Employment generation and productivity: Evidence from Turkish manufacturing industry

Studies on productivity at the firm level showed significant differences in productivity between large and small companies. These variations could be due to differences in the distribution of factors that affect Total Factor Productivity (TFP) across companies of different sizes, besides differences in the returns on these factors. In this thesis, we analyze the productivity of Turkey using a large dataset of the Turkish manufacturing industry, obtained from the Entrepreneur Information System (EIS), the data shows small firms are less productive than large firms, with a significant productivity gap between them. We use the Blinder-Oaxaca decomposition technique to identify the overall endowment and return effects and quantify each factor's contribution to the productivity gap. While many works use the Oaxaca-Blinder decomposition, our study tackles the identification issue that arises if the model utilizes more than one set of dummy variables, using the solution proposed by Yun (2005) to estimate the Turkish size productivity gap. To calculate TFP, we use the method proposed by Ackerberg et al. (2015), which considers the endogeneity issue that arises when firms seeking to maximize profits can observe their TFP early on and adjust their factor input decisions accordingly. The dissertation tracks 59,560 firms that existed in 2006 to investigate the contribution of firms of different sizes and ages to employment generation, taking into account the differences in survival rates of these firms. The Two-part model is used to estimate the growth rate of employment in 2020 and decompose it across groups of sizes and ages. Keywords: Blinder-Oaxaca decomposition, Identification Issue, TFP, Two-part model