Three essays in applied econometrics

Abstract

Institutions, circumstances and interactions between agents shape economic outcomes on the individual and aggregate level. In this thesis I explore three different set ups which combine a theoretical model and an empirical framework to better understand how the wider environment influences behavior and outcomes in markets. The following three papers focus on applications in the areas of economic growth, labor markets and health economics. The global network position of an economy has a profound impact on economic growth. A new measure of economic integration is implemented to characterize economic globalization. Descriptive statistics suggest that this new methodology offers superior possibilities to capture global trends which reflect patterns of interactions between firms and countries. Findings from a modified empirical growth model suggest that a more central global network position fosters economic growth. Robustness checks and alternative estimation strategies address issues of endogeneity and reversed causality in a dynamic panel framework. Social networks and in particular the interaction between applicants, workers and firms influence labor market outcomes. The behavior of firms, workers and applicants during the recruitment process is modeled in a bayesian signaling model which under certain conditions predicts a higher match quality between an applicant and a firm if employee referrals were used. Here, the theoretical model pays special attention to potential incentive problems due to nepotism and favoritism. Empirical results suggest a higher starting wage and a longer duration of the position as well as a different earnings path for workers who learnt about their job through a social network. Individual behavior in terms of consumption depends on the health status. The theoretical concept of state dependent utility functions illustrates that changes in circumstances impact individual behavior such that the health status influences the relative composition of the consumption basket over different categories of goods and services. Results from the empirical framework support this concept and show robust findings for changes in consumption in non durable and semi durable goods which can be linked to the individual health status measured in terms of functional problems to activities of daily living.