We study .rms.investment in internal control to reduce accounting manipulation. We
.rst show the peer pressure for manipulation: one manager manipulates more if he suspects
reports of peer .rms are more likely to be manipulated. As a result, one .rm.s investment in
internal control has a positive externality on peer .rms. It reduces its own manager.s manip-
ulation, which in turn mitigates the manipulation pressure on managers in peer .rms. Firms
don.t internalize this positive externality and thus under-invest in their internal control over
.nancial reporting. The under-investment problem provides one justi.cation for regulatory
intervention in .rms.internal control choices.