The impact of the brain drain on GDP per capita : a case study of Ethiopia

(1)

The impact of the brain

drain on GDP per capita

A case study of Ethiopia

Emilie Berkhout

10155120

Bachelor Thesis

Specialization: Economics

Supervisor: I. Rozentale MSc

June 2014

(2)

Abstract

The impact of the brain drain on the GDP per capita of Ethiopia is investigated with a case study, analyzing scientific literature and secondary data. In the literature on the magnitude of the brain drain it was found that, besides push factors in the source country, wage gaps between source and destination countries and higher relative earnings to skills in the

destination country increase the brain drain. The data on Ethiopia confirm these findings. For the impact of the brain drain on GDP per capita, the literature states that the brain drain creates an incentive effect, expanding the level of human capital if high-skilled emigration is between 20% and 30%. This increases GDP per capita. The case of Ethiopia shows that the incentive effect is too small, because the high-skilled emigration rate is below 20%. It cannot be concluded from the literature or the data whether skilled migrants remit more, so it does not have a clear effect on GDP per capita. Lack of returns for the investments in education due to lost taxes decreases the welfare in Ethiopia, although there is no evidence on the extent of this effect. Concluding, the brain drain has a negative effect on the GDP per capita of Ethiopia, but it is not large enough to create a problem for the educational investments of the government.

(5)

1. Introduction

1.1 Brain drain

The brain drain, or human capital flight, can be seen as one of the major aspects of globalization (Docquier & Rapoport, 2012; Di Maria & Lazarova, 2012). The proportion of emigrants that are highly skilled has increased extensively (Di Maria & Lazarova, 2012; Batista, Lacuesta & Vicente, 2012). A research by Docquier and Marfouk (2006) estimated that the number of tertiary educated workers that live in another OECD country than their country of origin had increased with 63.7% between 1990 and 2000, while this was only 14.4% for unskilled migrants (Di Maria & Lazarova, 2012; Batista et al., 2012).

Scientific research shows different views on the impact of the brain drain (Aredo, 2000). On the one hand, the main argument is that human capital is necessary for sustained economic growth, while developing countries lack human capital. The small amount they have, educated using scarce resources, emigrates to developed countries with better job opportunities. This holds back economic growth (Cattaneo, 2009). On the other hand, these skilled people earn more money in the developed countries than in their home countries, which they can send to their families as remittances. Also, the prospect of emigration may create a higher incentive for tertiary education. These arguments, together with the potential of the emigrants returning home after acquiring even more skills, states that there can be a brain gain and therefore a positive impact on economic growth. At this moment, there is still no generally accepted view on the brain drain and more research is being done. There are theoretical as well as empirical findings for both positive and negative effects of human capital flight on economic growth.

1.2 The case of Ethiopia

Among all country groups, Sub-Saharan Africa had experienced the largest growth in skilled emigration rates between 1975 and 2000. It grew from 6% to almost 12% and while Central America, South-East Asia, Northern Africa and the Middle East had a higher skilled

emigration rate in 1975, only Central America had a higher rate in 2000 (Beine, Docquier & Oden-Defoort, 2010). Of Sub-Saharan Africa, Ethiopia was the second largest country of source of the migration to the US between 1994 and 20041, which mostly consisted of highly educated individuals (Negash, Anteneh & Watson, 2012). The largest country of source was Nigeria, but what makes Ethiopia more interesting to investigate is that Ethiopia has been dedicated to higher education reforms from 1997 onwards in order to increase school enrollment rates and economic growth (Saint, 2004). Nigeria, however, struggles with a

1

The data used here comes from the US Yearbook of Immigration Statistics from 1994 to 2004 5

(6)

collapsing state of education (Odia & Omofonmwan, 2007). The increased tertiary enrollment rates due to the large investments of the Ethiopian government in its educational system may also increase the number of highly educated emigrants (Saint, 2004). Therefore, it is

interesting to investigate whether human capital flight has a positive or negative impact on economic growth in Ethiopia. The large increase in the skilled emigration rate allows for comparing the situation of little skilled emigration with the situation of larger skilled emigration in order to see positive and/or negative consequences.

1.3 The research

For investigating this possible problem, the following research question was stated in this thesis: How does the brain drain affect the GDP per capita of Ethiopia?

In order to answer this question, a case study approach is executed to get an

understanding of the different factors that influence the impact of the brain drain on GDP per capita. The reason for this is mainly because contextual conditions are relevant, since it does not occur in all countries and there are different views on the impact. Two information

sources were used for executing this approach, namely existing scientific literature and secondary data from the Institute for Employment Research (IAB), the OECD and the World Bank.

The next section describes the methodology of the case study. After that, the thesis is divided into two parts: a literature review and the case study of Ethiopia. Literature about the different views on the brain drain is studied, with a closer look at how these views came about. The emphasis is on factors that influence the magnitude of the brain drain and factors that influence the impact of the brain drain on GDP per capita, especially human capital formation and remittances. Important questions are whether the possibility of emigration increases human capital formation and whether high-skilled emigrants remit more than low-skilled emigrants. The third section gives a description of the theoretical findings, after which empirical findings are described in the fourth section. Subsequently, the fifth section

describes data on Ethiopia about the skilled migration, wage gaps, school enrollment rates, received remittances and GDP per capita, together with the political situation and the

educational policy. In the sixth section, the data is combined with the findings in the literature. Finally, it is concluded what the impact of the brain drain on GDP per capita of Ethiopia is and recommendations are made for further research.

2. Methodology

In order to answer the research question and get an understanding of the factors that

influence the impact of the brain drain on GDP per capita in Ethiopia, a case study approach 6

(7)

was chosen using existing literature and secondary data. The reasons for this are described in this chapter.

2.1 The case study approach

A definition for a qualitative case study approach as given by Baxter and Jack (2008) says that it is a research approach that facilitates exploration of a phenomenon within its context using a variety of data sources, such that the phenomenon is explored through different views and that different facets of it are revealed and understood. According to Yin (2003), there are four reasons when this approach should be used, namely when the research question is a ‘how’ or ‘why’ question, when you cannot manipulate the behaviour of those involved in the study, when you want to cover contextual conditions because you believe they are relevant to the phenomenon under study and/or when the boundaries are unclear between the phenomenon and context. The research question is a ‘how’ question, namely ‘how does the brain drain affect the GDP per capita of Ethiopia?’. The behaviour of the migrants cannot be manipulated by this research and contextual conditions are relevant to the brain drain, since not all countries experience a large outflow of human capital or experience the same impact. Because of the different views in the literature on the effect of the brain drain, boundaries between the phenomenon and the context are unclear. Thus, a case study approach for this research is useful.

The intention is to better understand the factors that influence the impact of the brain drain on GDP per capita and the case of Ethiopia in itself is of interest through the increase in the skilled emigration rate and the educational reforms. This is defined by Stake (1995) as an intrinsic case study approach. The used approach matches the definition of Baxter and Jack (2008), since different data sources are used, namely scientific literature and secondary data, and different views are described.

The impact on GDP per capita is assessed, because according to theory this can be seen as an indicator for the level of economic and human development of a country.

Although the indicator is incomplete, Ray (1998) found high correlations between per capita income and other variables that predict human development, even if he restricted the sample to only developing countries.

2.2 The literature

The criteria on which the selection of the literature is based are that it provides theoretical and empirical results on the magnitude of the brain drain and/or on the impact of human capital migration on GDP per capita, with clear distinctions between different influencing factors. All the literature used in the literature review is listed on the Tinbergen Institute Journal List, which lists academically accepted journals, or is cited by that literature. Most of the assessed literature was published after 2000, because skilled emigration data only

(8)

became available by then (Docquier & Rapoport, 2012). This literature is evidence-based and is valued more in the analysis. Other sources are influential institutions, which are the United Nations and the World Bank. Furthermore, articles from scientific journals about education are used.

2.2 Secondary data

With the findings about the relevant influences on the magnitude of the brain drain and economic growth, the case study of Ethiopia is executed. First the degree of the brain drain in Ethiopia is determined with data from the IAB. It contains the number of Ethiopian natives living in OECD countries, aged 25 years and older to filter out students that return after their study, as well as the skilled emigration rates between 1980 and 2010 with five year intervals. Distinction is made between a low, medium and high level of education of the migrants. The low education means primary education or no education, medium education means that the migrant has a high school certificate and tertiary education means a higher than high school certificate. The skilled migration rate is calculated as the proportion of migrants over the pre-migration population, defined as the sum of residents and migrants in each source country, belonging to the same level of education. The collection of the dataset was published in 2013. This is the newest dataset that is currently available (Brücker, Capuano & Marfouk, 2013). It covers 20 OECD destination countries, as shown in Appendix 2.

Second, to explain the movement of the brain drain of the last decades, wage gaps between Ethiopia and the United States (US), Canada, the United Kingdom (UK) and Sweden are examined by comparing the GDP per capita, since these are the largest destination countries. Portugal is one of the smallest destination countries and is used as a control country. The wage gap is seen as one of the main forces that drives human capital flight (Stark et al., 1997; Collier et al., 2004; Bénassy & Brezis, 2012; Docquier & Rapoport, 2012). In addition, the relative earnings to skills in the destination countries, found to be relevant by Grogger and Hanson (2011), are assessed to compare them. Data on this is given by the OECD (2012), only available between 2000 and 2010.

Finally, in order to see consequences of the brain drain, remittances send to Ethiopia are analyzed as well as primary, secondary and tertiary schooling rates, which can show the possible increased incentives for schooling through the possibility of emigration (Di Maria & Lazarova, 2012). They are compared to the skilled emigration and GDP per capita of

Ethiopia. The World Bank provided this data. By means of the literature a conclusion follows on what impact this should have had on GDP per capita. Because no empirical research is executed, it can only be concluded whether the brain drain possibly increases or decreases economic growth in Ethiopia.

(9)

3. Literature Review of the Theory

This chapter first gives a definition of human capital, brain drain and brain gain. After this, the theoretical and empirical findings about the role of human capital in economic growth are discussed. Finally, it gives a summary of the different theoretical views on the role of human capital flight in economic growth.

Human capital means educated labour, such that it has a higher productivity than without education. This includes labour that is skilled in creating or adopting technologies or in executing these technologies (Docquier & Rapoport, 2012). In educational terms this would mean that the workers have tertiary education (Collier et al., 2004).

The brain drain, also known as human capital flight, means that skilled labour emigrates, which is seen as a negative phenomenon for the source country. However, it is also argued by various literature that a brain gain may occur, which means that the

emigration of skilled labour creates positive effects for the source country.

3.1 The role of human capital in economic growth

New growth theories try to explain the role of human capital in technical progress and therefore in long-run per capita growth (Ray, 1998). The basic idea is that households can ‘save’ in two ways: in the form of capital or by investing in education. Mankiw, Romer and Weil (1992) add this idea to the Solow model2 and test it. They find that it explains almost 80% in the cross-country variation of per capita income, while without human capital accumulation the model explained about half of the income variation. It can be concluded that human capital plays a large role in per capita income growth.

In addition, Barro (1991) found that higher endowments of human capital contribute to per capita GDP by running regressions on the average growth on per capita real GDP over the period of 1965 until 1985. By including human capital in the regression, it predicted the growth in GDP per capita better in that period for Sub-Saharan African countries (Ray, 1998).

3.2 Theory about the magnitude of the brain drain

The magnitude of the brain drain is important to assess, because the impact of the brain drain on GDP per capita depends on its magnitude (Beine et al., 2010). The main motives for the brain drain from developing countries can be divided in 4 categories: demand for skilled workers in developed countries, socio-economic imbalances between regions, ineffective science and technology planning by national governments and political, religious and ethnic constraints in developing countries. The most important ones are higher wages, intellectual freedom, better working conditions and better career expectations in the destination

2_{See Appendix 1 for a clarification of the Solow model}

9

(10)

countries, while other factors include political instability and lack of research and other facilities in the source country (Aredo, 2000). More specifically, these push factors include shortfall of complementary facilities needed to carry out a specialized profession and political risks affecting abilities to utilize human capital (Collier et al., 2004). According to theory, the return on human capital should be high in countries with a low level of human capital, ceteris paribus, due to diminishing returns to scale. Apparently, this is not the case. Because GDP per capita also reflects policy effectiveness, human capital flight will decrease with higher GDP per capita to the extent that income reflects a good proxy for policy environment (Collier et al., 2004).

In addition to wage gaps, large relative earnings to skills in the destination country are a pull factor, since the high-skilled workers want a larger compensation for their high

productivity (Grogger & Hanson, 2011).

Collier et al. (2004) analyze human capital emigration with wage gaps, policy

effectiveness, chain migration effects and distance effects as explanations for human capital flight. Chain migration means that former migrants can supply new migrants with information, which facilitates human capital flight. The distance effects account for the costs of migration that increase with distance from the source country to the destination country.

3.3 Summary of the different views on the impact of the brain drain on

economic growth

3.3.1 The 1960s

In the 1960s, free migration of skilled workers was seen as a contribution to human capital as an international public good and the effect on the source country was assumed neutral with compensating remittances for any real loss of the brain drain as explanation (Docquier & Rapoport, 2012). The critique of Bhagwati and Hamada (1974) on this result is that the analysis and prescriptions are constrained by the simple theoretical model that is used by Grubel and Scott (1966), to whom this central result can be attributed. Because of the constraints and because this view does not concentrate on the source country, no more attention is given to this period.

3.3.2 The 1970s and 1980s

During this period, the view was mostly negative under the leadership of Jagdish Bhagwati (Docquier & Rapoport, 2012; Beine et al., 2010; Bénassy & Brezis, 2012). The literature introduced externalities of human capital flight and focused on the negative consequences for the source country (Docquier & Rapoport, 2012; Batista et al., 2012). The theory behind this negative view is that the home country would have had a larger skilled workforce without migration such that per capita output would have been higher (Stark et al., 1997).

(11)

Bhagwati and Hamada (1974) give more theoretical arguments that cause a negative view on the brain drain. If the social marginal product exceeds the private marginal product of the migrant as a result of strong externalities, the home country will experience a loss. This can be the case with emigrating doctors, because a healthy workforce is important for production. The composition of the brain drain goes beyond this research, so no more attention is paid to this. Another argument is that if the government had financed the education of the migrant, it will not get its investment back through taxes. Assuming a progressive tax system, the high-skilled emigrants would have to pay high taxes if they stayed in their home country, but they do not pay this when living in another country.

3.3.3 The 1990s onwards

In this period, the focus lies on the possible brain gain for the source country, recognized by Stark et al. (1997) for the first time (Di Maria & Lazarova, 2012). The argument they give is that the source country may end up with a higher level of human capital due to human capital flight, assuming that higher returns on education in foreign countries increase the incentive for education. Because of, for example, restrictions on emigration, not all the educated migrate (Beine et al., 2008). This consequence of the brain drain is called the incentive effect. Whether this effect increases the level of human capital, increasing GDP per capita as a consequence, depends on whether it is large enough to compensate for the skilled

emigration.

An additional influential factor is remittances sent back to the home country. Educated migrants earn more, but whether skilled migrants remit more is theoretically ambiguous (Catteneo, 2009; Docquier & Rapoport, 2012; Faini, 2007). It is more likely that skilled migrants come from wealthy families that could pay for the education and migration, so that the incentive to remit more is small. However, the family may ask a return for their

investment, increasing remittances (Faini, 2007). Remittances are seen as a stimulus to the economy through the consumption multiplier, so that it also increases welfare for non-migrant households (Catteneo, 2009). In addition, it has been argued that they relax credit

constraints on physical and human capital, which are a problem in many developing

countries (Docquier & Rapoport, 2012). But some arguments state that remittances can have a net negative effect on economic growth. They may decrease the incentive to work if the recipients can live on the amount received, while they may have educated themselves and worked otherwise (Le & Bodman, 2011). The effect of remittances on GDP per capita is therefore ambiguous as well.

Also return migration is an important positive factor of the brain drain, since the migrants may have accumulated additional knowledge and bring back financial capital, facilitating the access to and adoption of technologies (Stark et al., 1997; Docquier &

(12)

Rapoport, 2012). This consequence can also occur through technological diffusion as the migrants keep contact with the home country (Le & Bodman, 2011). However, since data on return migration and technological diffusion are limited, this is not assessed in the case study of Ethiopia.

4. Literature Review of the Empirical Findings

This chapter represents empirical evidence for the theoretical findings on the magnitude of the brain drain, human capital formation, remittances and the impact on GDP per capita and gives an overview of the important factors that influence the impact of the brain drain on GDP per capita.

4.1 Summary of the empirical evidence for the theoretical findings

4.1.1 Evidence on the magnitude of brain drain

Grogger and Hanson (2011) give evidence on the role of wage gaps in the magnitude of the brain drain by using data of migrants to OECD countries of 25 years and older. They found that the gap between high-skilled and low-skilled migrants tends to increase with the skill-related difference in wages between destination and source countries, supporting the theory. Belot and Hatton (2012) find similar results using the same data but a more advanced model. However, the introduction of poverty measures for taking credit constraints on migration into account was needed for this result. Since credit constraints are reasonable for developing countries, this finding can still be used for the case of Ethiopia. In addition, they find that educated migrants are more likely to emigrate to countries where the relative earnings to skills are high.

Beine, Docquier and Özden (2011) use migrant stocks in OECD countries, called diasporas, to investigate chain migration. They find that larger diasporas increase migration flows, but that the new migrants have a lower average educational level. This can be due to family members that are brought to the developed countries. The same result is found by Collier et al. (2004) by looking at the diaspora in the US, coming from 61 developing

countries. By comparing their results on human capital flight with general migration, they find that chain migration is more important to unskilled migration. This makes this determinant uninteresting to look at in the case study.

In addition, poor policy effectiveness, estimated by the Country Policy and

Institutional Assessment (CPIA), significantly increases human capital flight (Collier et al., 2004). This proves the effect of the push factors mentioned in the theory. However, skilled emigration from developing countries is less reactive to geographic variables such as distance (Docquier & Rapoport, 2012).

(13)

4.1.2 Evidence on human capital formation

There is evidence for the incentive effect in both macro- and micro-analyses. Beine, Docquier and Rapoport (2008) confirm the incentive effect with data on 127 developing countries, implying a positive and significant effect of the brain drain on premigration human capital formation through the prospect of emigration. This effect is stronger for countries with a low GDP per capita. The same conclusion was found controlling for whether the education was acquired in the home country or the destination country, using different measures for human capital and using alternative functional forms (Docquier & Rapoport, 2012). Micro-evidence for an incentive effect is found by Batista, Lacuesta and Vicente (2012), investigating the case of Cape Verde. They found that the brain drain not only has a net positive effect on the premigration level of human capital, it is also the main driver of human capital formation in the country.

Whether this effect is large enough to increase the post-migration level of human capital is investigated by Beine, Docquier and Oden-Defoort (2010) with a panel data

analysis, controlling for heterogeneity and endogeneity, The level of human capital increases through the incentive effect, but decreases with skilled emigration. A higher post-migration level of human capital can be realized if the skilled emigration rate is between 20% and 30%, depending on country characteristics. If it is below 20%, the incentive effect is too small to compensate for the emigration, but above 30%, too much of the increased human capital emigrates. As a consequence, the post-migration level of human capital will be lower, decreasing GDP per capita.

4.1.3 Evidence on remittances

The results on the relationship between skilled emigration and remittances are diverse, dependent on the focus and methodology of the study. Although macro-evidence finds that skilled emigrants remit less, micro-evidence suggests the opposite (Docquier & Rapoport, 2012). Both studies have limitations. At macro-level, the research suffers from the fact that the data for migration only contains OECD countries, while the remittances come from all over the world (Faini, 2007). At micro-level, the household surveys of immigrants in 11 destination countries may not represent the size and skill structure of global migration (Bollard, McKenzie, Morten & Rapoport, 2011). Thus, it cannot be concluded whether skilled migrants remit more. So even though there is enough evidence to conclude that remittances significantly increase economic growth (Le & Bodman, 2011), the relationship between skilled emigration and GDP per capita through remittances is ambiguous.

4.1.4 Evidence on the total influence of the brain drain on GDP per capita

Little of the literature focuses on the total impact of the brain drain on economic growth. Most of it only takes human capital formation into account. Even though this is seen as the largest

(14)

channel of the brain drain that affects economic growth, it does not give a complete picture. Therefore, the results should be considered with caution. Cattaneo (2009) does consider all the channels, except for return migration, but she only examines the impact of the brain drain on poverty. She finds that the intensity of the brain drain does not influence the income of the poor, implying that it does not have a significant effect on economic growth. However, this is probably because the poor do not have the assets to get educated and emigrate, so it is not representative for the effect on GDP per capita. By assessing only human capital formation, Di Maria and Lazarova (2012) found that 70% of the total population in their sample of developing countries suffers lower economic growth as a consequence of skilled migration, taking the composition of the migration into account with workers in science and technology in particular. These losses are concentrated in countries with low technological

sophistication. From this it can be concluded that the total impact of the brain drain on GDP per capita is ambiguous and more extensive empirical research is needed.

4.2 Overview of the important factors

Based on the literature review, several factors are discerned that can be seen as important for the analysis of the relationship between the brain drain and GDP per capita and hence will be considered in the case study of Ethiopia.

The most important determinants of the magnitude of the brain drain found in the theory and empirical results are wage gaps and policy effectiveness. The higher the wage gap between the destination and the source country, the wage gap between low and high-skilled workers in the destination country and the poorer the policy effectiveness, the larger the brain drain.

Several determinants of the impact of the brain drain on GDP per capita were found. The brain drain increases the post-migration level of human capital if the skilled emigration rate is between 20% and 30%. A higher level of human capital rises income per capita. It is ambiguous whether skilled migrants remit more, but it is found that remittances do increase GDP per capita. Finally, countries with low technological sophistication experience lower economic growth due to the brain drain.

5. The situation of Ethiopia

In this chapter, the political situation is described, with the emphasis on economic and

educational policy. Furthermore, the chapter describes the data on the brain drain in Ethiopia for the period 1980 to 2010, together with data on the relevant factors.

(15)

5.1 Political situation

By looking at the political situation of Ethiopia, push factors and other developments that have influenced its GDP per capita are found.

While Ethiopia was one of the ten poorest countries in 2011, it was the 12th fastest growing economy in the world in 2012. Despite large population growth, Ethiopia

experienced an average per capita growth in income of 8.3%, outperforming the average of Sub-Saharan Africa of 3.3%. The explanation for this growth is different for Ethiopia than for other Sub-Saharan Africa countries. While the latter grew primarily through private

investment and consumption, approximately two thirds of the extensive growth of Ethiopia can be explained by public investment (Geiger & Moller, 2013).

For almost the whole 20th century, Ethiopia had been ruled by highly centralized governments. In 1991, the Ethiopian People’s Revolutionary Democratic Front (EPRDF) took power and still is the ruling party. It aims at a democratic system of governance and

decentralizing authority. Although the state structure has been transformed, there are still challenges. With the parliamentary elections in 2010, the EPRDF won with 99.6% of the votes (Worldbank, 2013).

The government aims at macroeconomic stability and improving social services and protection through, among other things, increasing the access to education services. Educational quality has already increased and remains an important target of the

government (Worldbank, 2013). For higher education, the government has been engaged in expansion and reforms since 1997 (Saint, 2004). Aredo (2000) found push factors for skilled emigration at higher education institutions in Ethiopia. There was lack of research facilities, administrative support and clear policies. In addition, there was a heavy teaching load for the academic staff and the staff was demoralized due to lack of facilities. As a result of the expansion and reforms, there were eight universities in 2004 instead of two before 2000, which get more funding through block grants. Also private tertiary education has largely expanded. The universities are awarded substantial autonomy to choose their own institutional leaders. Graduate program enrollment has been encouraged to increase the supply of academic staff. However, the supply of academic staff may still be a problem since it should be more than doubled with the increase in universities. For boosting the quality of the academic program, new oversight agencies will monitor it. Before 2000, there was only a small supervisory department in the Ministry of Education, but this has been extended to three oversight agencies. Since international development assistance for tertiary education has been minimal, the government bears most of the costs (Saint, 2004).

However, there is still no complete intellectual freedom and freedom of speech. On April 30th 2014, demonstrating students got killed by security forces of the government for expressing their opinion on the plans of the government to expand the boundaries of the

(16)

capital Addis Ababa (Human Rights Watch, 2014).

The efforts of the government are intended to increase the enrollment rates in order to increase the supply of high-skilled workers. To keep this supply high, it is thought that the government needs to retain and effectively use the highly educated workers to avoid

emigration (Saint, 2004). A lot of skilled migrants are students who were sent abroad during their studies, but never returned. Therefore, the Addis Ababa University, the largest

university in Ethiopia, decided to build a PhD program in Information Systems in 2012 in order to prepare the students for a career in Ethiopia itself. The assumption is that if they get their whole education in Ethiopia, the probability that they will stay there is higher (Negash et al., 2012). The labour market for university graduates is limited due to the large dependency on agriculture (Saint, 2004). However, the service sector in Ethiopia is rapidly growing and explains most of the economic growth (Geiger & Moller, 2013). This creates employment opportunities for the highly educated.

The following sub-chapters describe the used data, which can be found in Appendices 2 to 5.

5.2 Emigration rates and number of migrants

In Appendix 2, Table 1 presents the emigration from Ethiopia to different destination countries, categorized by educational level. The data indicate that high-skilled emigration was 23.56 times larger in 2010 than in 1980. Table 2 shows that the emigration rate had increased from 4% in 1980 to 16,92% in 2010. Comparing this to the world average, the world high-skilled emigration rate increased from 4.94% in 1980 to 5.31% in 2010. So, while this rate did not differ much from the world average in 1980, it is more than three times higher in 2010. The low and medium skilled emigration rates only increased slightly, from 0.03% to 0.13% and from 0.26% to 1.14% respectively. Both had been lower than the world average for the whole period. Between 1990 and 2000, the high-skilled emigration rate had been approximately stable around 11%, with even a small decrease between 1990 and 1995. The absolute number of high-skilled migrants did increase, so this indicates that the pre-migration high-skilled workforce must have increased more. The highest increase in the high-skilled emigration rate in percentage points is found between 2000 and 2005. Although Ethiopia is the second largest source country for high-skilled migrants in the US, the general high-skilled emigration rate was not excessively high. According to Beine et al. (2008), the skilled migration rate is high if it exceeds 20%.

The US had been the largest destination country for high-skilled emigrants throughout the whole period. Canada comes in second place and the UK is third. However, Sweden comes third in total migration and the UK sixth. This makes it interesting to compare the UK to Sweden for analyzing the magnitude of the brain drain, since the finding suggests that there is a large incentive for highly educated workers to emigrate to the UK. For the UK, it

(17)

can be seen that the skilled emigration from Ethiopia increased after 1995. The proportion of high-skilled migrants in total migration more than doubled between 1995 and 2010.

The three countries with the least stock of Ethiopian migrants are Austria,

Luxembourg and Portugal. The United States and Canada are very large countries, which is likely to influence the number of migrants. To compare, Luxembourg is the smallest country in the sample and faces, after Austria, the smallest stock of Ethiopian migrants. The data show that there are no Ethiopian migrants in Austria. Since this is more likely to be due to a strict immigration policy or incomplete information, Portugal will be used as a control country when comparing wage gaps.

5.3 Comparing GDP per capita

The wage gaps between Ethiopia and the US, Canada, Sweden, the UK and Portugal are assessed with the GDP per capita, since there is no data available on skilled wages in Ethiopia. Although this does not represent the skilled wage gap, it gives an indication of the level of development of the country. It is likely that high-educated emigrants move to the most developed countries, because the demand for them is relatively high there (Aredo, 2000), indicating high skilled wages. As shown in Table 3, the difference in GDP per capita with the US had been the highest throughout the whole period, while the difference with Portugal had been around half of the difference with the other destination countries. The GDP per capita gap with the UK was the lowest of the most popular destination countries in 1981, but increased and has been higher than Canada’s since 1993. The wage gap with Sweden has been the second largest throughout the whole period, but the difference with the UK decreased since 1993.

As shown in Table 4, the wage gap between tertiary and below upper secondary educated workers is the largest for Portugal and the US. The UK comes third, while Sweden has the lowest.

5.4 Public spending on education and school enrollment rates

Table 6 indicates that public spending as a percentage of GDP as well as a percentage of government expenditures has practically doubled between 1980 and 2010. Public spending as a percentage of GDP increased only 0.65 percentage points until 1997, after which it increased with 2.7 percentage points until 2008. As a percentage of government

expenditures, public spending on education experienced its largest increase from 14.12% to 24.35% between 2000 and 2010. Data on public spending per student is only available for 2010. For primary, secondary and tertiary educated, it was respectively 19.17%, 10.39% and 24.64%. These findings demonstrate the intentions of the government described earlier in this chapter.

The payoff of these expenditures is reflected in the school enrollment rates in Table 5. 17

(18)

The gross primary enrollment rate fluctuated between 20% and 40% in the period of 1980 to 1997. Subsequently, it gradually increased to 79.23% in 2005. The net primary enrollment rate increased the most after 1997. The gross secondary enrollment rate started growing above the average of the former period after 2000 from 13.64% in 2000 to 24.95% in 2010. The net secondary enrollment rate is only available for 1999 till 2002. It shows a gradual increase, but cannot be compared to the period before 1997. For tertiary education, only the gross enrollment rate is available. The policy of the government is seen as well. Between 1997 and 2005, the tertiary enrollment rate grew 3.4 times larger. According to Saint (2004), the enrollment growth rate was possibly the highest in the world during the period of 1997 to 2003. Before that, the tertiary enrollment rate increased with 0.3 percentage point after 1984 and decreased with 0.2 percentage point between 1991 and 1994.

5.5 Development of received remittances and GDP per capita

In Table 7, it is seen that received remittances as a percentage of GDP were the lowest between 1987 and 1991. Since this is calculated by dividing the amount of remittances in current US dollars by GDP in current US dollars, the exchange rate bias is ruled out and the numbers can be compared. Periods of large increase are 1997 till 2000 with a growth from 0.11% to 0.66% and, after a drop in 2001, the largest increase is found between 2001 and 2007 from 0.23% to 1.89%. Still, these numbers are relatively low.

The data on GDP per capita in constant 2005 US dollars, shown in Table 8, confirm the large growth as mentioned in the situation of Ethiopia. By assessing the GDP per capita with 2005 as a base year, comparisons between years can be made without an exchange rate bias. From 2004, the GDP per capita grew with an average of 8.3% annually. Before that, it experienced alternate periods of growth and decline. The largest decreases can be found between 1983 and 1985 with annually 9.9% on average and between 1987 and 1992 with annually 5.8% on average.

6. Consequences of the brain drain in Ethiopia

By means of the findings in the literature and the data, consequences for the GDP per capita in Ethiopia are assessed.

6.1 The magnitude of the brain drain in Ethiopia

Low investment in education is seen as an important push factor (Collier et al., 2004). It is expected that, with the investments of the government since 1997, the impact of this push factor is diminished. The data indeed show that public spending on education increased the most after 1997. Also, in 2010, public spending per student as a percentage of GDP per capita was the highest for tertiary education. In addition, the increased employment

(19)

opportunities for the highly educated through the transition from agriculture to services in the period of high growth and the measures by universities to keep students in Ethiopia are expected to decrease the brain drain.

But, the data show that the human capital flight did not decrease, it even experienced its largest expansion in percentage points between 2000 and 2005. Different reasons for this are found in the analysis. The political situation of Ethiopia is a push factor for the high-skilled workers. Although the current ruling party claims to aim at a democratic state, there is still no freedom of speech, as shown by the killed demonstrating students. This is an important push factor mentioned by Aredo (2000) and Collier et al. (2004). Another explanation can be the large growth in gross tertiary enrollment from 1.19% to 2.79% and the slack high-skilled labour market (Saint, 2004). Since Ethiopia is still very dependent on agriculture, despite the transition to the service sector, the labour market for high-skilled workers is limited. The small decrease in both tertiary enrollment and the high-skilled emigration rate between 1990 and 1995 also supports this.

Another explanation are wage gaps, which is empirically proven by Grogger and Hanson (2011) and Belot and Hatton (2012) to influence the magnitude of the brain drain. It is found that the GDP per capita gap for the main destination countries is approximately two times larger than the gap with Portugal, which gives an explanation for the low human capital flight to Portugal. A more specific finding is that the GDP per capita gap with the UK started to increase faster after 1993 and between 1995 and 2010, the proportion of high-skilled migrants in total migration increased from 25.8% to 57.7%. Nevertheless, while Canada is the second largest destination country, it has the lowest wage gap apart from Portugal. The high migration to Canada is explained by the diplomatic relations from 1965 onwards (Government of Canada, 2014). The Ethiopian Embassy in Canada opened in 1989, after which a large increase in total migration is found compared to the period before 1989. In addition, the US have had diplomatic relations with Ethiopia since 1903 (Embassy of Ethiopia). The three largest destination countries for high-skilled workers are English-speaking countries. This represents another pull factor in the destination countries.

The high relative earnings to skills in the US and the UK give an explanation for the highest human capital flight to the US and the higher high-skilled emigration to the UK than to Sweden, which has the lowest relative earnings to skills. This supports the findings of Grogger and Hanson (2011). However, the data contradicts with the low emigration to Portugal and the high emigration to Canada. The explanation for Canada is already given. A possible explanation for Portugal is that high-skilled migrants value the international wage gap more. Thus, high relative earnings to skills are an important motivation for skilled emigration, but the difference in GDP per capita seems to be more important.

To sum up, the magnitude of human capital flight from Ethiopia depends mostly on 19

(20)

the slack high-skilled labour market, wage gaps, relative earnings to skills and diplomatic relations.

6.2 The impact of the brain drain on GDP per capita

For assessing the impact of the brain drain on GDP per capita, its effect on human capital formation is evaluated first. This is found in the literature to be the most influencing channel of the brain drain. The incentive effect should be strong for Ethiopia, since it is one of the poorest countries in the world (Beine et al., 2008; Geiger & Moller, 2013). Nevertheless, Beine et al. (2010) found that the incentive effect is too low to create a higher long-run, post-migration level of human capital if the skilled epost-migration rate is below 20%, which had been the case for Ethiopia for the whole period. Between 1980 and 1985, the high-skilled

emigration rate doubled, which can be an explanation for the increase in tertiary enrollment with approximately 70% between 1984 and 1987. This suggests an incentive effect. It is assumed that the incentive effect will occur after the emigration possibility is observed. However, the increase in the high-skilled emigration rate was relatively larger, implying a decrease in the level of human capital and supporting the results of Beine et al. (2010). Between 1990 and 1995, tertiary enrollment as well as the high-skilled emigration rate decreased slightly, and between 2000 and 2005, both tertiary enrollment and the high-skilled emigration rate faced a high expansion. Because these indicators seem to fluctuate together, they imply an incentive effect. Since the development of tertiary enrollment after 2000 is mostly explained by the efforts of the government, it cannot be concluded whether there was an incentive effect. The tertiary enrollment rate more than doubled, while the high-skilled emigration rate only increased with 46.7%. Even when assuming that the increased number of students graduate and migrate a period later, so between 2005 and 2010, the growth of the emigration rate is still smaller. Thus, the brain drain is likely to reduce the level of human capital in Ethiopia, decreasing GDP per capita. However, it is not large enough to create a substantial problem for the efforts of the government to increase the level of human capital. This is in line with Bénassy and Brezis (2012), who found that with a high level of

government intervention, the level of human capital will be high, taking the brain drain into account.

Comparing remittances as a percentage of GDP with the high-skilled emigration rate, both experienced their largest growth around the period from 2000 to 2005. However, while there was an increase in the high-skilled emigration rate between 1985 and 1990,

remittances as a percentage of GDP decreased and while the high-skilled emigration rate decreased slightly between 1990 and 2000, the remittances increased from 1990 onwards. Between 2005 and 2010, the findings contradict that human capital flight increases

remittances as well. So the relation between remittances and high-skilled emigration is 20

(21)

unclear, just like in the literature where micro- and macro-evidence differ. When comparing remittances as a percentage of GDP with high-skilled emigration as a percentage of total migration, they seem to fluctuate in the same direction. However, remittances as a percentage of GDP grew more than 8 times larger between 1990 and 1995, while the proportion of high-skilled emigration experienced its largest decrease in that period. Thus, there is no clear relationship between these indicators either. The poor policy of Ethiopia regarding remittances contributes to this observation. The World Bank says in its policy recommendations that Ethiopia should establish formal savings products for remittances, increase transparency of remittances-related fees and reduce costs through new technology such as mobile money transfers (Geiger & Moller, 2013). It should be noted that the data on remittances contains the whole world, while the data on migration only contains OECD countries.

Bhagwati and Hamada (1974) argue that, with progressive taxes, the investments in education will not be returned due to the fact that the skilled emigrants do not pay taxes to Ethiopia. Since Ethiopia has a progressive tax system (Federal Democratic Republic of Ethiopia, 2002) and invests a lot in education, this is welfare decreasing. However, no empirical evidence is found on the extent of this effect.

To sum up, the incentive effect is too low to increase the post-migration level of human capital. However, the brain drain is not large enough to create a real problem for the goal of the government. Considering remittances, no clear relation with human capital flight is found, so they are assumed to be an unimportant channel through which the brain drain can influence GDP per capita.

6.3 Future prospects

Although the brain drain is not a problem yet, it can become one if the high-skilled emigration rate keeps growing. Ethiopia should make sure that it is between 20% and 30% by

decreasing the push factors, since the wage gaps are huge. Programs such as the PhD in Informational Systems are a good initiative in this sense. However, this will not be enough to avoid migration. Ethiopia needs freedom of speech and a larger labour market for highly educated workers. To benefit the most from the skilled emigrants, Ethiopia should encourage remittances by implementing the recommendations of the World Bank.

7. Conclusion

The brain drain has increased extensively during the past decades and has become a

subject of interest for research. Because Ethiopia has invested a lot in its educational system since 1997, aims at high economic growth and experienced a large increase in its

(22)

skilled emigration rate, it is investigated how the brain drain affects the GDP per capita of Ethiopia.

The literature showed that the brain drain increases with push factors in the source country, large wage gaps between the source and destination countries and high relative earnings to skills in the destination country. The push factors include political instability, low institutional quality, a slack high-skilled labour market and lack of intellectual freedom and research facilities. The investments in education increased research facilities by building new universities and the quality is supervised by oversight agencies. However, there is no

intellectual freedom, because there is no full freedom of speech. Although the service sector has increased in Ethiopia, creating employment opportunities for high-educated workers, it is still largely dependent on agriculture. This implies a slack high-skilled labour market. Thus, despite the investments in education, there are still important push factors. With the data it is found that the largest OECD destination countries for high-skilled emigrants are the US, Canada and the UK. The development of GDP per capita gaps between them and Ethiopia and the relative earnings to skills in those destination countries confirm the findings in the literature that they are substantial pull factors.

The consequences of the brain drain for the level of human capital depend on the magnitude of it. The possibility to emigrate after education creates an incentive effect, increasing the tertiary enrollment rate. It is found in the literature that this effect is large enough to increase the long-run level of human capital if the high-skilled emigration rate is between 20% and 30%. Since this rate has been less than 20% for Ethiopia, the brain drain has decreased the level of human capital, which has a negative effect on GDP per capita. However, due to the investments of the government, tertiary enrollment increased much more than the high-skilled emigration rate after 1997. This indicates that the brain drain is not yet a considerable problem for the goal of the government to increase human capital.

The brain drain does have a negative effect on GDP per capita through lost taxes, since the government has invested a lot in the students that emigrate. However, the extent of this effect is not known.

Whether skilled emigrants remit more is unclear in the literature as well as the data. Therefore, it cannot be concluded whether remittances from the skilled emigrants can compensate for the lost human capital or taxes.

Concluding, it is found that the brain drain has had a negative effect on the GDP per capita of Ethiopia, because of the low incentive effect and the lost taxes. This is in line with the findings of Di Maria and Lazarova (2012), since Ethiopia is one of the poorest countries in the world and has low technological sophistication. If the high-skilled emigration rate keeps growing, the government should make sure that it does not become larger than 30% in order to benefit from the incentive effect. Since the GDP per capita gaps are very large and it is

(23)

very hard to diminish these extensively, the government should focus on decreasing push factors to control the high-skilled emigration. Otherwise, the brain drain may become a substantial problem.

7.1 Limitations and future recommendations

There are several limitations in this research due to unavailable data. First, only emigration data on 20 OECD destination countries is available with five year intervals and data on school enrollment, public expenditures on education and relative earnings to skills are not available for every year as well. This is also a limitation of the used literature. Assessing the skilled emigration to all other countries would be more complete. Better conclusions on remittances can then be made, since those data are from the whole world. Second, it was not possible to compare the skilled wage levels, only GDP per capita, since no data for Ethiopia was available. Comparing skilled wage gaps would give a better picture of the effect on the magnitude of the brain drain. Third, various researchers found that return emigration and technological diffusion, as consequences of the brain drain, have a positive effect on economic growth. However, no data is available on this as well. Fourth, the composition of human capital significantly influences growth (Di Maria & Lazarova, 2012). This implies that data is needed on the professions of the emigrants. Finally, empirical research to the total impact of the brain drain on GDP per capita, so considering technological diffusion and return emigration, has not been done yet and empirical research to the case of Ethiopia needs to be done to prove the found correlations and causal relationships. This would be more accurate if the missing data is collected.

(24)

References

Aredo, D. (2000). Human Capital Flight from Africa: An Assessment of Brain Drain from Ethiopia. Addis Ababa, Ethiopia.

Barro, R. J. (1991). Government spending in a simple model of endogenous growth. The Quarterly

Journal of Economics, 106(2), 407-443.

Batista, C., Lacuesta, A. & Vicente, P.C. (2012). Testing the ‘brain gain’ hypothesis: Micro evidence from Cape Verde. Journal of Development Economics, 97(1), 32-45.

Baxter, P., & Jack, S. (2008). Qualitative case study methodology: Study design and implementation for novice researchers. The qualitative report, 13(4), 544-559.

Beine, M., Docquier, F. & Oden-Defoort, C. (2010). A panel data analysis of the brain gain. World

Development, 39(4), 523-532.

Beine, M., Docquier, F., & Özden, Ç. (2011). Diasporas. Journal of Development Economics, 95(1), 30-41.

Beine, M., Docquier, F. and Rapoport, H. (2008). Brain Drain and Human Capital Formation in Developing Countries: Winners and Losers. The Economic Journal, 118(528), 631–652.

Belot, M. V., & Hatton, T. J. (2012). Immigrant Selection in the OECD. The Scandinavian Journal of

Economics, 114(4), 1105-1128.

Bénassy, J. & Brezis, E.S. (2012). Brain drain and development traps. Journal of Development

Economics, 102, 15-22.

Bhagwati, J., & Hamada, K. (1974). The brain drain, international integration of markets for

professionals and unemployment: a theoretical analysis. Journal of Development Economics, 1(1), 19-42.

Bollard, A., McKenzie, D., Morten, M and Rapoport, H. (2011). Remittances and the Brain Drain Revisited: The Microdata Show That More Educated Migrants Remit More. World Bank Economic

Review, 25(1), 132–56.

Brücker H., Capuano, S. and Marfouk, A. (2013). Education, gender and international migration:

insights from a panel-dataset 1980-2010. Norface Research Programme on Migration, Migration: New

developments, Spring 2013. Norface Migration: London. 31-32.

(25)

Cattaneo, C. (2009). International Migration, the Brain Drain and Poverty: A Cross-country Analysis.

The World Economy, 32(8), 1180-1202.

Collier, P., Hoeffler, A. & Pattillo, C. (2004). Africa's Exodus: Capital Flight and the Brain Drain as Portfolio Decisions. Journal of African Economies, 13(2), ii15-ii54.

Di Maria, C. & Lazarova, E.A. (2012). Migration, Human Capital Formation, and Growth: An Empirical Investigation. World Development, 40(5), 938-955.

Docquier, F., & Marfouk, A. (2006). International migration by educational attainment, 1990–2000. In C. Ozden, & M. Schiff (Eds.), International Migration, Remittances and the Brain Drain (pp. 151–199). New York: Palgrave Macmillan.

Docquier, F. & Rapoport, H. (2012). Globalization, Brain Drain, and Development. Journal of

Economic Literature, 50(3), 681–730.

Embassy of Ethiopia (n.d.). Ethiopia and the U.S. Retrieved from:

http://www.ethiopianembassy.org/AboutEthiopia/AboutEthiopia.php?Page=EthiopiaUS.htm

Faini, R. (2007). Remittances and the Brain Drain: Do more skilled migrants remit more? The World

Bank Economic Review, 21(2), 177-191.

Federal Democratic Republic of Ethiopia (2002, July 4). Income tax proclamation (Proclamation No. 286/2002). Addis Ababa, Ethiopia.

Geiger, M. & Moller, L.C. (2013). Ethiopia - Second economic update: laying the foundation for

achieving middle income status. Washington DC, World Bank Group.

Government of Canada (2014, March). Canada – Ethiopia Relations. Retrieved from:

http://www.canadainternational.gc.ca/ethiopia-ethiopie/bilateral_relations_bilaterales/canada_ethiopia-ethiopie.aspx?menu_id=7

Grogger, J., & Hanson, G. H. (2011). Income maximization and the selection and sorting of international migrants. Journal of Development Economics, 95(1), 42-57.

Grubel, H. B., & Scott, A. D. (1966). The international flow of human capital. The American Economic

Review, 56(1/2), 268-274.

(26)

Human Rights Watch (2014, May 5). Ethiopia: Brutal crackdown on protests. Retrieved from: http://www.hrw.org/news/2014/05/05/ethiopia-brutal-crackdown-protests

Le, T. & Bodman, P.M. (2011). Remittances or technological diffusion: which drives domestic gains from brain drain? Applied Economics, 43(18), 2277-2285.

Mankiw, N. G., Romer, D., & Weil, D. N. (1992). A contribution to the empirics of economic growth. The quarterly journal of economics, 107(2), 407-437.

Negash, S., Anteneh, S. & Watson, R.T. (2012). A PhD in information systems for emerging

economies: the Addis Ababa University model. Information Technology for Development, 18(3), 270-276.

OECD (2012). Education at a Glance 2012: OECD Indicators. Table A8.2a. Trends in relative earnings: Total population (2000-10). OECD Publishing.

Odia, L. O., & Omofonmwan, S. I. (2007). Educational system in Nigeria: Problems and prospects.

Journal of social science, 14(1), 81-86.

Ray, D. (1998). Development economics. New Jersey: Princeton University Press.

Saint, W. (2004). Higher education in Ethiopia: The vision and its challenges. Journal of Higher

Education in Africa, 2(3), 83-113.

Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.

Stark, O., Helmenstein, C., & Prskatwetz, A. (1997). A brain gain with a brain drain. Economic Letters,

55(2), 227–234.

United Nations Economic and Social Council (2000). Report of the Regional Conference on Brain

Drain and Capacity Building in Africa. Addis Ababa, Ethiopia.

Worldbank (2013, October). Ethiopia Overview. Retrieved from: http://www.worldbank.org/en/country/ethiopia/overview#3

Yin, R. K. (2003). Case study research: Design and methods (3rd ed.). Thousand Oaks, CA: Sage.

(27)

Appendices

Appendix 1: The Solow model

The Solow model uses the following equation to explain the growth of per capita capital stock:

𝟑𝟑. 𝟏𝟏 (1 + 𝑛𝑛)𝑘𝑘𝑡𝑡+1 = (1 − 𝛿𝛿)𝑘𝑘𝑡𝑡+ 𝑠𝑠𝑦𝑦𝑡𝑡

Where kt is the per capita capital stock at time t and kt+1 at time t+1. Rate n is the population

growth rate, 𝛿𝛿 is the capital depreciation rate and 𝑠𝑠𝑦𝑦_𝑡𝑡 is the per capita savings rate. So the Solow model assumes that the per capita capital stock grows with savings and declines by depreciation and population growth. This is combined with a Cobb-Douglas production function with diminishing returns to per capita capital in order to relate the per capita capital stock with per capita output or income. In the model, capital and labour produce total output together (Ray, 1998). As shown by Mankiw et al. (1992), by including human capital, the production function is as follows:

𝟑𝟑. 𝟐𝟐 𝑌𝑌(𝑡𝑡) = 𝐾𝐾(𝑡𝑡)𝛼𝛼_{𝐻𝐻(𝑡𝑡)}𝛽𝛽_{(𝐴𝐴(𝑡𝑡)𝐿𝐿(𝑡𝑡))}1−𝛼𝛼−𝛽𝛽

Where Y is output, K is capital, H is the stock of human capital, L is labour and A is the level of technology. It is assumed that L and A grow exogenously at respectively rates n and g. Together they represent the number of effective units of labour. The new Solow model, written somewhat in a different way, becomes:

𝟑𝟑. 𝟑𝟑 𝑘𝑘𝑡𝑡+1= 𝑠𝑠𝑘𝑘𝑦𝑦𝑡𝑡− (𝑛𝑛 + 𝑔𝑔 + 𝛿𝛿)𝑘𝑘𝑡𝑡 𝟑𝟑. 𝟒𝟒 ℎ𝑡𝑡+1= 𝑠𝑠ℎ𝑦𝑦𝑡𝑡− (𝑛𝑛 + 𝑔𝑔 + 𝛿𝛿)ℎ𝑡𝑡

Where k, y and h are quantities per effective unit of labour. The distinction between savings in the form of capital and savings in the form of investing in education can be seen in the difference between sk and sh. It is assumed that human capital depreciates at the same rate

as physical capital.

(28)

Appendix 2: Total emigration and emigration rates

Table 1

Emigration from Ethiopia between 1980 and 2010 with five year intervals, categorized by destination country and educational level.

Table 2

Emigration rates for Ethiopia and the world average between 1980 and 2010 with five year intervals, categorized by educational level. This is calculated by dividing the number of migrants of a particular level of education by the pre-migration stock of workers with that educational level.

Emigration rates

Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High

World 1,54% 1,26% 1,39% 4,94% 1,58% 1,26% 1,35% 4,78% 1,44% 1,10% 1,20% 4,17% 1,46% 1,12% 1,15% 4,11% 1,59% 1,26% 1,16% 4,14% 1,82% 1,34% 1,31% 4,96% 1,91% 1,40% 1,29% 5,31%

Ethiopia 0,07% 0,03% 0,26% 4,00% 0,17% 0,05% 0,50% 8,03% 0,29% 0,09% 0,75% 11,17% 0,39% 0,13% 1,02% 10,54% 0,42% 0,12% 1,02% 10,78% 0,55% 0,11% 1,15% 15,81% 0,62% 0,13% 1,14% 16,92%

2010 1980 1985 1990 1995 2000 2005

Emigration

Destination Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High Total Low Medium High

Australie 0 0 0 0 0 0 0 0 965 216 106 643 1647 433 107 1107 2552 580 267 1705 2860 579 276 2005 3149 520 289 2340 Austria 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Canada 575 120 105 350 1940 450 435 1055 8030 2555 2510 2965 11055 2720 3675 4660 11575 2725 3305 5545 16825 2330 5545 8950 19602 2338 4958 12306 Switzerland 0 0 0 0 447 255 109 83 1034 551 284 199 1456 867 332 257 2008 1260 411 337 2098 1298 457 343 2164 1292 507 365 Chile 10 0 0 10 8 0 0 8 13 3 5 5 44 10 20 14 108 21 54 33 142 24 73 45 201 28 105 68 Germany 0 0 0 0 0 0 0 0 0 0 0 0 10736 5241 3325 2170 10127 4984 3201 1942 7770 3238 2326 2206 7345 3541 2014 1790 Denmark 103 56 28 19 174 101 39 34 360 214 79 67 527 255 162 110 667 249 248 170 740 280 272 188 851 381 270 200 Spain 0 0 0 0 19 3 6 10 31 4 10 17 102 33 47 22 220 31 101 88 868 225 430 213 2619 505 1436 678 Finland 15 6 1 8 23 11 2 10 74 31 11 32 306 129 77 100 394 122 137 135 510 183 162 165 1004 524 237 243 France 0 0 0 0 406 242 66 98 916 476 160 280 1281 711 219 351 1598 886 267 445 1633 820 279 534 2384 1379 398 607 United Kingdom 0 0 0 0 1376 716 344 316 2855 1547 663 645 3472 1987 588 897 5709 2492 688 2529 6683 2568 778 3337 8618 2702 939 4977 Greece 0 0 0 0 706 400 247 59 608 323 222 63 776 344 290 142 1780 390 910 480 2554 460 1119 975 3216 407 1632 1177 Ireland 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 39 13 6 20 80 25 13 42 118 26 20 72 Luxembourg 4 2 1 1 5 3 0 2 6 3 0 3 8 7 0 1 21 13 4 4 21 13 4 4 24 13 6 5 Netherlands 2110 1687 278 145 2767 2051 440 276 3770 2566 688 516 4046 2558 790 698 4898 2692 1239 967 5689 2881 1450 1358 6570 2967 1923 1680 Norway 141 54 61 26 307 106 136 65 884 335 304 245 1324 344 627 353 1934 426 962 546 2664 505 1297 862 3588 592 1687 1309 New Zealand 147 69 42 36 144 33 42 69 174 42 33 99 249 111 39 99 450 147 165 138 339 42 111 186 561 149 209 203 Portugal 8 6 2 0 7 5 2 0 6 3 2 1 8 4 2 2 12 5 2 5 16 6 3 7 18 6 4 8 Sweden 863 488 265 110 1758 926 575 257 5411 2744 1809 858 8939 3977 3435 1527 9176 2667 4521 1988 9111 2366 4441 2304 11500 2953 5087 3460 United States 5152 864 912 3376 14941 2152 2888 9901 25098 2673 5135 17290 32843 2695 10603 19545 44522 3790 15551 25181 86380 6504 29072 50804 119340 11914 42780 64646 Total 9128 3352 1695 4081 25028 7454 5331 12243 50235 14286 12021 23928 78819 22426 24338 32055 97790 23493 32039 42258 146983 24347 48108 74528 192872 32237 64501 96134

Percentage of total migration 44,7% 48,9% 47,6% 40,7% 43,2% 50,7% 49,8%

2005 2010 1980 1985 1990 1995 2000

(29)

Appendix 3: Wage gaps and relative earnings

Table 3

Difference in GDP per capita in constant 2005 US dollars between Ethiopia and OECD countries. There is no information available on the GDP per capita of Ethiopia in 1980.

Wage gaps 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 United States 26346 25596 26535 28236 29200 29942 30693 31703 32570 32824 32378 33085 33538 34471 34987 United Kingdom 19735 20162 20883 21421 22168 23001 23991 25153 25668 25793 25394 25669 26494 27743 28647 Sweden 25483 25791 26264 27367 27936 28616 29468 30090 30692 30686 30141 29630 28836 29786 30797 Canada 23465 22518 22894 24011 24953 25295 26017 26970 27194 26842 25935 25857 26164 27165 27696 Portugal 10140 10298 10224 10000 10277 10690 11370 12250 13073 13626 14275 14459 14133 14232 14791 Wage gaps 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 United States 35887 37053 38266 39665 40831 40804 41148 41944 43127 44154 44886 45245 44671 43021 43718 United Kingdom 29565 30777 31795 32621 33924 34533 35179 36412 37344 38281 39041 40061 39428 37086 37377 Sweden 31235 32066 33409 34943 36442 36796 37592 38337 39796 40881 42392 43463 42844 40321 42591 Canada 27832 28708 29639 31025 32363 32602 33261 33559 34251 34928 35620 36040 35882 34471 35169 Portugal 15273 15880 16626 17208 17757 17974 18015 17789 18012 18106 18324 18710 18667 18088 18413 29

(30)

Table 4

Relative earnings by educational attainment from 2000 to 2010 for the total population between 15 and 64 years old. Post-secondary non-tertiary education = 100.

OECD 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Canada Below upper secondary 82 79 79 81 81 78 77 83 82 80

Canada Tertiary 142 141 135 138 137 135 136 140 138 138

Sw eden Below upper secondary 86 87 87 87 86 85 84 83 83 82

Sw eden Tertiary 131 130 128 127 126 126 126 126 126 125

United Kingdom Below upper secondary 69 70 68 69 69 71 71 70 71 70 67

United Kingdom Tertiary 160 160 157 162 157 158 160 157 154 159 165

United States Below upper secondary 68 66 66 65 67 66 65 66 64 66

United States Tertiary 176 172 172 172 175 176 172 177 179 177

Portugal Below upper secondary 67 67 68 68

Portugal Tertiary 178 177 177 169

OECD average Below upper secondary 80 79 79 79 78 78 78 78 78 77 76 Tertiary 149 145 148 147 155 151 157 154 153 155 159

The impact of the brain drain on GDP per capita : a case study of Ethiopia