We address the post-entry performance of new Portuguese firms by investigating the structural characteristics survival, using both non-parametric methods and semi-parametric methods, for the seven regions of Portugal (NUT II). We also provide a disaggregated analysis by sector and size class, for the period 1985-2007. In order to approach the prevalence of some stylized facts and determinants of new firm survival, a new entrepreneurship database was produced, using the administrative data of Quadros de Pessoal, a linked employed-employee micro-dataset, following the Eurostat/OECD´s (2007) internationally comparable business demography methodology ‚ÄúManual of Business Demography Statistics‚Äù. This has the advantage of international comparability, in line with the ‚ÄúEntrepreneurship Indicators Programme‚Äù (OECD/Eurostat 2008 and 2009). We thus obtained a specific subsample, consisting only of the population of active enterprises with at least one paid employee, the so-called employer enterprise population. In the non parametric estimation, we use Kaplan-Meier survival functions and the Nelson-Aalen hazard rates. The semi-parametric estimation (log-log complementary model and Cox proportional model) relates the survival capacity with seven explanatory variables and sector and year dummies. The explanatory variable are: firms´ start-up size, the number of employees reported at the year of measurement, firm‚Äôs entry rate, degree of concentration, growth, Entry Rate times Growth and Turbulence. The use of the proportional Cox Model allows us to observe an increase in the risk in the evolution of firm survival over time, when compared to the non-parametric estimation of Kaplan-Meier. By correcting heterogeneity with a different specification model, we obtain stronger magnitudes of the hazard ratios found previously. The main contribution of this work is the application of a recent internationally comparable methodology for entrepreneurship to provide a multidimensional regional overview of firm and survival dynamics for all NUT II regions, over an extended period of time, while guaranteeing regional, national and international comparability. To our knowledge, there is not such a study that contemplates all this aspects and grants international comparability, over such a long time frame, made for any country.
