Abreu-Matsushima mechanisms can be applied to a broad class of games t
o induce any desired outcome as the unique rationalizable outcome. We
conduct experiments investigating the performance of such mechanisms i
n two simple coordination games. In these games one pure-strategy equi
librium is ''focal''; we assess the efficacy of Abreu-Matsushima mecha
nisms for implementing the other pure-strategy equilibrium outcome. Ab
reu-Matsushima mechanisms induce some choices consistent with the desi
red outcome, but more choices reflect the focal outcome. Moreover, ''s
trengthening'' the mechanism has a perverse effect when the desired ou
tcome is a Pareto-dominated risk-dominated equilibrium. Journal of Eco
nomic Literature Classification Number: C7. (C) 1996 Academic Press, I
nc.