The determinants of research and development firms' location choice in Turkey: A Nested Logit Model application

The purpose of this dissertation is to analyse the location choice behaviour of the firms in the TURKSTAT Financial and Non-Financial Corporations Research and Development Activities Survey Micro Data Set (2019). For this purpose, the location choice of the firms is analysed empirically within the frame of their ability to identify, assimilate, and exploit knowledge which refers to absorptive capacity. The location choice model is based on the hypothesis that "the firms' absorptive capacity affects their location choice". The empirical analysis is conducted utilizing the Nested Logit Model estimation method. The cross-sectional dataset contains the variables specific to 5871 firms in the Micro Data Set, the sector in which they operate and the region in which they locate their R&D centres. The results show that the absorptive capacity is crucial in firms' location choices. Firms move according to their absorptive capacity to take advantage of the knowledge pool and spillovers in a region. They tend to locate regions in the western part of Turkey, such as Marmara, Western Anatolia, the Aegean and the Mediterranean. This indicates that knowledge is localized and sticky in particular regions of Turkey. Furthermore, firms are sensitive to agglomeration externalities, such as educated and skilled labour, sectoral and regional knowledge flows, forward and backward linkages, and industrial diversity. The tendency of firms towards developed and industrialized regions shows that a more balanced geographical distribution between regions is required. This is thought to be achieved through policies which address the human and physical capital needs of firms and take into account sectoral and regional differences. It is therefore suggested that sectoral and regional policies should be coordinated horizontal policies like health, transport, employment and culture, including education and science, R&D and innovation.