A Political Economy Model of Earnings Mobility and Redistribution Policy

• Author: RYO ARAWATARI,TETSUO ONO
• Date: 01 June 2015
• Published date: 01 June 2015
• DOI: http://doi.org/10.1111/jpet.12112

APOLITICAL ECONOMY MODEL OF EARNINGS MOBILITY

AND REDISTRIBUTION POLICY

RYO ARAWATARI

Nagoya University

TETSUO ONO

Osaka University

Abstract

This paper presents a politico-economic model that in-

cludes a mutual link between life cycle earnings mobility

and redistributive politics. The model demonstrates that

when an economy features a high opportunity of upward

mobility and high risk of downward mobility, it attains

a unique equilibrium where unskilled, low-income agents

support a low redistribution because of the hope of upward

mobility in future. In contrast, the economy attains multiple

equilibria when mobility opportunity and risk are low: one

is an unskilled-majority equilibrium defined by low mobil-

ity and the other is a skilled-majority equilibrium defined by

high mobility. The paper gives a comparison between the

political equilibrium and the social planner’s allocation in

terms of mobility, and shows that the skilled-majority equi-

librium realizes mobility close to the optimal one.

1. Introduction

Expectations of redistribution affect individuals’ decisions on educational

investment. Their decisions determine the distribution of skilled and

unskilled agents and thus inequality among agents, which in turn has an

Ryo Arawatari, Graduate School of Economics, Nagoya University, Furo-cho, Chikusa-ku,

Nagoya, Aichi 464-8601, Japan (arawatari@soec.nagoya-u.ac.jp). Tetsuo Ono, Graduate

School of Economics, Osaka University, 1-7, Machikaneyama, Toyonaka, Osaka 560-0043,

Japan (tono@econ.osaka-u.ac.jp).

The authors would like to thank associate editor and two anonymous referees for their

valuable comments and suggestions. Financial support of the Nomura Foundation is grate-

fully acknowledged.

Received October 5, 2013; Accepted October 11, 2013.

C2014 Wiley Periodicals, Inc.

Journal of Public Economic Theory, 17 (3), 2015, pp. 346–382.

346

Earnings Mobility and Redistribution Policy 347

impact upon individuals’ votes over redistribution. This feedback mecha-

nism between individual decisions and redistributive politics could produce

multiple equilibria (Glomm and Ravikumar 1995, Saint-Paul and Verdier

1997, Benabou 2000). Based on the concept of a stationary Markov-perfect

equilibrium, Hassler et al. (2003) and Hassler, Storesletten, and Zilibotti

(2007) capture the feedback mechanism and demonstrate multiple equi-

libria that explain the cross-country variations in welfare programs among

democratic countries sharing similar economic backgrounds.

While the analysis by Hassler et al. (2003, 2007) provided a key insight

into redistributive politics, it leaves the earning mobility issue untouched.

In their framework, agents who succeed in education during their youth

retain their skills over their life cycle without any additional effort, and thus

face no risk of downward mobility such as job loss or demotion. In addition,

agents who fail in education during their youth have no second opportu-

nity of becoming skilled at a later stage of their life and thus must accept

their low-income status throughout their life. In an earlier study (Arawatari

and Ono 2009), we considered a mutual link between upward mobility and

redistributive politics by introducing an upward mobility opportunity into

the framework of Hassler et al. (2007).1However, the downward mobility is

omitted from the analysis, and the normative implication of mobility in the

political equilibrium allocation in the presence of earnings mobility is left

untouched.

While previous studies contribute to our knowledge and understand-

ing of mobility and redistributive politics, the following issues still remain

unresolved: (i) how do redistributive politics interact with mobility and dis-

tribution of income in the presence of both upward mobility chance and

downward mobility risk, and (ii) how does the political equilibrium outcome

depart from the commitment solution (called the Ramsey allocation) with

respect to earnings mobility and redistribution? Answers to these questions

will provide more general insights into mobility and redistributive politics.

For the purpose of analysis, we adopt a framework based on that

developed by Hassler et al. (2007) and that extended by Arawatari and

Ono (2009). We further extend this framework by introducing a downward

mobility risk of agents. In particular, we consider agents living in two periods,

youth and old age. In youth, agents undertake educational investments that

determine whether their status in youth is skilled (i.e., rich) or unskilled (i.e.,

poor). At the beginning of old age, unskilled agents have an opportunity of

upward mobility with a probability γ, and can increase the probability of

becoming skilled via reinvestment in education. By contrast, skilled agents

are at risk of downward mobility with a probability γ×θ, but they can reduce

1Early work by Piketty (1995) and Quadrini (1999) considered the effect of earnings mo-

bility on agents’ preferences for redistribution. However, a mutual link between mobility

and redistributive politics is omitted from their analysis because of the assumption of ex-

ogenous mobility or idiosyncratic shocks to mobility.

348 Journal of Public Economic Theory

the probability of becoming poor by reinvestment in education. The expec-

tations of redistribution affect the agents’ decisions on education, which in

turn determines voting behavior over redistribution policy and thus mobility

in the economy.

Focusing on the two key parameters, γand θ, we first present the

political equilibrium allocation via majority voting and investigate how the

two parameters affect the political equilibrium outcome. When the upward

mobility opportunity is high such that γis above the threshold value, the

economy attains a unique, unskilled-majority equilibrium with no taxation

on the old, representing the United States. A high prospect of upward mo-

bility (POUM) in the future gives agents a disincentive to invest in education

in youth, but they support no taxation on the old because of the POUM in

the future (Benabou and Ok 2001, Alesina and La Ferrara 2005).

In contrast, when γis below the threshold value, the economy attains

multiple equilibria, representing some European countries. One equilib-

rium is an unskilled-majority equilibrium with taxation on the old, represent-

ing those Continental European countries that feature high inequality and

low mobility; and the other is a skilled-majority equilibrium with no taxation

on the old, representing Scandinavian countries defined by low inequality

and high mobility. Which equilibrium is realized as an outcome depends on

the expectations of the agents.

The parameter θ, representing the risk of downward mobility, also af-

fects agents’ decisions on education. In particular, a higher θgives agents a

disincentive to invest in education during youth because, for skilled agents,

one’s status in youth is less likely to persist into old age. Given this feature, a

natural prediction is that in an economy with a high prospect of downward

mobility the majority will be unskilled agents who support taxation on the

skilled old. However, this prediction is not true. A higher θimplies a lower

number of the skilled old and thus less redistributive benefit from taxation

on the skilled old. The redistributive benefit is outweighed by the expected

tax burden of the unskilled who may become the skilled via reinvestment

in education. Therefore, a higher downward mobility risk is more likely to

realize an equilibrium supporting the POUM hypothesis. Upward mobility

opportunity and downward mobility risk produce qualitatively similar prop-

erties with respect to political equilibrium characterization: this result was

not shown in Hassler et al. (2007) and Arawatari and Ono (2009).

Another noteworthy feature of the political equilibrium is that the rela-

tive number of upwardly and downwardly mobile agents are completely dif-

ferent between the two types of unskilled-majority equilibrium. The number

of upwardly mobile agents is larger than the number of downwardly mobile

agents in the unskilled-majority equilibrium with no taxation on the old,

representing the United States. However, the opposite result holds in the

unskilled-majority equilibrium with taxation on the old, representing some

Continental European countries. Whether the old are taxed or not criti-

cally affects the relative number of upwardly and downwardly mobile agents.