Using survival analysis in economics

The main objective of this paper is to examine methodological and applicative problems of survival analysis in the analysis of socio-economic phenomena. Although at the beginning the survival analysis was used to study death as an event specific to medical studies, as from the '70s these statistical techniques have been increasingly used in economics and social sciences. Besides the fact that survival data are not normally distributed, they often contain incomplete information, censored subjects. Censoring of subjects may be on the right or left. It is vital to include censored subjects in the statistical analysis. But, according to Greene (2003), a very large number of censored subjects may affect the accuracy of statistical tests. We presented in this paper methodological aspects of Kaplan-Meier analysis, and statistical significance testing for the resulted survival curves. We have also concentrated on the Cox regression, and we have set out the concept of hazard, baseline hazard, hazard rate, hazard rate interpretation. An application of the survival analysis in unemployment is presented.