UNIVERSITY OF GRONINGEN
FACULTY OF ECONOMICS AND BUSINESS
ASSESSMENT:
JAKUB PIEROŃ
Remittances in Sub-Saharan Africa:
CURSE OR BLESSING?
Master Thesis
Supervisor: PhD T.M. Harchaoui
Co-assessor : PhD A.A. Erumban
Student number: s2784734 e-mail: j.pieron@student.rug.nl
1
Table of Content
Abstract ... 2
Introduction ... 3
1. Remittances, migration and Dutch disease ... 7
1.1. Theory ... 8
1.2 Facts on remittances ... 11
1.3 Facts on remittances and migration ... 14
1.4 Dutch disease ... 18
2. Model and empirical analysis ... 19
2.1 Model ... 20 2.2 Econometric Methods ... 22 2.3 Data ... 22 2.4 Descriptive ... 24 2.5 Econometric Results ... 25 2.6 Robustness check ... 33
3. Conclusions and Discussion ... 35
References ... 37
2
Abstract
The research is analyzing the impact of remittances, searching for Dutch Disease symptoms. The values of remittances skyrocketed in last decades, reaching the size of even 10 percent of GDP for many countries. The impact of remittances on the economy is evitable. The study is made on 13 African countries (Botswana, Ghana, Egypt, Ethiopia, Kenya, Malawi, Mauritius, Morocco, Nigeria, Senegal, South Africa, Tanzania, Zambia). The analyzed sample consist of countries mostly form Sub-Saharan Africa region. This region is growing very fast in last decades, but is still the place where the most of the least developed economies are located.
First sections of the thesis describe facts and theory behind remittances. The analysis is performed by finding Spending Effect and Resource Movement Effect, which would suggest Dutch Disease. The effects of remittances are analyzed on REER, inflation and labor shift towards non-tradable sector of the economy with SUR estimation method on panel data for 1980-2010 period. The results are not stating clear if remittances are the source for Dutch Disease, even more, for some countries the influence can be perceived as positive.
The research is the contribution to the literature, helping to increase the effectiveness in policy-making for developing countries with high inflows of remittances.
3
Introduction
The land of hope, the land of contrast and the land of future growth. Africa has been named in many different ways for decades and centuries, changing its position on World scene from very important developing player, to forgotten and left alone for sluggish stagnation or even economic regress. After last, not encouraging decades of XX century African economy seems to strike back with the increase noticed in economic indicators and growth rates.
The first decade of XXI century, according to World Bank (2015), brought annual 5 percent growth in GDP for Sub-Saharan region. Compared to previous economic take-offs, this time region’s prospects for long-term economic growth are more solid. This is due to countries’ internal changes in economic and social oriented performance. The additional fuel to the fire of economic growth has been made by external, global incentives and trends focusing on this still unexplored developing region (McKinsey, 2010).
The best example of economic miracle in the region can be Mauritius, with spectacular growth and evolution from poor, sugar-based economy to upper-middle-income country with highly competitive economy (Svirydzenka & Petri, 2014; Wignaraja & O’Neil, 1999). Nigeria is another example of growth, with its largest regional economy and most populated area. It constantly grew from 1991 reaching 34% GDP growth in 2004 and continued to growth at 5-7 percent level during following years. In general, year 2004 was a period of improvement for Sub-Saharan Africa, where the average GDP growth rate for countries in the region was established at the impressive level of 10% and similar to Nigeria’s case, stayed positive till 2013. (World Bank Data,2015)
4
The result of migration are various, through the changes in labor force supply in the economy to the increase of private, unrequired transfers of money from migrants to relatives left in the country of origin – remittances. For many developing countries, the flows of remittances are exceeding the values of foreign direct investment or foreign aid received (Nyamongo et al. 2012). Countries in Sub-Saharan Africa have difficulties in attracting foreign investments due to low international competitiveness, which is defined as the set of institutions, policies and factors determining the level of productivity of a country (Chami et al. 2008, Schwab, 2014).
Remittances are the significant and one of the most important sources of new funds for developing countries, larger than traditional, official development assistance and aids, just second force after FDI (World Bank, 2003, 2004; Aggarwal et. al, 2006; Giulia & Zazzaro, 2011; Giuliano & Ruiz-Arranz, 2009; Rao & Hassan, 2011). The numbers are striking. Remittances to developing countries are expected to grow by an annual average of 8.8 percent, reaching the level of $515 billion in 2015, as presented in the press release of World Bank (2013). The share by regions in years 1980-2010, for developing countries by regions is presented on figure1.
Figure 1 "The size of remittances per region" source: World Bank. World Development Indicators 2015
5
regions. However, in this region remittances are extremely important source of foreign capital for many countries. For smaller countries like Lesotho, remittances account for more than 20 and in huge Nigerian economy are fluctuating between 4 – 13 percent of the GDP depending on the year (World Bank Data, 2015). This is leading to the other comparison of remittances between the same regions. The interesting and different trends can be visible, when remittances are compared according to GDP ratio. Complete data for all regions has been available in the 1995-2010 period. However the average ratio of remittances to GDP was analyzed in years 1980-2013 because the lack of data for regions was random and mostly individual for regions in taken years. The table 1 shows the change in the ranking of regions. At the end of the period, Sub-Saharan Africa is the second biggest receiver of remittances in relation to GDP, just behind Middle East and North Africa, where Egypt and Morocco are mostly responsible for the big share of it.
Table 1 "Ratio of remittances to GDP - regions"
Region / year 1995 2000 2005 2010 average
1980-2013
Middle East & North Africa 4,40% 2,94% 3,51% 3,04% 4,02%
Europe & Central Asia 1,40% 1,81% 2,06% 2,07% 2,16%
Sub-Saharan Africa 1,06% 1,54% 3,26% 2,28% 1,41%
Latin America & Caribbean 0,78% 0,99% 1,91% 1,16% 0,95%
East Asia & Pacific 0,70% 0,97% 1,12% 1,00% 0,81%
source: World Bank. World Development Indicators, 2015
Considering the increase of total value of remittances from US$ 37 billion in 1980, to US$ 460 billion in 2013 globally, where only in the last decade this number doubled, the impact of remittances on economy has to be an important issue, especially in African region, where the size of remittances is very high comparing to the GDP (World Bank Data, 2015).Due to this high ratio, the impact on this region should be different, probably stronger than on the other regions.
6
et al. 2010) using cross-country data. Studies on the improvement in health, expanding savings and investments or encouraging entrepreneurship is also visible (Woodruff & Zenteno, 2001; Maibo & Ratha, 2005; Yang, 2005), as much as financial development (Aggarwal et al. 2006; Chowdhury, 2011; Nyamongo et al. 2012) or income smoothing effect (Amuedo-Dorantes & Pozo, 2011).
Remittances can also be a curse for developing countries, affecting the international competitiveness of receiving economy. This research is trying to find the evidence of growth diminishing impact of remittances in form of Dutch Disease syndrome, analyzing cross-country dataset. More precisely, my research question is if workers’ remittance inflows can lead to spending effect (appreciation of real effective exchange rate and increase of inflation) and resource movement effect (labor shift towards non-tradable sector of the economy) which would be the symptoms of Dutch Disease. Remittances in this case are treated as the resource abundance and this fact leads to the broader scope of studies . In recent years, the literature on the impact of resource abundance on economies significantly expanded, strongly underlining the negative long-run effects on the economy (Frankel, 2010; van der Ploeg, 2011, Papyrakis & Raveh, 2013). Recent findings and already existing studies on remittances and Dutch Disease where the impact of remittances on exchange rate has been tested (Chowdhury & Rabbi, 2013; Owusu-Sekyere et al. 2011; Bourdet & Falck,2007), gives the incentive to contribute to this literature a new perspective, combining studies on the resource abundance and remittances. In addition to well-known remittances’ impact on exchange rate approach, used in many studies, this research is looking into other factors as well. The research is additionally testing for resource movement effects for economies, together with effects on REER and inflation, which will suggest the Spending effect. Controlling for these three factors gives broader perspective and should help to give the definite answer if remittances are causing Dutch Disease, or are the blessing for African economies.
7
competitiveness will be difficult to reach. The purpose of this research is to broaden the knowledge about developing markets by analyzing the effects of remittances in Sub-Saharan Africa on macroeconomic level and what is more, to indicate variations between countries and what is driving this difference.
The first chapter gives the overlook on the theory and facts about remittances, showing how workers’ financial support varies between countries and how numbers of migrants are different for each country. Further on, the possibilities of Dutch Disease symptoms are described and how it can be noticed by policy makers. The second part is providing empirical analysis for the sample of 13 African countries, which leads to concluding remarks in the third section of the study.
1. Remittances, migration and Dutch disease
Immigrants from less developed countries usually send regular remittances home to support family or friends. The support should improve the quality of life for remittance receivers, maximize private welfare by the distribution of revenues from hosting country. This would enable to acquire financial and non-financial assets in the country of origin (Bouhga-Hagbe, 2004). Despite the fact that remittances have helped to reduce poverty and inequality (Adams & Page, 2003; Fajnzyber and Lopez, 2008), the breadth of the influence, positive and negative as well, on recipient economies is rising more and more questions. The role of remittances in economic development has been noticed decades ago (Johnson and Whitelaw, 1974; Lucas and Stark, 1988), but remittances did not achieve such high amounts those times, at least not in the official data. This part is explaining the theory and facts about remittances, moving into trend analysis and migration impact on remittances.
8
1.1. Theory
To begin with the theoretical part of the research, one should have understand the meaning of the term “remittances”. Remittances, as mentioned in introduction are private and unrequired transfers of money from workers abroad to relatives left in home. The flow of remittances is recorded in national accounts of the receiving economy as a foreign transfer. Following this fact, this financial flow is tied with expenditure part of the economy, associated with consumption of goods and services, or savings and investment (Loser et al., 2006). The problem with estimating private transfers is significant. Many remittances are transferred by informal way.
According to World Bank (2011), Sub-Saharan Africa is significantly supported with remittances not only via formal channels. The informal channels can be responsible for over 50 percent of remittances inflowing to the African economies. This estimates suggest that official data are undervalued and the real impact of remittances can be much stronger (Nyamongo et al. 2012).
9
To find more about macroeconomics’ impact, the main questions about the influence of these flows would be addressed considering the remittance’s effect on external current account, exchange rate, GDP growth, investment and domestic savings (Loser et al. 2006). The stability of remittances during shocks affecting global economic conditions was noted in the literature (Ratha 2003, World Bank 2011). Comparing to other international financial flows during the worldwide financial crisis after 2007, remittances did not fall as significantly as FDI or financial aids. The indicator of remittances for developing countries dropped just by 5,2% while the same indicator for FDI suffered from declining by 39.7% between 2008 and 2009 (Yang, 2011).
This macro level impact of external transfers (FDI, foreign aid, remittances) and its welfare implications were already analyzed already decades ago by Keynes(1929) and Ohlin (1929). In their studies the main goal was to explain the adverse impact on economy caused by international financial flows (Chowdhury & Rabbi, 2012), but remittances were not the main issue. Nowadays, the increase in the flow of remittances can be even perceived as a new implication of globalization and its new phase due to the increase of cross-country migration. At this stage, the policy makers are analyzing remittances more carefully. Sometimes the flows of remittances are the reason of changing international financial relations between countries, for example via liberalization of capital flows (Singer et al., 2008). The fact that during last decades countries have been liberalizing their economic systems, making it more favorable for remitters to send money home, with trade openness and capital account liberalization, just increased the number of remittances recorded in official surveys. The trend of deeper international economic and political integration with the fruits of establishing regional agreements like ASEAN, UE, NAFTA, AGOA, or simply WTO just underpinned increasing flexibility in labor migration. Combining these facts, the progressive importance of remittances is evitable (Chami et al., 2008).
10
phenomenon of volatility of foreign inflows was studied in several articles (Amuedo-Dorantes & Pozo, 2011; Chami et al,2008; Yang, 2011), showing the stability of remittances during financial crisis and other disturbances on global markets. During the same market disturbances FDI and official aid flows had tendencies to significant downfalls.
Continuing analyzing the volatility, the time comes for income smoothing by remittances. The shortfall of social insurance programs, saving opportunities or borrowing constrains in developing world suggest one more function of remittances. Remittances have replaced and responded to these market failures as stabilizing household income of family members who stayed home in low living standards (Clarke & Wallsten, 2003; Amuedo-Dorantes & Pozo, 2011). However, the income smoothing does not have to be the only reason for sending remittances home. The positive outcomes can derive from family business support, expected returns on investments at home in fixed and tangible assets or building a goodwill with relatives by sending money back home (Amuedo-Dorantes & Pozo, 2006). Inflow of remittances can also be a link to financial inclusion for developing countries, like in case of Sub-Saharan Africa, where international remittances increase demand for saving instruments (Aga & Peria, 2014).
The impact of remittances on economic growth is found to be various among countries in different studies. The potential long-run economic growth based on worker’s remittances can be achieved by increase in investments and better human capital formation which would have the contribution in rising total factor productivity (Abdih et al, 2012; Chami et al,2008; Bouhga-Hagbe, 2004). In addition, the poverty reduction and increase in purchasing power in home countries link remittances to investments which were found in works of Giuliano & Ruiz-Arranz (2009) and Rao & Hassan (2011). Investments followed with financial development and potential growth enhancement.
11
shifts of resources allocation away from tradable sector of the economy. This positive change in price of non-tradable goods affects real exchange rates, with its appreciation (Lartey et al., 2012). The negative impacts of remittances and Dutch Disease is discussed more at the end of this chapter.
The impact of remittances can be stronger with each year, because of the increase of remittance flows in last decades, mainly supported by enhanced international migration. Demographic issues are suggesting increase migration even more in incoming years, as the demand for labor in OECD countries is not going to be satisfied by “Western society”. It is due to working-age population is expected to keep decreasing in OECD countries, whereas Africa is the region of opposite trends. The increase in migration to Developed Countries, can increase the numbers of remittances flowing to Africa, influencing local economies (Ratha et al. 2011).
1.2 Facts on remittances
Considering the increase of total value of remittances from US$ 37 billion in 1980, to US$ 460 billion in 2013 globally, where only in the last decade this number doubled, researchers and policy makers perceive immigrants’ remittances as very important issue and recently numerous publication have been tackling this problem ( Chami et al. 2008, Yang 2011, Amuedo-Dorantes et al. 2011, and many more).
12
The highest ratios has been recorded for small economies. Lesotho unquestionably stands out from the rest of the sample. The country which is land-locked surrounded by big, South Africa which has one of the regional top performing economy, is economically dependent on its neighbor (Lundahl, 2003). The similar isolated story is present for Comoros, the third smallest African country, located on the archipelago on Indian Ocean. For Comoros, the share of remittances was increasing with each year. From the level of 1,31 percent in 1980, remittances to GDP ratio has risen to 16,86 percent in 2010. Comoros is one of the poorest countries in the World (UN, 2015), high ratio for this country can support the theory about negative correlation between economic growth and the size of inflowing remittances.
The picture presented by Mauritius, shows opposite trends. The Mauritian ratio of remittances to GDP decreased significantly in the mid-2000’s after economic transformation (Yeung Lamko, 1998; Svirydzenka & Petri, 2014), but as seen, the negative correlation between inflows of remittances and GDP growth sustained visible. With the time passing, the ratio of remittances to GDP was decreasing. After the pick in terms of remittances received in 2004, this type of foreign inflows started decreasing significantly, in pair with increasing economic
1980 1985 1990 1995 2000 2005 2010 average 1980-2010 Cabo Verde 28,17 15,07 19,27 21,73 16,12 14,06 7,87 7,32 Comoros 1,31 3,69 3,97 5,27 .. 14,21 16,86 15,81 Egypt 11,77 9,26 9,93 5,36 2,86 5,59 5,69 7,29 Ethiopia 0,14 0,15 0,04 0,36 0,64 1,40 1,15 0,47 Gambia 0,16 .. .. .. .. 9,50 12,16 6,39 Ghana 0,02 0,09 0,10 0,27 0,65 0,92 0,42 0,33 Kenya 0,38 1,08 1,62 3,30 4,23 2,27 1,71 2,03 Lesotho 60,99 90,41 78,57 47,79 61,99 43,80 28,04 61,87 Malawi .. .. .. 0,04 0,04 0,82 0,40 0,22 Mali 3,32 5,10 4,42 4,55 3,02 3,34 5,02 3,93 Mauritius .. .. .. 3,27 3,86 0,01 0,01 2,53 Morocco 5,00 6,80 6,96 5,30 5,84 7,71 7,08 6,48 Nigeria 0,03 0,03 0,03 2,82 3,00 13,04 5,37 3,26 Senegal 2,20 2,68 2,49 2,99 4,99 9,06 11,43 4,79 South Africa 0,08 0,06 0,12 0,07 0,25 0,24 0,28 0,15 Sudan 3,44 2,09 0,50 2,50 5,23 2,65 1,68 3,00 Tanzania .. .. .. 0,02 0,08 0,14 0,18 0,12 Togo 0,87 2,02 1,65 1,15 2,65 9,10 10,61 3,97 Sub-Saharan Africa 0,63 0,64 0,68 1,06 1,54 3,26 2,28 1,37
Table 2 Remittances to GDP ratio for selected African countries
13
growth. The records at the end of the analyzed period showed almost invisible 0,01 remittances’ percent size of the GDP value. The case of Mauritius can be the interesting topic of further studies. More detailed analysis can answer if remittances supported the economic miracle of this country as a blessing, since in 1990’s remittances were fluctuating around 4 percent of Mauritian GDP and after a decade, the stronger economic performance was visible.
The similar story with even higher remittances’ ratio is presented by data on Cabo Verde. Till the end of XX century, remittances were the size of even 20 percent of GDP, after 2000’s rocketing GDP growth occurred. In Cape Verde, the evidence of remittances causing Dutch Disease were found, but the negative effects were limited by growth-oriented distribution policy (Bourdet & Falck, 2006).
Understanding such high ratios for small economies as Lesotho, surprising can be the significant ratio’s value for big African economies like Nigeria (3,23% on average) or Kenya (2,03% on average). Nigeria, which is currently the biggest African economy according to World Bank Data (over 460 billion USD in 2012, more than 500 billion USD in 2013 and still growing) received remittances worth more than 20 billion USD and this is only data on official inflows (World Bank Data, 2015).
On average in years 1980-2010, in the Sub-Saharan region remittances were accounted on the level of 1,36 percent of region’s GDP, according to official data (World Bank Data, 2015). As presented in previous part of this study, in table 1, when three following years are added to the analysis, this ratio is increasing to 1,41 percent. This shows how fast remittances are increasing from year-to-year. Remembering about the size of informal transfers (which are estimated to be the value of 50% of official ones), this high ratio is definitely underestimated.
14
1.3 Facts on remittances and migration
Official statistics of Word Bank and United Nation shows the incredible size of migration for African continent. For year 2010, the number of African emigrants, both voluntary emigrants and refugees, was estimated around 30 million, which was giving the size of 3% of total population of this region. This number is however considered to undervalue the scope and the significance of African migration within and outside the continent, because the movement of people who abandon their countries in fear of conflicts without reporting it is very high and not well reported (World Bank Data 2014).
Crossing the border in order to find a job can be divided for short-term and long-term periods. The first, can be connected with seasonal agricultural activities in neighboring countries, however if it is on regular basis, can become the second one. Immigrants often decide to work abroad in the response to the change in economic condition of their family, to come back into the financial balance. The other reason is gaining experience and allocating earnings for future investments. Destinations vary between countries and immigrants can decide to move alone, or with the closest relatives, heading towards neighboring country or, deciding to settle in the other continent – Europe, Asia or even America (Shaw, 2007).
15
Figure 2 Rate of migration, source: Ratha et al. , 2011
Data on migration from surveys in Burkina Faso, Ghana, Nigeria and Senegal gives a sample indicating that migrants from middle-income countries are more willing to migrate outside the continent, where emigrants form less developed countries usually cross the neighbor’s boarder. The choice of the destination country is affecting the financial support inflows to home country (Ratha et al., 2011). That gives the diversification in the size of remittance inflows between countries, even if their migrant stocks are comparable.
When moving to the other country is influencing life of the migrant himself, families and friends left behind usually experience support coming from abroad which is changing their lives as well. Remittances are send back to the home-country, and support loved ones significantly, where in some areas, receivers highly depend economically on this inflows (Adams & Page 2003, and many others).
16
migrant were increasing, with only one downfall in 2010, to rise once again in the end of 2013. Annually, an average migrant from African country, was sending between 680 USD and 880 USD in years 1990-2013. According to World Bank (2015), this is the size of the average annual household income for other, the poorest African countries like Congo, Liberia or Malawi. For Malawi, in 2013, an average migrant was sending over 100 USD annually home, which is worth almost 2-month salary in this country.
Average size of remittances annually sent home for whole sample is driven up by Nigeria, which has outstanding ratio of over 20 000 USD per migrant in last years. Such significant increase of this ratio, after 2000’s can be explained with improved data collecting system on remittances for Nigeria. The time Nigerian emigrants started to send more, or the records were taken more properly, Mauritius and Botswana have recorded downfalls in this matter. For these two countries, migrant stock abroad increased, but the flow of remittances decreased within last years. For Botswana, the flow of remittances was fluctuating significantly in last three decades, probably due to poor institution system that was recording the flows and worse economic performance of South Africa, which is the main migrant destination.
For all other countries in the sample, remittances per migrant were increasing with time. For Egypt, the second biggest (after Nigeria) receiver of remittances on African continent in last years, average migrant was sending annually over 5000 USD home, which is equal to more than 4 months average salary in this country. For every other country, the ratio is comparable at least with one month average salary. This is giving tremendous opportunities for families receiving remittances, if invested properly. The blessing and help coming together with foreign transfers can transform into the curse very easily. When families taste regular extra cash injections, they are tempted to spend more on luxurious goods, leading to the danger of Dutch Disease (Acosta et al. 2009).
18
The enormously high inflows of foreign exchange earnings can result in decrease in the international competitiveness of the economy of a country, causing Dutch Disease, for example by the appreciation of real exchange rate. This can lead to the decrease of the production of tradable goods and other manufactured products in a country (Acosta et al., 2009). The effect of Dutch Disease caused by remittances was studied in the various regions, giving not clear general answers (Acosta et al., 2009; Loser et al., 2006; Chowdhury et al., 2012 and more). Some of the research suggest growth decreasing effects of remittances by appreciation of real exchange rate, but in some cases remittances do not influence this rate significantly and are enhancing the economy.
1.4 Dutch disease
The name “Dutch Disease” derives from 1970s, when in the Netherlands new gas resources were found and effected in the economic turmoil. The consequences of new resource findings tend to increase the supply of foreign exchange, even on a permanent basis, which leads to the domestic currency appreciation and deterioration of trade in other goods than the “new” resource (Loser et al, 2006). The term is applied also to remittances, as the new source of foreign exchange inflows, when it occurs in significant amounts.
19
Governments by proper policy implementation can accelerate and manage the process of reallocation, encouraging mobility of production factors between sectors (Bourdet & Falck, 2007). Remittances influence the labor supply in general via income change. This is resulting in increasing reservation wage for remittance recipients, which is possible to lower labor supply. Therefore, lower labor supply heighten wages (so as well the price of final output of tradable goods), increasing costs of production and leading to shrink of tradable sector (Acosta et al.,2009; Hanson, 2007).
The long-run effect of remittance can be growth-reducing, because of lower willingness of labor participation of remittance receivers ( Bouhga-Hagbe, 2004). However, when long-run effect is considered in terms of real exchange rate appreciation and diminished international competitiveness, the mechanism should work in the opposite direction of spending effect, increasing capital accumulation and domestic savings. The final result of remittances would depend then on country specific characteristics, like policy implemented to enhance the growth. The effects of remittances vary between countries, leading to Dutch Disease or economic growth (Bourdet & Falck, 2007).
2. Model and empirical analysis
This section is introducing empirical test on the influence of remittances across African countries. With the focus on African continent, thirteen countries are being analyzed in pursuing of Dutch Disease symptoms, mostly (11) from Sub-Saharan region (Botswana, Ethiopia, Ghana, Kenya, Malawi, Mauritius, Nigeria, Senegal, South Africa and Zambia) and two North African countries of Egypt and Morocco. This selection of countries is based on limited data availability for economies on African continent. However, for these countries all data required for the research are provided by GGDC-10 Sector Database combined with World Bank Development Indicators.
20
to risk of lower international competitiveness by resource movement effect (RME) and its force of labor shifts from tradable to non-tradable sectors of the economy.
The expected is to find the evidence of remittances causing : exchange rate appreciation, increase of inflation and labor shift toward non-tradable sector of the economy. The aim of the empirical test is to find evidence of SE and RME across African continent as previously mentioned, caused by flows of received remittances. This would support the hypothesis that remittances can lead to Dutch Disease. To achieve it, cross-country regression for 13 countries is estimated, using panel data for 1980-2010 period.
This chapter provides information about models used in the empirical analysis. Following described models and methods, data used is discussed. Before reaching the ending point of presenting the empirical results, diagnostic checks are performed together with robustness checks.
2.1 Model
To perform the analysis of the influence of remittances on African economies, several regressions are performed in direction to capture the Dutch Disease mechanisms. The most important for this research is to find SE and RME, for country “i” at time “t”. This is possible since the annual time frame is used in panel data, described more in detail in further sections of this chapter. Expected is to find the evidence of Real Exchange Rate Appreciation, increase of inflation and the expansion of non-tradable sector, caused by remittances. The model is in line with literature combining the analysis of Dutch Disease symptoms caused by resource abundance (Papyrakis & Raveh, 2013) and remittances (Acosta et al. 2009). The study of Papyrakis and Raveh (2013), is indicating how inflation and resource movement to non-tradable sector are affected by resource abundance. The complementation of their study is using REER indicator, which is the common tool for testing Dutch Disease caused by remittances (Acosta et al. 2009).
21
possible impact is supported with empirical evidence. The economic characteristics for African countries are diverse, to check it, controlling for fixed effects was added to equations.
When To find county-specific effects of remittances further analysis was undertaken. Countries are analyzed in form of dummy variables, finding fixed effects and differences among the sample in equations 4-6.
(1) ln(𝑅𝐸𝐸𝑅)𝑡,𝑖= 𝛼0+ 𝛼1ln(𝑅𝑒𝑚𝑖𝑡𝑡𝑎𝑛𝑐𝑒𝑠)𝑖,𝑡 + 𝛼2𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛𝑖,𝑡+ 𝛼3ln(𝐿𝑎𝑏𝑜𝑟𝑀𝑜𝑣𝑒𝑚𝑒𝑛𝑡)𝑖,𝑡+ 𝜀𝑖,𝑡 (2) 𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛𝑡,𝑖= 𝛾0+ 𝛾1ln(𝑅𝑒𝑚𝑖𝑡𝑡𝑎𝑛𝑐𝑒𝑠)𝑖,𝑡+ 𝛾2ln(𝑅𝐸𝐸𝑅)𝑖,𝑡+ 𝛾3ln(𝐿𝑎𝑏𝑜𝑟𝑀𝑜𝑣𝑒𝑚𝑒𝑛𝑡)𝑖,𝑡+ 𝜀𝑖,𝑡 (3) ln(𝐿𝑎𝑏𝑜𝑟𝑀𝑜𝑣𝑒𝑚𝑒𝑛𝑡)𝑡,𝑖= 𝛽0+ 𝛽1ln(𝑅𝑒𝑚𝑖𝑡𝑡𝑎𝑛𝑐𝑒𝑠)𝑖,𝑡+ 𝛽3𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛𝑖,𝑡+𝛽4ln(𝑅𝐸𝐸𝑅)𝑖,𝑡+ 𝜀𝑖,𝑡
Resource abundance is represented in the equation in form of the value of remittances inflows to the country “i” in time “t”, which is rather exogenous to changes in other countries in the sample. The effects vary between countries are contemporaneous and heterogeneous, which is captured with country fixed effects.
Equations 1 and 2 focus on Spending Effect of remittances, when the 3rd equation is
supposed to capture the Resource Movement Effect.
After the first overview of interactions between indicators given by equations 1-3, the research takes steps to find effects of remittances in given countries form the sample. To do so, twelve dummy variables are introduced and are interacting independent variables. The interaction indicates how much remittances are influencing specific factor of the economy. The models for these equations look as follows:
22
2.2 Econometric Methods
To tackle the problem analyzed in this research, using OLS method would not bring proper estimation due to residuals’ heteroscedasticity. As the result of this, OLS would give
robust standard errors, with violation of the 3rd Gauss-Markov assumption (all error terms have
the same variance). To diminish the negative impact of not perfect model, the other estimation method was introduced.
Being aware of the general equilibrium nature of interactions within the variables used in the research model, where the same variables are present in each equation, with the different side-appearance, the Seemingly Unrelated Regression (SUR) method was used. This method is supported by the literature, since has been already used in the research seeking Dutch Disease symptoms in resource abundant regions (Papyrakis & Raveh, 2013). The Error terms in three estimated equations, at the same point of time are the examples of contemporaneous correlation. This means that error terms capture similar effects and will be correlated. With this assumption added to the model, additional information about driving forces is included, which makes it more precise than using separate least square estimations for three equations. SUR method fits more to this research purpose. However, the correlation ratios of residuals are still showing the correlation.
2.3 Data
The sample of countries used in the research is giving the opportunity to test the influence of remittances in countries across whole African continent with exception of Central Africa region. The economic analysis of the second biggest continent is challenging not only because of data availability, but also due to the diversity in social, economic and political matters of country level.
23
the estimates for total economy are aggregated at the sector level, so the numbers can vary from national accounts. What is more, the data on value added is given in national currencies. The division of sectors is presented in Table5. (The GGDC database is publicly available via http://www.ggdc.net/dseries/10-sector.html. Data on value added were complemented by gross value added at current prices in dollars.)
Tradable sector Non-tradable sector
Agriculture, hunting, forestry and fishing
Electricity, gas and water, public utilities Construction
Wholesale and retail trade, hotels and restaurants
Mining and quarrying Transport, storage and communication Finance, insurance, real estate and
business services
Manufacturing Community, social and personal services Government services
Table 6 Division of sectors ; source: GGDC 10-sector database, 2015
The measurement of currency overall alignment is made in form of real effective exchange rate (REER) performance. Measuring exchange rate in form of REER shows the average of bilateral real exchange rates between the country and its trading partners. The measure is weighted by the trade shares of trade partners. In the other words, REER can be defined as relative price of traded goods to non-tradables (Montiel, 1999; Montiel & Hinkle, 1999). During analysis of the index, the increase of REER indicate the appreciation of real exchange rate and what is more, the rise of opportunity costs for the production in tradable sector of the economy (Bourdet & Falck, 2008). This appreciation of real exchange rate implies the lowering of international competitiveness, because of the exogeneity of tradable goods prices. The evidence of decreaing REER, would be analogically showing deprecation of the currency and strengthening global position in the matter of competitiveness (Hassan & Holmes, 2012).
24
research, the same time period is analyzed as for the other indicators, which is 1980-2010. The year 2007 is a base year for estimating changes in REER, so the value in this year is 100. The database offers different levels of analysis depending on the number of trading partners taken under consideration during the assessment. The 67-partner dataset is the most accurate for countries in the tested sample and the research is based on these data.
World Development Indicators database, provided by World Bank gives a chance to conduct the study on Remittances and many other issues. Development indicators are coming from renowned, official sources and are presenting updated and accurate data available, on cross-country, cross-regional and global indicators, covering 1960-2014 time period for most of them.
The database is used to possess data on remittances received by a country, in current U.S dollars. Data are based on IMF’s Balance of Payments Manual, form which personal transfers and compensation of employees were taken to prepare the index.
World Bank is the source of data on inflation in this research as well. The research is using the index of inflation based on customer prices annual percentage change. This index reflects the change of costs for the typical consumer, acquiring a basket of goods and services. World Bank uses the Laspeyres formula to assess this dataset.
Sources mentioned above did not give a strongly balanced dataset of 13 countries in a time period of 1980-2010. It was because of data on remittances were missing for four countries (Malawi, Mauritius, Tanzania and Zambia) for some time periods. To balance the dataset, missing data ware estimated by running regressions on the interactions between GDP and remittances for these countries. Estimated missing values together with original data are presented in the Appendix.
2.4 Descriptive
25
The data on remittances, in the sample are present just for 335 observations, with the values between 500 thousand to 19 billion of USD. Due to the missing data on remittances, previously mentioned estimations were made. Since the fact that in the model logarithmic function on remittances is used, data were supplemented by estimated values in accordance to GDP performance. This process is documented in the appendix. Data on remittances were the only not complete for all 403 observations.
Inflation used in the research gives the wide scope of variation within this variable. The minimum recorded value for inflation is equal to (-9,7) , where maximum is equal to (183,3). Data on inflation are presenting annual percentage change in prices of goods, described before.
Real effective exchange rate data in the sample also is characterized with high variation. In this case, the records vary between value of (28,5) to (1957,2).
The last variable used in the regressions is the one responsible to indicate the labor shift to non-tradable sector of the economy. The variable called NONGDP is showing the number of labor force employed in non-tradable sector of the economy, divided by Gross Domestic Product and also contains very diversified scope of records.
Due to the variations mentioned above, for empirical tests logarithmic functions of Remittances, REER and NONGDP indicators were used.
2.5 Econometric Results
26
Table 7. SUR estimation method, first tests for Dutch Disease symptoms in African Economies. Regressions 1-3 without and 4-6 with fixed effects.
Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
Table 7 Regressions 1-6
After introducing country fixed effects to the model, results for the estimations changed significantly. When the impact of remittances on REER remained significant and negative, but with lower coeficient, the results for Inflation lost the significance and changed to be positive now. The impact on labor shift also changed its mark, from positive to the negative one. The SUR analysis showed the power of remittances influencing competitiveness of the economy.
The model with country fixed effects is much better tailored to analyzed data. R-squared statistics improved significantly. This suggests the importance of country fixed effects to be used in the analysis.
(1) (2) (3) (4) (5) (6)
lnREER Inflation lnNONGDP lnREER Inflation lnNONGDP
lnRemit -0.0401*** -1.780*** 0.128*** -0.0225** 0.414 -0.114*** (0.00649) (0.310) (0.0391) (0.00936) (0.453) (0.0324) Inflation -0.00524*** 0.0366*** -0.00131 0.0306*** (0.00105) (0.00603) (0.00108) (0.00360) lnLaborShift 0.123*** 2.426*** 0.167*** 5.489*** (0.00729) (0.399) (0.0129) (0.646) (0.112) (4.848) (0.350) lnREER -11.57*** 4.111*** -2.907 2.062*** (2.322) (0.243) (2.384) (0.160) Constant 6.145*** 114.3*** -27.05*** 5.650*** 36.84** -11.57*** (0.126) (14.53) (1.556) (0.169) (15.72) (1.096) Country fixed
effects used Yes Yes Yes
Observations 403 403 403 403 403 403
27
The results of the executed analysis are showing the diverse influence of tested indicators on macroeconomic performance between countries. For this reason dummy variables for countries were added to the model, showing the interactions between country fixed effects on independent variables. Regressions 7-9 are the base for the analysis and are presented in Table8.
28
Table 8. SUR estimation method, testing for cross-country Dutch Disease effects in African economies with country-specific effects
Dependent variable (7) (8) (9)
lnREER Inflation lnNONGDP
Fixed effect Ethiopia 1.793 -127.1 0.361
(1.994) (221.3) (7.272)
Fixed effect Ghana -6.578*** 95.23 13.58*
(2.276) (223.2) (6.923)
Fixed effect Kenya -1.116 252.3 6.802
(2.433) (233.0) (7.449)
Fixed effect Malawi 0.595 520.1* 9.967
(1.852) (278.0) (9.251)
Fixed effect Mauritius 0.520 155.4 -7.972
(2.579) (319.2) (9.359)
Fixed effect Nigeria -4.058** 67.70 5.880
(1.910) (209.6) (6.740)
Fixed effect Senegal 6.756** 128.4 -6.207
(2.646) (351.7) (8.029)
Fixed effect South Africa -1.252 101.7 -1.509
(2.557) (252.3) (8.475)
Fixed effect Tanzania 3.530* 178.2 -3.628
(1.904) (217.6) (7.266)
Fixed effect Zambia -30.28*** 180.6 157.9***
(7.657) (521.8) (8.223)
Fixed effect Morocco -0.976 78.47 -0.734
(2.868) (300.1) (9.591)
Fixed effect Egypt -1.760 -163.5 21.74**
30 lnLaborShift*Tanzania 0.456*** 7.459 (0.159) (10.96) lnLaborShift*Zambia 0.0129 -1.242 (0.0671) (4.109) lnLaborShift*Morocco 0.327 10.58 (0.324) (21.58) lnLaborShift*Egypt 0.130 6.155 (0.104) (5.291) lnREER*Botswana -4.441 -0.174 (36.84) (1.153) lnREER*Ethiopia 15.02 0.868 (38.00) (1.192) lnREER*Ghana 21.82 1.343 (37.34) (1.165) lnREER*Kenya -24.78 0.606 (41.04) (1.321) lnREER*Malawi -69.66 -0.118 (46.76) (1.526) lnREER*Mauritius -11.08 0.995 (59.18) (1.857) lnREER*Nigeria -11.66 1.259 (37.33) (1.163) lnREER*Senegal -7.675 0.933 (42.52) (1.266) lnREER*South Africa -6.991 2.023 (46.35) (1.309) lnREER*Tanzania -9.518 1.061 (38.06) (1.191) lnREER*Zambia -67.22* 0.835 (38.22) (1.206) lnREER*Morocco -3.868 1.457 (54.78) (1.548) lnREER*Egypt 1.771 0.996 (37.94) (1.187) Constant 5.486*** -1.546 -4.288 (1.800) (205.6) (6.631) Observations 403 403 403 R-squared 0.744 0.484 0.967
Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
Table 8 Regressions 7-9
31
To indicate country-specific effects of remittances on analyzed indicators, estimates responsible for changes in macroeconomic performance are presented below, in Table 9. Shaded cells are indicating the results which were statistically significant.
*** p<0.01, ** p<0.05, * p<0.1
Table 9 Country specific effects of Remittances
The impact of remittances on real exchange rate is significant just for 3 countries in the sample which are Ghana, Nigeria and Zambia. For all these countries remittances are increasing REER, which leads to decrease of international competitiveness due to making export more expensive, diminishing net export.
The increase of 1 percent in remittances received would increase REER in Ghana by 0,3325 percent, in Nigeria by 0,3255 percent and in Zambia by 1,7385 percent. The biggest impact in Zambia can be explained by the size of the economy, which is the smallest from all three countries mentioned. However, remittance inflows are not exceeding 1 percent of GDP for Ghana, suggesting further analysis of this county performance.
The effect on REER is suggesting the occurrence of Dutch disease in the form on Spending Effect. We can see the similar, positive results for other 5 countries in the sample, but the results were not statistically significant.
Influence of lnRemittances - estimated coeficients
id Country lnREER Inflation
32
The case of Nigeria is interesting and worth further analysis. Remittance flows to this economy are recorded on very high levels, in last two decades are oscillating between 2 and 12 percent of the GDP. One should keep in mind that Nigeria is the biggest African economy at this time, with GDP over 500 billion USD in 2013 and still growing (World Bank, 2015).
The independent variable which is representing the labor shift towards non-tradable sector of the economy ( number of labor employed over sector GDP), seems to be negatively influenced by remittances. The results are negative for 12 out of 13 African countries, where the only country with positive (but not statistically significant result) is Mauritius. Mauritian economy is characterized with strong labor shift towards service sector with booming GDP per capita growth a decade ago (Frankel, 2010) with huge migrant stock abroad (around 10 percent of population in last decades) (UN data on migration, 2015).
With the decisive number of countries where the impact of remittances on labor employed in Tradable sector is negative, there is no evidence of Dutch disease in form of Resource Movement Effect for these African countries. Statistically significant results for 6 countries, which are Ghana, Kenya, Malawi, Nigeria, Zambia and Egypt are the prove of this statement. The increase in remittances by 1 percent, causes the decrease of non-tradable sector indicator by ; 0,8554 percent for Ghana ; 0,4974 for Kenya; 0,6284 percent for Malawi; 0,5904 percent for Nigeria; 9,4364 percent for Zambia and 1,1424 percent for Egypt. The highest value of the influence on the economy is once again assigned to Zambia. One should keep in mind that the most of logarithmic values of remittances for Zambia were estimated in this research due to lack of data, so the results for this country should be analyzed carefully.
33
2.6 Robustness check
To check for robustness in the model, data was divided between two time periods, 1980-1994 and 1995-2010. The distinction between these two period is applicable, due to original data availability and the assumption that data before 1994 are not well estimated and collected. What is more, after second half of 1990’s the significant increase of remittances was noticed. The higher values can have much more significant impact on the economies of African countries from the sample. The methodology as used in main models stays unchanged. In this test equations which were the base of regressions 1-6 are used. Results are presented below in Table10 for 1980-1994 and in Table11 for 1995-2010 period.
Table 10. Robustness check, time period 1980-1994
(10) (11) (12) (13) (14) (15)
lnREER Inflation lnNONGDP lnREER Inflation lnNONGDP
lnRemit -0.0410*** -1.767*** 0.0697 -0.0369* 1.335 0.0473 (0.00983) (0.493) (0.0453) (0.0194) (1.075) (0.0375) Inflation -0.00606*** 0.0161** 0.0140*** -0.0295*** (0.00146) (0.00651) (0.00115) (0.00197) lnLaborShift 0.168*** 1.966** 0.489*** -23.94*** (0.0135) (0.794) (0.0247) (1.602) (0.134) (6.814) (0.258) lnREER -14.29*** 3.246*** 41.05*** 1.764*** (3.436) (0.261) (3.369) (0.0890) Constant 6.348*** 129.0*** -20.98*** 6.497*** -269.8*** -11.49*** (0.183) (22.29) (1.768) (0.351) (28.76) (0.880)
Fixed effects Yes Yes Yes
Observations 195 195 195 195 195 195
R-squared 0.215 0.034 0.137 0.703 0.507 0.936
Table 10 Regressions 10-15
34
Table11. Robustness check, time period 1995 - 2010
(16) (17) (18) (19) (20) (21)
lnREER Inflation lnNONGDP lnREER Inflation lnNONGDP
lnRemit -0.0125* -1.673*** 0.165** -0.0147 0.524 -0.105*** (0.00660) (0.322) (0.0662) (0.0122) (0.581) (0.0301) Inflation -0.00316** 0.0452*** -0.00106 0.0409*** (0.00132) (0.0132) (0.00145) (0.00304) lnLaborShift 0.0346*** 1.174*** 0.0149 14.06*** (0.00653) (0.342) (0.0269) (1.045) lnREER -8.320** 3.506*** -2.444 0.100 (3.464) (0.663) (3.360) (0.181) Constant 5.096*** 88.15*** -25.33*** 4.955*** 75.19*** -3.548*** (0.139) (18.00) (3.422) (0.221) (19.16) (1.044) Country fixed
effects Yes Yes Yes
Observations 221 221 221 221 221 221
R-squared 0.003 0.092 0.007 0.390 0.447 0.955
Table 11 Regressions 16-21
Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
The results of regressions 10-21 shows how the model is sensitive for changes. In case of both, new samples, R-squared indicator significantly improves, showing that country distinction should be used in the model. However, what is the most important, the signs of the main coefficients are changing with implemented changes.
As suspected, results for both sub-periods differ comparing to each other. The biggest difference is in the estimated coefficients for Labor shift. While in 1980-1994 period the coefficient is positive but not statistically significant, in 1995-2010 period it is negative and statistically significant at 1 percent significance level. What is more, the more recent results are closer to the results estimated in the main model of the research but less significant.
35
3. Conclusions and Discussion
Africa as the second largest continent is usually forgotten and left by itself. The variety of countries gives almost the unlimited diversification of factors influencing the prosperity and well-being or the opposite, failure and misery of African economies. In recent years part of Sub-Saharan region got back on tracks of economic growth, attracting the international attention. One of the factors which cannot be omitted analyzing the economy of the region are huge amounts of international remittances, inflowing to support local families.
Remittances are the significant part of the economy for many countries. For some the inflows are exceeding even 10 percent of GDP value and with increasing migration ratios and still not reliable estimates of remittances for most of African countries, remittances are expected to play even more important role in near future.
The question is if remittances are the blessing source of welfare and new steam-power for the economy, or the growth-reducing and international competitiveness-diminishing curse for receiving countries. This research shows how uneasy this question is to answer, giving the balanced outcomes, still left for further discussion.
Searching for the evidence of Dutch Disease for 13 African countries, the results vary not only between countries, but also between potential symptoms.
The statistically significant impact of remittances for specific countries in the sample has been found on all three tested indicators (REER, inflation and labor shift). However, the results are not straightforward in answering the research question. Spending effect, enhanced by remittances, has been found in the analysis of REER for Nigeria, Ghana and Zambia, suggesting Dutch Disease. The same factor is not showing the Resource Movement Effect for Ghana and Zambia which would confirm the Dutch Disease in both countries. What is more, remittances seem to reduce their non-tradable sector of the economy, which would be the opposite to the expected outcome.
36
significant and inflation-decreasing and together with negative estimates for labor shift, showed no evidence of the remittance’s curse in this country.
All in all, there is no strong evidence of received remittances diminishing international competitiveness of African countries from the analyzed sample. The final statement can be even more pointing the positive effects that remittances are bringing for the economy or balanced effects. The research was based on finding growth-diminishing symptoms for economies, but now it is clear, that it was not enough. The further research on combining possible Dutch Disease appearance and growth enhancing proxy would be strongly recommended.
The robustness of results can be the limitation of this research. The model is sensitive for implementing changes. Even after introducing changes in the model, using logarithmic functions of variables, presence of country-specific dummy variables and using the Seemingly Unrelated Regression (SUR) method. However, R-squared values for performed regressions are satisfyingly high, indicating well fitted data to the econometric model.
Comparing a group of African economies is not an easy task, knowing all variations in the economic and political environment occurring in the region. More detailed models can bring better understating of the process how remittances are influencing the economy in this growth-promising region.
Sending remittances home is the characteristic of migrants from developing countries. The global migration is increasing, leading to the increase of remittance flows. The influence and effects of this phenomenon have to be tested better to understand the growth possibilities
not only for Africa, but also other regions. With the annual value of remittances sometimes
37
References
Abdih, Yasser & Chami, Ralph & Dagher, Jihad & Montiel, Peter, 2012."Remittances and Institutions: Are Remittances a Curse?," World Development, Elsevier, vol. 40(4), pages 657-666.
Acosta, Pablo A. & Lartey, Emmanuel K.K. & Mandelman, Federico S., 2009, "Remittances and the Dutch disease," Journal of International Economics, Elsevier, vol. 79(1), pages 102-116, September.
Adams, R. 2005. ‘‘Remittances, Household Expenditure and Investment in Guatemala.’’ Policy Research Working Paper Series no. 3532, Washington, DC: The World Bank
Adams, R. and J. Page, 2003, “The Impact of International Migration and Remittances on Poverty,” paper prepared for DFID/World Bank Conference on Migrant Remittances, London, October 2003, (Washington: World Bank).
Aga, Gemechu Ayana & Soledad Martinez Peria, Maria, 2014. "International remittances and financial inclusion in Sub-Saharan Africa," Policy Research Working Paper Series 6991, The World Bank.
Aggarwal, R., A. Demirguc-Kunt, and M. Peria. 2006. ‘‘Do Workers’ Remittances Promote Financial Development?’’ World Bank Policy Research Working Paper no. 3957, Washington, DC: The World Bank.
Amuedo-Dorantes & Pozo, 2006. "Remittances as insurance: evidence from Mexican immigrants," Journal of Population Economics, Springer, vol. 19(2), pages 227-254, June.
Barajas, A., R. Chami, C. Fullenkamp, M. Gapen, and P. Montiel. 2010. Do Workers’ Remittances Promote Economic Growth? Washington, DC: IMF
Beck, T., A. Demirgiic-Kunt, and R. Levine. 2004. ‘‘Finance, Inequality and Poverty: Cross-Country Evidence.’’ NBER Working Paper no. 10979, National Bureau of Economic Research: Cambridge, MA 02138.
Bourdet, Yves & Hans Falck, 2006. "Emigrants' remittances and Dutch Disease in Cape Verde," International Economic Journal, Taylor & Francis Journals, vol. 20(3), pages 267-284.
Chami et al.Macroeconomic consequences of Remittances, 2008, Occasionally paper 259, IMF, Washington.
Chowdhury, M. 2011. ‘‘Remittances Flow and Financial Development in Bangladesh.’’ Economic Modelling 28 (6): 2600–8.
Chowdhury, Mamta B. & Fazle Rabbi, 2014. "Workers' remittances and Dutch Disease in Bangladesh," The Journal of International Trade & Economic Development, Taylor & Francis Journals, vol. 23(4), pages 455-475, June.
38
Docquier, Frédéric, Rapoport, Hillel & Salomone, Sara, 2011. "Remittances, Migrants' Education and Immigration Policy: Theory and Evidence from Bilateral Data"IZA Discussion Papers 6104, Institute for the Study of Labor (IZA)
Durand, J., E.A. Parrado, and D.S. Massey, 1996, “Migradollars and Development: A Reconsideration of the Mexican Case,” International Migration Review, Vol. 30, pp. 423–44.
Emmanuel K. K. Lartey & Federico S. Mandelman & Pablo A. Acosta, 2012. "Remittances, Exchange Rate Regimes and the Dutch Disease: A Panel Data Analysis," Review of International Economics, Wiley Blackwell, vol. 20(2), pages 377-395, 05.
Findley, Sally E. 2004. “Mali: Seeking Opportunity Abroad.” Migration Information Source. Migration Policy Institute. Washington, D.C.
Frankel, J. A., 2010. "Mauritius: African Success Story," NBER Working Papers 16569, National Bureau of Economic Research, Inc
Giulia, B., & Zazzaro,A.(2011). Remittances and financial development: Substitutes or compliments to economic growth. Bulletin of Economic Research, doi:10.1111/j. 1467-8586.2011.00398
Giuliano, Paola & Ruiz-Arranz, Marta, 2009, "Remittances, financial development, and growth," Journal of Development Economics, Elsevier, vol. 90(1), pages 144-152, September.
Hanson, G., 2007. Emigration, remittances, and labor force participation in Mexico. Integration and Trade Journal 27, 73–103.
Hassan, Gazi & Holmes, Mark, 2012. "Remittances and the real effective exchange rate"MPRA Paper 40084, University Library of Munich, Germany.
Johnson, G.E., and W.E. Whitelaw, 1974, “Urban-Rural Income Transfers in Kenya: An Estimated-Remittances Function,” Economic Development and Cultural Change, Vol. 22, pp. 473-79.
Keynes, J.M. 1929. ‘‘The German Transfer Problem.’’ Economic Journal 39: 1–17.
Loser, C., Lockwood, C. Minston, A. and Balcazar, L. (2006), "The Macro-Economic Impact of Remittances in Latin America: Dutch Disease or Latin Cure?" mimeo, Inter-American Dialogue.
Lucas, Robert E.B., and Oded Stark, 1988, “Migration, Remittances and the Family,” Economic Development and Cultural Chang, Vol. 36, pp. 465–81.
Lundhal, Mats (2003). “In the shadow of South Africa : Lesotho's economic future”, The Making of Modern Africa Series, Ashgate, Aldershot.
Maimbo, M.S., and D. Ratha. 2005. Remittances: Development Impact and Future Prospects. Washington, DC: The World Bank.
39
Montiel, P. and Hinkle, L. (1999) “Exchange rate misalignment: an overview”, in Hinkle, L. and Montiel, P. (Eds) Exchange Rate Misalignment: Concept and Measurement for Developing Countries. World Bank Research Publication, Oxford University Press, New York, NY, pp. 264–63.
Nyamongo, Misati, Kipyegon and Ndirangu, Remittances, financial development and economic growth in Africa, Journal of Economics and Business, Volume 64, Issue 3, May–June
2012, Pages 240-260, ISSN 0148-6195,
http://dx.doi.org/10.1016/j.jeconbus.2012.01.001.
Obeirai, A.S., and H.K.M. Singh, 1980, “Migration, Remittances and Rural Development: Findings of a Case Study in the Indian Punjab,” International Labour Review, Vol. 119, pp. 229–41.
Ohlin, B. 1929. ‘‘The Reparation Problem: A Discussion.’’ Economic Journal 39: 172–83. Owusu-Sekyere Emmanuel & Francis M. Kemegue & Renee van Eyden &, 2011."Remittances
and the Dutch disease in Sub-Saharan Africa. A Dynamic Panel Approach,"Working Papers 259, Economic Research Southern Africa.
Papyrakis, E. Raveh O. (2013). An empirical analysis of a regional Dutch Disease: The case of Canada. Oxford : Oxford Centre for the Analysis of Resource Rich Economies.
Rao, B. B., & Hassan, G. M. (2011). A panel data analysis of the growth effects of remittances. Economic modelling, 28, 701–709
Ratha Dilip, Sanket Mohapatra, Caglar Ozden, Sonia Plaza, William Shaw & Abebe Shimeles, 2011."Leveraging Migration for Africa : Remittances, Skills, and Investments," World Bank Publications, The World Bank, number 2300, 6.
Roxburgh C. and others, , 2010, “Lions on the move :The progress and potential of African
economies”, McKinsey Global Institute. Available on
http://www.mckinsey.com/insights/africa/lions_on_the_move
Schwab Klaus, 2014. “The Global Competitiveness Report 2014–2015: Full Data Edition”, World Economic Forum within the framework of The Global Competitiveness and Benchmarking Network.
Shaw, W. 2007, Migration in Africa: A Review of the Economic Literature on International Migration in 10 Countries. Development Prospects Group, World Bank: Washington, D.C.
Singer, D. A. (2008). Migrant Remittances, Financial Globalization, and Exchange Rate Regimes in the Developing World. Massachusetts Institute of Technology, Cambridge, MA.
Sofranko, A.J., and K. Idris, 1999, “Use of Overseas Migrants’ Remittances to the Extended Family for Business Investment: A Research Note,” Rural Sociology, Vol. 64, pp. 464– 81.
40
Svirydzenka Katsiaryna, Mr. Martin Petri, 2014. “Mauritius The Drivers of Growth—Can the Past be Extended?” IMF Working Papers 14/134 International Monetary Fund.
van der Ploeg F (2011) Natural resources: curse or blessing? Journal of Economic Literature 49(2):366–420
Wignaraja, G and O’Neil S (1999) ‘SME Exports and Public Policies in Mauritius’, Commonwealth Trade and Enterprise Paper No 1 London: Commonwealth Secretariat, London
Woodruff, C., and R. Zenteno. 2001. ‘‘Remittances and Microenterprises in Mexico.’’ Graduate School of International Relations and Pacific Studies Working Paper no. 43, University of California, San Diego.
World Bank (2013). Developing Countries to Receive Over $410 Billion in Remittances in
2013, Says World Bank. WASHINGTON, 2013 Retrieved from
http://www.worldbank.org/en/news/press-release/2013/10/02/developing-countries-remittances-2013-world-bank
World Bank. (2003). Workers’ remittances – an important and stable source of external development finance. Global development finance 2003. Washington, DC: World Bank. World Bank. (2004). Harnessing cyclical gains for development. Global development finance
2004. Washington, DC: World Bank.
World Bank. (2011). Migration and remittances factbook 2010. Washington: World Bank. Yang Dean, 2011. "Migrant Remittances," Journal of Economic Perspectives, American
Economic Association, vol. 25(3), pages 129-52, Summer.
Yang, D. 2005. ‘‘International Migration, Human Capital, and Entrepreneurship: Evidence from Philippine Migrants’ Exchange Rate Shocks.’’ World Bank Policy Research Working Paper no. 3578, The World Bank.
41
Appendix
1. The estimation of missing variables for logarithmic value of Remittances was based on
data available, and the correlation with variable on Gross Domestic Product. The regressions were run for four countries where data were missing, to estimate coefficients needed. Estimated values gave the origin of completed data used in the research models. Missing values were replaced. Tables below show all the values.
Mauritius Malawi
year lnRemit lnGDP estimated
