The Impact of Migrant Remittances on Economic Growth: Evidence from South Asiar oie_12008 985..998

of 14
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
The Impact of Migrant Remittances on Economic Growth: Evidence from South Asiar oie_12008 985..998
Document Share
Document Tags
Document Transcript
  The Impact of Migrant Remittances on EconomicGrowth: Evidence from South Asia roie_12008 985..998  Arusha Cooray* Abstract Incorporating migrant remittances among other variables into a growth model, and employing panel dataover the 1970–2008 period, this study investigates the impact of migrant remittances on economic growth inSouth Asia. Migrant remittances are found to have a significant positive effect on economic growth.A sig-nificant positive interactive effect of remittances on economic growth is detected through education andfinancial sector development. 1. Introduction Remittance inflows into the developing economies have increased tenfold fromUS$31,058 million to US$327,591 million over the 1990 to 2008 period, accounting forthe second largest foreign exchange inflow next to foreign direct investment, and insome cases the largest (World Bank, 2010). Latin America is the largest remittance-receiving region in the world, followed by South Asia. While much work has beenundertaken on remittance receipts into Latin America (Orozco, 2000; Woodruff andZenteno, 2007; Mundaca, 2009), there is little  empirical   work on remittance inflowsinto South Asia. Remittance receipts into South Asia have increased significantly overthe 2000 to 2008 period. In Pakistan, remittances have increased from a little overUS$1 billion in 2000 to US$6.4 billion in 2008; in Bangladesh, from US$1.9 to US$8.9billion; in India from US$12.9 billion to US$49.5 billion; in Nepal, from US$111million to US$2.7 billion; and in Sri Lanka, from US$1.2 billion to US$3.0 billion(World Bank, 2010). India is the largest remittance-receiving country in the worldaccounting for 73% of the flow into South Asia, with Bangladesh seventh and Paki-stan eleventh (Maimbo et al., 2005; World Bank, 2010).Remittances have been found to have a number of positive effects on the develop-ing economies. They have served as insurance policies against risks associated withnew production activities and reduced income inequality (Taylor, 1999), helped lowincome households to smoothen their consumption by reducing their vulnerability toadverse shocks (Yang and Choi, 2007), increased the propensity to save (Adams,2002), reduced poverty (Adams and Page, 2005) and even helped build schools andclinics (see Orozco, 2000; Martin et al., 2002). Remittances have also been found topromote economic growth (Mundaca, 2009), promote financial sector development(Aggarwal et al., 2006; Giuliano and Ruiz-Arranz, 2009) and reduce output volatility(Chami et al., 2009). * Cooray: University of Wollongong, Northfields Avenue, NSW 2522, Australia. Tel: 61 2 4221 4017; Fax:61 2 4221-3725; E-mail: I wish to thank two anonymous referees and Amnon Levy forvaluable suggestions. Review of International Economics, 20(5), 985–998, 2012 DOI:10.1111/roie.12008 © 2012 Blackwell Publishing Ltd  Given the positive impact of remittance flows into the developing economies, andrise in migrant remittances into South Asia, the contribution of the present study is toinvestigate the impact of migrant remittances on economic growth in this region.Thestudies of Mundaca (2009) and Giuliano and Ruiz-Arranz (2009) among otherssupport the argument that remittances promote economic growth.Several studies alsoshow that growth rates are higher in countries with a well developed financial sector(King and Levine, 1993; Cooray, 2009a), a high human capital stock (Mankiw et al.,1992) and well developed infrastructure. Studies further support the argument thatremittances have contributed to financing education (Cox et al., 2003; Ranasinghe,2007) and promoting financial sector development (Aggarwal et al., 2006). Thereforehigh-remittance receiving countries with comparatively better developed physical andhuman capital stocks, and financial systems, should be able to successfully channelremittance flows towards economic growth. Accordingly, this study also investigatesthe interactive effects of remittances on economic growth through human capital,financial sector development and the physical capital stock.The paper is organized as follows. Section 2 briefly discusses region specific charac-teristics. Section 3 presents the empirical model. Section 4 details the data and meth-odology. Section 5 discusses the empirical results and section 6 concludes. 2. Country Characteristics A series of economic reforms were undertaken under the auspices of the IMF and theWorld Bank in Sri Lanka in the 1970s, Bangladesh and Pakistan in the 1980s, India,Nepal and Bhutan in the 1990s. In the years following liberalisation, the growth ratesof these countries have accelerated, in particular, that of India. Labor migration wasencouraged through the introduction of several measures. 1 Bangladesh for example,introduced a special savings scheme in the form of Wage Earner Bonds to promotemigrant savings. In Pakistan, local migrants remitting US$2,500 per annum areentitled to duty-free imports of up to US$700 per year, and non-resident Pakistanis(NRPs) remitting a minimum of US$10,000 through banking channels, are entitled toduty-free imports of up to US$1,200 from 2001. NRPs also have access to a merit-based quota system in all public professional colleges and universities, are able to par-ticipate in a lottery for land in public housing schemes at concessionary rates if theypay in foreign currency, and buy shares in privatized companies. Sri Lanka offersmigrants pre-departure loans to cover travel costs, migrants and their families aregranted free life insurance and can maintain non-resident foreign-currency accountsthrough which remittances can be transmitted. In India, certain states such as Keralahave set up Human Resources Corporations to promote migration. Migrants are alsopermitted to transfer capital between their home country and destination country freeof government regulations (Khatri, 2007). Evidence shows that in Nepal, migrantremittances have led a decline in poverty from 42% in 1995–1996 to 31% in 2003–2004 (Pant, 2008). The stock of migrants from South Asia stands currently at 12.2million or 0.7% of the population compared to 215.8 million or 3.2% for the world(Ratha et al., 2011). Table 1 records remittance inflows as a % of GDP into the coun-tries under study.Note that remittances as a % of GDP have steadily increased into all of the econom-ies with the exception of Pakistan, which has experienced a decline in remittancereceipts since 1980, and the Maldives. In Pakistan this can be attributed primarily topolitical and economic instability.The Maldives, on the other hand, is a labor receivingrather than a labor sending country.Thus remittance inflows play an important role in986  Arusha Cooray © 2012 Blackwell Publishing Ltd  contributing to the economic growth of all South Asian economies apart from theMaldives, which is primarily a net outflow country. Remittances have also contributedto strengthening the balance of payments deficits in all countries with the exceptionof the Maldives. In the Maldives, remittance outflows were 27% of the trade deficit in2006 (de Mel and Jayaratne, 2009). Migrant remittances to GDP, moreover, exceedoverseas development aid and foreign direct investment to GDP into all South Asiannations with the exception of the Maldives. With the opening up of the economies inthis region, the growth rates of these countries have accelerated.Accordingly, investi-gating the role of migrant remittances in South Asia’s growth trajectory is particularlyrelevant. 3. The Model The model is based on the neo-classical production function.The production functionis specified as follows: Y A e Z K H L e it it it  it  it it  i it  =  − − 01 δ  ϕ  α  β α β  ε  (1)where  Y  it   is aggregate output,  A i 0  the level of technology,  K  it   the stock of physicalcapital,  H  it   the stock of human capital and  L it   the labor force of country  i  in period t  . The parameter  d  , captures the growth effects of omitted trended variables, Z it  represents the main variable of interest, the ratio of migrant remittances to GDP, andother control variables which contribute to the adoption of new technologies. Theparameter  j  i , captures the growth effects of the variables in  Z  it  . The term,  e  it  , is arandom disturbance that captures the aggregate effect of all other factors. Dividingboth side of (1) by  L ,  y A e Z k h e it i it it  it  it  i it  =  0 δ  ϕ  α  β  ε  (2)where  y  is output per capita,  k  is physical capital per capita and  h  is human capital percapita.Taking the natural logarithm transformation of (2) yields:ln ln ln ln ln .  y A t Z k h it i i it it it it  = + + + + + 0  δ ϕ α β ε   (3)Incorporating the components of   Z  it  , equation (3) can be specified as follows:ln ln ln ln  y a t REM GDP EX GDP FDI GDP  it i it it it  = + +  [ ]  +  [ ]  +  [ ] + 0 1 2 3 δ ϕ ϕ ϕ ϕ  44 5 2ln ln ln ln M GDP G GDP k h it it  it it it  [ ]  +  [ ]  + + + ϕ α β ε   (4) Table 1. Remittance Inflows into South Asia as % of GDP Country 1980 1990 2000 2008 Bangladesh 1.87 2.58 4.17 11.23India 1.20 1.50 2.80 4.11Nepal — 1.49 2.02 21.61Maldives — — 0.35 0.24Pakistan 8.64 5.01 4.91 4.26Sri Lanka 3.76 4.98 7.13 7.23 Source : World Development Indicators. REMITTANCES IN SOUTH ASIA  987 © 2012 Blackwell Publishing Ltd  where the  Z   includes the ratio of migrant remittances (REM) to GDP, and otherstandard control variables used in the growth literature including, the ratio of exports(EX) to GDP, the ratio of foreign direct investment (FDI) to GDP, the ratio of   M  2( M  2) to GDP and the ratio of government expenditure ( G ) to GDP. 4. Data and Methodology Data The data are annual and cover the 1970–2008 period for India, Sri Lanka, Pakistan,Bangladesh, Nepal and the Maldives. 2 The dependent variable in the study is thegrowth rate of output per capita. The main independent variable is the ratio of migrant remittances to GDP. These are formal remittances that are recorded in theNational Accounts and are from the World Development Indicators. Remittances aredefined as the addition of workers’ remittances, compensation of employees andmigrants’ transfers. It is estimated that a large proportion of remittance flows aretransmitted through informal channels.A limitation of the study, therefore, is that it isonly able to capture official flows that are transmitted through formal channels. 3 Thecapital stock is estimated using the perpetual inventory method. 4 Since the work of Mankiw et al. (1992) and Barro (1991), there has been a general consensus on thepositive role played by human capital in economic growth. Thus, human capital isemployed in the estimation, and is measured by the secondary enrolment ratio(Barro, 1991; Mankiw et al., 1992). Given the large literature on the positive associ-ation between financial sector development and economic growth—see King andLevine (1993) and Cooray (2009a) for example—the ratio of   M  2 to GDP is used as aproxy for monetary policy and the level of financial sector development. It can altern-atively be argued that a large money supply may reflect an irresponsible monetaryauthority or a low velocity of money. 5 The ratio of   M  2 to GDP is used to measurefinancial sector size and depth in the present study because: one, it is generally used inthe literature as a proxy for financial sector development; and two, financial deregula-tion in South Asia has contributed to a significant rise in deposit mobilization, leadingto increases in the ratio of   M  2 to GDP. 6 Moreover, this region has not in general facedhyper-inflation or episodes of price instability suggesting any irresponsibility on thepart of the monetary authorities.The ratio of government expenditure to GDP is usedto capture fiscal policy (Barro, 1991; Cooray, 2009b), as in developing nations, the gov-ernment plays an important role in the distribution and allocation of resources. Inaddition to the ratio of migrant remittances to GDP, the degree of openness of theeconomies is also measured by the ratio of exports to GDP and the ratio of FDI toGDP. Balassa (1985) shows that exports can provide greater access to internationalmarkets and hence economic growth. Similarly, both exports and FDI can promotefaster technological innovation and learning from abroad (Balasubramanyam et al.,1996). Evidence on the effects of FDI on economic growth, however, has been mixed.The polity index of Marshall and Jaggers (2010) is also used to control for institutionsas a further measure of robustness. 7 Table 2 presents summary statistics and sourcesfor the data used in the study. Methodology Several alternative methodologies are used to test the model.The preliminary estima-tion on the panel data is carried out using ordinary least squares (OLS).The model is988  Arusha Cooray © 2012 Blackwell Publishing Ltd  also tested using fixed effects, and the system general method of moments (GMM) tocheck the robustness of the results to the estimation method.The panel data model isexpressed by equation (5): ∆  y y X u it it it i t it  = + + + + − γ β µ η 1  (5)where  D  y it   is the first difference of output per capita for country  i  in period  t  . Allcontrol variables are captured by the vector  X  it  .  m  i  is a country specific effect and  h t,  afixed time effect.  u it   is a random error term that captures all other variables. Allvariables are converted into natural logarithms for the empirical estimation.Interactions terms are added to the above specification to investigate differentialeffects. Both fixed and random effects models were estimated. A Hausman testshowed greater support for the fixed effects model, therefore results are reported forthe fixed effects estimator.The explanatory variables are not strictly exogenous in this model. An approachthat allows controlling for the joint endogeneity of explanatory variables through theuse of internal instruments is the Arellano–Bover (1995) and Blundell–Bond (1998)system GMM estimator. Here the levels equation (6) is combined with a first differ-ence equation (7). The equation in levels, (6), is instrumented with lagged first differ-ences of the variables, while the equation in first differences, (7), is instrumented withlagged levels of the variables.  y y X u it it it i t it  = + + + + − γ β µ η 1  (6) Table 2. Descriptive StatisticsVariable Obs MeanStandarddeviation Min Max Source Per capita income(constant 2,000US$)237 530.42 554.90 138 3418 World DevelopmentIndicators (2010)Per capita capital 206 1.22e + 11 2.45e + 11 2.90e + 07 1.56e + 12 Authors own calculationEnrolment ratiosecondary(% gross)181 35.78 18.38 8.74 88.48 World DevelopmentIndicators (2010) M  2 (% of GDP) 242 34.30 12.88 8 73 World DevelopmentIndicators (2010)Governmentexpenditure(% of GDP)227 11.17 4.92 3 28 World DevelopmentIndicators (2010)Migrantremittances (%of GDP)187 3.86 3.77 0.18 23.82 World DevelopmentIndicators (2010)Exports (% of GDP)252 25.77 27.35 3 166 World DevelopmentIndicators (2010)FDI (% of GDP) 211 0.74 0.96  - 0.20 6.71 World DevelopmentIndicators (2010)Polity Index 229 0.67 7.30  - 10 9 The Polity IV Database,Marshall and Jaggers(2010)REMITTANCES IN SOUTH ASIA  989 © 2012 Blackwell Publishing Ltd
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!