Job and Worker Flows and Labor Market Dynamics in Turkey

Talk by Sinem Ayhan (IOS/IZA) as part of the Research Seminar Series of the IOS Economics Department.

This paper examines the dynamics of Turkey's labor market using job and worker flows. We analyze administrative data from 2006 to 2021, covering all non-financial firms and their employees registered with social security institutions. We investigate how firm age, size, sector, and region influence employment structure dynamics. We also explore the role of foreign trade in job creation and destruction, considering supply-side events and job switches. Our findings cover normal periods and significant shocks like the 2008 global recession, the local currency collapse in late 2018, and the first two years of the COVID-19 pandemic. Turkey's labor market exhibits high dynamism, with gross job reallocation rates of 34-44%, similar to Chile, Colombia, and Morocco but higher than Anglo-Saxon countries, and significantly higher than transition countries. The excess job reallocation rates are particularly high, pointing to a very dynamic and genuine reshuffling of the job structure in Turkey. Quarterly flow rates suggest round-tripping within each year, exceeding the annual rate. While the job creation rate generally exceeds the job destruction rate, we observe a reversal trend during economic downturns, with the currency crisis being particularly notable. Import intensity relates negatively to job creation and positively to job destruction, while export intensity shows no systematic relationship with job reallocation rates. Micro firms with up to 10 employees dominate in numbers and contribute significantly to job creation and destruction, even after accounting for firm age. Job flow rates decrease as firms grow larger or older, and jobs created by micro and small firms have lower persistence. Job turnover accounts for about 70% of total hires and separations, surpassing US and Colombia evidence, due to lower worker reallocation rates in Turkey. In addition, our findings support the firm-size ladder hypothesis, particularly for employees in the smallest firms and firms above a certain scale.