How Borrowing Constraints Hinder Migration: Theoretical Insights from a Random Utility Maximization Model*

• Published date: 01 April 2020
• DOI: http://doi.org/10.1111/sjoe.12355
• Date: 01 April 2020

Scand. J. of Economics 122(2), 732–761, 2020

DOI: 10.1111/sjoe.12355

How Borrowing Constraints Hinder

Migration: Theoretical Insights from a

Random Utility Maximization Model*

L´ea Marchal

Bielefeld University,DE-33615 Bielefeld, Germany

lea.marchal@ifw-kiel.de

Claire Naiditch

University of Lille, FR-59655 Villeneuve d’Ascq, France

claire.naiditch@univ-lille.fr

Abstract

We provide a theoretical framework to analyze how financial constraints hinder migration.

Introducing wealth heterogeneity and borrowing constraints into a random utility maximization

model of migration, we find evidence of multilateral resistance to migration stemming from

borrowing constraints. We calibrate the model on 22 European countries, and we show that

omittingthe constraints biases upward the estimation of bilateral migration rates. We then simulate

an increase in the bilateral cost of migration to the United Kingdom. We find that omitting the

constraints biases downward the change entailed by the cost increase in the bilateral rates of

migration to all destinations.

Keywords: Discrete choice model; financial constraint; multilateral resistance to migration

JEL classification:C63; F22; J61; O15

I. Introduction

About 386.1 million individuals were potential migrants in 2010. Yet, only

28.9 percent of them ended up migrating; the rest were individuals who

desired to move but did not (from the 2012 Gallup World Poll survey;

Docquier et al., 2015). This high discrepancy between migration intentions

and migration decisions can partly be explained by financial constraints

faced by potential migrants. The poorest individuals lack resources and

are caught in a poverty trap that does not permit them to migrate even

if they intend to do so. Wealthier individuals can afford to migrate to

*We thank two anonymous referees, Fr ´ed´eric Docquier, Hubert Jayet, Paul Klein, Glenn Rayp,

and participants at various conferences for their constructive remarks. We thank the institutions

that support our research: L. Marchal thanks Bielefeld University, the IfW Kiel. and the KCG;

C. Naiditch thanks Lille University and the LEM-CNRS.

C

The editors of The Scandinavian Journal of Economics 2019.

L. Marchal and C. Naiditch 733

some countries but not necessarily to all destination countries (Hatton and

Williamson, 2005).

In this paper, we provide a theoretical model to describe bilateral

migration flows and to explain the observed gap between migration

intentions and decisions. We set up a random utility maximization (RUM)

model of migration in which we explicitly introduce the role of borrowing

constraints into the migration decision. Within this framework, individuals

choose their destination in order to maximize their utility net of bilateral

costs of migration. Some of these costs must be paid upfront. Individuals

are heterogeneous in their wealth and only those who can afford the

upfront costs of migration to a potential destination are able to migrate to

that destination. Heterogeneity in wealth among individuals and in upfront

costs of migration across destinations both ensure that the set of potential

destinations differs across individuals. Wealthier individuals have broader

choice sets than poorer individuals. Thus, this model allows us to analyze

how migration decisions are made when individuals cannot borrow against

their future income.

At the aggregate level, we find that the bilateral rate of migration

between two countries depends on the attributes of the origin and the

destination countries, on the bilateral cost of migration between these two

countries, and on a borrowing constraint effect. The latter effect depends

on the attributes of alternative destinations. Therefore, bilateral rates of

migration do not exhibit the independence from irrelevant alternatives

(IIA) property and multilateral resistance to migration arises.1Besides, the

presence of the borrowing constraints breaks down the symmetry between

destinations so that one destination can be a close substitute to another,

whereas the opposite is not necessarily true.

We calibrate the model on 22 European Union (EU) member states in

2015, and we show that the omission of the borrowing constraints induces

a large upward bias in the estimation of bilateral migration rates (ranging

from 98.5 to 99.8 percent). We then perfor m a counterfactual exercise to

quantify the effect of the borrowing constraints on European migration.

Echoing current discussion on Brexit, we analyze the effect of an increase

of 1 percent in the bilateral cost of migration from Hungary to the United

Kingdom (UK). We find that omitting the borrowing constraints not only

counteracts multilateral resistance effects but also induces a downward bias

in the change in the bilateral migration rate from Hungary to the UK

entailed by the cost increase. We estimate that this downward bias ranges

from 26.31 to 50.22 percent.

1Multilateral resistance to migration can be defined as the influence that the attractiveness and

the accessibility of alternative destinations exert on the bilateral rate of migration between two

countries.

C

The editors of The Scandinavian Journal of Economics 2019.