Does Institutional Quality in Developing Countries Affect. Remittances?

Does Institutional Quality in Developing Countries Affect Remittances? Emmanuel K.K. Lartey California State University, Fullerton Evelina Mengova Governors State University February 23, 2015 Abstract This paper explores the role of institutions in driving the flow of remittances to developing countries. Based on data for a sample of 90 countries, and panel data estimation techniques, the findings show that improvement in the quality of institutions conducting monetary policy has a positive impact on remittances. The estimates also suggest that an improvement in the operations of institutions of government leads to an increase in the inflow of remittances. Furthermore, the results provide evidence for a positive relationship between the quality of legal systems and remittances, albeit a weak one. Thus, to the extent that remittances have become an important source of external finance, policies promoting sound and accountable local institutions should be a priority for policy makers in developing countries. JEL Classification: F22 F24 F43 O10 Keywords: Remittances, Institutions, Legal systems, Developing countries Correspondence: Department of Economics, California State University, Fullerton, 800 N. State College Blvd, Fullerton, CA 92834. E-mail: elartey@fullerton.edu, Tel: 657-278-7298, Fax: 657-278-3097. College of Business and Public Administration, Governors State University, 1 University Parkway, University Park, IL 60484; USA. E-mail: emengova@govst.edu; Tel.: 708-534-4964. 1

1 Introduction Remittances to developing countries are estimated to have been $327 billion in 2010 and $372 billion in 2011, reaching slightly over $406 billion in 2012. The most recent projections for 2014 and 2015 are $434 billion and $454 billion respectively, according to a World Bank report on global migration and remittances published in 2014. Remittances have become an important component in total international capital flows to developing countries over the last two decades, and in recent years, have more than doubled the level of official development assistance and have become the second largest source of external finance behind foreign direct investment (FDI). There is empirical evidence that remittances are associated with lower poverty indicators and high growth rates (Adams and Page, 2005; Acosta et al., 2008), and that they have served to finance investment in human capital and smooth household consumption (Gupta et al., 2009). There is also evidence, although to a lesser extent, that remittances have funded investment in recipient countries. This is one of the channels through which remittances have been found to positively affect economic growth - via the mitigation of credit constraints in developing countries. Other studies, on the contrary, have indicated that remittances may be harmful to the long-run growth of recipient economies through an appreciation of the real exchange rate, which tends to be detrimental to the tradable sector (Amuedo-Dorantes and Pozo, 2004; Acosta et al., 2009); and through a decline in labor supply, as well as labor market participation rate. Catrinescu et al. (2009), show that remittances are more likely to generate longer-term growth where the quality of political and economic institutions is higher - a result that should provide incentives to policy makers to pursue the development of high-quality, and accountable political and economic institutions. A concern that has been raised among policy makers, 2

though, is the possibility that the observed positive impact of remittances on economic growth and poverty alleviation might serve as a disincentive for institutional reforms and sound macroeconomic policy. Abdih et al. (2012), on the other hand, show that remittances affect the incentives faced by governments, and may therefore have important impact on the quality of domestic governance. They argue, in essence, that access to remittance income makes government corruption less costly for domestic households to bear, and hence such corruption is likely to increase. Rodrik (2004) notes it is a generally accepted view that institutional quality holds the key to prevailing patterns of prosperity across countries. He argues that wealthier countries attract investors because of the presence of effective property rights and the rule of law, and the existence of monetary and fiscal policies that are grounded in solid macroeconomic institutions, whereas poor countries are those where these arrangements are non-existent or ill-formed. He also emphasizes, however, that some empirical research has shown that institutions exert a very strong effect on aggregate incomes, such that poor countries that strengthen property rights of entrepreneurs and investors would likely experience a lasting increase in productive capacity. An interesting and yet important question that has not received as much attention in the literature though, is whether institutions matter for the flow of remittances to developing countries. This is the focus of our paper. Ratha (2003), in a less systematic study, notes that over the period 1996-2000, remittances averaged 0.5 percent of GDP in countries with corruption index that is higher than the median level, in comparison to 1.9 percent in countries with a level lower than the median. The quality of political institutions may therefore, influence the flow of remittances. Moreover, one of the most important determinants of remittances is the cost of transactions in the destination country. Conceivably, an improvement in economic institutions that facilitate economic freedom would serve to reduce such costs, 3

which affect both the volume and value of remittances. In this vein of thought, the quality of economic institutions in general, could be an important determinant of remittances in developing countries. Given that remittances are a significant source of external finance, the quality of institutions in the recipient countries can be expected to affect remittances, particularly in the case where they are driven by a portfolio choice motive, where migrants seek to exploit investment opportunities as a means to allocate their savings optimally between origin and destination countries. The presence of high quality institutions that favor foreign investments would, therefore, serve to attract remittances towards investment opportunities in the recipient country. This paper, thus, makes an essential contribution by analyzing the impact of the quality of institutions on the flow of remittances to developing countries, primarily focusing on the role of both political and economic institutions as determinants of remittances. In a related study, Freund and Spatafora (2008) explore the determinants of remittances, focusing on transactions costs. They find a statistically significant and economically meaningful negative impact of transactions costs on remittances to suggest that, when costs are high, migrants either refrain from sending money home or else remit through informal channels. They estimate that the impact of high transaction costs is that, it significantly reduces recorded remittances; and according to their calculations, a one percentage point reduction in transaction costs raises recorded remittances by 14 23%. Bang et al. (2013) also examine the impact of financial reform on formal remittance inflows to a sample of 84 countries over the period 1990-2005. They find that various dimensions of financial reform have a differential impact on remittances. In their study, increased economic freedom in the financial sector, captured by absence of direct government control over the allocation of credit has a positive and immediate impact. However, improved robustness of 4

financial markets, captured by the effective and apolitical regulations and other policies that enhance financial markets, is found to have a negative impact in the long term. They conclude that the net combined effect suggests that in the long-run, an across-the-board reform has a slightly negative impact on remittances. Our paper differs from the aforementioned ones, not only because of our focus on both political and economic institutions as determinants of remittances, but also because of the data for indicators of institutional quality that we employ in our analysis. Rodrik (2004) argues that institutional quality, as is typically measured and used in most of the empirical literature, does leave a lot of questions unanswered. The most commonly used indexes of institutional quality are based on surveys of foreign and domestic investors, in which the respondents in a particular country are asked whether they consider their investments safe, or how they would rate the rule of law (Kaufmann et al., 2002). The World Bank built on those empirical measures, and developed the Doing Business database, which also feeds into the Economic Freedom of the World (EFW) indicators. These indicators capture investors perceptions, rather than any of the formal aspects of the institutional setting of a given country. They measure how well the rules with regard to property rights are perceived to operate by investors and local businesses and not necessarily what those rules are, and are considered to offer a good description of the institutional environment of a country. We, therefore, utilize data on both economic and political institutions from the Economic Freedom of the World database (Gwartney et al., 2014), together with static and dynamic panel data estimation techniques, to study the dynamics of this relationship between institutional quality in developing countries and remittances. The rest of the study is organized as follows: the next section presents a description of the data and descriptive statistics, and discusses the empirical model and estimation techniques. Section 3 reports and analyzes the empirical results, and provides some policy implications 5

and recommendation. The final section offers some concluding remarks. 2 Data and methodology 2.1 Data and descriptive statistics We employ an unbalanced panel data set comprising 90 developing and transition countries for the period 1970-2012. 1 The data for the quality of institutions indexes, the main variables of interest, comes from the Economic Freedom of the World (EFW) database, whereas the data on remittances and all other variables come from the World Bank s World Development Indicators (WDI) database. We follow the literature to choose the set of control variables. Remittances tend to depend on economic conditions in the home country, and would be considered altruistic in nature where they are inversely related with the level of income, and may be characterized as self-interest where they exhibit a positive relationship with the level of income, as measured by (GDP per capita). Other factors that could potentially explain variability in remittances include domestic investment, domestic inflation and indicators of the quality of institutions, proxied by measures of the effectiveness of government, legal system, monetary policy institutions, and freedom to trade. 2 Table 1 shows descriptive statistics for the dependent variable, remittances, together with the main explanatory variables and the indicators for institutional quality. Of those indicators, the one for legal system and property rights has the lowest average at 1.54, while credit market regulations has the highest one of 1.95. 3 Table 2 displays the pairwise correlation co- 1 The sample was chosen on the basis of data availability. For the institutional quality indexes, there are 6 observations between 1970 and 2000 as data was only reported once every 5 year within that interval. Thus, we have 5 year averages for all other variables in order to match observations for institutional indexes for that period. 2 See for instance, Singh et al. (2011) for some details on the determinants of remittances. 3 All of the data, with the exception of the real interest rate, are expressed in logarithms. 6

efficients between the indicators of institutional quality, remittances and GDP per capita, and shows a positive and statistically significant correlation between remittances and GDP per capita. In addition, there is a positive and statistically significant correlation between remittances and each of the indicators of institutional quality, except the legal system index. The positive correlation with the indicator for sound money, which is a proxy for monetary policy, shows that remittances are positively associated with a stable monetary system, characterized by lower inflation and the freedom to own foreign currency back accounts. The positive association with the government variable also indicates that a well functioning political structure may be a relevant factor in attracting remittances to these countries. 2.2 Model specification and estimation In order to further examine the relationship between the quality of institutions and remittances, we specify and estimate a reduced form equation for the determinants of remittances, where the hypothesized general model, following prior studies, is given by, = ( ) (1) where is remittances, is GDP per capita, is domestic investment, is financial development index, is inflation and is institutional quality index. First, a static panel model is estimated using the fixed effects (within) estimator. The static model is given by, = 0 + + (2) where represents remittances, is a vector of explanatory variables featuring GDP per capita and indicators of institutional quality, is unobserved country specific effect, and 7

is an error term. GDP per capita, however, is potentially an endogenous regressor in the specified model, as the literature has documented the impact of remittances on GDP. Moreover, remittances may be serially correlated, in which case the static model presents some concerns in the form of omitted variable bias and hence a mis-specification error. Thus, subsequently, we specify a dynamic panel model by introducting the lagged level of remittances into a specification that is estimated using a generalized method of moments (GMM) estimator, and which is designed to deal with potential endogeneity in all explanatory variables. In particular, this estimator accounts for endogeneity due to the introduction of the lagged dependent variable as a regressor. In effect, this is done as a check for the consistency of the estimates from the static model. The dynamic equation is represented by an autoregressive-distributed lag model of the form = 1 + 0 ( ) + + (3) This is a dynamic model for,where 1 is the one period lag of, is a vector of other explanatory variables, and ( ) is a vector of associated polynomials in the lag operator. Applying the fixed-effects (within) estimator to equation (3) yields a biased and inconsistent estimate of the coefficient on the lagged dependent variable because it makes use of a transformation by which the country specific effect is eliminated, and which culminates in a correlation between the lagged dependent variable and the error term. We, therefore, utilize the System GMM estimator which combines an estimator in first-differences with an estimator in levels. The inclusion of a levels equation allows the use of information on cross- 8

country differences. 4 This estimation technique generates internal instruments by using the lagged levels and lagged differences of the explanatory variables as instruments under two conditions: i) that there is no serial correlation in the errors and ii) the differences of the explanatory variable and the errors are uncorrelated. Two specification tests, Sargan test for over-identifying restrictions and Arellano-Bond test for second-order serial correlation [(AR) 2] are applied to assess the validity of the instruments and consistency of the estimates. 3 Results 3.1 Fixed Effects Estimates The results obtained from the first set of regressions based on various specifications are presented in Table 3, and show that there is a positive and statistically significant relationship between sound monetary policy and remittances, as well as a positive and statistically significant relationship between the indicator of government effectiveness and remittances. The indicator for credit market regulations bears a positive and statistically significant coefficient, whereas the relationship between the indicator of legal system and property rights and remittances is not statistically significant. 5 The index for freedom to trade internationally, also, has a positive and significant coefficient. The sequential introduction of the indicators of freedom to trade, general regulations affecting the business environment, and the real exchange rate do not change how government effectiveness and macroeconomic stability via monetary policy impact the inflow of remittances. The initial observations can be recapitulated as follows. Government effectiveness, sound 4 These estimators have been widely applied and discussed in a number of studies. See Blundell and Bond (1998) for details on the system GMM estimators. 5 The indicator for legal system and property rights is found to be statistically insignficant in most specifications, so we drop it and only consider it in later analysis. 9

monetary policy, freedom to trade and credit market regulations are relevant drivers of remittances, as they bear coefficient estimates that are robust across specifications. Also, the financial development indicator, domestic credit, is found to have a statistically significant relationship with remittances. Table 4 presents the estimates for the second set of regressions, which are based on various specifications that introduce an additional financial system indicator, the real interest rate, while maintaining the key variables of interest. The coefficients on government and sound monetary policy remain consistent in sign and statistical significance across all specifications. Total domestic credit is significant only in the specification without the real interest rate, whereas the real interest rate bears a negative and statistically significant coefficient in all cases where it is considered. The index for credit market regulations is also positive and statistically significant across specifications, but in the case of the freedom to trade index, the coefficient estimates are sensitive to the model specification, such that each index is positive and significant in columns (5) and (6) only. Nevertheless, these results are generally consistent with those presented in Table 3. In summary, the estimates based on static models suggest that an improvement in key institutions in recipient countries have a positive impact on inflow of remittances. 3.2 System GMM Estimates The next set of results accounts for the potential endogeneity of the indicators of institutional quality, against the backdrop of the existing literature on the impact of remittances on sound macroeconomic policy and institutional reforms,aswellasthedocumentedimpactof remittances on both the level and growth rate of GDP. The specifications here also consider the potential persistence in remittances, and hence serve as a robustness check for the initial results using the static models. The results, which are given in Table 5, show that all the 10

variant specifications of the dynamic model, with the exception of column (1), satisfy the Sargan and AR(2) tests. Some of the specifications of the model incorporate the index for regulation of investment environment and the one-period lag of GDP per capita as regressors. The coefficient on the lagged dependent variable ( 1 ) is positive and statistically significant, which indicates that the dynamic model is the valid representation of the empirical model of remittances determination, and underscores the reliability of these estimates. 6 Government and sound monetary policy are positive and significant across all specifications, and trade freedom is positive and significant in some cases. Market regulation variables, however, are not statistically significant. With respect to the control variables, the indicators of financial development, domestic credit and real interest rate, are in general, negative and statistically significant. One interesting observation is the different signs on the contemporaneous (positive) and lagged (negative) GDP per capita, supporting the self-interest and altruistic motives for remittances, respectively. Table 6 presents estimates for each of the model specifications shown in Table 5, but with heteroscedasticity-consistent standard errors. Notably, the coefficient estimates for government and sound monetary policy are robust to these estimators, whereas freedom to trade index is significant in column (3) only. Moreover, although observed in column (3) only, the estimates for contemporaneous GDP per capita and the lagged counterpart are positive and negative respectively, and both are statistically significant. To further assess the robustness of the estimates capturing the relationship between institutional quality and remittances, the growth rate of GDP is introduced instead of GDP per capita, into a set of specifications of the dynamic model. The results are given in Table 7. Controlling for both the contemporaneous and lagged level of GDP growth rate, and using 6 This also implies the static model estimates are subject to omitted variable bias, and therefore potentially inconsistent, inefficient and unreliable. 11

estimation techniques with and without robust standard errors, the estimates show that government and monetary institutions indexes remain positive and statistically significant across different model specifications, likewise the index of freedom to trade. The index for credit market regulation, on the other hand, remains statistically insignificant in all the various representations of the model. Furthermore, in order to assess the existence of any potential nonlinear effects in some of the main institutional indexes on remittances, a dummy variable is generated, such that it is equal to 1 if the index is greater than the median observation and zero otherwise, and then interacted with the relevant indicator. This is done for the indicators of the quality of the legal system and sound monetary policy, and introduced into the final set of specifications. 7 The final set of estimates is presented in Tables 8 and 9. Based on the specification for which the instruments are valid [Table 8, column (3)], the estimates show that the legal system and property rights index is positive and statistically significant, and so is the interaction term. Sound monetary policy is statistically significant and positive in columns (3) and (4) of Table 9, but the interaction term is not significant in that case. It is noteworthy that government effectiveness and sound monetary policy coefficients are robust to the various specifications in this final set of regressions as well. 3.3 Analysis of Results Based on the estimator that accounts for potential endogeneity issues, and the specifications that potentially resolve the issue of omitted variable bias, the results indicate that an improve- 7 We limit this analysis to the two indexes for the reasons that follow. Given that the estimates for the legal system index were generally omitted from the analysis because they were observed to be statistically insignificant, and the likelihood that the effectiveness of institutions that represent the legal system do matter for remittances that are sent for investment purposes, we hypothesize that the relationship may be nonlinear. The choice of the monetary policy index was based on the fact that monetary policy is crucial to macroeconomic stability, and therefore an important index. 12

ment in the quality of institutions in charge of monetary policy, and hence, in the conduct of monetary policy has a positive impact on remittances, and that this impact increases with the quality of such institutions. In terms of the coefficient estimate, per Tables 5 and 6, an improvement in the quality of monetary policy institutions, as captured by a percentage point increase in the index of sound money, will result in an increase in remittances by, between 0.56% and 1.4%. Another main finding is that a percentage point increase in the effectiveness of government will be associated with an increase in remittances between 1.2% and 1.6%, which is consistent with the finding in Bang et al. (2013), and suggests that a decrease in direct government control or participation in the private sector has positive impact on remittances. In general, these results are robust to a variety of specifications of the dynamic model. Sound monetary policy and a well functioning government are critical to economic freedom, and these results seem to suggest that migrants tend to be motivated to send remittances in order take advantage of macroeconomic environments that are favorable to better economic performance. The quality of these institutions should therefore, be critical to the investment motive for remittances. The results provide evidence that weakly suggests that an improvement in the legal system and property rights may be associated with an increase in remittances, but that the effect is lower for countries with an index value greater than the median level. This potentially captures the relevance of improving such institutions to attract remittances for investment purposes in countries with low quality legal system. Notable also is the finding that credit market regulation is not a relevant driver of remittances to developing countries, which likely underscores the fact that financial markets are in general under-developed and such regulations have limited implications for remittances that are driven by investment motives. Another interesting result to note is the negative and significant coefficient on domestic credit, which is robust to variant specifications of the model. Given that this is an important 13

indicator of financial development, this finding may signal some substitutability between remittances and other forms of external finance in countries with better developed financial systems. An alternative explanation would be that, better functioning domestic banking system improves the availability of credit, and lowers returns on savings, which may discourage migrants who send remittances to take advantage of relatively higher returns. 3.4 Policy Implications The analysis presented in Ratha (2003), suggests that institutional quality could matter for the flow of remittances to developing countries. The empirical observations, on the basis of the results of this study, provide some support for that hypothesis, indicating that a policy environment that promotes a well regulated use of monetary policy, and hence, a stable and low inflation rate, would serve to attract more remittances. A better functioning government that encourages privatization of industries and less government investment and enterprises, will also foster the inflow of remittances, and thus provide additional source of external finance, which is key to the economic growth of many developing countries. Arguably, monetary policy institutions and institutions of government, in general, play an important role in these countries with implications for other institutions; and to the extent that they affect economic freedom, they should be important determinants of the amount of remittances sent to these countries. Furthermore, but to a lesser extent based on our findings, the case could also be made that policies that are geared toward enhancing judicial independence, impartial courts, better protection of property rights, contract enforcement, and the integrity of the legal system as a whole, could be essential elements in driving remittances to developing countries. 14

4 Conclusion This paper presented an empirical study of a group of developing and transition countries, exploring the impact of the quality of institutions on remittances, which are now considered an important source of external finance to developing countries. The focus of this research was on the quality of monetary institutions responsible for conducting sound monetary policy, effectiveness of institutions of government, and the quality of local legal systems, in charge of setting a predictable and stable legal framework, and enforcement of contracts, among others. Our results suggest that the quality of institutions related to monetary policy and institutions of government have a positive impact on remittances. The results also provide support, although a weaker one, for a positive relationship between the quality of legal systems and remittances. The effectiveness of these institutions would be an important element affecting the flow of remittances to developing countries, and therefore, investing in high-quality institutional framework should be a priority for policy makers in these emerging countries. References [1] Abdih, Y., R. Chami, J. Dagher and P. Montiel (2012). "Remittances and Institutions: Are Remittances a Curse?", World Development, 40 (4), 657 666. [2] Acosta, P., Calderón, C.; Fajnzylber, P., and López, H. (2008). What is the Impact of International Migrant Remittances on Poverty and Inequality in Latin America?, World Development, 36 (1), 89-114. [3] Acosta, P., Lartey, Emmanuel K.K., and Mandelman, F. (2009). Remittances and Dutch Disease., Journal of International Economics, 79, 102-116. 15

[4] Adams, R. and J. Page (2005). Do International Migration and Remittances Reduce Poverty in Developing Countries?, World Development, 33 (10), 1645 1669. [5] Amuedo-Dorantes, C. and Pozo S. (2004). Workers Remittances and the Real Exchange Rate: A Paradox of Gifts., World Development, 32 (8), 1407-1417. [6] Arellano, Manuel and Stephen Bond, Some tests of specification for Panel Data: Monte Carlo Evidence with an Application for Employment Equations, Review of Economic Studies, 58 (1991):277-97. [7] Bang, J., Mitra. A and Wunnava. P. (2013) Financial Liberalization and Remittances: Recent Longitudinal Evidence, IZA DP No. 7497 [8] Blundell, Richard and Stephen Bond, Initial conditions and Moment Restrictions in dynamic panel data models, Journal of Econometrics, 87 (1998):115-43. [9] Bond, Stephen R., Hoeffler, Anke and Temple, Jonathan R.W., (2001) GMM Estimation of Empirical Growth Models, CEPR Discussion Paper No. 3048. [10] Catrinescu, N., M. Leon-Ledesma, M. Piracha and B. Quillin (2009), "Remittances, Institutions and Economic Growth," World Development, 37 (1), 81 92 [11] Freund, C. and Spatafora, N. (2008) Remittances, transaction costs, and informality, Journal of Development Economics, 86 (2), 356-366. [12] Gupta Sanjeev, Catherine A. Pattillo and Smita Wagh (2009), Effect of remittances on poverty and financial development in Sub-Saharan Africa, World Development, 37 (1), 104 115. 16

[13] Gwartney, James D., Joshua C. Hall, and Robert Lawson (2014). Economic Freedom of the World: 2012 Annual Report. Vancouver, BC: The Fraser Institute. Data retrieved from www.freetheworld.com. [14] Kaufmann, D., Kraay, A. and Zoido, P. (2002) Governance Matters II: Updated Indicators for 2000-01, World Bank Policy Research Working Paper No. 2772. [15] Lartey, Emmanuel K.K.,"Remittances, Investment and Growth in Sub-Saharan Africa" (2013), Journal of International Trade and Economic Development 22:7, 1038-1058 [16] Ratha, D. (2003), "Workers Remittances: An Important and Stable Source of External Development Finance" Chapter 7) in Global Development Finance: Striving for Stability in Development Finance, Washington, DC: World Bank, pp. 157-175. [17] Rodrik, D. (2004). Getting institutions right: Institutions and economic performance (Forum). CESifo DICE report: Journal for Institutional Comparisons; the international platform of Ludwig-Maximilians University s Center for Economic Studies and the Ifo Institute, 2(2). [18] Singh, Raju J., Haacker, M.; Lee, Kyung-woo; Le Goff, Maelan, (2011) Determinants and macroeconomic impact of remittances in Sub-Saharan Africa, Journal of African Economies, 20 (2), 312-40. 17

A Appendix-Variables Description and Data Sources Personal remittances, received (current US$) Domestic credit provided by financial sector (% of GDP) Inflation, consumer prices (annual %) Real effective exchange rate index (2010 = 100) Real interest rate (%) GDP growth (annual %) GDP per capita (constant 2005 US$) Gross fixed capital formation (Investment) (% of GDP) Economic Freedom of the World (EFW) Index components summary: 1. Size of Government: a. Government Consumption, b. Transfers and Subsidies, c. Government Enterprises and Investment, d. Top Marginal Tax Rate. 2. Legal Systems and Property Rights: a. Judicial Independence, b. Impartial Courts, c. Protection of Property Rights, d. Military Interference in Rule of Law and Politics, e. Integrity of the Legal System, f. Legal Enforcement of Contracts, g. Regulatory Restrictions on the Sale of Real Property, h. Reliability of Police, i. Business Costs of Crime. 3. Sound Money: a. Money Growth, b. Standard Deviation of Inflation, c. Inflation: most recent year, d. Freedom to Own Foreign Currency Bank Accounts. 4. Freedom to Trade Internationally: a. Tariffs, b. Regulatory Trade Barriers, c. Black-market Exchange Rates, d. Controls of the Movement of Capital and People. 5. Regulations: a. Credit Market Regulations, b. Labor Market Regulations, c. Business Regulations. 18

B Appendix-List of Countries in the Sample Albania Guatemala Nepal Argentina Guyana Nicaragua Armenia Haiti Niger Azerbaijan Honduras Nigeria Bangladesh Hungary Oman Barbados India Pakistan Belize Indonesia Panama Benin Iran, Islamic Rep. Papua New Guinea Bolivia Jamaica Paraguay Botswana Jordan Peru Brazil Kazakhstan Philippines Bulgaria Kenya Poland Burkina Faso Kyrgyz Republic Romania Cambodia Latvia Russian Federation Cameroon Lesotho Senegal China Lithuania Sierra Leone Colombia Macedonia, FYR Slovak Republic Congo, Rep. Madagascar South Africa Costa Rica Malawi Sri Lanka Cote d'ivoire Malaysia Syrian Arab Republic Croatia Mali Tanzania Dominican Republic Mauritania Thailand Ecuador Mauritius Togo Egypt, Arab Rep. Mexico Trinidad and Tobago El Salvador Moldova Tunisia Estonia Mongolia Turkey Ethiopia Morocco Uganda Fiji Mozambique Ukraine Gabon Myanmar Venezuela, RB Ghana Namibia Zimbabwe 19

Table1 DescriptiveStatistics Variables OBS MEAN SD MIN MAX Remittances 1203 19.33 2.52 8.37 24.95 GDP per capita 1587 7.15 1.14 4.74 9.63 Domestic Credit 1507 3.54 0.77-1.14 5.67 Investment/GDP 1507 3.03 0.32 0.69 4.30 Government 1503 1.83 0.27-0.36 2.29 Legal System/Property Rights 1427 1.54 0.30 0.14 2.13 Sound Monetary Policy 1519 1.92 0.35-2.30 2.28 Freedom to Trade 1461 1.75 0.44-2.30 2.23 Credit Market Regulation 1528 1.95 0.43-1.61 2.30 Table 2 Correlation Matrix (1) (2) (3) (4) (5) (6) (7) (1) GDP per capita 1.000 (2) Remittances 0.178*** 1.000 (0.000) (3) Government -0.025 0.293*** 1.000 (0.350) (0.000) (4) Legal System 0.398*** 0.004-0.076*** 1.000 (0.000) (0.888) (0.005) (5) Sound Monetary Policy 0.231*** 0.231*** 0.196*** 0.187*** 1.000 (0.000) (0.000) (0.000) (0.000) (6) Freedom to Trade 0.320*** 0.171*** 0.283*** 0.368*** 0.394*** 1.000 (0.000) (0.000) (0.000) (0.000) (0.000) (7) Credit Market Regulation 0.257*** 0.158*** 0.370*** 0.238*** 0.431*** 0.600*** 1.000 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Notes: p-values are in parentheses. 20

Table 3 Static Model - Fixed Effects Estimator (I), Dependent variable: Remittances Regressors (1) (2) (3) (4) (5) Inflation, CPI -0.076* -0.077* -0.068-0.064-0.166*** (0.081) (0.077) (0.120) (0.142) (0.009) GDP per capita 2.220*** 2.284*** 2.130*** 1.918*** 1.905*** (0.000) (0.000) (0.000) (0.000) (0.000) Investment (% of GDP) 0.131 0.146 0.060 0.044 0.084 (0.483) (0.429) (0.746) (0.814) (0.779) Domestic Credit (% of GDP) 0.330*** 0.322*** 0.366*** 0.321*** 0.270* (0.002) (0.003) (0.001) (0.003) (0.091) Government 1.002*** 1.036*** 0.852*** 1.142*** 0.978*** (0.000) (0.000) (0.000) (0.000) (0.003) Legal System 0.043 (0.834) Sound Monetary Policy 0.414** 0.331 0.273 0.539*** 1.069*** (0.040) (0.101) (0.174) (0.005) (0.000) Credit Market Regulation 1.260*** 1.344*** 1.212*** 0.605** (0.000) (0.000) (0.000) (0.043) Freedom to Trade 0.617*** 0.387*** 0.840*** (0.000) (0.010) (0.005) Regulation 1.766*** (0.000) Real Exchange Rate -0.752** (0.029) Hausman (0.000) (0.000) (0.000) (0.000) (0.000) Observations 937 946 943 937 477 Countries 84 84 84 84 40 Notes: ***Significant at 1% level, **5% level, *10% level p-values are in parentheses 21

Table 4 Static Model - Fixed Effects Estimator (II), Dependent variable: Remittances Regressors (1) (2) (3) (4) (5) (6) GDP per capita 2.456*** 2.311*** 2.422*** 2.251*** 2.161*** 1.832** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Investment (% of GDP) -0.405** -0.397** (0.036) (0.040) Domestic Credit (% of GDP) 0.043 0.052 0.035 0.038 0.157 0.258* (0.720) (0.664) (0.764) (0.741) (0.266) (0.085) Real Interest Rate -0.025*** -0.024*** -0.021*** -0.019*** -0.026*** (0.000) (0.000) (0.000) (0.000) (0.000) Real Exchange Rate -1.214*** -0.676** (0.000) (0.025) Government 1.333*** 1.466*** 1.473*** 1.750*** 1.486*** 1.274*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Sound Monetary Policy 1.454*** 1.598*** 1.106*** 1.326*** 1.625*** 1.081*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Credit Market Regulation 0.843*** 1.263*** 1.042*** 0.978*** (0.001) (0.000) (0.001) (0.001) Freedom to Trade -0.029 0.006 0.048 0.152 0.931*** 0.844*** (0.852) (0.970) (0.756) (0.323) (0.006) (0.002) Regulation 0.983** 1.277*** (0.018) (0.002) Hausman (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Observations 814 816 859 861 444 535 Countries 80 80 82 82 41 42 Notes: ***Significant at 1% level, **5% level, *10% level p-values are in parentheses 22

Table 5 Dynamic Model - System GMM Estimates (I), Dependent variable: Remittances Regressors (1) (2) (3) (4) (5) Remittances (-1) 0.869*** 0.860*** 0.875*** 0.871*** 0.713*** (0.000) (0.000) (0.000) (0.000) (0.000) GDP per capita 0.150*** 0.154*** 1.615*** 0.162*** 0.311*** (0.000) (0.000) (0.000) (0.000) (0.000) GDP per capita (-1) -1.634*** (0.000) Investment (% of GDP) -0.018 (0.875) Domestic Credit (% of GDP) -0.124*** -0.201*** -0.075** -0.216*** -0.214*** (0.000) (0.000) (0.011) (0.000) (0.000) Real Interest Rate -0.006*** -0.006*** -0.003 (0.005) (0.003) (0.207) Real Exchange Rate -0.705*** (0.000) Government 0.907*** 1.156*** 0.866*** 1.170*** 1.400*** (0.000) (0.000) (0.000) (0.000) (0.000) Sound Monetary Policy 0.564*** 0.497*** 0.439** 0.552*** 1.441*** (0.001) (0.006) (0.034) (0.003) (0.000) Credit Market Regulation 0.075 0.244 0.168 (0.541) (0.147) (0.153) Freedom to Trade 0.466** 0.212 0.353** 0.218 0.858*** (0.019) (0.294) (0.048) (0.294) (0.004) Regulation -0.086 0.005 (0.741) (0.975) Serial correlation test (0.261) (0.494) (0.538) (0.525) (0.143) Sargan test (0.026) (0.111) (0.680) (0.112) (0.640) Instruments 40 40 40 40 41 Observations 781 673 827 673 358 Countries 83 72 85 72 36 Notes: ***Significant at 1% level, **5% level, *10% level; p-values are in parentheses; 23

Table 6 Dynamic Model - System GMM Estimates (II), Dependent variable: Remittances Regressors (1) (2) (3) (4) (5) Remittances (-1) 0.848*** 0.829*** 0.858*** 0.834*** 0.711*** (0.000) (0.000) (0.000) (0.000) (0.000) GDP per capita 0.180 0.219 1.534*** 0.247 0.325 (0.200) (0.170) (0.000) (0.122) (0.289) GDP per capita (-1) -1.669*** (0.000) Investment (% of GDP) -0.152 (0.371) Domestic Credit (% of GDP) -0.130* -0.211* -0.102-0.215* -0.245* (0.090) (0.070) (0.324) (0.057) (0.097) Real Interest Rate -0.001-0.001 0.000 (0.848) (0.832) (0.985) Real Exchange Rate -0.963** (0.035) Government 1.202** 1.523** 1.139** 1.573** 1.578** (0.021) (0.013) (0.025) (0.012) (0.050) Sound Monetary Policy 0.561 0.713* 0.683* 0.761* 1.406** (0.131) (0.085) (0.063) (0.079) (0.033) Credit Market Regulation 0.065 0.069 0.245 (0.763) (0.823) (0.290) Freedom to Trade 0.347 0.043 0.627* 0.044 0.924 (0.336) (0.929) (0.060) (0.928) (0.108) Regulation -0.323 0.003 (0.555) (0.996) Serial correlation test (0.253) (0.358) (0.514) (0.381) (0.167) Instruments 40 40 40 40 41 Observations 781 673 827 673 358 Countries 83 72 85 72 36 Notes: ***Significant at 1% level, **5% level, *10% level p-values are in parentheses Sargan test statistic is unavailable for one-step (robust) estimation in Stata. 24

Table 7 Dynamic Model - System GMM Estimates (III), Dependent variable: Remittances Regressors (1) (2) (3) (4) (5) (6) Remittances (-1) 0.867*** 0.837*** 0.858*** 0.836*** 0.846*** 0.821*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) GDP Growth 0.012*** 0.013*** 0.010*** 0.012*** (0.000) (0.002) (0.000) (0.007) GDP Growth (-1) 0.018*** 0.027*** (0.000) (0.000) Domestic Credit (% of GDP) -0.057** -0.072-0.092*** -0.088-0.062** -0.062 (0.030) (0.348) (0.000) (0.244) (0.028) (0.404) Real Interest Rate -0.002-0.000-0.003-0.001 (0.460) (0.996) (0.140) (0.884) Government 1.124*** 1.323*** 1.416*** 1.584*** 1.270*** 1.504*** (0.000) (0.008) (0.000) (0.009) (0.000) (0.010) Sound Monetary Policy 0.502*** 0.793** 0.446** 0.847** 0.663*** 1.012*** (0.006) (0.012) (0.019) (0.016) (0.003) (0.004) Credit Market Regulation 0.075 0.222 0.241 0.166 0.046 0.093 (0.571) (0.316) (0.178) (0.566) (0.794) (0.748) Freedom to Trade 0.425*** 0.546* 0.141 0.177 0.346** 0.081 (0.006) (0.089) (0.323) (0.640) (0.036) (0.836) Serial correlation test (0.413) (0.397) (0.546) (0.487) (0.219) (0.103) Sargan test (0.201) (0.121) (0.335) Instruments 39 39 40 40 40 40 Observations 835 835 681 681 680 680 Countries 86 86 73 73 73 73 Notes: ***Significant at 1% level, **5% level, *10% level p-values are in parentheses; columns (2), (4), (6) are estimates with robust standard errors Sargan test statistic is unavailable for one-step (robust) estimation in Stata. 25

Table 8 Dynamic Model - System GMM Estimates (IV), Dependent variable: Remittances Regressors (1) (2) (3) (4) Remittances (-1) 0.845*** 0.822*** 0.864*** 0.835*** (0.000) (0.000) (0.000) (0.000) GDP per capita 0.238*** 0.251* (0.000) (0.092) GDP growth 0.012*** 0.013*** (0.000) (0.002) Domestic Credit (% of GDP) -0.193*** -0.215* -0.046* -0.068 (0.000) (0.076) (0.090) (0.386) Government 1.051*** 1.317*** 1.046*** 1.308*** (0.000) (0.007) (0.000) (0.008) Sound Monetary Policy 0.452** 0.592 0.471*** 0.780** (0.011) (0.114) (0.008) (0.014) Credit Market Regulation 0.140 0.116 0.105 0.221 (0.233) (0.591) (0.408) (0.316) Freedom to Trade 0.405* 0.425 0.406** 0.541* (0.056) (0.262) (0.018) (0.086) Legal System 0.247** 0.147 0.245** 0.187 (0.021) (0.496) (0.032) (0.348) Dummy*Legal System 1-0.052* -0.041-0.076*** -0.050 (0.074) (0.466) (0.005) (0.364) Serial correlation test (0.255) (0.263) (0.424) (0.405) Sargan test (0.027) (0.223) Instruments 41 41 41 41 Observations 827 827 835 835 Countries 85 85 86 86 Notes: ***Significant at 1% level, **5% level, *10% level; p-values are in parentheses; 1 - Dummy=1 if Legal Systems and Property Rights Index is median. Columns (2), (4) are estimates with robust standard errors Sargan test statistic is unavailable for one-step (robust) estimation in Stata. 26

Table 9 Dynamic Model - System GMM Estimates (V), Dependent variable: Remittances Regressors (1) (2) (3) (4) Remittances (-1) 0.846*** 0.822*** 0.863*** 0.834*** (0.000) (0.000) (0.000) (0.000) GDP per capita 0.243*** 0.252* (0.000) (0.098) GDP growth 0.012*** 0.013*** (0.000) (0.002) Domestic Credit (% of GDP) -0.196*** -0.214* -0.047* -0.063 (0.000) (0.077) (0.073) (0.412) Government 1.123*** 1.333*** 1.118*** 1.331*** (0.000) (0.007) (0.000) (0.008) Sound Monetary Policy 0.386** 0.537 0.417** 0.663** (0.031) (0.150) (0.025) (0.041) Credit Market Regulation 0.094 0.110 0.083 0.221 (0.424) (0.612) (0.521) (0.312) Freedom to Trade 0.430** 0.429 0.428*** 0.536* (0.030) (0.268) (0.004) (0.090) Dummy*Monetary Policy 1 0.037 0.024 0.046 0.049 (0.246) (0.716) (0.107) (0.437) Serial correlation test (0.261) (0.272) (0.435) (0.437) Sargan test (0.027) (0.206) Instruments 40 40 40 40 Observations 827 827 835 835 Countries 85 85 86 86 Notes: ***Significant at 1% level, **5% level, *10% level; p-values are in parentheses; 1 - Dummy=1 if Legal Systems and Property Rights Index is median. Columns (2), (4) are estimates with robust standard errors Sargan test statistic is unavailable for one-step (robust) estimation in Stata. 27