Determinants of Migration in Pakistan

With an average population growth of more than 2.5 percent for decades, Pakistan is a labour surplus country. Therefore, like many developing countries, export of labour is one of its development strategies. It is one of the top 10 major emigration countries in the World. Export of labour reduces unemployment; increases wages and the remittances sent home improve the balance of payments and reduce poverty. However, much has been studied about remittances in the developing countries, their determinants and consequences because of their fiscal expediency and availability of data; little is known about determinants of international migration itself because of lack of data. The development of international data on migration during the last decade has led to a number of studies relating to either one-destination country like the US or for a few destination OECD countries. Moreover, these studies pool migrants from the developed and developing countries together. However, in Pakistan’s context studying only the determinants of migration in the OECD countries would be only a part of the story because majority of 60 percent of Pakistanis go to the Middle East countries. There is a huge difference in the structure and the quality of migrants in these two regions. The migrants to the Middle East/Gulf countries are mainly skilled or unskilled construction workers working on short contract periods, in response to the demand in these areas due to huge infrastructural development. In contrast, the migrants towards OECD countries are mainly educated people who go to these countries to get higher returns usually on permanent basis. These migrants are also selected by the immigration policies in the OECD countries, which prefer educated workers who could contribute in their economies. While workers in the non-OECD countries are comparatively from lower income groups who go without their families, the migrants to the OECD countries go with their families. Gulf countries are near and less costly to travel to and they have similar culture. The number and types of emigrants varies across destinations, reflecting differences in both their attractiveness and openness to international migrants. Given the broad diversity of migration patterns in Pakistan both with regard to the characteristics of migrants and of the countries of their destination, it would be interesting to identify the determinants of migration and to extend the literature on developing countries. The literature on international migration from developing countries is very limited. Some of these studies look at the supply side determinant of migration and find that low income, population pressure on resources and increase in the cohort of young potential migrants -‘demographic supply side pressure’ are the main driving force of migration in less developed countries Hatton and Williamson (2001, 2011). Other studies (for example Mayda, 2010) which are mostly from OECD host countries’ perspectives consider the demand side determinants and use some exogenous measures of immigration policies of the host countries which are subjective in nature. However, these studies overlook the demographic characteristics of host countries that could be significant determinants of the migration process and are important in the making of the immigration policies in the host countries. The main emphasis of this study is to look on the impact of previous migrant stock on potential emigration rate from Pakistan. It is now well documented in the literature that family and friends reduce the cost of potential migrants in several ways, like by providing them cost of travel, food and shelter and information about jobs and opportunities in destination countries,. A few studies also relate to the influence of the previous stock of migrants on the immigration policies of the destination countries, which encourage families and friends to migrate. Given the poverty levels in Pakistan, the role of previous stock of migrants is important in determining the propensity to migrate of future emigrants. Therefore, it is useful to study the determinants of migration from Pakistan. We also explore whether these networks have different impacts in both OECD and non-OECD regions, which have diverse demographic, socio-political and economic features that affect the demand for, labour both in numbers and in their skill levels. A modified gravity model is used to study the emigration rate from Pakistan to 175 countries for the period 1980-2000 obtained from the global database of the Development Research Centre on Migration, Globalisation and Poverty (Migration DRC). We explain the emigration rate (supply of migrants) from Pakistan by the income, population density, dependency rate and tertiary rate of education in the host countries (demand side determinants). Moreover to focus on the network effect on future migration, a lagged migrant stock variable has been used. The results of this study are consistent with the theory, which considers migration as a human capital investment, and imply that migration is more likely to get higher income. The coefficient on the income in the host country is positive and highly significant in all the specifications of regression models and thus an important determinant of migration from Pakistan. The results also support the view of the network theory of migration, the impact of lagged migrant stock on future emigration rate is positive and highly significant and these effects are positive in both OECD and the Middle East countries. Thus networks of family and friends previously migrated have a strong positive impact on current emigration rate. In addition, the coefficient on distance, which indicates the cost of migration, is negative in all regressions and significant. However, distance loses significance when lagged migration stock is included in the specifications, implying that these networks reduce the cost of migration. The findings of this study also indicate that high population density is a deterrent and an increase in the rate of tertiary education in the host country discourages emigrants. This study also finds mixed effects of traditional gravity variables on emigration rate like many earlier studies on migration. For example, the coefficient on common language is positive and significant as expected but the signs on the coefficients on Commonwealth and neighbouring country are mostly not according to the expectations. When Commonwealth countries are lumped together, their effect on emigration rate turned out to be unexpectedly negative. This may be due to the diverse nature of these countries as Commonwealth countries include both OECD and non-OECD countries. When a dummy variable is used for the UK, the effect turned out to be positive and significant as expected. This implies that migrants find UK more attractive than other Commonwealth countries. Similarly, the effect of China as a neighbouring country is not positive as is expected that it is less costly to travel to neighbouring countries. There are several other types of costs of mobility apart from distance in kilometres, like the type of border terrain and the policies of the neighbouring countries.