Working Papers

• Ottoman Commercial Agriculture and Its Determinants on the Eve of World War I (with Nurhan Davutyan and A. Önder Hanedar)

• Güzelsoy, H. & Tekgüç, H. (2025). Correcting for the Missing Rich: Imputing Missing Incomes into Household Data Proportional to Wealth Distribution. Luxembourg Income Study (LIS) Working Paper No. 903. 

Work in Progress

• Educational Assortative Mating in Turkey: Implications for Household Income Inequality (with Değer Eryar) 

  • This paper investigates the association between educational assortative mating and household income inequality in Turkey between 2003 and 2019. Income inequality is expected to be higher in countries where the individuals choose their partners according to the same level of educational attainment both at the top and at the bottom of the income distribution (homogamy). Recent empirical studies mostly from advanced economies show relatively minor impact of homogamy on income inequality. One possible factor accounting for the lack of a strong effect in advanced economies seems to be the high levels of female education and employment rates throughout the study periods. In contrast to developed countries, Turkey has experienced a significant rise in educational attainment of women as well as in their wage employment rate over the last two decades. Additionally, the share of singles with tertiary education and above average incomes have also increased in the same period. Therefore, we expect to find a significant association between educational assortative mating and the household inequality in Turkey differently from the evidence for advanced economies. Moreover, theorized relationship between educational assortative mating and household income inequality is concerned with gross (before tax, after transfers) income inequality while in practice most of the empirical studies use disposable (after tax and transfers) income due to data limitations. We analyze micro data from 2003, 2007, 2011, 2015, and 2019 Household Budget Surveys with variance decomposition and counterfactual approaches. We also estimate gross household income in addition to disposable income to examine whether the association between educational assortative mating and household income inequality are similar for both income types.

• Correcting the downward bias in inequality estimates for Turkey with house price data (with Çağlar Karabulut, H. Emre Almaz, and Asya Çalıkoğlu)

  • Household surveys often fail to capture incomes of top earners. Top earners are less likely to respond to surveys and more likely to not answer questions concerning entrepreneurial income, i.e. the “missing rich” problem. The most common method in the literature to solve these problems is the use of data obtained from income tax records. Using tax records for developing countries is problematic in two respects: (i) even summary measures of tax records are not publicly available in most of these countries and (ii) tax evasion is rampant and official tax records are probably not reliable even if they were available. Therefore, in this study, the “missing rich” problem is corrected by using house price data obtained from www.sahibinden.com to estimate top incomes. We estimate household incomes using house prices and append these estimates to the right tail of survey data. We estimate that both the Household Budget Survey (HBS) and Survey of Income and Living Conditions (SILC) undercounted approximately 10% of total households in 2019 and 2020, respectively. When the missing rich households included, the Gini index of household income in Turkey has increased from 0.353 to 0.522 for HBS and 0.402 to 0510 for SILC. We use these corrected income distribution data to estimate (i) the amount of income tax evasion (ii) propose a progressive property tax regime. Our conservative estimate for tax evasion is equivalent to 4.4% of 2019 GDP. We also show that it is possible to raise property tax revenue equivalent to 1.7-1.8% of GDP where majority of this tax is raised from top 5 % of the income distribution. 

• Does more schooling improve health? Evidence from Turkey (with Ezgi Cengiz and Serkan Demirkılıç)

  • We examine the health effects of Turkey’s 2012 education reform, which extended mandatory schooling from eight to twelve years. Using a quasi-experimental design, we exploit this reform to estimate the impact of education on health outcomes in early adulthood. Our results show that the reform increased educational attainment among high school-aged girls, whose schooling levels have historically been lower, while it had no significant impact on boys. Despite this increase in schooling, we find no robust effects of extending compulsory education on a wide range of individual and social health outcomes for women in early adulthood. We discuss potential reasons for these results. 

• Taxless Finance of Local Governments in Turkey: Corrupt but Competent (with H. Emre Almaz)

  • This study examines the impact of the revenue generated by the commercialization of urban lands by local governments in Turkey on local government budgets. Commercialization of urban land can benefit local government coffers via land sales and permitting fees levied on construction activities, especially during periods of rapid urbanization. Land sales and fees levied on such activities can increase local budgets without increasing taxes, i.e. taxless [public] finance. However, allocation of valuable public land for development can be rife with corruption. Hence the seemingly contradictory character of local politics in periods of rapid urbanization: corrupt but competent administrators. Rapid urbanization in China in the last 40 years is considered the epitome of such practices. This study contributes to literature by generating estimates for taxless finance for Turkey. In the empirical analysis, the share of such revenues within the local government budgets is estimated. For that, balance sheet data is combined with building usage and building permit data for the period 2007-2022. First of all, a real increase in local revenues until 2018 and sharp downturn afterwards parallel to building usage trends were observed. Then revenue shares from construction related activities estimated: average share of land commercialization related revenues are in the range of 6-46% for Istanbul municipalities. As a benchmark, Municipal revenues from land commercialization accounts for 16-46% of certain local government revenues in China in 2000s (Lin & Zhang, 2014).

• Alternative Working Arrangements in Turkey: A Study of Remote and Hybrid Job Outlook (with Atakan Açıkgöz and Gozde Corekcioglu)

  • This study explores the current state of remote and hybrid work opportunities in Turkey using job posting data from kariyer.net, examining how factors such as province, industry, occupation, experience, education, and task requirements influence flexible work options. Despite the post-pandemic, flexible working arrangements remain a critical element of modern labor markets and a core part of job contracts. First, we employ a task-based approach through Natural Language Processing (NLP) techniques, specifically Term Frequency - Inverse Document Frequency (TF-IDF) with cosine similarity, to categorize diverse and non-standardized job postings into a manageable number of job requirements. Then, we analyze job category requirements' impact on remote work availability. We find that industry, nature of tasks, education coupled with experience matter. The findings indicate that industries such as technology, finance, and professional services offer more remote options, along with jobs requiring higher education coupled with greater experience are more likely to be flexible. Additionally, jobs demanding interactive skills correlate strongly with remote work, while manual tasks remain on-site. This study provides valuable insights into labor market flexibility in a post-pandemic world, particularly in the context of an emerging economy like Turkey. We conclude that distribution of remote and hybrid work opportunities in Turkey align with global patterns in key dimensions albeit at lower levels. 

• 4+4+4 Compulsory Schooling Law and Its Effects on Schooling and Employment (with Başak Yılmaz)

• Ulusal Hesaplarda Emek, Toprak ve Sermaye Geliri [Labor, property, and capital shares in national accounts for Turkey] (with H. Emre Almaz and A. Göç)