INTRODUCTION
Problem statement
In recent decades, migration flows have significantly increased, leading to a rapid rise in remittances—funds sent by migrants to their home countries—which now exceed traditional financial channels such as official development assistance (ODA) and private capital (Dilshad, 2013) According to World Bank data, personal remittances reached one-third of foreign direct investment (FDI) in 2014 and were more than three times the combined total of ODA and official aid, growing from US$ 67 billion in 1990 to over US$ 500 billion.
Between 2000 and 2014, foreign direct investment (FDI) experienced significant fluctuations, while official development assistance (ODA) and other external funding sources either remained stable or decreased Notably, remittances showed a steady increase over time, despite a slight decline of nearly 5 percent in 2009 due to the global financial recession, marking the first drop since 2000 This decrease was minor compared to the 45 percent drop in FDI during the same year Following this, remittances fully recovered and continued to rise, whereas FDI faced further declines of nearly 10 percent in 2012 and 20 percent in 2014.
In recent decades, the stability and increase of remittance inflows have garnered significant attention from researchers and policymakers, leading to numerous studies exploring their relationship with various aspects of economic development While many investigations focus on the effects of remittances on growth, poverty, and education, there has been less emphasis on the connection between inward remittance flows and financial development This relationship is crucial, as financial development is well-documented to promote economic growth and alleviate poverty, according to extensive literature.
Kunt and Levine, 2004), across recipient countries despite their important contribution in total external financial sources
Remittance flows can significantly influence financial development, although empirical studies show varying results across different countries When recipients utilize formal financial channels for remittances, it can enhance financial services, allowing banks to reach unbanked individuals Additionally, the demand for banking products may grow as recipients seek to manage surplus income Remittances can also serve as collateral, increasing banks' willingness to issue loans to families with stable remittance income Even if remittances do not provide sufficient collateral, the influx of funds can boost overall credit availability in recipient communities.
Remittances can hinder the growth of financial sectors by easing recipients' budget constraints, leading to reduced demand for external credit (Martínez, Mascaró and Moizeszowicz, 2008; Brown et al., 2011) Many recipients view remittances as extra income, primarily using them for consumption rather than savings, while a lack of trust in financial institutions may prevent them from depositing money in banks, resulting in minimal demand for financial products (Chami et al., 2009) Consequently, the relationship between remittances and financial development in recipient countries remains complex, highlighting the need for further research to explore these dynamics.
In an effort to address the relationship between remittances and the development of financial sector in Asia, this study utilized a panel data from
From 1990 to 2014, a study analyzed personal remittances and financial development indicators, including domestic credit to the private sector and broad money, across thirty-seven Asian countries using various econometric methods such as fixed effects, random effects, and the system Generalized Method of Moments Additionally, the research investigates the varying impacts of remittance inflows on high, middle, and low-income countries, as classified by the World Bank.
Research objectives
This study is conducted in an attempt to:
Analyze the trend of remittance inflows and the situations of domestic credit to private sector provided by banks and broad money in Asia region,
Access the impact of remittances on financial development in Asia in general,
Evaluate different impacts of remittance flows on financial development in different income-groups of countries in Asia
From the findings, this study will propose recommendations in order to foster the effects of remittance inflows on financial development in Asian remittance recipient countries.
Scope and data of the study
This research examines the significant impact of remittances on financial development in Asian countries Over the past decade, remittances to Asia have surged, accounting for nearly 50% of global inflows Many middle and low-income nations have seen substantial contributions in both the numerical value and percentage of their Gross Domestic Product (GDP) from these remittances However, the effects of remittances on the financial sectors in this region have not been thoroughly analyzed.
Research on the impact of remittances on financial development has primarily focused on individual countries or specific regions in Asia For instance, Chowdhury (2011) identifies a direct positive correlation between remittance flows and the financial breadth and depth in Bangladesh Meanwhile, Noman and Uddin (2012) present evidence of an indirect effect of remittances on the banking sector and economic growth in selected South Asian countries.
Previous studies have categorized samples based on the distinction between developing and developed countries, which may skew results due to varying income tolerances within these groups Therefore, it is essential to explore the relationship between remittances and financial development in this context Specifically, analyzing the effects of remittance flows across different income brackets will offer valuable insights for policymakers.
Remittances significantly contribute to financial development in Vietnam, a middle-income country in Asia, with inflows increasing from US$1.34 million in 2000 to over US$13 million by 2015, representing nearly 7% of its GDP.
This study investigates the impact of remittances on various income groups in Asia, utilizing data from selected countries collected between 1990 and 2014 The analysis includes the financial openness indicator represented by the Chinn-Ito Index, while other data sources are drawn from the World Bank database.
Structure of the study
This thesis is structured into five chapters, beginning with an exploration of the relationship between remittances and financial development in Asian countries It continues with a review of key literature on the impact of remittances on financial sectors and empirical studies on their effects alongside other variables The methodology and data used in the research are detailed in Chapter 3 Chapter 4 analyzes the trends in remittances and financial development in Asia from 1990 to 2014, presenting empirical results from regression analysis The final chapter summarizes the main findings, offers policy recommendations, and discusses the limitations of the study and suggestions for future research.
LITERATURE REVIEWS
Theory of remittances and financial development
Remittances refer to the cross-border money transfers that migrants send back to their home countries, yet the term is often used without a clear definition These transfers can occur through official channels, such as banking systems and money transfer organizations, or unofficial channels, primarily involving cash sent through friends, family, or traditional methods like hawala, where funds are deposited with unlicensed entities in one country and withdrawn by partners in the recipient country.
Informal channels for transferring remittances are widely used worldwide due to their accessibility, anonymity, low cost, and reliability These channels allow both senders and receivers to bypass the need for bank accounts and complex procedures associated with official institutions Additionally, transaction details can remain confidential, providing a sense of privacy The reliance on family and friends for these transfers enhances trust, particularly for individuals who are not familiar with formal financial organizations.
Using informal channels for remittances can lead to several negative consequences Official organizations may struggle to accurately collect data and could miscalculate the total volume of remittances Additionally, these channels may facilitate money laundering by criminal organizations or individuals Ultimately, this undermines the intended financial development benefits of remittances, as highlighted by proponents of financial growth (Nyamongo, 2012).
The financial sector, as defined by the World Bank, encompasses institutions, instruments, markets, and the legal and regulatory framework that facilitate credit transactions Developing this sector involves minimizing costs associated with information acquisition, contract enforcement, and transaction execution This cost reduction fosters the emergence of financial contracts, markets, and intermediaries Variations in information, enforcement, and transaction costs, along with differing legal, regulatory, and tax systems, have led to the creation of unique financial contracts, markets, and intermediaries across different countries and historical contexts.
Financial development occurs through enhancements in financial instruments, markets, and intermediaries, which may or may not alleviate issues related to information, enforcement, and transaction costs (Levine, 2005).
2.1.3 The role of remittances in financial development
Previous studies have explored the various impacts of remittances on financial development in recipient countries, highlighting both demand and supply dynamics within the financial sector On the demand side, remittances can enhance financial literacy among recipients, especially when they utilize formal transfer channels This increased awareness drives the demand for banking products and services, as recipients seek information about available options Consequently, banks have the opportunity to reach unbanked individuals, expanding their outreach Additionally, even when remittances are not processed through banks, the surplus income generated may encourage recipients to seek banking services to store or invest their excess funds.
Remittances play a crucial role in enhancing the financial sectors of recipient economies by encouraging banks to expand their offerings and establish more branches to meet the needs of remittance recipients As these recipients become more attractive customers, banks may increase credit availability for those with stable remittance inflows Additionally, the influx of funds from remittance recipients can boost the loanable funds in banks, leading to increased credit availability within communities This dynamic contributes to the overall growth and depth of financial systems, making it vital to explore the impact of remittances on financial development in various countries.
Remittances can negatively impact financial development by easing budget constraints for recipients, leading to reduced demand for credit from financial institutions (Caceres and Saca, 2006; Martínez, Mascaró and Moizeszowicz, 2008; Aggarwal et al., 2011) This situation is exacerbated when recipients engage in conspicuous consumption and fail to cultivate saving habits, which are essential for fostering investments and driving economic growth.
Remittance inflows may not significantly stimulate financial development in countries with underdeveloped financial systems, as many migrants prefer informal transfer methods over formal banking channels (Brown et al., 2011) This preference leads to a lack of awareness and utilization of banking services among recipients, resulting in low financial literacy within communities Consequently, the demand for formal financial systems remains stagnant Additionally, remittance recipients often choose to keep excess funds at home or use them for consumption due to distrust in banks, limiting their engagement with savings products offered by financial institutions As a result, these behaviors contribute to a neutral impact on the performance of the financial sector.
Debates surrounding the impact of remittances on financial development also highlight adverse causality linked to varying levels of financial sector performance (Brown et al., 2011) A lack of a developed financial system and limited access to banks and formal financial institutions in remittance-receiving countries can lead to distrust among unbanked individuals, pushing them towards informal transaction channels Conversely, a higher level of financial development can enhance trust in financial sectors, increasing the demand for financial products and services Additionally, reduced costs associated with transferring remittances due to financial development may encourage both senders and receivers to utilize formal financial services.
Empirical studies
In recent decades, various methods and datasets have been employed to assess the impact of steadily increasing remittance inflows on global financial development Research on this relationship can be divided into two main approaches: indirect and direct The indirect approach examines how remittances contribute to economic growth at specific levels of financial development, while the direct approach evaluates the effects of remittances on financial sector performance, focusing on their role in enhancing and expanding financial development.
Studies on the relationship between remittances and economic growth reveal contrasting effects based on a country's level of financial development In nations with underdeveloped financial systems, remittances significantly contribute to capital markets, enhancing investment and fostering economic growth For example, incorporating a financial development variable alongside remittances in growth equations has been analyzed in a sample of over 100 countries, highlighting the positive impact of remittances in less developed financial contexts.
Between 1975 and 2002, Giuliano and Ruiz-Arranz (2009) demonstrate that remittance inflows significantly enhance economic growth by serving as an alternative financial source for investment and alleviating credit constraints in countries with underdeveloped financial systems Their findings remain consistent across various measures of financial system development, even after addressing endogeneity issues related to the causal link between remittances and financial growth The study also indicates that remittances can stimulate economic growth through an investment channel, particularly when the financial sector's credit offerings do not meet public demand However, the impact of remittances as an external investment source diminishes in countries with well-developed financial sectors, where the availability of credit reduces the reliance on remittances for investment purposes.
Ramirez and Sharma (2008) found similar results to Giuliano and Ruiz-Arranz (2009), suggesting that remittances can enhance economic growth more significantly in countries with weak financial development Their analysis, which utilized unit root tests, cointegration tests, and fully modified ordinary least squares methods, examined data from twenty-three upper and lower-income countries in Latin America and the Caribbean from 1990 to 2005 Additionally, the study revealed that the impact of remittances on economic growth is more pronounced in upper-income countries compared to lower-income ones Furthermore, it highlighted other channels through which remittances can positively affect economic growth, including education levels and economic liberalization.
The impact of remittances on economic growth is more pronounced in countries with developed financial systems, as access to financial products facilitates effective usage and investment of these funds Mundaca (2009) demonstrates this through a cross-country analysis of Latin America and the Caribbean from 1970 to 2002, utilizing First Difference GMM to address potential endogeneity The study concludes that remittances enhance growth in countries with robust financial systems capable of directing inflows towards technology and capital investments Additionally, it finds that incorporating remittance proxies into growth equations significantly increases the effect of per capita investment on economic growth Supporting findings from Ojeda (2003), Terry and Wilson (2005), and the World Bank (2006) further emphasize that channeling remittances into profitable investments via the formal financial sector amplifies their development impact.
Bettin and Zazzaro (2011) found similar outcomes in their study on the connection between remittance inflows and financial development levels, using a panel dataset that encompassed sixty-six developing countries over a specified period.
From 1970 to 2005, a study utilizing the system GMM method revealed that remittance inflows significantly enhance economic growth in countries with efficient banking systems This research highlights the importance of a well-functioning financial system in maximizing the benefits of remittances for economic development in remittance-receiving nations.
Recent studies have highlighted the significant impact of remittance inflows on financial development in recipient economies While initial research focused on the overall financial development level, the direct influence of remittances on the growth of financial sectors has gained attention As remittances continue to play an increasingly vital role, their effects on economic development are becoming more pronounced.
Recent empirical studies have shifted focus to the direct impact of remittances on financial development in recipient countries, yielding mixed results Notably, many recent papers indicate that remittances foster financial development across various nations For example, Gupta et al (2009) analyzed the relationship between remittances and financial development, as well as their effect on poverty, using panel data from 44 Sub-Saharan African countries between 1975 and 2004 Their findings, derived from advanced regression models, reveal that while remittances in this region are modest compared to other aid flows, their consistent and gradually increasing volume significantly enhances financial development and contributes to poverty alleviation Importantly, these results hold true even after addressing potential reverse causality in the relationship between remittances, financial development, and poverty.
Demirguc-Kunt et al (2010) provide evidence of a significant positive correlation between remittance inflows and the expansion of the banking sector in Mexico, based on county-level data from 2000 The study reveals that as the percentage of the population receiving remittances increases, there is a corresponding rise in the number of bank branches, bank accounts, and overall bank deposits Specifically, for every one percentage point increase in households receiving remittances, the utilization of financial products and services among households grows by approximately 0.16 to 0.19 percentage points, even after accounting for potential endogeneity.
Aggarwal et al (2011) investigate the relationship between remittance inflows and financial system development in 109 developing countries from 1975 to 2007, employing various research methods such as fixed effects estimations, dynamic system GMM, and instrumental variables (IV) estimations Their study reveals that worker remittances significantly and positively impact the aggregate bank credit and deposits to the private sector, as measured by their proportion of GDP, while accounting for variable omission, reverse causality, and measurement errors.
Cooray (2012) provides evidence of the relationship between remittances and the financial sector by analyzing an annual dataset of 94 non-OECD countries from 1990 to 2010 using pooled OLS and system GMM The study reveals that remittance inflows positively impact both the size and efficiency of financial systems Additionally, it explores how government ownership of banks in remittance-receiving countries influences these effects, finding that lower government ownership correlates with a greater expansion of the financial sector due to remittances Conversely, increased efficiency in financial sectors is associated with higher levels of government ownership in banks.
Chowdhury (2011) explores the impact of remittance flows on the financial sector's development in Bangladesh, utilizing annual data from 1971 to 2008 and employing Cointegration and Vector Error Correction Model methods The findings reveal a significant positive relationship between remittance inflows and the expansion and enhancement of the country's financial system Additionally, the study confirms that there is no reverse causation between remittance inflows and financial development indicators, addressing potential endogeneity bias.
While many studies support the positive link between remittances and financial development, some research indicates otherwise Brown et al (2011) found that remittances can have no effect or even hinder financial sector growth Their macro-level analysis, which included data from 138 countries between 1970 and 2005, utilized fixed effects and Probit models to assess the relationship Conversely, at the micro level, data from 3,899 households in Azerbaijan and 3,995 in Kyrgyzstan revealed varied impacts In Azerbaijan, remittances negatively affected financial development, while in Kyrgyzstan, they had a positive influence on both households and communities.
Other determinants of financial development
In addition to remittances, previous studies have highlighted several macroeconomic and openness variables that influence financial development, including country size, GDP per capita, inflation, financial openness, and trade openness Research by Goldsmith (1969) and Gurley and Shaw (1967) indicates that economic growth increases the demand for financial products and services, prompting the financial sector to adapt to these evolving requirements This rise in demand encourages the emergence of more sophisticated financial intermediaries to cater to the new needs for their offerings (Yartey, 2008).
Economic development and the quality of a country's legal institutions are closely linked, as higher income can lead to increased savings rates, fostering the growth of financial instruments that facilitate better investments (Kamar and Ben Naceur, 2007; Yartey, 2008) Conversely, poor institutional quality significantly hinders economic development, as it can create instability in savings and investments Research by La Porta et al (1997) highlights that differences in the protection of investor and creditor rights, as well as the enforcement of laws and regulations, can explain why some countries have more advanced financial systems than others.
High inflation negatively impacts financial properties, prompting investors to seek opportunities in other sectors rather than depositing money in financial institutions This trend can hinder the growth of the financial sector Recent studies, including those by Boyd et al (2001) and Naceur and Ghazouani, have demonstrated the adverse relationship between inflation and the performance of the financial sector.
Regarding financial openness, according to the findings of Chinn and Ito
(2002) and Baltagi et al (2009), the capital account openness enhances financial sector accountability and transparency, hence augmenting the access and utilization of financial products and services
Countries with higher trade openness are more likely to develop advanced financial systems that facilitate trade transactions Research by Rajan and Zingales (2003) suggests that in nations with low trade openness, established industries may obstruct the growth of the financial sector.
Despite previous studies highlighting the positive impact of remittances on financial development, inconsistent results arise from varying datasets and methodologies Additionally, while factors such as country size, GDP per capita, financial openness, and trade openness generally promote financial development, and inflation tends to hinder it, the specific effects of these determinants in Asia and across different income groups remain unclear Therefore, it is essential to separately analyze the influence of remittances and other factors on the financial sector's development in selected Asian countries, considering both general trends and income group distinctions.
MODEL SPECIFICATION AND DATA
Model specification
When analyzing the relationship between remittances and financial development, potential endogeneity due to measurement error, variable omissions and reverse causality should be taken in to account (Aggarwal et al.,
Remittances often face measurement errors due to informal transfer methods, such as those through friends, relatives, and Hawala-type organizations Understanding the causal relationship between remittances and financial development is crucial, as a higher level of financial development can lead to increased recorded remittances This can occur because financial development encourages remittance inflows or because a greater share of these funds is channeled through banks and formal financial institutions Additionally, improved financial development may lower the costs associated with transferring remittances, further boosting inflows However, neglecting key factors that influence remittance growth or financial sector performance can introduce biases in estimating the impact of remittances on financial development.
This study aims to explore the influence of remittances on financial development by utilizing established research to identify a suitable model and relevant variables Building on the methodologies of Aggarwal et al (2006) and Gupta et al (2009), a dynamic panel data model is employed to effectively tackle the identified issues The general structure of the dynamic panel data model is outlined below.
The equation Y it = α + γY i,t-1 + β 1 X it + n i + ε t + u it (3.1) models financial development indicators (Y it) for individuals indexed from 1 to n over time periods from 1 to T In this model, Y it-1 represents the lagged values of these financial indicators, while X it includes a range of explanatory variables such as remittances and other factors influencing financial development Additionally, n i accounts for unobserved country-specific effects, ε t captures time-specific effects, and u it denotes the error term.
Building on previous research related to financial development, remittance indicators serve as a primary explanatory variable; however, the model also incorporates additional factors that have been shown to influence financial development.
FDi,t = β1FDi,t-1 + β2REMITi,t + β3GDPPCi,t + β4LNGDPi,t + β5INFi,t + β6TRADEOPENNESSi,t + β7FINANCIALOPPENNESSi,t + β0 + ui,t (3.2) where i represents for country 1,2,…,37 and t represents for year from 1990 to 2014;
Financial development (FD) is assessed through two primary indicators: domestic credit to the private sector by banks as a share of GDP (CREDIT) and the ratio of broad money to GDP (M2) Domestic credit to the private sector, measured as a share of GDP, reflects the banking sector's efficiency in resource allocation and is crucial for evaluating financial development (Beck et al., 2000; Aggarwal et al., 2006) A higher ratio indicates increased funds allocated to the private sector, leading to greater domestic investment and, consequently, enhanced financial sector development.
Beck et al (2000) used the ratio of broad money to GDP to assess the size of the financial system, while Aggarwal et al (2006) argued that the growth of private financial assets, represented by broad money (M2 or M3), reflects the liquidity position of the financial system and indicates the level of monetization and financial market development Additionally, broad money as a percentage of GDP (M2/GDP) has been identified as an indicator of financial deepening (Brown et al., 2013).
REMIT, defined as personal remittances as a percentage of GDP, includes personal transfers and compensation of employees according to the IMF's BPM6 guidelines Personal remittances encompass a broader category than traditional worker's remittances, as they are not dependent on the migrants' earnings or their relationship with recipients Compensation of employees consists of wages and salaries, both in cash and in kind, along with social contributions from employers, covering border, seasonal, and short-term workers employed outside their home country, as well as residents working for nonresident entities (World Bank, 2016).
This study incorporates macroeconomic and openness variables known to impact financial development Economic development is represented by GDP per capita in thousands of constant 2010 US dollars, while country size is measured using the logarithm of GDP in constant 2010 US dollars Inflation is indicated by the annual growth rate of the GDP implicit deflator, reflecting overall price changes in the economy Trade openness is quantified by the sum of exports and imports of goods and services as a share of gross domestic product, utilizing the KAOPEN index developed by Chinn and Ito (2006) This index is based on four binary dummy variables that capture restrictions on cross-border financial transactions, as detailed in the IMF's Annual Report on Exchange Arrangements and Exchange Restrictions.
the existence of multiple exchange rates;
restrictions on current account transactions;
restrictions on capital account transactions; and
the requirement of the surrender of export proceeds
Chinn and Ito invert binary variables to assign a value of one when capital account restrictions are absent, subsequently deriving the first principal component, known as KAOPEN, which serves as their summary measure.
Data sources
This study investigates the influence of remittances on financial development by analyzing eight key variables, including two measures of financial development, personal remittances, GDP in thousands of constant 2010 US dollars, GDP per capita in thousands of constant 2010 US dollars, trade openness, and financial openness.
Most of data except for financial openness were gathered from World Development Indicators (WDI) Financial openness is presented by KAOPEN index, which is calculated by Chinn and Ito (2006)
This study aims to analyze data from thirty-seven Asian countries between 1990 and 2014, focusing on the availability of data across various variables Due to the lack of comprehensive data for all countries, the research addresses an unbalanced dataset.
Table 3.1: Definition and expected sign of variables
CREDIT Domestic credit to private sector by banks expressed as a percentage of GDP M2 Broad money measured as a proportion of GDP
Remittances REMIT Personal remittances comprising personal transfers and compensation of employees expressed as a percentage of GDP
Country size LNGDP Log of GDP in constant 2005
The economic development and quality of country legal institutions
GDPPC GDP per capita in thousands of constant 2010 US$
Inflation INF GDP deflator (annual %) Negative
Sum of exports and imports of goods and services expressed as a share of GDP
KAOPEN Index calculated by Chinn and Ito
To assess the impact of remittance inflows across various income groups, countries are categorized as high, middle, or low income based on the World Bank's annual income classifications This study employs two dummy variables, HIGH and MIDDLE, for its analysis.
HIGH equal 1 if that country is classified as high income country while HIGH equal 0 if that country is classified as middle or low income country
MIDDLE equal 1 if that country is classified as middle income country while MIDDLE equal 0 if that country is classified as high or low income country
In total of 704 observations, 121 observations are classified as high income country, 365 observations are classified as middle income country and the remaining are classified as low income country.
Estimation methods
This paper employs various empirical models to analyze the relationship between remittances and financial development Initially, the Pooled OLS, fixed effects model (FEM), and random effects model (REM) are utilized to evaluate panel data on remittances and their impact on financial development Additionally, the System General Method of Moments (GMM) is implemented to address issues related to causal relationships.
In dynamic panel data models, the inclusion of a lagged dependent variable as an explanatory variable can result in biased outcomes for Pooled OLS, Fixed Effects Model (FEM), and Random Effects Model (REM) To mitigate this issue, models that exclude the lagged dependent variable are employed for regression analysis Additionally, to assess the impact of remittance inflows across various income groups, interactive terms combining remittances with income group dummy variables are integrated into the model The results are presented through separate equations for two indicators of financial development.
CREDITi,t = β1REMITi,t + β2GDPPCi,t + β3LNGDPi,t + β4INFi,t + β5TRADEOPENNESSi,t + β6FINANCIALOPPENNESSi,t + β0 + ui,t (1)
CREDITi,t = β1REMITi,t + β2REMIT_HIGHi,t + β3REMIT_MIDDLEi,t + β4GDPPCi,t + β5LNGDPi,t + β6INFi,t + β7TRADEOPENNESSi,t + β8FINANCIALOPPENNESSi,t + β0 + ui,t (2)
M2i,t = β1REMITi,t + β2GDPPCi,t + β3LNGDPi,t + β4INFi,t + β5TRADEOPENNESSi,t + β6FINANCIALOPPENNESSi,t + β0 + ui,t (3)
M2i,t = β1REMITi,t + β2REMIT_HIGHi,t + β3REMIT_MIDDLEi,t + β4GDPPCi,t + β5LNGDPi,t + β6INFi,t + β7TRADEOPENNESSi,t + β8FINANCIALOPPENNESSi,t + β0 + ui,t (4)
Pooled OLS model is one of the most common employed methods with general form as below:
The equation Y it = β 0 + β 1 X it + u it represents a linear regression model where Y is the dependent variable, X is the independent variable, β 1 is the coefficient for X it, β 0 is the constant intercept, and u it is the error term In this model, 'i' denotes individual observations ranging from 1 to n, while 't' signifies time periods from 1 to T By assuming uniformity among individuals in the estimated cross-section, the Pooled OLS method is employed to estimate a common constant β applicable to all individuals within the cross-section.
To effectively adopt this method, several stringent assumptions must be met, including the requirement for model parameters to be linear and the error term \( u_{it} \) to be independently and identically distributed (iid) with \( u \sim iid(0, \sigma^2) \) Furthermore, it is essential that the error term remains uncorrelated with explanatory variables over time for each individual Additionally, the variance of \( u_{it} \) should exhibit homoscedasticity, and there must be no autocorrelation between \( u_i \) and \( u_j \) (where \( i \neq j \)).
In comparison with FEM and REM, the results derived from the regression of the Pool OLS the constant intercept and slope coefficients
Fixed effects regression is a statistical method used to assess the impact of explanatory variables on a dependent variable by analyzing changes over time This approach focuses on variations within entities, allowing researchers to control for unobserved characteristics that may influence the results.
The equation Y it = α i + βX it + ε it represents a model where Y is the explained variable, X is the explanatory variable, and β is the coefficient associated with X In this equation, 'i' denotes individual identifiers ranging from 1 to N, while 't' indicates time periods from 1 to T Additionally, α i signifies the unique intercept for each individual, and ε it represents the error component of the model.
The fixed effects model (FEM) assumes that the coefficients for explanatory variables remain constant while individual-specific effects vary across individuals but do not change over time By accounting for these individual-specific effects through the intercept, FEM allows for a correlation between time-invariant individual-specific effects and the explanatory variables, although it maintains that the idiosyncratic error (εit) should not correlate with the explanatory variables.
Fixed Effects Models (FEM) are designed to eliminate the influence of individual time-invariant characteristics, allowing for a clearer evaluation of the impact of predictor variables on the dependent variable However, it is crucial that these characteristics are unique to each individual and unrelated to those of others, as well as ensuring that the error terms among individuals are correlated If these conditions are not met, the use of FEM may be inappropriate, and alternative models, such as the Random Effects Model, should be considered.
Unlike Fixed Effects Models (FEM), Random Effects Models (REM) accommodate individual variations, treating these characteristics as random factors that do not influence the model's input or output variables The standard representation of REM is as follows:
The equation Y it = β 0 + β 1 X it + u it represents a regression model where Y is the dependent variable and X is the independent variable In this model, β 0 and β 1 are the intercepts, while u it is the error term that consists of two components: ε it, which accounts for cross-sectional errors, and v it, which captures errors related to both cross-section and time series The indices i and t denote individual observations and time periods, respectively, with i ranging from 1 to N and t from 1 to T.
In the random effect model, the country-specific effect is treated as a random variable that does not correlate with the regressors This allows for the inclusion of invariant variables, which cannot be added in a fixed effect model However, if omitted invariant variables are correlated with the regressors, it can lead to biased and inconsistent estimators.
3.3.4 Tests for choosing sufficient model
This study employs three statistical tests to identify the most suitable model among the fixed effects model (FEM), random effects model (REM), and pooled ordinary least squares (OLS) model The F-test is used to determine the better model between FEM and pooled OLS, while the Breusch–Pagan LM test assesses the appropriateness of REM compared to pooled OLS Additionally, the Hausman specification test is conducted to compare FEM and REM, ultimately selecting the more effective model between the two.
The F-test is utilized to determine the best approach between the fixed effects method and Pooled OLS based on goodness of fit The regression model for the fixed effects method is represented as Y_it = α + μ_i + βX_it + ε_it, which is used to test the hypothesis effectively.
H0: μ 1 = μ 2 =… = μ n-1 = 0 F-test is calculated as follows:
F= (RSS-URSS)/(N-1) URSS/(NT-N-K) ~ FN-1,N(T-1)-K
Where RSS is the restricted residual sum of squares obtained from the Pooled OLS model; URSS is the unrestricted residual sum of squares of FEM
If the hypothesis H0 is rejected, this means that at least one μ i is different from zero Consequently we can conclude that fixed effect model is favored over the pooled OLS
The Breusch-Pagan LM test is essential for deciding whether to use Random Effects Model (REM) or pooled Ordinary Least Squares (OLS) in this study The regression model for the random effects method is represented as Y_it = β_0 + β_1 X_it + u_it, where u_it is the sum of the error terms ε_it and v_it, to test the hypothesis effectively.
H0: var(u)=0 The LM test is computed as:
If the hypothesis H0 is rejected, this means that at least one variance component is different from zero or REM is better than the pool OLS
The Hausman Specification Test is utilized to determine the appropriate model between Fixed Effects Model (FEM) and Random Effects Model (REM) by assessing the correlation between time-invariant individual effects (ui) and regressors (xit) When ui is correlated with xit, FEM is both consistent and efficient, whereas REM lacks consistency Conversely, if ui is uncorrelated with xit, REM becomes consistent and efficient, while FEM is deemed inefficient.
The Hausman Test statistic is computed as:
The null hypothesis Ho: ui is not correlated with the xit or the REM should be employed
H1: ui is correlated with xit or the FEM should be employed
THE IMPACT OF REMITTANCES ON FINANCIAL
Overview of remittance inflows and financial development in Asia
In recent decades, official remittance inflows have significantly increased, particularly in Asia, where from 1990 to 2014, these inflows steadily rose, representing nearly 50% of global remittances.
Figure 4.1: Remittances received by areas in the world from 1990 to 2014 (US$ billion)
& CaribbeanNorth AfricaNorth AmericaPacificSub-SaharanAfrica
According to World Bank data from 2014, the top three remittance recipients were Asian countries, with India receiving $70.39 billion, followed by China at $29.91 billion, and the Philippines at $28.69 billion Additionally, Pakistan and Bangladesh also featured prominently on the list, receiving remittances of $17.24 billion and $14.99 billion, respectively.
Figure 4.2: Top 10 remittance recipient countries in 2014 (US$ billion)
In 2014, Asian countries dominated the list of top remittance recipients, with Tajikistan receiving remittances that contributed 43% to its GDP, followed by Kyrgyz Republic at 30.29% and Nepal at 29.18% Additionally, other middle-income nations such as Lebanon and Armenia also saw significant impacts from remittances, which accounted for over 15% of their GDP.
Asia dominates global remittance flows, as official data often underrepresents the total amounts due to limitations in government recording systems This discrepancy suggests that the actual remittance figures are likely higher than reported.
Figure 4.3: Top 10 remittance recipient countries in 2014 (% GDP)
South Asia and East Asia are the two largest recipients of remittance inflows in Asia, with South Asia experiencing a significant and continuous increase in remittances over the past decade This region accounts for nearly half of the total remittances sent to Asia While the volume of remittances may fluctuate, with some years showing a decrease, East Asia still receives about 35% of Asia's total remittances Overall, there is a noticeable upward trend in remittance inflows in both South Asia and East Asia.
From 2000 to 2014, South Asian countries, including India, Pakistan, Bangladesh, and Sri Lanka, saw a substantial increase in remittances, with India receiving nearly US$ 58 billion, while Pakistan and Bangladesh garnered over US$ 16 billion and US$ 13 billion, respectively In East Asia, China and the Philippines ranked highest in remittance receipts, each exceeding US$ 20 billion, despite a decline in the percentage of remittances to GDP in China In the Middle East, Lebanon stands out with a notable rise in remittances contributing significantly to its economy.
Jordan and Yemen have experienced an increase in remittance inflows; however, their contributions to GDP have significantly decreased, with current ratios at 19%, 18%, and 16%, respectively In Central Asia, Tajikistan, Kyrgyz Republic, and Armenia stand out with the highest remittance-to-GDP ratios in Asia, despite their remittance volumes being lower than those in other regions.
Figure 4.4: Remittances to areas in Asia from 1990 to 2014 (US$ billion)
Middle-income countries are the largest recipients of global remittances, as many migrant workers move to high-income areas for better living and working conditions, thereby increasing the funds sent back to their home countries High-income and low-income countries follow in second and third place, respectively This trend can be attributed to the declining number of low-income countries over time and the challenges in tracking remittance flows in nations with underdeveloped financial systems Notably, middle-income countries account for approximately 70% of total global remittances, with nearly 30% of remittances going to other income groups.
Asia South Asia Middle East Central Asia East Asia remittances flowing to high income countries while just 1% or 2% is recorded belong to low income countries
Figure 4.5: Remittances received by income groups in the world from
In Asia, middle-income countries receive the largest share of remittances, accounting for approximately 50% of the total inflow However, unlike the global trend, low-income countries in this region receive more remittances than high-income countries.
Between 1990 and 2006, remittances from India and Bangladesh significantly increased, with over 75% growth in later years and reaching as high as 90% in 2014 This surge can be attributed to the classification of these countries as middle-income economies starting in 2007 for India and 2014 for Bangladesh, positioning them among the largest recipients of remittances globally.
The absence of India and Bangladesh, coupled with a decrease in the number of low-income countries, has led to a significant decline in remittance volumes to these nations By 2014, remittances to low-income countries accounted for only about 2% of total remittance flows in Asia, highlighting a concerning trend in financial support for these economies.
High incomeLow income countries, high income economies have been receving more remittances in the period from 2007 to 2014 with around 8% since 2012
Figure 4.6: Remittances received by income groups in Asia from 1990 to
Asia is the leading region globally for remittance inflows, both in absolute value and as a percentage of GDP Over the past two decades, there has been a notable upward trend in remittances, particularly in South and East Asia, as well as among middle-income countries.
4.1.2 Overview of financial development in Asia from 1990 to 2014
Since the 1990s, Asia has experienced significant financial development, evidenced by the rise in domestic credit to the private sector by banks and the increase in broad money relative to GDP Most regions in Asia demonstrated a clear upward trend in these indicators until the onset of the financial crisis in 1997-1998.
The income distribution in East Asia has experienced a significant decline, while other regions have shown little to no change in their middle and low-income ratios This situation can be attributed to the Asian financial crisis, which arose from the debt crisis affecting East Asia during that period.
Figure 4.7: Domestic credit to private sector by banks (% of GDP) in Asia from 1990 to 2014
From 1998 to 2002, East Asia experienced a persistent decline in the ratio of domestic credit to the private sector by banks relative to GDP However, following this period, East Asia successfully reversed this trend, witnessing a significant increase in this ratio, while other regions began to see growth as well During the financial crisis of 2008-2009, disparities emerged across different Asian regions, with East Asia's domestic credit to private sector ratio sharply increasing, in contrast to a decline in the Middle East and South Asia.
Empirical results
Table 4.1 summarizes the descriptive statistics of the variables used in this study, revealing high standard deviations and significant gaps between minimum and maximum values The average ratio of domestic credit to the private sector by banks to GDP (CREDIT) is approximately 51.8%, with a standard deviation exceeding 46.6%, while broad money (M2) averages 65.5% with a standard deviation of 53.5% These indicators of financial development exhibit considerable variability compared to their mean values, highlighting a substantial disparity in both domestic credit and broad money to GDP ratios among countries This indicates a pronounced inequality in the development of financial sectors within the Asian region.
Remittances contribute an average of 4.7% to GDP, with a significant standard deviation of over 7.6%, highlighting vast disparities across countries; while some nations see remittances account for nearly 50% of GDP, others report figures close to zero In countries like China and Russia, despite high absolute remittance values, these amounts represent a small fraction of GDP Conversely, high-income countries such as Japan and Saudi Arabia attract fewer migrants, resulting in lower remittance inflows Additionally, the standard deviations of variables like inflation and GDP per capita exceed their mean values, further illustrating the economic development disparities across Asia.
Table 4.1: The summary statistics of variables
Variable Obs Mean Std Dev Min Max
Figures 4.1 and 4.2 illustrate the relationships between domestic bank credits to the private sector and broad money as a percentage of GDP, alongside personal remittances and other variables Notably, total remittance inflows show a negative linear correlation with domestic credit to the private sector, while the correlation with broad money remains ambiguous.
Figure 4.11: Correlation between domestic credit to private sector by banks (%GDP) and remittance inflows (%GDP) and other controlling variables
REMIT_HIGH CREDIT Fitted values
REMIT_MIDDLECREDIT Fitted values
Figure 4.12: Correlation between broad money (%GDP) with remittance inflows (%GDP) and other controlling variables
Remittances exhibit a positive correlation with domestic bank credit to private sectors in middle-income countries, despite an ambiguous relationship In contrast, both remittances and broad money in middle-income nations show a positive correlation, similar to that observed in high-income countries Other variables, aside from inflation, demonstrate similar fluctuation patterns with financial development indicators, although these connections are less pronounced regarding trade and financial openness Notably, inflation has a clear correlation with credit from local banks to private sectors and broad money Detailed numerical representations of these correlations can be found in the correlation matrix presented in Table 4.2.
Table 4.2: The correlation between variables
CREDIT M2 REMIT GDPPC LNGDP INF TRADE-
According to the World Bank (2016), there is a strong correlation between domestic bank credit to private sectors, while the relationship with remittance inflows is notably different, indicating weak connections between these indicators.
The descriptive statistics offer insights into the relationship between remittance inflows and various controlling variables, alongside two key indicators of financial development examined in the study While the scatter diagram and correlation matrix suggest minimal correlation, the relationships between remittances and financial development indicators appear to vary significantly across different income groups These connections will be explored in greater detail using economic methods in the subsequent section.
This study explores the relationship between remittances and the percentage of domestic credit to the private sector provided by banks, as well as broad money to GDP The analysis employs equations (1), (2), (3), and (4) using Pool OLS, Random Effects Model (REM), and Fixed Effects Model (FEM) To determine the most suitable model among these three, appropriate tests are conducted, with the findings detailed in Table 4.3.
Table 4.3: The results of tests for choosing models
F value P-value Chi-square P-value Chi-square P-value Equation (1) 64.61 0.0000 2976.63 0.0000 31.69 0.0000 Equation (2) 55.82 0.0000 2476.28 0.0000 41.94 0.0000 Equation (3) 230.23 0.0000 3114.32 0.0000 33.60 0.0000 Equation (4) 215.35 0.0000 3128.74 0.0000 44.01 0.0000
In Chapter 3, we analyzed model selection tests, revealing that the F-Test and Breusch-Pagan LM Test yielded p-values of 0.0000, indicating that both fixed effects and random effects models are preferable to the pooled OLS model To determine the most suitable model between the two, we applied the Hausman test, which also produced a p-value of 0.0000 across all four equations This consistent result suggests that fixed effects models are more appropriate than random effects models for the dataset used in this study.
Table 4.4 presents the findings from fixed effect models that robustly analyze the impact of remittances on bank credit to the private sector and the ratio of broad money to GDP in Asia, as indicated by equations (1) and (3) Additionally, the effects of remittances across various income groups within the region are detailed in the results of equation (2).
The findings from equations (2) and (4) are examined to confirm that the interaction terms between remittances and both the high-income and middle-income groups are not simultaneously equal to zero.
Remittance inflows show positive trends in financial development indicators; however, their coefficients are statistically insignificant, suggesting no relationship between remittance growth and financial development in the Asia region In contrast, for different income groups, remittances positively correlate with the ratio of domestic credit to the private sector by banks to GDP in high-income countries at a significant level of 5% Conversely, in middle-income countries, remittances negatively associate with the ratio of broad money to GDP, significant at the 10% level.
On the one hand, different impacts of remittances on two indicators are also though many previous studies yield analogous impacts of these two indicators on financial development
Table 4.4: The results of FEM with robust
Notes: *, **, *** denote statistical coefficients at 10%, 5% and 1% significance levels, respectively R-square (within) is reported.
Remittances have varying impacts on different income groups, particularly highlighting a trend in middle and low-income countries where the majority of remittance funds are allocated to consumption instead of savings or investments In contrast, high-income countries utilize remittances more productively, as households tend to deposit surplus funds in banks, leading to an increase in bank credit for the private sector (Brown et al., 2013).
The analysis reveals that while the size of a country, indicated by LNGDP, has a significant positive impact, inflation negatively affects financial development, albeit insignificantly across all equations Notably, the openness of the financial sector shows a strong positive correlation with domestic credit to the private sector as a percentage of GDP, significant at both the 5% and 1% levels, while its relationship with broad money to GDP is positive but insignificant Conversely, GDP per capita and trade openness exhibit a positive but insignificant correlation with bank credit to the private sector relative to GDP, yet they significantly associate with broad money to GDP at the 5% and 1% levels, respectively Overall, these findings suggest that financial sector development is influenced by economic growth, country size, and trade and financial openness, while inflation poses a hindrance to financial progress.
The findings from the FEM analysis may be skewed due to the reverse relationship between financial sector development and remittances, along with other variables in the study To address potential endogeneity issues more effectively, the system GMM method will be employed to examine the effects of remittance inflows on the financial sector.
Table 4.5: The results of system GMM
Notes: *, **, *** denote statistical coefficients at 10%, 5% and 1% significance levels, respectively R-square (within) is reported.
Table 4.5 presents the results of the system GMM analysis for equations (5), (6), (7), and (8), highlighting the overall impact of remittances on bank credit to the private sector and broad money to GDP in Asia, as indicated by equations (5) and (7) Additionally, the effects of remittances across various income groups within the region are detailed in the findings of equation (6).