Sebastian Buhai, Stockholm University / UC

Paper: Valuing Jobs with Unobserved Offers: Identification, Robustness, and Validation

Abstract: What does job mobility reveal about job values, when alternative job offers are unobserved? The empirical starting point is Danish work-environment surveys matched to comprehensive employer-employee registers, combining observed wage and employment histories with workers’ reporting on various dimensions of workplace conditions, including physical environment, ergonomics, skills and discretion, social environment, shift work, and employment prospects. These data motivate the central identification problem in this paper: distinguishing the values workers attach to jobs from their access to alternatives. In a finite-state model of forward-looking search, I show that different job valuations can rationalize the same observed wages, job durations, and transition destinations if offer arrival rates are unrestricted. Next, I characterize sharp identified sets under restrictions on the offer process, as well as quantify how much those restrictions must be relaxed to overturn a valuation conclusion. I further examine how information on accepted and rejected offers for selected origin-destination pairs can substitute for these restrictions. Given specified switching costs and a known relationship between job-value differences and the probability of accepting an offer, a connected set of offer comparisons identifies relative job values without requiring observation of every offer rate. Numerical experiments distinguish identification from fit and show how the choice of offer comparisons affects precision. My analysis thus separates recovering job values from attributing them to individual working conditions, making explicit the additional restrictions needed to interpret mobility as evidence on the value of a specific amenity.