This study examines the presence of convergence clubs in per capita electricity consumption across cities in Turkey over the period 2000–2020. Using the Phillips-Sul club convergence test and the clustering algorithm, we explore whether total, industrial, and residential electricity consumption across regions converge to the same long-run equilibrium or diverge. We also analyze the drivers of the club formation after identifying convergence clubs using ordered logit regressions. The findings support the full convergence in residential consumption while separate clubs with different convergence dynamics and steady-state levels are identified in industrial and total electricity consumption. Results indicate that there are two convergence clubs in total electricity consumption which may be clustered as high and low consumption regions. Our analysis also identifies two clusters in industrial consumption with a dominating club. Results from the ordered logit analysis suggest that income, industrial activities, and international trade are positively associated with club formation in total electricity consumption. Overall, our findings suggest that energy policies need to be tailored and implemented on a regional and sectoral basis since energy user groups have heterogeneous convergence behavior.

This study evaluates spatial market and storage efficiency in Turkish lemon markets using switching regime (SR) and threshold autoregression models. Our sample period includes a crucial regulatory reform aimed at improving the performance of fresh fruit and vegetable markets, shortening the production-consumption chain, and reducing retail prices. Using an extended SR model that allows for a gradual transition from the old to the new marketing regime, we test the hypothesis of no structural change in market efficiency and transaction costs in regional markets, including major consumption regions Istanbul, Ankara, and Izmir. Results indicate that the probability of efficient arbitrage regime is relatively higher in the post-reform period, albeit there is still room for improvement. Extended SR models show no significant change in average transaction costs, while threshold autoregression models suggest a slight increase. Furthermore, spatial prices adjust more quickly in the post-reform period, supporting relatively more efficient market functioning. While the regional markets seem to function more efficiently following the marketing reform, it has not led to permanent reductions in retail prices. Finally, our results from storage threshold autoregression models suggest that markets function reasonably efficiently, with no remarkable difference in storage behavior between the pre-reform and post-reform periods.

İktisatta son dönemde üç eğilim göze çarpmaktadır: (1) ampirizmin yükselişi, (2) ekonometrik analizde deney bazlı nedensel çıkarım çerçevesinin genel kabul görmesi, ve (3) makine öğrenmesi yaklaşımının yaygınlaşması ve ekonometri ile daha fazla etkileşimi. Bu çalışma bu ana eğilimler üzerinden ekonometrinin gelişim sürecini tartışmayı ve makine öğrenmesi ile ekonometri arasındaki etkileşimin doğasını anlamayı amaçlamaktadır. Ekonometri göreli kısa tarihi içinde önemli atılımlar gerçekleştirmiş ve yöntemsel/paradigmatik kırılmalar yaşamıştır. Günümüzde ekonometrik analizin temel amacı nedensel ekonomik ilişkilerin sapmasız/tutarlı ve etkin tahminine ilişkin yöntemler geliştirmektir. Makine öğrenmesinde (gözetimli) ise amaç kestirim/öngörü ve sınıflandırma problemlerinin çözümüne yönelik algoritmalar geliştirilmesidir. Ekonometrik analizdeki gibi sapmasız/tutarlı ve asimptotik etkin tahmin yapabilmek geri planda olduğu için makine öğrenmesi problemleri daha başarılı kestirim modelleri verebilmektedir. Ekonometri ile makine öğrenmesi etkileşiminin özellikle bu olağanüstü kestirim başarısı üzerinden şekillendiği söylenebilir. Devam eden bu etkileşim, nedensel çıkarım için yeni ekonometrik yöntemlerin geliştirilmesi ve mevcut olanların iyileştirilmesi ile sonuçlanmıştır.

This paper investigates the relationship between economic adversity and voter participation in Turkish parliamentary elections. We employ a dynamic model of voter turnout using a cross-regional panel dataset covering all parliamentary elections for the 1987–2018 period. In addition to lagged voter turnout to account for persistency in electoral participation, our models include two measures of economic adversity: inflation, and economic growth. Empirical results indicate that inflation has a positive and significant effect on voter turnout, while economic growth is insignificant in all specifications. Moreover, inflation is associated with higher voter turnout in provinces with a denser population, higher urbanization, and more abundant human capital. Regional voter turnout rates tend to persistently support the habit formation argument. Overall, our empirical results are consistent with the mobilization hypothesis, suggesting that adverse economic conditions lead to higher electoral participation by stimulating voters to punish the government.

This paper examines the impact of information and communication technology (ICT) on firm-level productivity in Turkey using a novel longitudinal data set.

This paper investigates club convergence in per capita income across 81 NUTS-III regions in Turkey over the 1987–2017 period.