Optimal interregional redistribution and local borrowing rules under migration and asymmetric information

• DOI: http://doi.org/10.1111/jpet.12412
• Author: Darong Dai,Liqun Liu,Guoqiang Tian
• Published date: 01 December 2019
• Date: 01 December 2019

J Public Econ Theory. 2019;21:1266–1285.wileyonlinelibrary.com/journal/jpet1266

|

© 2019 Wiley Periodicals, Inc.

Received: 6 November 2018

|

Accepted: 20 October 2019

DOI: 10.1111/jpet.12412

ORIGINAL ARTICLE

Optimal interregional redistribution and local

borrowing rules under migration and

asymmetric information

Darong Dai

1

|

Liqun Liu

2

|

Guoqiang Tian

3,4

1

Institute for Advanced Research (IAR),

Shanghai University of Finance and

Economics, Shanghai, China

2

Private Enterprise Research Center,

Texas A&M University, College Station,

Texas

3

Department of Economics, Texas A&M

University, College Station, Texas

4

Institute for Advanced Studies in

Finance and Economics, Hubei

University of Economics, Wuhan, China

Correspondence

Darong Dai, Institute for Advanced

Research (IAR), Shanghai University of

Finance and Economics, 200433

Shanghai, China.

Email: dai.darong@mail.shufe.edu.cn;

darongdai@hotmail.com

Abstract

Assuming two types of regions that differ only in the

discount rate, Huber and Runkel show that optimal

federal redistribution is from impatient to patient

regions, and optimal local public debt is higher in

impatient regions than that in patient regions. This

paper extends their analysis by allowing for interregio-

nal migrations and by considering two alternative

regional goals. When the regional governments max-

imize their respective residents’welfare, considering the

interregional migrations does not change Huber and

Runkel’s analysis. When the regional governments

maximize their respective natives’welfare, incorporat-

ing migrations would reverse Huber and Runkel’s

conclusions when migration intensity is sufficiently

high and the regional difference is sufficiently large.

1

|

INTRODUCTION

Interregional redistribution is implemented in developing countries, such as China, as an

important policy tool for fiscal equalization (e.g., Li, 2018) that helps achieve balanced

development (e.g., The Economist, 2016), and is also of policy relevance in developed

economies, such as Canada, France, the UK, and the US (see, Mélitz & Zumer, 2002). The early

literature either focuses on the cross‐region spillover effects due to interregional migrations

(e.g., Breuillé & Gary‐Bobo, 2007; Brown & Oates, 1987; Manasse & Schultz, 1999; Wildasin,

1991) or focuses on the information asymmetries between the federal government and local

governments (e.g., Bordignon, Manasse, & Tabellini, 2001; Cornes & Silva, 2002; Cremer,

Marchand, & Pestieau, 1996; Cremer & Pestieau, 1997; Dai, Liu, & Tian, 2019; Huber & Runkel,

2008; Oates, 1999), but not much is known about their joint effect on the optimal mechanism to

redistribute resources among regional governments. This paper represents an attempt to fill this

gap and focuses on pure redistribution among two regional governments in the presence of both

asymmetric information between the center and regions and the interregional migrations.

To this end, we extend Huber and Runkel’s (2008) two‐period model of a federation

consisting of a benevolent federal government and two regions that differ in the rate of time

preference—which is the private information of each region—by allowing individuals of each

region to migrate to the other region with a given probability and by entertaining each of the

two plausible regional goals in the presence of migration: Maximizing the welfare of the natives

of a region or maximizing the welfare of the residents of a region. We find that considering the

interregional migrations does not change Huber and Runkel’s analysis when the regional

governments maximize their respective residents’welfare. When the regional governments

maximize their respective natives’welfare, on the other hand, incorporating migrations could

reverse Huber and Runkel’s conclusions as summarized below.

First, in the first‐best optimum under full information, the prediction of Huber and Runkel

(2008), namely that the impatient region should borrow more than the patient region and the

federal government should redistribute from the impatient region to the patient region, holds

only if the intensity of mutual migration is low; otherwise, the patient region should borrow

more than the impatient region and the redistribution should be from the patient region to the

impatient region.

Second, in the second‐best optimum under asymmetric information, their prediction that the

impatient region should borrow more than the patient region and the redistribution should be

from the former to the latter holds only if either migration intensity is low or migration

intensity is high and the regional difference in discounting is small; otherwise, the patient

region should borrow more than the impatient region and the redistribution should be from the

patient region to the impatient region.

Third, although it is more likely to be the case that the second‐best optimum achieves less

interregional redistribution than does the first‐best optimum, we also identify conditions under

which the reverse holds true. That is, on the one hand, we extend the set of circumstances

in which the prediction of Huber and Runkel, namely that information asymmetry limits

the ability of the federal government to adopt a tax‐transfer system to redistribute resources

across heterogeneous regions, carries through. On the other hand, we identify reasonable

circumstances in which this prediction is overturned.

As such, the optimal interregional redistribution policy obtained by Huber and Runkel

remains optimal only in special cases of our more general model with interregional migrations.

Importantly, we establish under reasonable circumstances an optimal redistribution policy that

is exactly the opposite of theirs.

We choose the discount factor as the source of heterogeneity among regions due to these two

considerations. First, as argued by Huber and Runkel and empirically demonstrated by Evans

and Sezer (2004), the discount factor is indeed difficult to observe and is more likely to be the

private information of local regions. Second, it is a key factor affecting intertemporal resource

allocation, and hence is relevant in causing individual welfare disparity among regions.

The rest of the paper is organized as follows. Section 2 describes the model of the economy.

In Section 3, we derive optimal interregional redistribution and local borrowing policies under

mutual migration. Section 4 establishes the budget institutions arranged for local governments

so that the full‐information optimum can be implemented by decentralized debt decisions.

Section 5 concludes. Proofs are relegated to the appendix.

DAI ET AL.

|

1267