Journal Articles
In progress
Geography and Economic Development: Subnational Evidence from Türkiye [with Seda Başıhoş]
The model economy features two groups of workers. One group is working in complex and/or cognitive tasks, implying that industrial robots or artificial intelligence technologies cannot compete with them. The other group faces the automation risk because workers in this group can supply their labor endowment for only the routine and manual tasks. Downward (real) wage rigidity implies that, in general, the equilibrium is characterized by involuntary (technological) unemployment of workers in the second group, i.e., the one that is subject to the automation risk.
Completed or submitted
Türkiye’de Eğitim ve Siyasi Kültür: Birey Düzeyinde Ekonometrik Bulgular
Education and Political Culture in Türkiye: Individual-Level Econometric Evidence
Eğitim siyasi kültürü nasıl etkiler? Bu araştırma, Dünya Değerler Araştırması ve Avrupa Değerler Çalışması'ndaki tüm Türkiye gözlemlerini birleştirerek, eğitimin, (i) post-materyalist değerler, (ii) seküler değerler ve (iii) özgürleştirici değerler üzerindeki etkisi için basit ekonometrik modeller tahmin ediyor.
How does education affect political culture? Combining all Turkey observations from the World Values Survey and the European Values Survey, this research estimates simple econometric models for the impact of education on (i) post-materialist values, (ii) secular values, and (iii) emancipative values.
Quantifying the Unseen: Epidemiological Underestimation Problem for COVID-19 [with Ayça Tekin-Koru]
Abstract. Reliable epidemiological data is a prerequisite for meaningful economic analysis of pandemic-related policies, as it provides the foundation for evaluating public health measures and their economic impacts. In Türkiye, the government did not disclose the number of all confirmed COVID-19 cases for several months after the relaxation of initial mobility restrictions in June 2020, creating significant challenges for assessing the economic and health tradeoffs of these policies. This paper addresses this issue by developing a system dynamics approach that can identify and quantify epidemiological underestimation under extreme data limitations. Our simulation algorithm builds on a nonlinear dynamical model that explicitly accounts for individuals that are exposed but not yet infectious and requires only a few reliable data points. Results imply large deviations between official and estimated figures, and counterfactual experiments show that social distancing, if practiced well and long enough, would have been highly effective for the containment of COVID-19.