REGIONAL VARIATION IN NEW FIRM FORMATION IN TURKEY: CROSS-SECTION AND PANEL DATA EVIDENCE

Esma GAYGISIZ (METU)
Miyase Yeşim KÖKSAL (METU)

Since the 1980s new firm births especially the formation of small and medium sized firms, SMEs, have been considered as the main source of employment generation in developed countries. For this reason, government policies focus on stimulating startups. But the uneven regional firm birth rates within countries require for the examination of regional characteristics that are influential on firm births. Determinants of spatial variation in new firm formation have been examined for most of the advanced countries; France, Germany (West), Italy, Ireland, United Kingdom, Sweden, Greece, Finland, Spain and the United States in the literature. Since most of the analyses are related to developed countries, we prefer to conduct such an analysis for a developing country, Turkey.

One of the main problems of Turkey in a spatial context is the uneven regional economic development pattern and the gap between the most and the least developed regions continue to incerase. As an evidence, in 1990 the firm birth rate for Marmara Region with the highest rate is 5.5 , whereas it is only 0.1 for the Southeast Anatolian Region with the lowest rate. This substantial variation in new firm formation across regions of Turkey forms the basis of this research.

The aim of this study is twofold. First, the differences in regional variation in underlying firm birth processes between the developed and developing countries -assuming Turkey as a representative of a developing country- are investigated. Second, different start-up processes that are influential on new firm formation in each region are determined. The analysis is restricted to SMEs in manufacturing sector due to the problems in availability of data. The determinants of regional variation in new firm formation in Turkey are examined adopting both the cross-section and the panel data approach.

The findings indicate that there are significant differences in new firm formation processes between developed and developing countries. The results of the cross-section data analysis show that population density tends to be the most significant regressor, whereas variables proxing demand growth are surprisingly not significant. However, in most of the studies for advanced countries demand growth is found to be the most significant process explaining the regional variation in new firm formation. Also, panel data results differ from cross-sectional ones. Panel data evidence shows that the rate of unemployment tends to explain regional variation in new firm formation in Turkey.