Financial Development,Remittances and Economic Growth: Evidence Using a Dynamic Panel Estimation

Abstract

This study examines how remittances can influence economic growth under different levels of financial development. Using a dynamic panel estimation of 33 top remittance-recipient developing countries from 1979 to 2011, the results suggest that financial development neither works as a substitute nor a complement for the remittance–growth nexus. While remittances are effective in promoting economic growth, the influence of financial variables is found to be insignificant. More developed financial systems may attract more remittances; however, the interaction effect of financial development and remittances is not growth enhancing. Promoting financial literacy, reducing the cost of sending remittances through banks and encouraging the overall use of formal financial institutions may induce a stronger remittance–growth nexus.

JEL Classification: F24, F41, F63, F68

Keywords

Financial Development Remittance Economic Growth Generalised Method of Moments Dynamic Panel Estimation

1. INTRODUCTION

Since the last quarter of the twentieth century, there has been an accelerated growth in remittance flows to developing countries. Remittances constitute the second-largest source of external finance after foreign direct investment (FDI) (Aggarwal et al., 2011; Giuliano and Ruiz-Arranz, 2009; Glytsos, 2005). Over the past 40 years, total workers’ remittances to low- and middle-income countries went up from a modest US$0.3 billion in 1975 to more than US$404 billion in 2013 (World Bank, 2014). According to the World Bank (2013), remittance flows are expected to grow at an average of 8.8 per cent annually from 2013 to 2015, with growth in remittances to low-income countries estimated to grow at a faster rate of 12.3 per cent during this same period, as economic conditions strengthen in remittance-sending countries such as the United States. With one-seventh of the world’s population migrating, both internally and across international borders, in search of better economic and social opportunities, it is claimed that this has profound implications for economic development. Not surprisingly, this large increase in both migration and remittances has motivated researchers to explore the potential significance of remittance flows as a tool for economic development. While most studies have analysed different development aspects of this external flow, such as poverty, inequality and the development of human capital, the macroeconomic growth effects of remittances have not been adequately studied until recently.

Remittances can promote economic growth in developing countries through various channels (Loxley and Sackey, 2008; Pradhan et al., 2008; Ziesemer, 2006). First, remittances can finance much-needed consumption expenditures and increase the demand for goods and services in the economy. Many industries benefit from this increase in demand through the so-called multiplier effect, which increases economic growth (Stahl and Habib, 1986; Taylor and Dyer, 2009). Second, remittances can affect growth by increasing productive investments in developing countries. If households treat remittances as transitory income, according to the permanent income hypothesis (Friedman, 1956), the propensity to save out of remittances is much larger than other family income. The financial sector can play a key role in channelling these large savings into productive investments (Levine, 2005) and thereby promoting economic growth in developing countries. Hence, it is important to understand the relationship between remittances and economic growth while considering the development of the financial sector.

The relationship between remittances, financial sector development and economic growth is not immediately clear in the literature. Giuliano and Ruiz-Arranz (2009, p. 144) noted,

... well-functioning financial markets, by lowering costs of conducting transactions, may help direct remittances to projects that yield the highest return and therefore enhance growth rates. On the other hand, remittances might become a substitute for inefficient or nonexistent credit markets by helping local entrepreneurs bypass lack of collateral or high lending costs and start productive activities.

Although these connections have been adequately studied for other sources of external flows, such as FDI or official development assistance (ODA), the existing literature has not examined this relationship for remittances in detail.

This article attempts to fill the void by analysing the link between remittances, financial development and economic growth in three steps. First, we construct a panel dataset of 33 countries over the period 1979–2011. While the time period is constrained by the availability of data, we select countries based on their experiences in major increases in remittance flows over the past decade. Total remittance flow in all these countries over the past three decades was at least 1 per cent of their respective gross domestic productivities (GDPs; Table 1). Moreover, most of these countries have gone through the phase of financial liberalisation in the 1990s, which make this set an ideal choice to gauge the interaction between remittances and financial development. To overcome the limitation of a restrictive modelling framework, we have included a large number of control variables that may have potential influences on countries’ economic growth. In the second step, we conduct unit root tests to investigate the stationarity of the series. The following step involves using the two-step generalised method of moments (GMM) estimator as described in Baum et al. (2003, 2007) to extract consistent and information efficient estimates of the impact of financial development on remittance-growth dynamics. This GMM procedure allows for a dynamic specification of the dependent variable and controls for endogeneity of all the explanatory variables. Better understanding the relationship between remittances and economic growth under different levels of financial development would help policy makers design appropriate policies pertaining to the flow of remittances. Moreover, understanding the channels through which remittances affect economic growth is important in formulating appropriate policy to enhance the growth impact of remittances.

Table 1
List of Countries and Level of Remittances Received

Country Remittances (% of GDP) Country Remittances (% of GDP)

Algeria 1.0 Mali 4.0

Bangladesh 8.0 Mexico 2.0

Benin 3.0 Morocco 8.0

Bolivia 4.0 Niger 2.0

Botswana 1.0 Nigeria 7.0

Burkina Faso 1.0 Pakistan 4.0

Colombia 2.0 Paraguay 3.0

Costa Rica 2.0 Philippines 12.0

Cote d’ Ivore 1.0 Rwanda 1.0

Dominican Republic 8.0 Senegal 9.0

Egypt 5.0 Sri Lanka 8.0

El Salvador 16.0 Swaziland 3.0

Fiji 6.0 Syrian Arab Republic 2.0

Guatemala 10.0 Thailand 1.0

Honduras 16.0 Togo 9.0

India 3.0 Tunisia 4.0

Kenya 3.0

Country	Remittances (% of GDP)	Country	Remittances (% of GDP)
Algeria	1.0	Mali	4.0
Bangladesh	8.0	Mexico	2.0
Benin	3.0	Morocco	8.0
Bolivia	4.0	Niger	2.0
Botswana	1.0	Nigeria	7.0
Burkina Faso	1.0	Pakistan	4.0
Colombia	2.0	Paraguay	3.0
Costa Rica	2.0	Philippines	12.0
Cote d’ Ivore	1.0	Rwanda	1.0
Dominican Republic	8.0	Senegal	9.0
Egypt	5.0	Sri Lanka	8.0
El Salvador	16.0	Swaziland	3.0
Fiji	6.0	Syrian Arab Republic	2.0
Guatemala	10.0	Thailand	1.0
Honduras	16.0	Togo	9.0
India	3.0	Tunisia	4.0
Kenya	3.0

Source: Author’s calculation based on the World Development Indicator’s (WDI) online database.

The rest of the article is organised as follows. The second section presents the literature review. The third section outlines the estimation strategy. The fourth section presents the results and analysis and the fifth section concludes the article.

2. LITERATURE REVIEW

Whether remittances affect economic growth by exploiting the development of the financial sector is not clear in the existing literature. Freund and Spatafors (2008) and Giuliano and Ruiz-Arranz (2009) noted that remittances can positively affect both investment and economic growth if channelled to projects with higher returns in the presence of well-functioning financial markets that tend to reduce transaction costs. On the contrary, if remittances do not ease liquidity constraints in the financial system or are not used for productive investment, the growth impact of remittances through financial channels may be insignificant.

Two recent papers by Demirgüc-Kunt et al. (2011) and Aggarwal et al. (2011) found a positive association between remittances and financial sector development in developing countries. In Aggarwal et al. (2011), the level of financial development measured by bank deposits to GDP and bank credit to GDP increased significantly with the flow of remittances in most countries. Mundaca (2009) used a dataset of 39 Latin American and Caribbean countries over the period 1970–2002 and suggested a complementarity between remittances and financial development in enhancing economic growth. Nyamongo et al. (2012) supported this observation and found complementary effects of remittances and financial development when tested for a set of 36 Sub-Saharan African countries from 1980 to 2009. However, the effect of financial development on economic growth was weak for countries in their study. If the depth of the financial system has any impact on the remittance–growth nexus, accelerated financial development under the financial liberalisation initiative may have had some impact on the remittance–growth relationship.

Using data from 84 countries for the period 1986–2005, Bang et al. (2015) argued that financial reform, which accelerates financial development, should increase the flow of remittances via the investment motive. The authors argued that the relaxation of directed credit programmes, credit ceilings and greater autonomy for the banking sector have positive impacts in attracting remittances, while development of security markets, quality enhancement of banking supervision and the removal of restrictions on interest rate determination had a negative impact on remittances in the long run. Overall, the net impact of financial reforms on remittances is slightly negative in the long run.

A number of economists contested the view that remittances complement investment in financially developed economies. Giuliano and Ruiz-Arranz (2009) used macroeconomic data from 1975 to 2002 for 73 developing countries and employed a GMM technique. After controlling for endogeneity of remittances and financial development, their results suggested that remittances removed credit constraints, improved the allocation of capital and promoted economic growth in less financially developed countries. In addition, remittances were not found to act as a complement to magnify economic growth in financially developed economies. Bettin and Zazzaro (2012) used a panel of 66 developing countries for the period 1991–2005 and showed that the overall effect of remittances on economic growth was unclear.

The orthodox argument of increased domestic resource mobilisation due to the development of the financial sector has been long contested by critics from different schools of thought. The main arguments made by economists from different ‘non-mainstream’ schools are the following:

Informal financial sectors tend to generate more savings than formal sectors in developing countries. Serieux (2008, p. 5) cited the analysis of Taylor (1983), Van Wijnbergen (1983) and Buffie (1984) and noted,

... that the informal financial sector (curb market) was a more efficient conduit of savings. ... while financial liberalisation would lead to an increase in the share of savings intermediated through the formal financial sector, if that increase came at the expense of the informal financial sector, the net effect on output and, presumably, investment, would be negative.

A few studies have recently shown that remittances help the process of development by using the informal sector in different developing countries including Tonga (Brown and Connell, 1993) and Nepal (Seddon, 2004).

Most households in lower and lower-middle-income developing countries survive with a subsistence level of income. While refuting the success of financial reforms, Ogaki et al. (1996) used a simple endogenous growth model and showed that higher interest rates would have little impact on savings and economic growth in low-income developing countries. Any additional income is mostly used for increasing overall consumption. In the context of remittances, a large group of literature suggests that remittances are mostly used for consumption. In a recently published paper, Combes and Ebeke (2011) used a large panel of 87 developing countries over the period of 1975–2004 and showed that remittances reduced the households’ consumption instability. More importantly, the stabilising role of remittances was found to be stronger in less financially developed countries. Nishat and Bilgrami (1991) used data from Pakistan for the period 1959–60 to 1987–88 and applied the three-stage least squares (3SLS) technique within the standard Keynesian framework. Their results suggested that the largest effect of remittances was on private consumption while the smallest was on private investment. More recently, Quartey (2006) showed that remittances were countercyclical and helped minimise economic shocks in Ghana. These findings reinforce the argument that the overall effects of remittances may be larger on consumption than on savings.

To sum up, the interaction effect of remittances and financial development on the economic growth in the existing literature does not produce any unanimous result. Studies included a wide range of countries in their dataset, including countries with a very small percentage of their GDP (less than 0.001 per cent) as remittances to a large percentage of their GDP as remittances (more than 10 per cent). Several studies used very restrictive econometric models to examine the remittance–growth nexus. Many developing countries accelerated the financial development process in their financial sectors during the late 1980s and early 1990s. Using the available data, this study intends to examine whether accelerated financial development has any significant impact on the remittance–growth nexus.

3. DATA AND METHODOLOGY

3.1 Data Issues

Data on time series macroeconomic variables are available from 1979 for most of the top remittance-receiving countries. Our dataset goes from 1979 to 2011.¹ Annual data on real GDP growth (gdpg) has been collected from the ‘World Development Indicators’ (WDI) published by the World Bank (2013). There is no single measure of financial sector development in the existing literature (Wolde-Rufael, 2009). Following Bhattacharya and Sivasubramanian (2003), Al-Yousif (2002), Giuliano and Ruiz-Arranz (2009), Wolde-Rufael (2009), Kar et al. (2011) and Bettin and Zazzaro (2012), we employ the four most commonly used indicators as proxies for financial development. We find that the correlation among all four financial variables is moderately high (Table 2).

These variables are

Ratio of domestic credit to the private sector to GDP, pcredit, measures the extent to which the private sector relies on banks to finance consumption, working capital and investment.

Ratio of total domestic credit provided by the banking sector to GDP to the whole economy (private as well as public), tcredit, measures the strength of financial intermediation in the economy by the banking system.

The degree of monetisation in the economy, m2gdp, measured by the M2 to GDP ratio. This measure includes the currency plus demand and interest-bearing liabilities of banks and non-banks financial intermediaries to GDP. It is designed to show the real size of the financial sector of a growing economy.

Table 2
Correlations Matrix of Financial Indicators

Variables	pcredit	tcredit	m2gdp	m3gdp
pcredit	1.0000	—	—	—
tcredit	0.8816	1.0000	—	—
m2gdp	0.7594	0.7546	1.0000	—
m3gdp	0.6908	0.7063	0.9144	1.0000

Source: Author’s calculation based on the WDI online database.

Liquid liabilities to GDP, m3gdp, measured by the M3 to GDP ratio. M3 is the sum of demand, time, saving and foreign currency deposits. This measures the size of the banking system relative to the economy.

Data on the liquid liabilities to GDP ratio is obtained from the Financial Development and Structure Dataset. Beck et al. (2013) first published this dataset, which was subsequently updated by the World Bank (2013). Data on the three other measures of financial development are collected from the WDI online database (World Bank, 2013).

3.2 Model and Methodology

The basic neoclassical Solow model assumes that the production in an economy is a function of capital, labour and technology. However, due to its restrictive framework, the model fails to explain international differences in income across countries (Mankiw et al., 1992). In addition to capital and labour, empirical evidence supports the importance of foreign aid, human capital, remittances and the financial sector in explaining growth across countries (Hansen and Tarp, 2001; International Monetary Fund, 2005; Mankiw et al., 1992; Pradhan et al., 2008). We use a modified version of the Solow model where GDP growth is assumed to be affected by a set of control variables, including growth rates of gross capital accumulation (inv), the labour force (lfg), output of high-income countries (g_OECD), official development assistance (oda), all other external flows (ocf), remittances (rem), the financial variables (f) defined above (pcredit, tcredit, m2gdp, m3gdp) and the interaction effects of remittances and financial development (remfin).² It has long been argued that the output variable is persistent. In other words, growth in the current period may depend on the past year’s growth (Alesina et al., 1992; Bond et al., 2001). To capture the ‘memory effect’ of the output variable, we include lagged GDP growth (‘gdpg_–1’) as an explanatory variable.

We first consider the following AR(1) model with unobserved country-specific effects:

g d p g_{i, t} = β g d p g_{i, t - 1} + x_{i, t} + c_{i} + v_{i, t}

(1)

where i = 1, 2, ....N and t = 2, ......T

gdpg_i,t = Growth rate of GDP at t period

gdpg_i,t–1 = Growth rate of GDP at t–1 period

x_i_,t = Matrix of control variables

c_i = Time in-variant country specific effects

v_i_,t = Time-variant error component

$ε_{i, t} = c_{i} + v_{i, t}$ has the standard error component structure, which follows strict exogeneity,

E [c_{i}] = 0, E [v_{i, t}] = 0, a n d E [c_{i} v_{i, t}] = 0

The GMM dynamic panel estimator makes two assumptions: (i) transient errors are serially uncorrelated,

E [v_{i, t}, v_{i, s}] = 0 f o r i = 1, 2, ........ N a n d s \neq t

and (ii) the explanatory variables are not correlated with future realisations of the error component.

This GMM procedure allows us to use the lagged levels dated t – 2 and earlier as instruments (Bond et al., 2001). As we are primarily interested in exploring the relationship between remittances and economic growth, we estimate the following behavioural model:

\begin{matrix} \begin{matrix} g d p g_{i, t} = β_{0} + β_{1} g d p g_{i, t - 1} + β_{2} i n v_{i, t} + β_{3} l f g_{i, t} + β_{4} r e m_{i, t} \end{matrix} \\ \begin{matrix} + β_{5} g o e c d_{i, t} + β_{6} o d a_{i, t} + β_{7} o c f_{i, t} + β_{8} f_{i, t} + c_{i} + ε_{i, t} \end{matrix} \end{matrix}

(2)

To examine the significance of remittances on economic growth at different levels of financial development, we employ an interaction term for remittances and financial variables. A negative (positive) and significant interaction term will indicate the substitutability (complementarity) of remittances and financial development in economic growth. The behavioural model can be written as

\begin{matrix} \begin{matrix} g d p g_{i, t} = β_{0} + β_{1} g d p g_{i, t - 1} + β_{2} i n v_{i, t} + β_{3} l f g_{i, t} + β_{4} r e m_{i, t} + β_{5} g o e c d_{i, t} \end{matrix} \\ \begin{matrix} + β_{6} o d a_{i, t} + β_{7} o c f_{i, t} + β_{8} f_{i, t} + β_{9} (r e m f i n_{i, t}) + c_{i} + ε_{i, t} \end{matrix} \end{matrix}

(3)

β₉ measures the interaction effect of remittances and financial development to trigger economic growth.

Time-series data on the labour force is not available for most developing countries. Following Das and Paul (2011), we use the economically active population as a proxy for the labour force. This variable is defined as the size of the population in the age group 15–64 years (Das and Paul, 2011). A measure of human capital, such as the secondary-school enrolment rate would be a better proxy of labour force growth. However, Barro and Lee (2013) data are not available on a yearly basis, which limits our ability to include it in our model.

g_OECD is defined as the rate of GDP growth in high-income OECD countries and China. This variable is included to control for the effect of the relative demand for exports on economic growth. Most of the top remittance-recipient countries export to OECD countries and China. The demand for exports depends on the economic strength of importing countries. The proxy for trade openness such as the ratio of exports plus imports to GDP can be used as a measure; however, this may create simultaneity bias. As such, following Serieux (2008), this study uses the growth in the GDP of OECD countries and China as the proxy variable for openness. While the relevant effect of oda is an increased access to resources other than domestic sources, ocf is expected to affect economic growth by providing credit for public and private investment. Our interest variable rem is expected to provide additional resources for economic growth. Finally, financial sector variables (pcredit, tcredit, m2gdp, m3gdp) are included to control for the potential effect of channelling savings into investment and thus economic growth.

The inclusion of a lagged variable as a regressor makes Equation (1) dynamic. Estimation of the dynamic specification using ordinary least square (OLS) would produce inconsistent estimates of the coefficients (Greene, 2003). To obtain consistent coefficient estimates, we use an instrumental variable technique (i.e., the two-step GMM), which has a number of advantages over the OLS (Das and Paul, 2011). First, this technique uses the pooled cross-section and time series data to estimate the relationship between remittances and growth for a large set of countries over a long time period. Second, an appropriate GMM technique takes country-specific effects into account. Third, this approach controls for any potential endogeneity that may arise from explanatory variables. Finally, this technique outperforms the two-SLS technique, even with robust standard errors.

4. RESULTS AND ANALYSIS

The selection of an appropriate estimation strategy relies on the nature of the data. To determine the level of stationarity, we apply three different panel unit root tests: the Im et al. W-test (2003); Augmented Dickey–Fuller test (Dickey and Fuller, 1981; Said and Dickey, 1984); and the Phillips-Perron test (Phillips and Perron, 1988). The null hypothesis of the three tests is the non-stationarity of the series. The results are presented in Table 3.

Table 3
Stationarity Tests for Relevant Variables

Variables Im, Pesaran and Shin W-test ADF-Fisher Chi-square PP-Fisher Chi-square

gdpg –12.572* 449.381* 449.381***

inv –15.240* 670.137* 670.000***

rem –19.441* 836.364* 775.364*

g_OECD –16.625* 355.377* 355.376*

oda –18.793* 1022.751* 1023.700***

ocf –18.169* 893.029* 893.029***

lfg 1.674 196.04* 182.666*

pcredit –16.498* 748.121* 748.121***

tcredit –15.713* 714.546* 714.546***

m2gdp –17.703* 932.215* 932.215***

m3gdp –13.460* 625.494* 589.871***

Variables	Im, Pesaran and Shin W-test	ADF-Fisher Chi-square	PP-Fisher Chi-square
gdpg	–12.572***	449.381***	449.381***
inv	–15.240***	670.137***	670.000***
rem	–19.441***	836.364***	775.364***
g_OECD	–16.625***	355.377***	355.376***
oda	–18.793***	1022.751***	1023.700***
ocf	–18.169***	893.029***	893.029***
lfg	1.674	196.04***	182.666***
pcredit	–16.498***	748.121***	748.121***
tcredit	–15.713***	714.546***	714.546***
m2gdp	–17.703***	932.215***	932.215***
m3gdp	–13.460***	625.494***	589.871***

Source: Author’s calculation based on the WDI online database.

Note: ***, ** and * indicate significance at the 1, 5 and 10 per cent levels, respectively.

All three test results suggest that all variables but lfg are stationary at the 1 per cent level. lfg is found to be stationary by two of the three tests. Given these results, an approach that does not presume stationarity remains valid.

Table 4 reports different specifications of the growth equations. Specification (i), (ii), (iii) and (iv) are the empirical outcomes of Equation (2). However, the differences in specifications are based on the inclusion of a different financial variable as an indicator.

The study supports the ‘memory effect’, where the GDP growth in the last period affects GDP growth in this period. Further, as anticipated, the growth rate is positively (and significantly) related to the rate of investment. The outcome based on labour force growth is insignificant. Due to data unavailability, as we are using a proxy variable for labour force growth, the outcome based on the regression estimation should be considered cautiously. The impact of economic growth in OECD countries on the growth of remittance-receiving countries is insignificant. In most cases, ODA and other capital flows significantly influence economic growth.

Table 4

Determinants of GDP Growth

Dependent Variable: GDP growth rate (annual %)
Variables	(i)	(ii)	(iii)	(iv)
gdpg(t–1)	0.318***(0.078)	0.339***(0.087)	0.326***(0.090)	0.360***(0.091)
inv	0.335***(0.063)	0.343***(0.069)	0.338***(0.069)	0.311***(0.708)
lfg	1.141(1.039)	1.352(1.063)	1.171(1.041)	0.965(1.113)
g_OECD	0.003(0.002)	0.003(0.002)	0.002(0.002)	0.003(0.002)
rem	0.036**(0.016)	0.033**(0.016)	0.033**(0.015)	0.034***(0.017)
oda	0.013*(0.007)	0.012(0.007)	0.011(0.007)	0.012*(0.007)
ocf	0.004**(0.002)	0.004*(0.002)	0.004**(0.002)	0.005**(0.002)
pcredit	0.067(0.041)	—	—	—
tcredit		0.016(0.002)	—	—
m2gdp	—	—	0.034(0.040)	—
m3gdp	—	—		0.008(0.051)
Hansen J Statistic	5.337 (p value: 0.612)	4.068 (p value: 0.771)	4.750 (p value: 0.576)	3.381 (p value: 0.759)
Observations	742	724	748	652

Source: Author’s calculation based on the WDI online database.

Notes: (1) The endogenous variables are lag GDP growth and investment. Instruments are L2 GDP growth, lag investments, food production index, lag remittances, lag labour force growth and 1 lag of the financial variable.

(2) ***, ** and * indicate significance at the 1, 5 and 10 per cent levels, respectively.

Our main focus is on the link between remittances and GDP growth. All specifications establish a positive and significant relationship between remittances and GDP growth. The top remittance-receiving countries benefit from the flow of remittances from destination countries. The inclusion of financial variables is meant to test the Shaw (1973) proposition that private credits would have a significantly positive impact on the economic growth of the countries. Despite significant financial development, the estimated outcome of pcredit on growth is insignificant. This does not support Shaw’s (1973) contention that the increased provision of private credit through the formal financial sector is growth enhancing (Serieux, 2008). Moreover, the estimated outcome on the growth equation is insignificant for all other financial indicators. Table 5 reports the estimated outcomes based on Equation 3.

The specifications (v), (vi), (vii) and (viii) report the empirical results of Equation (3) with different financial variables. The estimated outcome is insignificant for all interaction terms which indicate that financial development fails to make remittance flows more productive. Remittances neither worked as a substitute nor as a complement under different levels of financial development. It is likely that remittances are not saved in the form of deposits in the financial system but held in cash or used by recipients immediately. The coefficient for the lagged growth is very significant at 1 per cent level in all specifications. These results strongly support the hypothesis of the persistence characteristics of economic growth as suggested by Alesina et al. (1992).

There could be various reasons behind the insignificant impact of financial development on the remittance–growth nexus. Despite accelerated financial development in the late 1980s and the early 1990s, the use of financial institutions by remittance recipients could still be small. According to the Inter-American Development Bank Report (2006), banks in many of the remittance-receiving countries operate only as payment agents. A very small percentage of remittances paid by banks actually enters the financial system through existing or new accounts. Remittances might not spur deposits or growth if access to physical banking outlets is limited in those countries. The distance to the nearest financial outlet could be an obstacle for remittance recipients to demand further credit from the banking system. In many cases, remittance recipients in those countries are less likely to receive remittances via banks. In countries where individuals receiving remittances through banks are more likely to open or maintain bank accounts and use other financial services, banks can play a larger role for credit expansion using remittance money. Moreover, banks can sell other financial products, which may spur economic growth. Due to the lack of business or investment experience among remittance-recipient households, remittances may not act as a source of capital for economic development. Moreover, there are obstacles to transferring remittances into a significant source of capital. The outcome is not surprising as many remittance-recipient households do not have sufficient knowledge about business opportunities, and therefore the use of remittance money for business is limited.

Table 5

Effects of Remittances and Financial Development on GDP Growth

Dependent Variable: GDP growth rate (annual %)
Variables	(v)	(vi)	(vii)	(viii)
gdpg(t–1)	0.315***(0.086)	0.339***(0.087)	0.318***(0.088)	0.351***(0.089)
inv	0.340***(0.067)	0.347***(0.069)	0.343***(0.067)	0.312***(0.066)
lfg	1.134(1.048)	1.370(1.067)	1.267(1.063)	1.019(1.122)
g_OECD	0.002(0.002)	0.003(0.002)	0.002(0.002)	0.003(0.002)
rem	0.034**(0.017)	0.034**(0.002)	0.030**(0.013)	0.033**(0.017)
oda	0.013*(0.007)	0.011(0.007)	0.011(0.007)	0.011*(0.007)
ocf	0.004**(0.002)	0.003*(0.002)	0.004*(0.055)	0.005**(0.002)
pcredit	0.066(0.041)	—	—	—
tcredit	—	0.019(0.032)		—
m2gdp	—	—	0.023(0.004)	—
m3gdp	—	—	—	0.008(0.051)
remfin	–0.041(0.119)	0.029(0.079)	–0.128(0.113)	–0.057(0.084)
Hansen J statistic	5.332 (p value: 0.502)	4.028 (p value: 0.672)	4.778 (p value: 0.572)	3.426 (p value: 0.752)
Observations	745	727	748	652

Source: Author’s calculation based on the WDI online database.

Notes: (1) Endogenous variables: lag GDP growth, investment. Instruments: L2 GDP growth, lag investments, food production index, lag remittances, lag labour force growth, 1 lag of financial variable.

(2) ***, ** and * indicate significance at the 1, 5 and 10 per cent levels, respectively.

The high cost of maintaining a bank account could be another reason for remittances not being used for the expansion of credit in remittance-recipient economies. Beck et al. (2006) observe that in many countries in South Asia, Sub-Saharan Africa and Latin America, maintaining a bank account and the fees associated with loans are comparatively higher than in developed countries. The sluggish growth of credit in recent years may be attributed to weaker creditor protection and poor contract enforcement (Peria et al., 2008).

5. CONCLUSION

This article re-examines the relationship between remittances, financial development and economic growth for the major remittance-recipient developing countries. We went beyond the direct effects of remittances on growth by estimating the interactive effects of remittances and financial variables. To do so, the study introduces several interaction terms between remittances and financial indicators and tests their relationship with GDP growth. Using growth equations, this study confirms the positive association between remittances and economic growth for the top remittance-recipient developing countries. The inclusion of financial variables in the growth equations asserts that none of the financial development indicators have any impact on the remittance–growth nexus.

Through financial development, which could be an outcome of several reforms initiative, developing countries intended to deepen their financial sector relative to output to increase private and overall saving and investment and improve the quality of investment. Such an increase in the volume of formal finance is expected to have a direct positive effect on output growth. Upon considering the interaction terms of financial variables and remittances, we cannot claim any impact of financial development on the remittance–growth nexus. This article is expected to enrich the existing literature in several ways. First, among the scholars who have examined the remittance–growth nexus, none has examined such a relationship among the top remittance-recipient developing countries. However, as remittances are a significant source of external financing for that group of countries, this study provides an important understanding about the remittance–growth nexus during the process of financial development. This study suggests that policy makers in top remittance-recipient developing countries should formulate policies for financial development such that these countries can reap the potential benefits of remittances on economic growth. As remittances can be used to relax individuals’ financing constraints, this may lead to a lower demand for credit and have a dampening effect on credit market development. Moreover, a rise in remittances failed to translate into an increase in credit to the private sector due to channelling into financing the government. Remittances may not have increased bank deposits if they are immediately spent on imported consumption goods or if remittance recipients distrust financial institutions and prefer other ways of saving these flows (Peria et al., 2008).

This study could be improved in various ways. The quality and coverage of data could be an issue. For many countries, due to the weakness of data collection, many types of formal remittance data remain unrecorded. Moreover, the flow of remittances through informal channels, such as unregulated money transfers by firms or families, remains unaccounted for. The World Bank (2006) reports that if remittances through informal channels are included in the data, total remittances could be as much as 50 per cent higher than the official record. Having better data could enrich our results. Understanding the distinctive nature of different countries/regions/income groups and the ways in which remittances are used and channelled might also shed more light on the nature of the relationship between remittances and financial indicators.

Footnotes

Acknowledgements

The author is indebted to John Loxley, John Serieux, Anupam Das and anonymous referees for their valuable comments and suggestions. This article was presented at the 49th Canadian Economics Association conference in Toronto, 2015 and at the Immigration Matters and Human Resource Management conference at Harvard University in 2015.

Notes

References

Aggarwal

, Demirgüc-Kunt

, & Peria

M.S.M.

(2011). Do remittances promote financial development? Journal of Development Economics, 96(2), 255–64.

Alesina

, Ozler

, Roubini

, & Swagel

(1992). Political instability and economic growth. NBER Working Paper No. 4173. Retrieved 14 April 2015, from http://www.nber.org/papers/w4173.pdf

Al-Yousif

Y.K.

(2002). Financial development and economic growth: Another look at the evidence from developing countries. Review of Financial Economics, 11(2), 131–50.

Bang

J.T.

, Mitra

, & Wunnava

P.V.

(2015). Financial liberalization and remittances: Recent panel evidence. The Journal of International Trade & Economic Development: An International Comparative Review, 24(8), 1077–102. doi: 10.1080/09638199.2014.1001772

Barro

R.J.

& Lee

(2013). A new data set of educational attainment in the world, 1950–2010. Journal of Development Economics, 104(September), 184–98.

Baum

C.F.

, Schaffer

M.E.

, & Stillman

(2003). Instrumental variables and GMM: Estimation and testing. The Stata Journal, 3(1), 1–31.

Baum

C.F.

, Schaffer

M.E.

, & Stillman

(2007). Enhanced routines for instrumental variables/GMM estimation and testing. Boston College Department of Economics Working Paper No. 667. Retrieved 14 April 2015, from http://fmwww.bc.edu/ec-p/wp667.pdf

Beck

, Demirgüc-Kunt

, & Peria

M.S.M.

(2006). Banking services for everyone? Barriers to bank access around the world. The World Bank Economic Review, 22(3), 397–430.

Beck

, Demirgüc-Kunt

, Levine

R.E.

, Cihak

, & Feyen

H.E.B.

(2013). Financial development and structure dataset. Retrieved 14 April 2014, from http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:20696167~pagePK:64214825~piPK:64214943~theSitePK:469382,00.html

10.

Bettin

, & Zazzaro

(2012). Remittances and financial development: Substitutes or complements in economic growth. Bulletin of Economic Research, 64(4), 509–36.

11.

Bhattacharya

, & Sivasubramanian

M.N.

(2003). Financial development and economic growth in India: 1970–1971 to 1998–1999. Applied Financial Economics, 13(12), 925–29.

12.

Bond

, Hoeffler

, & Temple

(2001). GMM estimation of empirical growth model. CEPR Discussion Paper No. 3048. Retrieved 14 April 2015, from http://www.nuffield.ox.ac.uk/users/bond/CEPR-DP3048.PDF

13.

Brown

R.P.C.

, & Connell

(1993). Migration and remittances in Tonga and Western Samoa. Report prepared for the ILO. Brisbane: The University of Queensland.

14.

Buffie

E.F.

(1984). Financial repression, the new structuralist, and stabilization policies in semi-industrialized economies. Journal of Development Economics, 14(3), 305–22.

15.

Combes

, & Ebeke

(2011). Remittances and household consumption instability in developing countries. World Development, 39(7), 1076–89.

16.

Das

, & Paul

(2011). Openness and growth in emerging Asian economies: Evidence from GMM estimations of a dynamic panel. Economics Bulletin, 31(3), 2219–28.

17.

Demirgüc-Kunt

, Cördova

E.L.

, Peria

M.S.M.

, & Woodruff

(2011). Remittances and banking sector breadth and depth: Evidence from Mexico. Journal of Development Economics, 95(2), 229–41.

18.

Dickey

D.A.

, & Fuller

W.A.

(1981). Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica, 49(4), 1057–72.

19.

Freund

, & Spatafora

(2008). Remittances, transaction costs, and informality. Journal of Development Economics, 86(2), 356–66.

20.

Friedman

(1956). The quantity theory of money—a restatement. In Friedman

(Ed.), Studies in the quantity theory of money (pp. 129–38). Chicago: University of Chicago Press.

21.

Giuliano

, & Ruiz-Arranz

(2009). Remittances, financial development, and growth. Journal of Development Economics, 90(1), 144–52.

22.

Glytsos

N.P.

(2005). The contribution of remittances to growth: A dynamic approach and empirical analysis. The Journal of Economic Studies, 32(6), 468–96.

23.

Greene

W.H.

(2003). Econometric analysis (5th ed.). Upper Saddle River, NJ: Prentice Hall.

24.

Hansen

, & Tarp

(2001). Aid and growth regressions. Journal of Development Economics, 64(2), 547–70.

25.

K.S.

, Pesaran

M.H.

, & Shin

(2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115(1), 53–74.

26.

Inter-American Development Bank (2006). Remittances 2005: Promoting financial democracy. Retrieved 20 April 2014, from http://www.iadb.org/am/2006/doc/StatisticalComparisons.pdf

27.

International Monetary Fund (2005). Two current issues facing developing countries. In World economic outlook, April 2005: Globalization and external imbalances. Washington, DC: World Economic and Financial Surveys.

28.

Kar

, Nazlioğlu

, & Ağir

(2011). Financial development and economic growth nexus in the MENA countries: Bootstrap panel granger causality analysis. Economic Modelling, 28(1–2), 685–93.

29.

Levine

(2005). Finance and growth: Theory and evidence. In Aghion

& Durlauf

(Eds), Handbook of economic growth (pp. 866–34). Amsterdam: North-Holland Elsevier Publishers.

30.

Loxley

, & Sackey

(2008). Aid effectiveness in Africa. African Development Review, 20(2), 163–99.

31.

Mankiw

N.G.

, Romer

, & Weil

D.N.

(1992). A contribution to the empirics of economic growth. The Quarterly Journal of Economics, 107(2), 407–37.

32.

Mundaca

(2009). Remittances, financial markets development and economic growth: The case of Latin America and Caribbean. Review of Development Economics, 13(2), 288–303.

33.

Nishat

, & Bilgrami

(1991). The impact of migrant workers’ remittances on the Pakistan economy. Pakistan Economic and Social Review, 29(1), 21–41.

34.

Nyamongo

E.M.

, Misati

R.N.

, Kipyegon

, & Ndirangu

(2012). Remittances, financial development and economic growth in Africa. Journal of Economics and Business, 64(3), 240–60.

35.

Ogaki

, Ostry

, & Reinhart

C.M.

(1996). Saving behavior in low- and middle income developing countries: A comparison. IMF Staff Papers, 43(1), 38–71.

36.

Peria

M.S.M.

, Mascaró

, & Moizeszowicz

(2008). Do remittances affect recipient countries’ financial development? In Fajnzylber

& Lopez

J.H.

(Eds), Remittances and development: Lessons from Latin America (pp. 171–215). Washington, DC: World Bank.

37.

Phillips

P.C.B.

, & Perron

(1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335–46.

38.

Pradhan

, Upadhyay

, & Upadhyaya

(2008). Remittances and economic growth in developing countries. The European Journal of Development Research, 20(3), 497–506.

39.

Quartey

(2006). The impact of migrant remittances on household welfare in Ghana. AERC Research Paper 158. African Economic Research Consortium, Nairobi, Kenya.

40.

Said

E.S.

, & Dickey

D.A.

, (1984). Testing for unit roots in autoregressive-moving average models for unknown order. Biometrika, 71(3), 599–607.

41.

Seddon

(2004). South Asian remittances: Implications for development. Contemporary South Asia, 13(4), 403–20.

42.

Serieux

(2008). Financial liberalization and domestic resource mobilization in Africa: An assessment. International Poverty Centre Working Paper No. 45. International Poverty Centre, Brasillia, Brazil.

43.

Shaw

E.S.

(1973). Financial deepening in economic development. New York, NY: Oxford University Press.

44.

Stahl

C.W.

, & Habib

(1989). The impact of overseas workers’ remittances on indigenous industries: Evidence from Bangladesh. The Developing Economies, 27(3), 269–85.

45.

Taylor

J.E.

, & Dyer

G.A.

(2009). Migration and the sending economy: A disaggregated rural economy-wide analysis. Journal of Development Studies, 45(6), 966–89.

46.

Taylor

(1983). Structuralist macroeconomics: Applicable models for the Third World. New York, NY: Basic Books.

47.

Van Wijnbergen

(1983). Interest rate management in LDCs. Journal of Monetary Economics, 12(3), 433–52.

48.

Wolde-Rufael

(2009). Re-examining the financial development and economic growth nexus in Kenya. Economic Modelling, 26(6), 1140–46.

49.

World Bank (2006). Global economic prospects: Economic implications of remittances and migration. Washington, DC: World Bank.

50.

World Bank (2013). World development indicators. Washington, DC: World Bank.

51.

World Bank (2014). Press release. Retrieved 25 May 2015, from http://www.worldbank.org/en/news/press-release/2014/04/11/remittances-developing-countries-deportations-migrant-workers-wb

52.

Ziesemer

(2006). Worker remittances and growth: The physical and human capital channel. UNU-Merit. Working Paper Series No. 020. United Nations University, Maastricht, The Netherlands.

Financial Development,Remittances and Economic Growth: Evidence Using a Dynamic Panel Estimation

Abstract

Keywords

1. INTRODUCTION

3. DATA AND METHODOLOGY

3.1 Data Issues

Table 2 Correlations Matrix of Financial Indicators

Footnotes

Acknowledgements

Notes

References

Table 2
Correlations Matrix of Financial Indicators