Risk aversion in the large and in the small

Estimates of agents' risk aversion differ between market studies and experimental studies. We demonstrate that the estimates can be reconciled through consistent treatment of agents' tendency for narrow framing, regarding integration of background wealth as well as across risky outcomes: Risk aversion is similar whenever similar degrees of narrow framing is assumed in either setting. Risk aversion, narrow framing, background wealth, laboratory experiments, market studies, equity premium puzzle