A Case of Framing Effects: The Elicitation of Time Preferences.

We compare three methods for the elicitation of time preferences in an experimental setting: the Becker-DeGroot-Marschak procedure (BDM); the second price auction; and the multiple price list format. The first two methods have been used rarely to elicit time preferences. All methods used are perfectly equivalent from a decision theoretic point of view, and they should induce the same ‘truthful’ revelation i dominant strategies. In spite of this, we find that framing does matter: the money discount rates elicited with the multiple price list tend to be higher than those elicited with the other two methods. In addition, our results shed some light on attitudes towards time, and they permit a broad classification of subjects depending on how the size of the elicited values varies with the time horizon.

Time Variation in the Dynamics of Worker Flows: Evidence from the US and Canada

VAR methods have been used to model the inter-relationships between inflows and outfl ows into unemployment and vacancies using tools such as impulse response analysis. In order to investigate whether such impulse responses change over the course of the business cycle or or over time, this paper uses TVP-VARs for US and Canadian data. For the US, we find interesting differences between the most recent recession and earlier recessions and expansions. In particular, we find the immediate effect of a negative shock on both in ow and out flow hazards to be larger in 2008 than in earlier times. Furthermore, the effect of this shock takes longer to decay. For Canada, we fi nd less evidence of time-variation in impulse responses.