As economists introduce more and more cognitive biases into economic models, an important question is whether we can distinguish these biases empirically. We present a simple model of welfare program participation that nests two well-documented cognitive biases: projection bias and present bias. We argue that agents with present bias and projection bias will exhibit di¤erent attitude toward time limits and other welfare eligibitlity restrictions, both before and after such restrictions are implemented. To the extent that such attitudes can be accurately elicited and measured, we argue that we can use attitudinal data to distinguish present bias and projection bias models. An earlier draft of the paper was presented in the conference Behavioral Public Finance: Toward a New Agenda at the University of Michigan on April 23-24, 2004. We are grateful to George Loewenstein, Edward McCa¤ery, Joel Slemrod and other conference participants for helpful comments. All remaining errors are our own.
Download Full PDF Version (Non-Commercial Use)