Kovacheva, Penka (2013). Essays on transition economies. Master's thesis / Doctoral dissertation.
This thesis consists of three chapters, which analyze how labor market and policy uncertainty affect education, employment, fertility, and well-being in the context of three transition economies. The first chapter examines the cause of the unprecedented, and still unexplained, 40% decline in total fertility rates across the former Communist Bloc during the 1990s. Using individual-level panel data from Russia and Ukraine, I show that a substantial increase in wage uncertainty in the 1990s triggered the fertility decline, and that a moderate decline in wage uncertainty in the 2000s prompted a partial fertility recovery. I build a dynamic structural model of the joint decisions over fertility and job sector: private, state, and non-work. In the model, wage uncertainty affects fertility because children are illiquid assets. I estimate the key parameters of the model exploiting panel variation in wage uncertainty across job sectors, regions, and time. I construct a measure of wage uncertainty which is a function of the incidence of job dismissal, administrative unpaid leave, and wage arrears. Counterfactual exercises show that lowering wage uncertainty in the private sector to the level of the state sector reduces the fertility gap between sectors by 51-57%. Lowering the levels of the 1990s uncertainty in both sectors to the levels of the late 2000s increases overall fertility by 26-30%. By contrast, I estimate much smaller effects from a traditional fertility stimulus in the form of more generous maternity benefits. These results suggest that in economies with poor labor law enforcement, targeting employment violations (such as wage arrears) can be part of effective fertility policies. More broadly, labor market uncertainty can impose significant negative externalities on irreversible investments such as children. The second chapter, co-authored with Xiaotong Niu, investigates the effects of a 1996 pension crisis in Russia on subjective well-being. We use a difference-in-difference strategy and an individual fixed-effects model, which isolate the pension crisis from other changes in the economic environment and account for individual heterogeneity in reported well-being. We find that the failure to receive one's pension during the crisis not only significantly reduces current life satisfaction, but also depresses expectations for the future. Additionally, the effect of the crisis extends to non-pensioners who live in households with pensioners in arrears: they experience an equally strong and significant decline in life satisfaction, even after accounting for their personal income. We also find that the crisis, despite being a purely monetary shock, affects well-being in ways that go beyond the monetary size of pension loss. Policies aimed to fully compensate for such disruptions in the redistribution system would need to take these externalities into account. The third chapter investigates the dynamics of human capital returns and their effect on wage inequality during Bulgaria's transition to a market economy. I use regression analysis to estimate the monetary returns to different individual characteristics, while accounting for the selection of individuals into the new market economy. I also use semi-parametric analysis to examine counterfactual earnings distributions under several scenarios for how individual characteristics could have evolved over the transition. I find that returns to human capital are not substantially affected by gender, ethnicity, or sector of employment but differ by age. Returns to work experience are stagnating as the market economy places almost no value to experience acquired under communism. Returns to education start low relative to both Western Europe and some of Central Europe but increase throughout the 1990s. These rising returns to education are the single most important driver behind the growth in wage inequality, accounting for over 75% of it. This leads to an interesting policy prescription: little if any policy action is needed si ce the rise in inequality can be expected to level off as subsequent generations get educated under the new system.