Working Paper: NBER ID: w11967
Authors: John Kennan
Abstract: Shimer (2005) pointed out that although we have a satisfactory theory of why some workers are unemployed at any given time, we don?t know why the number of unemployed workers varies so much over time. The basic Mortensen-Pissarides (1994) model does not generate nearly enough volatility in unemployment, for plausible parameter values. This paper extends the Mortensen-Pissarides model to allow for informational rents. Productivity is subject to publicly observed aggregate shocks, and to idiosyncratic shocks that are seen only by the employer. It is shown that there is a unique equilibrium, provided that the idiosyncratic shocks are not too large. The main result is that small fluctuations in productivity that are privately observed by employers can give rise to a kind of wage stickiness in equilibrium, and the informational rents associated with this stickiness are sufficient to generate relatively large unemployment fluctuations.
Keywords: informational rents; wage stickiness; unemployment fluctuations; Mortensen-Pissarides model
JEL Codes: E3; J6; D8
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| small fluctuations in productivity (O49) | wage stickiness (J31) |
| wage stickiness (J31) | significant fluctuations in unemployment rates (J64) |
| small fluctuations in productivity (O49) | significant fluctuations in unemployment rates (J64) |
| informational rents (D89) | wage stickiness (J31) |
| wage stickiness (J31) | creation of more vacancies (J63) |
| small fluctuations in productivity (O49) | creation of more vacancies (J63) |