Economics of Migration

University dissertation from Department of Economics, Lund University

Abstract: Abstract Population movements are more substantial today than at any other point in human history. If managed effectively, migration can be beneficial for all aspects of social and economic life. This thesis contains four papers, all of which are related to the economic consequences and determinants of migration, within and across countries. The first paper employs a household survey to explore whether education has an effect on the income attainments of rural households in China. The educational levels of people residing in the households and people who have moved from the households (migrants) are considered. A substantial flow of rural-urban migration, considerable drop-out rates and emerging evidence of the positive effects of education in China motivate the study. The findings generally show that the effects of education are positive. For household members, more education increases earnings from non-farm work. Permanent migrants with 7-9 years of schooling remit more money to the households than permanent migrants with 1-6 years of education. However, although temporary migrants are found to contribute to household incomes by bringing home money earned from non-farm work, the results do not indicate that more education increases the ability to do so. The second paper employs a household survey from Guatemala. Building on the general observation that distinct segments of the labor market are characterized by varying entry and exit conditions, the aim is to compare the effects of time in different segments of the labor market and to investigate whether the effects vary between migrants and natives. A particular focus lies on the dichotomous division between the informal and formal sectors. The labor market is divided into Agriculture (informal), Self-employment (informal), Uncovered Wage Work (informal) and Covered Wage Work (formal). This paper investigates how the amount of time individuals have spent at their current residence before obtaining their current job (YearsCurrent) has affected their probability of employment, and how the number of years of employment (Experience) affects earnings. Amongst other things, time can be a proxy for individual influence and knowledge. For all sectors except agriculture, it is found that YearsCurrent is positively correlated with the probability of employment, and that the effect is higher for migrants than natives. There is a reversed pattern in agriculture. Experience positively affects earnings for all covered wage workers, and there is a positive effect for migrants performing uncovered wage work. Experience has a negative correlation with earnings in agriculture for all workers. Two important findings are: 1) For natives, YearsCurrent has the strongest effect for uncovered wage workers, which could indicate a higher importance of social capital than in other sectors. 2) Migrants generally benefit more from time than natives, which contradicts migrant discrimination. The third paper employs macroeconomic data and simulates dynamic multiplier effects of international remittances in a panel of 120 countries for the period 1980 to 2006. Three structural equations are estimated to assess how a unit increase in remittances affects total income in the receiving country within one year (short run) and ten years (long run). When remittances are assumed to have the same spending patterns as other incomes (GDP), the findings suggest positive multiplier effects, which vary substantially across regions and income categories. North Africa and the Middle East and Latin America and the Caribbean had the greatest rise in income due to remittances since the 1980s (400 US$/Capita/Year). Among income categories, lower middle income economies benefited the most (330 US$/Capita/Year). Separating remittances from other incomes and allowing different spending patterns change the results quite dramatically. The results of this exercise, which must be interpreted very cautiously, suggest considerably different multiplier effects that are lower in some regions (sometimes even negative) and higher in others. The fourth paper, co-authored with Fredrik NG Andersson, explores the reverse causality problem of remittances in a new-fangled way; i.e. by testing the validity of two hypotheses about the determinants of remittances i.e. the altruism and insurance hypotheses. These predict the same short run, but different long run relationships between remittances and consumption. The hypotheses are tested by decomposing the data into cycles, representing the short run, and a trend, representing the long run. A macroeconomic panel with consumption and remittances data for 50 low and middle income economies between 1980 and 2006 is used for this purpose. This paper estimates Keynesian consumption functions with GDP and remittances per capita as explanatory variables for the full panel and for different sub panels: Declining economies, Rapidly Growing Economies, Low Income Economies and Middle Income Economies. Generally, the results provide more support for the altruism hypothesis than the insurance hypothesis.

  This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.