Annals of the University of Oradea: Economic Science (Jul 2013)

ENTREPRENEURIAL ACTIVITY IN ROMANIA – A TIME SERIES CLUSTERING ANALYSIS AT THE NUTS3 LEVEL

  • Sipos-Gug Sebastian,
  • Badulescu Alina-Daciana,
  • ,

Journal volume & issue
Vol. 22, no. 1
pp. 673 – 682

Abstract

Read online

Entrepreneurship is an active field of research, having known a major increase in interest and publication levels in the last years (Landström et al., 2012). Within this field recently there has been an increasing interest in understanding why some regions seem to have a significantly higher entrepreneurship activity compared to others. In line with this research field, we would like to investigate the differences in entrepreneurial activity among the Romanian counties (NUTS 3 regions). While the classical research paradigm in this field is to conduct a temporally stationary analysis, we choose to use a time series clustering analysis to better understanding the dynamics of entrepreneurial activity between counties. Our analysis showed that if we use the total number of new privately owned companies that are founded each year in the last decade (2002-2012) we can distinguish between 5 clusters, one with high total entrepreneurial activity (18 counties), one with above average activity (8 counties), two clusters with average and slightly below average activity (total of 18 counties) and one cluster with low and declining activity (2 counties). If we are interested in the entrepreneurial activity rate, that is the number of new privately owned companies founded each year adjusted by the population of the respective county, we obtain 4 clusters, one with a very high entrepreneurial rate (1 county), one with average rate (10 counties), and two clusters with below average entrepreneurial rate (total of 31 counties). In conclusion, our research shows that Romania is far from being a homogeneous geographical area in respect to entrepreneurial activity. Depending on what we are interested in, it can be divided in 5 or 4 clusters of counties, which behave differently as a function of time. Further research should be focused on explaining these regional differences, on studying the high performance clusters and trying to improve the low performing ones.

Keywords