Determinants of household savings behavior in Turkey

This thesis aims to investigate the effects of demographic and socio-economic factors on household savings behavior using a microeconometric technique and to see whether the life cycle hypothesis holds for the household savings behavior in Turkey. Domestic saving and investment play a crucial role in the long run growth of an economy. Developing countries, like Turkey, face significant difficulties in financing their investments, and consequently growth, during a sudden stop or outflow of international capital, which leads to financial crises, as a result of low levels of domestic saving. The dramatic fall in the growth rate of Turkey during the financial crisis of 2008-2009 mostly resulted from outflow of international capital. Turkey regained her growth by new capital inflows after crisis, however the rate has slowed down recently due to improved economic conditions in the U.S. and the Euro zone, which attracts international funds. These events, therefore, emphasized the critical role of domestic saving for the long run growth of Turkish economy. There exist limited number of studies which analyze the determinants of domestic savings in Turkey. Most of them are macro level studies, which analyze the saving behaviors by some macroeconomic indicators. Nevertheless, macroeconomic studies assume a representative household while analyzing the effects of monetary and fiscal factors, which ignores the differences across the household behaviors. Since micro level studies can explain the household savings behavior better via some social, economic, demographic and geographic factors; this study focuses on micro level determinants of household saving, such as age, sex, education level, employment status, marital status, household size and household income. The data drawn from the Household Budget Surveys (HBS) between the years 2003 and 2013, which are published by Turkish Statistical Institute (TurkStat), are analyzed via the instrumental variable quantile regression method, which has been widely used in recent years to estimate the variables like savings which vary across the households and have asymmetrical distribution. This study confers different results from the previous literature, which analyzes the issue at micro-level, as household savings behavior in this study is investigated by taking the dissimilarities among households, who have different levels of saving rate. It is observed that significant determinants of household savings behavior has varied effects on households with different levels of saving rate. Additionally, some variables, which are accepted as the statistically significant determinants of household saving rate by the previous literature, such as education level, household size, can not be identified as significant determinants for some of the saving rate quantiles. It is also concluded that the finding of the life cycle hypothesis, which implies that household saving rate follows a hump-shaped pattern over a life cycle, does not hold for household savings behavior in Turkey. This result most probably stems from the altruism of old-age people and the bequest motive. Moreover, precautionary saving motive does not hold in Turkey with one exception, savings behavior of entrepreneur household heads.