Spatial dynamics of agglomeration economies: Sectoral and regional analyses for Turkey

The spatial concentration of economic activities is a fundamental characteristic of the economic geography of countries and regions. This concept explores the reasons why certain geographic areas host more intense economic activities compared to others. In this context, understanding where and to what extent spatial concentration occurs in various sectors and regions is crucial for policymakers and entrepreneurs. This thesis aims to analyze Turkey's spatial structure using various index values on a regional and sectoral basis. The goal is to contribute to the development of regional development policies and strategies and enhance the understanding of the regional distribution of economic activities. Data from the Union of Chambers and Commodity Exchanges of Turkey (TOBB) and the Social Security Institution Presidency were used for this research. Address information and employment numbers for firms in the ready-made clothing, textile, and leather products sectors were obtained from TOBB's industrial information system. Additionally, sectoral employment data for all sectors, organized based on the NACE Rev. 2 classification and accessible through the website of the Social Security Institution, were also utilized in the analyses. The study employed both clustering-based indices and distance-based methods in its analyses. This approach contributes to the originality of the study and offers a broader and more comprehensive perspective on spatial agglomeration. Within the framework of sectoral concentration analyses, the highest levels of sectoral concentration have been observed in industries such as petroleum and natural gas extraction, airline transportation, manufacturing of other vehicles, and the publication of cinema films and sound recordings. According to regional specialization analyses, the provinces with the least diversity in employment across 88 different sectors have been identified as Batman, Şırnak, Yalova, and Hakkâri. In contrast, provinces such as İzmir, İstanbul, Ankara, Adana, Samsun, and Kocaeli exhibit considerable diversity in their economic structures. Based on analyses using the Kemp function, a distance-based method, accumulations at different distances have been detected in the TR31, TR32, and TRC1 regions. Additionally, using the spatial clustering index, index values for all Level 2 regions in Turkey have been determined, and index components for selected regions have been evaluated in detail. These analyses contribute to a better understanding of regional economic structures.