The Medical Brain Drain from Small States *

PAPER #2 The Medical Brain Drain from Small States * By Frédéric Docquier a and Maurice Schiff b a FNRS and IRES, Catholic University of Louvain b Development Research Group, World Bank March 2007 * This paper is a product of the Economic Policy and Debt Department of the World Bank, which launched a Research Program on Economic Growth and Integration of Small States to the World Economy (World Bank contract PO. 0007674813). It was presented at the World Bank Conference on Small States, Growth Challenges and Development Solutions, Washington, 7-8 December. We thank David McKenzie and Edgardo Favaro for their comments. 1

Foreword The Economic Policy and Debt Department of the World Bank launched a Research Program on Economic Growth and Integration of Small States to the World Economy, which includes a series of economic growth thematic, regional and case studies. The objective of these studies is to improve our understanding of the sources of and impediments to economic growth of small states in the Africa, Caribbean, Europe, Middle East, Asia and Pacific regions. Small states are grouped by continent and/or region to analyze more deeply the characteristics of countries in each group that might not emerge in a more aggregate study. The regional studies examine economic growth at the country level and distill lessons for other countries with similar characteristics in the region and elsewhere. In most cases, the analysis will not discuss all small states in the region; however, the coverage will be comprehensive in studying small states most important problems in each region. Recent changes in information and communication technologies (ICT) have contributed to a dramatic increase in the degree of integration and interdependency of countries, markets and people. Against this background, one aspect of particular concern for small states is the international movement of people. This report focuses on this particularly important aspect of globalization, with emphasis on the mobility of skilled people. This intermediate report includes three papers as described in the terms of reference. Paper #1 describes how data on skilled emigration rates are constructed for small states. Paper #2 focuses on the medical brain drain and depicts the situation of small states. Paper #3 analyzes the determinants of the brain drain, especially its relationship with country size. 2

1. INTRODUCTION The growth of international migration has accelerated in recent years, particularly for skilled labor. The same has occurred for remittances. This recent phenomenon has led to a renewed interest in the topic of migration and remittances by source and host country analysts and policy makers, as well as by multilateral, regional and bilateral development institutions (Ozden and Schiff 2006). Many economists have analyzed the possible impact of brain drain on the origin countries and inequality across nations. The early literature dates back to the 1960s and 1970s and supports the view that skilled migration is detrimental for those left behind. The main argument was that migrants' contribution to their origin economy, referred to as their social marginal product, is greater than their private marginal product because of the positive externalities they tend to generate for the rest of society. Such a view has been recently reformulated in an endogenous growth framework (Haque and Kim 1995). In the last decade, some channels through which the brain drain may positively affect the migrants home economies have been presented in the literature. These include a range of feedback effects such as remittances, return migration after additional knowledge and skills have been acquired abroad, creation of business and trade networks, and the effect of migration prospects on the expected return to education and human capital formation. In most of these studies, the brain drain is seen as the emigration of workers with high skills, whatever their field of study. However, if concentrated in certain fields, brain drain can induce specific occupational shortages that are harmful for developing countries. Recall the explosion of the space shuttle Challenger which was caused by the malfunctioning of small components called O-rings and which illustrates the complementarity between the various components, and basic and intermediate inputs, in the production process. Kremer (2000) proposed an ''Oring theory'' of economic development in which the production process consist of a series of tasks. Deficiencies in any of those tasks can lead to substantial reduction in the value of output. The same argument applies to the brain drain. This effect can be reinforced by the fact that individual governments have fewer incentives to provide internationally applicable education. Poutvaara (2004) addresses this important issue in a theoretical model where the possibility of a brain drain distorts provision of public education away from internationally applicable education towards country-specific skills. Countryspecific skills may include both tertiary education with national emphasis, such as degrees in 3

law and certain humanities, as well as secondary education which is less mobile internationally. Correspondingly, internationally applicable education may include, in addition to sciencebased, commercial and other internationally applicable degrees in tertiary education, those held in secondary education (like nurses), that are internationally mobile. At the end, this means educating too few engineers, economists, nurses and doctors, and too many lawyers. In many poor countries, shortages are particularly severe in the medical sector where the numbers of physicians and nurses per 1000 inhabitants are extremely low compared to industrialized countries. The brain drain of physicians and nurses to countries such as the US, Australia, Canada and the UK is one of the causes of shortage (Awases et al., 2004, Bhargava, 2005). Such a medical brain drain is likely to generate important effects on health outcomes and economic activity. This paper is concerned with the measurement of the medical brain drain, with a particular focus on the situation of small states. There are no official statistics on the emigration of healthcare staff. We gathered information from many medical associations and Statistics Institutes in OECD countries on immigration of doctors on an annual basis for the period 1991-2004. Then combining these data, we build a unique and original data set on the emigration of healthcare staff from developing to developed countries. 2. SMALL STATES This section first defines the small state concept and, second, identifies them as well as some of their characteristics. The measures and levels of medical brain drain are provided in Section 3. 2.1. Definition Small states can be defined according to population, GDP, size, and more. Moreover, various thresholds and base years can also be used. Since the three measures are highly correlated, 1 the findings are likely to be (approximately) invariant to the choice of country size measure. We follow the standard approach and use population size as the measure of country size. There is no special significance in the selection of a particular population threshold to define small states. The Commonwealth uses a threshold of 1.5 million people in its work on small 1 In 2000, the correlation rate between population and size is 84% and between population and GDP is 78%. 4

states, but it also includes larger member countries (Jamaica, Lesotho, Namibia and Papua New Guinea) because they share many of the same characteristics of smallness. The World Bank Task Force on small states used the same threshold as a convenient yardstick for classifying all small states, and only considers sovereign states. 2.2. Who Are the Small States and What Are Some of their Characteristics? Using the standard of a population below 1.5 million people in 2000, 45 developing countries are small, accounting for nearly one third of the total number of developing countries. They are home to 20 million people, less than 0.4 percent of the total population of developing countries. They range in size from micro-states like Cook Islands, Nauru, Niue, Palau, St. Kitts and Nevis, and Tuvalu (with fewer than 50,000 people each) to Botswana, Gabon, The Gambia, Guinea-Bissau, Mauritius, and Trinidad and Tobago (with more than 1 million people each). The per capita GNP in these countries also ranges widely, from less than $400 in several African countries (Comoros, The Gambia, Guinea-Bissau, and Sao Tome and Principe) to just $700-1,300 in such countries as Cape Verde, Guyana, Kiribati, Maldives, Solomon Islands, and Tuvalu; to more than $9,000 (The Bahamas, Brunei, Cyprus, Malta, and Qatar). There are small states in every geographic region, but most countries fall into three main groups: twelve states are in the Caribbean region, fourteen in East Asia and Pacific, and twelve in Africa. Of the remaining seven countries, two are in South Asia, two in the Middle East, and three in Europe. 3. MEASURING THE MEDICAL BRAIN DRAIN Following Carrington and Detragiache (1998, 1999) and Docquier and Marfouk (2006), we evaluate the medical brain drain in terms of stock and rates. The stock of physicians from country i and working abroad is denoted by M,. Let i t P, denotes the number of physicians working in the home country. The rate of medical brain drain from country i at time t can be written as m i, t = P i, t M i, t + M i, t The computation of such medical emigration rates requires combining many data sources and using certain methodological assumptions. i t 5

3.1. Source and methodology The number of physicians by country and by year can be obtained from the World Development indicators or from the World Health Organization Statistics. The number of emigrants is more difficult to assess. Here we opt for the following methodological options: - Emigration data, when available, are incomplete and imprecise. Hence, the only way to capture the medical brain drain is to collect new immigration data in the most important host countries. We focus on 16 important OECD countries where data on foreign doctors are available. The stock M, is then obtained by aggregating bilateral data from country i i t to the 16 receiving countries at time t. - We restrict our sample to general practitioners in activity in the host country. - We develop an annual data set and concentrate on the period 1991 to 2004. Data provided by national medical associations are available on an annual basis. Regarding data extracted from national censuses, we usually have two or three points. We interpolate the remaining years using a log-linear adjustment: n k k [ ] ln[ M ] [ M ] ln M i, t+ k = i, t + ln i, t+ n, for k=0 to n. n n - When the data are available, emigrants are defined according to their country of qualification. Such data can be obtained from the national medical associations and represent about 67 percent of our sample in 2004. This is the case for Canada, France, New Zealand, Norway, the United Kingdom and the United States. When the country of qualification cannot be determined, we define migrants according to their country of birth. Such data can be obtained from national censuses and registers. This is the case for Australia, Austria, Belgium, Denmark, Ireland and Sweden. They represent 20 percent of our sample in 2004. Finally, when the country of birth is not available, we define migrants according to their citizenship. This is the case in Italy, Germany, Portugal and Switzerland, i.e. about 13 percent of the sample. Table 1 provides the data sources. 6

Table 1. Data source Country Source Definition Available data Number of immigrants In % in 2004 Australia Australian Bureau of Census Country of birth (corrected) 1991, 1996, 2001 14805 9.9% Austria Statistik Austria Country of birth 1991, 2001 3940 2.6% Belgium Institut National de Statistiques Country of birth 1991, 2001 10618 7.1% Canada Canadian Medical Association Country of qualification From 1994 to 2004 15331 10.3% Denmark Statistics Denmark Country of birth 2004, 2005 350 0.2% France French Medical Association Country of qualification From 1991 to 2004 1585 1.1% Germany German Medical Association Country of citizenship From 1991 to 2004 11318 7.6% Ireland Central Statistical Office Country of birth 1991, 2002 2016 1.4% Italy Istituto nazionale di statistica Country of citizenship 1991 4910 3.3% New Zealand New Zealand Medical Association Country of qualification 1991, 2004 3994 2.7% Norway Norway Medical Association Country of qualification 2004 3752 2.5% Portugal Ordem dos medicos Lisboa Country of citizenship 2002, 2003 1989 1.3% Sweden Statistics Sweden Country of birth 1991, 2003 6134 4.1% Switzerland Office Fédéral de la Statistique Country of citizenship 1991, 2000 2206 1.5% United Kingdom General Medical Council Country of qualification From 1991 to 2004 4716 3.2% United States American Medical Association Country of qualification From 1991 to 2004 61396 41.2% 7

4. RESULTS In describing the results of Table 2, we denote to the number of doctors per 1000 inhabitants by ND, the brain drain by BD and the medical brain drain by MDB. Changes always refer to the period 1990-2000. The small states are disaggregated into three groups according to population size P (in millions), which are referred to as Group 1 (P <.5), Group 2 (.5 < P < 1), and Group 3 (1 < P < 1.5). Two other groups are Group 4 with 1.5 < P < 3, and Group 5 with 3 < P < 4. Finally, unless otherwise indicated, all figures refer to small states. Table 2 presents some interesting results on the MBD, several of which are rather alarming. First, ND fell from 1.09 to.98 (i.e., it fell by 10%). This decline was essentially due a large decline in Group 3, from 1.40 to.98 (by some 36%), given that ND increased in Group 1 (from 1.01 to 1.09) and Group 2 (from.80 to.97). The MBD increased from 15.9 to 24.3% (by 8.4 percentage points, equal to some 35%). This is essentially due to a large increase in Group 1, from 31.2 to 44.0% (by 12.8 percentage points, equal to 41%). The fact that Group 1 experienced an increase in both its MBD and its ND implies that the smallest of the small states were successful in substantially raising the number of doctors graduating from its medical schools. Second, the ND in 2000 was larger for the larger countries of Groups 4 and 5, at 1.30 for Group 4, or some 33% more than for small states as a whole. It was 1.93 for Group 5 or 97 more than for small states. Third, ND was larger for developing countries as a whole in 1990, at 1.19 (9% more than for small states) and in 2000, at 1.15 (17% higher). The MDB for all developing countries was 1.8% in 1990 or 11.3% of that for small states, and was 2.1% in 2000 or 8.6% of that for small states. Thus, even though the MBD increased for developing countries as a whole, it increased less than in small states (16.7% versus 35%). In other words, the situation of small states worsened relative to developing countries as a whole, from being 8.8 times larger in 1990 to 11.6 times in 2000. Fourth, the ND in high-income countries as a whole was 2.49 in 1990 (130% larger than the ND in small states) and 2.80 in 2000 (185% more than in small states). Surprisingly, the BD in high-income countries was larger than that for developing countries as a whole, at 2.7% versus 1.8% in 1990 (i.e., 50% more) and at 2.6% versus 2.1% in 2000 (24% more). 8

Fifth, quite surprisingly, the ND in the group of 37 small-island developing states in 2000 was twice as large as that for small states as a whole (1.96 versus.98), 70% larger than for developing countries as a whole, and 40% larger than the world average. Its MBD was less than half that of small states as a whole (10.9% versus 24.3%, or 55% lower). Sixth, of the eight small state country groups (3 population groups, 4 region/income groups, and the group of small-island developing states), the ND increased in five of them and fell in Group 3 by 36%, in East Asia and the Pacific by 32%, and in Sub-Saharan Africa by 6.4%. It fell in small states as a whole by 10%. Turning to Table 3, data on the MBD are available for 34 of the 46 small states. We find that, out of the 10 top MBD countries in 2000, 5 belong in the top 10 brain drain (BD) countries, 3 in the top 15 BD countries, and the other 2 in the top 20 BD countries. Thus, it would seem that there might be a high rank correlation between countries ranked according to the BD and according to the MBD. Looking at the countries with highest MBD, Dominica is the top one with an MBD of 98.1% (it had the tenth highest BD). Grenada is second with an MBD of 97.9% (and was second for the BD as well). Saint Lucia is third at 69.8%, Cape Verde is fourth at 54.2%, and Fiji is fifth at 50.1%. These levels of MBD are extremely high. The figures for Dominica and Grenada mean that they will keep 2 doctors out of 100 educated there, Saint Lucia will keep 30, Cape Verde 46 and Fiji 50. Surprisingly, Guyana is only ranked 25 th of the 34 countries with MBD data (at 2.7%), even though while it is ranked the highest in terms of the BD (at 89%). In other words, its brain drain is 33 times greater than its medical brain drain. Finally, figure 1 reveals that the medical brain drain has been rather stable over the last 15 years, except in Latin American and Caribbean countries. One explanation could be that, due to capacity restrictions in US medical schools, the outsourcing of medical training to Caribbean is more and more important. 9

Table 2. Medical brain drain and health staff in small countries Nb Physicians per 1000 people 2000 1990 Medical/general Physicians per brain drain 22+ 1000 people Medical brain drain Medical brain drain Medical/general brain drain 22+ Small States (pop < 1.5 million) 46 0.98 24.3% 80.5% 1.09 15.9% 43.1% By population size Population from 0 to 0.5 million 32 1.09 44.0% 151.4% 1.01 31.2% 91.9% Population from 0.5 to 1 million 8 0.97 8.3% 25.4% 0.80 9.0% 27.8% Population from 1 to 1.5 million 6 0.90 4.1% 14.1% 1.40 3.2% 5.6% By region / income East Asian and Pacific 12 0.26 28.8% 83.4% 0.38 21.5% 35.9% Latin America and Caribbean 10 0.73 64.7% 103.8% 0.62 44.9% 70.3% Sub-Saharan Africa 10 0.29 10.6% 33.2% 0.31 9.9% 29.0% High-income countries 12 2.17 9.4% 63.1% 1.78 12.0% 70.6% Other groups of interest Small Islands Developing States 37 1.96 10.9% 38.6% 1.57 10.3% 33.6% Population from 1.5 to 3 million 15 1.30 3.2% 23.8% 1.51 2.9% 19.3% Population from 3 to 4 million 13 1.93 7.5% 54.5% 2.74 7.8% 51.1% World average 192 1.41 2.3% 62.2% 1.42 2.1% 60.5% Total high-income countries 41 2.80 2.6% 112.0% 2.49 2.7% 111.3% Total developing countries 151 1.15 2.1% 40.5% 1.19 1.8% 33.3% Source: Bhargava and Docquier (2006) 10

Table 3. Small states medical brain drain in 2000 Country General General Medical Population 0+ 22+ Tuvalu 27.3% 23.8% - 11468 Nauru 34.5% 19.8% - 12809 Palau 26.1% 18.5% - 20016 San Marino 17.1% 14.9% - 28503 Saint Kitts and Nevis 78.5% 65.3% 35.5% 38836 Marshall Islands 39.4% 39.2% - 57738 Antigua and Barbuda 66.8% 49.6% 3.2% 68320 Dominica 64.2% 51.2% 98.1% 69278 Andorra 6.9% 4.6% 1.3% 69865 Seychelles 55.8% 47.5% 4.7% 80832 Grenada 85.1% 76.9% 97.9% 89357 Kiribati 23.1% 20.7% - 100798 Tonga 75.2% 58.8% - 101000 Micronesia, Federated States 37.8% 34.8% - 107000 Saint Vincent & Grenadines 84.5% 75.1% - 119000 Saint Lucia 71.1% 59.2% 69.8% 147000 Sao Tome and Principe 22.0% 20.0% 44.3% 149000 Samoa 76.4% 60.9% - 174000 Vanuatu 8.2% 4.7% - 195000 Belize 65.5% 47.0% 12.8% 238000 Barbados 63.5% 47.5% 13.8% 266000 Iceland 19.6% 15.8% 25.3% 282000 Maldives 1.2% 0.8% 1.4% 290000 Bahamas 61.3% 42.3% 0.9% 300000 Brunei 15.6% 9.7% 10.2% 335000 Malta 57.6% 44.1% 7.0% 389000 Suriname 47.9% 36.7% 4.9% 425000 Luxembourg 8.0% 5.8% 21.0% 436000 Solomon Islands 6.4% 3.5% - 437000 Cape Verde 67.4% 55.5% 54.2% 438000 Macao 14.4% 11.4% 2.0% 449000 Equatorial Guinea 12.9% 10.2% 1.6% 457000 Qatar 2.5% 1.9% 2.8% 582000 Djibouti 11.0% 7.5% 2.6% 666000 Bahrain 4.9% 3.5% 1.8% 677000 East Timor 15.5% 7.9% 16.2% 701000 Comoros 21.9% 13.1% 2.7% 706000 Guyana 89.0% 81.9% 2.7% 761000 Cyprus 33.2% 21.3% 5.8% 783000 Fiji 62.2% 44.5% 50.1% 813000 Mauritius 56.1% 45.1% 7.9% 1186000 Gabon 14.7% 8.4% 1.1% 1258000 Trinidad and Tobago 79.3% 67.5% 7.0% 1289000 Gambia 63.2% 60.4% 14.3% 1312000 Estonia 11.5% 9.4% 2.8% 1366000 Guinea-Bissau 24.4% 18.7% 1.1% 1367000 11

Figure 1. Evolution of the medical brain drain in small states (1991-2004) 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 East Asian and Pacific Latin America and Caribbean Sub-Saharan Africa High-income countries 5. CONCLUSION There are several small states groups where the problem of the decline in the number of doctors per 1000 inhabitants (ND) is worrisome. These are Group 3 (with the largest population among small states) where the ND fell by 36%, East Asia and the Pacific, by 32%, and Sub-Saharan Africa, by 6.4%. The ND also fell in small states as a whole by 10%. In other words, the four small state country groups where the ND fell dominated the five groups where it increased. As for the smallest of the small states (Group 1), the fact that they experienced an increase in both their MBD and ND implies that they were successful in substantially raising the number of doctors graduating from their medical schools. Thus, special efforts would have to be made in these regions. Though the decrease in the ND of Sub-Saharan Africa was relatively small, the region already had such a low ND in 1990 that any reduction in it might have dramatic consequences. Moreover, with the poaching of doctors and other healthcare providers in Sub-Saharan Africa in recent years, the situation may well have worsened since 2000. Both the low ND level and its decrease should convince 12

the international community that it should act more forcefully and probably more creatively in their attempt to reverse the trend. We now describe two policies that might be helpful for source countries. Source countries should generally benefit from programs established by host countries in cooperation with the former, whereby host countries would provide fellowships to study medicine or for doing the internship abroad, with a condition that the recipient would return home after graduation for a period of time before s/he would have the option to emigrate. Like any policy, implementation problems will have to be resolved. For instance, some countries might not have the resources to do so and might have to obtain financial support from private sources (foundations) or from the authorities of the countries where the students decided to go. Moreover, some students or interns might decide not to return despite their commitment to do so, and ensuring the program s success would require, say, for the host country to commit itself not to renew the visa of anyone not returning to their home country after completing their studies or internship. Second, the medical brain drain might be reduced if source countries were to improve doctors working conditions, in terms of pay as well as in terms of facilities (including equipment, medicines, basic implements, and more). Many small developing countries would be unable to finance such a program and would need support from host countries and from regional and multilateral development institutions. Third, source countries would benefit if host countries hiring contracts were temporary. They might stipulate that doctors might or would be able to return to the host country after a specified period in their home country. Such circular migration would benefit both source and host countries. REFERENCES Awases, M., A. Gbary., J. Nyoni, and R. Chatora. (2003), Migration of health professionals in six countries: A synthesis report, World Health Organization, Regional Office for Africa, Congo. Bhargava, A. (2005), "AIDS epidemic and health care infrastructure inadequacies in Africa: A socioeconomic perspective", Journal of AIDS, 40, 241-242. 13

Carrington, W.J. and E. Detragiache (1998), How big is the brain drain?, IMF Working paper WP/98/102. Docquier, F. et A. Marfouk (2006), International migration by educational attainment (1990-2000), in: Ozden, C. et M. Schiff (eds), International migration, remittances and the brain drain, Chap 5, Palgrave-Macmillan. Haque, N. U. and A. Jahangir (1999), The quality of governance: second-generation civil reform in Africa, Journal of African Economies 8, 65-106. Kremer, M. (1993), The O-Ring Theory of Economic Development, Quarterly Journal of Economics, 551-576. Ozden, C. et M. Schiff (2006), International migration, remittances and the brain drain, Palgrave-Macmillan. Poutvaara, P. (2004), Public education in an integrated Europe. Studying to migrate and teaching to stay, CESifo Working Paper, n. 1369. 14