Is the Rational Addiction model inherently impossible to estimate?

B-Tier
Journal: Journal of Health Economics
Year: 2017
Volume: 54
Issue: C
Pages: 161-175

Score contribution per author:

0.670 = (α=2.01 / 3 authors) × 1.0x B-tier

α: calibrated so average coauthorship-adjusted count equals average raw count

Abstract

The Rational Addiction (RA) model is increasingly often estimated using individual level panel data with mixed results; in particular, with regard to the implied rate of time discount. This paper suggests that the odd values of the rate of discount frequently found in the literature may in fact be a consequence of the saddle-point dynamics associated with individual level inter-temporal optimization problems. We report the results of Monte Carlo experiments estimating RA-type difference equations that seem to suggest the possibility that the presence of both a stable and an unstable root in the dynamic process may create serious problems for the estimation of RA equations.

Technical Details

RePEc Handle
repec:eee:jhecon:v:54:y:2017:i:c:p:161-175
Journal Field
Health
Author Count
3
Added to Database
2026-01-25