This research involves laboratory experiments to test job search models and econometric analysis of field experimental data to test parallelism of laboratory and field search behavior. In addition, experimental data will be used to evaluate econometric techniques that have been developed to cope with the limited information problems associated with naturally occurring-job markets. The tests include direct tests of the reservation wage property using a new experimental design with precommitment and econometric techniques that use either limited or full information on reservation wages. Experiments will also be conducted taking into account more complicated search environments which will include tests of models with variable search intensity, perfect and imperfect job recall, and search with unknown wage offer distributions. Evaluation of the limited information econometric techniques involves application of these methods to the subset of the laboratory data that corresponds to variables that are observable in uncontrolled markets. Econometric tests of parallelism will use results from field experiments conducted in the naturally-occurring labor market to provide empirical evidence about the qualitative similarity of laboratory search behavior and actual market search behavior. $$$ This project is concerned with testing theories about the job search behavior of unemployed workers using experimental methods. A key element in testing these theories is the minimum wage a worker is willing to accept in taking an available job. These theories will be tested in terms of this reservation wage property by exploiting information on the responses of subjects to actual financial offers received. The research will examine behavior in environments corresponding to labor markets in which individuals have imperfect notions about their job offer prospects, or in which they are able to vary the intensity in which they search for jobs. Econometric techniques have been developed to analyze labor market data in environments where key variables in the worker's decision problem are unobservable. The validity of these methods will be evaluated by determining whether the inferences generated from applying these techniques are contradicted by inferences made on the basis of the more complete information provided by laboratory experiments. This project is important because the results from the laboratory experiments may shed new light on the factors contributing to unemployment.
