abstract Bruce, Buck and Main (2005) offer two criticisms of agency theory as a valid model of executive behaviour. First, they suggest that because researchers have failed to find a strong empirical link between executive pay and firm performance, and since this research generally rests on models derived from agency theory, then we must question the theory. Second, they suggest that agency theory is under-socialized and therefore lacks generalizability to settings where social solutions would seem to eliminate the agency problem. In our response we make three points. First, agency theory rests on an assumption of self-interest that does not necessarily reflect opportunism. Second, agency theory does not make any reference to pay-performance sensitivity, and the failure of this research can be attributable to a variety of problems with the research. Third, we agree that agency theory does not explicitly recognize contextual factors, but suggest that this abstraction from context, gives agency theory greater generalizability. Finally, we review the UK and German contexts discussed by Bruce, Buck and Main to show that socialized solutions do not prevent the occurrence of agency problems.