Sectoral Business Cycle Asymmetries and Regime Shifts: Evidence from Turkey

The aim of this paper is to fit Markov regime switching behavior models to the sectoral GDP growth rates in Turkey for the period 1998: Q1 to 2019: Q2. The findings support the existence of two regimes as low-growth and high-growth for all three sectors. The mean growth rate of the total GDP is closer to the mean growth rate of the industry sector than to the mean growth rate of the agricultural and services sectors. Moreover, the regime volatilities are higher in the low-growth regime for the industry and services sectors and vice versa for the agricultural sector and the total GDP. The results also show that the high-growth regime periods are longer than the low-growth regime periods. Finally, it is observed that there are more frequent fluctuations in the agricultural sector than the other sectors' cycles based on the smoothed probabilities for low-growth regime. Moreover, since 2016 till now, the services sector's regime switching behavior is associated with the low-growth regimes of GDP, which indicates that Turkey's largely free-market economy is driven by the services sector. The findings also show that Markov switching model used in this study provides an advantage to model the nonlinearities in GDP fluctuations which assume different behaviors in different regime periods.