Markets, Governance, and Institutions in the Process of Economic Development 9780198812555, 0198812558
Written in honour of Kaushik Basu, Markets, Governance, and Institutions in the Process of Economic Development is a cel
282 116 23MB
English Pages 416 [486] Year 2017
Title Pages
Ajit Mishra and Tridip Ray
Title Pages
(p.i) Markets, Governance, and Institutions in the Process of Economic Development (p.ii) (p.iii) Markets, Governance, and Institutions in the Process of Economic Development
Title Pages
Dedication
Ajit Mishra and Tridip Ray
Dedication
(p.v) A Festschrift in Honour of Kaushik Basu
(p.ix) List of Figures
Ajit Mishra and Tridip Ray
(p.ix) List of Figures
(p.ix) List of Figures
(p.xi) List of Tables
Ajit Mishra and Tridip Ray
(p.xi) List of Tables
(p.xi) List of Tables
(p.xiii) List of Contributors
Ajit Mishra and Tridip Ray
(p.xiii) List of Contributors
(p.xiii) List of Contributors
(p.xiii) List of Contributors
(p.xiii) List of Contributors
(p.xiii) List of Contributors
(p.xiii) List of Contributors
(p.xvii) Foreword: Early Kaushik and the World Lately
Ajit Mishra and Tridip Ray
(p.xvii) Foreword: Early Kaushik and the World Lately
1
(p.xvii) Foreword: Early Kaushik and the World Lately
(p.xvii) Foreword: Early Kaushik and the World Lately
(p.xvii) Foreword: Early Kaushik and the World Lately
(p.xx) 2
(p.xvii) Foreword: Early Kaushik and the World Lately
(p.xvii) Foreword: Early Kaushik and the World Lately
(p.xvii) Foreword: Early Kaushik and the World Lately
Notes:
Introduction
Ajit Mishra and Tridip Ray
Introduction
Ajit Mishra
Tridip Ray
Abstract and Keywords
Introduction
Introduction
Introduction
Introduction
Introduction
Introduction
Introduction
Introduction
Notes:
Introduction
Introduction
Markets, States, and Institutions
Ajit Mishra and Tridip Ray
Markets, States, and Institutions
Joseph E. Stiglitz
Abstract and Keywords
Markets, States, and Institutions
Markets, States, and Institutions
Markets, States, and Institutions
Markets, States, and Institutions
1 New Understandings of Markets
Markets, States, and Institutions
1.1 Aggregation
Markets, States, and Institutions
1.2 Contract Enforcement
1.3 Competition and Power
Markets, States, and Institutions
2 Public Governance
Markets, States, and Institutions
Markets, States, and Institutions
Markets, States, and Institutions
3 Reducing the Likelihood of State Capture
Markets, States, and Institutions
Markets, States, and Institutions
4 Concluding Remarks
Markets, States, and Institutions
Acknowledgements
Markets, States, and Institutions
Markets, States, and Institutions
Markets, States, and Institutions
Markets, States, and Institutions
Notes:
Markets, States, and Institutions
On a Concept of Freedom
Ajit Mishra and Tridip Ray
On a Concept of Freedom
Prasanta K. Pattanaik
Yongsheng Xu
Abstract and Keywords
1 Introduction
On a Concept of Freedom
On a Concept of Freedom
2 Competitive Markets, Scarcity of Aggregate Resources, and Interdependence of Consumers’ Choices
2.1 Competitive Markets and Consumers’ Freedom
On a Concept of Freedom
On a Concept of Freedom
(p.36) 2.2 Conditionality of Freedom
On a Concept of Freedom
(p.37) 3 Strategic Interaction of Individuals and the Opportunity Set
3.1 Some Examples
On a Concept of Freedom
On a Concept of Freedom
3.2 Strategic Interaction and Freedom of Choice
On a Concept of Freedom
On a Concept of Freedom
On a Concept of Freedom
On a Concept of Freedom
On a Concept of Freedom
On a Concept of Freedom
(p.45) 4 Concluding Remarks
On a Concept of Freedom
Acknowledgements
On a Concept of Freedom
Notes:
On a Concept of Freedom
Multidimensional Poverty
Ajit Mishra and Tridip Ray
Multidimensional Poverty
Some Comments
Bhaskar Dutta
Abstract and Keywords
1 Introduction
Multidimensional Poverty
2 Basic Framework
Multidimensional Poverty
Multidimensional Poverty
3 Identifying the Poor
Multidimensional Poverty
Multidimensional Poverty
Multidimensional Poverty
4 The Aggregation Exercise
Multidimensional Poverty
Multidimensional Poverty
Multidimensional Poverty
Multidimensional Poverty
Multidimensional Poverty
4.1 Some Specific Multidimensional Poverty Measures
Multidimensional Poverty
Multidimensional Poverty
Multidimensional Poverty
5 Concluding Remarks
Acknowledgements
Multidimensional Poverty
Multidimensional Poverty
Notes:
The Quintile Income Statistic and Distributional Analysis
Ajit Mishra and Tridip Ray
The Quintile Income Statistic and Distributional Analysis
S. Subramanian
Abstract and Keywords
1 Introduction
The Quintile Income Statistic and Distributional Analysis
2 The Quintile Income Statistic
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
3 The Quintile Income Statistic and Money-Metric Poverty
3.1 Quintile Income as a Fuzzy Poverty Indicator
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
3.2 Towards a More Substantive Poverty-Related Interpretation of Quintile Income
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
4 Quantile Statistics and the Assessment of Inclusive Growth
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
5 Concluding Remarks
Acknowledgements
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
The Quintile Income Statistic and Distributional Analysis
Sensitivity of Stationary Equitable Preferences
Ajit Mishra and Tridip Ray
Sensitivity of Stationary Equitable Preferences
Tapan Mitra
Abstract and Keywords
1 Introduction
Sensitivity of Stationary Equitable Preferences
Sensitivity of Stationary Equitable Preferences
Sensitivity of Stationary Equitable Preferences
2 Axioms on Preferences
2.1 Basic Axioms
Sensitivity of Stationary Equitable Preferences
2.2 Equity
2.3 Stationarity
Sensitivity of Stationary Equitable Preferences
3 Sensitivity
Sensitivity of Stationary Equitable Preferences
3.1 A Classification of Alternatives
Sensitivity of Stationary Equitable Preferences
3.1.1 A Basic Classification
Sensitivity of Stationary Equitable Preferences
Sensitivity of Stationary Equitable Preferences
3.1.2 First Period Sensitivity and Strong Pareto
Sensitivity of Stationary Equitable Preferences
3.1.3 A Refined Classification of Alternatives
Sensitivity of Stationary Equitable Preferences
3.2 Representable SWOs and Insensitivity to the Present
Sensitivity of Stationary Equitable Preferences
3.3 Representable SWOs and Sensitivity: An Example
Sensitivity of Stationary Equitable Preferences
Sensitivity of Stationary Equitable Preferences
Sensitivity of Stationary Equitable Preferences
Sensitivity of Stationary Equitable Preferences
4 Concluding Remarks
Sensitivity of Stationary Equitable Preferences
Acknowledgements
Sensitivity of Stationary Equitable Preferences
Notes:
The Link between Preferences, Prices, Inequality, and Poverty
Ajit Mishra and Tridip Ray
The Link between Preferences, Prices, Inequality, and Poverty
The Indian Evidence
Ranjan Ray
Abstract and Keywords
1 Introduction
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
2 Methodologies and Concepts for the Welfare Analysis of Price Changes
2.1 A Welfare Consistent Measure of Price Changes: The TCLI
The Link between Preferences, Prices, Inequality, and Poverty
2.2 Evaluating the Distributive Consequences of Price Changes
The Link between Preferences, Prices, Inequality, and Poverty
2.2.1 Spatial Price Index between Regions within a Country
2.2.2 Equivalence Scales
The Link between Preferences, Prices, Inequality, and Poverty
2.3 Spatial Price Deflated Real Expenditure Comparisons between Regions
The Link between Preferences, Prices, Inequality, and Poverty
3 Selected Empirical Evidence on NSS data
3.1 Prices and Expenditure Inequality
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
(p.107) 3.2 Prices and Expenditure Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
3.3 Spatial Differences in Inequalities and Prices
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
3.4 Welfare Rankings of States in India
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
4 Concluding Remarks
The Link between Preferences, Prices, Inequality, and Poverty
Acknowledgements
The Link between Preferences, Prices, Inequality, and Poverty
The Link between Preferences, Prices, Inequality, and Poverty
Notes:
Private versus Public Monopoly
Ajit Mishra and Tridip Ray
Private versus Public Monopoly
Jörgen W. Weibull
Jun Chen
Abstract and Keywords
1 Introduction
Private versus Public Monopoly
Private versus Public Monopoly
2 Model
Private versus Public Monopoly
Private versus Public Monopoly
Private versus Public Monopoly
3 Unregulated Profit Maximization
Private versus Public Monopoly
Private versus Public Monopoly
Private versus Public Monopoly
Private versus Public Monopoly
4 First-Best
Private versus Public Monopoly
(p.129) 5 Intermediate Cases
Private versus Public Monopoly
Private versus Public Monopoly
(p.131) 5.1 Second-Best: Budget-Constrained Welfare Maximization
Private versus Public Monopoly
(p.132) 6 Numerical Simulations
Private versus Public Monopoly
6.1 Profit Maximization
Private versus Public Monopoly
Private versus Public Monopoly
Private versus Public Monopoly
Private versus Public Monopoly
6.2 First-Best
Private versus Public Monopoly
Private versus Public Monopoly
6.3 Second-Best
Private versus Public Monopoly
Private versus Public Monopoly
7 Extensions
Private versus Public Monopoly
Private versus Public Monopoly
8 Concluding Remarks
Private versus Public Monopoly
Acknowledgements
Private versus Public Monopoly
Notes:
Private versus Public Monopoly
US–Japanese Differences in Employment Practices
Ajit Mishra and Tridip Ray
US–Japanese Differences in Employment Practices
New Explanations and Predictions
Hodaka Morita
Abstract and Keywords
1 Introduction
US–Japanese Differences in Employment Practices
1.1 US–Japanese Differences in the Post-War Growth Period
US–Japanese Differences in Employment Practices
2 Explanations and Predictions
(p.148) 2.1 Outline of the Model and Its Analysis
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
2.2 Explanations for US–Japanese Differences
US–Japanese Differences in Employment Practices
2.2.1 Managerial Capability
US–Japanese Differences in Employment Practices
2.2.2 Entry Regulations
US–Japanese Differences in Employment Practices
2.3 Predictions for the Contemporary and Future Economy
US–Japanese Differences in Employment Practices
(p.155) 2.4 Contribution to the Literature
3 Concluding Remarks
US–Japanese Differences in Employment Practices
Acknowledgements
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
Notes:
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
Ajit Mishra and Tridip Ray
US–Japanese Differences in Employment Practices
New Explanations and Predictions
Hodaka Morita
Abstract and Keywords
1 Introduction
US–Japanese Differences in Employment Practices
1.1 US–Japanese Differences in the Post-War Growth Period
US–Japanese Differences in Employment Practices
2 Explanations and Predictions
(p.148) 2.1 Outline of the Model and Its Analysis
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
2.2 Explanations for US–Japanese Differences
US–Japanese Differences in Employment Practices
2.2.1 Managerial Capability
US–Japanese Differences in Employment Practices
2.2.2 Entry Regulations
US–Japanese Differences in Employment Practices
2.3 Predictions for the Contemporary and Future Economy
US–Japanese Differences in Employment Practices
(p.155) 2.4 Contribution to the Literature
3 Concluding Remarks
US–Japanese Differences in Employment Practices
Acknowledgements
US–Japanese Differences in Employment Practices
US–Japanese Differences in Employment Practices
Notes:
US–Japanese Differences in Employment Practices
On Quality Traps and Economic Development
Ajit Mishra and Tridip Ray
On Quality Traps and Economic Development
Patrick M. Emerson
Abstract and Keywords
I Introduction
On Quality Traps and Economic Development
On Quality Traps and Economic Development
On Quality Traps and Economic Development
2 Utility Functions with Supermodularity in the Quality of Complements
On Quality Traps and Economic Development
3 Equilibria when Complementary Quality is Multiplicative in Preferences
3.1 The Model
On Quality Traps and Economic Development
On Quality Traps and Economic Development
On Quality Traps and Economic Development
On Quality Traps and Economic Development
3.2 Illustrative Example
On Quality Traps and Economic Development
3.3 Extensions of the Model
On Quality Traps and Economic Development
On Quality Traps and Economic Development
On Quality Traps and Economic Development
4 Quality Traps and Development
On Quality Traps and Economic Development
5 Concluding Remarks
Acknowledgements
On Quality Traps and Economic Development
On Quality Traps and Economic Development
Notes:
Inequality and Neighbourhood Effects
Ajit Mishra and Tridip Ray
Inequality and Neighbourhood Effects
Market Access and Welfare of the Poor
Namrata Gulati
Tridip Ray
Abstract and Keywords
Inequality and Neighbourhood Effects
1 Introduction
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
2 The Set-up
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
3 Partial Market Access for the Poor and Complete Access for the Rich
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
4 Inequality, Market Access, and Welfare of the Poor
Inequality and Neighbourhood Effects
4.1 Effect of a Change in the Proportion of Poor
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
(p.188) 4.2 Effect of a Mean-Preserving Spread
Inequality and Neighbourhood Effects
(p.189) 4.3 Welfare of the Rich
Inequality and Neighbourhood Effects
(p.190) 4.4 Discussion and Policy Considerations
Inequality and Neighbourhood Effects
5 Poverty, Inequality, and the Possibility of Complete Exclusion of the Poor
Inequality and Neighbourhood Effects
5.1 Summary of Different Equilibrium Possibilities
5.1.1 Complete Market Access for both Rich and Poor
5.1.2 Complete Market Access for the Rich but No Access for the Poor
Inequality and Neighbourhood Effects
5.1.3 Partial Market Access for Poor
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
5.2 Implications of Poverty and Income Inequality
Inequality and Neighbourhood Effects
5.3 Possibility of Complete Exclusion of the Poor
Inequality and Neighbourhood Effects
(p.197) 5.3.1 Implication of the Income Gap
5.3.2 The Case of the Minority Poor
5.3.3 Policy Considerations in the Presence of a Complete Exclusion Possibility
Inequality and Neighbourhood Effects
5.4 Comparison with a Single-Income Neighbourhood
5.4.1 Feasibility Income Threshold in a Single-Income Neighbourhood
(p.199) 5.4.2 Comparing Single-Income with Mixed-Income Neighbourhoods
Inequality and Neighbourhood Effects
6 Concluding Remarks
Inequality and Neighbourhood Effects
Acknowledgements
Inequality and Neighbourhood Effects
Inequality and Neighbourhood Effects
Notes:
Inequality and Neighbourhood Effects
The Great Recession and Life Satisfaction
Ajit Mishra and Tridip Ray
The Great Recession and Life Satisfaction
The Unique Decline for Americans Approaching Retirement Age
John Ifcher
Amanda Cabacungan
Abstract and Keywords
1 Introduction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
(p.204) 2 Data and Descriptive Statistics
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
3 Main Result
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
4 Potential Mechanisms
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
5 Concluding Remarks
The Great Recession and Life Satisfaction
The Great Recession and Life Satisfaction
Notes:
The Great Recession and Life Satisfaction
Does Rising Inequality Delay Marriage?
Ajit Mishra and Tridip Ray
Does Rising Inequality Delay Marriage?
Evidence from India
Amalavoyal V. Chari
Annemie Maertens
Sinduja Srinivasan
Abstract and Keywords
1 Introduction
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
2 A Model of Marriage Market Search
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
3 Data
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
(p.223) 4 Analysis
4.1 Effect of Income Inequality on Marriage Rates
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
4.2 Effect of Income Inequality on Female Education
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
5 Concluding Remarks
Does Rising Inequality Delay Marriage?
Does Rising Inequality Delay Marriage?
Notes:
Does Rising Inequality Delay Marriage?
The Impact of Migration on Child Labour
Ajit Mishra and Tridip Ray
The Impact of Migration on Child Labour
Theory and Evidence from Brazil
Garance Genicot
Anna Maria Mayda
Mariapia Mendola
Abstract and Keywords
1 Introduction
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
2 Related Literature
The Impact of Migration on Child Labour
3 Theoretical Framework
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
4 Data
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
5 Empirical Analysis
The Impact of Migration on Child Labour
(p.245) 5.1 The Labour Market Effect at Municipality Level
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
5.2 Child Labour Estimates at the Individual Level
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
6 Concluding Remarks
Acknowledgements
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
The Impact of Migration on Child Labour
Notes:
The Impact of Migration on Child Labour
Relation-Based Governance and Competition
Ajit Mishra and Tridip Ray
Relation-Based Governance and Competition
Avinash Dixit
Abstract and Keywords
1 Introduction
Relation-Based Governance and Competition
Relation-Based Governance and Competition
Relation-Based Governance and Competition
2 Monopolistic Competition
2.1 The Circle Model
Relation-Based Governance and Competition
Relation-Based Governance and Competition
Relation-Based Governance and Competition
2.2 Relational Governance with Limited Span
Relation-Based Governance and Competition
Relation-Based Governance and Competition
Relation-Based Governance and Competition
Relation-Based Governance and Competition
3 Concluding Remarks
Relation-Based Governance and Competition
Relation-Based Governance and Competition
Notes:
Labour and Tenancy in Retrospect
Ajit Mishra and Tridip Ray
Labour and Tenancy in Retrospect
Two Bihari Villages in 1970
Clive Bell
Abstract and Keywords
1 Introduction
Labour and Tenancy in Retrospect
2 Employment and Markets
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
3 The Setting
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
4 The Two Villages: Population and Sample
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
5 Attached Farm Servants
Labour and Tenancy in Retrospect
5.1 Servants
Labour and Tenancy in Retrospect
5.2 Other AFS: Ploughmen
Labour and Tenancy in Retrospect
6 Day Labourers
6.1 Wage Rates
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
6.2 Evidence from Elsewhere
Labour and Tenancy in Retrospect
7 Tenancy
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
8 Draught Power
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
9 Risks
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
10 Brief Reflections on Theory
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
11 Concluding Remarks
Acknowledgements
Labour and Tenancy in Retrospect
Labour and Tenancy in Retrospect
Notes:
Labour and Tenancy in Retrospect
Holding India Together
Ajit Mishra and Tridip Ray
Holding India Together
The Role of Institutions of Federalism
Nirvikar Singh
Abstract and Keywords
1 Introduction
Holding India Together
2 Conceptual Background
Holding India Together
Holding India Together
3 Historical Developments
Holding India Together
Holding India Together
4 Current Institutions
Holding India Together
Holding India Together
Holding India Together
4.1 Finance Commission Transfers
Holding India Together
4.2 Planning Commission Transfers
4.3 Central Ministry Transfers
Holding India Together
4.4 State-Local Transfers
5 Holding Together—Politics, Law, and Bureaucracy
Holding India Together
Holding India Together
Holding India Together
Holding India Together
6 Holding Together—Fiscal Federalism
Holding India Together
7 Asymmetries and the Periphery
Holding India Together
Holding India Together
Holding India Together
8 Concluding Remarks
Holding India Together
Acknowledgements
Holding India Together
Holding India Together
Holding India Together
Notes:
Holding India Together
Holding India Together
Holding India Together
The Dividend of Diversity for India’s Corporate Sector
Ajit Mishra and Tridip Ray
The Dividend of Diversity for India’s Corporate Sector
Ashwini Deshpande
Abstract and Keywords
1 Introduction
The Dividend of Diversity for India’s Corporate Sector
The Dividend of Diversity for India’s Corporate Sector
2 Caste Discrimination and Exclusion
The Dividend of Diversity for India’s Corporate Sector
The Dividend of Diversity for India’s Corporate Sector
2.1 Caste Discrimination in the Labour Market
The Dividend of Diversity for India’s Corporate Sector
3 Diversity and Business Performance
The Dividend of Diversity for India’s Corporate Sector
The Dividend of Diversity for India’s Corporate Sector
3.1 Regulation of Discrimination in the Workplace
The Dividend of Diversity for India’s Corporate Sector
4 Dalit Entrepreneurship and Dalit Employment
The Dividend of Diversity for India’s Corporate Sector
(p.334) 4.1 Homophily
4.2 Discrimination in the Non-Farm Business Sector
The Dividend of Diversity for India’s Corporate Sector
5 Diversity and the Corporate Sector in India
The Dividend of Diversity for India’s Corporate Sector
5.1 Caste Data for the Private Sector
The Dividend of Diversity for India’s Corporate Sector
5.2 How Could the Private Sector be Made More Diverse?
The Dividend of Diversity for India’s Corporate Sector
5.2.1 Other Measures: Supplier Diversity and Beyond
The Dividend of Diversity for India’s Corporate Sector
6 Concluding Remarks
Acknowledgements
The Dividend of Diversity for India’s Corporate Sector
The Dividend of Diversity for India’s Corporate Sector
The Dividend of Diversity for India’s Corporate Sector
Notes:
The Dividend of Diversity for India’s Corporate Sector
A ‘Rights-Based’ Approach to Optimal Tax Policy
Ajit Mishra and Tridip Ray
A ‘Rights-Based’ Approach to Optimal Tax Policy
Eduardo Zambrano
Abstract and Keywords
1 Introduction
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
1.1 Related Literature
A ‘Rights-Based’ Approach to Optimal Tax Policy
(p.346) 2 The Setting
A ‘Rights-Based’ Approach to Optimal Tax Policy
3 Social Orderings
3.1 Efficiency Principles
4 Fairness
A ‘Rights-Based’ Approach to Optimal Tax Policy
4.1 Strong Principles of Fairness
A ‘Rights-Based’ Approach to Optimal Tax Policy
4.2 Fairness–Efficiency Trade-offs
4.3 Moderate Principles of Fairness
A ‘Rights-Based’ Approach to Optimal Tax Policy
4.3.1 Laissez-Faire Choices
A ‘Rights-Based’ Approach to Optimal Tax Policy
4.3.2 Fairness Relative to a Laissez-Faire Benchmark
A ‘Rights-Based’ Approach to Optimal Tax Policy
5 Robustness Principles
A ‘Rights-Based’ Approach to Optimal Tax Policy
6 Fair Social Orderings
A ‘Rights-Based’ Approach to Optimal Tax Policy
(p.355) 6.1 Remarks
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
7 Fair and Efficient Public Policy
7.1 The First Best
A ‘Rights-Based’ Approach to Optimal Tax Policy
7.2 Second-Best Policies when Preferences Are not Observable but Income Is Observable
A ‘Rights-Based’ Approach to Optimal Tax Policy
(p.359) 7.3 An Example
A ‘Rights-Based’ Approach to Optimal Tax Policy
7.3.1 Remarks
A ‘Rights-Based’ Approach to Optimal Tax Policy
8 Concluding Remarks
A ‘Rights-Based’ Approach to Optimal Tax Policy
Acknowledgements
A ‘Rights-Based’ Approach to Optimal Tax Policy
A1 Proof of Theorem 1
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
A2 Proof of Theorem 2
A3 Preparing for the Proof of Theorem 3
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
(p.368) A4 Proof of Theorem 3
A ‘Rights-Based’ Approach to Optimal Tax Policy
A ‘Rights-Based’ Approach to Optimal Tax Policy
Notes:
A ‘Rights-Based’ Approach to Optimal Tax Policy
The Many Faces of Corruption
Ajit Mishra and Tridip Ray
The Many Faces of Corruption
Which One(s) to Target?
Ajit Mishra
Abstract and Keywords
1 Introduction
The Many Faces of Corruption
The Many Faces of Corruption
2 The Many Faces
2.1 Embezzlement and Bribery
The Many Faces of Corruption
2.2 Bribery and Lobbying
(p.374) 2.3 Collusion and Extortion
The Many Faces of Corruption
3 The Model
The Many Faces of Corruption
3.1 Corruption
3.1.1 Preventing Corruption
The Many Faces of Corruption
(p.377) 3.1.2 Deterring Collusion
The Many Faces of Corruption
3.1.3 Deterring Extortion
3.2 Does Extortion Matter?
The Many Faces of Corruption
4 An Experiment
The Many Faces of Corruption
The Many Faces of Corruption
The Many Faces of Corruption
5 Concluding Remarks
Acknowledgements
The Many Faces of Corruption
The Many Faces of Corruption
Notes:
The Many Faces of Corruption
The Many Faces of Corruption
(p.385) Index
Ajit Mishra and Tridip Ray
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
(p.385) Index
Recommend Papers
![The Governance of Telecom Markets: Economics, Law and Institutions in Europe [1st ed.]
9783030581596, 9783030581602](https://ebin.pub/img/200x200/the-governance-of-telecom-markets-economics-law-and-institutions-in-europe-1st-ed-9783030581596-9783030581602.jpg)
![Globalisation and Economic Security in East Asia : Governance and Institutions [1 ed.]
9781134190447, 9780415373500](https://ebin.pub/img/200x200/globalisation-and-economic-security-in-east-asia-governance-and-institutions-1nbsped-9781134190447-9780415373500.jpg)
- Author / Uploaded
- Ajit Mishra (editor)
- Tridip Ray (editor)
File loading please wait...
Citation preview
Title Pages
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Title Pages Ajit Mishra, Tridip Ray
(p.i) Markets, Governance, and Institutions in the Process of Economic Development (p.ii) (p.iii) Markets, Governance, and Institutions in the Process of Economic Development
(p.iv) Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of
Page 1 of 2
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Title Pages Oxford University Press in the UK and in certain other countries © Oxford University Press 2018 The moral rights of the authors have been asserted First Edition published in 2018 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2017942090 ISBN 978–0–19–881255–5 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.
Access brought to you by:
Page 2 of 2
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Dedication
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Dedication Ajit Mishra, Tridip Ray
(p.v) A Festschrift in Honour of Kaushik Basu Kaushik Basu turned 65 earlier this year, on 9 January 2017. We are using this occasion to bring out this volume to celebrate his contributions over the last four decades. This book contains contributions from his present and past collaborators, and research students. He has been an inspiring intellectual and a dear friend; we are delighted to be able to honour him with this collection of essays. (p.vi)
Access brought to you by:
Page 1 of 1
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Figures
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
(p.ix) List of Figures Ajit Mishra, Tridip Ray
3.1 Competitive markets and individual freedom 34 5.1a Ratios of actual to warranted mean consumption expenditure (rural) 75 5.1b Ratios of actual to warranted mean consumption expenditure (urban) 75 7.1a Hasse diagrams for rural India 113 7.1b Hasse diagrams for urban India 114 8.1 The set of willing buyers 122 8.2 The effect of shared component precision 133 8.3 The effect of idiosyncratic component precision 134 8.4 The effect of value dispersion 135 8.5 Revenue and price given the signal precision 136 8.6 The effect of value dispersion on information acquisition 137 8.7 Pricing as a function of value dispersion 139 8.8 Information acquisition as a function of value dispersion 140 10.1 The firms’ reaction functions 167 11.1 Demand faced by Firm j 179 11.2 Aggregate market access of the poor 185 11.3 Aggregate consumer surplus of the poor 186 11.4 Different equilibrium possibilities 194 12.1 Nominal household income time series 209 12.2 Unemployment-rate time series 210 12.3 Life-satisfaction time series 210 12.4 Case–Shiller Housing Price Index 211 12.5 S&P 500 Index and life-satisfaction 211 12.6 Age 55–64 life-satisfaction time series 213 14.1a Effect of unskilled migration: Child labour decreases 240 Page 1 of 2
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Figures 14.1b Effect of unskilled migration: Child labour increases 240 14.2a Effect of skilled migration: On the skilled market 241 14.2b Effect of skilled migration: On child labour 241 (p.x) 15.1 Monopolistic competition on circular space 263 18.1 Log wage gaps between Others and SC–STs 328 19.1 Individual i pays ti in taxes, public capital good g 346 19.2 Equal pay for equal work 348 19.3 Weak social improvement 349 19.4 Liberal reward 350 19.5 Choices in a laissez-faire world 351 19.6 Laissez-faire reward 352 19.7 Amount individual j willing to pay for the right to consume bundle zj 354 19.8 Two ways of calculating uiL(zi) 356 19.9 The first-best policy 357 19.10 Second-best optimal policies 358 19.11 Proof of Theorem 1 363 19.12 Proof of Theorem 2 364 19.13 Proof of Lemma A.1 366 19.14 Proof of Theorem 3 368
Access brought to you by:
Page 2 of 2
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Tables
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
(p.xi) List of Tables Ajit Mishra, Tridip Ray
3.1 Freedom as the power to choose uncertain prospects 40 3.2 Limitations of Formulation 1 41 3.3 Equilibrium and freedom as the power to choose outcomes 42 3.4 Formulation 3 and contingent freedom 43 3.5 Limitations of Formulation 3 44 5.1 Some global statistics on the decile income 71 5.2 Some Indian statistics on the quintile consumption level 72 5.3a The Gini coefficient of inequality in the distribution of global income 76 5.3b Global statistics on decile income (1988–2008) 77 7.1 Quintile shares of total expenditure in rural areas 105 7.2 Quintile shares of total expenditure in urban areas 105 7.3 Nominal and real expenditure inequalities in rural areas 106 7.4 Nominal and real expenditure inequalities in urban areas 106 7.5 Headcount poverty rates 107 7.6 State specific and All-India temporal price indices 108 7.7 State specific spatial price indices 110 7.8 State specific and All-India Gini coefficients 111 12.1 Descriptive statistics 204 12.2 Standard ‘happiness’ regression 205 12.3 DD estimates 207 13.1 Descriptive statistics 221 13.2 Earnings distributions by caste 222 13.3 Effect of income inequality on marriage rates 224 13.4 Effect of income inequality on predetermined characteristics 225 13.5 Effect of income inequality on age at marriage 226 13.6 Are the results driven by unobserved tastes for education? 228 Page 1 of 2
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Tables 13.7 Effect of income inequality on school attendance and completed years of schooling 229 (p.xii) 13.8 Effect of income inequality on completion rates of schooling levels 230 14.1 Summary statistics 244 14.2 Labour market effect of internal migration: Adult 247 14.3 Labour market effect of internal migration on children: OLS 248 14.4 Labour market effect of internal migration on children: IV results 249 14.5 Labour market effect of internal migration on children: By skill 251 16.1 Size distribution of ownership holdings 277 16.2 Daily wage rates by operation 283 16.3 Size distributions of ownership and operational holdings 287 16.4 Quality of tenants’ holdings 290 20.1 Tobit regression on agreed bribe amounts in Round 2 381
Access brought to you by:
Page 2 of 2
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Contributors
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
(p.xiii) List of Contributors Ajit Mishra, Tridip Ray
Clive Bell is Professor of Economics Emeritus at Heidelberg University. His chief areas of research are rural economic organization and social cost-benefit analysis. His recent work deals with the long-running economic costs of low-frequency epidemics, especially AIDS in Africa, and the evaluation of India’s rural roads programme. He has taught at Johns Hopkins, Regensburg, Sussex, and Vanderbilt Universities, and was earlier a senior economist at the World Bank. Amanda Cabacungan is an Economic Consultant at Torrey Partners, San Diego, specializing in the quantification of intellectual property and contract damages. Her research interests include subjective well-being and behavioural economics. She graduated summa cum laude from Santa Clara University with Bachelor of Science Degrees in Economics and Mathematics. A. V. Chari is a Reader in the Department of Economics at the University of Sussex. His research interests include labour markets and health issues in developing countries. Jun Chen is a Post-Doctoral Fellow at Hanken School of Economics, Helsinki, with particular interests in information economics and microeconomic theory. He holds a PhD in Economics from Stockholm School of Economics, and an MA degree and BA degree in Economics from Renmin University of China. Ashwini Deshpande
Page 1 of 6
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Contributors is a Professor of Economics at Delhi School of Economics, University of Delhi. Her current research areas include economics of discrimination and affirmative action, with a special focus on caste and gender in India. She has previously worked in areas of international debt and aspects of the Chinese economy. Avinash Dixit is John J. F. Sherrerd ’52 University Professor of Economics Emeritus at Princeton University. His books on game theory, economics of uncertainty, and economics of law and governance are bestsellers worldwide. His research interests have included microeconomic theory, game theory, international trade, industrial organization, growth and development theories, public economics, political economy, and the new institutional economics. Bhaskar Dutta is a Professor of Economics at Warwick University and heads the Department of Economics at Ashoka University, India. He has served as President of the Society for Social Choice and Welfare and is a Fellow of the Econometric Society. His research interests include mechanism design, formation of groups and networks, cooperative game theory, political economy, and social choice theory. (p.xiv) Patrick M. Emerson is Professor of Economics at Oregon State University and Research Fellow at the Institute of Labor Economics (IZA). He received his PhD in Economics from Cornell University in 2000 under the supervision of Kaushik Basu. His research focuses on development, labour, growth, public policy, and industrial organization. Garance Genicot is an Associate Professor of Economics at Georgetown University, specializing in development economics. Her specific areas of research include risk sharing, informal credit markets, social networks, and inequality. She is a Research Associate at the National Bureau of Economic Research (NBER), a Fellow at the Bureau for Research and Economic Analysis of Development (BREAD), and a Research Fellow at the Institute of Labor Economics (IZA). Namrata Gulati is Assistant Professor of Economics at South Asian University, New Delhi, and is a former faculty member at Indira Gandhi Institute of Development Research (IGIDR), Mumbai. Her academic interest is in understanding the impact of inequality on the welfare of the poor. John Ifcher is an Assistant Professor of Economics at Santa Clara University. His recent research focuses on decision making on behalf of others, subjective well-being, and social welfare programmes. One recent paper investigates the relationship between neighbours’ income and Page 2 of 6
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Contributors subjective well-being; another investigates the effect of the Earned Income Tax Credit on happiness. Annemie Maertens is a Senior Lecturer in the Department of Economics at the University of Sussex. Her research is on agricultural markets, technology adoption, and gender. Anna Maria Mayda is an Associate Professor of Economics at Georgetown University. Her research mainly focuses on the political economy of trade and immigration policy. She is a Research Affiliate at the Centre for Economic Policy Research (CEPR), a Research Fellow at the Institute of Labor Economics (IZA), and an External Fellow at the Center for Research and Analysis of Migration (CReAM). Mariapia Mendola is an Associate Professor of Economics at the University of MilanBicocca. She is a Research Fellow at the Institute of Labor Economics (IZA) in Bonn, and a board member of the Italian Center for International Development (ICID) in Rome. Her main research interests are in the fields of development economics, household economics, and economics of migration. Ajit Mishra is Associate Professor of Economics at the University of Bath. His research interests include economic development, inequality, economics of corruption, and measurement of vulnerability. He graduated from the Delhi School of Economics and has taught at various places including Delhi School of Economics, Indira Gandhi Institute of Development Research (IGIDR), and Universities of Dundee and Edinburgh. Tapan Mitra is Goldwin Smith Professor of Economics and a faculty member of the Center for Applied Mathematics at Cornell University. He is an Alfred P. Sloan Fellow and a Fellow of the Econometric Society. His research interests are economic dynamics, social choice, and natural resource economics. Hodaka Morita is Professor of Economics at the University of New South Wales in Sydney. He received his PhD from Cornell University under supervision of Kaushik Basu. Hodaka’s research focuses on applied microeconomic theory, where his main fields of (p.xv) interests are industrial organization, organizational economics, labour economics, and international trade. Prasanta K. Pattanaik is Emeritus Professor at the University of California, Riverside. A fellow of the Econometric Society, he has served as President of the Page 3 of 6
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Contributors Society for Social Choice and Welfare and the Indian Econometric Society. His current research interests lie mainly in the areas of welfare economics and the theory of social choice, decision theory (including the theory of choice under uncertainty and the theory of fuzzy preferences), and the measurement of deprivation and the standard of living. Ranjan Ray is a Professor of Economics at Monash University, Australia. He also serves at the Editorial Board of Review of Income and Wealth and Journal of Human Development and Capabilities. His research concentrates on developmental issues in South Asia spanning overconsumption patterns, tax reforms, human development, gender issues, and intra-household decision making. Tridip Ray is a Professor of Economics at the Indian Statistical Institute (ISI), Delhi. He received his PhD in economics from Cornell University and taught at the Hong Kong University of Science and Technology before joining ISI Delhi. His research areas include economic development and growth, inequality, financial development, economics of education, and issues in applied microeconomics. Amartya Sen is Thomas W. Lamont University Professor, and Professor of Economics and Philosophy, at Harvard University. He was until 2004 the Master of Trinity College, Cambridge. He was formerly Professor of Economics at Jadavpur University Calcutta, the Delhi School of Economics, and the London School of Economics, and Drummond Professor of Political Economy at Oxford University. His research ranges over social choice theory, decision theory, ethics and political philosophy, welfare economics, measurement theory, development economics, public health, and gender studies. He was awarded the 1998 Nobel Prize in Economics for his contributions to welfare economics. Recipient of several awards and honours by many countries, Professor Sen received India’s highest civilian honour, the Bharat Ratna, in 1999. Nirvikar Singh is Director of the Center for Analytical Finance at the University of California, Santa Cruz. His current work includes projects on capital flows and monetary policy rules in India, including an NIPFP-DEA (National Institute of Public Finance and Policy) project funded by the Ministry of Finance, healthcare systems in Punjab, and Indian– American entrepreneurs in the US. Sinduja Srinivasan is an Economist at the United Nations Economic Commission for Latin America and the Caribbean, Santiago, Chile. She specializes in Page 4 of 6
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Contributors programme evaluation, the agricultural transition in Latin America, and data visualization. Joseph E. Stiglitz is Professor of Economics at Columbia University. He is also Co-Chair of the High-Level Expert Group on the Measurement of Economic Performance and Social Progress at the OECD, and Chief Economist of the Roosevelt Institute. A recipient of the Nobel Memorial Prize in Economic Sciences (2001) and the John Bates Clark Medal (1979), he is a former Senior Vice-President and Chief Economist of the World Bank and a former member and chairman of the (US president’s) Council of Economic Advisers. His work focuses on income distribution, asset risk management, corporate governance, market imperfections, and information economics. (p.xvi) S. Subramanian, a former professor of the Madras School of Economics, is an Indian Council of Social Science Research (ICSSR) National Fellow at Madras School of Development Studies. His research areas include aspects of social and economic measurement with specific reference to poverty, inequality, discrimination, and demographic indicators; development economics, collective choice theory, and welfare economics. Jörgen W. Weibull holds the A.O. Wallenberg Professorship in Economics at Stockholm School of Economics. He is a member of the Royal Swedish Academy of Sciences (KVA) and a fellow of the Econometric Society. His research areas are game theory, microeconomic theory, political economy, mathematical economics, decision theory, industrial organization, and evolutionary biology. Yongsheng Xu is a Professor of Economics at Andrew Young School of Policy Studies at Georgia State University. Xu’s research and teaching interests include microeconomic theory, particularly individual and collective choice theory, welfare economics, and information and organization. His current work focuses on the measurement of deprivation and the standard of living, individual choice behaviour, and tax and fiscal competition. Eduardo Zambrano is Professor of Economics and Jacobsen Faculty Fellow at Cal Poly in San Luis Obispo, California, with research interests in general equilibrium theory, game theory, decision theory, political economy, and welfare economics. He obtained his PhD in Economics in 1999 from Cornell University under the supervision of Basu and has worked with the United Nations Development Programmes (UNDP), United Nations Environment Programme (NEP), and Organisation for Page 5 of 6
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
List of Contributors Economic Co-Operation and Development (OECD) in matters regarding the measurement of human development, gender inequality. Homa Zarghamee is an Assistant Professor of Economics at Barnard College. Her recent research focuses on decision making on behalf of others, the impact of mood on economic behaviour, gender and competition, and determinants of subjective well-being. She received her PhD and MA in Economics and BA in Mathematics and Economics from Cornell University under Kaushik Basu’s supervision.
Access brought to you by:
Page 6 of 6
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
(p.xvii) Foreword: Early Kaushik and the World Lately Amartya K. Sen
1 It was in the autumn of 1972 that I first met Kaushik Basu at the London School of Economics, where he had just arrived to do a Master’s degree in Economics, to be followed by a doctorate. LSE, which I had joined only a year earlier, was a new base for me too. I had moved from the Delhi School of Economics where I had wonderful students, and my predatory inclination to look for talented and motivated students had been, by then, only partly met. As I chatted with Kaushik, it rapidly became clear to me that he had a brilliant mind. It was also easy to see that he was deeply interested in economics of a kind that would both demand analytical sophistication, and could make a big difference to the world. Kaushik came to my LSE room in the company of Siddiqur Osmani from Bangladesh (they were close friends already), and we chatted together. It did not take much time to see that in Kaushik and Siddiqur we had two of the cleverest and best-motivated economics students we could possibly find in the academic world. After their shining performances in the MSc, both Kaushik and Siddiqur went on to do imaginative and powerful work for their doctorate degrees (the two theses were published respectively by Cambridge University Press and Oxford University Press). I was privileged to supervise their research. In this Festschrift for Kaushik, I pick up the story from the early days of his doctoral research—this would have been during the academic year 1974–75—in Kaushik’s first year as a PhD student. A lot of work was going on already by then, on how to get information about a person’s preferences—and priorities and inclinations—from the choices he or she makes when there are several alternatives to choose from. Kaushik investigated a fresh set of probing Page 1 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately questions (indeed a new field of systematic enquiry—going beyond individual choices to governmental decisions): what can we read from the choices that (p.xviii) a government makes? What can we infer from them about the government’s preferences and values? What can we say about the ‘revealed preference of government’?1 To be sure, at an implicit—and often rather casual—level, people were used to commenting on what the ‘government really prefers’—basing their reading on the observation of public decisions. But as Kaushik noted, simple connections between choices and priorities are even more difficult to draw in the case of governmental choice than with individual choices, and it is notoriously difficult to decide what the government ‘really wants’. Kaushik discussed many difficulties that would be present in trying to jump from governmental choices to the ‘preferences’ of the government. And each of these difficulties turned out to be important issues to investigate. To consider a few of the problems, first, the government has different departments taking distinct decisions, and their priorities need not all be the same. There could be internal dissensions—even battles—within a government and the actual decisions may be a compromise—explicitly agreed or implicitly emergent. Second, there are many different people involved in the decisions taken even by one department, and their respective inputs can, at least to some extent, conflict with each other. ‘Power’, as Kaushik put it in Revealed Preference of Government, ‘is diffused in many quarters and the government is a nebulous organisation’ (1980: 6). Third, the information on the basis of which government agents make their decisions may not be very precise, and this lack of precision can make reasoned decisions imprecise as well. Fourth, to the extent that a government can be assumed to be maximizing aggregate welfare of the people involved (as is often assumed in ‘public finance’ and more generally ‘public economics’), there are problems of interpersonal comparison of well-being of which note has to be taken, and they may reflect inescapable ambiguities in comparing different persons’ well-being. Fifth, when information is incomplete, there can be decisions to be taken about investing in getting better—or fuller—data, which would involve an incurring of costs even as the resulting expansion of information may allow better decisions to be taken. There are real difficulties in deciding on how much to invest in informational explorations with uncertain rewards.
Page 2 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately Sixth, going more towards the foundations of the discipline of revealed preference, Kaushik also asked whether a government can be taken to be a coherent ‘agent’, as the discipline tends to demand from decision takers. There (p.xix) may be serious doubts on whether a government has, as he put it, ‘a unitary interest motivating its decisions’ (1980: 57). These are examples of some of the difficulties that Kaushik addressed (there were also other identified problems which he took on). What is truly remarkable is the way he pursues each of the problems, sometimes with informal reasoning, but quite often with extensive formal analyses. As it happens, many of these difficulties arise in one form or another even for individual decisional choices. As a result, in the process of examining the ‘revealed preference of government’, Basu ended up significantly illuminating the general exercise of relating observed choices to underlying valuations—for individuals, or groups, or entities (like a government). A large area of choice theory, and also what can be called utility theory (under a broad interpretation), in general, gets powerfully explored in this thesis ostensibly confined to ‘the revealed preference of government’. Along with this general ground clearing, Kaushik went on to study and often resolve particular—sometimes well-known—problems related to these general concerns, such as understanding the consequences of repeated ‘Prisoner’s Dilemma’, extending the reach of utilitarian reasoning in welfare economics, the use of trade-offs in the rights people respectively have, and the way ‘fuzziness’ can be systematically incorporated in representing—or reflecting—imprecise information and partly unclear priorities. Basu’s works on these distinct problems came out—often after his dissertation was completed (through the 1980s)—in papers published in a variety of journals, varying from quite technical ones such as Econometrica, Journal of Economic Theory, Review of Economic Studies, Theory and Decision, and Journal of Mathematical Economics, to more general ones, such as Economica, Quarterly Journal of Economics, Oxford Economic Papers, and even Explorations in Economic History. If Kaushik can be seen as spending most of his time in later years in being an applied economist, writing on such subjects as ‘India’s emerging economy’, or working out the demands of being ‘an economist in the real world’, and serving as an economic advisor or decision taker (including being the Chief Economist of India, and later, of the World Bank), this followed an earlier part of his life as an innovative theorist, even what can be called (following Francis Edgeworth, the economist, or Hans Kelsen, the legal theorist) a ‘pure theorist’. What is, however, particularly striking is that even in his pure theory work, Kaushik was typically being guided by his interest in practical problems in the world— research that would be important for his applied work many decades later. It is this combination of analytical force with practical motivation that I had found
Page 3 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately striking when I first came to know Kaushik at the London School of Economics, and it has been serving him very well through his illustrious professional life.
(p.xx) 2 Well, that was a quick glimpse of early Kaushik. What about the world lately—a world in which Kaushik is a big player? Aside from the new and original work that Kaushik has continued to do throughout his very successful life as a teacher, researcher, advisor, and decision maker, with great relevance to the problems of the contemporary world, his early research too has remained closely relevant to problems today. I have already commented on why and how Kaushik’s early works are of great use for addressing practical problems today, particularly in dealing with decisions taken by government—and other entities, including international institutions (like the World Bank). They are also relevant in understanding the demands of social choice and the pursuit of social welfare, involving complexities of the relation between choices and priorities. I turn now to a less explored connection—and a less examined problem: that of interpreting a consensus which is a central issue for political economy and social choice. In a justly celebrated paper, Cass Sunstein, a leading philosopher and legal theorist, has argued that political consensus may often be reached on the basis of ‘incompletely theorized agreements’.2 There may be something somewhat deceptive in the agreement. Indeed, the consensus arrived at may actually break down if a theorized agreement is sought, since the absence of a shared theory may help people to arrive at the same conclusion. The presence of fragility in the agreement may well be effectively obscured by not insisting on a theorized agreement. This is a very significant recognition, and its implications for political ethics can be quite extensive, as Sunstein has brought out with clarity and force. The questions I want to ask here are: (1) what is the relevance of Kaushik’s early analysis of ‘revealed preference’ to this issue? (2) what implications does that analysis have for the political economy of the contemporary world? Note, first, that while the form of Kaushik’s investigation began with observed choices and involved the exercise of extricating preferences and priorities from these observations, the analytical transition can be viewed in exactly the opposite direction as well. We can begin with priorities and objectives and ask what choices should follow from them. But—it can be asked—wouldn’t the reverse trajectory be plain sailing? Shouldn’t the choice decisions be immediate and obvious: pick the best alternative on the menu?
Page 4 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately (p.xxi) In fact, the reverse move might not be quite so easy if there is no best alternative on the menu, which can easily happen if the preference ranking is incomplete. The issues of incompleteness and its underlying causation are important, along with the consequent distinction between maximality and optimality. I have, however, discussed the complications raised by incompleteness in a series of publications from 1970 onwards, and so has Kaushik Basu.3 So let us overlook that problem in this paper. Is there any other problem? The big problem that exists even when the issue of incompleteness of preference rankings is dropped is one that Kaushik aired in the context of going from choices to preferences, but it has relevance in the opposite direction as well. The issue is that several different sets of priorities and preferences may have the same best choice, but for quite different reasons. Kaushik linked this with the main motivation for his thesis on the very first page of his monograph: Clearly, the precise preference of the government cannot, in general, be deduced by observing a single choice, because there will usually be a whole set of preference patterns that could make the chosen project optimal. But it has been alleged that if we observe many choices, we can narrow down the possible preference patterns and may even be able to reveal the exact one. This issue forms the central theme of our discussions.4 Basu does indeed pursue the task of trying to get precise preference patterns from observing multiple choices, and shows varying feasibility related to the actual circumstances. But how does all this relate to the Sunstein problem of ‘incompletely theorized agreements’? The first connection is immediate. If a group of people have ‘a whole set of preference patterns’ that make a particular alternative optimal, then they have every reason to agree on what to ask for, even without agreeing on the reasons for that choice. The interpretation can, then, take the form not so much of an incompletely theorized agreement, but a differently theorized agreement (combining possibly well theorized disagreements on reasons, along with well theorized agreement on what to choose). And unlike in the case of incompletely theorized agreements of the kind to which Sunstein has drawn attention, there may be no lacuna of theorization in Basu’s case. In the world in which we live, differently theorized agreements can be both frequent and perfectly fine. Consider the recent elections in France for the choice of a president. In the second round—that of the ‘run-off’ between Marine Le Pen and Emmanuel Macron—a vast majority buried their political differences and voted for Macron to keep Le Pen out of the presidency. Their (p.xxii) reasons for wanting to keep Le Pen out might well have differed greatly, but the Page 5 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately legitimacy of their agreed choice is not diminished as a result. In fact, to look for a theorized agreement on voting Le Pen out may well be both otiose and frustrating. It is, of course, also true, as Eric Maskin and I have argued in pleading for a pairwise majority contest,5 if we had a proper majority rule in France, each of a number of candidates (Macron, Melenchon, Fillon) would almost certainly have defeated Le Pen in a head-to-head contest. And there could also have been pairwise contests among the ones other than Le Pen, which would have brought out the differences between a very left-wing Melenchon, a moderate Macron, and a firmly right-wing Fillon. But given the French system, Macron being chosen over Le Pen in the run-off, while reflecting a differently reasoned agreement, cannot be taken to be incompletely theorized, or seen as revealing any lack of the backing of reason. I expect Macron would have been the winner as well in pairwise majority contests, even though the French system does not resolve that issue. But what it does resolve need not await completion through further theorizing. Something similar can be said about presidential elections in the United States. Donald Trump did not get a majority in the first seventeen of the Republican primaries, and very plausibly could have lost to one or more of the other candidates had there been pairwise majority votes (easily ascertainable by all the candidates being ranked). Had there been a French-type system of a ‘runoff’, it is quite plausible to expect that one of the other candidates would have defeated Trump in a head-to-head contest, because of the desire of many of the voters to keep Trump out—for possibly very different reasons. A similar problem arises in India, where the Hindutva-oriented Bharatiya Janata Party (the BJP) led by Mr Modi got 31 per cent of the votes in the general elections of 2014, but a majority of the seats in the Lok Sabha of the Indian Parliament. They were all multi-cornered fights, and even though a large majority were against the Modi-led BJP, it could still prevail. Had there been a coalition of the other parties, it is plausible to expect that the BJP would have been defeated, as happened in the following year in Bihar, even though the nonBJP voters, coming from different parties, would have differently theorized their votes. That possibility was present even in the Uttar Pradesh (UP) elections, in which BJP had a sweeping victory on the basis of a minority of votes in a field lacking in critically important coalitions. This issue will be engaged in, again in the general elections to come in 2019, where coalition building would be central to the electoral outcome. The many-one connection (p.xxiii) between preferences and choice, extensively studied by Kaushik in his first book, is an important point for reflection in Indian politics today.
Page 6 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Foreword: Early Kaushik and the World Lately In the world lately, whether in India, or in Europe, or America, the presence of many-one relations between preferences and choice remains a central diagnosis. It has practical importance, but also analytical significance in helping us to distinguish between differently theorized agreements and incompletely theorized agreements. It is interesting that Kaushik Basu’s research on preference and choice remains so relevant more than forty years after that work was undertaken. (p.xxiv) Notes:
(1) Kaushik Basu, Revealed Preference of Government (Cambridge: Cambridge University Press, 1980). (2) Cass R. Sunstein, ‘Incompletely Theorized Agreements’, Harvard Law Review 108 (1995). (3) See in particular Amartya Sen, Collective Choice and Social Welfare (San Francisco: Holden-Day, 1970; Amsterdam: North-Holland, 1979; extended edition, London: Penguin, and Cambridge, MA: Harvard University Press, 2017); and Basu, Revealed Preference of Government (1980), chaps 1 and 6. (4) Basu, Revealed Preference of Government (1980), p. 1. (5) Eric Maskin and Amartya Sen, ‘A Better Way to Choose Presidents’, New York Review of Books 64(10) (8 June 2017).
Access brought to you by:
Page 7 of 7
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Introduction Ajit Mishra Tridip Ray
DOI:10.1093/oso/9780198812555.003.0001
Abstract and Keywords This chapter looks at some of the key contributions by Kaushik Basu and discusses how the different chapters relate to his interests and contributions. Beginning with his early work on interlinked markets and usurious interest rates, to the recent work on child labour and shared prosperity, Basu’s research over the last four decades has contributed immensely to the way the field of development economics has been transformed. This chapter highlights this broad range of contributions and provides a brief narrative of the interactions between markets, governance, and institutions in the context of economic development. The chapter provides an introduction to each contribution, demonstrating their interrelations and complementarities. Keywords: Basu, markets, governance, institutions, development
When we started planning for this book to honour Kaushik Basu in his sixty-fifth year, we were unsure on several fronts. First, it wasn’t clear whether we would focus on development or on game theory or on normative economics. Basu’s range of contributions made it difficult for us. Second, Basu is a much loved and admired scholar and the set of potential contributors to a volume in his honour would simply be too big for us (two editors with little experience). The solution to the second problem was simple: we chose to make our task easier by inviting his collaborators, past and present, and his research students to contribute. The response has been amazing. For the first problem, we have left the canvas quite broad; a true reflection of Basu’s work.
Page 1 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction Beginning with his early work on interlinked markets and usurious interest rates, to the recent work on child labour and shared prosperity, Basu’s research over the last four decades has contributed immensely to the way the field of development economics has been transformed. In fact, his book The Less Developed Economy: A Critique of Contemporary Theory (Basu, 1984) changed the way development economics began to be taught in many universities. He has been one of the leading scholars responsible for reshaping development economics as a rigorous and theoretically well-founded discipline. The collection of papers in his Development, Markets, and Institutions (Volume 3, Collected Papers in Theoretical Economics, 2005) is a reflection of this. The current book takes inspiration from this and looks at the complex interactions between markets, governance, and institutions in the process of economic development. Basu’s range of contributions is evident in the four-volume set of collected papers (Basu, 2005) published by Oxford University Press. At one level, he has been involved in deep philosophical issues in welfare analysis and decision making, and theoretical analysis of human behaviour in different settings and (p.2) contexts. By his own admission, he has always been fascinated by the abstract, logical foundations of economics.1 The sheer beauty of deductive reasoning lured him to abandon his plans for a legal career in favour of a career in economics research. Our discipline, with its unique blend of positive science and normative philosophy, has certainly benefited from the legal profession’s loss. Several contributions in this book share his love for scientific pursuits. However, at the same time, he has been deeply concerned, both as a researcher and as a policy advisor, with several policy issues such as rent control, child labour, labour laws, harassment, shared prosperity, and gender empowerment. His compassion for fellow human beings and his desire to contribute to society has been evident in his working life. He mentions how he was in a quandary when he was invited to be Chief Economic Advisor to the Indian Government, as it was going to take him away from his research life; but he felt the need to give back to society.2 He has always contributed to society through his research, teaching and institution building;3 but this was his modest way of accepting the role of a responsible public servant, which after a few years he continued in the World Bank. The contributions from authors in this volume, theoretical as well as empirical, reflect this range of interests in some of the pressing and practical issues of development.4 The book has three parts. In Part I, contributors look at various foundational and measurement issues associated with economic development. Part II deals with the functioning (and non-functioning) of markets in the context of development. In the final part, Part III, contributors look at various issues related to governance and institutions.
Page 2 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction Joseph Stiglitz (Chapter 2) provides an overarching framework which explains the evolution of our thinking on markets, states, and institutions—from a singleminded focus on markets to a broader inclusion of institutions. Markets do not operate in a vacuum; they need well specified rules of the game, strong institutions including the state, and societal goods like trust. These institutions interact among themselves to create the right checks and balances. As we have seen, market economies, left untampered, lead to (p.3) forces which may threaten their own survival. Too often they lead to greater inequality and injustice, erosion of trust, and weakening of the state. The institutional checks and balances, so vital for the functioning of institutions and markets, may not work when there is excessive inequality. Prasanta Pattanaik and Yongsheng Xu (Chapter 3) discuss the conceptual foundations of the notion of freedom. Individual freedoms have received much attention in recent times both in welfare economics and in the context of development, following Sen’s capabilities and functioning approach (Sen, 1985). Developmental success has to be viewed not just in terms of what individuals achieve but also of what they are capable of achieving. However, as Pattanaik and Xu demonstrate, there are conceptual problems in arriving at a precise formulation of freedom. In fact, this is related to Basu’s earlier contribution (Basu, 1987) where he raises similar concerns and points out the difficulties that lie in measuring freedom. A related issue in the developmental context is the measurement of well-being. It is argued that well-being cannot be captured by income alone and we have to consider non-income attributes like health, education, mental well-being, and several others dealing with quality of life. This has obviously influenced the way we conceptualize and measure poverty—a shortfall from some stipulated level of well-being. While it is widely accepted that individuals’ shortfalls in these dimensions should be considered, it is not clear whether and how such information should be aggregated. A rich literature on multidimensional poverty measurement has sprung up recently following several key contributions.5 An individual is identified as poor or deprived if he/she falls short of these povertyline-like cut-offs along various dimensions.6 A key assumption is that these shortfalls are non-comparable; a shortfall in one dimension cannot be compensated by a gain in another. The chapter by Bhaskar Dutta (Chapter 4) reviews some of these axiomatic treatments and questions some of these assumptions. As he points out, there are other ways one can identify poor individuals even within the multidimensional framework. A similar measurement issue arises when we try to evaluate the well-being of society as a whole. A standard measure like per capita income or gross domestic product (GDP) may not be adequate as it does not capture distributional aspects of society which may be welfare relevant. Kaushik Basu (2001, 2006) suggested the use of the ‘quintile income statistic’, which is simply the mean income of the Page 3 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction bottom 20 per cent, as an indicator of the well-being of society. The chapter by S. Subramanian (Chapter 5) offers different (p.4) interpretations of this concept and explores its implications for the measurement of poverty and inequality. This concept lies at the heart of the more recent notion of ‘shared prosperity’ promoted by Basu as one of the main goals of the World Bank during his stint as its chief economist.7 The next chapter by Tapan Mitra (Chapter 6) deals with the deeper foundations of the notion of ‘sustainable development’. As he argues, sustainable development is not only recognizing the interdependence of resource use across generations, it is also about intertemporal social preferences which represent those of present and future generations. He shows how there is a fundamental difficulty in implementing such social preferences satisfying some notion of equity across present and future generations. This is related to his earlier work with Basu (Basu and Mitra, 2003). Despite its abstract nature, it has strong links with many real world problems dealing in intergenerational equity. The final contribution in this part is an empirical chapter by Ranjan Ray (Chapter 7) evaluating the role of relative price changes in welfare comparisons, more specifically on poverty and inequality.8 An interesting problem in welfare comparisons is the computation of a deflator to compute real incomes when people have different preferences leading to different consumption patterns. Using an average consumption basket (as is the practice) may not be ideal; Ray approaches this problem by assuming preference patterns to be homogenous within income quintiles within a region. He computes the intertemporal and spatial true cost of living index by estimating preference parameters for these region specific income quintiles through demand system estimation. Part II, Markets and Development, has seven contributions addressing various allocation and distributional issues in the context of different markets. Some of the essays analyse the market outcomes in general settings, while others focus on more specific markets. Several of these chapters seek to identify the role of policies or non-market mechanisms in different contexts. So in some sense, these essays go beyond the standard market mechanisms (to borrow and compare with Basu’s Beyond Invisible Hand [2011a]). Jörgen Weibull and Jun Chen (Chapter 8) look at private and public monopoly in a situation where the monopolist has to spend resources on finding out what product varieties individuals want.9 They analyse the monopolist’s endogenous information acquisition and choice of product variety in different (p.5) scenarios: a private monopoly maximizing profit and a public monopolist who maximizes a convex combination of profit and welfare under a budget constraint. They find that, broadly speaking, public monopoly is preferable in societies with a wide spread in income while private monopoly is better in societies with less
Page 4 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction inequity. This has clear implications for developing countries with greater inequalities. In Chapter 9, Hodaka Morita looks at an important issue in the labour market: variations in employment practices in terms of labour mobility, wage structure, and in-house training. Morita proposes an explanation based on ‘managerial capability’, using the interconnection between firm dynamics, labour mobility, and specific human capital to explain the different employment practices in the US and Japan. While most of the existing literature relies on information asymmetry regarding workers’ abilities, Morita’s model uses the strategic complementarity arising from the connection between continuous process improvement and firm-specificity of human capital. Complementarities and subsequent multiple equilibria have a central role in many of Basu’s theoretical contributions. Complementarity is also the key theme in the next chapter by Patrick Emerson (Chapter 10). He looks at the effect of complementarity that exists between goods and services in the quality space. He suggests that consumers tend to desire goods of the same quality and thus the price producers can charge for their good depends not only on its own quality, but also on the quality of other goods that will be used with it. This can lead to ‘quality traps’ where any one individual producer might not find it in its own interest to unilaterally increase quality; however, if all producers of the other complementary products increased quality at the same time it would be in all of their interests to do so. Lower levels of development can be associated with a low-quality equilibrium trap.10 Namrata Gulati and Tridip Ray, in Chapter 11, explore how neighbourhood effects interacting with income inequality affect poor people’s ability to access basic facilities like health-care services, schooling, and so on. They model the interaction by integrating consumers’ income distribution with the spatial distribution of their location11 and investigate the consequences of an increase in income inequality on the welfare of the poor in general, and their access to markets in particular. They find inverted U-shape relationships between income inequality and market access, and welfare of the poor owing to an (p.6) interesting trade-off between the positive provision effect and the negative price effect of the proportion of rich living in the neighbourhood. In a different but related context, John Ifcher, Homa Zarghamee, and Amanda Cabacungan (Chapter 12) study how subjective well-being of different age groups is affected by adverse economic shocks. They examine the subjective well-being effects of the 2007–09 recession on individuals aged 55 to 64 and find that individuals of pre-retirement age became significantly less satisfied with their lives compared to other age groups.12 This again calls for identification and differential treatment of various groups with differing vulnerabilities.
Page 5 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction Development discourses are incomplete without careful consideration of the states of children and women in these societies. In this context, both the marriage market and the market for child labour have attracted attention. A. V. Chari, Annemie Maertens, and Sinduja Srinivasan (Chapter 13) analyse the impact of changing income distribution on the marriage market. Using Indian data, they find that rising inequality at the top end leads to a delay in female marriage. Interestingly, this does not hold for rising inequality at the lower end of the distribution. The delay in female marriage has significant implications for poor countries, where early marriages and subsequent problems of health risks to both mother and child are serious issues. Basu’s work (Basu and Van, 1998 and Basu 1999) has spawned a massive literature on child labour and policies aimed at elimination of this practice. Garance Genicot, Anna Maria Mayda, and Mariapia Mendola (Chapter 14) analyse the impact of internal migration on child labour outcomes. As the authors point out, the impact of immigration on resident adults’ wages and employment has been studied at length, but how immigration affects child labour has not received proper attention. Child labour is more likely to be affected by immigration because of its low-skilled nature and higher elasticity of supply compared to adult labour. They find, using data from Brazil, that unskilled immigration has a significantly negative impact on child labour, but skilled immigration has the opposite effect. Part III has six chapters on governance and institutions. It is an area where Basu has written extensively in recent times. Governance has been a central concern in recent years, more so in the context of development. Avinash Dixit (Chapter 15) looks at relation-based governance and its potential implications. As he notes, relational governance works well when compliant behaviour (primary action as well as prescribed sanction against deviations) is well understood and more importantly, information regarding deviations from prescribed actions are quickly and accurately disseminated. For this reason, (p.7) relational governance works well in relatively small communities. Using a theoretical model, based on a spatial model of monopolistic competition, he explores the implications of relational governance. The requirement that firms should be close to each other for the effective spread of information limits their size and allows inefficient entry of too many firms, similar to the observation of too many firms in the informal sector of developing economies. Clive Bell (Chapter 16) looks at agrarian contracts involving land, labour, and credit from a historical perspective. Using primary household data from two villages in Bihar, India, in the year 1970, Bell examines the prevailing contractual relations in the absence of well-defined and functioning markets. These data, collected by the author himself, have never been the primary basis of any previously published work. It is interesting to note that many of these
Page 6 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction practices (as discussed by Bell in these villages), attracted the attention of development theorists in later years.13 Moving away from information or agrarian institutions, the next chapter by Nirvikar Singh (Chapter 17) analyses the role of institutions of federalism in India in managing heterogeneity and preserving national unity over the last five decades. His contribution reviews the conceptual and analytical underpinnings of the role of federal structures in sustaining unity, and summarizes historical developments and current institutional structures of the Indian case. Using a broad canvas, Singh portrays the political, legal, bureaucratic, and fiscal aspects of federalism in the Indian context. In a country as diverse as India, the role of these ‘holding together’ institutions can hardly be overestimated. The issue of ‘diversity’ is taken up in the next chapter by Ashwini Deshpande (Chapter 18) which argues the normative case for greater diversity in the workforce of private corporations in the specific context of caste disparities in India. Deshpande discusses in her essay how ensuring greater diversity, in addition to enabling social inclusion, would make good business sense. However, she finds that discrimination based on social identity manifests itself in different ways in different segments of the market depending on the size of the firm. For large firms, she finds evidence to support a positive association between profits and more diverse workforce teams. The same pattern is absent for small and micro-enterprises; such enterprises owned and populated by members of marginalized groups might face discrimination on account of their identity, adversely affecting their performance. We need institutions to promote diversity and unity, as has been argued in previous chapters. At the same time, there is a need for institutional structure protecting and promoting individual rights. In a related chapter, Eduardo Zambrano (Chapter 19) proposes a framework for evaluating policy options (p. 8) when these adopted policies are going to create winners and losers. He is concerned about the ranking of social states where priority (in the social evaluation) is given to the situation of those individuals who are being treated the most unfairly by policies, relative to what they have a right to. He goes on to examine how this is implemented in the more specific context of tax reforms. Combining liberal and libertarian principles of fairness, together with Pareto efficiency, he analyses the distributive impacts of taxation in the economy. In the final governance related essay, Ajit Mishra (Chapter 20) looks at the possible trade-offs that anti-corruption policy faces. As is well known, corruption has several faces, ranging from petty corruption where ordinary citizens have to pay bribes to get goods and services, to high level political corruption.14 It has been argued that attempts to control one form may promote another form of corruption. In such a context, petty corruption tends to be ignored. He argues that, despite its lower value and gift-like manifestation, petty corruption should Page 7 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction not be ignored as it creates a ‘culture’ of corruption and raises the tolerance levels. Finally, we would like to thank all the contributors for their help and patience in meeting our highly irregular demands. Our inexperience, if any justification is to be offered, can be blamed! We would like to thank Adam Swallow of Oxford University Press (OUP) for his advice and support throughout the entire process. Our special thanks are due to Alaka Basu, Karna Basu, Lui-Felipe Lopez Calva, Mausumi Das, Indranil Dutta, Shasi Nandeibam, Lucy O’Shea, Bharat Ramaswami, Anuradha Saha, Kunal Sen, Erik Thorbecke, and several other colleagues who have provided encouragement and guidance. Additionally, we wish to thank colleagues at our respective institutions and at Ashoka University, which proved to be an excellent sanctuary during the final stages of preparation of the manuscript. References Bibliography references: Alkire, S. and J. Foster (2011), ‘Counting and multidimensional poverty measurement’. Journal of Public Economics 95(7–8): 476–87. Atkinson, A. B. (2003), ‘Multidimensional deprivation: contrasting social welfare and counting approaches’. Journal of Economic Inequality 1(1): 51–65. Basu, K. (1984), The Less Developed Economy: A Critique of Contemporary Theory. Oxford and New York: Basil Blackwell. (p.9) Basu, K. (1987), ‘Achievements, capabilities and the concept of wellbeing’. Social Choice and Welfare 4: 69–76. Basu, K. (1988), ‘Why monopolists prefer to make their goods less durable’. Economica, 55(22): 541–6. Basu, K. (1990), Agrarian Structure and Economic Development. London: Harwood Academic Publishers. Basu, K. (1993), Lectures in Industrial Organization Theory. Oxford and Cambridge, MA: Basil Blackwell. Basu, K. (1997), Analytical Development Economics. Cambridge, MA: MIT Press. Basu, K. (1999), ‘Child labor: cause, consequence and cure, with remarks on international labor standards’. Journal of Economic Literature 37: 1083–119. Basu, K. (2001), ‘On the goals of development’, in G. M. Meier and J. E. Stiglitz (eds) Frontiers of Development Economics: The Future in Perspective. New York: Oxford University Press, pp. 61–86. Page 8 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction Basu, K. (2005), Collected Papers in Theoretical Economics, Vol. I–IV, New Delhi: Oxford University Press. Basu, K. (2006), ‘Globalization, poverty, and inequality: what is the relationship? what can be done?’ World Development 34(8): 1361–73. Basu, K. (2011a), Beyond the Invisible Hand: Groundwork for New Economics Princeton NJ: Princeton University Press. Basu, K. (2011b), ‘Why, for a certain class of bribes, the act of giving a bribe should be legal’. Working Paper. Government of India: Ministry of Finance. Basu, K. (2013), ‘Shared prosperity and the mitigation of poverty: in practice and in precept’. World Bank Research Working Paper 670. Washington, DC: World Bank. Basu, K. and Mitra, T. (2003), ‘Aggregating infinite utility streams with intergenerational equity: the impossibility of being Paretian’. Econometrica 71(5): 1557–63. Basu, K. and T. Mitra (2016), ‘Nash on a rotary: two theorems with implications for electoral politics’. Policy Research Working Paper No. 7701. Washington, DC: World Bank. Basu, K. and Van, P. (1998), ‘The economics of child labor’. American Economic Review 88(3): 412–27. Bourguignon, F. and Chakravarty, S. R. (2003), ‘The measurement of multidimensional poverty’.Journal of Economic Inequality 1(1): 25–49. Kremer, Michael. (1993), ‘The O-ring theory of economic development’. Quarterly Journal of Economics 58(August): 551–75. Sen, A. (1976), ‘Real national income’. Review of Economic Studies 43(1): 19–39. Sen, A. (1985), Commodities and Capabilities. Amsterdam: North-Holland. (p. 10) Notes:
(1) He used to take groups of students from his development class in the Delhi School of Economics for ‘field surveys’ in Hazaribagh district and both of us were fortunate to be part of such a trip. It was a unique opportunity for students to learn about the real issues in development. After a whole day’s work in the neighbouring village, we would return to the base camp and, as if he had had enough of the real world, in the evening Basu would initiate discussions and debate on some really deeper philosophical issues in economics!
Page 9 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction (2) This was eloquently communicated by him during his acceptance speech at the University of Bath, which honoured him with the degree of Doctor of Laws, 2016. (3) The Centre for Development Economics at the Delhi School of Economics was his brainchild and he was the first Executive Director. One of the editors has a personal account of the huge amount of effort that went into the establishment of this centre. (4) We have been mindful of his strong interests in abstract analysis and geometry; as a result, this volume has more figures than tables! (5) See, for example, Atkinson (2003), Bourguignon and Chakravarty (2003), Alkire and Foster (2011). (6) The actual number of dimensions depends on the precise formulation. In the union approach, an individual is identified as poor if he/she is deprived in any dimension. In the intersection approach, he/she is identified as poor only if he/ she is deprived in all dimensions. (7) It refers to the maximization of the mean income of the bottom 40 per cent of the population as an important goal of the society. See Basu (2013). (8) That this is related to Chapter 5 is no coincidence; both strands draw inspiration from Sen (1976). (9) In an earlier contribution, Basu (1988) explains why a monopolist would produce a less durable good even when it can improve durability at no cost. (10) This takes inspiration from earlier work by Kremer (1993) and several others. This has been used to explain underdevelopment in different dimensions; see Basu (1997). (11) Kaushik Basu is quite fond of the spatial model of competition (see, for example, Basu, 1993: Chap. 8; Basu and Mitra, 2016). (12) The authors draw attention to the significant boost in life satisfaction associated with those who enter the 65+ cohort—a club which Kaushik Basu joined this year! (13) Basu himself has contributed immensely to this literature on contractual practices; see his book Agrarian Structure and Economic Development (1990). (14) Failure to recognize this has created misconceptions and unnecessary debates surrounding anti-corruption reforms, as was evident from the popular debate which followed Basu’s suggestion that for certain kinds of bribes, bribepaying may be legalized (Basu, 2011b). Page 10 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Introduction
Access brought to you by:
Page 11 of 11
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Markets, States, and Institutions Joseph E. Stiglitz
DOI:10.1093/oso/9780198812555.003.0002
Abstract and Keywords The chapter describes both the evolution of thinking—from a single-minded focus on markets to a broader inclusion of institutions—and how even that broader perspective is excessively narrow and, too often, insufficiently nuanced. Markets do not operate in a vacuum; they need strong institutions including the state, interacting among themselves to create the right checks and balances. Market economies, left untampered, lead to forces that threaten their own survival. Too often they lead to greater inequality and injustice; erosion of trust; and weakening of the state. The institutional checks and balances that are so vital for the functioning of institutions and markets do not work when there is excessive inequality. Keywords: markets, institutions, government, checks and balances, inequality
It is a real pleasure for me to participate in this festschrift in honour of my good friend Kaushik Basu, to recognize his intellectual contributions, his contributions as a public intellectual, and his contributions as both a national and global public servant. The subject of my chapter is one to which Kaushik has made profound contributions: markets, states, and institutions. In particular, I want to highlight how our thinking about this subject has changed over the past third of a century; and to provide an overarching framework into which these changes can be placed—a framework that both helps explain why the approaches taken in the
Page 1 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions past have been less successful than was hoped in promoting development, and provides some guidance for policy reforms and research going forward. Earlier work both at the World Bank and within the development community more generally focused on necessary reforms to policy frameworks. These ‘reforms’—the now infamous Washington Consensus policies—mostly consisted of giving a larger role to markets in the allocation of resources. When these reforms were less successful than hoped, there was a switch to a focus on institutions, including those of the public sector. It was recognized that the policy reforms had to be instituted by governments, and that governments often failed to do what was required. Thus, even if the overall agenda was to place a greater emphasis on markets, to accomplish that end one needed reforms in at least one key institution—the government—to bring that about. There was a second rationale for a focus on institutions—there were pervasive market failures, and a hope that non-market institutions, on their own, would ‘step in’ to fill the gap. This belief was not based on any deep theory, but rather on the notion that with a market failure (say the absence of an insurance (p.14) market) there was an opportunity for a Pareto-improving non-market action. A strong Hayekian belief in decentralized evolution suggested that such evolution would lead society to higher and higher levels of well-being—especially to Pareto improvements. These beliefs were reflected, for instance, in the idea that nonmarket life insurance, say provided by the family or burial societies, would be adequate to address market deficiencies. No government intervention would be needed. This particular line of research, sometimes associated with Douglas North’s early work, was laid to rest when Arnott and Stiglitz (1991) showed that the Nash equilibrium with non-market institutions could be worse than without these institutions. There is an incentive for such institutions to be created, but they may actually displace the admittedly imperfect markets, in such a way as to lower welfare.1 Both markets and states are, of course, institutions—institutions through which we allocate resources. It used to be argued that, in thinking about the best way of organizing societal systems of resource allocation, one assessed in which sectors the market should dominate, and in which sectors the state. The perspective was that fully private goods should be produced by the private sector; those associated with the delivery of public goods should be produced by the state. Today we see the interaction in a more complicated way: in many cases, the two interact, in a complementary way, within the same sectors. For instance, there is the possibility of the separation of finance from production; government could provide finance for a typically publicly provided service, like education, but the production would be done through private enterprises. In the provision of Page 2 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions infrastructure, there has been great interest in public–private partnerships (PPPs). In the financial (and other sectors) which might have seemed to fall naturally within the private sector, there is an important role for government regulation. And in some areas, government has had to do more: underwriting mortgages, providing finance for small businesses, and—especially in many developing countries—providing long-term finance. While in recent years there has been a great deal of hyperbole over PPPs, in practice, there has been disappointment. PPPs often entail the government taking the risk, while the private sector takes the profits. So too, the conditions under which government can delegate to a private body the fulfilment of public objectives have been shown to be extraordinarily restrictive (Sappington and Stiglitz, 1987). (p.15) The standard argument for introducing a role for government began with the theory of market failures—the work of Arrow and Debreu identified a large variety of circumstances in which private markets do not lead to (Pareto) efficient outcomes. Subsequent work by Greenwald and Stiglitz (1986) showed that whenever information was incomplete (asymmetric) or markets incomplete —that is, always—markets were not efficient. The presumption that markets were efficient, which had reigned since Adam Smith, was reversed: the presumption now was that markets were inefficient. There was always a potential role for government. But while there was a potential role for government, it was not always obvious that government could fulfil this role. Attention shifted to government failure. While the theory of government failure is not as well-developed as that of market failure, it is clear not only that governments often fail, but also that such failures are not inevitable: even imperfect governments can result in an improvement in resource allocation. They can help markets work better. Indeed, it is hard to find any country that has had successful development in the absence of strong government interventions. But as our understanding of government failures has increased in recent years, so too has our grasp of the depth of market failures—highlighted by the financial crisis of 2008. More importantly, we have come to appreciate markets as institutions that must be structured. Markets do not exist in a vacuum. They are structured by public policy, by the rules of the game that are set by the government, for instance through laws that relate to corporate governance, competition policy, and labour market regulation.
Page 3 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions These then are the central messages of this chapter: (a) In any society, resource allocations occur within institutions, so that the rules governing the institution are critical, particularly the rules determining how decisions are made within it. Institutions consist of multiple individuals, with differences in preferences and beliefs. A critical issue is how these are ‘aggregated’, so that the institution reflects in some adequate way those within the institution. This was the central question posed by Arrow in Social Choice and Individual Values (1951). His results were deeply disturbing, for he showed that there was no way of aggregating the multiple preference orderings of the different individuals comprising an institution that had certain desirable properties (like transitivity), in the absence of some restrictions on preferences and/ or the choice set—other than dictatorship, where the actions chosen were those that reflected the preferences and beliefs of a single member. This negative result poses one of the great challenges for governance. (p.16) (b) Societal resource allocations are the result of the interaction among these institutions. In recent years, economists have given a great deal of attention to ‘mechanism design’; that is, to the design of allocation mechanisms with certain desirable (usually efficiency) properties. But the set of institutions in place in any economy is not the result of rational deliberation over alternative mechanisms (even if our politicians understood what that entailed). Rather, they have evolved, with adjustment of one set of institutions or another in response to changes in the world and changes in ideas, including learning from past successes and failures. As a result, there is no presumption that, in any country, the existing set of institutions or the rules governing their interactions are optimal in any sense, that they produce either efficient or equitable outcomes. A key concept in institutional design has been ‘checks and balances’: a recognition that within an institutional arrangement (say government), there is the danger of the aggrandizement of power in the hands of a subset of individuals, or even a single individual, resulting in decisions that reflect that individual’s or those individuals’ perceptions or interests. At the societal level, the same issues arise: we should see different institutions as providing checks and balances on each other. (c) The functioning of markets (both the decisions made by individual institutions and the outcomes of the interactions among the institutions) depends on the rules of the game specified by the political process, which in turn depends on the rules of the political game and underlying characteristics of society, most importantly, the magnitude of economic inequality and the degree of solidarity and political cohesion. But the functioning of markets also depends on trust. No economy can rely on the enforcement of contracts through the legal system. Trust, especially as it relates to the functioning of market institutions, depends in part on Page 4 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions perceptions of the legitimacy of the economic and political system, which in turn depends on perceptions of fairness and equity. In short, the functioning of the market depends on non-market institutions and beliefs and perceptions that reach beyond the market. By focusing too narrowly on markets, by creating markets that are seemingly disjointed from the rest of society, by taking excessively tolerant views of market abuses (of the kind that became rampant in financial markets before and during the 2008 crisis), market advocates may have actually undermined the success of markets.2 (p.17) (d) If a system of checks and balances among institutions within society is to work—to ensure that societal resource allocations do not come to reflect the interests and beliefs of a certain subset of individuals —then there cannot exist excessive economic inequality. For if there is excessive economic inequality, there is at least a risk that this economic inequality will get reflected in political inequality—in inequality in key public institutions. The voice of the wealthy will predominate both public choices (public allocations of resources) and in the setting of the rules of the game. In short, the emphasis of the World Bank and development economists more generally on the governance of public institutions is correct, but good governance is, in part at least, an endogenous variable. Lectures about good governance won’t succeed if the conditions for good governance aren’t there. Policy discourse should focus not just on what is entailed by good governance (e.g. transparency and accountability) but also on the conditions necessary to create and sustain good governance, e.g. reforms in economic policies that lead to greater equality both in market incomes and in income and wealth after taxes and transfers. (e) Everyone benefits from the good performance of the public sector— including having the rules of the economic game written in ways that support efficient and equitable outcomes. But since the public good is a public good, there will be an under-supply of efforts at maintaining good public governance, making it particularly easy for interest groups to capture the state. The rules of the game for the public sector have to recognize this and guard against it. We will discuss in Section 3 what this entails.
1 New Understandings of Markets Since the development of the Walrasian economic model,3 a particular view of the market economy has prevailed. It entails simplistic firms that maximize profits (or stock market value in a dynamic context), and households consisting of unitary actors, with households and firms interacting in competitive markets through a price mechanism. Economists celebrated the informational efficiency of prices, the ability of prices to provide requisite information from households to firms and vice versa: firms don’t have to have knowledge (p.18) of other firms’ technology or of households’ preferences, and similarly, households don’t Page 5 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions have to have any knowledge of technology or resource availability. It was all quite miraculous. This view of the market economy relied on three critical assumptions that are worth noting for the purposes of this discussion: (a) In each institution (treating firms and households as institutions), there was no problem of preference aggregation. Indeed, the issue was not even recognized; (b) Each institution faithfully carried out what was agreed—there was no problem of contract enforcement; (c) All markets were competitive—no one had market power. There were other assumptions, such as those relating to information, which are critical to the results concerning the efficiency of the resulting resource allocations. Advances in recent decades have shown the central role of oversimplistic information assumptions employed in standard analyses. For instance, even a small amount of information imperfection could generate a high level of market power, both within an institution (like a household or a firm; see for example Edlin and Stiglitz, 1995) or across institutions (monopoly and monopsony power; see Diamond, 1971; Stiglitz, 1985b, 2009, 2013). The absence of good information provides opportunities for one group to exploit others. Most fundamentally, as we have already noted, with even slight imperfections and asymmetries of information, there is a presumption that the market economy— even if it were competitive—is not efficient. There was, in the traditional analyses of a market economy, no discussion of institutions and institutional interactions simply because institutions didn’t matter. They were ‘superficial’. We saw the world through institutions. In agriculture there was the institution of sharecropping; in finance, the institution of banks. But economics looked deeper, beneath the surface, to the underlying economic forces—the laws of supply and demand. Simply by studying these, one could understand resource allocations (including distribution of income). In the past forty years, the foundations of all of these assumptions underlying the standard model have been challenged—and so too the belief that institutions didn’t matter. 1.1 Aggregation
One example of a challenge to these assumptions involves preference aggregation. The family is now seen as consisting of several members, with often conflicting preferences and beliefs. Household resource allocations cannot be described as if the family maximized some family social welfare function. It is better described as the result of a complex bargaining situation. This is important, because there are changes, which in standard theory would make no difference, but which might affect the bargaining power of each member (p.19) and which affect the resulting resource allocations. One of the reasons for the success of the Grameen microcredit programmes is that they changed the power Page 6 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions relationships within the household; so too for their mortgage programmes, which required the transfer of title to women, and which affected incentives for divorce. Kaushik’s pioneering work (Basu and Van 1998; Basu 1999) on child labour shows how changing the rules—not allowing child labour—can have general equilibrium effects, which are welfare enhancing.4 In the case of the firm, it has been shown that the conditions under which there is unanimity about what the firm should do are very restrictive.5 There has to be a full set of Arrow–Debreu securities.6 The modern theory of the firm (Berle and Means 1932) has emphasized that there are multiple stakeholders in the firm, and that managers do not necessarily and in general do not fully represent the interests of any group other than themselves. Because of imperfections of information, there has to be delegation of decision-making authority (Stiglitz 1985a), and managers will inevitably be decisive. Importantly for our purposes, society has not left the working-out of the power relationships within either the firm or the household to the ‘market’. Government sets the rules of the game, through family law and through corporate governance. Corporate governance restricts the power of managers, even as managers try to do what they can to increase their discretionary power (see, for example, Edlin and Stiglitz, 1995). And, of course, corporate executives fight hard for legal frameworks that give them more discretion. For instance, they fought hard against initiatives of ‘say in pay’. giving shareholders some say in the pay of the managers who are supposedly working for them (even when shareholder votes were non-binding). They suggested that the passage of say in pay would have a destructive effect on the functioning of the market economy: it would potentially curb their ability to extract rents from the corporation. 1.2 Contract Enforcement
Contract enforcement is of particular importance in the context of intertemporal contracts.7 Again, the role of government is crucial: the government sets (p.20) the rules determining how contracts will be enforced, and even what kinds of contracts can be enforced. Individuals cannot sell themselves into slavery, but student debt can essentially never be forgiven, with lenders being able to garnish 25 per cent of a worker’s wages for his/her entire life. When the costs of contract enforcement are very high, de facto it is as if they are only enforced through reputation mechanisms. 1.3 Competition and Power
The strength of the underlying competitive paradigm in explaining resource allocation is so strong and pervasive that we forget that a country like the United States developed on the basis of slavery—the involuntary provision of Page 7 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions labour. For long periods of time, various forms of feudalism, restricting contracts that individuals or classes of individuals can undertake, persisted around large parts of the world. And even today, such restrictions effectively exist in some places. In many places, to increase the market power of one group, there were restrictions on what others could do (e.g. land ownership in apartheid South Africa). In some cases, the deviations from competition were enforced through what might be viewed as a market mechanism as part of a repeated game. It is easy to see how ‘cooperative’ behaviour among one group can be used to exclude others. Social capital may strengthen the functioning of society, but it can and has been used to enforce power relationships (see Dasgupta and Serageldin, 2000; Dasgupta, 2005, 2012). In many cases, government actions were pivotal. In some cases, as in the case of racial discrimination in the United States, the two interacted. This then is the fundamental dilemma: the government is often complicitous in one group’s exploitation of another. But the government is the only means to ‘tame the jungle’—to prevent powerful groups from exploiting others.
2 Public Governance The previous discussion should have made it clear that the notion of a market economy without government intervention is a phantasm. Without government— one way by which society sets the rules of the game—there is a jungle. Power triumphs—until it is overcome by some stronger power. In recent decades, economists have focused on the need for collective action. Society is better off if or when it acts collectively—through the provision of public goods, proscribing activities that give rise to negative externalities, and encouraging those that give rise to positive externalities. There can be Pareto improvements. But the most important arena for collective action is (p.21) the establishment of the rules of the game, enabling a market economy to function, enforcing contracts, and preventing the abuse of power, whether within an institution (e.g. through corporate governance rules) or within society (through antitrust laws). I need to emphasize: these rules relate to both equity and efficiency—an obvious observation in the context of competition policy. Without such policies, there is a tendency in market economies for the growth of market power; and that results both in Pareto-inefficient allocations and in distributions of income in which those with market power gain at the expense of the rest. In the standard economic model, the importance of these rules of the game was given short shrift. If the assumptions of the standard model were always satisfied, then these rules of the game might matter little. Unfortunately, as we have already noted, both for the economics profession and our society, those Page 8 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions assumptions do not hold, and the formulation of economic policies on the belief that they do has had sometimes disastrous effects. To be fair, much of the policy advice (especially in the context of development) over the past half century recognized that markets often didn’t work well— though typically the onus was placed on government, blaming it for intervening in one way or another. The presumption was that if only government got out of the way, we would wind up in a world well described by the competitive ideal. That idea was, of course, absurd: in the absence of government intervention, markets do not gravitate towards the ideal, but rather in the opposite direction. The reason that countries have enacted competition laws is precisely that in the absence of government action, there is a tendency for excessive market concentration. Historically, many government interventions have arisen out of public demands seeing massive market failures: interventionist (Keynesian) macro-policies from the excessive volatility of market economies, with persistent high unemployment; social security from the failure of the private sector to provide annuities at reasonable transaction costs; unemployment insurance and other social insurance from the failure of the market to provide adequate risk mitigation instruments, etc. Historically, many government interventions have arisen from public demands after massive market failures: interventionist (Keynesian) macro-policies in response to the excessive volatility of market economies, with persistent high unemployment; social security in response to the failure of the private sector to provide annuities at reasonable transaction costs; unemployment insurance and other social insurances in response to the failure of the market to provide adequate risk mitigation instruments, etc. But here, standard economics often makes two other mistakes: (a) assuming that so long as market imperfections are not too large, the economy can be well described by the competitive equilibrium model; and (b) ignoring the theory of the second-best. (p.22) Rothschild and Stiglitz (1976) and Diamond (1971) laid to rest the first idea: even arbitrarily small information imperfections can have very large effects on the nature of market equilibrium. Policy analysts understood that it was impossible to achieve anything like the ideal world envisioned in the competitive equilibrium model. There would be information imperfections and incomplete markets. These were inherent market failures—not in any way related to government actions. But they assumed that moving towards that ideal would lead to better outcomes—and if we got close enough, we would achieve something like the ideal results envisaged by Adam Smith. Not only was there no support for this conclusion, there was also a strong theoretical literature (beginning with Meade, 1955 and Lipsey and Lancaster, 1956) showing that these results were not true. Later work showed that free trade—in the absence of good risk markets—could make everyone worse off Page 9 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions (Newbery and Stiglitz, 1984). Capital market liberalization could lead to more economic volatility (a theoretical proposition—see Stiglitz, 2008) supported now by a wealth of empirical evidence (Rodrik, 1998). Before the 2008 financial crisis, there was an agenda called ‘completing markets’—creating new structured financial products, arguably trying to get closer to the Arrow–Debreu complete set of risk markets. But more recent theoretical analyses have explained how this actually contributes to economic volatility (Guzman and Stiglitz, 2016a, 2016b), and there is now a consensus that these products were an important factor in giving rise to the crisis (FCIC, 2011). Thus, the standard competitive equilibrium model is not the ‘right’ model for thinking about much of what goes on in the economy. Not surprisingly, it is particularly unsuited for reaching an understanding either of recent macroeconomic volatility or of the large increase in inequality. But it is not even the best model for thinking about the slowdown in economic growth, for explaining the growth of short-termism in the economy and its increasing financialization. These changes are not the result of enhanced understandings of economics, leading to improved strategies by firms or improved policies by government. An important development in economics in recent decades has been game theory—predicated on the belief that what matters is strategic interactions of a kind that simply don’t exist in the standard competitive model. An important insight of game theory is that the rules of the game matter. Market participants know this. There is thus a metagame over the determination of the rules of the game. The rules of the game are set by the state, by government. Much of the fight today is over the determination of the rules governing labour—with corporations attempting to eviscerate the power of unions; competition (p.23) policy—with big behemoths arguing that their anti-competitive behaviour is really efficiencyenhancing and in the interests of ordinary consumers; and corporate governance —with those controlling large corporations trying to ensure that they have freer rein in doing what they want to do, including seizing for themselves a larger share of corporate revenues. In Rewriting the Rules (Stiglitz et al., 2015), we argued that beginning around a third of a century ago, the rules of American capitalism (and those of much of the rest of the advanced world) were rewritten, changed in ways that favoured the powerful at the expense of the rest. The liberalization agenda was actually a ‘special interest’ agenda, allowing, for instance, those in the financial sector to reap huge rewards from excessive risk taking, with the downside risks being borne by the public.
Page 10 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions The adverse effects of these rewritten rules were even greater because they led not only to more inequality, but also to lower growth, as they encouraged firms to focus on short-term financial returns, and to use their scarce capital for purposes other than investment in productivity enhancement.
3 Reducing the Likelihood of State Capture The central issue, as we noted in the introduction, is that, while good governance is essential for a well-functioning society, ensuring the public good— which a good set of rules does—is itself a public good, from which all benefit. There is always an undersupply of public goods (on their own), and this includes efforts at maintaining good public governance. By contrast, there are ample incentives to subvert good governance, resulting in efforts at rent seeking and state capture—using the power of the state, including its powers of compulsion, to advance particular interests. There is no easy or simple resolution to this problem. Some countries have done a reasonably good job of ensuring that the state advances the public interest; some have failed. Out of this wealth of experience—backed by a modicum of theoretical analyses—there are some precepts that may be useful. First, transparency is essential. Transparency in the public sphere is what good information is in the private sphere: without transparency, it is easy for special interests to divert state resources for their own purposes. Transparency includes the right to know (see e.g. Florini, 2007 and Stiglitz, 1999)—to know what the government is doing; and the right to disseminate (i.e. a free press)—the right to tell (Islam 2002). But transparency also entails more: the provision of information is itself a public good, so that there will be an undersupply in the absence of government action to correct for this market failure (see Stiglitz 2002). (p.24) Transparency is often linked with accountability: there have to be consequences for one’s actions. In the absence of transparency, there cannot be accountability. The problem though is that the relevant outcomes (societal outcomes, or even more limited outcomes, say in the education sector) are the results of actions taken by multiple individuals, and it is typically impossible to parse out the (marginal) contribution of any single individual. And in many areas, the lags between the actions and the consequences may be large, with many intervening events. Among the most important aspects of the design of the public sector is a system of checks and balances, to reduce the risk of capture—with checks and balances, state capture requires a hold over multiple branches of government. But while a system of checks and balances makes capture more difficult, special interests have still managed to overcome the obstacle.
Page 11 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions Thus, the system of checks and balances has (so far) prevented one branch of the government dominating over another; but it has not prevented powerful groups from capturing the entire government, or to put it more mildly, from exercising disproportionate influence, of a kind inconsistent with democratic values. This failure can be traced to the failure of a broader set of checks and balances—within our society. As inequality grew in the United States during its gilded age, it became increasingly clear that excessive income and wealth inequality would lead to excessive political inequality. The reforms of the Progressive Era, including antitrust measures, were motivated by an understanding of the political process more than by insights from competitive market analyses. Their architects realized that ordinary sensibilities about what democracy and the principle of one-person-one vote mean were undermined in societies in which there is excessive inequality.8 It was apparent that economic inequality was being translated into political inequality. The United States shows the dangers of economic inequality getting translated into political inequality, as the Republican Party (disproportionately representing those at the top)9 has in many states engaged in a strategy of disenfranchisement and disempowerment, making it more difficult for those at the bottom to vote and more likely that if they do vote, their vote won’t matter. They have openly engaged in gerrymandering. The Republican-appointed justices of the Supreme Court, in its infamous Citizens United decision in 2010, seemingly argued that money was not corrupting the political process. Elsewhere, I argued that these and other changes in America’s political processes meant that it might better be described as a democracy with ‘one dollar, one vote’ rather than ‘one person, one vote’ (Stiglitz, 2012). (p.25) Because the public good is a public good, society ought to do what it can to promote civic engagement in promotion of the public good. We now recognize the central role of civil society—groups within society getting together, voluntarily, to advance their conception of the public interest.10 That is, collective action occurs not just through national governments, but through a host of institutional arrangements, some government (local governments) and some non-governmental. The government can (and should) subsidize and encourage, in a variety of ways, these organizations, because they are the means by which the voice of various groups within society can get injected into the political process. This is one of the ways in which thinking about development (markets, states, and institutions) has changed: as we noted in the introduction, we used to view society in dichotomous terms—states or markets (sometimes emphasizing their complementarity). But there are a host of other institutional arrangements and players—the most successful institutions in the United States are arguably notfor-profit educational institutions. In fact, for-profit universities are among the Page 12 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions least successful institutions. Even in the United States, in many areas, cooperatives play an important role (credit unions and agriculture cooperatives are two examples).
4 Concluding Remarks This essay is about both economic policy and economic methodology. I have argued that the standard workhorse model of economics, the competitive equilibrium model, provides a poor description either of advanced or developing economies, and policy frameworks based on that model have proved less effective than hoped. In one of his last acts as chief economist of the World Bank, Kaushik brought together past chief economists as well as other development experts to see whether there was a post-Washington Consensus consensus. There was— articulated in the Stockholm Consensus. The theoretical models underlying that new consensus go well beyond the standard competitive equilibrium model. So too, the policy advice goes beyond ‘improve markets’ and ‘increase resources’. I have argued that the pervasive imperfections of competition imply the relevance of game theory. And the constant changing in the economy—both in response to new ideas and new technologies—suggests the relevance of ideas borrowed from evolutionary theory. But it is evolution without teleology, without any notion that we are necessarily moving in directions that make everyone in society better off. Indeed, in some cases, societies can get (p.26) caught in lowlevel equilibrium traps; in others, some gain, but at the expense of others.11 This chapter argues that economists should strive to base their policy advice on a broader set of models—of understandings that go beyond economics narrowly defined. But it also argues that much of the standard policy advice is not based on a deep grasp of economics. Standard policy has failed to understand the implications of the theory of the second-best, the lack of robustness of the standard model—where small deviations from idealized assumptions have great effects—and has not really acknowledged that markets need to be structured. The standard model portrays the market economy as a fine-tuned machine. Economists’ job is only to keep it well oiled, and more importantly, to stay out of the way—to make sure that government intervention didn’t muck up the works. But upon closer examination, it is clear that market economies, unless tempered, create a dynamic that may not be consistent with their own (successful) survival: a selfishness, which breeds inequities and injustices; a lack of trust and dishonesty, which undermines the functioning of markets themselves; and a weakening of the state, which makes it unable to govern the market and to make investments that can sustain learning. Can we have cheap labour and wellheeled consumers? For a time, perhaps, but recent history suggests there will eventually be an unravelling.
Page 13 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions The rewriting of the rules of the market economy a third of a century ago, to advance the interests of the wealthiest, impaired the functioning of the market. This led to slower growth and more instability—including the largest economic crisis in three quarters of a century. Markets can be self-destructive. The market needs to be saved from itself. But the political process that might do this has increasingly been captured by the wealthy in society. The United States prided itself in creating a system of checks and balances between branches of government. While the dangers of gridlock may not have been fully anticipated, many in the elites may find the dysfunction to their liking: a political system too weak to stop their exploitation of others, too weak to even impose a tax system with a modicum of progressivity. (In the United States, unlike almost any other advanced country, the very rich actually pay quite a low effective tax rate.)12 (p.27) But in a deeper sense, in terms of the functioning of society and the political system as a whole, there is an absence of checks and balances—no way, short of a wholesale recommitment to an agenda of greater equality, of preventing those at the top from continuing their aggrandizement of power; no way to prevent the concentration of economic and political power; no way to ensure a true democracy even in the market place of ideas. In the introduction, we noted that markets do not exist in a vacuum. They have to be structured, and they have to be seen in the context of the richer ecology of institutions within our society; so too for non-market institutions, most importantly the state. Kaushik is one of the few economists who has seen markets, the state, and institutions within this holistic framework, and who has demonstrated the ability to use models in the way they should be. Simple models can provide important insights, such as his path-breaking work on child labour. One needs precisely the right degree of complexity—to capture that which is relevant, excluding the extraneous; and the right degree of simplicity—so that one can thoroughly understand what is going on. But too often, economists who have proven their mettle in model formulation and analytics lose the ability to exercise judgement when it comes to policy. They do not understand the limitations of each model and how to blend the insights of various models together coherently, with the whole being greater than the sum of the parts. It is because Kaushik was able to combine these deep analytic skills with this superb judgement that he has been such an outstanding public servant. Acknowledgements I wish to acknowledge the multiple discussions with Kaushik Basu on the topics of this chapter. I also want to acknowledge conversations with Ravi Kanbur, Karla Hoff, Roumeen Islam, and many others. Some aspects of the issues Page 14 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions discussed here were analysed in Stiglitz (2016). I am indebted to Matthieu Teachout and Eamon Kircher-Allen for research and editorial assistance. References Bibliography references: Arnott, R. and J. E. Stiglitz (1991), ‘Moral hazard and non-market institutions: dysfunctional crowding out or peer monitoring’. American Economic Review 81(1): 179–90. [Reprinted in Selected Works of Joseph E. Stiglitz, Volume I: Information and Economic Analysis. Oxford: Oxford University Press, 2009, pp. 589–603.] Arrow, K. J. (1951), Social Choice and Individual Values. New York: John Wiley & Sons. (p.28) Basu, K. (1999), ‘Child labor: cause, consequence and cure, with remarks on international labor standards’. Journal of Economic Literature 37(3): 1083–119. Basu, K. and P. H. Van (1998), ‘The economics of child labor’. American Economic Review 88(3): 412–27. Berle, A. A. Jr and G. C. Means (1932), The Modern Corporation and Private Property. New York: Macmillan. Dasgupta, P. (2005), ‘The economics of social capital’. Economic Record 81(Issue Supplement S1): S2–S21. Dasgupta, P. (2012), ‘Dark matters: exploitation as cooperation’. Journal of Theoretical Biology 299: 180–7. Dasgupta, P. and I. Serageldin (eds) (2000), Social Capital: A Multifaceted Perspective. Washington: World Bank. Diamond, P. A. (1971), ‘A model of price adjustment’. Journal of Economic Theory 3: 156–68. Eaton, J. and M. Gersovitz (1981), ‘Debt with potential repudiation theoretical and empirical analysis’. Review of Economic Studies 48(2): 289–309. Eaton, J., M. Gersovitz, and J. E. Stiglitz (1986), ‘The pure theory of country risk’. European Economic Review 30(3): 481–513. Edlin, A. and J. E. Stiglitz (1995), ‘Discouraging rivals: managerial rent-seeking and economic inefficiencies’. American Economic Review 85(5): 1301–12.
Page 15 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions FCIC (Financial Crisis Inquiry Commission) (2011), The Financial Crisis Inquiry Report: Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States. Washington, DC: Government Printing Office, January. Florini, A. (ed.) (2007), The Right to Know: Transparency for an Open World. New York: Columbia University Press. Greenwald, B. and J. E. Stiglitz (1986), ‘Externalities in economies with imperfect information and incomplete markets’. Quarterly Journal of Economics 101(2): 229–64. Grossman, S. and J. E. Stiglitz (1977), ‘On value maximization and alternative objectives of the firm’. Journal of Finance 32(2), May, pp. 389–402. Grossman, S. and J. E. Stiglitz (1980), ‘Stockholder unanimity in the making of production and financial decisions’. Quarterly Journal of Economics 94(3): 543– 66. Guzman, M. and J. E. Stiglitz (2016a), Pseudo-Wealth and Consumption Fluctuations. Cambridge, MA: National Bureau of Economic Research. Guzman, M. and J. E. Stiglitz (2016b), ‘A theory of pseudo-wealth’, in Joseph E. Stiglitz and Martin Guzman (eds) Contemporary Issues in Macroeconomics: Lessons from The Crisis and Beyond. IEA Conference Vol. No. 155-II. Basingstoke and New York: Palgrave Macmillan, pp. 21–33. Hoff, K. and J. E. Stiglitz (2001), ‘Modern economic theory and development’, in G. Meier and J. Stiglitz (eds) Frontiers of Development Economics: The Future in Perspective. Oxford: Oxford University Press, pp. 389–459. Islam, R. (2002), The Right to Tell: The Role of Mass Media in Economic Development, ed. R. Islam. Washington, DC: World Bank. Lipsey, R. G. and K. Lancaster (1956), ‘The general theory of the second best’. Review of Economic Studies 24(1): 11–32. Meade, J. E. (1955), Trade and Welfare. Oxford: Oxford University Press. (p.29) Newbery, D. and J. E. Stiglitz (1984), ‘Pareto inferior trade’. Review of Economic Studies 51(1): 1–12. Rodrik, D. (1998), ‘Who needs capital-account convertibility?’, in Peter Kenen (ed.) ‘Should the IMF pursue capital-account convertibility?’ Princeton Essays in International Finance 207. Princeton, NJ: Princeton University, pp. 55–65.
Page 16 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions Rothschild, M. and J. E. Stiglitz (1976), ‘Equilibrium in competitive insurance markets: an essay on the economics of imperfect information’. Quarterly Journal of Economics 90(4): 629–49. Sappington, D. and J. E. Stiglitz (1987), ‘Privatization, information and incentives’. Journal of Policy Analysis and Management 6(4): 567–82. Stiglitz, J. E. (1985a), ‘Credit markets and the control of capital’. Journal of Money, Banking, and Credit 17(2): 133–52. Stiglitz, J. E. 1985b), ‘Equilibrium wage distributions’. Economic Journal 95(379): 595–618. Reprinted in The Selected Works of Joseph E. Stiglitz, Volume II: Information and Economic Analysis: Applications to Capital, Labor, and Product Markets, Oxford: Oxford University Press, 2013, pp. 559–80. Stiglitz, J. E. (1994), Whither Socialism? Cambridge, MA: MIT Press. [Expanded from a paper presented at the Wicksell Lectures, May 1990.] Stiglitz, J. E. (1999), ‘On Liberty, the Right to Know and Public Discourse: The Role of Transparency in Public Life’. Oxford Amnesty Lecture, Oxford, January 1999, subsequently published as Chapter 8 in The Rebel Within, Ha-Joon Chang (ed.), London: Wimbledon Publishing Company, 2001, pp. 250–78 and in Globalizing Rights, Matthew Gibney (ed.), Oxford: Oxford University Press, 2003, pp. 115–56. Stiglitz, J. E. (2000), ‘Challenges in the analysis of the role of institutions in economic development’, in Gudrun Kochendörfer-Lucius and Boris Pleskovic (eds) Inclusion, Justice and Poverty Reduction. Villa Borsig Workshop Series 2000: The Institutional Foundations of a Market Economy. Berlin: Deutsche Stiftung für Internationale Entwicklung, pp. 15–28. Stiglitz, J. E. (2002), ‘Transparency in government’, in R. Islam (ed.) The Right to Tell: The Role of Mass Media in Economic Development. WBI Development Studies. Washington, DC: World Bank, pp. 27–44. Stiglitz, J. E. (2008), ‘Capital market liberalization, globalization, and the IMF’, in J. E. Stiglitz and J. A. Ocampo (eds) Capital Market Liberalization and Development. New York: Oxford University Press, pp. 76–100 [Updated version of ‘Capital-Market Liberalization, Globalization and the IMF’. Oxford Review of Economic Policy (2004) 20(1): 57–71.] Stiglitz, J. E. (2009), Selected Works of Joseph E. Stiglitz, Volume I: Information and Economic Analysis. Oxford: Oxford University Press. Stiglitz, J. E. (2010), Freefall: America, Free Markets, and the Sinking of the World Economy. New York: W.W. Norton.
Page 17 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions Stiglitz, J. E. (2012), The Price of Inequality: How Today’s Divided Society Endangers Our Future. New York: W.W. Norton. Stiglitz, J. E. (2013), The Selected Works of Joseph E. Stiglitz, Volume II: Information and Economic Analysis: Applications to Capital, Labor, and Product Markets. Oxford: Oxford University Press. (p.30) Stiglitz, J. E. (2015), ‘A progressive agenda for the twenty-first century’, in David B. Woolner and John M. Thompson (eds) Progressivism in America: Past, Present and Future. New York: Oxford University Press, pp. 215–32. Stiglitz, J. E. (2016), ‘The state, the market, and development’, in Tony Addison and Finn Tarp (eds) (forthcoming) Mapping Development Economics: The Past, Present and Future. Helsinki: UNU-WIDER [WIDER working paper 2016/1, January 2016, originally presented at UNU-WIDER 30th Anniversary Conference held September 2015 in Helsinki], available at https://www.wider.unu.edu/sites/ default/files/wp2016-1.pdf, accessed 27 June 2017. Stiglitz, J. E. with N. Abernathy, A. Hersh, S. Holmberg, and M. Konczal (2015), Rewriting the Rules of the American Economy: An Agenda for Growth and Shared Prosperity. New York: W.W. Norton and the Roosevelt Institute. World Bank (2001), World Development Report 2002: Building Institutions for Markets. Washington, DC: World Bank. Notes:
(1) See also Stiglitz (2000) and World Bank (2001). This result only holds if nonmarket insurers have no better information than market insurers. Given the restrictive conditions under which Nash equilibria within market economies achieve Pareto efficiency, there was little grounds for the presumption that this broad Nash equilibrium, involving market and non-market institutions, would be efficient. For a broader critique of these naive evolutionary ideas, see Stiglitz (1994). (2) That is, when markets are viewed as non-competitive, when they abuse the consumers that they are supposed to serve, when they are able to extract excessive rents, markets lose their legitimacy as mechanisms for allocating resources, and there will be less voluntary compliance with the terms of (implicit or explicit) contracts. In Freefall (Stiglitz, 2010) I detail the host of abusive practices engaged in by the financial sector in the years surrounding the financial crisis. (3) There were many key contributions over the more than a century during which that model evolved, including formalizations by Arrow, Debreu, and Samuelson.
Page 18 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Markets, States, and Institutions (4) In this work, it is the rules governing society as a whole that matter. (5) The conditions under which there is unanimity that it should maximize stock market value are even more restrictive. (6) See Grossman and Stiglitz (1977, 1980). (7) The importance of contract enforcement is highlighted by sovereign wealth debts, where the ability to enforce contracts is particularly limited, e.g. through reputation mechanisms (implicit contracts). See Eaton and Gersovitz (1981); Eaton et al. (1986). (8) See Stiglitz (2015). (9) Though as the 2016 election showed so clearly, with significant support from other segments of society. (10) Of course, special interests often try to cloak themselves in the guise of civil society, and it is not always easy to distinguish between the two. (11) For a brief discussion of these traps, especially viewed within an evolutionary context, see Hoff and Stiglitz (2001). (12) This low effective tax rate was underlined by the revelation in 2012 that Mitt Romney, the Republican presidential candidate at the time, had paid less than 14 per cent of his reported income in taxes. Romney practically boasted about the low rate he enjoyed, saying in a debate that ‘I pay all the taxes that are legally required and not a dollar more.’ It is a pathology of American political discourse that bragging about gaming the system for personal gain has become, in some quarters, the mark of a ‘good businessman’ and not a sign that someone is deficient in civic responsibility. This pathology was of course on display during the campaign of the current US President, Donald Trump.
Access brought to you by:
Page 19 of 19
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
On a Concept of Freedom Prasanta K. Pattanaik Yongsheng Xu
DOI:10.1093/oso/9780198812555.003.0003
Abstract and Keywords A considerable segment of the recent analytical literature on freedom visualizes an individual’s freedom as his/her ability to choose any one of several alternative (mutually exclusive) outcomes available to him/her. A major application of this formulation of freedom is to be found in the functioning and capability approach where an individual’s capability or opportunity set is taken to be a set of functioning bundles any one of which the individual is able to choose at will. Following the lead provided by an early contribution from K. Basu, this chapter explores the limitations of this conception of freedom, both in the framework of perfectly competitive markets and in cases where an individual’s life is determined through strategic interaction among several individuals. Keywords: freedom, outcomes, capability, functioning bundles, competitive equilibrium, strategies, game form, power of a player
1 Introduction There is now a significant literature on freedom in welfare economics.1 It is increasingly being accepted that, apart from the effect that greater or less freedom of choice may have on the well-being of an individual via its effects on the choices that the individual actually makes, freedom of choice also matters directly for an individual’s well-being since it is an essential dimension of it.2 This idea has been particularly influential in the functioning and capability approach to well-being due to Sen (1985, 1987) and Nussbaum (1988, 2000).
Page 1 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom The basic analytical model in recent literature on freedom visualizes an individual’s freedom as his/her opportunity to choose from a set of options. It is assumed that, at any given time, the individual is faced with a set of (mutually exclusive) options which are feasible for him/her. We shall call this set the individual’s opportunity set. An individual’s freedom is conceived as his/her ability to choose at will any of the options belonging to the opportunity set. At an abstract level, many contributions to the literature do not commit themselves to any particular interpretation of these options, leaving this to be chosen depending on the specific context. Thus, the options can be simple options as for example ‘pasta for lunch’, ‘fish for lunch’; alternatively, the options may be complex ‘outcomes’ for the individual, such as alternative (p.32) commodity bundles for a consumer. In the functioning and capability approach to the notion of an individual’s well-being, which has received extensive attention in recent decades and where the concept of freedom of choice has played a very important role, the options have been interpreted as outcomes, specified in a particular and very comprehensive fashion: the options for an individual are taken to be different functioning bundles, functionings being attributes, such as being healthy, being well-nourished, being protected from the elements, interaction with family and friends, living without shame about one’s position in society, participation in the affairs of one’ community, and so on, which people value in their lives. In this chapter, we focus on this particular interpretation of options that is used in the functioning and capability approach. We do so for two reasons. First, the concept of freedom of choice has played a very conspicuous role in the functioning and capability approach. Second, the functioning and capability approach uses an explicit and very clear interpretation of the options as alternative functioning bundles, while elsewhere in the literature the nature of the options is sometimes not clearly specified. Our emphasis on the notion of an opportunity set as it is conceived in the functioning and capability approach, however, does not imply that our discussion is relevant only for the concept of freedom in the functioning and capability approach. Virtually everything we say will remain intact even if the options before the agent are not functioning bundles but ‘outcomes’ (as distinct from actions or strategies) for the individual, specified in some other way. The purpose of this chapter is to examine, in some detail, certain conceptual issues which arise when an individual’s freedom is conceived in terms of his/her ability to choose any element of a set of functioning bundles available to him/ her.3 One can think of many instances where such a conception of freedom turns out to be problematic either because, given the scarcity of aggregate resources in society, it is not possible to envision individuals as independently choosing functioning bundles for themselves, or because strategic interaction among individuals makes the outcome for any given individual dependent on the actions of other individuals as well as his/her own action. As far as we know, Kaushik Page 2 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom Basu’s (1987) perceptive review of Sen’s (1987) was the first contribution to draw attention to the problem of how, in a perfectly competitive economy, overall scarcity of resources causes difficulties for the conception of a consumer’s freedom as his/her ability to choose a commodity bundle from his/her budget set defined by the prices in a (p.33) competitive equilibrium.4 In subsequent sections, we examine Basu’s (1987) claim that scarcity of resources in a perfectly competitive economy leads to interdependence of consumers’ choices. We also consider problems with the notion of an agent choosing a functioning bundle from a set of feasible functioning bundles, when we go beyond the standard model of a perfectly competitive economy so as to permit strategic interaction among the individuals in society. The plan of the chapter is as follows. In Section 2, we consider the problem of formulating the notion of a consumer’s freedom in a perfectly competitive economy as either his/her freedom to choose a consumption bundle or as his/her freedom to choose a functioning bundle. In Section 3, we consider the difficulties with the notion of an individual’s opportunity set as a set of functioning bundles in the presence of strategic interaction among individuals. We conclude in Section 4.
2 Competitive Markets, Scarcity of Aggregate Resources, and Interdependence of Consumers’ Choices 2.1 Competitive Markets and Consumers’ Freedom
In this section we examine Basu’s (1987) claim that, given the scarcity of aggregate resources in a perfectly competitive economy, it is not possible for a consumer to choose freely any consumption bundle in his/her budget set in a competitive equilibrium independently of other consumers’ choices from their respective budget sets. Basu (1987) considered a two-person pure exchange economy to support his argument. For our purpose here, it will be useful to start with a slightly different model, which is based on Pattanaik (1994). Consider a pure exchange private-ownership economy, with two commodities, , and two types (Type I and Type II) of consumers, with
‘identical’
consumers of each type. Figure 3.1 depicts a competitive equilibrium for this economy, where, for each consumer of Type r,
,
is the initial endowment bundle of each
is the consumption bundle of each consumer of Type r in
the competitive equilibrium, and p is the equilibrium price vector. We assume that there are exactly two functionings, functioning
and
, and that, to generate q units of
each consumer needs q units of commodity cj (j = 1, 2) and
nothing of the other commodity. Note that, given these assumptions, the distinction between functioning bundles and commodity bundles practically disappears, and, for our purpose, we can talk (p.34) interchangeably about a consumer’s ability to choose a commodity bundle and his/her ability to choose a functioning bundle. It is tempting to say that, in the depicted competitive Page 3 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom equilibrium, the set of commodity bundles available to each consumer of Type I is OIab and the set of commodity bundles available to each consumer of Type II is OIIad. First, consider the case where so that there are exactly two consumers in the economy, consumer 1 being the single consumer of Type I and consumer 2 being the single consumer of Type II. In what sense does consumer 1 have the freedom to choose any commodity bundle in OIab, and consumer 2 have the freedom to choose any bundle in OIIad? Clearly, in this two-person pure exchange economy, it is not possible for consumers 1 and 2 to simultaneously choose
and
, respectively, though
the initial competitive equilibrium and
belongs to consumer 1’s budget set in belongs to consumer 2’s budget set in
the initial equilibrium. Starting with the initial competitive equilibrium, where the price vector is p = (p1, p2), if consumer 2 continues to demand consumption bundle
, but consumer 1 seeks to buy
then, at the equilibrium price vector
p, there will be positive excess demand for commodity
and that will destroy
the initial equilibrium itself. (p.35) It may be argued that the difficulty here arises because we are considering a perfectly competitive economy with only two consumers and that although formally there is no reason why one cannot have a perfectly competitive economy with only two consumers, our intuition about perfect competition is linked to the assumption that the number of agents is large. When we have an indefinitely large number of consumers in our pure exchange economy, no single consumer’s desire to shift from his/her equilibrium consumption bundle to some Figure 3.1 Competitive markets and other consumption bundle in individual freedom his/her budget set at the equilibrium will have any effect on the equilibrium prices. This is true, but the difficulty with the formulation of a consumer’s freedom of choice in terms of his/her ability to choose from his/her budget set does not disappear completely when we introduce a large number of consumers. Consider again our pure exchange economy and let
(the number
of consumers of each type) be indefinitely large. Now, if a single consumer of Type I seeks to switch from
to
then it will not make any difference to the
Page 4 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom prices. The same, however, cannot be said when, given the price vector
a
‘significant’ number of consumers of Type I seek to shift, simultaneously, from to
, and the remaining consumers of Type I, as well as all consumers of
Type II, want to continue with their original consumption bundles in the competitive equilibrium. In this case we shall have a problem exactly analogous to what we saw in the two-person case. The basic point is this. When we conceive the freedom of each individual j belonging to a group of some specified set,
as his/her ability to choose any element , of mutually exclusive options, then, irrespective of
whether the options are taken to be consumption bundles, functioning bundles, or something else, intuitively that conception of freedom seems to imply that: (1) for all
such that, for all
to choose, simultaneously,
,
, it is possible for respectively.
If the opportunity set of each individual is, say, {believing in God, not believing in God, being an agnostic}, then there would not be any tension between the conception of an individual’s freedom of choice as his/her ability to choose any element of his/her opportunity set and (1). But, if the individuals are consumers in a competitive economy and the opportunity set of each consumer in a competitive equilibrium is taken to be his/her budget set at that equilibrium, then, as we have seen earlier, the conception of a consumer’s freedom of choice as the ability to choose any element of the opportunity set cannot satisfy (1), given the economy’s aggregate resource constrains.5 (p.36) 2.2 Conditionality of Freedom
In informal discussions, we have come across an argument consisting of the following assertions: (2) Freedom of choice of an individual is usually subject to conditions which are often omitted in statements about such freedom if those conditions are obvious. (3) When one says that, in a competitive equilibrium, every consumer has the ability to switch from his/her equilibrium consumption bundle to any other consumption bundle in his/her budget set, implicit in that statement there is an obvious qualification, namely, that not enough consumers should want, simultaneously, to switch from their respective equilibrium consumption bundles to other commodity bundles in their budget sets in a way that will generate an excess demand or excess supply for any commodity at the given equilibrium prices. We believe that (2) is a valid claim. When we talk about the freedoms enjoyed by individuals in a given society to do/be/achieve certain things, we often do so with Page 5 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom the implicit assumption that their physical environment and personal features, as well as the legal and political institutions of society, will continue unchanged; it is taken for granted that an individual’s ability to achieve anything can hardly remain unaffected in the event of major changes in his/her natural, political, and legal environments. But it seems much less usual to talk about the freedom of each individual in a society to choose any consumption bundle from his/her budget set provided other individuals do not choose certain consumption bundles from their budget sets. We also believe that it will be a valid statement if, as suggested in (3), we say that: (4) In a competitive equilibrium, every consumer has the freedom to choose any commodity bundle in his/her budget set defined by the equilibrium prices provided not enough consumers decide to exercise their freedom of choice by seeking to deviate simultaneously from their respective equilibrium consumption bundles to other commodity bundles in their budget sets in a way that will generate excess demand for or excess supply of any commodity at the initial equilibrium prices. We do not know how obvious the qualification, ‘provided…equilibrium prices’, in (4) is, but, explicitly stated, it does serve the useful purpose of reminding ourselves about the limitations and fragility of the concept of freedom of choice with respect to consumption bundles that a consumer is often assumed to enjoy in a competitive economy.6
(p.37) 3 Strategic Interaction of Individuals and the Opportunity Set In Section 2, we saw that, even in the framework of a model, where each consumer’s functioning bundle depends exclusively on the commodity bundle that he/she buys, the notion of an individual’s freedom of choice as the ability to choose any one of several available functioning bundles becomes problematic and needs to be suitably qualified. The difficulties become more serious when, instead of perfectly competitive markets, we have situations where individuals strategically interact with each other. We now consider such situations. 3.1 Some Examples
It may be useful to start with a few examples. Example 1. Suppose a team of five persons in a corporation has been given a project. If at least a majority of the team members work extra hours, then the project will be completed by a target date and each member of the team will receive a large cash reward. On the other hand, if only a minority of the team work extra hours, then the project will not be completed by the target date and no one in the team will receive any award. In this situation, the outcome for each individual member of the team (receiving a cash award and working extra hours; receiving a cash award and not working extra hours; not receiving a cash award Page 6 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom and working extra hours; and not receiving a cash award and not working extra hours) depends on whether or not he/she works extra hours and also on how many other members of the team decide to work extra hours. In this case, it will be rather odd to talk about an individual’s freedom to choose an outcome for him/herself though it does makes sense to talk about each individual’s freedom to choose whether to work extra hours or not to work extra hours. The outcomes in this example involve extra income or the absence of it and the amount of leisure, and have no direct reference to functionings. But if the leisure–income combination that an individual will have depends on his/her teammates’ choices of their actions (to work extra hours or not to work extra hours), then it is clear that an individual’s functioning bundle will also depend on his/her action as well as the actions of other individuals in his/her team. Example 2. Consider the functioning of interaction with family and friends. Note that, by definition, how much interaction a person will have with (p.38) his/her family and friends cannot be determined by him/her alone; it will also depend on whether family and friends want to interact with him/her. Example 3. If one considers a family as a mini society, then several functionings, such as housing, constitute a public good for this ‘society’. If the spouses are both earning members of the family, then how much of this public good one of the spouses can have cannot be independent of the amount of income that the other spouse is willing to contribute to generate this functioning.7 Example 4. The presence or absence of sexual or racial harassment in workplaces, caste- or religion-based discrimination, payment of less than minimum wages to undocumented migrant workers, can have much importance for a person’s ability to live without shame about his/her position in society. Such harassment, discrimination is illegal in many countries. But, even in such countries, these are not matters where a victim can control the outcome for him/ herself. Even if he/she lodges a formal complaint with appropriate authorities, the outcome of the complaint for him/her will also depend on the actions of several other people, including the persons who harass him/her or discriminate against him/her, his/her colleagues who may be asked to give evidence, and the authorities who will adjudicate the complaint. These are just a few examples of strategic interactions among individuals, where the individual’s functioning bundle is determined by the actions of several other individuals as well as his/her own action in situations. Numerous such examples can be found, especially in traditional rural societies where the operation of the competitive market mechanism is often severely limited and many components of the functioning bundle of an individual are determined through the strategic interaction of a small number of persons including the individual under consideration.
Page 7 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom Since game theory is our standard tool for modelling such interactive situations, a natural question that arises is how one should think about an individual’s freedom of choice in a game. Following an early contribution by Pattanaik and Xu (2009), we discuss this issue in Subsection 3.2. 3.2 Strategic Interaction and Freedom of Choice
To discuss how the concept of an individual’s freedom of choice can be interpreted in the presence of strategic interaction among individuals, we need some additional notation. A strategic game form G is characterized by
where N is
the non-empty set of players, X is a non-empty set of outcomes, for all is (p.39) the (non-empty) set of strategies of player i, and g is the outcome function from
to X. A strategic game H is characterized by , where G is a strategic game form, and for all
,
is a real-
valued pay-off function of i defined over X. Let
be the set of all possible
functioning bundles. For every
denote
elements of
being denoted by
, we let
, the
, etc. In the rest of this section, we
consider strategic game forms and strategic games, such that: (5) the set,
of outcomes is
and (6) for each such that, for all
there exists a real-valued function ,
defined over
,
,
so that outcomes are n-tuples of functioning bundles with exactly one functioning bundle for each individual and every individual’s pay-off from an outcome depends only on his/her own functioning bundle and not on any other player’s functioning bundle. In general, in a strategic game/game form as specified above, the functioning bundle of an individual will depend on other players’ strategies as well as his/her own strategy. Therefore, the conception of an individual’s opportunity set as a set of functioning bundles, any one of which the individual can choose at will, will not in general be a very fruitful concept if the individuals are players in a game. The game has to be very unusual for a player to have that degree of control over his/her functioning bundle. We now consider several formulations of the influence/power/control that an individual player may have with respect to his/her functioning bundle in a game or game form. All of them capture some aspect or other of the influence or control that a player may have on his/her functioning bundle. But, as can be expected, none of them can be interpreted as implying the type of power for an Page 8 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom individual that is envisioned in the literature on the functioning and capability approach, namely, the individual’s ability to choose a functioning bundle from a set of functioning bundles available to him/her. Formulation 1 For every
and every strategy
{
let
some
if
uses strategy
denote the set ,
and
Thus,
then, no matter what strategies others use, i’s functioning
bundle will be an element of
and, for every
, there exists
, such that i’s functioning bundle figuring in
will be a.
is to be intuitively thought of as an uncertain prospect that i faces if he/ she chooses the strategy
(the uncertainty involved is non-probabilistic in (p.
40) nature): if i chooses strategy
, then, depending on the strategies chosen by
other players, i may end up with any functioning bundle in time of his/her choosing strategy
, and, at the
, he/she may not know which functioning
bundle he/she would end up with. The opportunity set of i can be interpreted as the set of all such uncertain prospects corresponding to the different strategies of i. More formally, the opportunity set of i can be taken to be the class subsets A of
such that, for some
of all
. Under this formulation
of an individual’s opportunity set, an individual, when choosing a strategy, really chooses an uncertain prospect belonging to
(of course, the uncertain
prospect may be a trivial uncertain prospect represented by a set containing a single functioning bundle). Two points may be noted here. First, the real-valued pay-off functions figuring in the specification of the game H do not play any role in this formulation of an opportunity set, only the game form implicit in the game H being relevant here. Second, the formulation has intuitive links with the notion of an α-effectivity function that has been used to represent the power of a coalition in a game (see Moulin, 1983). Under Formulation 1 of a player’s power, for every
and every
, i has the power to ensure that the outcome of
the game will lie in A. As noted above, one can think of of i, each element of
as the opportunity set
being interpreted as an uncertain prospect for i. In Table
3.1, which presents a two-person strategic game form, Table 3.1 Freedom as the power to choose uncertain prospects
Page 9 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom
Formulation 1 has its limitations. Saying that the individual can choose at will any one of the uncertain prospects belonging to
may not tell us much when
these uncertain prospects involve a large number of different functioning bundles. Also, Formulation 1 is not particularly discriminating. For example, under Formulation 1, the opportunity set of player 1 consists of the single uncertain prospect {
,
} for the game form represented in Table 3.2 as well
as for the game form given in Table 3.1. Table 3.2 Limitations of Formulation 1
Intuitively, however, the two game forms are very different in terms of the power that player 1 has in them. In the game form in Table 3.1, player 1’s choice of a strategy is of little consequence for player 1’s functioning bundle (it is player 2’s choice of strategy that completely determines the functioning bundle of each player), while in the game form presented in Table 3.2, player 1’s choice of his/ her strategy does matter for player 1’s functioning bundle. (p.41) Formulation 2 This formulation of the opportunity set starts with an equilibrium of the game and introduces the concept of a person’s opportunity set, given this initial equilibrium of the game. Though we have a multitude of equilibrium notions available to us, for our purpose it is not important to choose any specific equilibrium notion. Intuitively, it is enough for our purpose to assume that there is some generally acceptable notion of equilibrium and that the initial situation is an equilibrium n-tuple,
of strategies. Given this
Page 10 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom equilibrium, if player
shifts to a strategy
when the other
players continue to use their respective strategies figuring in the initial equilibrium, then the functioning bundle of i will be component of the vector
is the i-th
. Given the initial equilibrium s,
Formulation 2 identifies component of
, where
{
for some
a is the i-th
as the opportunity set of player i (
Intuitively,
if we start with an initial equilibrium, then the opportunity set of player i is the set of all functioning bundles that player i can secure for him/herself by playing the different alternative strategies available to him/her when the other players continue to play their initial equilibrium strategies. Table 3.3 Equilibrium and freedom as the power to choose outcomes
(In Table 3.3, the cell corresponding to each pair of strategies specifies the pair of functioning bundles, as well as the pair of pay-offs of the two players.) Consider the game in Table 3.3. Assume that Nash equilibrium is the relevant notion of equilibrium, and the initial situation is
which constitutes
a Nash equilibrium of the game. Then player 1’s opportunity set under Formulation 2 is
.
The intuitive basis of Formulation 2 seems to be analogous to the intuition underlying the reasoning that we examined earlier, namely that, in a competitive equilibrium of a pure exchange economy, every consumer has the freedom to choose any consumption bundle in his/her budget set defined by the competitive equilibrium prices. In Section 2, we discussed the difficulties with this line of reasoning in the context of a competitive equilibrium. Analogous difficulties arise in the case of Formulation 2 of the concept of a player’s opportunity set in a game. To say that, at the equilibrium
player 1’s (p.42) opportunity set is the
set of functioning bundles that player 1 can achieve by suitably choosing the alternative strategies available to him/her when player 2 is assumed to hold his/ her (i.e., player 2’s) strategy fixed at his/her equilibrium strategy
will be
Page 11 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom persuasive only if there is some convincing reason for making that assumption. It is difficult to think of any such reason. Just as player 1 can shift from his/her equilibrium strategy
to another strategy, say, strategy
deviate from his/her equilibrium strategy
, player 2 also can
It is not clear why one should rule
out the possibility of player 2’s shifting from his/her equilibrium strategy, some other strategy of his/hers when player 1 switches from
to
. On the
other hand, if player 1 switches from his/her equilibrium strategy player 2 also simultaneously decides to switch from will end up with the functioning bundle rather than the functioning bundle
to, say,
to
to
and
, then player 1
which does not belong to
,
.
Formulation 3 Consider any
and any given
. It is clear that, if individuals in
choose their respective strategies in the vector outcome of the game will lie in the set {
then the final
: for some
,
}, and, hence, the functioning bundle of i will be some element of the set : for some
and some
. Thus, it is tempting to say that, for each suitably choosing a strategy from in
and, hence, given
and
and each
, given
by
can have any specific functioning bundle
,
can be said to be the opportunity set
of i. Thus, it would seem that, for every opportunity set of
,
, Formulation 3 specifies an
the elements of which are functioning bundles that i can
possibly have, given
. This statement, however, needs further scrutiny. What
exactly is meant by saying that, given functioning bundle in
, i can choose to have any
by suitably choosing a strategy from
Note
that, in general, at the time of choosing his/her own strategy, a player, i, in a strategic game will not know the strategies that the other players are deciding to adopt. Therefore, even if other players decide to adopt their respective strategies in
, at the (p.43) time of choosing his/her own strategy i will not
even know that the set of functioning bundles that he/she can possibly hope to have is restricted to their strategies in
by virtue of the decisions of other players to adopt ; of course, without knowing the strategies chosen by other
players, there is no way in which i can choose to have any specific functioning bundle in
by an appropriate choice of his/her own strategy. Given this,
even if players in call
decide to choose their strategies in
, one can hardly
i’s opportunity set in the sense of i being able to have any
functioning bundle in that set by suitably choosing his/her own strategy. Thus, what Formulation 3 does is to specify, for each
the functioning bundles
that will be possible for i when other players decide to adopt the strategies in Page 12 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom 8
(though, as we have said, at the time of choosing his/her own strategy, i will
not know how exactly the decisions of other players has affected the set of possible functioning bundles for him/her). Like Formulation 1, Formulation 3 also involves the problem of not being able to discriminate between game forms which have significant differences in terms of an individual’s control over his/her functioning bundle. Consider the two game forms in Tables 3.4 and 3.5. In the game form in Table 3.4, given any strategy of player 2, the set of functioning bundles possible for player 1 is
So
also is the case in the game form in Table 3.5. However, it can be argued that the game form in Table 3.4 offers individual 1 more power than the game form in Table 3.5. Table 3.4 Formulation 3 and contingent freedom
Table 3.5 Limitations of Formulation 3
It may be noted that Formulations 1 and 3 are based on game forms,9 while Formulation 2 is based on games and the notion of equilibrium. A game form (p. 44) does not have information about players’ preferences, while preferences of the players constitute an important component of a game and the notion of equilibrium. Given this difference between a game form and a game, Formulations 1 and 3 reflect a view of an individual’s power or control over his/ her functioning bundle, in which an individual’s power or control is independent Page 13 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom of the preferences of individuals, while Formulation 2 seems to suggest that there may be a role of preferences in conceptualizing the individual’s control over his/her functioning bundle. In the existing literature on the measurement of freedom (see Pattanaik and Xu (forthcoming) for an overview on this and related issues), there has been some discussion of the issue of whether preferences should matter at all for such measurements. It seems that a similar issue also arises when we seek to impose a formal structure on the concept of control over his/her own functioning bundle that a player enjoys in a situation of strategic interaction. As a closing remark of this section, we note that the outcome function figuring in a game form or a game is an important feature of the underlying institution governing the players, and as such, it can impact significantly an individual’s opportunity and achievement. Though in our discussion on possible formulations of an individual’s control/influence over his/her functioning bundles, we have explicitly focused on and seen how interactions among the players via their strategies and/or the notion of equilibrium employed in games shape the individual’s control over his/her functioning bundle, it is clear that the outcome function in a game form also plays an important role in conceptualizing an individual’s opportunity set: game forms in Tables 3.1 and 3.2, and in Tables 3.4 and 3.5, respectively, demonstrate the role that the outcome function plays in formulating an individual’s control over his/her functioning bundles.
(p.45) 4 Concluding Remarks The recent literature on freedom of choice in welfare economics, especially the analytical literature on the functioning and capability approach to human wellbeing, has done much to clarify the important role of freedom in the ethical assessment social states. An individual’s freedom of choice, however, has sometimes been conceived in this literature as the agent’s ability to choose any element of a set of comprehensively specified outcomes (e.g. a set of consumption bundles, a set of functioning bundles) for the individual concerned. This raises certain conceptual problems. We have discussed some of these problems, which come up in the conventional framework of a perfectly competitive economy and also in situations involving strategic interaction among individuals. In particular, though we considered several possible formulations of an individual’s control over his/her functioning bundle in the presence of strategic interaction among individuals, none of them requires that the individual has the ability to choose a functioning bundle from a set of functioning bundles available to him/her—the ability that the individual is supposed to have in the conceptualization of his/her freedom in the functioning and capability approach. Freedom is an exceptionally rich and complex concept. Therefore, it is not surprising that difficulties have come up in attempts to provide a precise formulation of freedom. These difficulties, however, do not detract from the Page 14 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom importance of such attempts. What these difficulties point to is the need for much further work on the notion of an individual’s freedom. We hope that our analysis in this chapter, which draws on an insightful observation by Basu (1987), clarifies some aspects of this rich concept. Acknowledgements We are grateful to Wulf Gaertner and Maurice Salles for their helpful comments. References Bibliography references: Barbera, S., W. Bossert, and P. K. Pattanaik (2004), ‘Ranking sets of objects’, in S. Barbera, P. J. Hammond, and C. Seidl (eds) Handbook of Utility Theory, Vol. 2. Boston: Kluwer Academic Publishers, pp. 893–977. Basu, K. (1987), ‘Achievements, capabilities and the concept of well-being’. Social Choice and Welfare 4: 69–76. (p.46) Basu, K. and L. F. López-Calva (2011), ‘Functionings and capabilities’, in K. J. Arrow, A. Sen, and K. Suzumura (eds) Handbook of Social Choice and Welfare, Vol. 2. Amsterdam: North-Holland, pp. 153–87. Braham, M. (2006), ‘Measuring specific freedom’. Economics and Philosophy 22(3): 317–33. Dowding, K. and M. van Hees (2009), ‘Freedom of Choice’, in P. Anand, P. K. Pattanaik, and C. Puppe (eds) The Handbook of Rational and Social Choice. Oxford: Oxford University Press, pp. 374–92. Fleurbaey, M. and W. Gaertner (1996), ‘Admissibility and feasibility in game forms’. Analysis & Kritik 18: S.54–66. Foster, J. E. (2011), ‘Freedom, opportunity, and well-being’, in K. J. Arrow, A. Sen, and K. Suzumura (eds) Handbook of Social Choice and Welfare, Vol. 2. Amsterdam: North-Holland, pp. 687–728. Miller, A. (2001), Interview with A. Miller by W. R. Ferris for the March–April 2001 issue of Humanities magazine, reported on the National Endowment for the Humanities website, available at https://www.neh.gov/about/awards/jeffersonlecture/arthur-miller-interview, accessed 3 December 2016. Moulin, H. (1983), The Strategy of Social Choice. New York: North-Holland. Nussbaum, M. (1988), ‘Nature, functioning and capability: Aristotle on political distribution’, in J. Annas and R. H. Grimm (eds) Oxford Studies in Ancient Philosophy: Supplementary Volume. Oxford: Oxford University Press. Page 15 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom Nussbaum, M. (2000), Women and Human Development. Cambridge: Cambridge University Press. Oppenheim, F. E. (1961), Dimensions of Freedom: An Analysis. Oxford: Blackwell. Oppenheim, F. E. (2004), ‘Social freedom: definition, measurability, valuation’. Social Choice and Welfare 22(1): 175–85. Pattanaik, P. K. (1994), ‘Rights and freedom in welfare economics’. European Economic Review 38: 731–38. Pattanaik, P. K. and Y. Xu (2009), ‘Conceptions of individual rights and freedom in welfare economics: a re-examination’, in P. Dumouchel and R. Gotoh (eds) Against Injustice: Ethics, Economics, and the Law. Cambridge: Cambridge University Press, pp. 187–218. Pattanaik, P. K. and Y. Xu (forthcoming), ‘On capability and its measurement’, in E. Chiappero-Martinetti, S. Osmani, and M. Qizilbash (eds) The Cambridge Handbook of the Capability Approach. Cambridge: Cambridge University Press. Sen, A. K. (1985), Commodities and Capabilities. Amsterdam: North-Holland. Sen, A. K. (1987), ‘The standard of living’, in G. Hawthorn (ed.) The Standard of Living. Cambridge: Cambridge University Press, pp. 1–38. Notes:
(1) For some overviews from this literature, see, among others, Barbera et al. (2004), Dowding and van Hees (2009), Basu and López-Calva (2011), and Foster (2011). (2) Further, it can also be argued that the freedom enjoyed by an individual in a society should have an intrinsic value in ethical assessments of the state of affairs of the society in addition to the instrumental value that it may have as a determinant of the individual's well-being (see Pattanaik and Xu, forthcoming). (3) Note that a fairly large part of recent literature on freedom deals with the problem of comparing levels of freedom, which different opportunity sets offer to a given agent, and the problem of comparing the freedoms of different individuals. In this chapter, we are not concerned with this problem of measurement; instead, we are concerned exclusively with the more primitive issue of how the notion of freedom of choice should be formulated. (4) Basu (1987) discusses a competitive consumer’s options interpreted as commodity bundles, but his reasoning is equally applicable to the case where
Page 16 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On a Concept of Freedom one interprets the options as functioning bundles (see Pattanaik (1994) for an early elaboration on this). (5) In a different context concerning individual rights formulated in a game form and the feasibility of those rights, Fleurbaey and Gaertner (1996) note a similar problem where a market outcome depends crucially on every agent’s action in the economy and a planned action by one individual may not bring about any outcome (allocation) that is feasible due to resource constraints. (6) Speaking about his experience during the Great Depression, Arthur Miller (2001) once said, ‘[I]t left me with the feeling that the economic system is subject to collapse at any particular moment…and that security is an illusion which some people are fortunate enough not to outlive’. Miller was talking about security in a market-based economic system. But the idea that every consumer in a perfectly competitive economy has freedom to choose any commodity bundle in his/her budget set defined by the prices in that competitive equilibrium may also be illusory in some ways. (7) See Basu and López-Calva (2011) for another example of interdependent choices of the members of a family. (8) Insofar as, under Formulation 3, the set of functioning bundles available to an individual is contingent on other individuals? strategies, Formulation 3 is reminiscent of the notion of contingent freedom discussed by Oppenheim (1961, 2004). (9) Braham (2006) develops a measure of freedom based on game forms where, intuitively, an individual’s freedom (of choosing an outcome a) is viewed as the conditional probability that gives the expectation that the individual can ensure the occurrence of a by choosing his/her strategies.
Access brought to you by:
Page 17 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Multidimensional Poverty Some Comments Bhaskar Dutta
DOI:10.1093/oso/9780198812555.003.0004
Abstract and Keywords A rich literature on multidimensional poverty measurement has recently sprung up. In this multidimensional setting, an individual is identified as poor or deprived if he or she falls short of these poverty-line-like cut-offs along various dimensions. A key assumption is that these shortfalls are non-comparable—a shortfall in one dimension cannot be compensated by a gain in another. This chapter reviews some of these axiomatic treatments and questions some assumptions. It points out that there are other ways in which one can identify impoverished individuals even within the multidimensional framework. Keywords: poverty, multidimensional poverty, monotonicity, focus axiom, identification, aggregation
1 Introduction A virtual consensus in the recent approach to the measurement of poverty is that individual well-being does not depend on income alone.1 Many non-income attributes related to, for instance, health and education determine the quality of life. This is despite the fact that those with higher incomes may be able to improve the level or value of several non-monetary variables. However, markets for many non-income attributes such as access to safe drinking water may not exist.2 Even when markets exist, they may be highly imperfect. So income as the sole indicator of well-being is inappropriate and has to be supplemented with other variables such as life expectancy, proximity to government health-care centres, access to safe water, and so on.
Page 1 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty Since well-being depends on a variety of attributes, a shortfall from any given level of well-being—the level of deprivation—must also be multidimensional. Since deprivation and poverty are essentially two ways of visualizing the same concept, this provides the rationale for measuring poverty from a multidimensional perspective. The steps involved in the construction of a multidimensional measure of poverty follow Sen (1976), who described two distinct stages in the measurement of poverty. The first step is an identification exercise and consists of specifying who among the entire population are the (p. 48) poor people. The second stage is an aggregation exercise which combines the relevant information about the poor into a scalar measure of poverty. Suppose the set of attributes which are key determinants of well-being have been identified.3 Analogous to the one-dimensional exercise, the typical approach is to specify threshold levels for each of the attributes. Given the threshold level(s), the identification exercise in the one-dimensional case is trivial—a person is deemed poor if her actual level of consumption or income falls below the threshold level. A main reason for writing this note is to point out that there are several alternative ways of conducting the identification exercise in a multidimensional case. Once the poor have been identified, the aggregation exercise will combine the vector of individual deprivation levels of the poor in different attributes into a scalar measure of poverty. This is usually done through the use of axioms which represent properties that a ‘sensible’ or ‘good’ measure should satisfy. Of course, here too, different sets of axioms can and have been used.4 This issue will be discussed in Section 4. Section 2 describes the basic framework and some notation. This is followed by Sections 3 and 4 on identification and aggregation exercises.
2 Basic Framework Let n represent the number of persons in a given society, and let d be the number of dimensions or attributes under consideration. Typical attributes are income or expenditure, achievement levels in health and education, access to safe drinking water, housing, and so on. Notice that while some attributes like income can be any real number, others (for example access to safe drinking water) may be binary—an individual either has access to water or does not. In such cases, we will denote 1 and 0 to denote access and no access respectively. The nature of the attributes plays an important role since this determines the kind of properties or axioms that one may find desirable in a ‘sensible’ measure of poverty. For any individual i, a d-dimensional vector of achievement in the d dimensions. So in dimension j. An
will represent the individual’s level
is the achievement of individual
matrix
represents society’s level of
achievement in these different dimensions. Since we may want to measure Page 2 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty poverty in societies of different size, let n vary but keep the number of dimensions d fixed. Let
be the set of achievement matrices
under consideration. (p.49) A measure of poverty P will assign a real number to each matrix of achievement levels. Occasionally it will be more convenient to represent the data in the form of deprivation levels rather than achievement matrices. For any dimension j, let represent a threshold value such that any individual whose achievement less than
is
is deemed deprived. If attribute j can take only binary values, then
. Anand and Sen (1997) point out that the threshold value itself may depend on the context or environment where the measure is going to be applied. For instance, it may not make sense to use the same threshold value of literacy or life expectancy in, say, both India and Japan. Of course, this makes intercountry comparisons of poverty levels problematic though such comparisons are frequently performed—the World Bank’s use of the $1 per day comparison being an obvious example. For any given y, let when vector of
and ,
denote a
deprivation matrix, where
when i is not deprived in dimension j. Then, the i-th row
, is person i’s deprivation vector—it describes the dimensions in
which person i is deprived. In one of the approaches to be described in Subsection 4.1, it will be useful to use a column vector c of deprivation counts, where Notice that
represents the number of deprivations suffered by person i. and c are well defined even when all attributes have only ordinal
significance since they remain invariant to monotonic transformations of y and z. In the special case when all attributes are cardinal, other transformations of the deprivation matrix also play a role in the construction of poverty indices. For any achievement matrix y and associated matrix
, let
be the matrix of normalized
poverty gaps, where
Thus, each
is non-negative, and is a measure of the extent to which person i is
deprived in dimension j. If individual i is not deprived in dimension j, then
.
It is a normalized measure since the extent of deprivation is expressed as a ratio of the poverty threshold by
. In general, for any
, define the matrix
.
Page 3 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty 3 Identifying the Poor The identification exercise in the unidimensional case is straightforward. Suppose the relevant unidimensional variable is income. This involves the specification of a poverty line, the interpretation being that this is the income required in order to sustain a minimum acceptable level of living. Of course, the choice of any particular poverty line is normative. But, once this is specified, (p. 50) anyone with income below the poverty line is deemed to be deprived and hence poor. Once the set of poor people is identified, the aggregation exercise converts the individual poverty gaps (the difference between the poverty line and the individual income) into a scalar measure of poverty. Notice that the unidimensional variable need not be a single indicator such as income or expenditure. Instead, it may well be a composite aggregate or index of fundamentally different indicators. For example suppose the principal determinants of individual well-being are deemed to be income, an indicator of health—say life expectancy; and education—for instance number of years of schooling. Then, one possible approach is to ‘substitute’ an individual’s achievement levels in these dimensions into a well-being function in order to derive the level of well-being. Specify a cut-off level of well-being—the analogue of the poverty line—and then judge the individual to be poor if their level of wellbeing is below the threshold level. We discuss a variant of this, attributable to Decancq et al. (2014), later in this section. The identification exercise in the multidimensional case is more complicated. Analogous to the unidimensional case, one could start with a vector of minimum acceptable levels Given z, let
,
being the ‘poverty line’ for dimension j.
be the level of poverty associated with the achievement
vector y. There may well be individuals who will be deprived in some dimensions but may be ‘over-achievers’ in other dimensions. Will overachievement in some dimension compensate for deprivation in some other dimension? In other words, is some trade-off allowed across dimensions? The following axiom rules out any such trade-offs. Definition 1 Strong Focus (SF): For any (i) for any i such that
, where
,
, if
, (ii)
for all k and for all
, then
The axiom states that if individual i is not deprived in some dimension j, then any further improvement in her achievement in that dimension has no effect on the level of poverty. A similar poverty axiom is common in the literature on unidimensional poverty measurement.5 The idea in the unidimensional context is Page 4 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty straightforward. An individual is either poor—if his/her income is below the poverty line; or non-poor—if his/her income is above the poverty line. So, if the level of income of someone who is not poor goes up, this can only influence the incidence of unidimensional poverty if the income levels of the non-poor are allowed to affect the level of poverty. Since there is (p.51) no transfer of income across individuals, it makes eminent sense to assume the focus axiom. However, the implicit assumption underlying the SF axiom in the multidimensional framework is significantly stronger. No amount of overachievement in any dimension can compensate for deprivation in other dimensions, even for the same individual. It is stronger than its unidimensional cousin because no interpersonal transfers are involved. The rationale for this somewhat extreme assumption has not been very well spelt out in the literature. The essential problem seems to be that although a great deal of formalism has been incorporated into the multidimensional measurement of poverty through various axioms, there is no formal definition of the basic concept of well-being itself! Notice that the premise that well-being depends on many attributes is perfectly consistent with the notion that wellbeing itself is a one-dimensional concept like utility or happiness. Once this interpretation is introduced, it makes sense to conceive of a well-being function, whose arguments are relevant monetary and non-monetary attributes, and such that the function is increasing in each argument. Within this framework, there are only two conditions under which compensation across dimensions can be ruled. The first such condition is if each dimensional threshold also represents a level of satiation, so that individuals derive no marginal benefit once their ‘consumption’ crosses this level. But this is an unrealistic assumption. The poverty threshold is supposed to represent a ‘minimum acceptable level’ in the sense that anyone with an achievement below this is deprived. Clearly, this cannot also be a satiation limit. Alternatively, the well-being function can be such that all attributes are perfect complements of one another. This too is an extreme assumption. Obviously the multidimensional approach assumes that achievement levels in different dimensions cannot be aggregated into a single index of well-being. The following quotation from Alkire and Foster (2011a) (henceforth AF) is an apt summary of the opposition to the unidimensional approach. However, the uni-dimensional form of identification entails a host of assumptions that restrict its applicability in practice, and desirability in principle. (2011a: 478)
Page 5 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty One justification for ruling out the possibility of such an aggregation exercise is that this requires the existence of appropriate weights which can be used to determine the relative importance of the deprivation in one dimension versus that in another dimension, and that these weights simply do not exist. For instance, even in the extreme case where goods and services constitute all d dimensions, appropriate prices may not exist because markets may be missing. Atkinson (2003) provides an interesting example of an instance where although markets exist, social norms may prevent some individuals from accessing the market. He points out that a minority family may have to live (p.52) in inferior housing despite having adequate income because discrimination may prevent the family from moving to localities with good housing facilities. A second argument against the unidimensional approach is that this entails ‘the loss of information on the dimension-specific shortfalls’.6 AF also argue that if dimensions are independently valued so that dimensional deprivations are inherently undesirable, then a deprivation in one dimension cannot be offset by an excess of achievement over the threshold level in another dimension. How strong are these arguments against the unidimensional approach? First, note that whenever a vector of achievement levels is converted into a scalar measure of poverty, there is some aggregation of deprivation levels. The only way this can be done is by using a system of weights. Viewed from this perspective, the difference between the unidimensional and multidimensional approaches is not whether aggregation across dimensions is possible or not, but whether the aggregation should be attempted before or after the identification exercise. It is not clear why one approach is better than the other. Second, consider the argument that there is a loss of information about dimension-specific shortfalls under the unidimensional approach. This is not very persuasive either because the multidimensional approach too can suffer from this problem unless one uses what has been called the union approach, to be discussed later in this section. To see this, consider an extreme example in which a person is deemed poor only if he/she is deprived in all dimensions. Consider an achievement matrix y where everyone is deprived in all but one dimension. Possibly, the dimension in which individual i is deprived is different from the dimension in which j is deprived. Then, no one is deemed poor, although there can be deprivations in many dimensions. Moreover, if n is large, then the aggregate deprivation in each dimension can be quite large. Even when such trade-offs are ruled out, the identification exercise is not straightforward and alternative definitions are possible. In the ‘union’ approach, an individual is identified as poor if he/she is deprived in even a single dimension. Of course, if there are a large number of dimensions, then the number of the poor according to the union approach can be quite large. Some of them may not be even considered ‘poor’—but may be deprived in one dimension Page 6 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty as a matter of choice. For instance, Banerjee and Duflo (2011) provide anecdotal evidence that malnourished individuals may decide to spend on consumer durables such as radios and TV sets. Should such persons be deemed ‘poor’? It does seem somewhat paternalistic to do so! At the other extreme is the ‘intersection’ approach in which a person is deemed poor if he/she is deprived in every dimension.7 There is a sense in (p.53) which the intersection approach identifies a situation of extreme poverty or deprivation —one has to be deprived in all dimensions to be considered poor. The intersection approach obviously suffers from the defect that it excludes individuals from the set of poor if the individual escapes deprivation in even one dimension. An intermediate approach specifies a number
and identifies
an individual as poor if he/she is deprived in k or more dimensions. Obviously, the set of poor will be nested in the sense that the set of poor for k will contain that for
. The choice of the appropriate value of k is, of course, an ethical
issue. Alkire and Foster (2011a) mention that this ‘counting approach’ is widely used in empirical work. Decancq et al. (2014) follow a different approach. They start from the premise that an individual’s own preferences should be used to determine his/her wellbeing. This in turn implies that individual well-being (and deprivation) is also determined by one’s own preference ordering. Let of individual i. They postulate a bundle of attributes individual i with achievement level
feels deprived if
the indifference contour passing through
be the preference ordering for each i, so that . In other words,
represents the threshold level of
well-being, and anyone with any achievement vector on a lower indifference contour is poor. Two points are worth noting at this stage. First, this approach obviously violates the SF axiom by allowing for trade-offs across dimensions. In fact, since the individual preference orderings aggregate the different dimensions into a one-dimensional measure of well-being, this is essentially a one-dimensional approach, but with the important difference that it recognizes that individual well-being depends on a variety of attributes. Second, allowing individual preferences or perceptions to influence the identification exercise implies that two individuals with the same achievement vectors may feel different levels of deprivation. In particular, one individual may feel deprived while the other does not! The authors follow an axiomatic approach in order to make consistent inter-preference poverty comparisons. Since interpersonal comparisons are involved, these axioms involve ethical judgements. Of course, an alternative approach would be to assume the existence of a well-being function, uniform across all individuals. This would also involve value judgements, but has the advantage of being somewhat more straightforward.
4 The Aggregation Exercise
Page 7 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty One of the first papers in the literature on multidimensional poverty was that of Anand and Sen (1997), who constructed a Human Poverty Index (HPI). Anand and Sen started with the observation that the principal sources of deprivation used in the construction of such an index must be context-dependent; that is, (p. 54) the attributes which are the building blocks of the index must depend on the community for which the index is going to be used. For instance, deprivation in poor, developing countries is more likely to be due to hunger, poor sanitary conditions, inadequate access to health care, and the like. In contrast, issues such as social inclusion are more important determinants of deprivation in developed countries. Anand and Sen focus on attributes which are similar but not identical to those used in the construction of the Human Development Index.8 The variable used to represent deprivation in health is the proportion of people who are expected to die before age 40 at the current age-specific mortality rates (s), while illiteracy (k) is the variable used to measure educational deprivation. The third component of deprivation relates to economic provisioning and is represented by a simple average of (i) the percentage of the population without access to health care (h); (ii) the percentage without access to safe water (w); and (iii) the percentage of children who are undernourished (n). Letting
, the HPI is given by
(1)
Hence, the index gives equal weight to the three components of well-being. This is a special case of a more general class of poverty indices which take the form
where
stands for the level of poverty in component i. This allows for a more
general weighting scheme and also permits greater flexibility in the construction of the poverty subindices. Note the nature of the aggregation exercise conducted in the Anand and Sen (1997) approach. They first derive the aggregate level of deprivation in each attribute, and then combine the attribute-specific levels of deprivation into an overall level of deprivation. This approach is different from the identification exercise described earlier, for instance in the counting approach, where the first step is to identify the poor, and then aggregating the levels of deprivation of the poor to arrive at an overall level of poverty or deprivation. Unlike Anand and Sen, several recent papers have adopted the axiomatic route, with the axioms often being natural analogues of those used in the unidimensional case. In an influential paper, Bourguignon and Chakravarty (2003) defined some of these axioms.
Page 8 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty Definition 2 Weak Focus (WF): For any , if for some i,
,
, and
for all k; and (i) and for any j = 1, …
j, yij > xij, (ii) yil = xil for all l ≠ j, and (iii) xjk = yjk for all , then
, for all
.
(p.55) WF is considerably weaker than SF and requires that the value of the poverty index should be independent of the achievement levels of those persons who are deemed to be non-poor. This is the natural generalization of the focus axiom in the unidimensional case.9 Definition 3 Subgroup Decomposability (SD): For any and poverty bundle where
,
is the population size of the kth group and
.
This is an important axiom.10 Decomposability requires that the overall level of poverty is the weighted average of subgroup poverty levels, where the weights are the relative population shares of each group. This decomposability property is very convenient for policy analysis since it conveys information on how much for instance different social groups (partitioned say according to different religions, or geographical regions) contribute to overall poverty. Repeated application of SD also obviously implies that overall poverty is the simple average of individual poverty levels.
where
if individual i is not poor.
Clearly, the Anand and Sen approach does not satisy SD. Definition 4 Symmetry (S): For all permutation of y, then
, if x is obtained by a .
This is a relatively mild assumption and states that characteristics of persons other than the levels ‘consumed’ of the attributes used to define poverty are irrelevant for measuring poverty. Definition 5 Scale Invariance (SI): For all for all positive scalars
and poverty threshold z, .
This captures the idea that poverty or deprivation is a relative concept and so what matters is the relative distance between an individual’s level of achievement and the poverty threshold.
Page 9 of 17
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Multidimensional Poverty Definition 6 Continuity (C): For all z,
is continuous on Y.
This is essentially a technical requirement, and is quite common in the literature on unidimensional poverty. (p.56) Definition 7 Weak Monotonicity (WM): For all bundle
and poverty
, if
and
for all other pairs
, then
.
WM states that if an individual’s level of achievement goes up in some dimension and nothing else changes between two achievement vectors, then poverty decreases weakly. Note that if the improved achievement is experienced by someone who is non-poor, then poverty cannot go down absolutely—otherwise, there would be a violation of the WF axiom. On the other hand, a decrease in deprivation in some attributes for someone who is poor should translate into an absolute decrease in the incidence of poverty. This is captured by the next axiom. Some additional notation is required to state the axiom. Fix any system or method of identification I. Let
denote the set of individuals who have been
identified as poor according to the method of identification I, the achievement matrix
and the poverty threshold z.
Definition 8 Monotonicity (M): For all yij < zj, and xkl = ykl for all pairs
, for all
, if xij
0 implies that the official revision of the poverty line leads to a downward bias in the poverty rates, while the reverse is indicated if (PtR – PtN) < 0. The bias in nominal poverty rates (PtN) in relation to the real expenditure poverty rates (PtR) is due to the use in Page 4 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty calculating the former of a household invariant temporal adjustment to the household expenditures to compensate for price movements that may not reflect the true nature of price inflation faced by the individual households. The headcount poverty rate was used in the poverty calculations reported in Section 3.2. 2.2.1 Spatial Price Index between Regions within a Country
We can adapt the TCLI defined in the temporal context to estimate spatial differences in prices, with p1 and p2 denoting the vector of prices in two regions. The initial and given years will be reinterpreted, respectively, as the numeraire or base region, and the given region. If region 1 is the numeraire region, then the TCLI will yield the spatial price index of the given region, 2, with respect to region 1. Alternatively, if 1 refers to the country as a whole, then the spatial equivalent of the temporal TCLI yields the spatial price index of each constituent region with the country as the numeraire. One can further extend the TCLI concept, interpreted spatially across countries, to calculate the purchasing power parity (PPP) of the various countries with country 1’s currency adopted as the numeraire currency. 2.2.2 Equivalence Scales
Since the information on expenditure is at the household level, we need to deflate the household expenditures by the household’s ‘equivalence scale’ which takes account of the size and composition of the household. That requires knowledge of the equivalence scales which involves estimation of the parameters of the equivalence scale function from the available information. The TCLI, used in the household context, yields an analogous expression for the ‘exact household equivalence scale’ as follows:
(4)
z1 and z2 are the demographic vectors of the reference and comparison households respectively. Equation (4) will yield the general equivalence scale of household 2 evaluated at price vector, p, at reference utility level, ū, with respect to household 1 as the numeraire. Note that the extension of the TCLI concept to the measurement of equivalence scales, implied by Equation (4), requires the ad hoc specification of the demographically extended utility (p.103) function, u (x, z, p). Barten (1964) proposed a functional form that assumed quasi-price household composition effects: u (x, p*), where p* is the vector of the inner product of the vectors of item specific prices and item specific equivalence scales (pi zi). This form proved quite popular in welfare analysis8 since the assumed quasi-price demographic effects allowed the estimation of price elasticities from single cross-section data with no temporal price variation but with demographic variation between households. Alternatively, on pooled crossPage 5 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty section data over several time periods, the Barten (1964) model allowed the temporal variation of item prices to identify the item specific equivalence scales that are unidentified on single cross-section data where all households face the same prices. The Barten model underlined the parallel between price and demographic effects but, as Muellbauer (1977) found, the strong relation assumed by Barten between the two was rejected on pooled data from the UK Family Expenditure Surveys. To make the concepts operational, we need to assume specific functional forms for the utility, demand, and equivalence scale functions. In his study of the distributional consequences of UK inflation during 1964–70, Muellbauer (1974b) assumed the restrictive Stone Geary Linear Expenditure System (LES), which was subsequently relaxed in Ray (1985). In the studies on India reported below, the recent Quadratic Almost Ideal System (QAIDS) attributable to Banks et al. (1997) has been used to estimate the preference parameters required in the TCLI along with a linear functional form for the equivalence scale specification to estimate the household size deflator to convert the aggregate household expenditures to per capita terms. 2.3 Spatial Price Deflated Real Expenditure Comparisons between Regions
The TCLI defined in Equation (1) in the temporal context can be adapted to the spatial context to provide a spatial price index. Such an index is typically normalized at 1.0 for the whole country, with the spatial price index recording a value greater than one for a region with above average prices compared to the other regions, and less than 1.0 otherwise. As described in Section 3.4, such an index can be used in conjunction with the methodology described in Sen (1976) to provide a welfare ranking of the regions in each time period. The methodology proposed by Sen (1976) for real income comparisons between countries, and illustrated in that paper by applying it to studying regional differences in rural standard of living in India, was used in Majumder et al. (2015) to compare real expenditure among the constituent states of the Indian union. Sen (1976) proposes, as a welfare measure, the inequality (p.104) corrected nominal expenditure: , where and
is mean of the nominal expenditures
in state r,
is the Gini inequality measure of nominal expenditures in that state. The
spatial price of state r can be used to convert the welfare measure from nominal to real terms by defining expenditures
,
where
is the mean of the real
the corresponding spatially corrected real
expenditure inequality, and Sr is the spatial price of state r with respect to the All India figure which is normalized at 1.
Page 6 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty Rather than simply ranking the states on the basis of welfare measures, , Sen (1976) proposes that the welfare measure in nominal terms,
,
for region r is calculated not only at that region’s prices (pr), but also at other region’s prices, (ps), i.e.,
. Sen’s methodology consists
of constructing the matrix W from these spatially corrected welfare values, with the diagonal elements Wii being the values of the measure, states evaluated at that state’s prices, i.e.,
, in the various
, and the off diagonal elements
denoting the corresponding values evaluated at other states’ prices, i.e., the (s,r)th element denotes
. Majumder et al. (2015) adopt Sen’s
recommendation to rank states from the values of the W matrix as follows: ‘if the value of the diagonal element for any state 1 is larger than the value in the same row for another state 2, then we conclude that in terms of consumption state 1 has a higher rural standard of welfare’ (Sen [1976: 35]). This gives us a ‘partial ordering of a complete welfare indicator rather than a complete ordering of a partial welfare indictor’ (32).These pair-wise comparisons may not yield unambiguous rankings—for example, state i may have a higher welfare than state j with both states’ expenditures evaluated at state i’s price, while state j may have a higher welfare than state i with both expenditures evaluated at state j’s price. The welfare rankings can be presented in the form of Hasse diagrams as show in Figures 7.1a, and 7.1b.
3 Selected Empirical Evidence on NSS data 3.1 Prices and Expenditure Inequality
Tables 7.1 and 7.2 (from Mishra and Ray [2011]) present the expenditure shares in rural and urban areas, respectively, of households in the five quintiles of the expenditure distribution, arranged in ascending order of household expenditure per adult equivalent. The tables report the shares of the quintiles in terms of both nominal expenditure per adult equivalent and real expenditure per adult equivalent. There has been expenditure redistribution in both rural and urban areas from the bottom three quintiles to the top quintile throughout the reforms period and beyond (1993/94–2004/05). The expenditure distribution in both nominal and real terms is more unequal in urban (p.105) areas compared to the rural, as reflected in the lower share of the bottom three quintiles in the urban sector. A comparison of the nominal and real expenditure shares suggests that the price movements have been progressive over this period, since the real expenditure shares of the lower quintiles exceed the corresponding nominal expenditure shares in NSS rounds 55 and 61,9 and this is true in both rural and urban areas. This is not surprising if we recall that, during this period, the price of the composite group of luxury items called Miscellaneous increased more than those of the items of necessities, notably, Food.
Page 7 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.1 Quintile shares of total expenditure in rural areas Quintile
Nominal Expenditure Share
Real Expenditure Share
50th
55th
61st
50th
55th
61st
1
10.237
9.746
9.188
10.237
9.813
9.374
2
14.344
13.858
13.145
14.344
13.945
13.377
3
17.837
17.495
16.785
17.837
17.582
16.955
4
22.443
22.415
21.820
22.443
22.479
21.925
5
35.139
36.485
39.062
35.139
36.182
38.368
Source: Mishra and Ray (2011).
Page 8 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.2 Quintile shares of total expenditure in urban areas Quintile
Nominal Expenditure Share
Real Expenditure Share
50th
55th
61st
50th
55th
61st
1
9.039
8.477
7.792
9.039
8.580
7.854
2
13.399
12.940
11.593
13.399
13.065
11.679
3
17.250
16.945
15.874
17.250
17.064
15.968
4
22.621
22.657
22.446
22.621
22.735
22.558
5
37.691
38.981
42.295
37.691
38.556
41.941
Source: Mishra and Ray (2011).
Page 9 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty The progressive nature of price movements in India during the 1990s and the early part of the new millennium is seen more directly from Tables 7.3 and 7.4 which present the nominal and real expenditure inequalities in the two sectors. These tables report the inequality estimates calculated using the Gini inequality measure and the decomposable Generalized Entropy (GE) inequality index at varying levels of distribution sensitivity. The qualitative picture on inequality is generally robust to the inequality measure employed. Consistent with the evidence presented in Tables 7.1 and 7.2, inequality has been increasing in both rural and urban areas. (p.106)
Page 10 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.3 Nominal and real expenditure inequalities in rural areas Rounds
Nominal Gini
Real Generalized Entropy
Gini
GE(0)
GE(1)
GE(2)
Generalized Entropy GE(0)
GE(1)
GE(2)
50th
0.2482
0.1009
0.1097
0.1805
0.2482
0.1009
0.1097
0.1805
SEa
0.0014
0.0015
0.0045
0.0489
0.0014
0.0015
0.0045
0.0489
95% UB
0.2510
0.1039
0.1186
0.2764
0.2510
0.1039
0.1186
0.2764
95% LB
0.2455
0.0979
0.1008
0.0846
0.2455
0.0979
0.1008
0.0846
55th
0.2660
0.1159
0.1254
0.1844
0.2634
0.1136
0.1220
0.1727
SEa
0.0012
0.0012
0.0030
0.0219
0.0011
0.0011
0.0026
0.0173
95% UB
0.2683
0.1184
0.1312
0.2272
0.2655
0.1158
0.1271
0.2066
95% LB
0.2638
0.1135
0.1196
0.1416
0.2612
0.1114
0.1170
0.1389
61st
0.2962
0.1439
0.1665
0.2724
0.2876
0.1354
0.1534
0.2324
SEa
0.0013
0.0015
0.0031
0.0161
0.0012
0.0013
0.0025
0.0109
95% UB
0.2988
0.1469
0.1726
0.3041
0.2900
0.1379
0.1582
0.2537
95% LB
0.2936
0.1410
0.1603
0.2408
0.2852
0.1329
0.1486
0.2110
(a) SE = bootstrap standard error of the estimate; UB = Upper bound; LB= Lower bound. Source: Mishra and Ray (2011).
Page 11 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.4 Nominal and real expenditure inequalities in urban areas Rounds
Nominal Gini
Real Generalized Entropy
Gini
GE(0)
GE(1)
GE(2)
Generalized Entropy GE(0)
GE(1)
GE(2)
50th
0.2848
0.1336
0.1395
0.1892
0.2848
0.1336
0.1395
0.1892
SEa
0.0014
0.0015
0.0032
0.0206
0.0014
0.0015
0.0032
0.0206
95% UB
0.2876
0.1366
0.1458
0.2295
0.2876
0.1366
0.1458
0.2295
95% LB
0.2821
0.1307
0.1332
0.1489
0.2821
0.1307
0.1332
0.1489
55th
0.3045
0.1584
0.1866
0.7078
0.2998
0.1534
0.1762
0.5666
SEa
0.0034
0.0046
0.0166
0.2516
0.0030
0.0040
0.0139
0.1851
95% UB
0.3112
0.1675
0.2191
1.2009
0.3057
0.1613
0.2034
0.9294
95% LB
0.2978
0.1493
0.1542
0.2147
0.2939
0.1455
0.1490
0.2039
61st
0.3439
0.1924
0.2135
0.3365
0.3404
0.1884
0.2075
0.3173
SEa
0.0018
0.0022
0.0045
0.0270
0.0017
0.0020
0.0041
0.0227
95% UB
0.3474
0.1967
0.2223
0.3894
0.3437
0.1923
0.2155
0.3618
95% LB
0.3405
0.1882
0.2047
0.2836
0.3371
0.1844
0.1995
0.2727
(a) SE = bootstrap standard error of the estimate; UB = Upper bound; LB= Lower bound. Source: Mishra and Ray (2011).
Page 12 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty (p.107) 3.2 Prices and Expenditure Poverty
Table 7.5 presents the headcount poverty rates during the three NSS rounds, 50 (July 1993–June 1994), 55 (July 1999–2000), and 61 (July 2004–June 2005). The introduction of adult–child relativities via the estimated equivalence scales described in Section 2.2 leads to a sharp reduction in the nominal poverty rates from the per capita based figures in both areas. The boot strapped standard errors show that the poverty rates are well determined and highly significant. While the rural poverty rates register a steady decline throughout the period covered, the nominal per capita urban poverty rates record a sharp rise between 1999/2000 and 2004/05. The overall picture conveyed by Table 7.5 is one of declining poverty in rural areas and stagnant or rising poverty in urban areas. A comparison between the nominal and real poverty rates in Table 7.5 confirms the sharp effect of relative price changes on the poverty rates in both rural and urban areas. A comparison between the poverty rates based on four items— which are mostly necessities—and that on all items suggests that the nonessential items explain the bulk of the poverty in both sectors of the Indian economy.
Page 13 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.5 Headcount poverty rates Rounds
Rural
Urban
Over-expenditure on four included commodity groups
Over all items
a
Over-expenditure on four included commodity groups
Over all items
a
Nominalb Poverty Rate (per equiv.)
Real Poverty Rate Nominal Poverty (per equiv.) Rate (percapita)
Nominalb Poverty Rate (per equiv.)
Real Poverty Rate Nominal Poverty (per equiv.) Rate (percapita)
50th
0.0942
0.0942
0.2394
0.1130
0.1130
0.2162
SEc
0.0011
0.0011
0.0016
0.0013
0.0013
0.0020
95% UB
0.0964
0.0964
0.2424
0.1156
0.1156
0.2202
95% LB
0.0920
0.0920
0.2363
0.1103
0.1103
0.2122
55th
0.1123
0.0684
0.1922
0.1236
0.0792
0.1609
SEc
0.0011
0.0010
0.0014
0.0017
0.0014
0.0016
95% UB
0.1144
0.0703
0.1950
0.1269
0.0820
0.1641
95% LB
0.1102
0.0664
0.1894
0.1203
0.0765
0.1577
61st
0.0485
0.0308
0.1758
0.1102
0.0977
0.2689
SEc
0.0009
0.0006
0.0013
0.0015
0.0013
0.0017
95% UB
0.0503
0.0319
0.1783
0.1131
0.1002
0.2722
95% LB
0.0468
0.0297
0.1732
0.1074
0.0952
0.2655
(a) These included groups of items are: Food; Fuel and Light; Clothing, Bedding and Footwear; Miscellaneous.
Page 14 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty (b) The nominal poverty lines used in these calculations were obtained by scaling down the official poverty lines by multiplying them by the median budget share of the four commodity groups in total expenditure (0.944 for rural and 0.919 for urban) in the 61st round. (c) SE = bootstrap standard error of the estimate; UB = Upper bound; LB = Lower bound. Source: Mishra and Ray (2011).
Page 15 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty (p.108) A comparison of the nominal and real poverty rates in Table 7.5 based on expenditures on the four included items shows that the nominal poverty rates that use the official poverty lines had an upward bias in relation to the real poverty rates. This is also evident from the leftwards shift in the 95 per cent confidence interval as we move from the nominal to the real poverty rate estimates. This parallels the earlier result that the price movements had a progressive, inequality reducing effect through the realignment of relative prices. The narrowing of the difference between the nominal and real poverty rates in both areas between the fifty-fifth and the sixty-first rounds suggests, however, that the progressive nature of the relative price changes weakened in the second half of our chosen period. This is also evident in a similar narrowing of the difference between the nominal and real expenditure inequalities between these two NSS rounds, evident from Tables 7.3 and 7.4. 3.3 Spatial Differences in Inequalities and Prices
India is a large and diverse country with considerable heterogeneity in preferences and prices. If one recalls the discussion of the TCLI in Section 2.1, a (p.109) combination of sharp differences in prices and preferences between the principal states of the Indian union will imply differences in the TCLI between them. Table 7.6 (from Chakrabarty et al. [2015]) provides evidence on this issue from rural India. Table 7.6 presents the temporal ‘exact’ price indices for each state and for All India in NSS rounds 66 (July 2009–June 2010) and 61(July 2004–June 2005) with respect to NSS round 55 (July 1999–June 2000) as the base year. To see how these indices compare with the official Consumer Price Indices (CPIs) reported by the Labour Bureau, Government of India, we also report the official CPIs for the corresponding years. Table 7.6 shows that in both cases the second half (2004/05–2009/10) witnessed a much larger increase in prices than the first half (1999/2000–2004/05) of the decade. The similarity between the qualitative pictures painted by the NSS, ten items based ‘exact’ indices and the official cost of living estimates, evident from Table 7.6, confirms that the prices of the excluded items have not moved so differently from those of the included items so as to have large distributional implications that could question the robustness of the principal welfare conclusions of this study. Table 7.6 also underlines the spatial dimension in the price increases by recording considerable variation between the principal states in their temporal price inflation. As inflation accelerated sharply from the first half to the second half of the decade, so did the spatial dispersion in the temporal price indices between states. By the end of the decade, a wide gulf had opened up with, for example, Punjab recording almost a doubling of prices over the period in contrast to Kerala which recorded a much lower rate of inflation. The evidence on spatial price differences in India is consistent with similar findings for Brazil (Aten and Menezes [2002]) and Indonesia (McKelvey [2011]).
Page 16 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.6 State specific and All-India temporal price indices: rural sector base: NSS 55th round State
‘Exact’ Indices
Official Estimates*
NSS Rounds
2004–2005
2009–2010
55th
61st
66th
CPIAL
CPIRL
CPIAL
CPIRL
Andhra Pradesh
1.000
1.092
2.010
1.126
1.123
1.741
1.730
Assam
1.000
1.049
1.616
1.078
1.084
1.615
1.632
Bihar
1.000
1.056
1.727
1.149
1.148
1.773
1.761
Gujarat
1.000
1.071
1.703
1.115
1.114
1.713
1.708
Haryana
1.000
1.125
1.862
1.147
1.150
1.879
1.857
Karnataka
1.000
1.036
1.631
1.126
1.118
1.772
1.757
Kerala
1.000
1.058
1.545
1.093
1.086
1.545
1.549
Madhya Pradesh
1.000
1.034
1.718
1.065
1.073
1.694
1.700
Maharashtra
1.000
1.072
1.844
1.155
1.155
1.855
1.838
Orissa
1.000
1.011
1.666
1.053
1.053
1.628
1.632
Punjab
1.000
1.098
1.975
1.123
1.122
1.854
1.828
Rajasthan
1.000
1.021
1.665
1.113
1.106
1.842
1.817
Tamil Nadu
1.000
1.070
1.888
1.161
1.164
1.719
1.702
Uttar Pradesh
1.000
1.097
1.755
1.140
1.142
1.777
1.756
Page 17 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
State
West Bengal
‘Exact’ Indices
Official Estimates*
NSS Rounds
2004–2005
55th
61st
66th
CPIAL
CPIRL
CPIAL
CPIRL
1.000
1.062
1.671
1.140
1.147
1.726
1.727
2.84
7.48
2.97
2.96
5.51
4.91
1.076
1.790
1.125
1.124
1.743
1.729
Coefficient of variation (%) ALL INDIA (Rural)
2009–2010
–
1.000
(*) These were calculated from the published figures from Consumer Price Index for Agricultural Labourers (CPIAL) and Consumer Price Index for Rural Labourers (CPIRL) for the years corresponding to the three NSS rounds with 1986–87 as base. The figures in this table were obtained by dividing the 2004–05 and 2009–10 figures by the 1999–2000 figures for each state and All India. Source: Chakrabarty et al. (2015).
Page 18 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty Table 7.7 presents evidence of regional diversity of prices in India by reporting the spatial price indices estimated for each of the major states in India (rural areas) in NSS round 66 (July 2009–June 2010). This table also shows that the estimated spatial prices are quite robust between alternative procedures.
Page 19 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.7 State specific spatial price indices with respect to All India: 66th round, rural sector State
(Set 1)
(Set 2)
Evaluated using Exact Affine Stone Index (EASI) demand parameters estimated at All India level and
Evaluated using Exact Affine Stone Index (EASI) demand parameters estimated at state level and
Cox and Wohlgenant, Hoang Unit value
Deaton Unit value
Cox and Wohlgenant, Hoang Unit value
Deaton Unit value
Andhra Pradesh
1.389
1.354
1.215
1.163
Assam
1.179
1.197
1.098
1.109
Bihar
0.881
0.915
0.926
0.957
Chattisgarh
1.032
1.029
1.021
1.004
Gujarat
0.979
0.989
1.001
1.021
Haryana
0.847
0.857
0.911
0.957
Jharkhand
0.934
0.983
0.960
0.983
Karnataka
0.991
1.014
0.993
0.976
Kerala
1.384
1.272
1.198
1.136
Madhya Pradesh
0.789
0.828
0.874
0.910
Maharashtra
1.025
1.041
1.022
1.037
Orissa
0.884
0.877
0.932
0.909
Punjab
0.859
0.867
0.925
0.954
Rajasthan
0.776
0.783
0.866
0.881
Page 20 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
State
(Set 1)
(Set 2)
Evaluated using Exact Affine Stone Index (EASI) demand parameters estimated at All India level and
Evaluated using Exact Affine Stone Index (EASI) demand parameters estimated at state level and
Cox and Wohlgenant, Hoang Unit value
Deaton Unit value
Cox and Wohlgenant, Hoang Unit value
Deaton Unit value
Tamil Nadu
1.351
1.297
1.190
1.127
Uttar Pradesh
0.731
0.767
0.834
0.872
Uttaranchal
0.918
0.926
0.955
0.957
West Bengal
1.037
1.055
1.024
0.999
All India (Rural)
1.000
1.000
1.000
1.000
Coefficient of Variation
0.2129
0.1839
0.1201
0.0921
Source: Chakrabarty et al. (2015). See that paper for details of the procedures, Cox and Wohlgenant, Hoang, Deaton referred to in the table.
Page 21 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty Table 7.8 presents the Gini measure of the nominal and real expenditure inequalities (household level) both by state and for each time period. In this table, the nominal inequality refers to the case where all households within a state face the same price, while real inequality refers to the case where we allow the prices to differ between households by quartiles. Note that the two sets of inequalities are equal in the base year, 1999/2000. The following features are worth noting. First, there is considerable variation in the magnitude of the inequalities between states. This is true of both nominal and real expenditure inequalities. Second, while in most states inequalities were static or even recorded a decline during 1999/2000–2004/05, there was a sharp increase in inequality, in both nominal and real terms, in most states (p.110) during the second half, 2004/05–2009/10. The increase in inequality was particularly large in the cases of Kerala and Punjab, making them two of the most unequal states in India at the end of our sample period. While the sharp increase in the case of Kerala is possibly due to the increased inflow of remittances from countries in the Persian Gulf that favoured some households over others, the inequality increase in Punjab reflects the gain for richer farmers that benefited from growth-enhancing reforms and the large subsidies for diesel and fertilizers. The increase in inequality in nearly all the states during the period, 2004/05– 2009/10, is reflected in the sharp increase in inequality recorded by the All India figures in both nominal and real terms. Third, neither the magnitude nor the direction of change in inequality over the two sub-periods is identical for all the states nor is it robust between nominal and real expenditure inequality. For example, in Gujarat, while nominal inequality increased sharply during the period between NSS rounds 61 and 66, real expenditure inequality declined. In Haryana, while (p.111) there was a sharp increase in nominal inequality over this sub-period, real expenditure inequality remained unchanged.
Page 22 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Table 7.8 State specific and All-India Gini coefficients (nominal and temporal price deflated): Rural sector State
Gini coefficient *
Gini coefficient: Temporal Price Deflated (with respect to 55th Round)
(nominal)
Within a state all households face the same price (nominal)
Within a state all households within a quartile face the same price (real)
55th Round
61st Round
66th Round
61st Round
66th Round
Andhra Pradesh
0.226
0.204
0.265
0.202
0.250
Assam
0.189
0.141
0.232
0.128
0.219
Bihar
0.192
0.175
0.227
0.167
0.226
Gujarat
0.221
0.204
0.256
0.240
0.221
Haryana
0.243
0.232
0.287
0.260
0.260
Karnataka
0.228
0.195
0.252
0.192
0.221
Kerala
0.283
0.249
0.351
0.256
0.341
Madhya Pradesh
0.222
0.211
0.305
0.225
0.318
Maharashtra
0.240
0.207
0.246
0.214
0.235
Orissa
0.205
0.193
0.267
0.190
0.253
Punjab
0.221
0.205
0.313
0.179
0.258
Rajasthan
0.222
0.205
0.272
0.233
0.275
Tamil Nadu
0.264
0.204
0.290
0.213
0.273
Uttar Pradesh
0.232
0.211
0.253
0.226
0.253
West Bengal
0.202
0.187
0.232
0.174
0.233
Page 23 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
State
Gini coefficient (nominal)*
ALL INDIA (Rural)
Gini coefficient: Temporal Price Deflated (with respect to 55th Round) Within a state all households face the same price (nominal)
Within a state all households within a quartile face the same price (real)
55th Round
61st Round
66th Round
61st Round
66th Round
0.222
0.215
0.290
0.235
0.288
(*) The ‘nominal’ and ‘temporal price deflated’ Gini coefficients are the same for the 55th Round. Source: Chakrabarty et al. (2015).
Page 24 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty 3.4 Welfare Rankings of States in India
Let us now turn to the evidence on the spatial price deflated real expenditure comparisons between regions following the procedure due to Sen (1976) outlined earlier in Section 2.3. The evidence is contained in the Hass diagrams presented in Figures 7.1(a) and 7.1(b). State rankings and changes in rankings are brought out clearly by the Hasse diagrams for the different rounds presented in Figures 7.1(a) (rural), and 7.1(b) (urban). The diagrams are based on the W matrix (constructed from Laspeyre’s index) and the rule suggested by Sen (1976) on how to rank the states using the values of the distribution sensitive mean expenditure of a state evaluated at all the states’ prices, including its own prices. The Hasse diagram provides a clear representation of 210 pair-wise comparisons of the states’ welfare levels, ‘with a downward path (p.112) indicating superiority in the standard of welfare’ (Sen, 1976) under the assumption that all states have the same welfare function. A comparison of Figure 7.3 (from Sen [1976]) with Figure 7.1(a) of our chapter brings out several similarities and some sharp differences. Kerala was ranked near the bottom in Sen’s rankings based on NSS rounds 16 (1960–61) and 17 (1961–62), but it has moved up sharply to be at or near the top in Figure 7.1(a) in this chapter. Punjab has slipped slightly from its pre-eminent position in Sen’s study, with its top ranking taken by Haryana which was carved out of the erstwhile state of Punjab. Figures 7.1(a) and 7.1(b) in our chapter reveal several cases of changes in state rankings over the period spanned by the four NSS surveys. They also reveal several rural urban differences in the Hasse pictures. For example, in sixty-first round, rural Punjab is ranked quite highly among the rural states, but slips down several steps in the corresponding urban rankings. Overall, however, there are no major changes in the rankings over the period, 1993/94–2009/10, though the structure of the Hasse pictures has changed during this period. Figures 7.1(a) and 7.1(b) reiterate the evidence on significant spatial heterogeneity in prices and welfare reported and discussed in Section 3.3.
Page 25 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty
Figure 7.1a Hasse diagrams for various NSS rounds—rural India Source: Majumder et al. (2015): State names have been abbreviated in these diagrams. For state names corresponding to these abbreviations see Majunder et al., 2015: App. 2.
Page 26 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty 4 Concluding Remarks This chapter documents the shift in the literature on prices from being exclusively a macrotopic featuring in the study of inflation, national income accounting, and cross country income comparisons to one that is firmly rooted in microinvolving economic analysis of household behaviour, welfare, and the distributional implications of changes in relative prices. This shift has come in the wake of developments in the theory and application of price indices, most notably the TCLI that is based on utility specification, and the increasing availability of micro data sets that allow such applications. This has
Figure 7.1b Hasse diagrams for various NSS rounds—urban India Source: Majumder et al. (2015): State
allowed the treatment of prices to be more policy friendly
names have been abbreviated in these diagrams. For state names corresponding
widening the appeal of the topic of price measurement.
to these abbreviations see Majumder et al., 2015: App. 2.
This chapter brings together the results from some of the recent studies on Indian NSS data that examine the effect of price changes on inequality and poverty. It also contains evidence on spatial prices in the context of a large heterogeneous country such as India. The evidence on intra-country heterogeneity in preferences, price changes, and welfare, and the large changes in welfare based state rankings suggest that one needs to move beyond the traditional unitary treatment of a country in exercises such as estimation of PPP and GDP comparisons. There is now greater emphasis on calculating subnational PPPs as we move to the next round of the International Comparison Project (ICP). (p.113) (p.114) India has considerable information in its NSS data sets down to the level of unit records which has encouraged the distributional analysis of price changes. The discussion also suggests scope for improvements in the data. These include provision of panel data at the household level and an attempt to coordinate the NSS data sets with the National Family Household Survey (NFHS) data sets containing information on non-monetary deprivation (p.115) indicators. The NSS data have been repeated across section rather than panel Page 27 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty which makes it difficult to assess if the distributional changes over time are genuine or if they merely reflect changes in sampling design over time. Distributive analysis of price changes will remain incomplete if it is conducted on expenditure data alone. In its recent report, the World Bank appointed Commission of Global Poverty has recommended coordination of monetary and non-monetary measures of deprivation. Nowhere is this more desirable than in India which has two high quality data sets (NSS, NFHS) but with little coordination between the two. As the theory of price measurement develops further, it should be accompanied by greater attention to data provision to take the empirical literature forward. Acknowledgements I am pleased to contribute this chapter to the Festschrift in honour of my close friend and ex-colleague at Delhi School of Economics, Kaushik Basu. I have admired Kaushik for the range of his outstanding contributions which had such a huge impact on the profession. Kaushik has also been a source of great inspiration to his students, friends, and colleagues. References Bibliography references: Aten, B. and T. Menezes (2002), ‘Poverty price levels: an application to Brazilian metropolitan areas’. Paper presented to the World Bank ICP Conference, Washington, DC, March 11–15. Atkinson, A. (1970), ‘On the Measurement of Inequality’. Journal of Economic Theory 2: 244–63. Balk, B. (1995), ‘Axiomatic price index theory: a survey’, International Statistical Review 63: 69–93. Banks, J., R. Blundell, and A. Lewbel, A. (1997), Quadratic Engel curves and consumer demand’. Review of Economics and Statistics 79: 527–39. Barten, A. P. (1964), ‘Family composition, prices and expenditure patterns’, in P. E. Hart, G. Mills, and J. K. Whittaker (eds) Econometric Analysis for National Economic Planning. London: Butterworth. Basu, K. (2011), ‘Understanding inflation and controlling it’. Economic and Political Weekly 46(41): 50–64. Bhattacharya, N. and G. S. Chatterjee (1971), ‘Consumer prices and per capita household consumption in rural India’. Economic and Political Weekly 6(44): 2249–53.
Page 28 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty Chakrabarty, M., A. Majumder, and R. Ray (2015), ‘Preferences, spatial prices and inequality’. Journal of Development Studies 51: 1488–501. (p.116) Clements, K., I. H. Y. Izan, and E. A. Selvanathan (2006), ‘Stochastic index numbers: a review’. International Statistical Review 74: 235–70. Diewert, W. E. (1981), ‘The economic theory of index numbers: a survey’, in A. S. Deaton (ed.) Essays in the Theory of Consumer Behaviour. Cambridge: Cambridge University Press. Hicks, J. R. (1958), ‘Measurement of real income’. Oxford Economic Papers 10: 125–62. Konus, A. A. (1939), ‘The problem of the true index of the cost of living’ (1924 trans.). Econometrica 7: 10–29. McKelvey, C. (2011), ‘Price, unit value and quantity demanded’. Journal of Development Economics 95(1): 157–69. Majumder, A., R. Ray, and K. Sinha (2012), ‘The calculation of rural urban food price differentials from unit values in household expenditure surveys: a new procedure and comparison with existing methods’. American Journal of Agricultural Economics 94(5): 1218–35. Majumder, A., R. Ray, and K. Sinha (2015), ‘Spatial comparisons of prices and expenditure in a heterogeneous country: methodology with application to India’. Macroeconomic Dynamics 19(5): 931–89. Mishra, A. and R. Ray (2011), ‘Prices, inequality and poverty: methodology and Indian evidence’. Review of Income and Wealth 57: 427–48. Muellbauer, J. (1974a), ‘Household composition, Engel curves and welfare comparisons between households’. European Economic Review 5: 103–22. Muellbauer, J. (1974b), ‘Prices and inequality: the United Kingdom experience’. Economic Journal 84: 32–55. Muellbauer, J. (1977), ‘Testing the Barten model of household composition effects and the cost of children’. Economic Journal 87: 460–87. Ray, R. (1985), ‘Prices, children and inequality: further evidence for the United Kingdom, 1965–82’. Economic Journal 95: 1069–77. Samuelson, P.A. (1950), ‘Evaluation of real national income’. Oxford Economic Papers 2: 1–29. Sen, A. (1976), ‘Real national income’. Review of Economic Studies 43(1): 19–39.
Page 29 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Link between Preferences, Prices, Inequality, and Poverty World Bank (2013), Measuring the Real Size of the World Economy: The Framework, Methodology, and Results of the International Comparison Program —ICP. Washington, DC: World Bank. Notes:
(1) See Samuelson (1950); Hicks (1958). (2) See for example World Bank (2013). (3) See Basu (2011) for an insightful analysis of inflation in the traditional aggregate country context emphasizing the role of theory in reaching policy conclusions, and drawing a distinction between overall inflation and changes in relative prices. (4) There are several excellent reviews of price indices available. These include Diewert (1981); Balk (1995); Clements et al. (2006). (5) Sen (1976: App.) turns to the NSS to illustrate the empirical usefulness of his methodology. See also Bhattacharya and Chatterjee (1971) for an early price based empirical welfare analysis on Indian NSS data. (6) I draw on my recent work on Indian data with Manisha Chakrabarty, Amita Majumder, Ankita Mishra, and Kompal Sinha. (7) See Muellbauer (1974b). (8) Majumder et al. (2012) use the Barten (1964) model to estimate item-wise rural urban price differentials in India. (9) Since the prices are normalized at unity in the base round 50, the nominal and real expenditure shares are the same in that round. This remark also holds for inequality and poverty rates.
Access brought to you by:
Page 30 of 30
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Private versus Public Monopoly Jörgen W. Weibull Jun Chen
DOI:10.1093/oso/9780198812555.003.0008
Abstract and Keywords We compare private and public monopoly with respect to how much resource each spends on finding out what product varieties people want. We propose a simple model in which a monopolist supplies one variety of a good. This variety is chosen by the monopolist, and consumers differ in their valuations of the good and preferences over product varieties. The monopolist does not know the preference distribution, but can, at a cost, acquire more or less precise information about this distribution. We analyse the monopolist’s endogenous information acquisition and choice of product variety and find that, broadly speaking, public monopoly is preferable in societies with a wide spread in income and/or wealth while private monopoly is better in societies with less inequity. Keywords: uncertainty, information acquisition, monopoly, Ramsey–Boiteux price, regulation, inequity
1 Introduction We here address the following question: Does a monopolist spend too little, just about right, or too much—from a welfare point of view—to find out what product varieties consumers want? We posit a simple model in which a monopolist chooses a product variety from a continuum range, to supply to the market. Consumers have idiosyncratic Euclidean preferences over product varieties, and differ in their willingness to pay. For example, one rich individual may prefer a certain variety x and is willing to pay much for this personal ideal variety, while another, poor, individual may prefer a variety y (which may or may not be the same as x), but is willing to pay less for his/her personal variety. The monopolist Page 1 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly does not know the preference distribution in the population. It only has a vague prior belief about it. However, the monopolist may, at a cost, choose to obtain more precise information, where higher precision costs more. We represent the monopolist’s information about demand as a noisy signal about some parameter in the preference distribution. The monopolist may be an unregulated profit maximizer, an unconstrained welfare maximizer (‘first-best’), or, more generally, it may have a goal function that is an arbitrary convex combination of profit and welfare. This chapter belongs to the small literature dealing with firms’ choice of price and location, or product variety, under uncertainty. Basu (1988) studied a monopolist’s choice of price and product durability when consumers had heterogenous preferences. There is also a strand of literature studying duopolist’s choice of price and location, under given uncertainty about consumers’ preferences (see Vives, 1984; Basu, 1988; Harter, 1997; CasadoIzaga, 2000; Meagher and Zauner, 2004, 2005, 2011; Król, 2012). Unlike this literature about firms’ choice under exogenous uncertainty, this chapter analyses monopoly under endogenous uncertainty. The paper most similar to ours is Meagher (1996). (p.120) However, that paper does not address the question of optimal information acquisition. Instead, it considers a dynamic environment in which consumer preferences change over time and are not directly observable, and where the firm can conduct market research. We proceed first to analyse an unregulated profit-maximizing monopolist’s information gathering, price setting, and choice of product variety (or location). Second, we perform the same analysis on an unconstrained welfare maximizing monopolist, where welfare is defined as the sum of profit and consumer surplus. Such a monopolist will finance its activity by a lump-sum tax on all citizens. Third, and last, we analyse a monopolist who maximizes expected welfare under the constraint that its expected profit reaches an arbitrary pre-specified level. This setting is in line with the classical Ramsey–Boiteux approach (see Ramsey, 1927; Boiteux, 1971), and we obtain results for a wide range of second-best cases. In our simple model, the monopolist’s choice of product variety, at any given level of its uncertainty about consumer preferences, is always socially efficient; irrespective of whether its goal is profit or welfare or some convex combination thereof, the monopolist always strives to supply a variety that will attract as many consumers as possible, at any given price. By contrast, the different types of monopoly differ in the extent of their information acquisition. An unregulated profit-maximizing monopolist will typically either spend too little or too much resource on information acquisition, compared with first-best. The wider is the dispersion of consumers’ valuations of their ideal product varieties, the further is the unregulated private monopoly from first-best information acquisition. In a society with a wide spread in valuations, the unregulated private monopolist Page 2 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly therefore underinvests in information acquisition about consumer preferences, while in societies with little spread in valuations it overinvests, compared to first-best. It is only at an intermediate knife-edge case of dispersion of valuations that an unregulated private monopolist achieves the first-best. Broadly speaking, public or regulated private monopoly is preferable in societies with a wide dispersal of valuations, as would be expected in societies with a wide spread in income and/or wealth (say, India). By contrast, in societies with modest spread in valuations, such as would be expected in societies with fairly equal disposable incomes or wealth (say, Sweden), private monopoly is better than public monopoly. Our chapter thus also relates to the literature on monopoly regulation; see Armstrong and Sappington (2007). In this study of monopoly, we neglect many important aspects. Perhaps the most glaring omission is that this study presumes that the management of the monopoly, whether private or public, is fully rational and does not (try to) extract any private rents, such as shirking from work, taking bribes, using funds for luxurious offices, extravagant dinners, expensive but ill-motivated (p.121) trips, etc. Our focus is entirely on the monopolist’s incentives to acquire information about consumer tastes. To the best of our knowledge, this is the first paper studying this aspect of monopoly behaviour. We begin by setting up the model in Section 2. In Section 3 we analyse the case of an unregulated private monopolist. Here, the monopolist strives to maximize its expected profit, defined as its revenue from sales, net of production costs and its costs for information acquisition about consumer preferences over product varieties (or locations). In Section 4 we consider the first-best case of a monopolist who strives to maximize expected social welfare, defined as the sum of its profit and consumer surplus, without any budget constraint. In Section 5 we consider a range of intermediate cases, between the previous two extremes, and obtain the second-best solution, that of a monopolist who strives to maximize expected social welfare under a budget constraint. All numerical results are shown in Section 6. Section 7 briefly discusses some potential extensions of the model and Section 8 concludes.
2 Model Consider, thus, a monopolist in a market for a differentiated good sold in indivisible units. The monopolist has to choose a product variety a price cost
and
for this variety. The monopolist has a constant unit production . In this analysis, we abstract from fixed production costs. There is a
continuum of consumers, each with unit demand for the good. Every consumer has Euclidean preferences over varieties, with a personal ideal product variety . A consumer type is a pair utility for a consumer of type
, where V is an interval. The from buying one unit of product variety x
at price p is Page 3 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly
(1)
We will refer to
as the consumer type’s valuation (of its ideal product
variety). The utility from not buying is normalized to zero. A buyer of type τ buys a unit if and only if
.
Each consumer’s type τ is his or her private information. The population is treated as a continuum with unit mass, and v and θ are treated as statistically independent random variables. The cumulative distribution function (CDF) for valuations, v, is denoted density
, where F is absolutely continuous with
and such that the mean value
exists and exceeds the
unit production cost c:
(p.122)
Each consumer’s ideal product variety, θ, is the sum of a shared component,
,
common to all consumers in the population, and an idiosyncratic component ξ:
(2)
These two components are statistically independent and normally distributed, both with mean value zero. The shared component, the idiosyncratic component, ξ, has variance
, has variance
and
. Aggregate demand for
each product variety x is thus given by the demand function defined by
(3)
The integrand represents the mass of willing consumers for each valuation v. This is illustrated in Figure 8.1. The area above the parabola is the set of ‘willing consumer types’ when the price is at the horizontal dashed line and the supplied variety at the vertical dashed line. The monopolist’s revenue from sales is thus its price multiplied by the population mass of willing consumers, those with types inside the parabola.
Page 4 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly The monopolist knows all of the above. In addition, it can, at a cost, acquire more information about the shared component, , of consumer preferences over product varieties. We represent such information as a signal
where the error (or noise) term ε is statistically independent of all other random variables and is normally distributed with mean value zero and variance , where the precision
Figure 8.1 The set of willing buyers is
chosen by the monopolist, with
representing ‘no information acquisition’ or
‘no signal’. The cost for (p.123) signal precision q is is twice differentiable with
, where
,
and
for all
. The timeline is as follows: the monopolist first chooses its signal precision, then observes the signal realization s and thereafter chooses its product variety x and a price p. After these events, consumers observe the monopolist’s product variety and price, sales occur, and profits and utilities are realized. A strategy for the monopolist is a signal precision and, given any signal precision and any subsequently observed signal value, a product variety and price. Formally, a strategy is a pair
, where
precision-signal pair
and
some variety-price pair
assigns to each .
For any signal precision q, signal value s, product variety x and price p, let be the monopolist’s posterior expected profit (after the signal has been received), and let
be the posterior expected welfare, where
welfare is defined as consumer surplus plus monopoly profit. Let
and
be the associated ex ante expected profit and welfare (before the signal is received) under strategy
.
The monopolist’s goal is to maximize a convex combination of these two latter expectations. Formally, it solves the optimization programme
(4) Page 5 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly where
is an exogenous parameter. At one end of the parameter
spectrum,
, we find the unregulated profit-maximizing monopolist while at
the opposite end of the spectrum,
, we find the welfare-maximizing
monopolist, the monopolist that implements the first-best solution (in terms of welfare). As will be shown in Section 5 for an intermediate parameter value, , we will find the second-best public monopoly, that is a welfaremaximizing monopolist that faces the budget constraint that its ex ante expected profit be non-negative. We note that optimality in Equation (4) requires that the monopolist’s choice of product variety and price be optimal after (almost) all possible signal realizations.1 We also note that, since we have normalized the consumer population to unity, all expected values above are bounded from below by
and
from above by consumers’ average valuation, , of their ideal product varieties minus the marginal cost of production. The latter claim follows from the observation that in the ‘best of worlds’ production costs are nil and each consumer obtains his or her ideal product variety for free, in which case . Hence, in general (p.124)
We proceed first to analyse the case to all intermediate cases
, then the case
, and finally turn
. Henceforth, let Φ denote the CDF of the
standard normal distribution
and let ϕ be its density.
3 Unregulated Profit Maximization We here consider an unregulated profit-maximizing monopolist; the case
.
Optimality in Equation (4) requires that the monopolist’s choice of product variety and price be optimal after (almost) all possible signal realizations. Suppose that the monopolist has chosen signal precision q and observed a signal value s, and is about to choose a product variety and price. In such a situation, the monopolist will strive to maximize its conditionally expected profit,
(5)
The following result establishes that the monopolist’s choice of product variety is a random variable that depends linearly on the signal, while its price is deterministic and independent of the signal value. Moreover, we find that the choice of product variety is more sensitive to the signal the higher its precision. In particular, if the monopolist has (previously) chosen to acquire no Page 6 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly information,
, then it will opt for product variety
, the ex ante
expected mean value of consumers’ ideal product variety. Proposition 1 For any signal precision value
and any observed signal
, the profit-maximizing monopolist (
) will choose product
variety
(6)
and set its price
so that
(7)
Moreover, the set on the right-hand side of Equation (7) is a non-empty and compact subset of the open interval
.
Proof Conditional upon an observed signal value s, the monopolist’s posterior for
is normally distributed,
(p.125)
By assumption, the idiosyncratic taste parameter ξ is statistically independent of both
and s, so
and thus
The conditional random variable around its mean value,
is symmetrically and unimodally distributed , so for any
choose a product variety x such that, for any valuation interval of ‘willing buyer types’,
, the monopolist will , the associated , is centred on μ.
Hence, it is optimal for the monopolist to select product variety
as given in
Equation (6). Given this choice of product variety, and conditional upon the realized signal s, the monopolist chooses its price p so as to maximize
Page 7 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly where
is the variance of
. After a change of
variables one obtains Equation (7). The maximand in Equation (7) is a continuous function of p that vanishes at
.2 Moreover, the
and as
maximand is positive at intermediate values of p. Hence, by the Weierstrass maximum theorem, the set of maximizers is a non-empty and compact subset of the open set V. Q.E.D. Hence, the monopolist’s optimal variety/price strategy for all signal values,
and all for all
henceforth thus write
for
satisfies
, and
is constant across
and s,
. We will
. Since every optimal price is interior
(with respect to the support of the valuation distribution), the price p has to satisfy the first-order (p.126) condition that the derivative of the maximand in Equation (7) with respect to p, is zero. This condition in fact uniquely determines the pricing strategy
, to be specified in the next result. Let
(8)
This is the square root of the precision of the posterior (after the signal) estimate of individuals’ ideal product varieties. Proposition 2 For every
the monopolist’s price,
, is the
unique solution of the equation
(9)
Moreover, the solution lies in the open interval
.
Proof: The left-hand side of Equation (9) is the derivative of the maximand in Equation (7) with respect to p. Writing Equation (9), note that
for the left-hand side in
is continuous and satisfies
. Moreover, the integrand in Equation (9) is point-wise decreasing in p, for any given
, and the interval of integration is shrinking, so k is strictly
decreasing. This establishes the claimed uniqueness. Existence follows from Proposition 1 and the fact that every profit-maximizing price is interior and hence necessarily meets the first-order condition in Equation (9). Q.E.D. Equation (9) determines the optimal price by comparing the benefits and costs of changing price, and the optimal price is determined by the condition that Page 8 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly there are no gains by changing price. However, the benefits and costs of changing price are affected by the valuation distribution; for more details, see Weibull and Chen (2016). Therefore, other effects on the price will be affected by the valuation distribution. From Equation (9) we know that price is impacted by
, the square root of the precision of the posterior estimate of the
consumers’ ideals. However, the effects of
is affected by the valuation
distribution. More precisely, if consumers’ valuation is concentrated around the equilibrium price or lower than the equilibrium price, increases in
force the
monopolist to decrease the price; on the other hand, if consumers’ valuation is concentrated above the equilibrium price, increases in
will force the
monopolist to increase the price. For details, see Weibull and Chen (2016). We are finally in a position to consider the monopolist’s choice of signal precision, or, in other words, how well-informed it wants to be about consumer preferences. The monopolist will evidently never choose a signal precision so high that its cost cannot be recovered by its sales. Since sales revenues (p.127) cannot exceed consumers’ mean valuation, , and since information costs are non-negative and strictly increasing, the Weierstrass maximum theorem implies that the set
of the monopolist’s optimal signal precisions
is non-empty
and compact and can be characterized as follows:
(10)
where
. Moreover, since the marginal cost of information
acquisition by assumption is zero at zero precision, any optimal signal precision must be positive. In sum: Proposition 3 The set interval
. If
is a non-empty and compact subset of the open , then
satisfies the first-order
condition
(11)
evaluated at
.
In sum, then, the monopolist’s decision problem has a solution, and once a signal precision price
has been found, Equation (9) will produce a unique monopoly associated with that signal quality.
Page 9 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly 4 First-Best We now turn to the first-best case,
, that of a monopolist striving to
maximize welfare, but otherwise is identical with a private monopoly. In particular, the monopolist faces for example the same informational constraints and costs as before. Moreover, this public monopolist does not face any budget constraint; it is as if its costs can be covered by a lump-sum tax on all individuals in the consumer population, irrespective of whether or not they actually consume the good or not. As we will see, this type of monopolist will set the price of the good at unit production cost and raise all funds by way of a lumpsum tax. Just as in the case of private monopoly, we solve the monopolists’ decision problem by backward induction. Suppose, thus, that the monopolist has already chosen its signal quality
and observed the signal s. For any product variety
x and price p he/she may choose, the resulting expected welfare, conditional upon
, is
(12) (p.128)
where H is the CDF of the conditional random variable
. It is not difficult to
verify that a welfare-maximizing monopoly, once it has chosen its signal precision and observed its signal, will choose the same product variety as does a profit-maximizing monopolist, but will set its price equal to unit production cost. Proposition 4 For any given signal precision signal value
and any observed
, the welfare-maximizing monopolist (
) will choose
product variety
and sets price The intuition behind the difference between the private monopolist’s and the public monopolist’s pricing is simple; for the public monopolist a higher price has no benefit, since any potential rise in its profits would be matched by an equally large reduction in consumer surplus. Hence, a higher price is only potentially harmful for the unconstrained welfare-maximizing monopolist. In effect, it covers its costs by a lump-sum tax on all consumers. Moreover, the monopolist does not want to produce for the consumers whose valuation is less than the marginal cost of production since that consumption decreases social welfare.
Page 10 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly Using Proposition 4, one immediately obtains the following expression for the ex ante expected welfare (as expected before the signal has been observed):
(13)
where
is the optimal decision rule defined in Proposition 4 and the expectation
on the right-hand side is taken over the (normally distributed) signal. The double integral in Equation (13) is differentiable and strictly increasing in
. Hence
this term is strictly increasing and differentiable in q. Let denote the set of socially efficient signal precisions. The Weierstrass maximum theorem guarantees that there exists at least one such signal precision, and, in fact, that this set is compact. Proposition 5 The set or
is non-empty and compact. If
for some
, then either
that satisfies
(14)
Clearly, the welfare-maximizing monopolist’s signal precision
is positive if the
marginal cost of information acquisition at zero precision is zero, which is the case on our parametric specifications in the numerical simulations.
(p.129) 5 Intermediate Cases We now turn to the canonical intermediate case when spans from private monopoly (
. This case thus
, Section 3) to first-best public monopoly (
, Section 4). As will be seen, we obtain the solutions for all public monopolies that face any exogenous budget constraint. This setting is in line with the classical Ramsey–Boiteux approach; see Ramsey (1927) and Boiteux (1971). More precisely, the monopolist here strives to maximize expected welfare, under the constraint that its expected profit reaches a pre-specified level, to be denoted B; the arguably most relevant case being that of selffinancing, that is,
. But first we solve Equation (4) for any given
.
As in the two preceding cases, we solve this monopolist’s decision problem by backward induction. We have already shown that for any given signal precision q and price p, and after observing the signal s, welfare and profits are both maximized at product variety
. The following result is thus a
direct application of those conclusions.
Page 11 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly Corollary 1 For any given signal precision value
and any observed signal
, the optimal product variety for the generalized monopoly is
Given the optimal product variety (a random variable), the social welfare conditional on signal s is given by
where the first term,
, is expected consumer surplus, conditional
on the signal, and the last two terms together make up the profit of the monopolist (and these do not depend on the signal). For any signal quality q that the monopolist may have chosen, it should thus set its price
so that solves
(15)
Taking derivative of the maximand in Equation (15), we obtain a characterization of the optimal price. Proposition 6 Given any price
and
, the monopolist’s optimal
is the unique solution of the equation
(p.130)
(16)
In sum, then, our qualitative result for the strategy monopolist holds for all
of the profit-maximizing
: the monopolist’s optimal variety/price
strategy ψ always satisfies
for all
and all
, and
is always constant across signal values.
Page 12 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly Equation (16) defines the optimal pricing rule of notation, write as
, which we, with a slight abuse
. The marginal benefit of increasing price for the
monopolist is a weighted sum of the benefit to social welfare and the benefit to the monopolist’s profit. As we have seen, increasing the price generally decreases welfare, so the marginal benefit of price to the monopolist in this intermediate case is smaller than for the pure profit-maximizing monopolist, who places weight
on consumer surplus. This implies that, given any signal
precision q, the intermediate monopolist will charge a lower price than the profit-maximizing monopolist. The monopolist should choose its signal precision
so that it solves
(17)
where
and
since both
and
are independent of s in our
model. Applying the envelope theorem and differentiating, we obtain a characterization of the optimal signal precision for any given Proposition 7 Given or
:
, the monopolist’s signal precision is either
for some
that satisfies the equation
(18)
The marginal benefit of private information acquisition is also a weighted sum of the benefit to social welfare and the benefit to the private monopolist’s profit. Sections 3 and 4 show us that the comparison between the marginal benefit to social welfare and the marginal benefit to the firm’s profit depends on the spread of the consumers’ valuation. Therefore, the inefficiency of the profitmaximizing monopolist’s level of information, as compared with the first-best information level, is arguably mainly determined by this spread. (p.131) 5.1 Second-Best: Budget-Constrained Welfare Maximization
We now turn to the case of a budget-constrained welfare-maximizing monopolist, or, more precisely, a monopolist that chooses its strategy 〈 expected welfare,
〉 so as to maximize
, subject to the constraint that its expected profit,
, is at least B, where
is exogenously given. The case
is of
Page 13 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly particular interest, since it represents a familiar second-best situation, that of a monopolist who strives to maximize welfare under the requirement that it should cover its costs by its own revenues. Let
be the profit obtained by the monopolist in Section 3 (
the profit obtained by the monopolist in Section 4 ( , let
), let
be
), and for each
be the profit obtained by the monopolist in the present section. We
then have
, and
is continuous in γ for
. (The last
claim follows from Berge’s maximum theorem.) Hence, by continuity there exists at least one
for which
Proposition 8 Let
.
be such that
. A monopolist that maximizes
subject to the budget constraint same strategy 〈
〉 as defind above, for
will choose the , and the budget
constraint will be precisely met. That is to say, we find the budget-constrained monopolist’s behaviour by picking the parameter value γ so that the budget constraint is exactly met. The intuition behind this result is simple: the Lagrangian associated with the budgetconstrained monopolist’s decision problem can be written as
where
is the Lagrangian multiplier associated with the budget constraint.
By setting
, maximization of
, given
identical with solving our intermediate monopolist’s problem when by definition of this particular γ-value,
, is . And
, so the budget constraint is
met. To see the intuition why the budget constraint is necessarily precisely met, suppose, by contradiction, that the constraint were not binding, so that the monopolist now earns a profit above
. Then, given its signal precision
,
the monopolist could reduce its price slightly without violating the budget constraint (by continuity), and this way obtain a slightly higher social welfare, since the latter is monotonically decreasing in the price, as shown in the proof of Proposition 4.3
(p.132) 6 Numerical Simulations In order to illustrate the above results, we will now consider numerical simulation results for the special case of a log-normal value distribution and information costs that are a positive power of the signal precision. As will be seen, we will be able to numerically identify the monopolist’s unique signal precision and price, and make comparative statics experiments with these.
Page 14 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly 6.1 Profit Maximization
We now turn to our numerical simulation results. Assume that some
and
for
. Furthermore, the consumers’ valuation is assumed to be and standard deviation σ.4
log-normally distributed with mean
The first batch of our numerical simulations is focused on comparative statics when consumers’ valuation distribution is given. During the second batch, we assume that consumers’ valuation is log-normally distributed with mean and standard deviation
.
Figures 8.2a and 8.2b show the optimal information acquisition and optimal price as functions of α when
,
,
and
. Figure 8.2a shows that a
more precise prior for the shared component crowds out the monopolist’s incentive for information acquisition. Moreover, with linear information acquisition cost function, when α is small enough, the private information acquisition and α are perfect substitutes. Furthermore, information acquisition decreases at a lower speed than the increase of α such that the monopolist has a more precise posterior estimate of the consumers’ ideal products. With our parameter, this decreases the monopolist’s price in equilibrium.
Page 15 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly Figures 8.3a and 8.3b show the numerical solutions as functions of β when and
,
,
. Figure 8.3a shows
that the more concentrated are consumers’ idiosyncratic ideal product varieties, the more private information the monopolist will acquire. The intuition behind the result is that, with more concentrated ideal product varieties, private information is more valuable since it helps estimate more consumers’ ideal products. Therefore, large concentration on consumers’ ideal products increases the monopolists’ precision of posterior estimate of consumers’ ideal products, and this will decrease the monopolist’s price in equilibrium.
Figure 8.2 The effect of shared component precision
Page 16 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly We now turn to our second batch of numerical simulations, in which we study the effect of a mean-preserving spread of consumers’ valuation distribution. The reason why this may be interesting is that in general a monopolist’s profit, at the monopoly price, is decreasing in the spread of consumers’ (p.133) valuation for the good in question. To see this, suppose, first, that all consumers have virtually the same valuation,
, and that
they all have product variety as their ideal, and furthermore the unit production is 0. If the monopolist could know this, he/she could extract almost all consumer surplus by setting the price p just below 50 and make profits close to 50. Next, suppose instead that consumer valuations are uniformly distributed between zero and 100, and that again all consumers’ ideal is the variety . A monopolist would know
Figure 8.3 The effect of idiosyncratic this and could only extract half component precision the consumer surplus. He/she would again (p.134) optimally set p close to 50, but now only half the consumers would buy, so his/her profits would be only 25. Hence, one may ask if this qualitative relationship holds also in our more general model. If it does, then the monopolist might invest less in information acquisition. To find this out, we now assume that
and
consumers’ valuation v is log-normally distributed with mean
, and that and variance
.
For our purpose, for each , we investigate how the change of σ, when keeping (p.135)
constant, would affect the equilibrium information acquisition and
price. Figure 8.4a shows that the qualitative relationship in the above very parsimonious model still holds in our more general settings: more dispersion in
Page 17 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly consumers’ valuation decreases monopolist’s incentive in private information acquisition. Figure 8.4b shows the non-monotonic effects of more dispersion in valuation distribution on equilibrium price. This figure integrates two effects of more dispersion on the equilibrium price: the first one is that more dispersion decreases the equilibrium information acquisition, and therefore decreases the (p.136) posterior estimate precision of the individual’s ideal products; the second one is that by changing the consumer’s valuation distribution, the monopolist needs to adjust the price. The first effect on the price may be nonmonotonic, as we have said before, depending on the valuation distribution. However, the second effect may dominate the first one for the dispersion effects on price. Figure 8.5 shows that, given
Figure 8.4 The effect of value dispersion
Page 18 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly ,
and
, how
optimal price is adjusted as the standard deviation of consumers’ valuation distribution changes. We can see that, consistent with Figure 8.4b, the monopolist decreases the price when σ is between 0 and around 20 and then increases the price when σ is above 20. To see the effects of σ on price shown in Figure 8.5, we can interpret the adjustment process in terms of Equation (9). When σ is zero, all consumers have valuation setting
Figure 8.5 Revenue and price given the signal precision
and the optimal price is shown in Equation (9) by
and ignoring the integral. As σ increases, the mode of consumers’
valuation distribution decreases (see footnote 4), so the monopolist prefers a lower price to attract consumers around the mode. As σ increases, the mode of the valuation distribution is so low that it becomes less valuable to attract consumers with valuation around the mode; the monopolist prefers to increase the price to serve only the consumers with high valuation. 6.2 First-Best
With numerical methods, we can compare the monopolist’s optimal information acquisition with the first-best. Weibull and Chen (2016) have shown that, although α, β, a, and c all have some effect on the inefficiencies of the monopoly in acquiring private information, they do not play a critical role in (p.137) determining whether the monopolist acquires too much or too little private information. Figure 8.6 shows that the dispersion of consumers’ valuation distribution determines the inefficiencies of the monopolist’s information acquisition. When consumers are homogeneous in their valuation of the good, the monopolist invests too much in information acquisition. On the other hand, when consumers are very heterogeneous in their valuation, the monopolist has not enough incentive to acquire private information.
Page 19 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly To see why the valuation dispersion affects the inefficiencies of monopolist’s information acquisition, we can compare the monopolist’s and first-best marginal benefit of information acquisition. Note that the marginal benefit of information acquisition in Equation (14) comes from attracting more consumers with a more precise estimate of individual’s ideal product. For each valuation v above the unit production cost c, the marginal benefit from attracting more
Figure 8.6 The effect of value dispersion on information acquisition
consumers by investing in information acquisition is
(19)
Furthermore, by a change of variables, the corresponding first-order condition Equation (11) for the private monopolist’s information acquisition can be reformulated as
This is the marginal benefit derived from attracting more consumers by investing in information acquisition multiplied by the benefit of selling out one more (p.138) unit. From the expression above, we can see that for each valuation v, the marginal benefit from attracting more consumers by investing in information acquisition is approximately to be
(20)
Comparing the marginal benefit of attracting more consumers in Equations (19) and (20), we know that for each v, the first-best marginal benefit of attracting more consumers monotonically increases in v, and as v goes to infinity, it goes to . The private monopolist’s marginal benefit of attracting more consumers is monotonically decreasing in v when v is large enough, and it tends to zero as v goes to infinity. Therefore, there exists one
such that for all
Page 20 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly , the first-best marginal benefit is larger than the private monopolist’s, and the discrepancy between these two increases. As σ increases, increases for high valuation, implying more consumers endowed with high valuation. Therefore, the discrepancy between the total marginal benefits from attracting more consumers is increasing in σ. Furthermore, in our numerical simulations, generally
is much larger than 1; this increases monopoly’s marginal benefit
of information acquisition. When σ is small, the discrepancy between the total marginal benefits from attracting more consumers is not large enough to offset the high price in equilibrium; in this case, the monopoly has a higher marginal benefit of information acquisition. On the other hand, when σ is large enough, the discrepancy is large enough to offset the effect of high price; in this case, the monopoly has not enough incentive to acquire enough private information. Remark 1 The numerical results in this subsection are based on the lognormally distributed valuation; however, since the comparison between the marginal benefit of attracting more consumers is independent of consumers’ valuation distribution, the results on the effects of the dispersion of consumers’ valuation on inefficiencies in monopoly’s information acquisition are robust to any kind of distribution. 6.3 Second-Best
In this subsection we numerically solve for information acquisition and pricing when the goal function is a convex combination of the monopolist’s profit and social welfare. The numerical solutions are shown in Figures 8.7 and 8.8; we assume a quadratic information acquisition cost function and that v is log-normally distributed with mean
and
,
. We further write
and assume that both the variances of shared component and idiosyncratic component are 1. We will refer to the parameter κ, which is a monotone transformation of the parameter γ, as the (p.139) welfare weight in the monopolist’s goal function. We will analyse the case when so that we can see how the changes in weight affect the equilibrium pricing and information acquisition.
Page 21 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly Figure 8.7 shows that the monopolist caring about both profit and welfare prefers a price between the private monopolist’s optimal price and the first-best price. When γ is close to 1, the price is close to the first-best price; on the other hand, when γ is not large enough, the price is close to the private monopolist’s optimal price; more specifically, the monopolist’s optimal price is monotonically increasing in the weight γ. Furthermore, since the profit-maximizing monopolist’s optimal price may
Figure 8.7 Pricing as a function of value dispersion for different welfare weights
Figure 8.8 Information acquisition as a function of value dispersion for different welfare weights
be non-monotonic in the valuation dispersion, and the marginal benefit of the monopolist caring about both welfare and profit is a weighted sum of the marginal benefit to private monopoly and marginal benefit to social welfare, the optimal price for the monopolist caring for both social welfare and profit and specifically the second-best price, depending on the weight γ and other parameters, may also be non-monotonic in valuation dispersion. Figure 8.8 illustrates the relationship between information acquisition and valuation dispersion σ for different values of κ. When κ is large enough, the monopolist puts more weight on social welfare and the information acquisition in equilibrium is more close to the first-best information acquisition; on the other hand, when κ is very small, the monopolist behaves more like the private monopolist in information acquisition. Furthermore, although both the private monopolist’s and first-best information acquisition are monotonic in consumers’ valuation dispersion, the information acquisition of the monopolist caring about both profit and social welfare may be non-monotonic in the (p.140) dispersion. Specifically, Figure 8.8 shows that when κ is not large enough, the monopolist’s information acquisition is increasing in the dispersion when the dispersion is not large enough; and, contrarily, when the dispersion is large enough, the monopolist’s information acquisition is decreasing in the dispersion. One final comment is that, although the marginal benefit of information acquisition for the monopolist is a weighted sum of the marginal benefit to profit and the marginal benefit to social welfare, the optimal information acquisition of the monopolist who cares about both the profits and social welfare may be higher than both the private monopolist’s and first-best information acquisition. Page 22 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly This happens when the weight is not large enough and consumers’ valuation dispersion is intermediate. The intuition of this result is that, when the dispersion σ is relatively large, the marginal benefit from attracting more consumers is larger than the private monopolist; at the same time, the price of the monopolist is much higher than the first-best: in this case, the marginal benefit of information acquisition is higher than both the marginal benefit of the private monopolist and the first-best marginal benefit, resulting in an information acquisition which is higher than both private monopolist’s optimal information acquisition and the first-best information acquisition. Furthermore, when the dispersion is quite small, the dispersion is not large enough to generate a high marginal benefit to social welfare by attracting more consumers to offset the effect of the high price; in this case, the monopolist caring about both profits and social welfare acquires less information than the private monopolist and more information than the first-best. On the other hand, when the dispersion is large enough, the (p.141) marginal benefit to social welfare from attracting more consumers is large enough to offset the effect of high price; in this case, the monopolist caring about both profits and social welfare acquires more information than the private monopolist and less information than the first-best.
7 Extensions Our model is very simple and based on heroic assumptions. We briefly discuss here a few directions in which the model may be generalized: the dimensionality of product variety, multiplicity of product variety, and oligopolist competition. For while the monopolist in our model only supplies one variety and the space of varieties is unidimensional, in practice monopolists usually provide a whole menu of product varieties where each variety has many attributes and hence is multidimensional. Moreover, while competition is absent from our model, in practice there is competition or at least the threat of potential entrants. While an analysis of the mentioned general cases appears quite a challenge and falls outside the scope of the present study, we briefly sketch here how the present model can be generalized. First, in order to capture multiplicity and multidimensionality of varieties, the most natural generalization is arguably to let the monopolist choose a finite menu
of product varieties and prices for these but
maintain the hypothesis of unit demand. In such a setting each consumer chooses which variety, if any, to buy. Thus let each variety
be a vector in
for some positive integer k, the dimensionality of varieties, and assume a fixed cost
associated with each variety, alongside its unit cost of production (where costs thus may depend on the variety). The
monopolist’s decision problem then is to choose a menu M from among the set of all finite menus, or, in other words, how many and what varieties, and what price for each variety. Given a goal function of the sort Page 23 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly analysed in the preceding sections, this is a challenging optimization problem. However, given such a menu
, offered by the
monopolist, each consumer’s choice is simple, namely, to either buy no units or to buy one unit of the variety that gives most utility. To be more specific, let a consumer type be a pair type
and let the utility for a consumer of
from purchasing any variety
at any price
be defined as in
Section 2;
and let
(p.142)
Thus
is the utility a consumer of type
obtains from any menu M
offered by the monopolist. The consumer will buy a unit if and only if . As in Section 2, one may, for analytical tractability, treat θ and v as statistically independent random variables, each with an absolutely continuous probability distribution. If θ is multivariate normal and the monopolist’s signal is additive with normally distributed noise, and if the fixed and marginal costs are the same for all varieties, then much of the preceding machinery applies, although with substantial mathematical challenge. However, for any given fixed production cost, d (assumed to be the same for all varieties), an upper bound on the number m of varieties is
. In order to solve (at least numerically) the
monopolist’s decision problem, one may then proceed by solving it first for each , and thereafter choose the right number, of varieties. Moreover, at least in the case of unidimensional varieties ( the following heuristic could be explored. For that is, choose variety distance from
. For
),
solve as in Sections 3 to 5,
, place instead the two varieties at equal
, and optimize over this distance. For
, choose variety 1
to be and place varieties 2 and 3 on each side of 1, at equal distance from , and optimize over this distance, etc. Second, an analytically more tractable model extension that allows for multiple, but unidimensional, varieties would be to let the monopolist select a whole ‘product line segment’
as in Mussa and Rosen (1978), and
require it to charge a uniform price p for all varieties in this line segment L. In this case, it is easily verified that the monopolist’s optimal location of the product line segment is to centre it on
, that is, to pick
follows from the analysis above of the special case
. This
in combination with the
observation that now the demand function generalizes to
Page 24 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly
The optimality of the choice
holds when the width
of the
product line segment is exogenous. Suppose the width is endogenous and if the monopolist does not only have a constant unit cost an increasing cost
associated with the width
of production but also of its product line, then t
can be determined by a first-order condition equating the marginal cost of t, , to its marginal benefit to the monopolist (see the models in Mussa and Rosen (1978) and Spence (1975)). As shown in Spence (1976a, 1976b), this choice of t will typically be socially inefficient when the monopolist’s information about consumer preferences is fixed. An interesting avenue for further research is thus to explore this potential inefficiency in the case of endogenous information. (p.143) Third, a few words about oligopolistic competition. An extension in such a direction would be extremely interesting. The fundamental underlying question is then the effect of competition on producers’ incentives to find out about consumer preferences. More precisely, suppose now that there are two profit-maximizing firms, each modelled along the lines of the profit-maximizing monopolist in Section 3, and that each firm first has to choose its signal precision, then its product variety and price. A number of alternative scenarios are possible here. Arguably, the most basic and canonical scenario is that of simultaneous-move duopoly, that is, two firms who simultaneously first choose their signal precisions, keep their signal precisions and signal realizations as their private information, and then simultaneously choose their own product variety and price. What can then be said? A conjecture for symmetric Nash equilibrium is that they will each choose the same signal precision,
,
that they will choose their own product variety as under monopoly, that is and price,
. If so, what is then
, and that they will choose the same and
, and how do these relate to the
profit-maximizing monopolist’s choice and to the first-best monopoly choice? And is the conjecture at all true? To explore these questions appears as a natural next step in this exploration of producers’ endogenous information about consumer preferences.
8 Concluding Remarks This chapter examines a monopolist’s production variety choice, price setting, and information acquisition. In our model the marginal cost of production of the good is constant, and the monopoly will choose the efficient variety. Compared with first-best, the profit-maximizing monopoly sets too high a price. However, the inefficiency in private monopoly’s information acquisition depends on the parameters, especially on the consumer’s valuation distribution. With lognormally distributed valuation, our numerical results show that there is a critical Page 25 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly point of the spread such that the private monopoly acquires too much information when the spread is smaller than the critical point and the opposite is true when the spread is larger than the critical point. Compared with the second-best price setting and the information acquisition where the monopoly’s profit is constrained, the profit-maximizing monopoly will generally set too high a price, and the inefficiency in information acquisition is still determined by the consumers’ valuation spread. The results in the chapter are still not as general as we would wish, and some of them rely on numerical method. More general results on the monopoly’s choice under consumers’ preference uncertainty need further investigation. For instance, while we focus here on the entire analysis of endogenous (p.144) information about consumer preferences over product varieties, a relevant consideration is that of endogenous information about consumers’ valuations of their ideal product variety. There are other important factors that may affect the value of monopoly’s private information. For instance, threat and competition from a potential entrant may also affect the monopoly’s information acquisition choice (see Dimitrova and Schlee, 2003). To incorporate these factors into the model studying monopoly’s information acquisition would be extremely interesting. Acknowledgements The authors are grateful for comments from Francois Salanié, Xavier Vives, Paul Segerstrom, Jens Josephson, and Chloé Le Coq. A preliminary version was circulated under the title ‘How interested is a monopolist in consumer preferences?’. Financial support from Knut and Alice Wallenberg Research Foundation is gratefully acknowledged. References Bibliography references: Armstrong, M. and D. E. Sappington (2007), ‘Recent developments in the theory of regulation’, in M. Armstrong and R. Porter (eds) Handbook of Industrial Organization, Vol. 3. Amsterdam: North-Holland. Basu, K. (1988), ‘Why monopolists prefer to make their goods less durable’. Economica 55(220): 541–6. Boiteux, M. (1971), ‘On the management of public monopolies subject to budgetary constraints’, Journal of Economic Theory 3(3): 219–40. Casado-Izaga, F. J. (2000), ‘Location decisions: the role of uncertainty about consumer tastes’, Journal of Economics 71(1): 31–46.
Page 26 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly Dimitrova, M. and E. E. Schlee (2003), ‘Monopoly, competition and information acquisition’, International Journal of Industrial Organization 21(10): 1623–42. Harter, J. F. (1997), ‘Hotelling’s competition with demand location uncertainty’, International Journal of Industrial Organization 15(3): 327–34. Król, M. (2012), ‘Product differentiation decisions under ambiguous consumer demand and pessimistic expectations’, International Journal of Industrial Organization 30(6): 593–604. Meagher, K. J. (1996), ‘Managing change and the success of niche products’. Working Paper 96-08-066. Santa Fe, NM: Santa Fe Institute. Meagher, K. J. and K. G. Zauner (2004), ‘Product differentiation and location decisions under demand uncertainty’, Journal of Economic Theory 117(2): 201– 16. Meagher, K. J. and K. G. Zauner (2005), ‘Location-then-price competition with uncertain consumer tastes’, Economic Theory 25(4): 799–818. Meagher, K. J. and K. G. Zauner (2011), ‘Uncertain spatial demand and price flexibility: A state space approach to duopoly’, Economics Letters 113(1): 26–8. (p.145) Mussa, M. and S. Rosen (1978), ‘Monopoly and product quality’, Journal of Economic theory 18(2): 301–17. Ramsey, F. P. (1927), ‘A contribution to the theory of taxation’, Economic Journal 37(145): 47–61. Spence, A. M. (1975), ‘Monopoly, quality, and regulation’, Bell Journal of Economics 6(2): 417–29. Spence, M. (1976a), ‘Product differentiation and welfare’, American Economic Review 66(2): 407–14. Spence, M. (1976b), ‘Product selection, fixed costs, and monopolistic competition’, Review of Economic Studies 43(2): 217–35. Vives, X. (1984), ‘Duopoly information equilibrium: Cournot and Bertrand’, Journal of Economic Theory 34(1): 71–94. Weibull, J. and J. Chen (2016), ‘Private versus public monopoly’. mimeo. Notes:
(1) Since the signal by assumption has an absolutely continuous probability distribution, the goal function is unaffected by deviations from optimality on any subset of signal values with Lebesque measure zero.
Page 27 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Private versus Public Monopoly (2) To see that the maximand tends to zero as p tends to plus infinity, write the maximand as and note that
Since
, the last integral has to converge to zero as
. To see the
latter claim, note that
where the first integral converges to
as
.
(3) Actually for any B less than the profit of the unregulated private monopoly and more than the profit of public monopoly, there is one such that , the budget constraint will be precisely met. (4) In this case, denote the consumers’ valuation by v, then distributed with mean
is normally
and variance
Access brought to you by:
Page 28 of 28
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
US–Japanese Differences in Employment Practices New Explanations and Predictions Hodaka Morita
DOI:10.1093/oso/9780198812555.003.0009
Abstract and Keywords This chapter shows that a model that captures the interconnections between firm dynamics, labour mobility, and specific human capital provides new explanations for and predictions on the US–Japanese differences in labour mobility, wage structure, and firm-sponsored training, based on cross-country differences in the importance of managerial capability. My argument is based on the idea that managerial capability increases its importance as an economy or an industry approaches the technological frontier. It also provides complementary explanations and predictions based on governmental interventions in firm dynamics, given that a guiding principle of Japanese industrial policy has been the regulation of so-called ‘excessive competition’. Keywords: entry regulations, firm dynamics, labour mobility, managerial capability, specific human capital, US–Japanese differences, wage structure
1 Introduction US–Japanese differences in employment practices have attracted significant attention in the literature of international comparison as to how internal labour markets operate (see for example Waldman, 2013). The United States (US) and Japan have been considered representing two contrasting employment systems, and the Japanese system was often celebrated as a major source of the success of Japan’s economic development in the post-war growth era (Kambayashi and Kato, 2017). In contrast, however, Japan’s Great Recession (which lasted Page 1 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices approximately from 1992 to 2007) has led many to question the effectiveness of the Japanese employment system for contemporary and future economy (Ono and Rebick, 2003; Kambayashi and Kato, 2017). This chapter proposes a new theoretical explanation for US–Japanese differences in the post-war growth period, where I focus on the differences in labour mobility, firm-sponsored training for specific human capital, and wage structures. It then discusses its predictions and implications for the differences in the contemporary and future economy. As a theoretical framework, I use Morita’s (2012) model of firm dynamics that incorporates workers, their accumulation of specific human capital, and their mobility. Firm dynamics generate labour mobility because workers move from contracting and existing firms to expanding firms and new entrants. Also there are connections between labour mobility and firm-specific human capital investment because specific human capital possessed by a worker loses its value if the worker leaves his (p. 147) current employer. Elaborating on these interconnections, I show that the model provides new explanations for the differences. 1.1 US–Japanese Differences in the Post-War Growth Period
In the process of Japan’s economic development in the post-war period, labour mobility was lower and earnings-tenure profiles were more steeply sloped in Japan than in the US, and Japanese firms tended to provide higher levels of specific human capital through intensive firm-sponsored training compared to their US counterparts. Concerning labour mobility, it was found that the labour turnover rate was much higher in the US than in Japan (see Hashimoto and Raisian, 1985; Mincer and Higuchi, 1988). Hashimoto and Raisian (1985) for instance found that the 15-year job retention rates of the male population between the early 1960s and the late 1970s were much higher in Japan than in the US across all age groups. Regarding wage structures, Hashimoto and Raisian (1985) analysed data from the 1980 Basic Survey of Wage Structure for Japan, and data from the 1979 Current Population Survey for the US, and found that earnings-tenure profiles were more steeply sloped in Japan than in the US (see also Mincer and Higuchi, 1988). That is, they found that growth rates in earnings attributable to firm-specific tenure were substantially greater in Japan than in the US for all firm-size groups. Concerning firm-specific human capital, Koike (1977, 1988) found, in his comparative analysis of Japanese and US industrial relations, that Japanese workers acquired more firm-specific human capital through rotation among related jobs (see also Dertouzos et al., 1989; Ito, 1992). Also, levels of firmsponsored training were higher in Japanese firms than in US firms as pointed out by a number of authors such as Mincer and Higuchi (1988), Dertouzos et al. (1989), Ito (1992), and MacDuffie and Kochan (1995). According to MacDuffie and Kochan (1995), the International Assembly Plant Study, a survey of ninety motor vehicle assembly plants in twenty-four countries, found that Japanese Page 2 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices companies’ plants in Japan gave their newly hired workers an average of 364 hours of training compared to only forty-two hours in US companies’ plants in North America. The study also found a significant difference in the training of experienced workers.
2 Explanations and Predictions This section first presents an outline of Morita’s (2012) model and its analysis. It then discusses explanations for and predictions on the US–Japanese differences that the model yields, and its contribution to the existing theoretical explanations for the differences. (p.148) 2.1 Outline of the Model and Its Analysis
Consider an industry that produces a homogeneous good in a two-period setting, where the industry faces a downward-sloping demand schedule in each period , given by
. Entry and exit of firms are free in each period,
where each firm is of measure zero so that it is too small to affect prices. The production requires labour input; a firm can produce one unit of the good in a period if it employs one worker in that period, and the firm can produce nothing otherwise. There is a large number of ex ante identical individuals. Each individual can earn a reservation wage of
per period in a competitive
labour market outside this industry. Firms and individuals are both risk neutral. Each firm’s production efficiency is determined by its managerial capability and the level of its worker’s firm-specific human capital, where managerial capability is interpreted as representing the capability of a firm’s top management to develop an effective strategy and create a unique competitive position. Let a denote the managerial capability and
denote the realization of firm i’s
managerial capability, which is a random draw from a uniform distribution between 0 and 1. Assume that
is ex ante unknown to all agents including firm i
itself and becomes common knowledge at the end of the first period of firm i’s operation. This specification is consistent with the widely held view that the capability of a firm’s top management is mostly innate and difficult to observe or assess ex ante. At the beginning of period 1, firms simultaneously make first-period wage offers to the individuals. Each individual can apply to a firm for first-period employment, and each firm employs one individual from the applicants. If an individual is not employed by the firm, he/she can earn a reservation wage for period 1. If firm i employs a worker, it can provide the worker with a level of specific human capital denoted (
by incurring a convex cost of
) per employee. The level of firm-specific human capital is
observable but not verifiable, and so wage contracts contingent upon it are not feasible. To keep the analysis simple, assume that firm-specific human capital affects the production efficiency only in the next period. Page 3 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Each firm i that employed a worker in period 1 produces one unit of the good. At the same time, firm i’s managerial capability
is realized and becomes common
knowledge. If a firm that operated in period 1 continues to operate in period 2, the return from its first-period investment in specific human capital is shared with its employee through the generalized Nash bargaining process. At the same time, a firm that did not operate in period 1 can employ an individual by making wage offers. Each firm that employed a worker in period 2 produces one unit of the good. Each firm i’s per-unit production cost in a period is given by and
, where
are given constants and λ captures the relationship (p.149)
between specific human capital and the production efficiency of a firm. In particular, assume that
if firm i employs worker j in that period
and employed the same worker in the previous period, while
if firm i
employs worker j in that period but did not employ the worker in the previous period. Notice that x and y capture the importance of managerial capability and firm-specific human capital, respectively, for production efficiency. In the remainder of this subsection, I present an outline of the equilibrium characterization of the model.1 Consider a perfect foresight equilibrium that is characterized by a price sequence . In the equilibrium, all agents make optimal decisions based on the anticipation of a particular price sequence , and their behaviour does in fact give rise to the same
. Given
free entry and exit of firms, for every entrant in period t (=1, 2) the present value of its expected overall profit is zero in the equilibrium. Also, given that there is a large number of ex ante identical and risk-neutral individuals, and that every individual can earn a reservation wage
per period in a competitive
labour market outside the industry, the present value of every individual’s expected overall wage is
in period 1 in the equilibrium. The market clears in
each period in the equilibrium. Morita (2012) focuses on perfect foresight equilibria in which a strictly positive number (measure) of firms enter and exit the industry at the beginning of period 2, given that in reality entries to and exits from firms are common in most industries. Throughout the analysis, it is assumed that H is large enough so that the optimal level of firm-specific human capital is interior in the equilibrium. Suppose that there exists a perfect foresight equilibrium characterized by . In the equilibrium, each second-period entrant employs a worker at the reservation wage w, and its expected production cost is
.
Since each second-period entrant earns zero expected profit in the equilibrium, must hold. Consider firm i that employed worker j and chose in the equilibrium. If firm i continues to employ worker j in period 2, its secondPage 4 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices period production cost (excluding the wage bill) is
and the firm
must pay at least w (the reservation wage) in order to employ worker j in period 2. Then firm i continues to operate in period 2 if and only if , where
if
, and 0 otherwise. Suppose that firm i’s managerial capability
turns out to be greater than or
equal to the cut-off level . Then, at the beginning of period 2 firm i and worker j bargain over worker j’s second-period wage, where the worker’s bargaining power is
and his/her threat point is the reservation wage w.
Firm i’s outside option is to exit the industry and (p.150) earn zero profit or to employ another worker with wage w and continue operating in period 2. This leads us to find worker j’s second-period wage (which is determined as the outcome of the generalized Nash bargaining process)
, where
. Here, , which is the surplus gained by reaching an agreement in the bargaining. This in turn implies that firm i’s second-period production cost plus wage bill is where
,
.
In period 1, firm i chooses the level of firm-specific human capital knowing its own managerial capability
without
. Firm i makes this choice under the
anticipation that it will continue to operate in period 2 and earn second-period profit of
if it realizes managerial capability
and will exit the industry if maximize
, where
,
. Hence firm i chooses
to
is defined by . Through the
maximization exercise we find that
must hold, and every firm i
that employed a worker in period 1 chooses
in the equilibrium, where
. This implies that the firms’ exit rate is where
in the equilibrium,
.
Every firm that employs a worker in period 1 offers the same first-period wage (denoted by
) in the equilibrium, given that firms and individuals are ex ante
identical. A worker employed by firm i in period 1 anticipates that his/her second-period wage will be
if firm i realizes
, and w if
. The first-period present value of the worker’s expected overall wage is
Page 5 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices
then
. Given that there is a large number
of ex ante identical and risk-neutral individuals, and that every individual can earn a reservation wage
per period outside the industry, firms in the
equilibrium choose
such that
holds. Hence we find that . In the equilibrium, every firm i that employs a worker at the worker with level
in period 1 provides
of specific human capital by incurring
training cost. Since its expected production cost is its first-period cost is
as a
, the expected value of
. Firm i operates in period 2 if
it realizes the managerial capability of with the second-period total cost of . Hence the first-period present value of its overall expected cost is determines the level of
. Then the following zero profit condition .
(p.151)
Given (
)=(
), the demand for the good in period t (
) is
.
Since the exit rate is , firms continue to operate in period 2. Since the second-period demand for the good is , the measure of the second-period entrants is
in the equilibrium. Then
must hold for a strictly positive measure for firms to enter at the beginning of period 2 in the equilibrium. This condition, along with
and
level denoted by
, holds if and only if y is smaller than a unique threshold . Morita (2012) shows that
is not only necessary
but also sufficient for the model to have perfect foresight equilibria in which a strictly positive number (measure) of firms enter and exit the industry at the beginning of period 2. I will discuss comparative statics results of the model in this range of parameterizations. 2.2 Explanations for US–Japanese Differences
Page 6 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices 2.2.1 Managerial Capability
Acemoglu et al. (2006) argue in their analysis of technology frontiers and firm selection that managerial skill is more important for undertaking innovative activities than for adopting and imitating existing technologies from the world technology frontier. They then point out, based on their analysis of the correlation between distance to the frontier and R&D intensity using data from the OECD sectoral database, that innovation becomes more important as the economy approaches the world technology frontier and there remains less room for adoption and imitation. Following their analysis and argument, I maintain that the importance of managerial capability was substantially lower in Japan than in the US in the process of Japan’s economic development in the post-war period when most Japanese industries were catching up with the West (Okimoto, 1989). The model yields the following prediction regarding the importance of managerial capability. As the importance of managerial capability (captured by x in the model) decreases, labour mobility decreases, firms provide their workers with higher levels of specific human capital, and the slope of the earnings-tenure profile becomes steeper in equilibrium. Then, lower importance of managerial capability in Japan than in the US provides an explanation for the US–Japanese differences in labour mobility, wage structure, and firm-sponsored training of specific human capital in the post-war period. The logic for the explanation is as follows. In period 1, several firms hire workers, provide them with specific human capital, and produce outputs. At the beginning of period 2, some first-period entrants that have realized relatively low managerial capabilities exit the industry, whereas several second-period entrants hire workers from exiting firms or outside the industry to produce outputs in period 2. A first-period entrant exits the industry in period 2 if its (p. 152) second-period cost is higher than second-period entrants’ expected cost. This is because, due to free entry of firms, second-period entrants earn zero expected profits in equilibrium, implying that second-period entrants’ expected cost is equal to the second-period equilibrium price. Each first-period entrant has an advantage over second-period entrants because it has a worker who has already accumulated specific human capital during period 1. Nevertheless, a first-period entrant exits the industry if its managerial capability turns out to be relatively low, lower than a threshold value
.
As the importance of managerial capability decreases, the threshold value decreases. This is because, as managerial capability becomes less important, a first-period entrant’s advantage associated with its worker’s specific human capital increases, implying that it does not exit the industry unless its managerial capability turns out to be very low. Lower
means lower labour
mobility because fewer first-period entrants exit the industry and hence fewerworkers switch their employers between periods 1 and 2 as
decreases.
Page 7 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Lower mobility, in turn, increases first-period entrants’ incentives to provide their workers with specific human capital, because it is more likely for them to survive and continue employing their period 1 workers in period 2. Under lower mobility, each worker anticipates, when he is employed by a firm in period 1, that his period 1 employer provides a higher level of specific human capital, and there is a higher chance for the employment to continue in period 2. This increases each worker’s expected second-period wage. Since workers anticipate higher second-period wages, firms can hire workers at lower wages in period 1, implying a steeper earning-tenure profile under lower mobility. 2.2.2 Entry Regulations
In the process of Japan’s economic development in the post-war period, the Japanese government invoked a fairly sophisticated system of interventionist economic policy to promote steady growth, technological innovations, and international competitiveness (Suzumura and Okuno-Fujiwara, 1987). A guiding principle of Japanese industrial policy has been the regulation of so-called ‘excessive competition’ (Suzumura and Kiyono, 1987), and the Japanese government often restricted entries into industries. On the other hand, the degree of such governmental intervention has been much lower in the US.2 How do entry restrictions affect welfare and the nature of employment practices? To provide an answer to this question, Morita (2012) incorporates a government that can impose entry restrictions in an extension of the model. Suppose that, at the beginning of period 1, the government announces the maximum number of firms that can enter the industry in period 2. Such an (p. 153) entry restriction reduces the second-period supply of the good, which in turn implies that the second-period equilibrium price becomes higher than the price in the free-entry equilibrium. The higher second-period price increases attractiveness of the second-period operation, which results in the following trade-off from the welfare standpoint. On the one hand, some first-period entrants with low managerial capabilities that would not survive in the freeentry equilibrium would survive in the equilibrium with an entry restriction. The resulting cost inefficiency works in the direction of decreasing welfare. On the other hand, the higher survival rate of the first-period entrants means lower labour mobility, which increases firms’ incentives for providing firm-sponsored training. This effect works in the direction of increasing welfare because it mitigates the underinvestment problem in specific human capital.3 Morita (2012) shows that the government can increase welfare by choosing a degree of entry restriction so that the latter positive welfare effect dominates the former negative one. Entry restrictions reduce labour mobility, which in turn increases firms’ incentives to provide their workers with specific human capital. And the higher level of specific human capital steepens the slope of the earning-tenure profile. Hence entry restrictions that have often been imposed by the Japanese Page 8 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices government provide an explanation, complementary to the explanation based on the importance of managerial capability, for US–Japanese differences in the postwar growth period. In addition, the model yields another prediction regarding managerial capability. Lower importance of managerial capability reduces the cut-off managerial capability
and entry restrictions have the same effect on
.
This leads to a prediction that the average managerial capability (which represents a firm’s capability to develop an effective strategy) is lower in typical Japanese firms than in their US counterparts. This is consistent with Porter et al. (2000) who argue through a number of case studies that, although many Japanese firms have achieved very high operational effectiveness, they are less capable of identifying and implementing innovative strategies to sustain their competitive advantage. 2.3 Predictions for the Contemporary and Future Economy
The Japanese economy has already caught up with the West, and most Japanese industries have got much closer to the world technology frontier. This increases the importance of managerial capability which is crucial for undertaking innovative activities. Also, the degree of entry restrictions imposed by the Japanese government has become lower in light of the recent trend of (p.154) deregulation that started in the early 1990s.4 The theoretical model then predicts that the degree of US–Japanese differences tends to become smaller. That is, labour mobility increases, the slope of the earnings-tenure profile becomes flatter, and the level of specific human capital accumulation declines in Japan. In reality, however, such transitions in Japan are likely to take place rather slowly due to the vested interests and institutional inertia embedded in the Japanese economic system. Lincoln (2001) argues that deregulation in Japan has proceeded slowly because of the strong vested interests of government bureaucrats. Several empirical studies have been undertaken recently to find out whether or not the Japanese employment system has exhibited significant changes. Regarding the earnings-tenure profile, Hamaaki et al. (2012) used micro-level data from the Basic Survey on Wage Structure for the period 1989 to 2008, and found that the age-wage profile of Japanese employees has become flatter in recent years. This finding follows up previous studies by Clark and Ogawa (1992) and Hashimoto and Raisian (1992), who found that the slope of the earnings-tenure profile for large Japanese firms became somewhat flatter in the 1980s compared with the 1970s. Findings of recent studies on labour mobility are less clear-cut. Kambayashi and Kato (2017) conducted a cross-national analysis of micro-data from Japan’s Employment Status Survey (ESS) and its US counterpart, Current Population Survey (CPS). They calculated the ten-year job retention rates of Japanese employees in the private sector for the four time periods, 1982–92, 1987–97, 1992–2002, and 1997–2007, and then used various CPS Supplements to Page 9 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices calculate the comparable ten-year job retention rates for US employees. Kambayasi and Kato found that, on the one hand, core employees (aged 30–44 with at least five years of tenure) in Japan continued to enjoy consistently much higher job stability than their US counterparts over the last twenty-five years. On the other hand, job stability for mid-career hires and youth employees has deteriorated in Japan over the last twenty-five years, whereas there was no comparable decline in job stability in US counterparts.5 Kawaguchi and Ueno (2013) analysed Japan’s ESS to examine the change of age-specific mean tenure by birth cohort. They found that male workers born in 1970 have experienced about 20 per cent fewer years of job tenure at a given age than have those born in 1994.6 (p.155) 2.4 Contribution to the Literature
Several authors have previously proposed models characterized by multiple equilibria to explain lower labour mobility and more intense provision of firmsponsored training in Japan than in the US (see Prendergast, 1989; Chang and Wang, 1995; Acemoglu and Pischke, 1998). These are adverse selection models, where informational asymmetry on workers’ abilities plays a central role in explaining differences. Morita (2001, 2005) proposes models in which strategic complementarity arises from the connection between continuous process improvement and firm-specificity of human capital, and the resulting multiplicity of equilibria provides an explanation for a set of stylized differences in US– Japanese work organizations and labour market practices. This chapter contributes to the literature by offering complementary explanations. In existing theoretical explanations for the difference in labour mobility, some workers voluntarily move from their current employers to other existing firms for better remuneration, where separation is caused by worker– firm match qualities and/or information asymmetry. However, this chapter focuses on involuntary separation caused by firm dynamics. By incorporating firm dynamics, the model yields new explanations and predictions based on managerial capability and entry regulation. An advantage of previous explanations based on multiplicity of equilibria is that they do not rely on differences in exogenous parameters. But the advantage is accompanied by difficulties in equilibrium selection, which is an important element for explaining differences.
3 Concluding Remarks This chapter has proposed new theoretical explanations for US–Japanese differences in employment practices based on Morita’s (2012) model of firm dynamics that incorporates workers, their accumulation of specific human capital, and their mobility. Several papers have previously offered theoretical explanations for the differences based on multiplicity of equilibria, where labour mobility is generated by worker–firm match qualities and/or information asymmetry. Complementary to them, my explanations are based on exogenous Page 10 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices differences regarding managerial capability and entry regulation, where labour mobility is generated by firm dynamics. The model predicts that lower importance on managerial capability leads to lower labour mobility, higher levels of firm-specific human capital (p.156) acquisition, and steeper earnings-tenure profiles. This prediction, combined with the supposition that the importance of managerial capability was substantially lower in Japan than in the US during the process of Japan’s economic development in the post-war period, provides an explanation for the differences. The model also yields a complementary explanation based on entry restrictions, which reduce labour mobility and increase firms’ incentives for investment in specific human capital. Now that the phase of Japan’s economic development has changed from one that is catching up to one that has caught up, most Japanese industries have become much closer to the world technology frontier. This implies that the relative importance of managerial capability has increased in Japan. Also, the degree of entry restrictions imposed by the Japanese government has become lower in the recent trend of deregulation. The model then predicts that the degree of US– Japanese differences tends to become smaller. Several recent empirical results regarding the earnings-tenure profiles are consistent with this prediction, whereas recent findings on labour mobility are less clear-cut. Will US–Japanese differences disappear, or will the differences persist although they may become smaller? Answers to this question are left to future research. Acknowledgements I wish to thank Hiroyuki Chuma, Jed DeVaro, Arghya Ghosh, Junichiro Ishida, Shingo Ishiguro, Hideshi Itoh, Takao Kato, Edward Lazear, Hiroyuki Okamuro, Hideo Owan, John Roberts, Kotaro Suzumura, Hirotaka Takeuchi, Katsuya Takii, Michael Waldman, and Yijiang Wang for comments, discussions, and encouragement. Financial support from the Australian Research Council is gratefully acknowledged. References Bibliography references: Acemoglu, D. and J-S Pischke (1998), ‘Why do firms train? Theory and evidence’. Quarterly Journal of Economics 113: 79–119. Acemoglu, D., P. Aghion, and F. Zilibotti (2006), ‘Distance to frontier, selection, and economic growth’. Journal of the European Economic Association 4: 37–74. Chang, C. and Y. Wang (1995), ‘A Framework for understanding differences in labor turnover and human capital investment’. Journal of Economic Behavior and Organization 28: 91–105. Page 11 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Clark, R. L. and N. Ogawa (1992), ‘Employment tenure and earnings profiles in Japan and the United States: Comment’. American Economic Review 82: 336–45. (p.157) Dertouzos, M. L., R. L. Lester, and R. M. Solow (1989), Made in America. Cambridge, MA: MIT Press. Djankov, S., R. La Porta, F. Lopez-de-Silanes, and A. Shleifer (2002), ‘The regulation of entry’. Quarterly Journal of Economics 117: 1–37. Hamaaki, J., M. Hori, S. Maeda, and K. Murata (2012), ‘Changes in the Japanese employment system in the two lost decades’. Industrial Labor Relations Review 65: 810–46. Hashimoto, M. and J. Raisian (1985), ‘Employment tenure and earnings profiles in Japan and the United States’. American Economic Review 75: 721–35. Hashimoto, M. and J. Raisian (1992), ‘Employment tenure and earnings profiles in Japan and the United States: Reply’. American Economic Review 82: 346–54. Ito, T. (1992), The Japanese Economy. Cambridge, MA: MIT Press. Kambayashi, R. and T. Kato (2017), ‘Long-term employment and job security over the last twenty-five years: A comparative study of Japan and the U.S.’. Industrial and Labor Relations Review 70: 359–94. Kato, T. (2001), ‘The end of lifetime employment in Japan?: Evidence from national surveys and field research’. Journal of the Japanese and International Economies 15: 489–514. Kawaguchi, D. and Y. Ueno (2013), ‘Declining long-term employment in Japan’. Journal of the Japanese and International Economies 28: 19–36. Koike, K. (1977), Shokuba no rodokumiai to sanka [Labor Union of Workplace and Participation]. Tokyo: Toyokeizai Shinposha. Koike, K. (1988), Understanding Industrial Relations in Modern Japan. New York: St. Martin’s Press. Lincoln, E. J. (2001), Arthritic Japan: The Slow Pace of Economic Reform. Washington, DC: Brookings Institution. MacDuffie, J. P. and T. A. Kochan (1995), ‘Do U.S. firms invest less in human resources? Training in the world auto industry’. Industrial Relations 34: 147–68. Mincer, J. and Y. Higuchi (1988), ‘Wage structures and labor turnover in the United States and Japan’. Journal of the Japanese and International Economies 2: 97–133.
Page 12 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Morita, H. (2001), ‘Choice of technology and labour market consequences: an explanation of U.S.-Japanese differences’. Economic Journal 111: 29–50. Morita, H. (2005), ‘Multi-skilling, delegation, and continuous process improvement: a comparative analysis of US-Japanese work organizations’. Economica 72: 69–93. Morita, H. (2012), ‘Firm dynamics and labor market consequences’. Journal of Law, Economics, & Organization 28: 235–64. Okimoto, D. I. (1989), Between MITI and the Market: Japanese Industrial Policy for High Technology. Stanford, CA: Stanford University Press. Ono, H. (2010), ‘Lifetime employment in Japan: concepts and measurements’. Journal of the Japanese and International Economies 24: 1–27. Ono, H. and M. Rebick (2003), ‘Constraints on the level and efficient use of labor’, in M. Blomström, J. Corbett, F. Hayashi, and A. Kashyap (eds) Structural Impediments to Growth in Japan. Chicago, IL: University of Chicago Press, pp. 225–57. Porter, M. E., H. Takeuchi, and M. Sakakibara (2000), Can Japan Compete? London: Macmillan. (p.158) Prendergast, C. (1989), Theories of internal labor markets. PhD dissertation. New York: Yale University. Suzumura, K. and K. Kiyono (1987), ‘Entry barriers and economics welfare’. Review of Economic Studies 54: 157–67. Suzumura, K. and M. Okuno-Fujiwara (1987), ‘Industrial policy in Japan: overview and evaluation’, in R. Sato and P. Wachtel (eds) Trade Friction and Economic Policy. Cambridge: Cambridge University Press, pp. 50–79. Waldman, M. (2013),‘Theory and evidence in internal labor markets’, in R. Gibbons and J. Roberts (eds) The Handbook of Organizational Economics. Princeton, NJ: Princeton University Press, pp. 520–74. Notes:
(1) See Morita (2012) for details. (2) See Djankov et al. (2002) for an example. (3) A standard underinvestment problem occurs in the model because the return from specific human capital is split between the employer and the worker through bargaining.
Page 13 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices (4) Suzumura and Okuno-Fujiwara (1987) argued that the anti-competitive nature of Japanese industrial policy became weaker after the first oil crisis that occurred in 1973. (5) See also Kato (2001) for an earlier contribution. (6) See also Ono’s (2010) study on lifetime employment in Japan. Through the use of multiple data sets and methods, Ono found that no more than 20 per cent of workers in Japan were likely to be employed under informal lifetime employment contracts. He found that evidence regarding changes in lifetime employment was mixed. The share of workers in the core was declining, but the probability of job separations remained stable for those who were already in the system. There is also evidence that the economic stagnation of the 1990s disproportionately affected female and younger workers.
Access brought to you by:
Page 14 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
US–Japanese Differences in Employment Practices New Explanations and Predictions Hodaka Morita
DOI:10.1093/oso/9780198812555.003.0009
Abstract and Keywords This chapter shows that a model that captures the interconnections between firm dynamics, labour mobility, and specific human capital provides new explanations for and predictions on the US–Japanese differences in labour mobility, wage structure, and firm-sponsored training, based on cross-country differences in the importance of managerial capability. My argument is based on the idea that managerial capability increases its importance as an economy or an industry approaches the technological frontier. It also provides complementary explanations and predictions based on governmental interventions in firm dynamics, given that a guiding principle of Japanese industrial policy has been the regulation of so-called ‘excessive competition’. Keywords: entry regulations, firm dynamics, labour mobility, managerial capability, specific human capital, US–Japanese differences, wage structure
1 Introduction US–Japanese differences in employment practices have attracted significant attention in the literature of international comparison as to how internal labour markets operate (see for example Waldman, 2013). The United States (US) and Japan have been considered representing two contrasting employment systems, and the Japanese system was often celebrated as a major source of the success of Japan’s economic development in the post-war growth era (Kambayashi and Kato, 2017). In contrast, however, Japan’s Great Recession (which lasted Page 1 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices approximately from 1992 to 2007) has led many to question the effectiveness of the Japanese employment system for contemporary and future economy (Ono and Rebick, 2003; Kambayashi and Kato, 2017). This chapter proposes a new theoretical explanation for US–Japanese differences in the post-war growth period, where I focus on the differences in labour mobility, firm-sponsored training for specific human capital, and wage structures. It then discusses its predictions and implications for the differences in the contemporary and future economy. As a theoretical framework, I use Morita’s (2012) model of firm dynamics that incorporates workers, their accumulation of specific human capital, and their mobility. Firm dynamics generate labour mobility because workers move from contracting and existing firms to expanding firms and new entrants. Also there are connections between labour mobility and firm-specific human capital investment because specific human capital possessed by a worker loses its value if the worker leaves his (p. 147) current employer. Elaborating on these interconnections, I show that the model provides new explanations for the differences. 1.1 US–Japanese Differences in the Post-War Growth Period
In the process of Japan’s economic development in the post-war period, labour mobility was lower and earnings-tenure profiles were more steeply sloped in Japan than in the US, and Japanese firms tended to provide higher levels of specific human capital through intensive firm-sponsored training compared to their US counterparts. Concerning labour mobility, it was found that the labour turnover rate was much higher in the US than in Japan (see Hashimoto and Raisian, 1985; Mincer and Higuchi, 1988). Hashimoto and Raisian (1985) for instance found that the 15-year job retention rates of the male population between the early 1960s and the late 1970s were much higher in Japan than in the US across all age groups. Regarding wage structures, Hashimoto and Raisian (1985) analysed data from the 1980 Basic Survey of Wage Structure for Japan, and data from the 1979 Current Population Survey for the US, and found that earnings-tenure profiles were more steeply sloped in Japan than in the US (see also Mincer and Higuchi, 1988). That is, they found that growth rates in earnings attributable to firm-specific tenure were substantially greater in Japan than in the US for all firm-size groups. Concerning firm-specific human capital, Koike (1977, 1988) found, in his comparative analysis of Japanese and US industrial relations, that Japanese workers acquired more firm-specific human capital through rotation among related jobs (see also Dertouzos et al., 1989; Ito, 1992). Also, levels of firmsponsored training were higher in Japanese firms than in US firms as pointed out by a number of authors such as Mincer and Higuchi (1988), Dertouzos et al. (1989), Ito (1992), and MacDuffie and Kochan (1995). According to MacDuffie and Kochan (1995), the International Assembly Plant Study, a survey of ninety motor vehicle assembly plants in twenty-four countries, found that Japanese Page 2 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices companies’ plants in Japan gave their newly hired workers an average of 364 hours of training compared to only forty-two hours in US companies’ plants in North America. The study also found a significant difference in the training of experienced workers.
2 Explanations and Predictions This section first presents an outline of Morita’s (2012) model and its analysis. It then discusses explanations for and predictions on the US–Japanese differences that the model yields, and its contribution to the existing theoretical explanations for the differences. (p.148) 2.1 Outline of the Model and Its Analysis
Consider an industry that produces a homogeneous good in a two-period setting, where the industry faces a downward-sloping demand schedule in each period , given by
. Entry and exit of firms are free in each period,
where each firm is of measure zero so that it is too small to affect prices. The production requires labour input; a firm can produce one unit of the good in a period if it employs one worker in that period, and the firm can produce nothing otherwise. There is a large number of ex ante identical individuals. Each individual can earn a reservation wage of
per period in a competitive
labour market outside this industry. Firms and individuals are both risk neutral. Each firm’s production efficiency is determined by its managerial capability and the level of its worker’s firm-specific human capital, where managerial capability is interpreted as representing the capability of a firm’s top management to develop an effective strategy and create a unique competitive position. Let a denote the managerial capability and
denote the realization of firm i’s
managerial capability, which is a random draw from a uniform distribution between 0 and 1. Assume that
is ex ante unknown to all agents including firm i
itself and becomes common knowledge at the end of the first period of firm i’s operation. This specification is consistent with the widely held view that the capability of a firm’s top management is mostly innate and difficult to observe or assess ex ante. At the beginning of period 1, firms simultaneously make first-period wage offers to the individuals. Each individual can apply to a firm for first-period employment, and each firm employs one individual from the applicants. If an individual is not employed by the firm, he/she can earn a reservation wage for period 1. If firm i employs a worker, it can provide the worker with a level of specific human capital denoted (
by incurring a convex cost of
) per employee. The level of firm-specific human capital is
observable but not verifiable, and so wage contracts contingent upon it are not feasible. To keep the analysis simple, assume that firm-specific human capital affects the production efficiency only in the next period. Page 3 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Each firm i that employed a worker in period 1 produces one unit of the good. At the same time, firm i’s managerial capability
is realized and becomes common
knowledge. If a firm that operated in period 1 continues to operate in period 2, the return from its first-period investment in specific human capital is shared with its employee through the generalized Nash bargaining process. At the same time, a firm that did not operate in period 1 can employ an individual by making wage offers. Each firm that employed a worker in period 2 produces one unit of the good. Each firm i’s per-unit production cost in a period is given by and
, where
are given constants and λ captures the relationship (p.149)
between specific human capital and the production efficiency of a firm. In particular, assume that
if firm i employs worker j in that period
and employed the same worker in the previous period, while
if firm i
employs worker j in that period but did not employ the worker in the previous period. Notice that x and y capture the importance of managerial capability and firm-specific human capital, respectively, for production efficiency. In the remainder of this subsection, I present an outline of the equilibrium characterization of the model.1 Consider a perfect foresight equilibrium that is characterized by a price sequence . In the equilibrium, all agents make optimal decisions based on the anticipation of a particular price sequence , and their behaviour does in fact give rise to the same
. Given
free entry and exit of firms, for every entrant in period t (=1, 2) the present value of its expected overall profit is zero in the equilibrium. Also, given that there is a large number of ex ante identical and risk-neutral individuals, and that every individual can earn a reservation wage
per period in a competitive
labour market outside the industry, the present value of every individual’s expected overall wage is
in period 1 in the equilibrium. The market clears in
each period in the equilibrium. Morita (2012) focuses on perfect foresight equilibria in which a strictly positive number (measure) of firms enter and exit the industry at the beginning of period 2, given that in reality entries to and exits from firms are common in most industries. Throughout the analysis, it is assumed that H is large enough so that the optimal level of firm-specific human capital is interior in the equilibrium. Suppose that there exists a perfect foresight equilibrium characterized by . In the equilibrium, each second-period entrant employs a worker at the reservation wage w, and its expected production cost is
.
Since each second-period entrant earns zero expected profit in the equilibrium, must hold. Consider firm i that employed worker j and chose in the equilibrium. If firm i continues to employ worker j in period 2, its secondPage 4 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices period production cost (excluding the wage bill) is
and the firm
must pay at least w (the reservation wage) in order to employ worker j in period 2. Then firm i continues to operate in period 2 if and only if , where
if
, and 0 otherwise. Suppose that firm i’s managerial capability
turns out to be greater than or
equal to the cut-off level . Then, at the beginning of period 2 firm i and worker j bargain over worker j’s second-period wage, where the worker’s bargaining power is
and his/her threat point is the reservation wage w.
Firm i’s outside option is to exit the industry and (p.150) earn zero profit or to employ another worker with wage w and continue operating in period 2. This leads us to find worker j’s second-period wage (which is determined as the outcome of the generalized Nash bargaining process)
, where
. Here, , which is the surplus gained by reaching an agreement in the bargaining. This in turn implies that firm i’s second-period production cost plus wage bill is where
,
.
In period 1, firm i chooses the level of firm-specific human capital knowing its own managerial capability
without
. Firm i makes this choice under the
anticipation that it will continue to operate in period 2 and earn second-period profit of
if it realizes managerial capability
and will exit the industry if maximize
, where
,
. Hence firm i chooses
to
is defined by . Through the
maximization exercise we find that
must hold, and every firm i
that employed a worker in period 1 chooses
in the equilibrium, where
. This implies that the firms’ exit rate is where
in the equilibrium,
.
Every firm that employs a worker in period 1 offers the same first-period wage (denoted by
) in the equilibrium, given that firms and individuals are ex ante
identical. A worker employed by firm i in period 1 anticipates that his/her second-period wage will be
if firm i realizes
, and w if
. The first-period present value of the worker’s expected overall wage is
Page 5 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices
then
. Given that there is a large number
of ex ante identical and risk-neutral individuals, and that every individual can earn a reservation wage
per period outside the industry, firms in the
equilibrium choose
such that
holds. Hence we find that . In the equilibrium, every firm i that employs a worker at the worker with level
in period 1 provides
of specific human capital by incurring
training cost. Since its expected production cost is its first-period cost is
as a
, the expected value of
. Firm i operates in period 2 if
it realizes the managerial capability of with the second-period total cost of . Hence the first-period present value of its overall expected cost is determines the level of
. Then the following zero profit condition .
(p.151)
Given (
)=(
), the demand for the good in period t (
) is
.
Since the exit rate is , firms continue to operate in period 2. Since the second-period demand for the good is , the measure of the second-period entrants is
in the equilibrium. Then
must hold for a strictly positive measure for firms to enter at the beginning of period 2 in the equilibrium. This condition, along with
and
level denoted by
, holds if and only if y is smaller than a unique threshold . Morita (2012) shows that
is not only necessary
but also sufficient for the model to have perfect foresight equilibria in which a strictly positive number (measure) of firms enter and exit the industry at the beginning of period 2. I will discuss comparative statics results of the model in this range of parameterizations. 2.2 Explanations for US–Japanese Differences
Page 6 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices 2.2.1 Managerial Capability
Acemoglu et al. (2006) argue in their analysis of technology frontiers and firm selection that managerial skill is more important for undertaking innovative activities than for adopting and imitating existing technologies from the world technology frontier. They then point out, based on their analysis of the correlation between distance to the frontier and R&D intensity using data from the OECD sectoral database, that innovation becomes more important as the economy approaches the world technology frontier and there remains less room for adoption and imitation. Following their analysis and argument, I maintain that the importance of managerial capability was substantially lower in Japan than in the US in the process of Japan’s economic development in the post-war period when most Japanese industries were catching up with the West (Okimoto, 1989). The model yields the following prediction regarding the importance of managerial capability. As the importance of managerial capability (captured by x in the model) decreases, labour mobility decreases, firms provide their workers with higher levels of specific human capital, and the slope of the earnings-tenure profile becomes steeper in equilibrium. Then, lower importance of managerial capability in Japan than in the US provides an explanation for the US–Japanese differences in labour mobility, wage structure, and firm-sponsored training of specific human capital in the post-war period. The logic for the explanation is as follows. In period 1, several firms hire workers, provide them with specific human capital, and produce outputs. At the beginning of period 2, some first-period entrants that have realized relatively low managerial capabilities exit the industry, whereas several second-period entrants hire workers from exiting firms or outside the industry to produce outputs in period 2. A first-period entrant exits the industry in period 2 if its (p. 152) second-period cost is higher than second-period entrants’ expected cost. This is because, due to free entry of firms, second-period entrants earn zero expected profits in equilibrium, implying that second-period entrants’ expected cost is equal to the second-period equilibrium price. Each first-period entrant has an advantage over second-period entrants because it has a worker who has already accumulated specific human capital during period 1. Nevertheless, a first-period entrant exits the industry if its managerial capability turns out to be relatively low, lower than a threshold value
.
As the importance of managerial capability decreases, the threshold value decreases. This is because, as managerial capability becomes less important, a first-period entrant’s advantage associated with its worker’s specific human capital increases, implying that it does not exit the industry unless its managerial capability turns out to be very low. Lower
means lower labour
mobility because fewer first-period entrants exit the industry and hence fewerworkers switch their employers between periods 1 and 2 as
decreases.
Page 7 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Lower mobility, in turn, increases first-period entrants’ incentives to provide their workers with specific human capital, because it is more likely for them to survive and continue employing their period 1 workers in period 2. Under lower mobility, each worker anticipates, when he is employed by a firm in period 1, that his period 1 employer provides a higher level of specific human capital, and there is a higher chance for the employment to continue in period 2. This increases each worker’s expected second-period wage. Since workers anticipate higher second-period wages, firms can hire workers at lower wages in period 1, implying a steeper earning-tenure profile under lower mobility. 2.2.2 Entry Regulations
In the process of Japan’s economic development in the post-war period, the Japanese government invoked a fairly sophisticated system of interventionist economic policy to promote steady growth, technological innovations, and international competitiveness (Suzumura and Okuno-Fujiwara, 1987). A guiding principle of Japanese industrial policy has been the regulation of so-called ‘excessive competition’ (Suzumura and Kiyono, 1987), and the Japanese government often restricted entries into industries. On the other hand, the degree of such governmental intervention has been much lower in the US.2 How do entry restrictions affect welfare and the nature of employment practices? To provide an answer to this question, Morita (2012) incorporates a government that can impose entry restrictions in an extension of the model. Suppose that, at the beginning of period 1, the government announces the maximum number of firms that can enter the industry in period 2. Such an (p. 153) entry restriction reduces the second-period supply of the good, which in turn implies that the second-period equilibrium price becomes higher than the price in the free-entry equilibrium. The higher second-period price increases attractiveness of the second-period operation, which results in the following trade-off from the welfare standpoint. On the one hand, some first-period entrants with low managerial capabilities that would not survive in the freeentry equilibrium would survive in the equilibrium with an entry restriction. The resulting cost inefficiency works in the direction of decreasing welfare. On the other hand, the higher survival rate of the first-period entrants means lower labour mobility, which increases firms’ incentives for providing firm-sponsored training. This effect works in the direction of increasing welfare because it mitigates the underinvestment problem in specific human capital.3 Morita (2012) shows that the government can increase welfare by choosing a degree of entry restriction so that the latter positive welfare effect dominates the former negative one. Entry restrictions reduce labour mobility, which in turn increases firms’ incentives to provide their workers with specific human capital. And the higher level of specific human capital steepens the slope of the earning-tenure profile. Hence entry restrictions that have often been imposed by the Japanese Page 8 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices government provide an explanation, complementary to the explanation based on the importance of managerial capability, for US–Japanese differences in the postwar growth period. In addition, the model yields another prediction regarding managerial capability. Lower importance of managerial capability reduces the cut-off managerial capability
and entry restrictions have the same effect on
.
This leads to a prediction that the average managerial capability (which represents a firm’s capability to develop an effective strategy) is lower in typical Japanese firms than in their US counterparts. This is consistent with Porter et al. (2000) who argue through a number of case studies that, although many Japanese firms have achieved very high operational effectiveness, they are less capable of identifying and implementing innovative strategies to sustain their competitive advantage. 2.3 Predictions for the Contemporary and Future Economy
The Japanese economy has already caught up with the West, and most Japanese industries have got much closer to the world technology frontier. This increases the importance of managerial capability which is crucial for undertaking innovative activities. Also, the degree of entry restrictions imposed by the Japanese government has become lower in light of the recent trend of (p.154) deregulation that started in the early 1990s.4 The theoretical model then predicts that the degree of US–Japanese differences tends to become smaller. That is, labour mobility increases, the slope of the earnings-tenure profile becomes flatter, and the level of specific human capital accumulation declines in Japan. In reality, however, such transitions in Japan are likely to take place rather slowly due to the vested interests and institutional inertia embedded in the Japanese economic system. Lincoln (2001) argues that deregulation in Japan has proceeded slowly because of the strong vested interests of government bureaucrats. Several empirical studies have been undertaken recently to find out whether or not the Japanese employment system has exhibited significant changes. Regarding the earnings-tenure profile, Hamaaki et al. (2012) used micro-level data from the Basic Survey on Wage Structure for the period 1989 to 2008, and found that the age-wage profile of Japanese employees has become flatter in recent years. This finding follows up previous studies by Clark and Ogawa (1992) and Hashimoto and Raisian (1992), who found that the slope of the earnings-tenure profile for large Japanese firms became somewhat flatter in the 1980s compared with the 1970s. Findings of recent studies on labour mobility are less clear-cut. Kambayashi and Kato (2017) conducted a cross-national analysis of micro-data from Japan’s Employment Status Survey (ESS) and its US counterpart, Current Population Survey (CPS). They calculated the ten-year job retention rates of Japanese employees in the private sector for the four time periods, 1982–92, 1987–97, 1992–2002, and 1997–2007, and then used various CPS Supplements to Page 9 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices calculate the comparable ten-year job retention rates for US employees. Kambayasi and Kato found that, on the one hand, core employees (aged 30–44 with at least five years of tenure) in Japan continued to enjoy consistently much higher job stability than their US counterparts over the last twenty-five years. On the other hand, job stability for mid-career hires and youth employees has deteriorated in Japan over the last twenty-five years, whereas there was no comparable decline in job stability in US counterparts.5 Kawaguchi and Ueno (2013) analysed Japan’s ESS to examine the change of age-specific mean tenure by birth cohort. They found that male workers born in 1970 have experienced about 20 per cent fewer years of job tenure at a given age than have those born in 1994.6 (p.155) 2.4 Contribution to the Literature
Several authors have previously proposed models characterized by multiple equilibria to explain lower labour mobility and more intense provision of firmsponsored training in Japan than in the US (see Prendergast, 1989; Chang and Wang, 1995; Acemoglu and Pischke, 1998). These are adverse selection models, where informational asymmetry on workers’ abilities plays a central role in explaining differences. Morita (2001, 2005) proposes models in which strategic complementarity arises from the connection between continuous process improvement and firm-specificity of human capital, and the resulting multiplicity of equilibria provides an explanation for a set of stylized differences in US– Japanese work organizations and labour market practices. This chapter contributes to the literature by offering complementary explanations. In existing theoretical explanations for the difference in labour mobility, some workers voluntarily move from their current employers to other existing firms for better remuneration, where separation is caused by worker– firm match qualities and/or information asymmetry. However, this chapter focuses on involuntary separation caused by firm dynamics. By incorporating firm dynamics, the model yields new explanations and predictions based on managerial capability and entry regulation. An advantage of previous explanations based on multiplicity of equilibria is that they do not rely on differences in exogenous parameters. But the advantage is accompanied by difficulties in equilibrium selection, which is an important element for explaining differences.
3 Concluding Remarks This chapter has proposed new theoretical explanations for US–Japanese differences in employment practices based on Morita’s (2012) model of firm dynamics that incorporates workers, their accumulation of specific human capital, and their mobility. Several papers have previously offered theoretical explanations for the differences based on multiplicity of equilibria, where labour mobility is generated by worker–firm match qualities and/or information asymmetry. Complementary to them, my explanations are based on exogenous Page 10 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices differences regarding managerial capability and entry regulation, where labour mobility is generated by firm dynamics. The model predicts that lower importance on managerial capability leads to lower labour mobility, higher levels of firm-specific human capital (p.156) acquisition, and steeper earnings-tenure profiles. This prediction, combined with the supposition that the importance of managerial capability was substantially lower in Japan than in the US during the process of Japan’s economic development in the post-war period, provides an explanation for the differences. The model also yields a complementary explanation based on entry restrictions, which reduce labour mobility and increase firms’ incentives for investment in specific human capital. Now that the phase of Japan’s economic development has changed from one that is catching up to one that has caught up, most Japanese industries have become much closer to the world technology frontier. This implies that the relative importance of managerial capability has increased in Japan. Also, the degree of entry restrictions imposed by the Japanese government has become lower in the recent trend of deregulation. The model then predicts that the degree of US– Japanese differences tends to become smaller. Several recent empirical results regarding the earnings-tenure profiles are consistent with this prediction, whereas recent findings on labour mobility are less clear-cut. Will US–Japanese differences disappear, or will the differences persist although they may become smaller? Answers to this question are left to future research. Acknowledgements I wish to thank Hiroyuki Chuma, Jed DeVaro, Arghya Ghosh, Junichiro Ishida, Shingo Ishiguro, Hideshi Itoh, Takao Kato, Edward Lazear, Hiroyuki Okamuro, Hideo Owan, John Roberts, Kotaro Suzumura, Hirotaka Takeuchi, Katsuya Takii, Michael Waldman, and Yijiang Wang for comments, discussions, and encouragement. Financial support from the Australian Research Council is gratefully acknowledged. References Bibliography references: Acemoglu, D. and J-S Pischke (1998), ‘Why do firms train? Theory and evidence’. Quarterly Journal of Economics 113: 79–119. Acemoglu, D., P. Aghion, and F. Zilibotti (2006), ‘Distance to frontier, selection, and economic growth’. Journal of the European Economic Association 4: 37–74. Chang, C. and Y. Wang (1995), ‘A Framework for understanding differences in labor turnover and human capital investment’. Journal of Economic Behavior and Organization 28: 91–105. Page 11 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Clark, R. L. and N. Ogawa (1992), ‘Employment tenure and earnings profiles in Japan and the United States: Comment’. American Economic Review 82: 336–45. (p.157) Dertouzos, M. L., R. L. Lester, and R. M. Solow (1989), Made in America. Cambridge, MA: MIT Press. Djankov, S., R. La Porta, F. Lopez-de-Silanes, and A. Shleifer (2002), ‘The regulation of entry’. Quarterly Journal of Economics 117: 1–37. Hamaaki, J., M. Hori, S. Maeda, and K. Murata (2012), ‘Changes in the Japanese employment system in the two lost decades’. Industrial Labor Relations Review 65: 810–46. Hashimoto, M. and J. Raisian (1985), ‘Employment tenure and earnings profiles in Japan and the United States’. American Economic Review 75: 721–35. Hashimoto, M. and J. Raisian (1992), ‘Employment tenure and earnings profiles in Japan and the United States: Reply’. American Economic Review 82: 346–54. Ito, T. (1992), The Japanese Economy. Cambridge, MA: MIT Press. Kambayashi, R. and T. Kato (2017), ‘Long-term employment and job security over the last twenty-five years: A comparative study of Japan and the U.S.’. Industrial and Labor Relations Review 70: 359–94. Kato, T. (2001), ‘The end of lifetime employment in Japan?: Evidence from national surveys and field research’. Journal of the Japanese and International Economies 15: 489–514. Kawaguchi, D. and Y. Ueno (2013), ‘Declining long-term employment in Japan’. Journal of the Japanese and International Economies 28: 19–36. Koike, K. (1977), Shokuba no rodokumiai to sanka [Labor Union of Workplace and Participation]. Tokyo: Toyokeizai Shinposha. Koike, K. (1988), Understanding Industrial Relations in Modern Japan. New York: St. Martin’s Press. Lincoln, E. J. (2001), Arthritic Japan: The Slow Pace of Economic Reform. Washington, DC: Brookings Institution. MacDuffie, J. P. and T. A. Kochan (1995), ‘Do U.S. firms invest less in human resources? Training in the world auto industry’. Industrial Relations 34: 147–68. Mincer, J. and Y. Higuchi (1988), ‘Wage structures and labor turnover in the United States and Japan’. Journal of the Japanese and International Economies 2: 97–133.
Page 12 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices Morita, H. (2001), ‘Choice of technology and labour market consequences: an explanation of U.S.-Japanese differences’. Economic Journal 111: 29–50. Morita, H. (2005), ‘Multi-skilling, delegation, and continuous process improvement: a comparative analysis of US-Japanese work organizations’. Economica 72: 69–93. Morita, H. (2012), ‘Firm dynamics and labor market consequences’. Journal of Law, Economics, & Organization 28: 235–64. Okimoto, D. I. (1989), Between MITI and the Market: Japanese Industrial Policy for High Technology. Stanford, CA: Stanford University Press. Ono, H. (2010), ‘Lifetime employment in Japan: concepts and measurements’. Journal of the Japanese and International Economies 24: 1–27. Ono, H. and M. Rebick (2003), ‘Constraints on the level and efficient use of labor’, in M. Blomström, J. Corbett, F. Hayashi, and A. Kashyap (eds) Structural Impediments to Growth in Japan. Chicago, IL: University of Chicago Press, pp. 225–57. Porter, M. E., H. Takeuchi, and M. Sakakibara (2000), Can Japan Compete? London: Macmillan. (p.158) Prendergast, C. (1989), Theories of internal labor markets. PhD dissertation. New York: Yale University. Suzumura, K. and K. Kiyono (1987), ‘Entry barriers and economics welfare’. Review of Economic Studies 54: 157–67. Suzumura, K. and M. Okuno-Fujiwara (1987), ‘Industrial policy in Japan: overview and evaluation’, in R. Sato and P. Wachtel (eds) Trade Friction and Economic Policy. Cambridge: Cambridge University Press, pp. 50–79. Waldman, M. (2013),‘Theory and evidence in internal labor markets’, in R. Gibbons and J. Roberts (eds) The Handbook of Organizational Economics. Princeton, NJ: Princeton University Press, pp. 520–74. Notes:
(1) See Morita (2012) for details. (2) See Djankov et al. (2002) for an example. (3) A standard underinvestment problem occurs in the model because the return from specific human capital is split between the employer and the worker through bargaining.
Page 13 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
US–Japanese Differences in Employment Practices (4) Suzumura and Okuno-Fujiwara (1987) argued that the anti-competitive nature of Japanese industrial policy became weaker after the first oil crisis that occurred in 1973. (5) See also Kato (2001) for an earlier contribution. (6) See also Ono’s (2010) study on lifetime employment in Japan. Through the use of multiple data sets and methods, Ono found that no more than 20 per cent of workers in Japan were likely to be employed under informal lifetime employment contracts. He found that evidence regarding changes in lifetime employment was mixed. The share of workers in the core was declining, but the probability of job separations remained stable for those who were already in the system. There is also evidence that the economic stagnation of the 1990s disproportionately affected female and younger workers.
Access brought to you by:
Page 14 of 14
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
On Quality Traps and Economic Development Patrick M. Emerson
DOI:10.1093/oso/9780198812555.003.0010
Abstract and Keywords This chapter considers the interdependence among the quality levels of government institutions. Citizens of democratic societies are consumers of institutional output and the quality they demand from individual institutions is posited to be a function of the joint quality of all institutional output. Specifically, the quality of institutions is hypothesized to enter into consumers’ preferences in a supermodular fashion. An implication of this is that citizens will tend to desire institutions of the same quality; thus resource constrained democratic governments will tend to match the quality level of their complementary institutions. The Nash equilibrium concept is employed to show that multiple equilibria will result, and that a stable equilibrium exists at a low level of quality. Keywords: quality, complementarity, supermodularity, underdevelopment, multiple equilibria
I Introduction Do low-income economies get stuck in quality traps? This chapter posits that they can through a type of coordination failure caused by the natural inclination of consumers to consume complementary goods of similar quality. The basic idea is simple: Germans who enjoy high quality Autobahns value higher quality cars more than do consumers in countries where road quality is poor and consumers of auto tyres will value high quality tyres less if the quality of the car is low. Computer users will value fast computers less if their internet access is limited or slow, and consumers of software will value CPU intensive programs less if the quality of the computer is low. This can have consequences for development, if the quality of infrastructure or other complementary products is low; businesses that are attempting to climb the value added ladder might face difficulties Page 1 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development selling higher quality goods. Globalization is a partial antidote, freer trade and more efficient transportation allows consumers to consume in a global market; but infrastructure limitations are more endemic and without substantial investment might keep economies trapped in a low quality equilibrium. This chapter proposes a model of quality consumption where consumers value the quality of complementary goods and prefer to match quality across complementary goods. This behaviour can have the effect of creating an externality for producers of goods that are used together with other goods in that the price they can charge for their good depends not only on its own quality, but also on the quality of other goods that will be used with it. This can lead to ‘quality traps’ where any one individual producer might not find it in its interest to unilaterally increase quality; however, if all producers of the (p.160) other complementary products increased quality at the same time it would be in all of their interests to do so. This explanation is in no way limited to developing countries. A cursory examination of the way complementary products are bundled together all over the world reveals an interesting phenomenon: these bundles tend to be of similar quality. Thus, new Mercedes saloons come equipped with top-of-the-line tyres and sound systems, while KIAs come equipped with lower quality tyres and a lower quality radio. This is true as well for component stereos and computers: Sony does not try to sell its top quality audio player with its lowest quality headphones; portable computers that contain the latest, most powerful processor generally come equipped with a top quality display screen. So a consumer that wishes to spend less money on a portable computer will have a hard time finding models that bundle a top processor with a poor quality screen. Rather, they will generally find a modest processor bundled with a modest quality screen. The examples are endless: excellent restaurants serve top quality wines, sports shops sell low quality bicycle frames together with low quality components, and high resolution digital cameras come bundled with high quality lenses. In many cases, multiple equilibria can coexist in the same economy where lowincome consumers can consume lower quality goods and high-income consumers can consume high quality goods. In developing countries, however, the average consumer has a lower income than an average consumer from a developed country. This difference need not be great to eliminate the high quality equilibrium, and even though there might be sufficient demand for middle quality goods, the economy may be forced into low quality equilibrium. In addition, it is often the government that provides essential infrastructure like phone lines, roads, television and radio signals, electricity, etc., as well as often running leading industries like steel, concrete, petroleum, and others. If what the government is supplying is of low quality, which is more likely in developing countries where government revenues and wherewithal are low, then this could Page 2 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development keep entire networks of industries in low quality equilibria. Finally, capital constraints on the part of just one firm that produces a complementary good to other firms (which, due to imperfect capital markets, is more likely to affect firms in developing countries) can cause quality traps to arise and persist, thus making it more likely that developing countries will fall into them. For such a widespread phenomenon, there has been surprisingly little investigation into the root causes of this seemingly empirical regularity. Complementary goods are a well-studied subject in economics, as is strategic complementarity of one firm’s output to another’s. However both of these concepts deal with complementarity in quantity. That is, if you consume one left shoe then, naturally, you would like to consume a right shoe for an extra (p. 161) left shoe is not very useful. Or, if one firm is producing more blackboards, a chalk company finds it in its interest to produce more chalk. Little has been studied, however, about the quality of complements and strategic complementarity in the quality of complementary goods. The purpose of this chapter is to provide an initial examination of how quality matters in complementary goods and to show how government provision of a crucial good and the generally high level of industrial concentration in developing countries can bring about low quality traps. The theory is built around the hypothesis that consumers care about the quality of complementary goods and will, all else being equal, strive to consume similar qualities across complementary goods. This stems from the supermodular way in which the quality of complementary goods enters the utility function.1 What this chapter argues is that this is a common feature of human preference. Firms understand that preferences are supermodular in complementary quality. And when choosing to produce a good, firms often choose not only the quantity of the good to supply to the market, but also the quality of the good. When one firm’s output is complementary to another firm’s this decision then becomes dependent on the other firm’s quantity and quality decisions. This being the case, there are strategic implications for firms that produce a good to be used with another good. This chapter shows that in such a framework quality becomes a strategic complement for these firms, and that in monopolistic markets, multiple equilibria can arise where the low quality equilibrium is stable.2 This chapter is structured as follows. Section 2 studies the class of utility functions that have the property of supermodularity in the quality of complementary goods. Section 3 proposes a market equilibrium framework where the quality of complementary goods enters a potential consumer’s utility multiplicatively. It is assumed that two mediocre goods cannot be substituted for one high quality good (you can only drive one car at a time and two poor quality transmissions cannot equal one good quality transmission; or, you only use one computer at a time and two poor quality operating systems cannot be Page 3 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development substituted for one good quality one, etc.). The chapter then describes and characterizes solutions to the market equilibrium problem using the Nash equilibrium concept, and shows that multiple equilibria result where a stable equilibrium occurs at a low quality level for a class of cost functions and that a sufficiently large unilateral deviation is required by one of these firms to escape this equilibrium. Finally, extensions of the model such as a Stackelberg leader, a unique monopolist, and an oligopoly, are discussed. Section 4 discusses how this model can help explain quality traps (p.162) in developing countries. In Section 5, the findings of the chapter are summarized and future avenues of exploration are discussed.
2 Utility Functions with Supermodularity in the Quality of Complements As stated in Section 1, utility functions of the type u(x1, x2) where x1 is the quantity of good 1 and x2 is the quantity of good 2 and the two goods are complements have been thoroughly studied in economics. This chapter instead focuses on utility functions of the type u(q1, q2) where q1 is the quality of good 1 and q2 is the quality of good 2, and the two goods are complements for consumers who consume exactly one of each good. This type of utility function seems not to have been previously examined. Quality in production and the production of quality have been examined previously, however. Kremer (1993) and Kremer and Maskin (1996) look at quality in the production process (where quality is multiplicative in tasks) and call these ‘O-ring’ production processes. Following the Kremer formulation, Gabszewicz and Turrini (2000) look at worker quality in a vertically differentiated industry. Economides (1999) examines quality in the production of component goods and compares independent vertically related monopolists to a single vertically integrated monopolist. This chapter’s focus is on the way the quality of complementary goods interacts in consumers’ preferences and can be seen as a generalization of the above analyses. The central hypothesis of this chapter is that the quality of complementary goods enters consumers’ utility functions in a supermodular manner. More specifically, it is assumed that
. It is, of course, also being assumed that
, since greater quality is, by definition, good. This means that the marginal utility a consumer derives from the consumption of the quality of good 1 increases in the quality of its complement, good 2. In addition, as is shown formally in Section 4, for one type of utility function that has this property, this implies that consumers, ceteris paribus, prefer to match qualities of complementary goods. This comes about for a similar reason to the one in Becker’s (1982) marriage model where individuals prefer to wed those with similar attributes.3
Page 4 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development There are many common utility functions that have the supermodularity attribute, Leontief, Cobb-Douglas, and Constant Elasticity of Substitution are the three most common. In addition, a utility function that has the supermodular feature as well as the feature that if any one of a bundle of complementary (p. 163) goods is defective it renders the bundle unusable, is the multiplicative utility function. This utility function takes the product of all the quality levels of a bundle of complementary goods. This is simply a positive monotonic transformation of a Cobb-Douglas utility function, and it is the one that is studied for the remainder of the chapter. Section 3 utilizes the multiplicative utility function in a model where producers are monopolists in the production of their own good and each good is a complement to the others.
3 Equilibria when Complementary Quality is Multiplicative in Preferences If quality enters consumer’s utility functions multiplicatively (similar to the way quality enters into the production process in the O-ring production function of Kremer (1993) and Kremer and Maskin (1996), then this can induce reaction functions where one firm’s quality is a strategic complement to another firm’s and multiple equilibria can arise where all producers will produce similar quality. 3.1 The Model
Consider consumers with identical preferences over n complementary goods, where each good is produced by a single firm, who are represented by a utility function where the quality of the complementary goods enters multiplicatively. Therefore, the consumer’s utility function is
(1)
where xi is the amount of good i and is restricted to taking on values of either 0 or 1 for every i, qi is good i’s quality and qi ∈ [δ, β] for every i, where δ can be interpreted as the minimum quality attainable if you choose to make a good and β the maximum quality level possible. In what follows, I assume δ>0, and I will normalize by setting β=1. One interpretation of quality level, therefore, is the probability of non-failure, or the probability that the good will not break (fail to operate properly). M is a numeraire good or money. The consumer’s budget constraint is given by
(2)
where Y is the consumer’s exogenously given income level and pi is a quality specific price for good i. This constraint will clearly bind by non-satiation. It is Page 5 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development immediately evident that a consumer will choose to buy either one of each (p. 164) complementary good or none at all. Note that while, in this case, the consumer is restricted to buying either one or none of the good, the consumer could replicate this decision as long as total purchases respect the budget constraint. That is, one could simply add another
term to
the utility function, or that utility is additively separable in bundles of complementary goods. Also note that an analysis of total market demand is not needed as these producers are not competing in the quantity dimension. For ease of exposition, only the case of two complementary goods shall be considered, but note that all of the analysis is immediately generalizable to ngoods. In this context, we will restrict our attention only to consumers that buy one of each complementary good (this is a similar approach to that of Gabszewicz and Thisse, 1979, 1980, and Shaked and Sutton, 1982, 1983, in their studies of differentiated product markets) and note that for consumers who purchase one unit of each of the two complementary goods, the utility function reduces to
(3)
and the budget constraint reduces to
(4)
Due to non-satiation in the preferences, the budget constraint will bind at the solution. Given this set-up we can look at the two possible choices for this consumer. If the consumer decides not to buy the complementary goods the consumer’s utility is given by
(5)
If the consumer does buy the complementary goods his/her utility is given by
(6)
Clearly the consumer will purchase the complementary goods if and only if , or
(7)
Page 6 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Since the analysis is the same for both firms, consider firm 1. Since these firms are monopolists in the production of their own good, they make take-it-or-leaveit offers of price-quality pairs (p1, q1), and (p2, q2). Since firms will want to extract the entire surplus from the consumer, they will set price and quality such that (7) binds with equality. The profit function of firm 1 is given by
(8) (p.165)
where
, and j = 1, 2. Hence, given q2 and p2,
we can define p1(q1), the highest price firm 1 can charge for a given quality, as
(9)
Knowing this, we can re-write profit as
(10)
Firm 1 maximizes profit by selecting a quality level, so the solution to firm 1’s maximization problem is to choose a q1 such that,
(11)
subject to
(12)
The interpretation of Equation (11) is straightforward. The right-hand side is the marginal cost of increasing quality while the left-hand side is the marginal benefit. The marginal benefit depends on the quality of all the other complementary goods (in this case only good 2). Equation (12) is the condition that the firm will only produce the good if it is making non-negative profits. Equilibrium, therefore, is defined as a (p1, q1), and a (p2, q2) which satisfy the following five conditions:
(13)
(14)
Page 7 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development
(15)
(16)
(17)
What remains, then, is to search for Nash equilibria. To do so each firm’s reaction function in quality must be defined. To find firm1’s reaction function Φ1 (q2), we need only to invert the marginal cost function and to prevent the firm’s quality level from exceeding 1 or falling below δ.
(18)
With these reaction functions in place, finding Nash equilibria simply requires finding solutions to the system of two equations. The most interesting case is where convex reaction functions result in multiple equilibria. This result will be stated as a proposition. (p.166) Proposition 1 For the class of cost functions that result in convex reaction functions for both firms, multiple equilibria in quality space may result, with the feature that low quality equilibrium is stable. Proof Note that when q = 0, Φ′(q)= 0 for both firms. This means that, in q1, q2—space, one reaction function (firm 2’s) starts with zero slope and the other (firm 1’s) starts with infinite slope. If the reaction functions are convex, their slope increases in q. Again, in q1, q2—space, as q1 becomes larger, firm 2’s reaction function becomes more steeply sloped to a limit of a vertical slope, and conversely as q2 becomes larger, firm 1’s reaction function becomes less steeply sloped to a limit slope of zero. Thus, due to the convexity of the reaction functions, there exist dual crossing conditions of the functions in q1, q2—space. Specifically, there exists an interval [0, α), to which q1 belongs, where to which q1 belongs, where
, and there exists another interval .
There are really three possible outcomes depending on the properties of the reaction functions: they can cross once, twice, or not at all. From the simple dynamics of the reaction functions it is easy to see that if they cross once, the single equilibrium is at the crossing point (assuming it is in the [δ,1] interval for both q’s). If they do not cross, the equilibrium is at q1, q2 = 1, the high quality Page 8 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development level. If they cross twice, the first (low) crossing point is a stable equilibrium (because firm 1’s reaction function cuts firm 2’s from below) unless it is below the point q1, q2 = δ, in which case q1, q2 = δ is a stable low quality equilibrium. The second crossing is unstable (because firm 1’s reaction function cuts firm 2’s from above) and if it occurs below the point q1, q2 = 1, then it is an equilibrium along with the point q1, q2 = 1 (which is stable). If it occurs above the point q1, q2 = 1, then only the low crossing (or q1, q2 = δ) is an equilibrium. Q.E.D. It is also worthwhile to note that for the class of cost functions that result in a concave reaction function, there exists a stable single equilibrium, typically not at the high quality level.4 The question of how surplus is split among the producers remains. The result will again be stated as a proposition. Proposition 2 Any division of the surplus is a Nash equilibrium. Proof Consider the case where one producer captures the entire surplus and all other producers get zero profit. The producer who captures the entire surplus clearly does not want to deviate, but neither do the others. This is because a unilateral deviation in price upwards would cause all consumers to abandon the market and the producer would end up with zero surplus as (p.167) before. A unilateral deviation in price downwards would cause negative profits since producers see no increase in demand for their component. It should be clear from this extreme case that any division of the surplus is a Nash equilibrium. Q.E.D. In the repeated playing of the game we would expect to see strategies evolving that guarantee an even distribution of the surplus. Therefore, for the purposes of this chapter, it will be assumed that any surplus is split evenly. 3.2 Illustrative Example
The above propositions and their proofs become much clearer when an explicit example is studied. For illustrative purposes consider the two producer case where
, i = 1, 2. This specification of the cost
function produces the reaction function for firm 1 and
for firm 2. Here, δ is the
minimum possible quality level attained by producing the good and in this case is attainable at no cost. Figure 10.1 graphs these reaction functions in (q1, q2)space. Note that as long as a firm’s total revenue is at least as great as total cost, a firm will never want to drive a complementary producer out of business because this would ruin the market for its own good, so the reaction functions have a lower bound of assumed that
, and , and
. In Figure 1 it is . Given this specification, three
Page 9 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Nash equilibria result and these are illustrated in Figure 10.1. Note that equilibrium B is an unstable equilibrium, while A and C are stable. The two stable Nash equilibria are a low quality and high quality equilibrium respectively (in fact, the equilibrium at C is where each firm produces the highest quality (p. 168) possible). Note that to move from the low quality equilibrium to the high quality, a large unilateral deviation must occur (one that moves one firm’s quality level beyond that of equilibrium B). As this analysis is extended to n firms it is clear that each firm’s reaction function becomes increasingly convex as n gets larger. This results from the multiplicative nature of quality in consumer’s preferences. Therefore as the number of complementary producers gets larger, the set of cost functions that produce concave reaction functions becomes larger also. Note also that as n becomes
Figure 10.1 The firms’ reaction functions
larger the marginal benefit of quality, described by
, for
firm j, that accrues to any individual producer gets smaller as long as the industry is not at the high quality equilibrium. This translates to a strengthening of the low quality equilibrium (meaning that a larger unilateral deviation is now needed to move to a high quality equilibrium) as each individual firm tries to set its quality level such that marginal benefit equals marginal cost. It therefore becomes harder and harder for the economy to escape the low quality equilibrium. 3.3 Extensions of the Model
This section presents extensions of the model. These three extensions are presented as propositions. The first result considers the case of a Stackleburg leader. In this case, there are again two producers of complementary goods, each of whom choose a pricequality pair, but they do so sequentially. Proposition 3 If one of the firms described above is allowed to choose its price–quality pair first, it will be able to extract the entire available surplus.
Page 10 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Proof This result can be seen immediately when one recognizes that the second mover will produce (stay in the market) as long as it enjoys nonnegative profits. The first mover, knowing this, will make its optimal decision (thus choosing the optimal quality level) and will extract the entire surplus for itself by choosing a price that leaves no surplus for the other firm, pushing the second mover to the break-even point. Q.E.D. The next result considers the case of multiple complementary goods, where there exist multiple producers for each good but one. This last good is produced by a monopolist. Proposition 4 If there exists a single producer in only one of the complementary goods, the lone monopolist will extract the entire available surplus. Proof As firms are price–quality setters (where price is the independent variable), multiple producers in any one complementary good will immediately drive profits of those firms to zero through Bertrand price competition. This means (p.169) that if a monopolist exists in the production of only one of the goods, they are able to extract the entire surplus because they can maintain a price above marginal cost. Q.E.D. The final result considers a market in which two complementary goods are made, but where there are two producers of each good. Firms that produce the same good shall be called ‘competitors’ and the firms that make complementary goods shall be called ‘complementors,’ following the terminology of Brandenburger and Nalebuff (1996). Consumers are free to choose between two competing firm’s products for each good, but they take the quality of each good as given. For each firm the cost of producing a good will depend only on the quality of the good and, for ease of analysis, assume that each firm has a choice of producing at only one of two quality levels: low and high. The producers make their quality decisions simultaneously, and in the second stage they choose prices. The second stage game is as before: there will be Bertrand price competition among producers of the same good that have chosen the same quality level. Otherwise, producers act as monopolists and extract all of the surplus if their complementors have chosen the same quality level, or split the surplus if their complementors have chosen different quality levels (that is, if their complementor is also a monopolist). As for consumers, they shall consume the price–quality pair that gives them the highest non-negative surplus (otherwise they will not consume at all), and if there exists more than one bundle that gives them the same surplus, it is assumed that consumers divide themselves evenly over the bundles.
Page 11 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Proposition 5 If there exist two producers of each good, and two possible quality levels, then the two producers of each good will not produce the same level of quality. Proof To find the sub-game perfect Nash equilibria, note first that if both sets of competitors choose different quality levels, then they all earn positive profits. Note also that if the two sets of competitive firms choose the same quality (it does not matter if the quality is the same across complementors) then all firms earn zero profits. Finally if one set of competitors chooses the same quality and the other set chooses different qualities, then the same quality competitors earn zero profits and the different quality competitors earn positive profits. So it is never in a producer’s interest to produce the same quality level as its competitor. It can always do better by producing a different quality, regardless of whether its complementors are producing the same quality or, if so, which level it is. Thus the only Nash equilibrium is for each set of complementors to produce different qualities. Q.E.D. So both quality levels of both goods will be available on the market, but note that for a wide class of cost functions, consumers will never consume a mixed (p.170) bundle, for if they try and do so, they will end up worse off than not consuming at all. Consumers of low quality bundles spend less money on them but the net effect on their utility is the same in both cases—zero. What is particularly interesting here is that even though there is competition in the production of each good, differentiation through quality allows each firm to act as a monopolist and, as a result, any surplus is extracted by producers. This is bad for consumers, for if there were competition at each quality level all of the surplus would go to them. Two insights come out of this simple example. One is that producers spread out on the quality spectrum so as to find a unique match with a complementor. This is somewhat similar to the product differentiation literature (for example, Dixit and Stiglitz, 1977; Dixit, 1979), where differentiation through quality allows firms to capture more surplus. But in that literature differentiated consumers is what drives the result; here all consumers are identical. This also seems to correspond to the experience of the real world. Firms tend to build reputations as high or low quality producers, for example Sony and Visio in consumer electronics. Sony is known as a high quality brand and consumers know that they will have to pay more for a Sony TV than they will for a Visio TV with similar attributes. The second observation is that consumers never mix bundles. This helps explain why pre-packaged bundles almost always contain items of similar quality and why it is rare to find a consumer who purchases, for example, a high quality computer and amplifier, and a low quality monitor.
Page 12 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Section 4 discusses the unique attributes of developing countries that may lead to the creation and persistence of these quality traps.
4 Quality Traps and Development The model presented in this chapter implicitly assumes that the consumer has a sufficient income to afford the high quality pair. In developing countries, due to general poverty, this may not be true for many different goods. In this case even though consumers could afford quite a high level of quality they may not be able to afford the high quality available at the high quality equilibrium. In this case, it is possible that the high quality equilibrium becomes just out of reach, but because of the dynamics of the model, the only equilibrium available is the low quality one. Another feature of developing countries is that key infrastructure and strategic industries are more likely to be in the hands of one company or the government in developing countries than in developed.5 The model (p.171) above shows that if a key good is produced by a monopolist, the quality they produce will determine the quality of all other complementary goods. Governments and monopolistic producers who do not understand the implications of their quality decisions can be responsible for keeping a sector in a low quality trap. This cuts both ways however. There is scope for the same ‘leading sector’ to push the industry to the high quality equilibrium. Imagine a new fibre-optic highspeed data system installed throughout India. Suddenly owning high-speed computers, a digital television recorder and game consoles provide consumer with much more added utility. Thus manufacturers would now step in and take advantage of the new greater surplus to be had and shall find it worthwhile to invest in quality improvements. This is somewhat akin to the big push argument in development economics. In these sectors, a big push by at least one firm can wrest the entire industry out of the low quality equilibrium and provide consumers with an overall increase in quality. In these situations, capital constraints may play a large role in the fact that developing countries find themselves more often, as compared to developed countries, in low quality ‘traps’. At the same time there can be lagging industries that prevent the complementors from reaching the high quality equilibrium. This can happen if just one of a group of industries is capitally constrained, which is arguably more common in developing country industries than in developed country industries. One can see the effect of capital constraints in this model by using the same stage game among monopolist complementary producers as in Section 3. By adding an additional stage that proceeds the quality setting stage where firms must make fixed capital investments that determine the maximum quality level they will be able to produce in the next period, the result is immediately clear. Though it may be all in a firm’s interest to invest in a certain level of quality Page 13 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development capital to reach profitable high quality equilibrium, capital constrains that prevent just one firm from doing so can force the entire economy into low quality equilibrium. This is equilibrium where no one producer has incentive to deviate to a higher quality. Since developing economies are often characterized by capital constraints, the above extension to the basic model provides an explanation of how developing countries with imperfect capital markets can become mired in low quality equilibrium. In addition, and perhaps more interestingly, this low quality equilibrium can persist even after a loosening of capital constraints. The quality push of leading sectors could be the catalyst for the movement to high quality equilibria, as was the case in Japan for example. The government, working closely with conglomerates, brought about remarkable increases in the quality of Japanese goods in the second half of the century. (p.172) As it is often government that provides essential infrastructure in most developing countries, the model leads to the insight that there may be external benefits beyond those that directly accrue when, say, a phone service is improved that could potentially benefit many domestic firms. This insight is instructive when designing optimal policy instruments and has the potential of helping developing country governments decide where to best apply their limited resources.
5 Concluding Remarks This chapter has presented a model where quality of complementary goods is supermodular in consumer’s utility functions and can lead to multiple equilibria in quality space. The model has the characteristic that for a broad range of cost functions low quality equilibrium is stable and a high quality equilibrium is attainable only after a large deviation on the part of at least one producer. This model helps explain several casual observations about the world: that goods tend to be bundled and sold together with similar quality complements; that firms tend to produce goods of similar qualities and develop quality niches; and the paradoxical observation that developing countries seem to be stuck in low quality equilibria across vast numbers of goods. This chapter is best viewed as a first step into the exploration of this seemingly unexplored relationship among the qualities of complementary products. Acknowledgements The idea for this chapter was first proposed by Kaushik Basu in his graduate development economics class at Cornell University in 1996. For useful comments and advice, I would like to thank Robert Masson, Jonathan Price, Ramsey Shehadeh, André Portela Souza, and especially Kaushik Basu. I would also like to thank participants in the Cornell TWIPS workshop, especially
Page 14 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Andrew Miller, Tridip Ray, and Eduardo Zambrano. All remaining errors are my own. References Bibliography references: Becker, Gary (1982), Treatise on the Family. Cambridge, MA: Harvard University Press. Brandenburger, Adam M. and Barry J. Nalebuff (1996), Co-opetition. New York: Doubleday. (p.173) Bulow, Jeremy I., John D. Geanakoplos, and Paul D. Klemperer (1985), ‘Multimarket oligopoly: Strategic substitutes and complements’. Journal of Political Economy 93, No. 3 (Jun., 1985): 488–511. Dixit, Avinash (1979), ‘Quality and quantity competition’. Review of Economic Studies 46(October): 587–99. Dixit, Avinash and Joseph E. Stiglitz (1977), ‘Monopolistic competition and optimum product diversity’. American Economic Review 67(June): 297–308. Economides, Nicholas (1999), ‘Quality choice and vertical integration’. International Journal of Industrial Organization 17: 903–14. Gabszewicz, J. Jaskold and J.-F. Thisse (1979), ‘Price competition, quality and income disparities’. Journal of Economic Theory 20: 340–59. Gabszewicz, J. Jaskold and J.-F. Thisse (1980), ‘Entry (and exit) in a differentiated industry’. Journal of Economic Theory 22: 327–38. Gabszewicz, J. and A. Turrini (2000), ‘Workers’ skills, product quality and industry equilibrium’. International Journal of Industrial Organization 18: 575– 93. Kremer, Michael (1993), ‘The O-ring theory of economic development’. Quarterly Journal of Economics 58 (August): 551–75. Kremer, Michael and Eric Maskin (1996), ‘Wage inequality and segregation by skill’. NBER Working Paper 5718. Sattinger, M. (1993), ‘Assignment models of the distribution of earnings’. Journal of Economic Literature 31 (June): 831–80. Shaked, Avner and John Sutton (1982), ‘Relaxing price competition through product differentiation’. Review of Economic Studies 49: 3–13.
Page 15 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
On Quality Traps and Economic Development Shaked, Avner and John Sutton (1983), ‘Natural oligopolies’. Econometrica 51(September): 1469–83. Topkis, Donald M. (1998), Supermodularity and Complementarity. Princeton, NJ: Princeton University Press. Tybout, James R. (2000), ‘Manufacturing firms in developing countries: how well do they do, and why?’ Journal of Economic Literature 38(1): 11–44. Notes:
(1) For an excellent treatise on supermodularity and complementarity, see Topkis (1998). (2) For more on the idea of strategic substitutes see Bulow, Geanakoplos and Klemperer (1985). (3) See also Kremer (1993); Sattinger (1993); Topkis (1998). (4) For example, with power cost functions (like those examined by Economides [1999]) of the type , increasing and convex reaction functions result when
, and increasing and concave reaction functions result when .
(5) For example Tybout (2000) surveys evidence of the amount of industrial concentration in developing countries as opposed to developed.
Access brought to you by:
Page 16 of 16
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
Inequality and Neighbourhood Effects Market Access and Welfare of the Poor Namrata Gulati Tridip Ray
DOI:10.1093/oso/9780198812555.003.0011
Abstract and Keywords The key insight in our research is to recognize inequality–neighbourhood interaction: neighbourhood effects interacting with income inequality may affect poor people’s ability to access basic facilities like health-care services, schooling, and so on. While Gulati and Ray (2016) model this interaction on a monopolist service provider in a neighbourhood structured as a linear city where rich and poor consumers live side by side, in this chapter we extend the analysis to a competitive framework with free entry and exit where the natural neighbourhood structure is a circular city. We find inverted-U shape relationships between income inequality and market access and welfare of the poor: if we compare a cross-section of societies, the poor community as a whole is initially better off living in relatively richer societies, but, beyond a point, the aggregate market access and consumer surplus of the poor starts declining as society becomes richer. We identify the possibility of complete exclusion of the poor from the market: a scenario where the service providers cater only to the rich and the poor have absolutely no market access, and find that it is the higher income gap between rich and poor that exposes the poor to this unfortunate outcome. Keywords: inequality, neighbourhoods, welfare of the poor, market access, inverted U, complete exclusion of the poor
Page 1 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects 1 Introduction The key idea in our research is to recognize the inequality–neighbourhood interaction. For important services like hospitals, schools, or banks that are required at regular intervals, distance to the service provider is an important factor determining accessibility. On the other hand, private service providers take both the location and income mix of people into account while making strategic decisions like whether to enter into a neighbourhood at all, and, upon entry, what price and quality to choose for their products and services. Thus, market access and welfare of an individual are considerably influenced by the type of neighbourhood in which the individual lives. There is a substantial body of evidence showing how neighbourhood poverty affects poor people’s ability to access facilities such as health care and schooling (see Gulati and Ray (2016) for details). Although the evidence is compelling, there seems to be very little analytical research to demonstrate how neighbourhood effects interacting with income inequality might affect poor people’s ability to access these basic facilities. Our research is an early attempt to model this interaction by integrating consumers’ income distribution with the spatial distribution of their location. While Gulati and Ray (2016) models this interaction on a monopolist service provider in a neighbourhood structured by a linear city where rich and poor consumers live side by side, in this chapter we extend the analysis to a competitive framework with free entry and exit where the natural neighbourhood structure is a circular city. We explore the consequences of an increase in income (p.175) inequality on the welfare of the poor in general, and their access to market in particular. We consider a homogeneous product or service in a competitive framework with free entry and exit. It is very interesting to investigate the interaction of inequality and neighbourhood effect in such an ideal market structure. The inequality–neighbourhood interaction is captured by the spatial structure where the neighbourhood is a circular city across which consumers are uniformly distributed, with rich and poor consumers living side by side. The preference structure reflects the higher willingness to pay of the richer consumers and consumers’ reluctance to travel farther to access the product or service under consideration. The industrial structure is characterized by the presence of a fixed cost of production. The set-up is a two-stage game. In the first stage, the potential providers of the product or service decide whether to enter into the neighbourhood or not; in the second stage, the entering firms choose their prices simultaneously. In this set-up we explore the interaction of income inequality with the neighbourhood effect in determining market outcomes and their consequences on the market access and welfare of the poor. We find an inverted-U shape relationship between income inequality and the welfare of the poor: if we compare a cross-section of societies, the poor community as a whole is initially better off living in relatively richer societies; Page 2 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects but, beyond a point, the aggregate consumer surplus of the poor starts declining as society becomes richer. Interestingly the same inverted-U shape relationship is also observed between income inequality and market access of the poor. The reason for this inverted-U shape relationship can be traced to the opposing welfare impacts of income inequality working through equilibrium price and number of firms. Consumers benefit from the increase in numbers of firms as it increases their market access, but lose from the increase in price. As the neighbourhood of the poor becomes richer, both price and number of firms increase steadily. For the poor community effect of the number of firms initially dominates: the poor residing closer to firms get to access the product as the number of firms increases. But, beyond a certain point, the adverse price effect takes over. In order to examine the role of inequality in its purest form, we also analyse the effect of the mean-preserving spread: increases in rich income, together with a decrease in the proportion of rich, keeping fixed the poor income and the average income of society. We find that the effect depends on the initial proportion of poor in the neighbourhood. If the initial proportion of poor is to the left of the peaks of the inverted-U relationships, then both market access and consumer surplus of the poor increase at the beginning, reach a maximum, and then fall as the spread increases. If this initial proportion of the poor is to the right of the peaks, then both market access and consumer surplus of the poor decrease steadily as the spread increases. As the spread (p.176) increases through a decrease in the proportion of rich, firms are forced to lower prices and some firms leave as they find it unattractive to serve the neighbourhood. A fewer number of firms reduces poor people’s market access and hence their welfare, whereas lower prices increase market access and welfare. To the left of the peaks of the inverted U price effect dominates, while the number of firms effect dominates to the right of the peaks. We have also identified the possibility of complete exclusion of the poor from the market: a scenario where service providers completely ignore the presence of the poor and choose the price as if there are only rich individuals residing in the neighbourhood. We have isolated the higher income gap between rich and poor as the key factor that exposes the poor to the possibility of complete exclusion. We have also found that the poor are more likely to be completely excluded when they are in a minority: firms may completely ignore the poor even when the rich are not ultra-rich just because the rich are greater in number. Finally, we compare a mixed-income neighbourhood where rich and poor live side by side with a single-income homogeneous neighbourhood and find that the poor are better off living in the mixed neighbourhood as long as the poor income is below a certain feasibility threshold.
Page 3 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects The impact of income inequality on various socio-economic outcomes is quite well-known in the literature.1 There exists an extensive body of work that shows that higher inequality is positively correlated with higher infant mortality (Waldmann, 1992), lower economic growth (Alesina and Rodrik, 1994; Persson and Tabellini, 1994), violent crime (Fajnzylber et al., 2002), subversion of institutions (Glaeser et al., 2003), and so on. The present work complements this literature by pointing out a new channel through which income inequality may impact on economic outcomes—a channel that works through firm behaviour and industrial organization. To the best of our knowledge, Atkinson (1995) is the only work that investigates the implications of inequality operating through industrial structure. But Atkinson (1995) considers only a monopolist firm and does not allow free entry. The tension between price and number of firms effects is the key feature that is highlighted in our chapter. The idea that people with higher income generally have higher willingness to pay and that firms do take this into account while making strategic decisions was developed by Gabszewicz and Thisse (1979) and extended by Shaked and Sutton (1982, 1983). In our model, consumers differ both in terms of their income and location. The basic horizontal product differentiation model was introduced by Hotelling (1929) and later developed by Salop (1979). Subsequent works in this literature (for example, Neven and Thisse, 1990; (p.177) Economides, 1993) look at product specifications combining both the vertical and horizontal characteristics. However, this kind of industrial organization literature does not explore the implications of income inequality. The chapter is organized as follows. Section 2 outlines the set-up with the spatial structure capturing the inequality–neighbourhood interaction. Section 3 analyses the generic scenario where the poor have partial market access while the rich have complete access. The effect of inequality on market access and welfare of the poor is investigated in Section 4. In Section 5 we characterize all the equilibrium possibilities highlighting the role of inequality in generating the possibility of complete exclusion of the poor. Comparison with the single-income neighbourhood is also discussed in this section. Finally, we conclude in Section 6. The detailed derivations are relegated to the Appendix (Gulati and Ray, 2017).
2 The Set-up Our model adapts the framework of Salop (1979).2 There is a circular city of circumference 1 unit. Two types of consumers, rich and poor, are uniformly distributed along the circumference of the city: there is
proportion of poor with
income
Obviously
and
proportion of rich with income
The total number of consumers is normalized to 1. There are
private establishments in the city providing a homogeneous product
or service. Examples of such establishments are private schools, hospitals, banks, and so on. For the sake of brevity let us refer to them as firms. These Page 4 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects firms are located equidistant from each other around the circle so that the distance between adjacent firms is
The number of firms is not fixed; it is
determined endogenously from free entry and exit condition.3 Each consumer buys either one unit of the homogeneous product from his most preferred firm, or does not buy the product at all. Let consumer with income
be the gross utility a
enjoys from consuming the product. Here
is a
preference parameter indicating consumers’ valuation of the product. Since this formulation of gross utility captures the feature that willingness to pay is higher for the rich. This formulation also ensures that the preference is non-homothetic. Preference non-homotheticity and income heterogeneity, implying that changes in prices may affect rich (p.178) and poor consumers differently, allow us to explore the role of income distribution.4 Let us use the notations
for location of firm
A consumer at location access the product or service from firm transportation cost of
for the price it charges,
has to travel a distance
to
and he/she incurs a travel or
Of course he/she has to pay the price
the net utility of a consumer at location firm
and
with income
. Hence
and purchasing from
is given by
If a consumer does not buy the product, his/her utility, the reservation utility, is 0. This formulation of the utility function helps to model the interaction of neighbourhood effects with income inequality in a simple and tractable way. While gross utility captures the higher willingness of the rich to pay, the presence of travel cost reflects the disutility if the facility is not available in the nearby neighbourhood. Unlike the industrial organization literature where distance reflects horizontal product differentiation, we treat the distance literally as physical distance from the facility. For facilities like schools or hospitals the importance of distance or accessibility is undeniable. Production requires fixed costs; in order to produce any output at all, each firm must incur a fixed cost
Further, there is a marginal cost of production,
which is independent of output. Profit of firm
charging a price
is then given
by
Page 5 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects where
denotes demand faced by firm
elaborate in the next sections how
Given the spatial structure, we
depends on firm ’s own price,
the prices of the two adjacent firms,
and on
and
The set-up is a two-stage game. In the first stage, firms decide whether or not to enter, and entering firms locate equidistantly around the circumference of the city. In the second stage, firms choose their prices simultaneously.
3 Partial Market Access for the Poor and Complete Access for the Rich In order to understand the implications of the inequality–neighbourhood interactions, in this section we analyse the generic scenario where all the (p. 179) rich are served by the market, whereas, for the poor, some are served while others are left out.5,6 Consider firm
located between two firms
and
. As all the rich are
served, demand from the rich is given by the distance of the marginal rich who are indifferent between two adjacent firms. Let firm
denote the distance from
of the marginal rich who are indifferent between firms
and
that is,
It follows that
Similarly
so that firm ’s demand from the rich is that is,
Since only some of the poor are served while others are left out, whether a poor person will buy the product from firm
depends on whether he/she is better off
from buying. Hence the demand from the poor is determined by the distance of the poor who are indifferent between buying and not buying from firm denote this distance, that is,
Clearly firm ’s total demand from the poor is
Let
It follows that
that is,
Figure
11.1 illustrates this demand structure where the continuous bold arch depicts the segment of demand coming from the rich, whereas the dashed bold arch shows the part of demand coming from the poor.
Page 6 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects (p.180) Since there are proportion of poor and proportion of rich, total demand facing firm
is given by
Figure 11.1 Demand faced by Firm j
(1)
The first segment of demand comes from the rich, the second segment from the poor. Notice that the rich segment of demand is independent of rich income. Since the rich have complete market coverage, the exact level of the rich’s income is irrelevant to firms. On the other hand, the size of the poor segment is determined by the poor income. An increase in the number of firms reduces the size of the rich segment while the poor segment remains unaffected. The own price effect dominates the cross price effect within the rich segment, and the presence of the poor segment reinforces this domination. The price response to demand is given by
Note that since
demand loss due to increased price is larger in the poor segment, an increase in the proportion of poor increases the price response to demand. On the other hand an increase in travel cost makes it costlier to access the facilities, and this in turn reduces the price response to demand. To determine the equilibrium price and number of firms we proceed in the standard backward fashion. In stage 2, given the entry decision in stage 1, firm chooses its price to maximize profit,
The first-order condition with respect
to price implies
It is easy to see that this linear system has a unique solution,7
(2)
Page 7 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects In stage 1, firms’ entry decision is determined by the zero-profit condition (up to the integer problem8). Using (1) and (2) the common expression for profit becomes
(p.181) Hence the zero-profit condition implies
(3)
Using (2) and (3) we derive the equilibrium price and number of firms:
(4)
(5)
Note that since the exact level of the rich’s income is irrelevant to firms, both price and number of firms are independent of the rich income. Price is also independent of the poor income. But, since the poor form a captive market for firms whose size is restricted by their income, the number of firms increases with the poor’s income. As poor income increases, the demand size of each firm increases, and, with price remaining the same, each firm makes more than normal profit. This super-normal profit attracts fresh entry of firms into the neighbourhood. It is easy to see that while equilibrium price increases with fixed cost, the number of firms decreases. This result is fairly intuitive.9 We show (see Gulati and Ray, 2017: App 7.2.1) that the scenario under consideration, partial access for the poor and complete access for the rich, arises only when poor and rich incomes are such that
(6)
and
Page 8 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects
(p.182)
where
is the average neighbourhood income. That is, we
have identified an upper income threshold and a lower income threshold such that if the poor’s income is in between these two thresholds then firms do not compete with the adjacent firms for poor consumers but do so only for rich consumers. All rich consumers are served by the market, but some poor are left out—only those poor who are located closer to the firms get served. These income thresholds are shown in Figure 11.4. It is interesting to observe that the lower income threshold for the poor depends on the average neighbourhood income (on the income of the rich, to be more precise). Since the lower income threshold is higher when the rich’s income is reasonably high. In Gulati and Ray (2017: App. 7.2.1) we show that when rich income is high enough so that the firms find it optimal to increase prices to cater only to the rich when poor’s income is between and With the help of these two income thresholds we can now see how equilibrium price and number of firms respond to changes in the proportion of poor. This will be useful in understanding the mechanism of the impact of income inequality on the welfare of the poor analysed in the Section 4. We find that both equilibrium price and number of firms increase steadily as the proportion of poor (f) decreases from 1 to 0.10
and
remaining the same, as
decreases the
society or neighbourhood becomes richer and the average willingness of society to pay increases. This induces the existing firms to charge a higher price, and, at the same time, attracts fresh entry into the neighbourhood. The following proposition summarizes the discussion in this section. Proposition 1 In equilibrium the rich have complete market access while the poor have partial access only when rich and poor incomes are such that condition (6) holds. Only those poor residing closer to the facilities have access to them, others are excluded. (p.183) (a) Equilibrium price and number of firms are given by Equations (4) and (5) respectively.
Page 9 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects (b) Equilibrium price and number of firms are independent of the rich income; while price is also independent of the poor’s income, the number of firms increases with the poor’s income. (c) With an increase in fixed cost, equilibrium price increases and number of firms decreases. (d) Both equilibrium price and number of firms increase as the proportion of poor decreases.
4 Inequality, Market Access, and Welfare of the Poor Now we use this generic case to analyse the impact of income inequality on the market access and welfare of the poor. Consider market access of the poor consumers first. Not all the poor can afford to buy the product: only the poor up to the distance buying the product; those in between the distance
from any firm are and
cannot afford
it. Hence the aggregate market access of the poor community as a whole is
The tension between price and number of firms is clear: an increase in the number of firms increases market access while a price increase reduces it. Since price and number of firms are endogenous, substituting their equilibrium values from Equations (4) and (5) we derive the expression for aggregate market access of the poor community solely in terms of the parameters of the model:
(7)
Note that aggregate market access of the poor increases as poor’s income increases. In fact it is easy to check that lower income threshold), whereas
as as
income threshold); in between these two thresholds
(the (the upper increases steadily as
increases. There are two effects at work. First is the direct positive effect following from the non-homothetic preference structure. Second is the indirect effect working through the increase in number of firms as poor income increases. Both effects work in the same direction reinforcing each other. Since price is independent of income, there is no counteracting force at (p.184) work. Note also that the aggregate market access of the poor decreases as the fixed cost increases. As noted in Proposition 1(c), an increase in the fixed cost
Page 10 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects increases price and reduces the number of firms. Both effects work in the same direction to decrease the poor’s market access. Next consider the aggregate consumer surplus of the poor. To calculate the aggregate consumer surplus of the poor community as a whole we proceed as follows. Surplus to a poor consumer located at a distance which it is buying is market coverage of
11
Although there are
from the firm from firms each with a
on either side of its location, we have to keep in mind
that the poor in between the distance
and
from any firm do not buy
the product, so that the consumer surplus of these poor consumers is zero. Hence the aggregate consumer surplus of the poor is
Similar to aggregate market access, an increase in the number of firms increases aggregate consumer surplus of the poor while a price increase reduces it. Substituting the equilibrium values we derive
(8)
Similar to market access, the aggregate consumer surplus of the poor also increases steadily as the poor’s income increases in between the lower and upper bounds, and decreases with the increase in fixed cost. 4.1 Effect of a Change in the Proportion of Poor
Now to see the effect of income inequality on market access and welfare of the poor we first conduct the following comparative static analysis: we vary keeping
and
fixed. That is, we follow the poor with the same income level
and compare the aggregate market access and consumer surplus of the poor community as a whole when they live in relatively richer societies (as decreases from 1 to 0). For this comparative static exercise let us rewrite the expression for aggregate market access of the poor as (p.185)
Page 11 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects where than
captures, in a nutshell, all the parameters of the model other This expression becomes quite handy in depicting market access of the
poor as a function of
treating
as the parameter. Note from condition (6) that
Figure 11.2 depicts the aggregate market access of poor as varies from 0 to 1 for some illustrative values of parameter In Figure 11.2 higher values of
shifts the
within this range.
curve upwards illustrating the
point mentioned above that market access of the poor increases as the poor’s income increases, or as fixed cost decreases. Similarly we can express the aggregate consumer surplus of poor as (p.186)
Figure 11.2 Aggregate market access of the poor
Figure 11.3 illustrates this relationship as a function of values of
as in Figure 11.2. Again higher values of
for the same parameter
shifts the
curve
upwards illustrating that consumer surplus of the poor increases as the poor’s income increases.12
Page 12 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects It is interesting to observe the ‘inverted-U’ shape relationships between the proportion of poor people in the neighbourhood ( ) and their aggregate (p.187) market access (
) and
consumer surplus (
).13 That
is, if we compare a cross-section of neighbourhoods, the poor community as a whole is initially better off living in relatively richer neighbourhoods (as
decreases
from 1). But, beyond a certain point, both aggregate market access of the poor and their consumer surplus start
Figure 11.3 Aggregate consumer surplus of the poor
declining as the neighbourhood becomes richer. Instead of comparing a cross-section of societies if we consider the same society then this result can be interpreted as follows. Since
and
remain the same, as the proportion of poor ( ) decreases society becomes richer. If we restrict our attention to those who still remain poor, then their market access and consumer surplus demonstrate inverted-U shape relationships as society becomes richer. The reason for these inverted-U shape relationships can be traced to the behaviour of equilibrium price and number of firms. As established in Section 3 (see Proposition 1(d)), both price and number of firms increases steadily as decreases from 1 to 0. Consumers benefit from the increase in number of firms but lose from the increase in price. For the poor community as a whole the number of firms effect initially dominates: the poor located closer to firms get to consume the product and the number of poor served increases as the number of firms increases. But, beyond a certain point, the adverse price effect takes over. It is important to highlight the role of the spatial structure, in particular to point out that we are getting the inverted-U shape in both market access and consumer surplus because the number of firms is also changing endogenously. When we conduct the same analysis with the number of firms fixed, both market access and consumer surplus of the poor decreases steadily as
decreases; that
is, we do not see any inverted-U shape in the relationships. The reason is that price increases steadily without any compensating increase in the number of firms.
Page 13 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects The following proposition summarizes the relationships between the proportion of poor and the aggregate market access and welfare of the poor community as a whole. Proposition 2 When rich and poor incomes are such that the rich have complete market access while the poor have only partial access, then there exists an ‘inverted-U’ shape relationship between the proportion of poor people in the neighbourhood ( ) and their aggregate market access ( and consumer surplus (
): as
decreases from 1 to 0, both
)
and
initially increase, reach a maximum, and then fall. (p.188) 4.2 Effect of a Mean-Preserving Spread
In the comparative static exercise conducted above note that since
and
remain the same, as the proportion of poor ( ) decreases society becomes richer. In order to capture the role of inequality in its purest form let us next examine the effect of a mean-preserving spread: we increase an increase in
keeping
together with
and the average income of the society, (
), fixed. The result of this comparative static exercise follows in a straightforward way from the last two propositions. Note from Equations (7) and (8) that neither market access nor consumer surplus of the poor depends on the rich’s income. Hence the effect of the mean-preserving spread works only through the increase in the proportion of poor. Because of the inverted-U relationship encountered in Proposition 2 it follows that the effect of the mean-preserving spread depends on which part of the inverted U we start from. If the initial proportion of poor is to the left of the peak of the inverted U, then both market access and consumer surplus of the poor will increase for a substantial range of increase in the spread before reverting back to the downward trend. On the other hand, if the initial proportion of poor is to the right of the peak of the inverted U, then market access and consumer surplus of the poor decrease with the increase in the spread. The intuition for this result can again be traced to the behaviour of equilibrium price and number of firms. Observe that the mean-preserving spread works again only through the increase in the proportion of poor since neither price nor number of firms depends on the rich’s income. It follows that an increase in the mean-preserving spread decreases both price and number of firms. We have already noted that to the left of the peak of the inverted U price effect dominates while the number of firms effect dominates to the right of the peak. The following proposition summarizes the effect of this mean-preserving spread.
Page 14 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects Proposition 3 When rich and poor incomes are such that the rich have complete market access while the poor have only partial access, then the effect of mean-preserving spread depends on the initial proportion of poor. (a) If the initial proportion of poor is to the left of the peaks of the inverted-U relationships, then both market access and consumer surplus of the poor increase at the beginning, reach a maximum, and then fall as the spread increases. (b) If the initial proportion of poor is to the right of the peaks of the inverted-U relationships, then both market access and consumer surplus of the poor decrease steadily as the spread increases. (p.189) 4.3 Welfare of the Rich
Finally, consider the welfare of rich consumers. Recall that the rich have complete market access. Since there are
firms each with market coverage of
on either side of its location, the aggregate consumer surplus of the rich community as a whole is
As expected, consumer surplus increases with income ( (
), and decreases with travel cost ( ) and price (
) and number of firms
). Substituting the
equilibrium values of price and number of firms we get
(9)
Similar to that of poor people, we find an ‘inverted-U’ shape relationship between the proportion of poor people in the neighbourhood ( ) as well as the aggregate consumer surplus of the rich (
). This is the neighbourhood
externality for richer people working once again through the positive provision externality and negative price externality. But the trade-off (via the tension between the two externalities) is somehow relatively less acute for the rich than for the poor.14 The reason is that the rich always have full coverage whereas the poor only have partial coverage rendering the provision externality less relevant for the rich. It is interesting to note that the consumer surplus of the rich increases even when the income of the poor increases. As noted in Section 3, as poor income increases the price remains the same but number of firms increases. An increased number of firms implies greater accessibility to the product or service
Page 15 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects (leading to less travel cost) for the rich and hence their consumer surplus increases. Before we move on to conclude the welfare analysis, we would like to make the following remark. It should be noted that the welfare effects discussed in this section do not correspond to either Pareto improvements or losses.
and
measure the aggregate welfare of the rich and poor communities, respectively, by integrating over the whole distribution. For example, when falls, some of the individual rich and poor who were already enjoying the service provision may actually end up being worse off as a consequence of the higher equilibrium price that will prevail, even if this results in higher aggregate welfare measured by
and
(p.190) 4.4 Discussion and Policy Considerations
Our analysis of the inequality–neighbourhood interactions so far can be summarized as follows. On the one hand, the poor benefit from the presence of the rich: the higher willingness of the rich to pay for service leads to a larger number of firms providing the service, allowing the poor who live sufficiently close to the firm to purchase the service. This is the positive ‘provision effect’. On the other hand, the larger the proportion of rich, the higher the equilibrium price, hurting the poor, and possibly leading them to choose not to buy the service. This is the negative ‘price effect’. We find that the positive ‘provision effect’ dominates the negative ‘price effect’ when the fraction of poor in the neighbourhood is large, while the opposite holds when this fraction is small. Thus, there is an inverted-U shape relationship between the proportion of poor people in the neighbourhood and their market access and welfare. Our theoretical framework highlights the positive as well as negative externalities imposed by rich agents upon the poorer agents living in the same neighbourhood. Such externalities are extremely relevant and assume special significance in many emerging economies such as Brazil, Russia, China, and India. These economies have experienced prolonged periods of economic boom over the last two decades, accompanied by a widening of the income gap between rich and poor and a conscious policy initiative towards privatization of key merit goods like health care or education (higher education, in particular). This has led to widespread concern within the profession about the welfare of the poor in these rapidly growing economies. Our analysis shows that under private provisioning, the externalities imposed by richer agents on their poorer neighbours could be positive or negative depending on whether the provision effect dominates the price effect, or the other way round. For example, Estache et al. (2001) report that privatization of public utilities has led to an expansion of the outreach of such services in Latin American countries, indicating that the positive provision effect has dominated.
Page 16 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects It is important to recognize that whichever of these two externalities dominates would depend crucially on the proportion of rich vis-à-vis poor agents in the neighbourhood. This implies that welfare of the poor should be nonmonotonically related to the degree of inequality. Indeed, studies by Feng and Yu (2007) and Li and Zhu (2006) lend strong empirical support to our results. Using individual data from the China Health and Nutrition Survey (CHNS), they show that there exists a statistically significant inverted-U relationship between selfreported health status and neighbourhood level inequality. One would expect a similar inverted-U relationship to prevail between access to merit goods under private provisioning and the degree of neighbourhood inequality in many emerging economies which have moved to private provisioning of such merit goods in recent years. (p.191) Finally, our model contributes to several policy debates in the context of public service delivery. In the absence of any agency problem related to service quality, our model suggests that a voucher system would be quite effective in ensuring access to merit goods by the poor (see, for example, Blank, 2000). Such voucher systems are commonly used in many countries to facilitate the outreach of public services for lower income groups. For example, Section 8 housing vouchers in the US provide low-income families with a subsidy that they can use for housing rental. Food stamps provide a voucher for low-income families that can be used to purchase additional food items at the grocery store. The Rashtriya Swasthya Bima Yojana (National Health Insurance Plan) recently initiated by the Government of India allows people below the poverty line to access any hospital, private or public, by paying only a nominal registration fee to join the insurance plan. It is easy to see that targeted subsidized prices of these types increase access of the poor in our model. It has the additional effect of enhancing the welfare of the rich. As demand from the poor increases, additional firms enter the market to serve them resulting in greater accessibility to the product or service for the rich. Since service quality is not an issue in our model, we do not argue for a fully fledged public provision of merit goods or services. We recommend instead an indirect intervention whereby the government encourages the entry of private providers by subsidizing the fixed entry cost. In many developing countries, providing land for private schools or hospitals at subsidized rates is a common practice. Such policies would work in our model since subsidizing the entry cost would lower the equilibrium price, increase the number of firms and, in the process, increase market access of the poor and enhance the welfare of both rich and poor.
5 Poverty, Inequality, and the Possibility of Complete Exclusion of the Poor In Sections 3 and 4 we analysed in detail the generic scenario in which all rich consumers are served but only some poor consumers are served, while others are left out of the market. The model encompasses several other equilibrium Page 17 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects possibilities as in both rich and poor have full market coverage, complete coverage for the rich but no coverage for the poor, and so on. Analyses of these cases are similar and, for the sake of brevity, we do not repeat the detailed analyses in the text and relegate it to Gulati and Ray (2017: App. 7.2). Instead, in this section we first summarize the income ranges of rich and poor under which different equilibrium possibilities arise, and then discuss the implications of poverty and income inequality in characterizing the nature of (p.192) equilibrium. We highlight in particular the possibility of complete exclusion of the poor: a scenario where the service providers cater only for the rich and the poor have absolutely no market access. 5.1 Summary of Different Equilibrium Possibilities
In characterizing the equilibrium below we always consider the scenario where the rich have complete market coverage. Then, depending on whether the poor have complete or partial coverages, we have the following equilibrium possibilities to consider. 5.1.1 Complete Market Access for both Rich and Poor
Complete market access for the poor occurs when their income is high enough. Analysis of this scenario leads to the following proposition. Proposition 4 (a) Firms compete for both consumer types—rich and poor, and all consumers of each type are served only when the rich and poor incomes are such that
(b) Firms compete for the rich, but the marginal poor who are indifferent between two adjacent firms are also indifferent between buying and not buying only if rich and poor incomes are such that
although all consumers of each type are served. 5.1.2 Complete Market Access for the Rich but No Access for the Poor
On the other extreme, complete exclusion of the poor occurs when their income is low enough. When the poor’s income is low, and, at the same time, firms charge a high enough price, it becomes impossible for a poor consumer to afford the product even when they are located at the same location as the firm. Firms completely ignore the presence of the poor and choose the price as if there are
Page 18 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects only rich individuals residing in the neighbourhood. Analysis of this scenario can be summarized in the following proposition. Proposition 5 Firms compete only for the rich and all rich consumers are served, whereas all poor consumers are left out only when rich and poor incomes are such that (p.193)
5.1.3 Partial Market Access for Poor
In Section 3 we discussed one situation of partial market access for the poor when the rich’s income is high enough that firms compete for all rich consumers. Another case of the poor’s partial market access arises when the rich’s income is not that high; it is reasonably high in the sense that the marginal rich consumer is also indifferent between buying and not buying. The following proposition summarizes the analysis of this scenario. Proposition 6 Firms have a monopoly over the poor, and, at the same time, the marginal rich are also indifferent between buying and not buying only when rich and poor incomes are such that
and
All rich consumers are served however. The poor have partial access—only those residing closer to the facilities are served, others are excluded. The results for different equilibrium possibilities are summarized in the following corollary and illustrated in Figure 11.4 by plotting the lower and upper thresholds of incomes for different values of
Page 19 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects the proportion of poor in the neighbourhood. Corollary The results for different equilibrium possibilities can be summarized in the essential ranges of poor income (complemented by the rich or average neighbourhood incomes whenever necessary) as follows. 1. Complete market access for both rich and poor: (a) with the marginal poor who are indifferent between two adjacent firms
Figure 11.4 Different equilibrium possibilities
strictly preferring buying to not buying: (b) with the marginal poor who are indifferent between two adjacent firms indifferent between buying and not buying:
2. Complete market access for the rich but only partial access for the poor: (a) with the marginal rich who are indifferent between two adjacent firms strictly preferring buying to not buying: (i) with high average neighbourhood income, :
(p.194) (ii) with
:
Page 20 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects
(b) with the marginal rich who are indifferent between two adjacent firms indifferent between buying and not buying requiring
:
(p.195) 3. Complete market access for the rich but no access for the poor: 15
and 5.2 Implications of Poverty and Income Inequality
Our analysis of the different equilibrium possibilities summarized in Subsection 5.1 has a number of implications for poverty and income inequality. Upper Threshold for YP From Propositions 1 and 4 it is clear that there exists an upper income threshold for
call it
defined by
such that all poor consumers are served only if Proposition 4(a) shows the existence of another income threshold, such that if income of the poor is above this threshold, then not only all poor consumers are served but, in addition, each firm has to compete with its adjacent firms for both poor and rich customers. Equilibrium price and number of firms reflect this competition (see Gulati and Ray, 2017: App. 7.2.2). Lower Thresholds for YP There are three lower income thresholds for the poor,
such that no poor
consumer is served if Interestingly, whichever threshold is relevant depends on the income of the rich. When the rich’s income is high enough so that firms are competing for the rich, then it follows from Proposition 1 that the lower income threshold for the poor is
Page 21 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects
if rich income is quite high, while the threshold becomes
when rich income is not that high. On the other hand, when rich income is reasonably low in the sense that the marginal rich are indifferent between buying and not buying (Proposition 6), then this lower income threshold becomes (p.196)
Presence of thresholds for poor income emphasizes the implications of absolute levels of poverty in our model: no poor get served by the market when poor income is below the lower threshold, whereas all the poor are served only if poor income is above the upper threshold. When poor income is between the upper and lower thresholds, there are pockets of the neighbourhood where the poor are left out of the market: only those poor who are located closer to the facilities get served, others are excluded. The size of these exclusion pockets increases as the poor’s income decreases. This point can be illustrated using Figure 11.4. Suppose, for example, that there is
proportion of poor, and there is
proportion of rich with income given by the height of point R2. The poor have full market access only if poor income is above the height of point P3. The poor have no market access if poor income is below the height of point P2. In between P2 and P3, both market access and welfare of the poor decrease as the poor’s income decreases. Implication of Income Gap Between Rich and Poor Notice that that is, the lower income threshold of the poor is reduced when the rich’s income is reasonably lower. Thus the poor are better off when the income gap between the rich and poor is lower. The higher income gap between rich and poor has a quite striking implication in generating the possibility of the complete exclusion of the poor as elaborated in the following Subsection 5.3. 5.3 Possibility of Complete Exclusion of the Poor
Interestingly this model demonstrates the possibility of the complete exclusion of the poor from the market: a scenario where service providers completely ignore the presence of the poor and choose the price as if there are only rich individuals residing in the neighbourhood. Proposition 5 identifies the ranges of Page 22 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects rich and poor incomes such that firms exclude poor customers completely when the rich’s income is high enough while the poor’s income is low enough. Figure 11.4 illustrates these income ranges. For example, with
proportion of poor in
the neighbourhood, complete exclusion of the poor occurs when the rich’s income is above R1 (for example, R2) while the poor income is below P2. This possibly complete exclusion of the poor is similar to the findings in the literature that argues that income inequality may lead to social segmentation and club formation to ensure private provision of public services exclusively for the rich (see, for example, Graham, 1998; Jaramillo et al., 2003; Bhattacharya et al., 2016). (p.197) 5.3.1 Implication of the Income Gap
Note once again the implication of a higher income gap between rich and poor. With the poor’s income just below P2, if the rich’s income were below R1, then this possibly complete exclusion of the poor would not have arisen. Similarly, with the rich’s income at R2, the complete exclusion possibility does not arise if the poor’s income is above P2. It is the higher income gap that exposes the poor to this vulnerable situation. With a lower income gap firms find it optimal to enlarge service provision to at least some of the poor by lowering the price, while with a higher income gap they cater only for the rich keeping the price sufficiently high.16 5.3.2 The Case of the Minority Poor
The poor are more likely to be completely excluded when they are in a minority, that is, when
is low: firms may completely ignore the poor even when the rich
are not ultra rich just because the rich are more in number. For example, in Figure 11.4, with the same income levels, R2 for rich and just below P2 for poor, the complete exclusion possibility does not arise when the proportion of poor is but it does arise when the proportion of poor is
.
5.3.3 Policy Considerations in the Presence of a Complete Exclusion Possibility
In a situation of the possibility of complete exclusion of the poor, the role of government policy would be to eliminate (or at least reduce) this possibility. Consider first the targeted subsidized pricing policy discussed in Subsection 4.4. This policy is quite effective in reducing the complete exclusion possibility. Suppose the targeted price subsidy is
per unit. Then the lower threshold for
poor income in Proposition 5 reduces to while the upper threshold for rich income remains unchanged. In Figure 11.4, the upward-sloping curve depicting the lower threshold for poor income shifts down, while the other upward-sloping curve showing the upper threshold for rich income remains unchanged. Then the complete exclusion possibility associated with income distribution,
and
just below
is eliminated.
Page 23 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects On the other hand, the effectiveness of the policy of subsidizing the fixed entry cost is ambiguous. In Figure 11.4 this subsidy shifts down both the upwardsloping curves depicting the lower threshold for poor incomes and the upper threshold for rich incomes. While the downward shift of the lower (p.198) threshold for poor incomes works in eliminating some complete exclusion possibilities (similar to the effect of price subsidy discussed above), the downward shift of the upper threshold for rich incomes works in the opposite direction. For example, without this subsidy, the income distribution, below
and
just
does not result in complete exclusion of the poor. With the
subsidy, if the downward shift of the lower threshold for poor incomes is not large enough so that the point P1 continues to be below the lower threshold, then the complete exclusion possibility is generated. Since this subsidy is not targeted towards generating access for the poor, it might have the adverse impact of facilitating the creation of a club exclusively for the rich. Clearly the way out is to make the entry cost subsidy policy also a targeted one. For example, in India private schools or hospitals get land at a subsidized price on condition that 25 per cent of the capacity (enrolment for schools, beds for hospitals) is targeted towards the economically weaker section of the population. A clause like this facilitates achieving the desirable outcome where at least some of the poor have access to the facilities. 5.4 Comparison with a Single-Income Neighbourhood
In Section 4 we identified scenarios where the poor could be better off living in relatively richer societies. To see how the possibility arises in the simplest possible way it is interesting to compare our model economy with two income groups with a single-income neighbourhood. A single-income neighbourhood refers to a city inhabited by a single income group; that is, there is a measure 1 of consumers with the same income
distributed uniformly along the city
circumference. The single-income neighbourhood model is analysed in Gulati and Ray (2017: App. 7.3) and the relevant comparison is highlighted below. 5.4.1 Feasibility Income Threshold in a Single-Income Neighbourhood
In a single-income neighbourhood it is not feasible for any firm to operate unless the common income is at least
If the income is below this
feasibility threshold, the willingness to pay is so low that it is not possible for the firms to recover the fixed cost of production. The implication for a single-income poor neighbourhood with common income product or service when
is that nobody gets to enjoy the .
17
(p.199) 5.4.2 Comparing Single-Income with Mixed-Income Neighbourhoods
As opposed to a single-income neighbourhood, a mixed-income neighbourhood is the one that we are considering so far where there are
proportion of poor with
income
distributed uniformly
and
proportion of rich with income
along the circumference of the city. Since all three lower income thresholds of Page 24 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects
the poor,
are strictly less than the feasibility threshold, it is clear that the poor are better off staying in the mixed-income
neighbourhood as long as the poor’s income is below this feasibility threshold. At least some poor get to enjoy the product or service in the mixed-income neighbourhood as the firms recover their fixed costs due to the rich’s higher willingness to pay. This is not possible in a single-income poor neighbourhood.
6 Concluding Remarks The main contribution of our research (in this chapter as well as in Gulati and Ray, 2016) is to model the interaction between neighbourhood effects and income inequality in a simple and tractable way by integrating consumers’ income distribution with the spatial distribution of their location. While the basic analytical structure is adapted from the industrial organization literature (Salop, 1979; Bhaskar and To, 1999, 2003; Brekke et al., 2008), this literature does not explore the implications of income inequality. On the other hand, the literature on income inequality has not typically investigated the implications of inequality operating through industrial structure. Our research complements this literature by exploring the impact of income inequality working through price and number of firms. While Gulati and Ray (2016) models the inequality–neighbourhood interaction with a monopolist service provider in a neighbourhood structured by a linear city, in this chapter we extend the analysis to a competitive framework with free entry and exit where the natural neighbourhood structure is a circular city. We find inverted-U shape relationships between income inequality and market access and welfare of the poor. If we compare a cross-section of societies, the poor community as a whole is initially better off living in relatively richer societies by having access to a wider variety of products and services. But, beyond a certain point, the aggregate consumer surplus of the poor starts declining as society becomes richer: the welfare gain from an increase in access to wider varieties of products and services is not enough to offset the corresponding rise in price. Our results square well with the inverted-U relationship between health status and neighbourhood inequality found by Feng and Yu (2007) and Li and Zhu (2006) using the China Health and Nutrition Survey (CHNS) data. We have also identified the possibility of complete exclusion of (p.200) the poor from the market and isolated the higher income gap between rich and poor as the key factor that exposes the poor to this unfortunate possibility. Finally we compare a mixed-income neighbourhood where rich and poor live side by side with a single-income homogeneous neighbourhood and find that the poor are better off living in the mixed neighbourhood as long as the poor’s income is below a certain feasibility threshold.
Page 25 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects Acknowledgements Kaushik Basu is quite fond of the spatial model of competition (see, for example, Basu, 1993, chap. 8; Basu and Mitra, 2016). We are very happy to contribute this chapter to this volume in his honour where we have integrated consumers’ income distribution with the spatial distribution of their location. For helpful discussions and suggestions we thank Kaushik Basu, Mausumi Das, Indraneel Dasgupta, Arghya Ghosh, Parikshit Ghosh, Ashok Kotwal, Dilip Mookherjee, Priya Ranjan, Prabal Roy Chowdhury, and E. Somanathan. All remaining errors are our own. References Bibliography references: Alesina, A. and D. Rodrik (1994), ‘Distributive politics and economic growth’. Quarterly Journal of Economics 109: 465–90. Atkinson, A. B. (1995), ‘Capabilities, exclusion, and the supply of goods’, in K. Basu, P. Pattanaik, and K. Suzumura (eds) Choice, Welfare and Development: A Festschrift in Honour of Amartya Sen. Oxford: Oxford University Press, pp. 17– 31. Atkinson, A. B. and F. Bourguignon (2000), ‘Income distribution and economics’, in A. Atkinson and F. Bourguignon (eds) Handbook of Income Distribution. Amsterdam: North-Holland, pp. 1–58. Basu, K. (1993), Lectures in Industrial Organization Theory. Oxford and Cambridge, MA: Basil Blackwell. Basu, K. and T. Mitra (2016), ‘Nash on a rotary: two theorems with implications for electoral politics’. World Bank, Policy Research Working Paper No. 7701. Bhaskar, V. and T. To (1999), ‘Minimum wages for Ronald McDonald monopsonies: a theory of monopsonistic competition’. Economic Journal 109: 190–203. Bhaskar, V. and T. To (2003), ‘Oligopsony and the distribution of wages’. European Economic Review 47: 371–99. Bhattacharya, S., S. Saha, and S. Banerjee (2016), ‘Income inequality and the quality of public services: a developing country perspective’. Journal of Development Economics 123: 1–17. Blank, R. M. (2000), ‘When can public policy makers rely on private markets? The effective provision of social services’. Economic Journal 110: C34–C49.
Page 26 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects Brekke, K. R., L. Siciliani, and O. R. Straume (2008), ‘Competition and waiting times in hospital markets’. Journal of Public Economics 92: 1607–28. (p.201) Economides, N. (1993), ‘Quality variations in the circular model of variety-differentiated product’. Regional Science and Urban Economics 23: 235– 57. Estache, A., A. Gomez-Lobo, and D. Leipziger (2001), ‘Utilities privatization and the poor: lessons and evidence from Latin America’. World Development 29: 1179–98. Fajnzylber, P., D. Lederman, and N. Lloayza (2002), ‘Inequality and violent crime’. Journal of Law and Economics 45: 1–40. Feng, J. and Y. Yu (2007), ‘Income inequality and health in rural China’. Economic Research Journal 42: 79–88. Foellmi, R. and J. Zweimueller (2006), ‘Income distribution and demand-induced innovations’. Review of Economic Studies 73: 941–60. Gabszewicz, J. J. and J. F. Thisse (1979), ‘Price competition, quality and income disparities’. Journal of Economic Theory 22: 340–59. Glaeser, E., J. Scheinkman, and A. Shleifer (2003), ‘The injustice of inequality’. Journal of Monetary Economics 50: 199–222. Graham, C. (1998), Private Markets for Public Goods—Raising the Stakes in Economic Reform. Washington, DC: Brookings Institution Press. Gulati, N. and T. Ray (2016), ‘Inequality, neighbourhoods and welfare of the poor’. Journal of Development Economics 122: 214–28. Gulati, N. and Ray, T. (2017), ‘Inequality and neighbourhood effects: market access and welfare of the poor,' mimeo, New Delhi: Indian Statistical Institute, available at http://www.isid.ac.in/~tridip/Research/ Festschrift_KaushikBasu_WithAppendix.pdf accessed 26 July 2017. Hotelling, H. (1929), ‘Stability in competition’. Economic Journal 39: 41–57. Jaramillo, F., H. Kempf, and F. Moizeau (2003), ‘Inequality and club formation’. Journal of Public Economics 87: 931–55. Li, H. and Y. Zhu (2006), ‘Income, income inequality, and health: evidence from China’. Journal of Comparative Economics 34: 668–93. Neven, D. and J. F. Thisse (1990), ‘On quality and variety competition’, in J. J. Gabszewicz, J. F. Richard, and L. A. Wolsey (eds) Decision-Making: Games, Econometrics and Optimization. Amsterdam: North-Holland, pp. 175–99. Page 27 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects Persson, T. and G. Tabellini (1994), ‘Is inequality harmful for growth?’. American Economic Review 84: 600–21. Salop, S. (1979), ‘Monopolistic competition with outside good’. Bell Journal of Economics 10: 141–56. Shaked, A. and J. Sutton (1982), ‘Relaxing price competition through product differentiation’. Review of Economic Studies 49: 3–14. Shaked, A. and J. Sutton (1983), ‘Natural oligopolies’. Econometrica 51: 1469– 83. Tirole, J. (1988), The Theory of Industrial Organization. Cambridge, MA: MIT Press. Waldmann, R. (1992), ‘Income distribution and infant mortality’. Quarterly Journal of Economics 57: 1283–302. Notes:
(1) For an extensive review of this literature see Atkinson and Bourguignon (2000). (2) Our adaptation of the Salop (1979) framework is similar to Bhaskar and To (1999, 2003) and Brekke et al. (2008). (3) In this chapter we are not modelling firms’ location choice, rather our interest is to analyse the extent of entry. It is the extent of entry that determines the market access of the poor and hence their welfare. Our justification for this modelling structure is similar to Tirole (1988): ‘Omitting the choice of location allows us to study the entry issue in a simple and tractable way’ (1988: 283). (4) It is well understood that when preferences are identical and homothetic, income distribution does not matter. (5) This is similar to the scenario focused on by Bhaskar and To (1999, 2003) and Brekke et al. (2008). (6) Analysis of the other cases is similar, and we summarize and discuss the relevant results in Section 5. (7) The coefficient matrix of this system of equations forms a circulant matrix (in a circulant matrix each row vector is rotated one element to the right relative to the preceding row vector). The solution is unique since the determinant of a circulant matrix is non-zero if the sum of the elements of a row is non-zero. (8) See, for example, Tirole (1988: 283).
Page 28 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
Inequality and Neighbourhood Effects (9) For the comparative static exercises discussed here and summarized in Proposition 1, we consider large enough changes so that potential discrete jumps in prices (resulting from finiteness of number of firms) can be ignored. (10) While it is obvious from eqution (4) that
, from equation (5) we derive
since, for the scenario under consideration,
implies that
(11) Recall that the reservation utility of the consumer is 0. (12) Note that Figure 11.3 shows the aggregate consumer surplus of poor in proportion to
. To read the consumer surplus from the figure, we have to
multiply the height of each point on the figure by on the inverted-U shape of the curve.
. This will have no impact
(13) These ‘inverted-U’ shape relationships are established in details in Gulati and Ray (2017: App. 7.1). (14) Note the presence of the square term in the expression for
and its
absence from (15) It is shown in Gulati and Ray (2017: App. 7.2.1, footnotes 20 and 21) that the condition condition
in 2(a)(i) is equivalent to the in 3. So the scenario under 2(a)(i) nicely
complements the scenario under 3. (16) This result and mechanism is similar to the dilemma identified in Foellmi and Zweimuller (2006) where also non-homothetic preferences play an important role. (17) Note that
Access brought to you by:
Page 29 of 29
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
Markets, Governance, and Institutions in the Process of Economic Development Ajit Mishra and Tridip Ray
Print publication date: 2017 Print ISBN-13: 9780198812555 Published to Oxford Scholarship Online: December 2017 DOI: 10.1093/oso/9780198812555.001.0001
The Great Recession and Life Satisfaction The Unique Decline for Americans Approaching Retirement Age John Ifcher
, Homa Zarghamee Amanda Cabacungan
DOI:10.1093/oso/9780198812555.003.0012
Abstract and Keywords Using data from the US Centers for Disease Control and Prevention’s Behavioral Risk Factor Surveillance System, we examine the impact of the Great Recession on subjective well-being (as measured by life satisfaction) and attempt to identify disparate effects by age. We find that those approaching retirement age (aged 55 to 64) experienced reduced life satisfaction after the recession, whereas younger working-aged adults did not. The disparate effects by age cannot be explained by income or unemployment trends, but may be explained by wealth effects. For example we find that the life satisfaction of those approaching retirement age, but not of younger working-age adults, is closely correlated with wealth indices (e.g. the Case–Shiller Housing Price Index and the S&P 500 Index). Keywords: retirement, life satisfaction, Great Recession, subjective well-being, wealth effects
1 Introduction During the 2007–09 Great Recession, the American economic environment was bleak: unemployment roughly doubled, median household incomes fell 5 per cent, average household net worth declined by a third, and consumer spending dropped markedly. Each month, the Bureau of Labor Statistics reported massive lay-offs, disappointing job creation numbers, and a dismal outlook for future job growth. The literature studying the impact of the Great Recession on American Page 1 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction households finds that those nearing retirement age were particularly hard hit. For example, using data from the American Life Panel, Hurd and Rohwedder (2010) find that 25 per cent of respondents aged 50–59 lost at least 35 per cent of their retirement savings, and many took early retirement due to unemployment. Chakrabarti et al. (2015) corroborate these findings using data from credit report records and various household surveys. Using asset and labour market data from the Health and Retirement Study, Gustman et al. (2012) find that those approaching retirement age during the Great Recession lost retirement wealth, whereas older cohorts gained retirement wealth when they had approached retirement age prior to the Great Recession. That the above impacts of the Great Recession would be accompanied by a reduction in subjective well-being (SWB) is suggested by the strong positive relationship between income and SWB both within and across countries (e.g., Diener et al., 1995); the strong positive relationship between wealth and SWB (p.203) within countries, controlling for income (e.g., Senik, 2014); and the strong negative relationship between unemployment and SWB, controlling for income (e.g., Di Tella et al., 2001). Further, Reeves et al. (2012) report that the number of suicides in the US in 2007–10 exceeded trend predictions by 4,750; the authors attribute the increase to the recession. Kerr et al. (2016) report an increase in suicides among 40–64 year-olds since 2007 and find a positive relationship between suicide and foreclosure rates for this age group during this period; this relationship does not hold for other age groups, nor is a significant relationship identified between suicide and unemployment rates. Deaton (2012) uses data from the Gallup Healthways Well-Being Index (Gallup Daily Poll) to explore the impact of the Great Recession on SWB in the US. He examines the relationship between SWB and various economic indicators, e.g., the unemployment rate and S&P 500 Index, between 2008 and 2010. The Gallup Daily Poll surveys a random sample of 1,000 Americans each day and started including SWB items in 2008. Deaton identifies a strong positive relationship between the S&P 500 Index and a range of SWB measures using daily data (controlling for income) and using monthly data (controlling for income and unemployment). In contrast, the relationship between unemployment and SWB, controlling for income and the S&P 500 Index, is only significant (and positive) using a life-satisfaction measure; for other SWB measures (e.g. stress experienced yesterday), the unemployment coefficient is insignificant. We explore the effects of the Great Recession on the SWB of adult working-age Americans and conduct various analyses to examine whether those approaching retirement age were more adversely impacted.1 We use a difference-indifferences (DD) approach, comparing the change in pre- to post-recession SWB of those approaching retirement age to younger working-age adults. For younger working-age adults, we find no difference in their pre-to post-recession SWB. In contrast, we find that the post-recession SWB of those approaching Page 2 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction retirement age was significantly lower than pre-recession. We explore channels through which the Great Recession may have differentially impacted the SWB of those approaching retirement age and find evidence suggestive of wealth effects. The result and mechanism are specific to a context in which the institution of retirement is the norm and is funded with personal wealth; this is increasingly relevant as countries develop economically, and older adults become less likely to finance consumption with labour income.
(p.204) 2 Data and Descriptive Statistics We use data from six waves (2005–10) of the Behavioral Risk Factor Surveillance System (BRFSS), the world’s largest telephone survey, conducted by the US Centers for Disease Control and Prevention (CDC). The BRFSS has traditionally collected information on health risk factors, preventive health practices, and access to health care. Between 2005 and 2010, the following life-satisfaction item was included: ‘Overall, how satisfied are you with your life?’ Possible responses were: very satisfied, satisfied, dissatisfied, or very dissatisfied. The survey also collects detailed demographic information, including age, gender, income, marital status, parental status, employment, and education. Table 12.1 presents descriptive statistics.2,3,4 Column 1 includes the entire adult sample (n = 2,005,144), and Columns 2–5 restrict to age cohorts (age 18–34, 35– 54, 55–64, and 65+). Table 12.2 presents the results of estimating a standard ‘happiness regression’. Specifically, life satisfaction is regressed on a host of SWB-correlates identified in the literature, including state and (p.205) (p. 206) month/year fixed effects, and standard results are obtained: life satisfaction positively correlates with income, education, and being female, married, and employed. Further, the standard U-shaped relationship between age and life satisfaction, with a nadir in the 40s age group, holds. For the occasion celebrated by this book, we call the reader’s attention to the quantifiable boost in life satisfaction associated with entering the 65+ cohort. Happy Birthday, Kaushik!5
Page 3 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
Table 12.1 Descriptive statistics
Life satisfaction
Female
Age
Income < $10K
$10K < income < $15K
$15K < income < $20K
$20K < income < $25K
All
Age 18–34
Age 35–54
Age 55–64
Age 65+
(1)
(2)
(3)
(4)
(5)
3.388
3.368
3.374
3.401
3.448
(0.001)
(0.002)
(0.001)
(0.002)
(0.002)
0.504
0.491
0.500
0.501
0.543
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
46.11
26.82
44.49
59.21
74.17
(0.027)
(0.019)
(0.013)
(0.008)
(0.017)
0.051
0.063
0.041
0.046
0.061
(0.000)
(0.001)
(0.000)
(0.001)
(0.001)
0.051
0.054
0.036
0.047
0.087
(0.000)
(0.001)
(0.000)
(0.001)
(0.001)
0.071
0.083
0.051
0.058
0.110
(0.000)
(0.001)
(0.001)
(0.001)
(0.001)
0.087
0.100
0.064
0.075
0.136
(0.000)
(0.001)
(0.001)
(0.001)
(0.001)
Page 4 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
$25K < income < $35K
$35K < income < $50K
$50K < income < $75K
Income > $75K
Did not complete HS
HS graduate
Some college
College graduate
All
Age 18–34
Age 35–54
Age 55–64
Age 65+
(1)
(2)
(3)
(4)
(5)
0.114
0.126
0.088
0.105
0.167
(0.000)
(0.001)
(0.001)
(0.001)
(0.001)
0.149
0.155
0.136
0.155
0.168
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
0.169
0.166
0.182
0.184
0.128
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
0.308
0.252
0.403
0.330
0.145
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.103
0.113
0.085
0.085
0.146
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
0.276
0.283
0.254
0.265
0.333
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.268
0.295
0.261
0.269
0.235
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.353
0.309
0.400
0.382
0.286
Page 5 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
Married
Parent
Employed
Unemployed
Homemaker
Student
Retired
Unable to work
All
Age 18–34
Age 35–54
Age 55–64
Age 65+
(1)
(2)
(3)
(4)
(5)
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.621
0.455
0.724
0.706
0.583
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.440
0.605
0.589
0.121
0.043
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.624
0.681
0.782
0.578
0.143
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
0.0601
0.085
0.063
0.052
0.013
(0.000)
(0.001)
(0.001)
(0.001)
(0.000)
0.076
0.087
0.077
0.053
0.073
(0.000)
(0.001)
(0.001)
(0.001)
(0.001)
0.041
0.127
0.008
0.002
0.001
(0.000)
(0.001)
(0.000)
(0.000)
(0.000)
0.150
0.001
0.012
0.214
0.733
(0.000)
(0.000)
(0.000)
(0.001)
(0.001)
0.050
0.021
0.057
0.101
0.038
(0.000)
(0.000)
(0.000)
(0.001)
(0.001)
Page 6 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
All
Age 18–34
Age 35–54
Age 55–64
Age 65+
(1)
(2)
(3)
(4)
(5)
Observations
2,005,144
278,235
755,635
429,516
541,758
Note: Standard errors in parentheses. Source: Author’s work.
Page 7 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
Table 12.2 Standard ‘happiness’ regression, with dependent variable life satisfaction (1) Female
0.0219*** (0.002)
Age
−0.0126*** (0.000)
Age squared
0.0001*** (0.000)
Log income
0.0957*** (0.002)
Income top code
0.0858*** (0.003)
Never completed HS
−0.0135*** (0.004)
Some college
0.0021 (0.002)
College graduate
0.0639*** (0.002)
Married
0.1730*** (0.002)
Parent
−0.0080*** (0.002)
Unemployed
−0.2120*** (0.005)
Homemaker
0.0229*** (0.003)
Student
0.0392*** (0.007)
Retired
0.0399***
Page 8 of 18
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. Subscriber: OUP-Reference Gratis Access; date: 12 August 2020
The Great Recession and Life Satisfaction
(1) (0.003) Unable to work
−0.3490*** (0.005)
Constant
2.5270*** (0.025)
Observations R-squared
2,005,144 0.112
Note: Robust standard errors in parentheses. *** p