The Determinants and Consequences of Remittances: With Special Reference to Kosovo

Abstract
The aim of this thesis is threefold: to investigate, from the viewpoint of the home country, the effects of macroeconomic variables and policies on the flow of remittances; and to investigate the microeconomic foundations of the effects of remittances on both household expenditure and the labour force, with a specific application to Kosovo. For the macroeconomic determinants of remittances, a unique dataset is assembled using World Bank data with policy variable data collected from individual countries in the sample. For the implications of remittances on household expenditure and the labour force the United Nations Development Program Remittances Survey is used.
The estimates suggest that the key determinants of remittances/GDP and remittances per capita are real living standards in both the home (negatively) and the main host country (positively). The effect of GDP in the host country is not linear, but increases at decreasing rate. An original contribution of this section is the inclusion of specific policies applied to migrants and their remittances. However, despite the support for such policies in the literature, the policy variables modelled are statistically insignificant across several specifications.
In considering the effect of remittances on consumption, whether the household receives remittances or not is interacted with the log of income to examine if the spending of remittances is different from other sources of income. This variable is statistically insignificant for the share of expenditure on current consumption and education. However, the effect for the share of expenditure on durable goods is positive, albeit small. The model extends the literature by considering the migrant‟s involvement in various ways with the home family as possible control mechanisms for the use of remittances. The results give three significant effects suggesting that, compared to other households, when income increases: households who receive advice on spending remittances increase the share of expenditure on current consumption; households who get more visits decrease the share of expenditure on current consumption and increase the share of expenditure on education.
In addition, addressing the impact of remittances on the labour supply of individuals in Kosovo, the results suggests that remittances per capita are insignificant with regard to working age members being active in labour market and being employed for males and females. However, when the total value of remittances per household is used, the estimate suggests an increase in the probability of males becoming active in the labour market, though the effect is very small.
Overall, the findings suggest that the policies aimed at increasing remittances from migrants are not effective. Furthermore, there is little evidence that changes in income levels for households receiving remittances are spent very differently from the households not receiving them; small differences are found for the durable goods share and if the migrant has some direct involvement with the home family. Finally, contrary to expectations, remittances are not found to affect the labour supply in the majority of the empirical work in this thesis. Such findings may indicate that the policy relevance for remittances is not as important as suggested by the literature for developing countries, especially for those with a similar profile to Kosovo.