HO STE D BY Available online at www.sciencedirect.

com

ScienceDirect
EconomiA 18 (2017) 147–155

The impact of remittances on economic growth: An econometric

model
Dietmar Meyer a , Adela Shera b,∗
a Department of Economics, Faculty of Economics and Social Sciences, Budapest University of Technology and Economics, Hungary
b Pedagogue Faculty of Economics, University of Tirana, Albania

Received 20 April 2015; accepted 24 June 2016

Available online 14 July 2016

Abstract
Remittances in the world represent one of major international financial resources, which sometimes they exceed the flows of
foreign direct investment (FDI). For centuries, there have been heated debates on the sources of economic growth in developing
economies and also why some countries reflect strong economic growth comparing to others.
This study aims to observe the impacts of remittances on economic growth, using panel data set of six high remittances receiving
countries, Albania, Bulgaria, Macedonia, Moldova, Romania and Bosnia Herzegovina during the period 1999–2013. These countries
have experienced a major increase in remittance inflows, and at this time accounts for the bulk of total remittance receipts, compared
with other regions. Most countries, remittances represent the largest source of foreign exchange earnings and represent more than
10 percent of GDP.
In other words, the econometric analysis will be based on those six remittance receiving countries. The paper is then to review the
empirical literature devoted to the impact of remittances on economic growth, in order, to identify empirically if there are significant
relationships between remittances and growth in these countries. The results suggest that remittances have a positive impact on
growth and that this impact increases at higher levels of remittances relative to GDP.
© 2016 The Authors. Production and hosting by Elsevier B.V. on behalf of National Association of Postgraduate Cen-
ters in Economics, ANPEC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

JEL classification: C3; 01; F3

Keywords: Remittances; Economic growth; Panel data

Resumo
As remessas no mundo representam um dos principais recursos financeiros internacionais, que por vezes ultrapassam os fluxos de
investimento diretto estrangeiro (IDE). Durante séculos, tem havido acalorados debates sobre as fontes de crescimento econômico
em economias em desenvolvimento e também porque alguns países refletem forte crescimento econômico em comparação com
outros.
Este estudo tem como objetivo observar os impactos das remessas sobre o crescimento econômico, usando o conjunto de dados
de seis países receptores de remessas, a Albânia, a Bulgária, a Macedônia, a Moldávia, a Romênia e a Bósnia-Herzegovina durante

∗ Corresponding author.
E-mail address: adela [email protected] (A. Shera).
Peer review under responsibility of National Association of Postgraduate Centers in Economics, ANPEC.

http://dx.doi.org/10.1016/j.econ.2016.06.001
1517-7580 © 2016 The Authors. Production and hosting by Elsevier B.V. on behalf of National Association of Postgraduate Centers in Economics,
ANPEC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
148 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155

o período 1999-2013. Esses países experimentaram um grande aumento nos fluxos de remessas e, neste momento, respondem pela
maior parte das receitas totais de remessas, em comparação com outras regiões. Na maioria dos países, as remessas representam a
maior fonte de divisas que correspondem a mais de 10% do PIB.
A análise econométrica será baseada nos seis países receptores de remessas. O trabalho objetiva rever a literatura empírica
dedicada ao impacto das remessas sobre o crescimento econômico, a fim de identificar empiricamente se há relações significativas
entre remessas e crescimento nesses países receptores de remessas. Os resultados sugerem que as remessas têm um impacto positivo
no crescimento e que esse impacto aumenta com o nível de remessas em relação ao PIB.
© 2016 The Authors. Production and hosting by Elsevier B.V. on behalf of National Association of Postgraduate Cen-
ters in Economics, ANPEC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

Palavras-chave: Remessas; Crescimento econômico; Análise de dados de painel

1. Introduction

Remittances are a new financial phenomena and one of the main important sources of incomes based on it seize and
economic impact in the world. Data from (World Bank, 2014) indicates that global remittance is $430 billion dollar in
2011 and remittance is 0.31% of global GDP in 2009. The impact of remittance on economy system is more profound
in developing countries because, they receive $307.1 billion of the total N416 billion inward remittances, which is
about 74 percent. Remittance is also 27 percent of the GDP of developing countries. According to the World Bank,
remittances flows to the developing world have reached USD 414 billion in 2013 (up 6.3 percent over 2012), and are
now, behind foreign direct investment, the second largest source of external financial flows to developing countries.
Given the 232 million international migrants and the almost 70 million internal migrants, the earnings generated and
transferred by migrants are projected to reach USD 540 billion by 2016. Importance of remittances is increasing
potentially and they are becoming one of main important sources of foreign financial flows, especially in developing
countries, both in size and growth rate. The true size of remittances as well as unrecorded flows through formal and
informal channels is believed to be significantly large (Fig. 1 and Graphs 1–3).
Recorded remittances are more than twice as large as official aid and nearly two-thirds of foreign direct investment
(FDI) flows to developing countries. The enormous upward movement in remittances payments may be attributed
largely to two factors, namely; immigration between developing and developed countries has increased dramatically

Fig. 1. Residuals Control for the ‘Fixed Effects’ model. The first graph (above) shows the histogram of the distribution of the model residuals.
Comment: Form approaching normal distribution. Second graph: Comparison of theoretical quintiles. We see that the points fit very well with the
normal line. So this shows that our model is valid.
 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155 149

Graph 1. Observes a series of random performance residuals.

Graph 2. We see that the residuals series does not dhow any sign correlation since all lines are within the confidence limits.

Graph 3. Always accept the null hypothesis of absence of autocorrelation Ljung Box test (see p-value are observed all over the limit of 5%). So
there is no dependency between residuals.

in the past 20 years and decline in transaction costs as technological improvements have allowed for faster, lower cost
mechanisms for the international transfer of payments between individuals. Remittance is different from other external
capital inflow like foreign direct investment, foreign loans and aids due to its stable nature.
The purpose of this paper is to examine whether remittances has a positive effect or negative effect in the increase
in GDP per capita in the developing countries. By using empirical method in six developing countries in Europe, the
region with countries of high receiving ratio in remittances to GDP, the impact of remittances is to be observed.

2. Empirical literature review

Many economists and analysts have realized several large empirical studies on various aspect of remittance, such as
motivation of remittance senders, impact of remittance on economic growth, cost of remittance, etc. There are diverse
opinions on the impact of remittances on economic growth.
150 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155

2.1. Positive impact

There is empirical evidence that remittances contribute to economic growth, through their positive impact on
consumption, savings, or investment.
In this regard, several studies report supporting evidence on the positive impact of remittances in accelerating
investment in Morocco, India and Pakistan and in Mediterranean countries.
Adams and Page; Acosta et al. and World Bank argued that migrant remittances impact positively on the balance of
payments in many developing countries as well as enhance economic growth, via their direct implications for savings
and investment in human and physical capital and, indirect effects through consumption.
Ratha (2003) concludes that remittances increase the consumption level of rural households, which might have
substantial multiplier effects, because they are more likely to be spent on domestically produced goods.
Giuliano and Ruiz-Arranz had worked on data set of more than 100 developing countries from years 1975–2002
and found that remittances can enhance economic growth only in less financially developed countries.
The positive developmental effects of remittances focuses on the multiplier effects of consumption, development
of the financial institutions that handle remittance payments, use of remittances as foreign exchange, and the role of
remittances as an alternative to debt that helps alleviate individuals credit constraints in countries where micro-financing
is not widely available.
Barajas et al. explained that remittances are likely to expand the quantity of funds flowing through the banking
system. This in turn may lead to enhanced financial development and thus to high economic growth through one or
both of two channels: (1) increased economies of scale in financial intermediation, or (2) a political economy effect,
whereby a larger constituency (depositors) is able to pressure the government into undertaking beneficial financial
reform.
Remittances provide the catalyst for financial market and monetary policy development in developing countries.
Guilano and Arranz study found that remittances improve credit constraints on the poor, improve the allocation of
capital, substitute for the lack of financial development and thus accelerate economic growth.
Iqbal and Sattar found that in the absence of worker remittances, it was likely that exchange rate, monetary and
fiscal policies will come under pressure.
Rao and Hassan explained the effects of remittances on growth by using the Solow growth model. The study
found that migrant remittances have positive but marginal effect on growth. World Bank and IMF findings show that
remittances indirectly increase the growth rate by reducing output volatility.
Most recently, Nsiah and Fayissa had investigated the relationship between economic growth and remittances
through panel data of 64 different countries of African, Asian, and Latin American-Caribbean from 1987–2007. They
had employed panel unit root and panel co-integration tests to investigate the exact relationship between remittances and
economic growth. They found that there is positive relationship between remittances and economic growth throughout
the whole group.

2.2. Negative impact

Conversely, Amuedo-Dorantes and Pozo and López et al. posited that remittances, like capital flows can appreciate
the real exchange rate in recipient economies and therefore generate a resource allocation from the tradable to the
non-tradable sector.
Rodrik provided evidence that real exchange rate overvaluation undermines long-term economic growth, particularly
for developing countries, in that in those countries tradable goods production suffers disproportionately from weak
institutions and market failures.
Lipton, Ahlburg and Brown and Ahlburg argued that remittances undermine productivity and growth in low-income
countries because they are readily spent on consumption likely to be dominated by foreign goods than on productive
investments.
Nevertheless, Barajas et al. pointed out that the more highly integrated an economy is with world financial markets,
and the more highly developed the domestic financial system, the less likely it is that remittance receipts will stimulate
investment by relaxing credit constraints. Using, estimated dynamic simultaneous Keynesian type model for investi-
gating the impact of remittances on consumption, investment, imports and output for eight countries including Algeria,
Egypt, Greece, Jordan, Morocco, Portugal, Syria and Tunisia for the period of 1969–1993 and then further extended
 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155 151

in the other study that is, 1969–1998, Glytsos findings for both studies pointed out that the effect of remittances on
growth is partial and in several years negative impact of remittances to growth is observed.
Chami and Jahjah found that migrants remittances have negative impact on growth in per capita incomes. The study
reported three stylized facts: first, that a “significant proportion, and often the majority,” of remittances are spent on
consumption; secondly, that a smaller part of remittance funds goes into saving or investment; and thirdly, the ways in
which remittances are typically saved or invested – in housing, land and jewelry – are “not necessarily productive” to
the economy as a whole.
Empirical results also indicate that remittances may indirectly affect real exchange rate leading to the “Dutch
Disease” phenomenon, where remittances inflow causes a real appreciation, or postpones depreciation, of the
exchange rate. Exchange rates appreciate in countries with large remittances which will in turn hurt the economic
growth.

3. An empirical model of economic growth with remittances

3.1. Evaluation methods

Model 5.1, presented above, is just the general shape of the equation that we will test. The following two techniques
are characteristic of panel data study. Panel data consisting of repeated observations of the same unit, in our case, the
regional countries. Observations occur in different time, annual data that we analyze. The observation period is 14
years break, from 1999 to 2013. Our ‘Panel’ is balanced, which means that the period is the same for every state. Panel
Study data allows control of variables, which cannot be observed or measured, for example, cultural factors between
countries. This type of analysis is also performed to study the variables that change over time, but not from one country
to another, as for example, the agreements.
A fixed effects model (FE) to analyze the Panel data considers explanatory variables as non-random. This fact is
in contrast with the pattern of ‘random’ effects (RE), which considers as explanatory variables derived from random
events. In panel data analysis, the term ‘fixed effects estimator’ (or ‘Within estimator’) is used to identify appraisers
regression model parameters. A FE model is used to identify the type of impact variables, which change over time.
This technique explores the relationship between independent variables and ‘output’ within the same unit (partner
country). Every country has its own individual characteristics that probably can affect the independent variables. When
using FE, we assume that the variables are influenced by an individual characteristic and need to check for this. In
econometric terms, we talk about the correlation between the error term and independent variables. FE techniques
destroy the effect of these characteristics (independent of time) of the independent variables, so we can estimate the
net effect of independent variables. Another important hypothesis of the FE model is that these features are individual
for each country, so there should not be correlated with characteristics of other countries. Every country is different
from the other, so the error term is the constant (which captures the individual characteristics) should not be correlated
with the others. So, in substance, FE models are constructed to study the factors of change within an entity (place).
One characteristic that does not change with time may not have caused this change because it is constant for each
country.
The logic behind the ‘random’ effects model (RE) is based on the assumption that the difference between the units
(countries) is casual (random) and uncorrelated with the independent variables included in the model. According to
Green the fundamental difference between the fixed effects and the ‘random’ is the link between individual effects and
regressors in the model. These effects may be random or correlated with the independent variables. One advantage
of the technique is the inclusion of RE independent variables in modeling time. In the FE model, these variables are
‘within’ a constant term.
In this study data for all variables is collected from the publications of World Bank data set “World Development
Indicators”. Data set covers most recent year’s annual data from 1999–2013. Gujarati recommended that standard
tests of stationary are mostly applicable for large sample size and as the sample size in the current study is not so
huge that is way researcher have not employed any test for stationary. In order to investigate the impact of worker
remittances on economic growth of Albania, Bulgaria, Macedonia, Moldova, Romania and Bosnia Herzegovina panel
data regression analysis are employed. Some studies had been employed multiple regression analysis to investigate the
impact of worker remittances on economic growth (Chami et al., 2003).
152 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155

To determine the responsiveness of income growth rate to remittances and the traditional sources of economic
growth we give the equation:

GDPGROWTH = β0 + β1 WORREM + β2 GCF + β3 CONSUM + β4 ENR + β5 FDI + β5 TRADE

+ β6 POP + β7 REAL EXCH + β8 DEBT + ε (1)

Data description
Variable description
GROWTHi = Natural log of GDP per capita
WORREMi = Worker Remittances Received to GDP
CAP FIX GDPi = Gross Capital fix formation as a percentage of GDP
FCONSUMi = Household final consumption expenditure as a % of GDP
SCHOOLi = Ratio of school enrollment percetange to GDP
TRADEi = Current account BAL or balance as a percentage of GDP
POPGi = Population growth as a percentage of GDP
REAL EXCHi = Real exchange rate
DEBTi = Government debt as a percentage of GDP
β0 and ε = Intercept term and error term in the model

Proxies and expected relationship of all the variables is provided in Table 1. Here, GROWTHi is the dependent
variable which is measured in percentage. WORREMi is a stand for worker remittance is an independent variable and
taken as ratio of worker remittances received to GDP. The worker remittances and economic growth has a positive
relationship. While the remaining six variables are control variables which are add in the model to control for the effect
of other most important variables that effects the economic growth.
Where GCFit is the gross fixed capital formation as a percent of real GDP used as a proxy for investment in physical
capital; ENRit is log of secondary school enrollment used as measure of investment in human capital which has a
positive effect on the economic growth of developing countries (Stark and Lucas, 1988). TRADEit is in the terms of
trade for each country under consideration, measured by the ratio of the export to import price indices to capture the
impact of trade, or openness of the economy on economic growth. FCONit is the Final Consumption expenditures.
The theoretical relationship between house hold consumption expenditure and economic growth is positive. POPit
is the population rate. Hence, we expect the sign of the Population growth is to be negative. REAL EXCHit is the
real exchange rate and we expect a positive or negative sign. And finally we have the DEBTit Government debt as a
percentage of GDP which have a negative impact on economic growth.
To estimate the parameters corresponding to variables of interest from the data under consideration, we employ a
panel data estimation, an empirical exposition of which is provided in Eq. (1).

Table 1
Variables descriptions and their relationship.
Variable Proxy or definition Expected sign

GROWTHi GDP per capita growth in current %

WOR.REMi Workers’ remittances received as a % of GDP +
CAP FIX GDPi Gross capital fix formation % of GDP +/−
SCHOOLi Ratio of school enrollment percentage to GDP +
FCONSUMi Household final consumption expenditure as a % of GDP +
TRADE i Trade as a % of GDP +
POP G i Population growth as a % of GDP −
REAL EXCHi Real exchange rate −/+
DEBTi Total debt as % of GDP −
 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155 153

Table 2
Descriptive statistics.
gdp g remit cap fix gdp school trade

Minimum −6.799000 0.174000 13.516000 66.443000 49.316000

Maximum 10.100000 40.622000 34.104000 96.136000 145.370000
1. Quartile 1.841000 3.658000 18.829750 80.472750 84.680750
3. Quartile 6.200000 16.473500 25.470500 88.930750 123.342250
Mean 3.809711 10.935800 22.704244 84.301800 102.373256
Median 4.663500 7.840500 22.367000 82.663000 103.403500
Sum 342.874000 984.222000 2043.382000 7587.162000 9213.593000
SE mean 0.372534 1.012369 0.524288 0.645475 2.595133
LCL mean 3.069495 8.924244 21.662496 83.019255 97.216782
UCL mean 4.549928 12.947356 23.745993 85.584345 107.529729
Variance 12.490324 92.240196 24.738969 37.497438 606.124176
Stdev 3.534165 9.604176 4.973828 6.123515 24.619589
Skewness −0.881502 0.959023 0.472403 −0.268572 −0.034701
Kurtosis 0.553712 0.075541 −0.487697 0.427969 −1.075932

cons pop g real exch debt

Minimum 62.397000 −1.911000 62.244000 12.622000

Maximum 109.740000 3.008000 146.700000 150.739000
1. Quartile 72.325500 −0.640000 86.488250 23.864000
3. Quartile 88.721750 −0.068500 102.780750 54.670750
Mean 80.678922 −0.316433 96.610900 39.150433
Median 80.072500 −0.262500 99.656000 32.488000
Sum 7261.103000 −28.479000 8694.981000 3523.539000
SE mean 1.085882 0.075283 1.605631 2.300026
LCL mean 78.521298 −0.466019 93.420546 34.580330
UCL mean 82.836546 −0.166847 99.801254 43.720537
Variance 106.122555 0.510080 232.024567 476.110953
Stdev 10.301580 0.714199 15.232353 21.819967
Skewness 0.365813 1.320110 0.388189 1.837883
Kurtosis −0.414512 5.470410 1.492561 5.963035

Source: Own calculations.

3.2. Empirical results and interpretations

Several versions of Eq. (1) are tested in order to obtain a model which yields robust results and best fits the data.
Accordingly, column 2 of Table 3 presents the estimation results of a quasi fixed-effects panel with heteroskedasticity
corrected standard errors, whereas column 3 presents the estimation results for the random- effects model with bootstrap
standard errors. The correction for heteroskedasticity and the presence of the initial income converts the pooled
regression with heteroskedasticity corrected standard errors into a quasi fixed-effects model. Apart from the magnitude
of the coefficients, the results reported in columns 2 and 3 are comparable.
This portion of paper describes the descriptive diagnostics, outcomes of multiple regression analysis, results of
diagnostic tests and then discussion on these outcomes. First of all result of descriptive diagnostics is provided in
Table 2.
Similarly the average, standard deviation, largest and smallest value of independent and control variables is given
in this table.
After descriptive diagnostics the diagnostics tests are employed in order to check the assumptions of OLS. The
diagnostics tests are applied to investigates the OLS assumptions like; Multicollineartiy, Autocorrelation and Het-
eroskedasticity.
We present in Table 3, the evaluations based in three different methods (OLS, fixed effects, random effects):
Fixed effects method results: The above model expresses the impact of Remittances on economic growth, where
it is clear that Remittances have a positive impact on economic growth. Adjusted R2 means that the estimated model
explains 33.85% of the variance of the dependent variable (GDP growth). With a p-value (or alpha) which is always
154 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155

Table 3
Evaluations of the model.
Estimation method OLS Fixed effects Random effects

Dep. variable: gdp g Estimate P-value Estimate P-value Estimate P-value

remit 0,157 0,06 0,293 0,00 0,24 0,01

cap fix gdp 0,048 0,01 0,003 0,00 0,011 0,06
school 0,263 0,05 0,305 0,00 0,298 0,00
trade 0,009 0,01 0,058 0,00 0,044 0,00
cons 0,037 0,04 0,081 0,01 0,054 0,02
pop g −1,203 0,11 −1,037 0,19 −0,904 0,23
real exch −0,047 0,10 −0,041 0,02 −0,046 0,08
debt −0,031 0,10 −0,005 0,01 −0,017 0,07
F-statistic 7,527 0,00 9,542 0,00 8,526 0,00
Adjusted R-squared 0,1543 0,3385 0,2848
Number of observations 90 90 90

Source: Own calculations.

less than 1 percent of all parameters are statistically valid. Remittances coefficient is statistically significant with
a positive sign means that remittances reduces economic growth, an increase in remittances with a unit increases
GDP by 0.293%. The estimated coefficients for population growth, government spending, education (School Enroll),
Investments, Trade (Opening), exchange rate and fixed capital formation are statistically significant and have the
expected signs. A growing population with a unit of GDP decreases by 1.037%. While a growing trade with one unit of
GDP rose to 0.58% respectively. So in the above table clearly seen as statistically variables affecting economic growth.
Reject the hypothesis that the estimated model is statistically invalid because alpha test F is equal to zero (i.e. less than
5%).
Hausman test can be used to compare the two methods, ‘fixed effects’ and random effects. Null hypothesis of this
test is: individual effects are not correlated with other model regresor. If they are correlated, then the null hypothesis is
rejected, then a random effects model is not suitable. In this case, it would be a handy fixed effects model. Below we
present and output of our program, associated with this test. We can observe, that p-value is less than 1%, then reject or
accept the hypothesis zero fixed effects model. So, this test confirms again that the best method to use is what ‘Fixed
effects’.

>phtest(ModRandom,ModWithin)
Hausman Test
data: gdp g ∼ remit + cap fix gdp + school + trade + cons + pop g + real exch + debt
chisq = 21.4524, df = 8, p-value = 0.00604
alternative hypothesis: one model is inconsistent

4. Evaluation of residuals of fixed effects method

To prove again if our evaluation of data by ‘Fixed Effects’ method is available, check the Software through our
respective residues. So, the analysis of the following model performs better with fixed effects assessment.
To see more clearly the performance of standard residuals, perform the test and Ljung-Boxin see the structure of
interdependence through global autocorrelation functions.

5. Ljung Box test

Ljung-Box test (is named for Greta M. Ljung and George EP Box) is a type of statistical test that analyzes whether
the presence of autocorrelation is statistically valid in a time series. He tests the overall variability based on a number
of delays. This test is sometimes known as the Ljung-Box Q, and it is closely related to the Box-Pierce test (which is
named after George EP Box and David A. Pierce). Box-Pierce test statistic is a simplified version of the Ljung-Box test
 D. Meyer, A. Shera / EconomiA 18 (2017) 147–155 155

statistic for which subsequent studies have shown poor performance. Ljung-Box test is widely applied in econometrics
and other models of time series analysis.

6. Conclusions

This study is conduct to explore the impact of worker remittances on economic growth of Albania and five regional
countries by employed the annual panel data from 1999–2013. In order to explore the relationship between worker
remittances and economic growth multiple regression analysis is utilized. Different diagnostic tests are applied in
order to confirm the major assumption of multiple regression analysis like multicollinearity, heteroskedasticity and
autocorrelation. After employing all these tests multiple regression analysis is conducted which shows that worker
remittances is positively and significantly contribute in the economic growth of six countries. So, contribution of
worker remittance is the significant and most important in economic growth. But its productive use can help the
economy of these countries to maintain and improve the economic growth by investing this money into consumption
and investments. This study has being focusing only on relationship between worker remittance and economic growth
and the upcoming studies must investigates the relationship of worker remittances with other macroeconomic indicators
that have a great impact on economic growth.

References

Chami, R., Fullenkamp, C., Jahjah, S., 2003. Are Immigrant Remittances Flows a Source of Capital for Development, IMF Working Paper
(WP/03/189).
Ratha, D., 2003. Workers’ Remittances: An Important and Stable Source of External Development Finance. In Global Development Finance 2003,
Striving for Development Finance. The World Bank, Washington, DC.
Stark, O., Lucas, R.E.B., 1988. Migration, remittances, and the family. Econ. Dev. Cult. Change 36 (3), 465–481.
World Bank, 2014. Global Economic Prospectus. World Bank, Washington, DC.