6 Conclusion
A model is proposed where the cost of unemployment to society considers an additional factor, the psychological cost of unemployment. This factor is shown by empirical Öndings to be highly signiÖcant, on a level comparable to e§ects of ill health to an individualís wellbeing. More empirical evidence suggests that this cost is higher the less unemployment there is around, indicating that the high volatility of unemployment also a§ects the volatility of this psychological cost throughout the business cycle. I incorporate those facts in a DSGE model with search unemployment and nominal rigidities in the form of a social norm and Önd that it improves the performance of the model signiÖcantly. A lower rate of unemployment within a household increases the support of the norm and puts greater pressure upon the individual unemployed. This extra psychological cost imposed by the norm counterbalances the bargaining power gained by decreasing unemployment duration during an expansion, preventing the wage from crowding out vacancy creation and boosting the modelís ampliÖ- cation mechanism. The unemployment norm provides enough ampliÖcation to the model to match second moments in the data and can be used to provide ampliÖcation in any search and matching framework. The nominal rigidities (Calvo contracts), especially in the labor market, help account for other features of the data such as persistence. Price rigidity in the intermediate sector not only contributes in creating ináation inertia but also aids in keeping the wage response low and smooth. In a monetary model with no nominal rigidities in the labor market, the wage implied is too volatile.
The model can match the impulse responses implied by a structural VAR on quarterly US data from 1967:Q3 to 2012:Q1 and can be used for optimal monetary policy analysis. The goal is to match the data by assuming moderate degrees of price rigidity and without the aid of wage rigidity or unreasonable calibration, mechanisms which in any case have been heavily criticized. Among others, I address the Shimer puzzle by employing monetary shocks while avoiding stickiness in wages and high unemployment beneÖts. This paper attempts to create a monetary model that can be used for policy analysis, as the cost of unemployment incorporates also the associated psychological cost, which is extremely important to workers. This speciÖcation makes the trade-o§ between ináation and unemployment more challenging and should be important for any central bank conducting monetary policy.
