Reliable quantitative estimates are not available of: (1) the quality
of civil service performance and changes therein as the result of deve
lopment projects, or (2) the importance of civil service performance f
or various development outcomes. Nonetheless, anecdotal evidence indic
ates that in some countries government performance has indeed collapse
d, with calamitous effects on development. Although poor government pe
rformance is theoretically overdetermined-there are many possible caus
es, which we cannot disentangle in practice-a plausible story can be t
old based on institutional economics, using such concepts as informati
on, incentives, and credible commitment. This version of both problems
and solutions is supported by examples of successful reforms. The art
icle argues that ''institutional adjustment' deserves more considerati
on as a basis for reforms. Two practical examples are discussed in som
e detail: improving incentives in the public sector and strategies to
combat corruption. (C) 1997 John Wiley & Sons, Ltd.