Abstract
The Diamond-Mortensen-Pissarides model has been the primary workhorse for analyzing the dynamics of unemployment, vacancies, and market tightness over the business cycle. However, it predicts a near-perfect comovement between these variables and labor productivity, whereas the empirical correlation is only mild. We resolve this discrepancy by extending the model to incorporate sunk entry costs and finitely elastic vacancy creation, and by carefully distinguishing between business opportunity destruction and match separation as distinct sources of job loss. These features render vacancies a partially predetermined, positively valued stock variable. If the destruction rate is low, then most vacancies are inherited from the past and reflect historical rather than current productivity, breaking the tight unemployment-productivity link, while preserving strong correlations among labor market variables. We show that, when calibrated to information on job turnover and recall rates, the model reproduces the empirical contemporaneous and dynamic correlations between labor market variables and productivity while preserving the strong correlation between unemployment, vacancies, and the market tightness observed in the data.
| Original language | English |
|---|---|
| Article number | 105205 |
| Number of pages | 52 |
| Journal | Journal of Economic Dynamics and Control |
| Volume | 181 |
| Early online date | 3 Nov 2025 |
| DOIs | |
| Publication status | Published - Dec 2025 |
User-Defined Keywords
- Entry costs
- Unemployment
- Aggregate fluctuations
- Dynamic correlations
- Job destruction