A gravity model of remittance determinants: evidence from Latin America and the Caribbean

ABSTRACT

A gravity model of remittance determinants: evidence from Latin America and the Caribbean. Regional Studies. This paper constructs a microeconomic model of the motivation for remittances and uses it to explore the macroeconomic determinants. In addition, a new measure of bilateral remittances is used to estimate a gravity model of remittances for 27 Latin American and Caribbean countries and 18 industrialized countries. The results suggest remittances are motivated by a combination of altruism and self-interest, both of which are encapsulated by economic and non-economic variables.

摘要

汇款决定因素的重力模型：来自拉丁美洲与加勒比海地区的证据。区域研究。本文建构一个汇款动机的微观经济模型，并以此探讨巨观经济的决定因素。此外，本文运用衡量双向汇款的崭新方法，评估二十七个拉丁美洲与加勒比海地区国家，以及十八个工业化国家的汇款重力模型。研究结果显示，汇款是受到利他与利己的综合因素所驱动，而两者同时以经济及非经济的变因进行概括。

RÉSUMÉ

Un modèle de gravité des déterminants des transferts de fonds: des résultats provenant de l’Amérique latine et des Caraïbes. Regional Studies. Cet article construit un modèle microéconomique de la motivation d'effectuer des transferts de fonds et l'emploie afin d'en examiner les déterminants macroéconomiques. En outre, on se sert d'une nouvelle mesure des transferts bilatéraux pour estimer un modèle de gravité des transferts relatifs à 27 pays de l'Amérique latine et des Caraïbes et de 18 pays industrialisés. Les résultats laissent supposer que les transferts de fonds sont motivés par à la fois l'altruisme et les intérêts personnels, qui sont rendus tous les deux par des variables économiques et non économiques.

ZUSAMMENFASSUNG

Ein Schwerkraftmodell für die Determinanten von Überweisungen: Belege aus Lateinamerika und der Karibik. Regional Studies. In diesem Beitrag wird ein mikroökonomisches Modell für die Motivation von Überweisungen entwickelt und zur Untersuchung von makroökonomischen Determinanten genutzt. Zusätzlich kommt ein neuer Maßstab für bilaterale Überweisungen zur Schätzung eines Schwerkraftmodells für Überweisungen in 27 Staaten von Lateinamerika und der Karibik sowie in 28 Industrieländern zum Einsatz. Aus den Ergebnissen geht hervor, dass Überweisungen durch eine Kombination von Altruismus und Eigeninteresse motiviert werden, die beide mit wirtschaftlichen und nichtwirtschaftlichen Variablen verknüpft sind.

RESUMEN

Un modelo de gravedad para los determinantes de remesas: ejemplo de Latinoamérica y el Caribe. Regional Studies. En este artículo construimos un modelo microeconómico de la motivación de remesas que nos sirve para analizar los determinantes macroeconómicos. Además, una nueva medida de remesas bilaterales nos permite calcular un modelo de gravedad de las remesas de 27 países latinoamericanos y caribeños y 18 países industrializados. Los resultados indican que las remesas están motivadas por una combinación de altruismo e interés propio; ambos factores quedan incluidos en las variables económicas y no económicas.

KEYWORDS:

• gravity model
• bilateral remittances
• Latin America

关键词:

• 重力模型
• 双向汇款
• 拉丁美洲

MOTS-CLÉS:

• modèle de gravité
• transferts de fonds bilatéraux
• Amérique Latine

SCHLÜSSELWÖRTER:

• Schwerkraftmodell
• bilaterale Überweisungen
• Lateinamerika

PALABRAS CLAVES:

• modelo de gravedad
• remesas bilaterales
• Latinoamérica

JEL:

• C23
• J61
• O11

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed at http://dx.doi.org/10.1080/00343404.2015.1133904

Notes

1. Based on the authors’ own calculations. Data are taken from World Development Indicators (see http://data.un.org/DataMartInfo.aspx#WDI) (data accessed on June 25, 2015).

2. See http://www.worldbank.org/en/news/feature/2012/11/20/remesas-america-latina (accessed on May 7, 2013).

3. Argentina, Belize, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Honduras, Haiti, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, St. Lucia, St. Vincent and the Grenadines, Suriname, Uruguay and Venezuela.

4. Mathematically, .

5. It can be shown that as long as is small enough, the overall effect of an increase in or is negative. The condition (along with the previous condition) is:for which a sufficient condition is:or simply .

6. It also depends on the cost of obtaining credit.

7. A similar correlation is found for Jamaica.

8. Models 3 and 4 are estimated and the fitted values obtained. These values are squared and cubed and subsequently added to the original model. The models without the fitted values are the restricted models and the models with the fitted values are the unrestricted models. The joint significance of the fitted squared and cubed terms is then tested in the unrestricted models. The null hypothesis is that they are jointly equal to zero. Non-rejection of the null hypothesis indicates that the restricted models (without non-linear terms) are acceptable.

9. The results for the components of natural disasters are available from the authors upon request.

10. The results for the components of political risk are available from the authors upon request.

11. The interest rate variable is dropped because the first-stage F-statistic is less than 10. Since this variable was insignificant to begin with, the removal is justified. All results are available from the authors upon request.