Entropy in the Assessment of the Labour Market Situation in the Context of the Survival Analysis Methods

Since Shannon's pioneering work, the concept of entropy has been used in many major scientific fields. It is therefore a universal concept but also defined in different ways. Entropy is used in studies of system complexity and to investigate the information content of probability distributions. One of the areas of its applications is human lifespan, i.e., the link between entropy and the methods of survival analysis. These methods are also used in assessing the duration of any socio-economic phenomenon. The aim of this article is to assess the market situation on the basis of the entropy of duration in unemployment. This study determines the Shannon entropy, residual entropy, past entropy, and cumulative residual entropy under the assumption of an exponential distribution of duration. Ward's hierarchical clustering and the Dynamic Time Warping measure were used to analyse entropy and its relationship with the unemployment rate. It was shown that not all of the analysed models determine the entropy of duration in unemployment well for an exponential distribution. It was substantiated that there is a similarity between the formation of the entropy of duration in unemployment and the registered unemployment rate. It is shown that high unemployment rates in the labour market are a destabilising element of the labour market, more so than crises.