How not to extend models of inequality aversion

Several authors have made attempts to improve the explanatory power of models of inequality aversion, in particular the one by Fehr and Schmidt (1999), by adding concerns for total surplus or efficiency. In this note, I point out that these attempts are misguided because they are equivalent to a much simpler change, not requiring an additional parameter, unless we simultaneously consider games with different numbers of players. In the latter case, however, such an approach yields implausible predictions.

► Some authors suggest combining inequality aversion with efficiency concerns.
► I show that this approach is theoretically flawed and empirically useless.
► A simpler change has typically the same explanatory power.
► When the approach adds explanatory power, its implications are implausible.