5. Conclusions and discussion
This study is one ofthefirstlarger scale empirical studies to provide evidence on the effects ofthe use of performance measurement systems in public sector organizations. Our results provide some interesting insights into the functioning of these organizations. First, we find a positive association between contractibility and performance. This finding is consistent with a large literature documenting the positive performance effects of clear and measurable goals (Locke and Latham, 2002). While the importance of clear and measurable objectives is recognized by NPMadvocates, it is generally seen as a choice variable: if goals are ambiguous, (political) management should make them clear (cf. Dewatripont et al., 1999; Hood, 1995; Wilson, 1989). It is undoubtedly true that goal ambiguity is deliberately created in some public sector organizations for political or self-serving purposes (Kerr, 1975; Hofstede, 1981). But NPM seems to ignore that contractibility is often not a choice variable at all, and that objectives often are inherently ambiguous in public sector organizations (cf. Burgess and Ratto, 2003; Cavalluzzo and Ittner, 2004; Dixit, 1997, 2002).