When industries decline, workers often lose their jobs. Demands for compensation often arise when political decisions, e.g. international trade deals or environmental regulations, are at the origin of the decline. This requires a calculation of the welfare costs related to job loss.

Using a simple model of probabilistic transitions, we account for the fact that the value of jobs depends on individuals’ age, experience, occupation. Based on exhaustive administrative micro-data on employment biographies, we determine probabilities for individuals of different characteristic of becoming or leaving unemployment, moving between jobs and retiring. We also make transitions dependent on the business cycle.

We apply the model to assess the welfare costs of job loss in the German coal industry. We contrast predicted income streams for different workers under the alternative scenarios of a rapid coal exit versus business as usual. The results also enable policy-makers to assess the costs and benefits of policy instruments aimed at compensating workers, such as early retirement policies, cash transfers or job guarantees.