteenth ­century. In one particularly prominent example from 1841, Henry Clay, Whig senator from Kentucky, frustrated by Senate Demo­crats’ derailing of the majority Whigs’ legislative agenda, proposed both a one-­hour individual debate limit and a resurrection of the previous question motion. Clay was unsuccessful in adopting his favored rules changes,13 but while the Demo­crats delayed action on his legislation program, much of it (with the notable exception of a bill resurrecting a national bank) was eventually enacted.14 Obstruction was also pres­ ent in debates over slavery and its proxy cousin, territorial expansion; between 1845 and 1861 ­there ­were five separate filibusters related to new territories.15 In the post–­Civil War period, however, the incidence of obstruction increased significantly. Scholars generally agree that the changing nature of the chamber’s workload was largely responsible for this escalation, but they posit dif­fer­ent mechanisms for why new responsibilities led senators to engage in more extended debate. B ­ inder and Smith argue that the Senate’s agenda expanded in the postwar period and that more issues requiring attention meant more opportunities to obstruct.16 Broader responsibilities also raised the chamber’s profile, which, coupled with a closer alignment of partisanship and preferences, created electoral pressures on members to exploit their procedural rights in pursuit of legislative victories. Koger describes how a larger agenda can increase the value of floor time: when t­ here are more issues to which the Senate must attend, the opportunity cost of spending time on any one ­matter also grows.17 ­Under this logic, when floor time is scarce, a bill’s proponents are less willing to let it languish for an extended period, since t­ here are other impor­tant issues with which the chamber could be dealing. Knowing this, opposition legislators w ­ ill engage in more obstruction, with the expecta­ ill be more successful. Fi­nally, Wawro and Schickler emphasize tion that they w a par­tic­u­lar cause of the Senate’s expanded agenda—­the entry of new states into the union—­and argue that as the chamber’s membership increased, it became more difficult to sustain a set of norms ­under which senators refrained from fully exploiting their available opportunities to obstruct legislation.18 Nineteenth-­century Senate obstruction, both pre– and post–­Civil War, relied heavi­ly not only on extended speech-­making but also on dilatory motions. Vari­ous motions can be considered dilatory, or “[intended only] to consume



Limiting the Unlimited

13

the time of the Senate in rounding up a quorum and taking a vote.”19 Mea­ sures that ­were targeted by t­ hese kinds of motions included a bill admitting California to the ­union in 1850 and an antidiscrimination amendment in 1872, and a remarkable 114 dilatory motions ­were offered during the efforts to or­ga­nize the Senate ­after the election of 1880. The heavy use of ­these tactics to effect obstruction during the nineteenth ­century was accompanied by attempts to limit their use, often involving rulings from the chair. In 1844, for example, ­after John Berrien, Georgia’s Whig senator, moved to ­table a nomination, the presiding officer held that the motion was not debatable. A vote to appeal that ruling followed, and the chair’s decision was upheld, establishing a pre­ce­dent that debate on an appeal could be cut off by moving to ­table that appeal.20 Rulings from the chair ­were similarly at the center of the ultimately unsuccessful attempts to impose majority cloture in the Senate during debate in 1890 and 1891 on the so-­called Force Bill, which provided for federal supervision of southern congressional elections.21 The ability of senators to use ­these kinds of motions to extend debate means that any consequential limit on debate must also restrict them; as shown below, this dynamic ­shaped the structure of majoritarian exceptions during the twentieth ­century. By the early twentieth ­century, senators found themselves trying to legislate in the face of higher levels of obstruction, as effected through extended speech-­making, repeated introduction of amendments, and pre­sen­ta­tion of other dilatory motions. The adoption of Rule 22, with its cloture provision, in 1917 represented a major effort to stem this tide—­a reform for which scholars have again offered dif­fer­ent explanations. One account ties the success of the reform to the failure, at the hands of Demo­cratic filibuster, of a mea­sure authorizing the arming of merchant ships for protection against the Germans. President Wilson, outraged at the obstruction, called immediately and publicly for a change to the chamber’s rules, connecting closely the prospects of a par­tic­ u­lar policy with the procedures u­ nder which it would be considered. Both the adoption of the rules change and passage of the Armed Ship Bill followed quickly, suggesting that “the adoption of a cloture rule in the Senate conforms to the pattern of Senate reform—or lack thereof—in response to short-­term, pragmatic po­liti­cal considerations.”22 Wawro and Schickler attribute the reform not to a specific pending policy change but rather to “a dramatic increase in the membership of the Senate due to the addition of new states [which] undercut the conditions for relation-­based legislating to work effectively.”23 In

14

EXCEPTIONS TO THE RULE

this new legislative environment, the threat from a committed majority was no longer sufficient to keep obstruction in check, and formal rules to address the situation ­were required.

THE EMERGENCE OF STATUTORY LIMITS ON DEBATE IN THE SENATE

While existing scholarly explanations differ in the weight they assign to short-­ term policy and po­liti­cal goals in shaping the evolution of unlimited debate in the nineteenth and early twentieth centuries, they all aim to explain why the Senate has changed its broadly applicable rules. The cloture motion a­ dopted in 1917, ­after all, can be used to end debate on any debatable question in the Senate—­a broad category that includes bills, resolutions, amendments, conference reports, treaties, and vari­ous other questions.24 Beginning in the nineteenth ­century, the Senate also periodically implemented narrower procedural innovations, aimed at limiting debate on debatable questions in a specific policy context only. Each of ­these individual changes may have limited reach, but they can have significant consequences in the issue area to which they apply. The first of ­these involved the counting of electoral votes in the ­middle to late nineteenth ­century. Beginning in 1865, Congress adjudicated disputed electoral votes ­under the Twenty-­Second Joint Rule, ­under which electoral votes could be rejected if only one h ­ ouse of Congress objected to receiving them; the elections of 1864, 1868, and 1872 all saw the votes of southern states rejected ­under this rule, but in none of ­those elections did the exclusion affect the ultimate outcome. By 1875, however, some in Congress felt that the Twenty-­Second Joint Rule made it too easy to reject electoral votes, and in late 1875 and early 1876, the Senate debated a bill written by the Republican senator from Indiana, Oliver Morton, that would have required the assent of both chambers to a refusal to count a state’s votes.25 When the ­houses separated to deliberate over an objection, each representative or senator would be able to speak for ten minutes, and total debate would be capped at ten hours.26 The bill did not pass, and while the Senate managed to repeal the joint rule separately, “the nation headed into the 1876 election without any mea­sure in place” for resolving Electoral College disputes.27 What was the rationale for preventing unlimited debate on objections to electoral votes? During debate on the 1875–76 version of the proposal, Senator John Sherman, Ohio’s Republican senator, argued that “it is not an improb-



Limiting the Unlimited

15

able suggestion that in high party times, ­under ­great excitement, one ­house might thus neglect or refuse to direct the main question to be put. . . . ​If you allow for more than two hours or open the subject for indefinite debate, you may defeat the object of the law.”28 As it turned out, the 1876 election, between Republican Rutherford  B. Hayes and Demo­crat Samuel  J. Tilden, provided exactly that kind of “­great excitement.” Four states—­Florida, Louisiana, South Carolina, and Oregon—­submitted two or more conflicting slates of electoral votes, and unlike in earlier elections, the way ­those disputes ­were adjudicated would determine the outcome of the election. Dif­fer­ent interpretations of the Twelfth Amendment, coupled with divided control of the House and Senate, left Congress deadlocked. To break the gridlock, Congress established the Electoral Commission in December 1877. Made up of five senators, five representatives, and five Supreme Court justices, the commission would rule on each of the disputed states. Their decisions would be final, u ­ nless both ­houses of Congress overturned them; debate on the commission’s rulings would be limited to two hours, with each member permitted to speak for ten minutes.29 The commission ruled for Hayes in each of the four states, but the resolution pro­cess was thrown into further turmoil when a previously unknown alternative slate of electors for Vermont was delivered to the joint session that was counting the electoral votes just days before the new president was to be inaugurated. It took the Compromise of 1877 (whereby Hayes promised to pull federal troops from southern states) and several key parliamentary decisions by Speaker of the House Samuel Randall (D-­Pa.) to ultimately bring the contest to a close in ­favor of Hayes.30 Although the mea­sure creating the Electoral Commission in 1877—­with its debate limit—­helped resolve the disputed election of 1876, it applied only ­ fter another closely contested election in 1884, Congress to that election. A passed a permanent mea­sure, the Electoral Count Act, in 1887.31 The act—­ which still governs the counting of electoral votes ­today—­created the so-­ called safe-­harbor deadline: If a state experiences a dispute over its election results but manages to adjudicate that contest using a pro­cess in place before votes ­were cast in advance of the deadline (which is six days before the electors are constitutionally required to meet), then Congress is bound to accept that state’s resolved results.32 In addition, the act delineates Congress’s role if the state is unable to ­settle the ­matter by the safe-­harbor deadline and sets out procedures for members of Congress to object to the counting of one or more electoral votes. An objection must be submitted in writing and, while it must

16

EXCEPTIONS TO THE RULE

be endorsed by at least one legislator in each chamber, the chambers must vote on the objection separately. Debate on an objection is capped at two hours, with each member able to speak once for not more than five minutes.33 When the time for debate has expired, “it s­ hall be the duty of the presiding officer of each House to put the main question without further debate.”34 Although the procedures for considering objections to electoral votes are certainly impor­tant—­indeed, they have come into play three times (in 1969, 2005, and 2017) since their creation35—­they do not concern a policymaking responsibility of Congress. The first known statutory limit on debate involving congressional policymaking—­the kind of procedure that is the focus of this study—­d id not occur for another fifty years, and it involved questions of improving the administrative per­for­mance of the federal government by reor­ga­niz­ing the executive branch.36 In 1937 two reports ­were released—­one conducted at the Brookings Institution on behalf of Congress and one written by the President’s Committee on Administrative Management—­making recommendations for improved management. The latter called specifically for a renewal, with some revisions, of authority granted to the president to restructure the executive branch. (Congress had previously delegated this ability to the executive branch, but that authority had expired in 1935.) Congress, however, skittish ­after President Franklin Roo­se­velt attempted to restructure the federal judiciary with what was perceived as his court-­packing plan, failed to enact a reor­ga­ni­za­tion bill in 1937 or 1938. When Congress reconvened ­a fter the 1938 midterm elections, however, it quickly ­adopted a narrower reor­ga­ni­za­tion proposal, the Reor­ga­ni­za­tion Act of 1939. Among the key differences between the failed and successful mea­sures was a provision that allowed Congress to reject a reor­ga­ni­za­tion plan developed by the president by means of a concurrent resolution agreed to within sixty calendar days of its issuance. The notion of a congressional disapproval mechanism for executive branch restructuring plans was not new, as it had been part of the first statute in which Congress explic­itly delegated reor­ ga­ni­za­tion authority, enacted in 1932.37 What was innovative, however, was the presence of a limitation on debate for the resolution that would prevent the president’s plan from being implemented. The Reor­ga­ni­za­tion Act of 1939 typifies the dif­fer­ent components of a majoritarian exception. Once a member introduces a resolution of disapproval, that mea­sure is referred to committee; if the committee fails to report out the resolution within ten calendar days, any member favoring the mea­sure may



Limiting the Unlimited

17

make a highly privileged motion to discharge it from committee. If the resolution is reported or discharged, a highly privileged, nondebatable, nonamendable motion to proceed to its consideration is in order. Debate on the mea­sure is capped at ten hours, and no amendments are permitted. Other motions are ­either not in order (such as a motion to recommit or a motion to reconsider) or not debatable (including a motion to postpone and appeals from decisions of the chair).38 The debate limitation was controversial, with one senator proclaiming that “when [the Senate] ceases to have the utmost freedom of motion, of resolution, of discussion, and of debate, it . . . ​has become merely a legislative mill, ­under the control of any majority that may be in possession of it.”39 What, then, was the rationale for including it? During debate on an amendment that would have required Congress to affirm a reor­ga­ni­za­tion plan before it could be implemented (rather than halt its execution), several senators discussed the under­lying policymaking challenge facing the Congress: legislators may have supported reor­ga­ni­za­tion in princi­ple but ­were concerned about potential negative effects of restructuring on specific agencies. Colorado’s Demo­cratic senator, Alva Adams, described this situation in general terms: It seems to me that Congress could well work out a plan of reor­ga­ni­za­ tion; but we have not done it. . . . ​We all say that reor­ga­ni­za­tion is de­ oing it; and t­ here are very practical sirable, but we do not set about d reasons why we do not do it. If we set about a ­wholesale reor­ga­ni­za­tion, we bring in enough bureaus and enough outside influences to destroy ­every effort we make. . . . ​If we cut out the functions or the personnel of a bureau, down upon us come all the friends of the bureau.40 Other senators ­were more direct. Francis Maloney, a Connecticut Demo­ crat, explained that “Congress finds itself unable to cope successfully with the ­matter of reor­ga­ni­za­tion . . . ​­because of the wide difference of interests over the land, ­because of sectionalism, or ­because one Senator or another has a favorite bureau, institution, commission, or organ­ization.” 41 Montana’s senator Burton Wheeler, also a Demo­crat, admitted that his concerns about the pro­cess stemmed from the possibility that certain agencies, such as the Bureau of Reclamation and the Forest Ser­vice, would be reor­ga­nized in a way that hurt his constituents; an unamendable proposal of which Congress could only disapprove and on which debate was limited, argued Wheeler, would strip him and his western state colleagues of their ability to serve the citizens they represented.42

18

EXCEPTIONS TO THE RULE

­These senators’ comments indicate that reor­ga­niz­ing the executive branch was, in the language of social science, a collective action prob­lem.43 The coun­ ere restructured to be more try would be better off if the federal bureaucracy w efficient and effective, but individual senators, concerned about their specific constituents and the “friends of the bureau” (what we might ­today call interest group allies), had an incentive to oppose a reor­ga­ni­za­tion plan. By restructuring the procedural environment, however, the majoritarian exception included in the 1939 Reor­ga­ni­za­tion Act helps solve this prob­lem. By giving the president, rather than Congress, the authority to develop a restructuring proposal, legislators make someone e­ lse responsible for the hard work of coming up with ideas about how to reor­ga­nize the federal government. Generating a plan, however, is only half the ­battle. Restricting amendments and limiting debate reduce the opportunities for specific senators to act on their individualized objections to the under­lying policy change, easing its path to implementation—­which, ­after all, is the best outcome for the country as a ­whole. On this dimension, the 1939 act might be considered a success, as it shepherded to completion five separate presidential reor­ga­ni­za­tion plans, including one that formally or­ga­nized the Executive Office of the President and another that constituted the pre­de­ces­sor to the Department of Health and ­Human Ser­vices, the Federal Security Agency.44 What is more, it established an impor­tant pre­ce­dent: majoritarian exceptions are often used to make it easier to achieve change in policy areas afflicted with collective action prob­lems.

WHAT IS A MAJORITARIAN EXCEPTION?

The 1939 Reor­ga­ni­za­tion Act contains but one example of a majoritarian exception, and at this point, it is useful to pause and outline, in some detail, what exactly constitutes such a procedure. My definition is as follows: Majoritarian Exception: A majoritarian exception is a provision, included in statutory law, that exempts some ­future piece of legislation from a filibuster on the floor of the Senate by limiting debate on that mea­sure. First, a note on terminology. While o­ thers have explored ­these procedures—­ and classes of procedures that are similar but not identical—­this term is novel. The most common existing language describes ­these procedures as “statutory limit[s] on debate,” 45 which is descriptively accurate. The limitation on debate



Limiting the Unlimited

19

takes effect, ­after all, ­because it is included in statutory law. My reasons for adopting a dif­fer­ent term are twofold. First, the decision to characterize the procedures as majoritarian reflects the fact that removing the need to build a supermajority co­ali­tion to end debate on a mea­sure empowers ­simple majorities in the Senate. Second, the choice to denote the procedures as exceptions signifies the degree to which eliminating the need to invoke supermajority cloture is anomalous, especially in the post-1970 period ­under study ­here. Let us take up several components of the definition in turn, again using the 1939 Reor­ga­ni­za­tion Act as an illustrative example. A majoritarian exception places a limit on debate using language similar to the following: Debate on the resolution ­shall be limited to not to exceed ten hours, which ­shall be equally divided between t­hose favoring and ­those opposing the resolution. A motion to further limit debate ­shall not be debatable.46 In practice, the length of this debate limit varies. It can be as long as fifty hours, as in the case of the annual congressional bud­get resolution, but its presence is what prevents senators who oppose the mea­sure from obstructing it by engaging in unlimited debate. As noted above, however, actually speaking at length on the Senate floor during debate on a bill or resolution is not the only mechanism by which senators engage in obstruction. Majoritarian exceptions also anticipate the possibility of obstruction on the motion to proceed: When the committee has reported, or has been discharged from further consideration of, a resolution with re­spect to a reor­ga­ni­za­tion plan, it ­shall at any time be in order . . . ​to move to proceed to the consideration of such resolution. Such motion ­shall be highly privileged and ­shall not be debatable. No amendment to such motion ­shall be in order and it s­hall not be in order to move to reconsider the vote by which such moton [sic] is agreed to or disagreed to.47 Such exceptions also anticipate the possibility of obstruction on other dilatory motions: No . . . ​motion to recommit, the resolution ­shall be in order, and it ­shall not be in order to move to reconsider the vote by which the resolution is agreed or disagreed to.48

20

EXCEPTIONS TO THE RULE

All motions to postpone . . . ​and all motions to proceed to the consideration of other business, ­shall be deci­ded without debate. All appeals from the decisions of the Chair relating to the application of the rules of the Senate . . . ​­shall be deci­ded without debate.49 By restricting the opportunities for senators to offer some dilatory motions, and by rendering other, potentially dilatory motions nondebatable, then, majoritarian exceptions also prevent senators who oppose the under­lying mea­sure from obstructing it in ways other than just extended speaking. Two additional components of the 1939 Reor­ga­ni­za­tion Act illustrate other common, but not required, components of majoritarian exceptions. Many majoritarian exceptions contain provisions similar to ­these, but a set of legislative procedures need not do e­ ither of the following to constitute a majoritarian exception: No amendment to the resolution ­shall be in order.50 If the committee to which a resolution with re­spect to a reor­ga­ni­za­tion plan has not reported it before the expiration of ten calendar days a­ fter its introduction . . . ​it ­shall then (but not before) be in order to move ­either to discharge the committee from further consideration of such resolution. . . . ​Such motion may be made only by a person favoring the resolution, ­shall be highly privileged . . . ​and debate thereon s­hall be limited to not to exceed one hour. . . . ​No amendment to such motion ­shall be in order and it ­shall not be in order to move to reconsider the vote by which such motion is agreed to or disagreed to.51 The first of ­these provisions corresponds to the fact that any majoritarian exception prevents the designated legislation from being amended, e­ ither in committee or on the floor. Roughly 86 ­percent of majoritarian exceptions prevent amendments to the filibuster-­proof mea­sure on the floor. An additional 4  ­percent permit only germane amendments. Both of t­hese restrictions on amendments represent a significant departure from the Senate’s usual procedures, which allow for amendments to bills regardless of topic (see Senate Rule 16).52 The second provision, meanwhile, reflects the fact that many majoritarian exceptions provide a mechanism to prevent obstruction of the legislation by the committee with jurisdiction over the mea­sure. In some cases, the time the panel can spend deliberating over the legislation is often limited by a firm day



Limiting the Unlimited

21

limit. A ­ fter a prescribed number of days have elapsed, e­ ither the bill is automatically placed on the Senate calendar for debate or it can be forcibly discharged from the committee by a privileged, filibuster-­proof motion. Other instances, moreover, stipulate that the protected mea­sure be reported directly to the floor, bypassing committee consideration altogether.53 In addition to this language setting out exactly how debate is to be limited and how consideration is other­wise expedited, majoritarian exceptions also often (but not always) contain some variant of what Bruhl refers to as the “disclaimer clause”: The following sections . . . ​are enacted . . . ​with full recognition of the constitutional right of ­either House to change such rules (so far as relating to the procedure in such House) at any time, in the same manner, and to the same extent as in the case of any other rule of such House.54 This clause, and other related language, is discussed in detail below in the context of the strategy for identifying majoritarian exceptions in the historical rec­ord. Its presence also denotes another impor­tant feature of majoritarian exceptions: that they have “the same force and effect as the standing rules ­either ­house adopts by ­simple House or Senate resolution to govern its own organ­ization and procedure.”55 As a result, they can be changed by the same mechanism. Making clear that majoritarian exceptions are functionally equivalent to all other rules that govern debate helps insulate them from the criticism that they violate the Supreme Court’s prohibition on legislative entrenchment56—­that is, one Congress binding the legislative authority of its successors. As Bruhl describes, The hallmark of entrenchment is the enactment of a directive that can be changed only by a pro­cess more burdensome than that which created it. Yet [a statute containing a majoritarian exception] does not purport to be irrepealable by a f­ uture statute. It can be amended or repealed just like any other.57 A final key definitional point involves the distinction between a majoritarian exception itself, on one hand, and the legislation to which it applies, on the other; I refer to the latter as a mea­sure “covered” by a majoritarian exception. The majoritarian exception represents the procedure itself, ­adopted as part of a statute by a given Congress at a given point in time. Measures—­whether

22

EXCEPTIONS TO THE RULE

they be bills or resolutions—­that are covered by the exception may appear on the Senate’s calendar at any point ­after the procedure is created. Covered legislation generally takes two forms, which I refer to as approval mea­sures and disapproval mea­sures. The former function similarly to regular bills and resolutions in that they contain legislative language, and what­ever policy changes are articulated in that legislative language do not take effect ­unless Congress affirmatively approves of mea­sure.58 The disapproval form, meanwhile, functions differently, in that it delineates a par­tic­u­lar policy change that w ­ ill not take effect if the mea­sure is agreed to. This, for example, is how the majoritarian exception covering proposals to close military bases (explored in chapter 3) operates. An outside commission develops a list of which military installations should be closed or realigned. The commission delivers its recommendations to Congress, which can disapprove of the proposal in its entirety, without amendment. Only if the president signs that mea­sure, or if a veto is overridden, is the proposed policy change rejected.59 We can divide majoritarian exceptions into two substantive categories. Roughly 40 ­percent of the proposed procedural changes are del­e­ga­tion exceptions. ­These are procedures that give power to a designated actor to develop a proposal for policy change and then prevent that mea­sure from being obstructed on the floor of the Senate. Approximately 53  ­percent are executive branch oversight exceptions, in which Congress gives procedural protection to certain mea­sures that approve or disapprove of presidential actions.60

HOW ARE MAJORITARIAN EXCEPTIONS DIF­F ER­E NT FROM OTHER CONGRESSIONAL PROCEDURES?

The preceding section has established what a majoritarian exception is. It might now be helpful to outline what distinguishes it from other, related congressional procedures—­that is, to explain what majoritarian exceptions are not. To some, this may seem like an odd choice. Why devote several pages to describing what is not being studied? The foregoing explanation is meant to accomplish two goals. First, it helps distinguish what the par­tic­u­lar contribution of this account is by differentiating majoritarian exceptions, and their method of adoption, from other procedures explored previously. Second, ­because concepts relevant to this analy­sis come from multiple academic disciplines as well as from Capitol Hill prac­ti­tion­ers, discussing key overlaps and differences ensures that all readers begin the following chapters from a shared perspective.



Limiting the Unlimited

23

First, majoritarian exceptions differ from many other congressional procedures in their method of adoption. ­Under the Constitution, each ­house of Congress has plenary power over its own rules, and ­t here are several ways in which each chamber may alter ­t hese internal procedures. In the case of the Senate, the first of t­ hese involves changing the Standing Rules of the Senate. This is usually done using a s­ imple resolution (denoted S. Res.). Ultimate adoption of a change to the Standing Rules requires a ­simple majority vote, but ­because a s­ imple resolution is debatable, a supermajority may be required to end debate if a sufficient minority chooses to obstruct the proposed reform. While invoking cloture generally requires three-­fifths of the full membership of the Senate, the threshold is higher for changes to the Standing Rules. Invoking cloture on such a change requires two-­thirds of senators voting, or sixty-­seven votes, if all members are pres­ent and voting. In addition, the motion to proceed to debate on a resolution altering the Standing Rules is also debatable, and invoking cloture on that motion also appears to require two-­thirds of ­those members pres­ent and voting. In effect, then, “in cases in which opponents are willing to carry on a filibuster, it can become necessary to obtain supermajority support (by two-­thirds of Senators pres­ent and voting) in order to bring the Senate to the point at which it can vote on a proposal to amend Senate Rules. An ordinary ­simple majority of Senators voting can then adopt the proposal itself.” 61 One such change to Senate rules a­ dopted in this manner in January 2013 was a reform to consolidate the motions necessary to send a bill to conference.62 A second method of rule change that uses a ­simple resolution as its vehicle is a standing order. Some standing o­ rders remain in effect in­def­initely, barring a decision by the Senate to change them, such as the standing order ­adopted in 2011 designed to limit secret holds; a full list of such ­orders appears in the Senate Manual.63 Other standing ­orders, such as one limiting debate on the motion to proceed and guaranteeing opportunities for minority party amendments, ­adopted in 2013, are only in effect for the single Congress in which they are approved.64 Unlike changes to the Standing Rules, invoking cloture on a resolution implementing a standing order (and on the motion to proceed to its consideration) requires only three-­fi fths of the Senate’s membership (that is, sixty votes). A third technique by which the Senate can change its procedures is to establish new pre­ce­dents, that is, to “reform by ruling.” 65 To illustrate how this

24

EXCEPTIONS TO THE RULE

approach proceeds, a review of the 2013 modification of Rule 22 for executive branch and some judicial branch nominations is useful.66 On November  21, 2013, ­a fter introducing a successful motion to reconsider a failed cloture vote on the nomination of Patricia Millett to serve as a judge on the D.C. Cir­cuit, Senate majority leader Harry Reid (D-­Nev.) raised a point of order that “the vote on cloture ­under rule 22 for all nominations other than for the Supreme Court of the United States is by majority vote.” 67 The chair ruled against the point of order, ­after which Reid appealed the ruling. ­Because the under­lying question was nondebatable, the appeal was also nondebatable, prompting an immediate vote on which a ­simple majority was necessary to overturn the chair’s ruling. Fifty-­two senators voted in f­ avor of Reid’s appeal, thus creating a new pre­ce­dent ­under which invoking cloture on a nomination (except for a nominee to the Supreme Court) requires only a ­simple majority.68 This example illustrates the steps necessary to achieve reform by ruling. First, a supporter of the new pre­ce­dent raises a point of order asserting a new interpretation of an existing rule or pre­ce­dent. Second, the chair—­who is required “to rule in accordance with existing rules and their established pre­ce­ dential interpretation” 69—­issues a ruling on that point of order. ­Here, the chair has two options: to interpret the existing rules in a way that f­ avors the new pre­ ce­dent or in a manner that opposes it. In the case of the former, we would expect the opponents of procedural change to appeal the ruling, at which point a ­simple majority of supporters could approve a nondebatable motion to ­table the appeal, thus establishing a new pre­ce­dent by ­simple majority vote. In the 2013 case, the latter option occurred. The chair ruled against the new reading of Rule 22, placing the burden on supporters of the new pre­ce­dent to overturn the chair’s ruling. ­Because the appeal was nondebatable, this vote required only a ­simple majority to be successful; if, as is generally the case, the appeal had been debatable, overturning the ruling of the chair would have been more difficult, since debate could have continued in­def­initely. In sum, then, reform by ruling requires several steps, with the key one being the placement of “a question before the Senate that, if agreed to by a ­simple majority, would result in limiting or closing debate on the rules change proposal.” 70 If executed correctly, reform by ruling requires only a ­simple majority. Enacting a majoritarian exception, however, looks more like changing the Standing Rules or adopting a standing order. ­Because majoritarian exceptions are contained in statute, they are subject to all the same hurdles as other pieces of legislation. They must gain, at minimum, the support of a majority in the



Limiting the Unlimited

25

House of Representatives, a majority in the Senate, and the signature of the president. If a sufficient minority chooses to obstruct a mea­sure that would create a majoritarian exception in the Senate, then its supporters must overcome the threat of a filibuster. And if the president opposes the procedural change—­which, as chapter 6 explains, he might well have reason to—­then its congressional champions need an even larger co­a li­tion of two-­thirds of the members in each chamber to create the new procedures. While its tailored nature may make it appear at first blush that creating a majoritarian exception might be an easier way to limit debate, the coalition-­building challenges associated with moving one to passage make clear that may not be true. This distinction between kinds of procedural changes based on their method of adoption is not the only important difference between majoritarian exceptions and other, related procedures governing how mea­sures are considered in the Senate. ­There are also several key conceptual differences worth exploring; in each of ­these cases, majoritarian exceptions represent a subset of a broader class of legislative procedures. ­Here, I consider three, drawn, in turn, from po­ liti­cal science, ­legal scholarship, and congressional prac­ti­tion­ers. Legislative Veto Legislative vetoes have been defined as “statutory provisions that reserve for Congress the ability to overrule or other­wise affect policy-­making powers that have been delegated to executive officials without having to pass subsequent legislation.” 71 Legislative vetoes share the “basic general characteristic of making delegated executive power or discretion directly contingent on further legislative review and control,” 72 but the exact mechanism varies. In some cases, overturn­ ouses of ing an action of the executive branch requires the consent of both h 73 ­ ouse must act. In addition, a Congress, while in other instances, only one h legislative veto provision may give congressional actors the power to prevent an executive action through a disapproval mea­sure, or it may prohibit the executive from implementing a decision u ­ nless Congress affirmatively approves of it. Some legislative veto provisions do indeed limit debate on the approval or disapproval resolution addressing an executive action. ­Others, however, do not specify the procedures ­under which Congress would consider a relevant mea­sure.74 When ­there is no additional language governing how such a mea­ sure would be considered, an opponent of a disapproval resolution (in other words, a supporter of the secretary’s proposal) would be ­free to obstruct it.

26

EXCEPTIONS TO THE RULE

Legislative veto provisions ­were quite common before 1983, when the Supreme Court declared them unconstitutional in INS v. Chadha, and, in strategically modified form, they continue to be an impor­tant part of congressional policymaking.75 Many, however, do not qualify as majoritarian exceptions ­because they do not limit debate on the mea­sure overturning the president’s action. Between 1981 and 1983, for example, ­there ­were 397 separate bills containing legislative veto provisions, but only 150 (or roughly 38 ­percent) of ­these contained language providing for expedited consideration of the mea­sure approving or disapproving of the president’s action.76 Statutory Legislative Procedures A second category of procedures—­perhaps most familiar to Capitol Hill prac­ti­tion­ers—­that is related to, but broader than, majoritarian exceptions are what the parliamentarian of the House of Representatives refers to as “statutory legislative procedures”; a list of ­these procedures appears in the House Rules and Manual.77 Many of the rules listed by the parliamentarian do satisfy the definition of a majoritarian exception b­ ecause they contain language limiting debate. This inventory plays a key role in our strategy for identifying majoritarian exceptions in the historical rec­ord. Not all procedures on the statutory legislative procedures list, however, are majoritarian exceptions. Some, though they provide for an expedited path for a bill or resolution through committee or to floor consideration (or both), do not contain an overall limit on debate.78 Thus like with the legislative veto provision described above, obstruction would remain pos­si­ble. In addition, ­because the list is included in a House document and not a Senate one, it includes procedures that limit debate in the lower chamber but not the upper one. One such example was enacted in the 96th Congress as part of the Coastal Zone Improvement Act of 1976. According to this act, any regulations promulgated by the secretary of commerce ­under the earlier Coastal Zone Management Act of 1972 could be disapproved if both chambers ­adopted a concurrent resolution to that effect, but only in the House was debate on that legislation limited (to ten hours).79 A second example comes from the Medicare Prescription Drug, Improvement, and Modernization Act of 2003, in which the president was granted the power to develop a proposal to address pending Medicare shortfalls.80 Debate on that mea­sure is limited (to five hours of general debate and ten hours on amendments) in the House.



Limiting the Unlimited

27

In the Senate, however, while debate is limited to two hours on a motion to discharge the bill from the Finance Committee, ­there is no time limit on floor consideration of the legislation itself.81 Related Concepts from the L ­ egal Lit­er­a­ture In addition to ­these narrow, legislative concepts to which majoritarian exceptions are related, but not identical, the l­ egal lit­er­a­ture explores several relevant but not identical notions. The first of ­these is what Bruhl calls “statutized rules,” instances in which “the legislature uses a statute to prescribe its internal procedures.” 82 The rationale for creating such rules, argues Bruhl, is that they “prevent . . . ​Congress from engaging in certain types of procedural opportunism”; and by enshrining them in statute Congress signals a more power­ful commitment to avoiding mischief.83 Several of Bruhl’s examples of statutized rules, including the procedures for considering trade agreements, appear in ­later chapters as impor­tant majoritarian exceptions, but ­others fall outside our definition. The Legislative Reor­ga­ni­za­tion Act of 1946, which laid the groundwork for the modern committee system, for example, fits Bruhl’s definition but not ours. Both Garrett and Tiefer articulate similar concepts—­framework legislation and laws about lawmaking, respectively—in which the unit of analy­sis is the law rather than the rule.84 Garrett defines framework laws as ­those that “establish internal procedures that ­will shape legislative deliberation and voting with re­spect to certain laws or decisions in the f­ uture.” 85 Tiefer describes laws about lawmaking as “structural laws that facilitate the enactment of other­wise controversial non-­consensus bills by sparing them from some of the deliberative steps.” 86 Both of ­these concepts share with majoritarian exceptions a focus on the applicability of procedures created by statute to specific, ­future decisions.87 At the same time, a piece of legislation need not contain a debate limitation to qualify as ­either framework legislation or a law about lawmaking. A law can spare a ­future mea­sure from “some of the deliberative steps” 88 without ­going so far as to limit floor debate. Garrett, in par­tic­u­lar, explores several pieces of framework legislation that “shape legislative deliberation and voting” 89 without limiting debate, such as the Unfunded Mandates Reform Act and the Bud­get Enforcement Act. Whereas statutized rules, framework legislation, and laws about lawmaking are all broader categories that include majoritarian exceptions, the final

28

EXCEPTIONS TO THE RULE

relevant ­legal concept, recusal legislating, is narrow.90 In situations where “Congress recognizes that its own structural weaknesses prevent meaningful action on an issue on which ­there is policy consensus,” recusal legislating helps Congress address “the institutional or po­liti­cal hurdles that often seem to prevent the passage of impor­tant legislation.”91 Recusal legislation accomplishes this by delegating responsibility for developing a proposal to an actor other than a federal agency; that proposal then takes effect automatically, ­unless Congress, using expedited procedures, disapproves of the action. This setup should feel familiar from the description of the 1939 Reor­ga­ni­za­tion Act outlined above, and indeed, many majoritarian exceptions follow this basic setup—­but not all do. Many majoritarian exceptions are aimed at ensuring that Congress has a say in policy areas usually dominated by the executive (see chapter 6, this volume). Although many of the benefits of recusal legislating identified by Teter, including the ability to engage in specialized policy development without interference, may also apply to majoritarian exceptions, the procedures ­under study ­here represent a broader phenomenon.

IDENTIF YING MAJORITARIAN EXCEPTIONS IN THE HISTORICAL REC­O RD

Previous work on procedural change in Congress that explores multiple instances across time tends to examine w ­ hether we observe new procedures being 92 created in a given Congress. By comparing sessions in which procedures ­were altered to counterfactual observations in which the procedures w ­ ere left alone, ­these accounts test ­whether vari­ous Congress-­level ­factors, such as the relative strength of the majority and minority parties, are the determinants of procedural change. The unique status of majoritarian exceptions within the body of congressional procedure, however, allows us to identify an alternative set of counterfactual observations: changes to procedures that ­were actually introduced but ­were not successfully enacted into law. Unlike many other procedural innovations, majoritarian exceptions’ path to creation goes through the regular legislative pro­cess with an ordinary piece of legislation as the vehicle. The search strategy outlined below is, like any attempt to identify all pos­si­ble examples of congressional phenomenon, imperfect. It does not, for example, capture all proposals that ­were only formally proposed as amendments.93 We also might have reason to believe that restricting the analy­sis to only successful and un-



Limiting the Unlimited

29

successful proposals might obscure impor­tant variations in which potential exceptions are actually made into formal proposals. (In the empirical chapters to follow, I take steps to address this concern.) Despite ­these inevitable shortcomings, the data used in this analy­sis represent the most comprehensive look at ­these procedures to date. Focusing on proposals, rather than simply congress-­ level observations, also allows us to explore the policy content—­and thus the consequences—of individual changes. The mea­sures that enact majoritarian exceptions generally carry special language that makes clear the constitutional source of the authority used to change Congress’s internal procedures through the legislative pro­cess.94 This fact served as the first component of my search strategy. Generally, this authority derives from two sources. In the first instance, the existence of an exception in a bill is often signaled by language stating that a given provision in the bill is enacted “as an exercise of the rulemaking power of the Senate and House of Representatives.”95 A second clause indicates that the exception is a­ dopted “with full recognition of the constitutional right of e­ ither House to change the rules”—­a right that comes from Article I, Section  5 of the Constitution, which states that “each House may determine the Rules of its Proceedings.”96 Several slight variations of ­these phrases exist—­for example, reversing the order of the chambers in the first phrase—­but the language provides the basis for a set of search terms that can be used in the ProQuest Congressional database of all full-­text bills introduced in both the House and Senate dating back to 1789. (A full list of search terms appears in appendix ­table A1.) The second component of my search enabled me to confirm that search terms drawn from ­these two clauses identify as much of the relevant universe ­ oward this end, I relied on two sources delineating known, of cases as pos­si­ble. T successful exceptions. The House Manual contains a section titled “Legislative Procedures Enacted Into Law.” Many of ­these procedures meet the definition of a majoritarian exception. In addition, in their brief case study of enacted limitations on debate in the Senate, B ­ inder and Smith provide examples of ­these provisions.97 Neither source serves as a comprehensive list of successful enactments for our purposes. The House list, for example, omits previously created procedures that are no longer in force. And B ­ inder and Smith note that their list is made up of “selected laws.” Nonetheless, known cases serve as a useful benchmark for evaluating the success of the l­egal clauses as initial search terms.

30

EXCEPTIONS TO THE RULE

Using the disclaimer clauses as search terms, I located 83 ­percent of the laws containing Senate-­related provisions included in the House Manual’s list. Although this coverage rate is reasonable, I improved it by leveraging the fact that many majoritarian exceptions refer to other existing provisions. In par­ tic­u­lar, rather than stating specifically the debate limit for a ­future protected bill, new exceptions often cite the existing procedures established in an entirely separate context. For example, rather than stating that debate on a protected bill is capped at twenty hours, a mea­sure might say “the provisions of section 151 of the Trade Act of 1974” apply to the ­future resolution.98 To identify additional terms, then, I consulted the legislation located using the first-­ stage search of disclaimer clauses, compiling a list of other bills to which they referred to establish the particulars of a new set of procedures. (­These additional, bill-­referent search terms are also listed in appendix ­table A1.) Before the Supreme Court’s decision in INS v. Chadha, ­there was substantial overlap between legislative veto provisions and majoritarian exceptions. I leveraged this for the final component of my search strategy. For the 96th, 97th, and 98th Congresses, the Congressional Research Ser­v ice published reports detailing e­ very bill, both proposed and enacted, that contained a legislative veto provision.99 Although in a majority of ­these mea­sures, the legislative veto did not incorporate a majoritarian exception, the reports served as a useful check on the coverage of the search terms outlined in appendix ­table A1 (and, in some cases, informed the inclusion of additional reference provisions).

MAJORITARIAN EXCEPTIONS’ TIME HORIZON

Both chapter  1 and the historical discussion above pres­ent an account that emphasizes how short-­term po­liti­cal considerations have ­shaped choices about procedural change. Some might argue that the policy-­specific nature of majoritarian exceptions makes this claim less plausible. If the current Senate majority seeks a procedural change to help it achieve some partisan goal, it must reasonably expect that the situation anticipated by the exception w ­ ill come to pass in short order. The new procedures are of ­little value to the current Senate majority party if the mea­sure they protect does not appear on the agenda in the near ­future. We know, for example, from the 2013 change in the cloture threshold for executive branch and lower court nominations that it is pos­si­ble to observe effects of procedural innovation quickly.100 How might we forecast when parti-



Limiting the Unlimited

31

san advantages might accrue from majoritarian exceptions? Senators’ expectations ­will vary across instances of the procedures. On one end are exceptions that, in delegating responsibility for developing a policy proposal to a set of internal or external actors, provide a hard deadline for when that proposal is to arrive in the Senate for consideration. The statute authorizing the most recent round of military base closings, for example, carried a deadline for reporting the recommendations to Congress of September  8, 2005.101 Any electoral consequences of the base closings, then, could reasonably be expected to be felt in the 2006 Senate elections. A second relevant example is more recent. The Bud­get Control Act (BCA) of 2011—­best known for ending a showdown between Congress and the president over ­whether to raise the debt ceiling—­contained a provision creating the Joint Select Committee on Deficit Reduction, a congressional entity tasked with developing a proposal to cut the federal deficit by $1.2 trillion over ten years. This package of bud­getary reforms was to be completed in the fall of 2011 and then approved by both chambers by January 2012;102 in the Senate, the mea­sure was covered by a procedural exception with a limit on floor debate of thirty hours. However, the committee failed to reach agreement on a set of cuts,103 rendering moot the existence of the special procedures to consider them on the Senate floor. Although choices by the members of Congress who w ­ ere developing the proposal meant that the procedures ­were never actually used, only a few months elapsed between the creation of the procedures and the time at which they would have been deployed. For senators hoping to claim credit for cutting the deficit or forestalling the deep automatic spending cuts realized in the absence of a congressional proposal ­under the BCA, it was clear that any potential electoral gains of revisiting the issue would come before the 2012 elections. On the opposite end of the spectrum are exceptions that increase the Senate’s capacity to check the unilateral power of the president but for which the timing of the president’s next relevant action is uncertain. Vari­ous provisions of the Arms Export Control Act of 1976, for example, give Congress the opportunity to prohibit proposed arms sales from proceeding. The president is not required, however, to sell defense articles on a prescribed timetable—or ever, if he wishes to take U.S. security policy in that par­tic­u­lar direction. Creating the exception ensures that Congress ­will be able to weigh in on the president’s decisions in the f­ uture, but it does not necessarily increase the institution’s knowledge of when that opportunity for input ­will pres­ent itself. For ­these more uncertain policy areas, it is difficult, if not impossible, to mea­sure the beliefs of the Senate at the time of the passage of the procedural

32

EXCEPTIONS TO THE RULE

change about when the topic would reemerge on the congressional agenda. To investigate the reasonableness of my claim that the chamber assumes that issues ­will recur quickly, I examined each enacted procedural change to determine how soon a­ fter its creation it could have been used by Congress to affect policy outcomes. In the most straightforward cases, this date of first relevance represents the ­actual introduction of a resolution ­under the special, filibuster-­proof procedures prescribed in the legislation. In the case of the Arms Export Control Act, for example, the first attempt by Congress to halt an arms sale proposed by the president occurred in September  1976, when Senator Nelson (D-­Wis.) introduced a resolution that would have prevented a proposed sale of defense equipment to Saudi Arabia.104 Although Nelson’s effort to use the new procedural exception was unsuccessful, the brief period of time that elapsed between the procedure’s creation and its application—­less than three months—­suggests that, before passage of the act, senators would most likely have believed that opportunities to weigh in on arms sale policy would be soon in coming. Similarly, in September 1996 Senator Lott (R-­Tenn.) introduced the first resolution to overturn a rule proposed by the executive branch (regarding hospital reimbursement ­under Medicare) ­under the provisions of the Congressional Review Act, which had been enacted five months earlier.105 That resolution was not successful—­indeed, only one mea­sure exercising Congress’s authority u­ nder the CRA had even been signed into law by the end of the Obama administration106—­but its existence is a reliable indicator of the return of the under­lying issue to the congressional agenda. That it took only five months for a senator to use the new oversight powers afforded to him ­under the CRA suggests that, before the law’s enactment, members of the Senate could have reasonably expected that any electoral rewards could be quickly realized. In other cases, a policy change to which an exception could apply is proposed, but Congress declines to take the procedurally privileged action afforded to it in response. If t­ hese situations arise close to the enactment of the procedural change, even absent affirmative steps by the Senate to use the new procedures, the reemergence of the issue implies that legislators could have sensibly assumed that any electoral gains from revisiting the topic would be quick in coming. Take, for example, the provision included in the International Security and Development Cooperation Act of 1980 that endowed Congress with a procedurally protected resolution of disapproval of export licenses for commercial sales of defense articles or ser­vices valued at



Limiting the Unlimited

33

over $100 million to other, non-­NATO countries.107 Before the legislation, arms sales carried out by the government ­were subject to potential veto by Congress, but the 1980 law gave the legislature the same authority to prevent wholly private transactions.108 The first deal eligible to be reviewed by Congress ­under this law was a sale to Sweden in early 1981, which the legislature allowed to proceed unimpeded.109 The mere existence of the transaction, however, provides a reliable indication of the frequency with which the issue—­large defense deals executed by private companies—is likely to recur on the congressional agenda. A second illustrative example of this dynamic involves U.S. counternarcotics efforts in Colombia, which, since the mid-1990s, have involved aid to both the Colombian National Police and the Colombian military. In 2000, when President Clinton proposed funneling increased aid to the Colombian military as part of this program (a proposal designated Plan Colombia), some members of the Senate ­were concerned about increasing American involvement in what they characterized as an internal civil war.110 To ensure that the president did not expend more resources on Plan Colombia than expressly authorized, the mea­sure allocating funds for Plan Colombia for 2001 included a provision that prohibited the executive branch from expending any additional resources in fiscal year 2001 ­unless Congress explic­itly approved as such in a joint resolution; that mea­sure, moreover, would be exempt from a filibuster on the floor of the Senate. Ultimately, neither President Clinton nor President Bush sought t­hese supplemental funds through the pro­cess proscribed. One possibility is that the funds appropriated in the original bill ­were sufficient to achieve the country’s strategic goals for the year. Alternatively, however, the Senate could have used the new special procedures to prevent the president from accessing additional funds. In other words, as with presidential vetoes,111 that the Senate did not use this par­tic­u­lar procedural exception does not mean that the procedure was irrelevant for U.S. policy t­oward Colombia in 2000 and 2001. In addition, b­ ecause the exception covered a clearly defined period of time, the members of Congress who created it in the summer of 2000 knew that if they w ­ ere ­going to get an opportunity to claim credit for revisiting policy ­toward Colombia, that opportunity would come ­after the 2000 election and before the 2002 midterms. ­These examples illustrate a set of three general, hierarchical princi­ples that guided my collection of examples of when a procedural exception became relevant following its creation:

34

EXCEPTIONS TO THE RULE

• Did the exception contain a specific date by which the proposal covered by the special procedures was to be introduced? • If no specific date was included, when was the first resolution ­under the auspices of the procedural exception introduced? • If no formal resolution was introduced, when did the first policy change to which an exception could have been applied occur? Using a combination of congressional documents, such as the Congressional Rec­ord; other federal government resources, such as reports from the General Accounting (now Government Accountability) Office; academic articles; and other secondary source publications, I was able to identify examples of when successfully created exceptions became relevant to decisions in the policy area to which they applied. (A complete list of ­these situations, and the sources from which I identified them, appears in auxiliary appendix ­table AA2-1.)112 Of the 161 successfully created exceptions, I was able to identify examples of relevance using the criteria outlined above for 140 of the procedures. In general, the remaining 21 involve some sort of exogenous trigger, the occurrence of which is nearly unpredictable; they include, for example, provisions requiring Congress to approve compensation plans for victims of nuclear and commercial space accidents.113 For the median procedure that has become pertinent to policymaking, meanwhile, only a year passed between creation and relevance; the distribution of this elapsed time appears in figure 2.1. As the figure indicates, the majority of procedural exceptions become relevant for policymaking quite soon a­ fter they are created; indeed, 76 ­percent are ­either originally slated to be used, are actually used, or have the potential to be used within two years of their enactment. Although these data are descriptive (and certainly prone to some mea­sure­ment error) rather than causal, at the very least they provide a useful foundation for the argument about the role of majority party electoral considerations that unfolds in the ensuing chapters.

THE ISSUE-­S PECIFIC NATURE OF MAJORITARIAN EXCEPTIONS

As can be seen from this contextual discussion, a key way in which majoritarian exceptions differ from many other congressional procedures is that they are issue specific. Any investigation of their creation, then, must engage with



Limiting the Unlimited FIGU RE 2-1 .

35

Years a ­ fter enactment that majoritarian exceptions become relevant, 1969–2014

ENACTED EXCEPTIONS

40

30

20

10

0

5

10

15

20

YEARS

questions of how the majority party’s success at adopting special procedures is related to the under­lying politics of the policy area in which f­ uture legislation ­will be protected from obstruction. In this way, my analy­sis ­here joins a growing lit­er­a­ture on the role of issues in elite po­liti­cal be­hav­ior in the United States, in which other scholars have highlighted the way issue dynamics have impor­tant consequences for the policymaking pro­cess. This relationship begins at the campaign stage, as legislators are more likely to focus on the issues they discussed in their campaigns once they assume office.114 It continues at the agenda construction stage, where party leaders choose to focus on issues that most effectively balance members’ individual electoral concerns, the desire to build a rec­ord of legislative accomplishment, and the need to draw clear distinctions between them and their partisan opponents.115 Other work on the congressional agenda also highlights the degree to which legislators tend to focus on ­those issues that must be addressed to build a reputation as a collectively effective institution.116 Issues can also play an impor­tant role in interbranch dynamics, where policy areas on which the president focuses become more partisan simply ­because of executive attention.117

36

EXCEPTIONS TO THE RULE

­ ere, by arguing that the majority party chooses to make some issues and not H ­others the focus of procedural change that expedites ­future legislative consideration, I demonstrate that issue dynamics also affect questions of procedural choice. Central to this research agenda is the Policy Agendas Proj­ect, ­housed (as of this writing) at the University of Texas at Austin, which carefully categorizes a wide range of inputs to and outputs from the American policymaking pro­cess into a set of consistent policy topics. At vari­ous points in this book, I use three data series constructed using Policy Agendas Proj­ect issue codes: congressional bills,118 survey response to Gallup’s Most Impor­tant Prob­lem question, and the public’s mood in specific policy areas.119 Although t­ hese coding efforts are an invaluable piece of our investigation, two alterations to that categorization scheme are necessary for our purposes ­here. First, the concept of issue owner­ship, the notion that voters have durable ­ andling of issues, is central to our argument about perceptions about parties’ h when it is advantageous to the majority party to create a majoritarian exception ­because the electoral advantages the party accrues are not the same across all policy areas. In some situations, the majority party ­will want to expedite action on par­tic­u­lar issues ­because voters associate the policy area with the party, while in ­others, using special procedures obscures congressional action and thus makes it harder for voters to notice that a party is acting on an issue which they trust the party to ­handle. ­Because the Policy Agendas Proj­ect does not classify its policy areas into issues owned by Demo­crats and Republicans, I turn to an alternate source, work by Patrick Egan.120 The principal difference between Egan’s classification of issues and the one developed by the Policy Agendas Proj­ect is that the former disaggregates macroeconomic policy into several component parts, including deficit-­debt, taxes, inflation, and Social Security. Given the centrality of ­these issues to the politics of majoritarian ­ ere, following Curry, for a mapping of Polexceptions, I retain Egan’s coding h icy Agendas Proj­ect topics to issues owned by the two parties (or neither).121 Second, b­ ecause the provisions creating special procedures are often contained in broader pieces of legislation, I examined each individual provision to ensure that the coding of the entire bill reflected the content of the specific procedural change. For roughly 80 ­percent of the observations, I retained the original Policy Agendas Proj­ect code (subject to the modifications discussed in the previous paragraph). Approximately 5 ­percent of the observations ­were contained in legislation not coded by Adler and Wilkerson.122 For the remain-



Limiting the Unlimited

37

ing 15 ­percent (148 observations), the code assigned to the overall legislation does not reflect the issue with which the procedural provision deals; ­these changes are summarized in auxiliary appendix t­ able AA2-3. The search strategy and issue coding outlined ­here identifies 972 majoritarian exceptions proposed between the 91st  and 113th (1969–2014) Congresses in twenty-­five policy areas.123 Some policy areas have seen more proposals to create special procedures than o­ thers. Roughly half of the offered changes fall into three topical categories—­government operations, trade, and defense issues; on the latter two especially, majoritarian exceptions have played an impor­tant role. ­These policy areas are prominent for both types of majoritarian exceptions, making up approximately 47 ­percent of del­e­ga­tion exceptions and roughly 58 ­percent of executive branch oversight exceptions. Given the prevalence of debate limitations in each of t­ hese policy areas, we might expect that the special procedures ­will have par­tic­u­lar consequences for decisionmaking on ­these issues. Indeed, the case study of trade policymaking at the conclusion to chapter 3 suggests this is the case. At the same time, at least one exception has been proposed in ­every policy area discussed above. Proposals to create expedited procedures are offered somewhat more frequently when Demo­crats control the Senate (60 ­percent of observations) than Republicans (40 ­percent). The difference in success rates, however, is similar: 19  ­percent of procedural changes offered ­under Demo­ cratic control ­were ­adopted, as compared with 13 ­percent of ­those proposed when Republicans held the majority. Once ­t hese majoritarian exceptions are in place, they can have consequences for the policies Congress produces. What determines when they are ­adopted? We begin to answer that question in chapter 3.

Three OBSCURING THE CAUSAL CHAIN MAJORITARIAN EXCEPTIONS AS A BLAME AVOIDANCE TOOL

T

wenty-­five years ­after its creation in 1970, Amtrak found itself in dire financial straits. Its operating deficit had climbed to nearly $200 million by 1994, thanks in part to Congress’s cutting capital funding while si­mul­ta­ neously insisting that Amtrak continue to run unprofitable routes popu­lar with their constituents.1 Even when Amtrak attempted to cut costs, congressional meddling mitigated the effects. In October  1996, for example, the railroad announced it was eliminating five expensive routes, but Congress insisted that operations continue for another five months. The federal subsidy to cover this extension, moreover, fell $13 million short of the cost of actually ­doing so.2 The situation came to a head in late 1997, when Amtrak announced that without access to additional funds, it would need to seek an extension from creditors to pay personnel and continue operations.3 Congress managed to pass a mea­sure providing an influx of capital funds, and President Clinton signed it into law in early December. Building a co­ali­tion in support of Amtrak, however, required significant promises of reform. ­Because legislators’ desire to keep underused routes ­running in their states and districts was a large driver of Amtrak’s financial situation, members strug­gled to devise a mechanism to guarantee that changes would be made. One proposal, to use a commission to recommend specific train routes to be eliminated or restructured, had 39

40

EXCEPTIONS TO THE RULE

been rejected during consideration of the fiscal year 1998 transportation ­appropriations bill in July.4 Ultimately, a panel—­known as the Amtrak Reform Council—­with the power to consider a broader set of reforms won the day as part of the Amtrak reform mea­sure. The council would be responsible for “evaluat[ing] Amtrak’s per­for­mance” and “mak[ing] recommendation to Amtrak for achieving further cost containment and productivity improvements, and financial reforms.”5 ­After two years, if the council determined that the railroad was ­either falling short of financial per­for­mance goals or was on track to need federal operating subsidies ­after fiscal year 2002, it would “develop and submit . . . ​an action plan for a restructured and rationalized national intercity rail system,” to which Amtrak would respond by “develop[ing] and submit[ting] . . . ​an action plan for the complete liquidation of Amtrak.” 6 Congress could prevent the liquidation by adopting a disapproval resolution, on which debate in the Senate was limited to twenty hours.7 Why did Congress willingly give up some of its power to an outside actor? Had Congress attempted Amtrak reform on its own, electorally minded legislators from the areas hurt by reforms would most likely have attempted to undo any bargain reached to deliver the broad-­based benefits of a more financially sound passenger rail system. By giving responsibility for the hard work of reforming Amtrak to actors outside the chamber and easing the path of the associated legislative components, Congress managed to (potentially) deal with a prob­lem for which the regular legislative pro­cess was ill equipped.8 This attempt at reforming Amtrak is an example of one of the two major categories of majoritarian exceptions. I refer to this group of procedures as “del­e­ga­tion exceptions,” and they represent instances in which the chamber uses procedural change to resolve collective action prob­lems with which the regular legislative pro­cess cannot deal effectively. Most existing work that addresses when Congress creates this kind of procedural change tends to focus on one specific policy area, such as military base closures or trade.9 While deeply descriptive, that research has not offered or tested a crosspolicy area account of when ­these special procedures are created. In this chapter, I provide such an investigation. I begin by outlining Arnold’s concept of “traceability,” which describes the conditions u­ nder which legislators would be concerned about the electoral consequences of their choices in office.10 Next, I examine the policy prob­lems addressed by del­e­ga­ tion exceptions, arguing that they fit Arnold’s criteria for issues on which poli-



Obscuring the Causal Chain

41

cymakers ­will want to minimize traceability. I then describe how, if del­e­ga­tion exceptions are a tool enabling the majority party to achieve its policy and po­liti­cal ends, t­here are two main issue-­specific ­factors—­salience and issue owner­ship—­that should predict which proposals are ­adopted and which proposals are not. Building on this set of expectations, I use the data described in the previous chapter on proposed and enacted majoritarian exceptions to test them. I also explore ­whether the effects of salience and issue owner­ship vary with the electoral circumstances facing the majority party and ­whether the two parties behave similarly in the context of del­e­ga­tion exception proposals. I discuss ­whether ­t here may also be strategic be­hav­ior when del­e ­ga­t ion exceptions are proposed before concluding with a case study involving trade policy from about 1995 to 2015, which illustrates an alternative way to conceptualize the salience of a par­tic­u­lar policy area. Chapter 2 outlined my operational definition of a majoritarian exception: a provision, enacted as part of statutory law, that designates some ­future piece of legislation as exempt from a filibuster. In addition, majoritarian exceptions also may include provisions related to committee consideration, to allowable amendments, and to other aspects of floor consideration. Del­e­ga­tion exceptions are distinguished by one additional, impor­tant feature: they reallocate agenda-­setting power by designating a specific actor or actors to develop a proposal that achieves policy change. That proposal is followed by legislation, ­either approving or disapproving it, to which procedural protections apply. This special agenda setter can take a variety of forms. Within the institution, it may be a new committee, such as the Senate Bud­get Committee, which was created and charged with developing the yearly bud­get resolution by the Congressional Bud­get Act of 1974. Existing committees may also be empowered by del­e­ga­ tion exceptions, as is the case with the bud­get reconciliation pro­cess, by which individual standing committees are given the authority to develop proposals to amend mandatory spending programs.11 Outside of Congress, the new agenda setter may be in the executive branch. As part of the 2010 Dodd-­Frank Wall Street Reform and Consumer Protection Act, for example, the secretary of the trea­sury, in conjunction with the president, is responsible for determining the maximum amount of debt that the Federal Deposit Insurance Corporation is permitted to guarantee u­ nder an emergency financial stabilization program during a ­f uture economic crisis.12 Presidents may, however, be required to consult with Congress

42

EXCEPTIONS TO THE RULE

during the development of the proposal, as is the case with fast-­track trade authority. Fi­nally, the specific identity of the actor may be jointly determined by the president and Congress. The In­de­pen­dent Payment Advisory Board, created by the 2010 Patient Protection and Affordable Care Act and empowered to make changes to Medicare, has fifteen members nominated by the president and confirmed by the Senate. In par­tic­u­lar, the president is required to consult with the Senate majority leader, the Speaker of the House, and the Senate and House minority leaders on the se­lection of three members.13 Regardless of the identity of the special agenda setter, however, their power is largely the same across all del­e­ga­tion exceptions: they develop a proposal to make a specified kind of policy change and then watch a mea­sure that ­either approves or disapproves that proposal come to the floor of the Senate u­ nder procedural protections from a filibuster and, often, amendments.14

WHEN DO VOTERS RESPOND TO LEGISLATORS’ ACTIONS?

To understand when the Senate’s majority party adopts special procedures to help it pursue continued majority status, it is helpful to outline how voters are likely to respond to policy changes initiated by Congress, regardless of the procedures used to take that action. According to Arnold, for a given policy choice to affect an electoral outcome, voters must be able to “plausibly trace an observed effect first back to a governmental action and then back to a representative’s individual contribution.”15 Arnold argues that for something to be traceable, three conditions must hold: voters must perceive the effect of a policy; they must attribute ­those results to an identifiable action; and their legislator must have made a vis­i­ble contribution to that decision. If voters recognize positive effects of a specific governmental action to which their representative clearly contributed, they can reward the incumbent with reelection. If the traceable consequences are negative, however, the incumbent is apt to be punished. For a given policy change, the degree to which voters are capable of observing a decision depends on the costs and benefits associated with it. Citizens, argues Arnold, are more likely to detect the consequences of a choice if it conveys costs or benefits that are specific—­that is, borne by a par­tic­u­lar group of which they are a member or a certain geographic region in which they live—­ than if the effects are general, or felt uniformly by all members of society. In



Obscuring the Causal Chain

43

addition, voters w ­ ill be likely to notice larger costs and benefits more than smaller ones and ­will recognize a cost more readily than a benefit of equal size. ­Under this typology of costs and benefits, then, voters are most likely to perceive policy changes that convey large, specific costs, even if they are accompanied by diffuse, general benefits. To reduce the chances that incumbent legislators are punished for a decision that imposes ­these kinds of specific costs, representatives may attempt to alter the second two links in the traceability chain. Del­e­ga­tion exceptions can achieve this goal, principally by weakening the identifiability and visibility links. By separating the choice to vest a special agenda setter with the power to develop a proposal from the ultimate legislative action that ratifies that plan, del­e­ga­tion exceptions make it harder for voters to identify exactly which choice by policymakers initiated new policy. This is particularly true for del­e­ga­tion exceptions that cover disapproval resolutions; in t­ hose situations, Congress is able to completely isolate itself from the details of the new policy simply by not even scheduling a vote that would prevent the new policy from ­going into effect. In addition, by giving a special agenda setter the ability to develop a proposal that changes current policy, the del­e­ga­tion exception reduces the visibility of a legislator’s contributions to the development of the proposal, ­either as the original sponsor, as a cosponsor, or as a member of the committee that worked on the mea­sure. This is especially true if, as is often the case, the agenda setter is outside the chamber. Second, by protecting the proposal from amendments and filibusters on the floor of the Senate, del­e­ga­tion exceptions eliminate the expectation that individual members ­will exploit their vari­ous procedural rights to the benefit of their constituents—­a presumption that has only grown in the Senate over the course of the twentieth ­century16—­because their ability to do so has been severely curtailed. If legislators are limited in their ability to change or obstruct the proposal, they cannot make vis­i­ble contributions to its passage or its defeat. At their most extreme, del­e­ga­tion exceptions that involve a disapproval resolution reduce almost completely the possibility that constituents detect individual contributions by their representatives. The special procedures for closing military bases, for example, dictate that the proposal developed by the Base Realignment and Closure (BRAC) Commission ­will be implemented by the Department of Defense ­unless Congress enacts a joint resolution vetoing the recommendations in total.17 Specific decisions about which bases to close, then, can be blamed on the outside agenda setter and not congressional incumbents.

44

EXCEPTIONS TO THE RULE

Of course, punishment by voters is only one side of the traceability story. Citizens may also reward incumbent legislators for actions, but again, only if the benefits of their choices are perceptible, identifiable, and vis­i­ble. It is difficult, argues Arnold, to convince policymakers to devote attention and resources to delivering diffuse, general benefits precisely ­because ­those choices are less traceable. Consider, for example, deficit reduction. The economic consequences of large bud­get deficits are complicated, and as a result, any positive effects of ameliorating them by cutting the deficit w ­ ill be hard for an individual voter to understand—­assuming the voter is even able to connect a specific governmental decision to ­those consequences. If legislators do attempt to deliver diffuse benefits, then, they may need to be especially conscious of minimizing traceability. Policymakers rarely get credit for widely distributed benefits, so any perceptible costs ­will quite likely outweigh ­those benefits in voters’ judgments. Given this logic, if del­e­ga­tion exceptions minimize traceability, then one would expect them to be offered—­and, in some cases, a­ dopted—­when solving a policy prob­lem requires imposing specific costs in exchange for general benefits. ­These are exactly the kinds of issues on which expedited procedures are proposed.

DEL­E ­G A­T ION EXCEPTIONS AND THE COSTS AND BENEFITS OF POLICY CHANGE

Between the 91st and 113th Congresses (1969–2014), I identified a total of 390 del­e­ga­tion exceptions proposed in the House and/or Senate. As figure 3.1 illustrates, almost 70 ­percent of ­these fall into just five of the policy areas outlined in chapter 1: government operations (118), trade (51), deficit-­debt (42), health (32), and energy (27). In some of t­hese policy areas, the existence of specific costs and general benefits is well established. In the case of deficit reduction, for example, a conflict arises between the macroeco­nom­ically beneficial goal of balancing the bud­get and shrinking the national debt and the steps necessary to achieve that end—­that is, cutting or eliminating individual federal programs. Although programs ­will vary in the po­liti­cal engagement of their beneficiaries and the quality of their repre­sen­ta­tion by or­ga­nized interests, Congress has repeatedly strug­gled to enact meaningful deficit reduction as individual legislators seek to avoid reductions in programs that benefit their constituents.18 The collective action prob­lems inherent in trade policy are similarly

46

EXCEPTIONS TO THE RULE

similarly hampered by a tension between collective benefits and individualized costs. Efforts to improve and reor­ga­nize the pro­cess by which the federal government implements its programs date to the early twentieth ­century, and by the late 1970s ­these efforts became increasingly imbued with a distrust of government.20 The answer, reor­ga­ni­za­tion proponents argued, was to eliminate what ­were considered unnecessary government functions, and, beginning in the early 1990s, del­e ­ga­t ion exceptions represented a part of this overall strategy. With suggested names such as the Commission to Eliminate Waste, Fraud, and Abuse, the Federal Government Streamlining Commission, and the Commission on the Accountability and Review of Federal Agencies, the proposals, none of which w ­ ere ultimately successful, varied in their specifics but shared a general framework. Each would delegate power to an in­de­pen­ dent commission that would identify agencies or programs that could be eliminated, based on criteria such as w ­ hether an initiative was “ ‘duplicative . . . ​ wasteful . . . ​inefficient . . . ​outdated . . . ​irrelevant . . . ​or failed.’ ”21 As is the case with deficit reduction, using a del­e­ga­tion exception would allow the proponents of the overall goal of reducing the size of the federal government to achieve the broad benefits associated with smaller, more efficient operations while also preventing individual legislators whose constituents relied on the “wasteful” and “inefficient” programs from derailing the reform effort. A close examination of ­these proposals on energy policy reveals that, while the overall policy area might not be associated with the presence of collective action prob­lems, del­e­ga­tion exceptions have been proposed to address specific aspects of the issue where any negative consequences of policy change ­were likely to be felt keenly by individual constituents of specific senators. Energy-­ related proposals fall into two general categories. The first, prevalent in the 1970s, involves vari­ous responses to the energy crisis. Price controls in the oil market had helped create shortages, but legislators ­were reluctant to undertake deregulation, which would create immediate costs in the form of higher prices in exchange for the diffuse ­future benefit of reduced demand. The ­public was also broadly supportive of gasoline rationing as a mechanism for allocating scarce fuel during shortages, but legislators strug­g led to design a policy that actually limited specific consumers’ access to gasoline.22 In both cases, Congress eventually turned to del­e ­ga­t ion exceptions, with the president as the special agenda setter, to resolve its collective action prob­lems. The second set of energy-­related proposals occurred between 1993 and 2001, when eight separate bills w ­ ere introduced that would create a special



Obscuring the Causal Chain

47

commission to make recommendations on restructuring the Department of Energy’s (DOE) National Laboratories; the proposal would be subject to congressional approval, but it could not be amended and could not be filibustered in the Senate.23 Originally created before World War II as part of the large ­ ere the sites of impor­tant nuclear federal investment in basic science, the labs w weapons research during the Cold War but subsequently expanded to also cover work in all of the DOE’s mission areas (science and technology, national security, energy resources, and environmental quality). Beginning in the early 1990s, the DOE generally, and the National Labs specifically, came ­under increasing criticism for mismanagement and misplaced focus and for performing functions that could be better handled by other federal research and development programs or in the private sector.24 ­There ­were twenty-­two facilities in the National Labs system in 1993, and any regular legislative proposal to close them would have met with strong opposition from members in places where they ­were located, including California, Illinois, New Mexico, and New York. In 1996, for example, the member representing California’s Tenth District, home to the Lawrence Livermore National Laboratory, took to the floor of the chamber to defend the fa­cil­i­ty. “­These workers,” argued Representative Baker (R-­Calif.), “are truly national assets . . . ​[and] while bud­getary bottom lines may sometimes seem cold, a responsible government treats its workers as national assets to be valued and esteemed.”25 None of the eight proposals to create a del­e­ga­tion exception for restructuring the National Labs ­were ultimately enacted, and an attempt in 2003 to consolidate the operations of the Idaho National Laboratory and the Oak Ridge National Laboratory was unsuccessful, thanks to ongoing lobbying by Idaho’s all-­Republican congressional del­e­ga­tion, Representative Simpson and Senators Craig and Crapo.26 The final major category, health policy, illustrates an impor­tant dynamic of what we mean by “specific costs.” Up to now, our examples of specific costs—­higher gas prices, reductions in benefits, the elimination of jobs in a par­tic­u­lar state—­have been tangible and relatively easy to describe. In the case of health policy, however, even if legislators cannot quantify up front precisely what the par­tic­u­lar costs of a policy change ­will be, a more abstract description of them may be enough to convince voters that costs are coming—­and thus sufficient to scare policymakers away from attempting to deliver the general benefits. ­There ­were several prominent health reform plans in the early 1990s: consider, for example, the proposed Federal Health Board, modeled

48

EXCEPTIONS TO THE RULE

roughly a­ fter (depending on the description) the Securities and Exchange Commission27 or the Federal Reserve Board.28 The health board would be responsible for determining what ser­vices would be covered in a “standard benefit plan,”29 ­after which its decisions would come to the Senate for pos­si­ble disapproval ­under expedited procedures.30 By the fall of 1993, this notion had been transformed in the public discourse to the fear that citizens would, to quote the famous “Harry and Louise” ads, be “ ‘force[d] to pick from a few health care plans designed by government bureaucrats.’ ”31 While the evidence that the “Harry and Louise” ads affected health reform’s success may be overstated,32 their central conceit illustrates well that voters may see a dismantling of existing benefits and programs as specific costs, even if they w ­ ill ultimately be better off with the general benefits produced by the new policy. The composition of the “other” category in figure 3-1 also helps illustrate how del­e­ga­tion exceptions involve policy issues in which concentrated costs are imposed in exchange for general benefits. The ten issues that are the least frequent targets of del­e­ga­tion exceptions over the period 1969–2014 are generally not characterized by ­these kinds of policy prob­lems; ­these policy areas, with number of proposals, are as follows: • law and crime: 5 • jobs and unemployment: 5 • education: 4 • public lands: 4 • general macroeconomics: 2 • space, science, and technology: 2 • civil rights: 1 • environment: 1 • poverty: 1 • inflation: 1 For some of ­these issues, such as crime, education, and poverty, solving the under­lying policy prob­lem generates specific benefits for ­those citizens affected by the policy change and also produces gains for society as a ­whole. The costs,



Obscuring the Causal Chain

49

meanwhile, would most likely come in the form of relatively small (for each individual voter) tax increases. In other policy areas, such as space, science, and technology and civil rights, it is difficult to assign concrete costs and benefits to the prob­lems that need to be solved, suggesting that traceability concerns may not be as relevant for policymakers. Del­e­ga­tion exceptions, then, are clearly targeted at policy areas, or specific components thereof, where a tension between broad-­based benefits and narrowly imposed costs makes the regular legislative pro­cess ill equipped for enacting changes to the status quo. This descriptive evidence tells only part of the story, however. Collective action prob­lems appear to affect the generation of proposed del­e­ga­tion exceptions, but once a procedural change has been proposed, the tension between costs and benefits does not necessarily explain ­whether it is successfully enacted into law. To make predictions about ­whether a given procedural change is likely to be approved once proposed, we need to explore other features of the under­lying policy change that it would shepherd to passage in the ­future.

THE ROLE OF ISSUE SALIENCE AND OWNER­S HIP

As discussed in chapter 2, the focused nature of majoritarian exceptions is a key feature that separates them from other kinds of procedural innovations in Congress, and certainly, the degree to which a par­tic­u­lar issue area involves imposing specific costs in exchange for general benefits is an impor­tant component of the issue-­specific politics at hand. The component of Arnold’s account that emphasizes the differences in dynamics across policy areas,33 however, is largely time invariant. If a policy prob­lem can be characterized as imposing particularized costs in exchange for general benefits, then that feature is likely to be a permanent and unchanging characteristic of the issue. What may change over time, however, is the salience of the subject among citizens. As the importance that voters place on an issue increases, the attractiveness of a del­e­ ga­tion exception is affected. When voters evaluate their representatives, not all issues are created equal. As Kingdon argues, it is members’ be­hav­ior on the most salient issues that constituents are most likely to use as the basis for reward or punishment.34 Similarly, Fiorina describes how issue salience may be especially impor­tant when some or all of a legislator’s constituents are likely to oppose the legislator’s action on an issue.35 Arnold, moreover, applies this logic in the specific

50

EXCEPTIONS TO THE RULE

context of minimizing traceability.36 When legislators are making decisions about ­whether to obscure their actions, he argues, one relevant consideration is ­whether they believe that ­future opponents are likely to try and highlight the damaging consequences of that choice in the next election cycle. For the kinds of issues covered by del­e­ga­tion exceptions—­where the costs of policy change are noticeable and the benefits diffuse—­any perceptible effects are, on average, negative. Government programs that are considered “wasteful,” for example, have some beneficiaries who w ­ ill feel the consequences of their elimination. Along the same lines, when military bases are shuttered, the communities in which they are located may experience less overall economic activity, spillover job losses, or decreased government revenue.37 The existence of potentially harmful effects, then, makes the policy changes covered by them attractive targets for emphasis by f­ uture challengers. When t­ hese negative consequences occur in a policy area that is particularly salient to voters, opponents should find them even more appealing as a campaign issue. As a result, legislators may find it more impor­tant to minimize traceability on salient issues than on nonsalient ones. When an issue is salient, moreover, the majority party may find it easier to attract the support of a small number of minority members to support the del­ e­ga­tion exception proposal. In a sixty-­vote Senate—­a characterization that applies to much of the period ­under study ­here—­the majority party ­will almost certainly require the aid of a few minority party legislators to adopt legislation establishing expedited procedures. While other features of issues (including owner­ship, as discussed below) ­will make it easier or harder to attract minority party support, the same dynamics that drive majority party members to minimize traceability—­a concern that voters ­will punish them for making policy changes that impose specific costs—­may also lead a small number of minority party legislators to join the effort. The notion that members should fear electoral punishment for their be­ hav­ior on issues impor­tant to voters is widely supported by empirical evidence. Voters, for example, punish incumbent representatives for unpop­u­lar voting be­hav­ior on high-­profile issues, such as gay marriage, taxes, abortion, and immigration.38 Related work suggests that this penalty operates through a signaling mechanism. When a Demo­cratic legislator casts a roll call vote in support of a liberal position on a salient issue (for example, health care in 2009 and 2010), Republican and in­de­pen­dent voters are less likely to vote for the legislator ­because they interpret that choice as a signal that the member is



Obscuring the Causal Chain

51

ideologically extreme. On the same issue, moreover, t­ here is no evidence that copartisans reward their incumbent for a liberal stance. That punishment seems to outweigh reward on salient issues is consistent with our argument that minimizing traceability to avoid electoral penalty w ­ ill be more likely on priority 39 issues. Moving beyond a conception of salience as relevant only on “impor­tant” issues yields additional evidence that when voters find an issue to be relevant, they are more responsive to an incumbent’s actions on it. Voters in congressional districts40 and states41 with higher numbers of Iraq War deaths, for example, punished Republican House and Senate incumbents, respectively, for t­hose outcomes. As Berry and Howell document in the context of local test scores and school board elections, media attention to an issue makes it more salient to voters, who then appear to be more responsive to that information.42 This suggests that legislators have reason to be concerned about the possibility that instigators are drawing attention to negative consequences of their policy choices. Given this worry, del­e­ga­tion exceptions ­will be attractive when minimizing traceability is impor­tant. Together, ­these theoretical accounts and empirical evidence suggest the following hypothesis: Salience hypothesis: A proposed del­e­ga­tion exception is more likely to be enacted into law if it deals with a salient issue. Variations in issue salience help us understand when individual, reelection-­ minded legislators ­will prefer to see proposed del­e­ga­tion exceptions enacted into law. However, the shared interests of congressional parties ­will most likely affect the success of ­these prospective changes to procedures. Both parties are attempting to gain or maintain majority status—­the benefits of which are well documented.43 To achieve this goal, each must balance two considerations: building a strong, positive party brand44 and ensuring the individual reelection of its incumbent members. Just as issues vary in the importance the public places on them, they also vary in the degree to which they are associated with the party brand-­building exercise. In par­tic­u­lar, policy areas are often characterized by ­whether they are “owned” by one party or the other. Owner­ship is generally thought of as a long-­term, positive association between voters and a given party on an issue and is generally mea­sured by assessing citizens’ beliefs on which party w ­ ill do 45 a better job ­handling that policy area. U ­ nder this definition, Demo­crats, for

52

EXCEPTIONS TO THE RULE

example, own issues like health care and the environment, while Republicans own defense policy and immigration. Certain issues, such as transportation, remain per­sis­tently orphaned (that is, not owned by ­either party). ­Others move between owned and unowned status over time; using Egan’s owner­ship mea­sure,46 for example, neither party owned trade as an issue ­until the beginning of this ­century, when it became viewed as a Republican competency. How might an issue’s owner­ship status be associated with the prospects of a proposed del­e­ga­tion exception concerning that policy area? Some existing work posits that when a party controls Congress, it should take advantage of that opportunity to pay more attention to the issues it owns. Egan, for example, documents how parties engage in more federal spending and pass more landmark legislation in owned-­issue areas.47 At the same time, other work on issue attention and legislative productivity, such as Adler and Wilkerson, finds ­little evidence of a relationship between owner­ship and legislative output.48 In the case of del­e­ga­tion exceptions, however, the relationship between issue owner­ship and partisan incentives might be somewhat more complicated. First, the legislation creating a del­e­ga­tion exception can conceivably be filibustered, so even if the majority party wants to see a procedural change ­adopted, it must frequently seek the cooperation of some minority party members to do so. Issues owned by the majority party should be the ones on which obtaining this support is most difficult. For minority party members, ­there are few electoral gains to be had from acting in policy areas owned by the majority party, as any collective benefits from policy change are likely to be felt by the majority, rather than the minority. In addition, to the extent that del­e­ga­tion exceptions reduce traceability, the minority party may be loath to help the majority party obscure potentially unpop­u­lar policy actions in ­those policy areas that ­matter most to the majority party’s core voters. Minority party Demo­ crats, for example, might be wary of helping majority party Republicans adopt a del­e­ga­tion exception that closes military bases, since Republican voters would be more likely to punish their legislators for that policy choice. Second, work by Harbridge suggests that majority party leaders may approach action on owned issues differently when engagement with the policy area creates cross-­pressures between individual members’ reelection goals and their parties’ desire to gain or retain majority status.49 For Harbridge, cross-­ pressuring is related to electoral circumstances, but it might also be related to policy prob­lems and solutions. The combination of general benefits—­the delivery of which helps party goals—­a nd specific costs—­the imposition of



Obscuring the Causal Chain

53

which hurts individual members—­indicates that the policy changes achieved through del­e­ga­tion exceptions, by definition, create cross-­pressures. Certainly, for a given policy prob­lem, not all legislators ­w ill experience ­t hese cross-­ pressures equally. A pro–­f ree trade senator from a state with a large number of manufacturing jobs—­say, a Republican from the Rust ­Belt—­certainly feels ­these tensions more acutely than a colleague from a state with a dif­fer­ent economic base, such as one in the Mountain West. We need not assume that all legislators are cross-­pressured equally, however, to believe that party leaders are sensitive to ­these concerns. When leaders are constrained by tensions between the incentives of individual members and the collective party brand, argues Harbridge, ­there are fewer partisan gains to be realized through action on owned issues. Some members may find themselves on opposite sides of the owned issue, and “­there is no partisan advantage to be gained if the issue splits the majority or puts individual members at electoral risk.”50 Together, t­ hese dynamics suggest the following: Issue owner­ship hypothesis: A proposed del­e­ga­tion exception is less likely to be enacted into law if it deals with an issue owned by the majority party.

DATA AND ESTIMATION

To evaluate ­these pos­si­ble explanations, I return to the data on the universe of proposed majoritarian exceptions described in chapter 2. ­Here, I use only the data on proposed procedural changes that would delegate authority to a special agenda setter and grant that proposal privileged procedural rights. First, I code each observation for ­whether it was enacted into law. Between 1969 and 2012 (the 91st to 112th Congresses), 371 such proposals ­were introduced, 39 of which ­were successfully enacted into law.51 To determine which proposals ­were enacted into law, I use Adler and Wilkerson’s Congressional Bills Proj­ ect data set.52 It is worth noting that this success rate of roughly 11 ­percent indicates that a proposed del­e­ga­tion exception is less likely than an oversight exception to be enacted into law; the latter category of exceptions, explored at length in chapter 6, has a success rate of roughly 21 ­percent. Senators, in other words, are quite unwilling to give up some of their power over developing, amending, and ultimately approving a proposal even when solving the

54

EXCEPTIONS TO THE RULE

under­lying policy prob­lem involves confronting distributional issues for which the regular procedures are likely to prove insufficient. Since I have argued that legislators’ motivation for adopting special procedures is associated with the anticipated electoral consequences of acting in dif­fer­ent issue areas, testing my claims requires mea­sure­ment at the level of the issue with which the exception deals. To test the salience hypothesis, I rely on two separate mea­sures. The first involves data from the Gallup Poll showing the proportion of respondents who reported a given policy area as the most impor­tant prob­lem facing the nation. Using ­every Gallup survey made available by the Roper Center for Public Opinion Research at Cornell University, I coded each survey response into one of the policy categories described in chapter 1. (A full list of which response items fall into each policy area appears in auxiliary appendix ­table AA3-1.) Once each response was matched to an issue area, I followed the Policy Agendas Proj­ect’s methodology for aggregating responses by year; this involves taking the sum of all reported response percentages in a given topic in an individual poll, dividing each individual percentage by a weight based on the summed total of all percentages for that poll, and then averaging all weighted percentages across each Congress.53 ­Here, I follow Jones, Larsen-­Price, and Wilkerson and assume that the relationship between salience as mea­sured using Most impor­tant prob­lem and policymaking activity is contemporaneous.54 While the mea­sure of most impor­tant prob­lem is prob­ably the most common tool scholars use to capture issue salience, it has its shortcomings. Wlezien, for example, argues that using responses to the question of most impor­tant prob­lem to mea­sure salience conflates two dif­fer­ent dimensions of issues: the importance voters place on them and the degree to which respondents perceive them to be a prob­lem.55 In addition, in our theoretical account above, salience ­matters ­because we expect voters to notice action more readily on issues that are most relevant to them. While individuals’ rating of a given issue as most impor­tant should certainly be related to ­whether they ­will perceive policy change, ­there are other ways to mea­sure this as well. As a result, I use a second mea­sure to tap the importance of an issue to voters: the public’s mood, its “general dispositions” ­toward “policies and publics,” as described by Stimson.56 A range of scholars have used mood as a mea­sure of “public preferences for more or less government action,”57 arguing that when voters are demanding more government activity, Congress should respond by being more involved.58 Most previous work examining the relationship between



Obscuring the Causal Chain

55

public mood and legislative outputs has relied on the concept of a global mood, which is meant to capture the public’s overall demand for a more-­or less-­activist government. Voters’ preferences, however, may not be the same in one policy area as in ­others, so that mea­sur­ing them as such may obscure impor­tant variation.59 In the context of mood specifically, Atkinson and ­others document how, when the global mood mea­sure is disaggregated into a set of policy-­specific moods, ­there is some variance in how closely individual moods are related to the overall mood; they also find that the correlations between mood and government activity vary across issues.60 Since much of our story h ­ ere is concerned with the politics of individual issues—­especially minimizing the traceability of policy changes within them—­a mea­sure of the public’s demand for more or less government in a par­tic­u­lar area is especially attractive. Luckily, the increasing availability of high-­quality public opinion questions of the kind that inform Stimson’s mood calculations means that it is pos­si­ble to mea­sure public mood in a large number of specific policy areas. Using the custom series application made available by the Policy Agendas Proj­ect,61 we are able to calculate mood mea­sures that match the issue typology described in chapter 1. A full description of which questions are used to mea­ sure mood in each issue area appears in auxiliary appendix t­ able AA3-2.62 Fi­nally, to test the issue owner­ship hypothesis, I turn to the results of Egan’s analy­sis of survey data since 1970, where respondents ­were asked which party was better equipped to ­handle each of a set of consensus issues.63 (Consensus issues are ­those on which a broad swath of Americans share a par­tic­u­ lar goal, such as lower crime or a more educated populace, even if they disagree about how to achieve that end.) Egan operationalizes this concept, determining ­whether survey respondents have a (statistically significant) preference for one party to address the issue over the other. If this occurs, the policy is considered owned by that party. If neither party has an advantage, the issue is considered nonowned. I then create an indicator variable, coded 1 if the Senate majority party owns the issue and 0 if it does not.64 In addition to t­ hese mea­sures testing my two key hypotheses, I include a series of control variables that capture other features of the institutional environment likely to affect the probability that a del­e­ga­tion exception proposal is successfully ­adopted. First, I control for ­whether the procedural change was proposed by a member of the Senate’s majority party.65 The Senate’s majority party leadership enjoys a range of procedural advantages over its peers in the minority party that make this pro­cess easier. ­These include a preferential right

56

EXCEPTIONS TO THE RULE

of recognition on the floor for the majority leader and the leader’s ability to fill the amendment tree, restricting potential changes to the under­lying bill. ­These advantages translate to, for example, a greater likelihood that majority-­sponsored bills ­will be scheduled for consideration and overcome filibusters; majority-­ sponsored amendments also have a higher probability of being ­adopted.66 Second, I control for w ­ hether the president and the majority party in the House are of the same party as the Senate majority. ­Because majoritarian exceptions are enacted as part of statutory law, they also need to gain the approval of the House of Representatives and the president to be created successfully. Third, I include a variable mea­sur­ing the Senate majority party size, since larger Senate majority parties should find it easier to adopt del­e­ga­tion exceptions. In the earlier part of the time series, when the use of the filibuster on ordinary legislation was less common, a larger majority party co­ali­tion would allow the majority party to lose the support of more of its own members (perhaps t­hose most concerned about cross-­pressures) and still adopt expedited procedures.67 In the ­later part of the time period ­under study, when the use of the filibuster was more routine, a larger majority party co­ali­tion would have to gain the support of fewer minority party members to see the new procedures ­adopted. To estimate the probability that a proposed del­e­ga­tion exception ­will be enacted into law, I use a logistic regression. ­Because most Congresses feature multiple proposals of dif­fer­ent kinds, we might expect the errors to be correlated across all bills in a given session, so I cluster the standard errors by Congress.68 Results ­Table 3-1 pres­ents the results for my main analy­sis. The variables relevant to our three main hypotheses appear in the first three rows. The results of the estimation are consistent with our expectations in all three cases; the coefficients on Most impor­tant prob­lem and Public mood are positive, while the estimated effect of Issue owner­ship is negative. ­Because coefficients in a logistic regression model can be difficult to interpret directly, let us consider, in turn, the predicted probability of a proposal to be enacted at vari­ous levels of each variable.69 Figure 3-2 displays the effect of issue salience. (On this and all other figures, the shaded area represents a 95 ­percent confidence interval.) As we move from policy areas where essentially no respondents to the Gallup Poll have



Obscuring the Causal Chain

57

­ able 3-1.  Probability of adoption of proposed del­e­ga­tion exceptions, T proposal-­level analy­sis, 91st to 112th Congresses, 1969–2012a Public mood

2.893** (1.380)

Most impor­tant prob­lem

8.253* (4.702)

Issue owner­ship

−2.045* (1.079)

Majority

2.875** (1.123)

House

−1.740*** (0.409)

President

−1.159** (0.451)

Majority seat share

31.524*** (6.772)

Constant Log pseudolikelihood Pseudo R 2 Observations

−22.027*** (3.933) −102.45 0.18 371

a. Standard errors clustered by Congress in parentheses. ***p