The comparative study of public policy once promised to make major contributions to our understanding of government. Much of that promise now appears unfulfilled. What accounts for this decline in intellectual fortunes and change in intellectual fashion? Comparing Public Bureaucracies seeks to understand why. One of the principal answers is that there is no readily accepted and dependent variable that would allow comparative public administration to conform to the usual canons of social research. In contrast, comparative public policy has a ready-made dependent variable in public expenditure.

Peters discusses four possible dependent variables for comparative public administration. The first is personnel—the number and type of people who work for government. Second, the number and type of organizations that form government can suggest a great deal about the structure of government. Third, the behavior of members is obviously important for understanding what actually happens in government—such as the extents to which bureaucracies approximate the budget-maximizing behavior posited by economists. Ginally, the relative power of civil servants in the policymaking process is a major factor in institutional politics in contemporary industrial societies.