The determinants of new firm formation dynamics

Explaining uneven economic growth and development within a country or a state or even within a metropolitan area has drawn attention of many scholars as well as public officials around the world. Most of the theories proposed so far have not been able to explain clearly the processes that underlie economic development and growth, making them unsuitable for policy actions. Regional planners and government officials are more interested in identifying factors that drive economic growth and development at the local or regional level. This paper explores regional factors as determinants of new firm formations (hereafter, NF2) that are central to the phenomenon of economic growth and development. Specifically, it explores regional factors as determinants of NF2 for the Texas manufacturing sector for the 1976-91 period. All twenty-seven (27) Texas metropolitan areas have been used to account for the variation in NF2 over time and space. The results are of practical importance to policy makers, economists, and would be entrepreneurs alike. In the absence of specific theories of new firm formations, this study draws on neoclassical theory as well as ideas of Adam Smith, Alfred Marshall, and Joseph Schumpeter. A historical evaluation of entrepreneurs and associated entrepreneurial innovations is carried out to better understand the role of entrepreneurship in influencing economic development and growth in the present context. Empirical studies of new firm formations, though mostly from European countries, have been reviewed to get a first hand look at current status of research in the area. The problem of appropriate data with which one can measure the key processes undergirding economic dynamism and change is explained in detail. In this context, relative superiority of the data set used in this study to the data sets used in most of the previous empirical studies is explained in greater depth. Using Hicks' micro-level data on establishment entries and exits, this study develops pooled time-series data on entries and exits of firms collected over a period of 16 years (1976-91) using sales tax files for the manufacturing sector. The issues of: (1) the treatment of missing data, (2) the selection of referents for regional and time dummy variables, and (3) violations and remedies of OLS assumptions (heteroscedasticity, multicollinearity, and serial correlation) are addressed to maximize the reliability of regression models developed for further analysis. Fixed-effects regression models with region and time as dummy variables have been formulated to control for region-specific and time-specific influences, respectively. The fixed-effects models are estimated using two different regression estimation techniques, robust regression and estimated weighted least squares. Time-series analyses is employed to test for ten hypotheses proposed, one for each variable included in the model. All regression results are found to be consistent with their respective hypothesis. Robust regression model results indicated a positive and independent influence of mean establishment size, entry rate, exit rate, and per capita bank deposits on NF2; whereas change in unemployment rate indicated a negative influence on NF2. The empirical findings provide strong support for the role of deep churning - turnover and replacement in a business base - in understanding the forces shaping economic development and growth. Moreover, the findings provide a counterpoint to neoclassical theory by suggesting a symbiotic, rather than an adversarial, relationship between small and large firms.