This paper introduces productivity-dependent firing costs into an otherwise standard endogenous
separations matching model. We suggest an alternative to the standard fix cost approach and account
for empirical evidence emphasizing that firing costs vary across workers. We show that the model
with firing costs outperformes the model without firing costs and replicates the empirical facts fairly well. Furthermore, we present cross-country evidence that countries with stricter employment
protection have a weaker Beveridge curve relation and surprisingly more volatile job flow rates.