The first essay of this dissertation examines the role of teacher characteristics in schools on student outcomes using datasets from TIMSS 1999 and TIMSS 2007 international tests. Taking an advantage that students have to take both mathematics and science subjects from different teachers, I use the method of First Difference (FD) analysis in order to remove the potential biases between teacher attributes and unobserved student characteristics. The findings show some contradictory outcomes between the FD analysis and ordinary least squares (OLS) analysis. The second essay looks into the problem of recruitment of well-qualified high school and college graduates to work as primary and secondary school teachers. I compare teacher salaries and benefits vis-à-vis other mathematics and science-oriented professions namely medical professions, engineers, accountants, scientists and nurses. In addition, I compare incomes between people who graduate from teacher colleges and non-teacher colleges. Using data from Thailand Labor Force Survey from 1985 to 2005, I find that teachers are the most poorly paid of all professions, including nurses. The difference in terms of an opportunity cost between male and female teachers is also striking. Among the graduates from teacher colleges, male graduates earn more than their peers if they chose other occupations whereas female graduates earn less if they make other choices. The third essay looks at the reasons teachers choose part-time jobs, the type of jobs they choose, and the amount of income they receive from these jobs, as well as factors influencing these decisions. I find that approximately 20-25% of Thai teachers participated in moonlighting activities. The majority of them have part-time jobs including tutoring, selling food and other products, and farming. Low salaries and high level of indebtedness are the most important factors associated with the increased likelihood of having a part-time job. However, economic status does not correlate significantly with their decision to tutor as their part-time job.

In order to fulfil the requirements for the degree of Doctor of Philosophy in Economics and Finance, at the University of St.Gallen, after the completion of the course-work and research phases, I hereby to provide the cumulative dissertation. The dissertation consists of three essays on macroeconomic policy, development economics and economic growth. The first chapter examines empirically the relationship between the cyclical pattern of fiscal policy in democratic developing countries and their quality of political institutions. This chapter uses updated data to analyse 63 developing countries from 1980 to 2013 and robustly shows that pro-cyclical fiscal policy exist in both democratic and non-democratic developing countries. The main contributions of this chapter are controlling the endogeneity issue by using the instrumental variable method and investigating the interactions between political institutions variables and the cyclicality of fiscal policy. The results suggest that an improvement in the level of institutional quality plays an important role to restrain pro-cyclical fiscal policy. These effects are larger in democratic countries than non-democratic ones. Additionally, the maturity and stability of a democratic regime influence in restraining the implementation of pro-cyclical fiscal policy. The second chapter studies the effects of demographic transition and unfunded pension system in affecting the economic growth of Thailand. The chapter adds the aspect of the informal sector and investments in children's education into the Overlapping Generation (OLG) model. We modify the model to allow individuals heterogeneity in their ages and exogenously assign to work either in the formal or informal sector. The results from our model suggest that the non-contribution pension system or old-age security schemes hinder the economic growth of the economy with a high share of the informal sector and low fertility rate like Tha.