Abstract
A substantial literature exists on the economic analysis of bureaucracy, which includes formal models of bureaucratic behaviour by Niskanen (1971) and Dunleavy (1991). This article develops hypotheses from those models about changes in size of bureaux over time and tests them against data on the Australian federal budget sector in 1982–83 and 1991–92. The models predict that bureaucrats usually will be able to influence the size or structure of their bureaux in line with their personal preferences. However, the Australian data suggest that the extent of such influence may be less than is assumed in either model and that the preferences of politicians are more influential than the models allow. One conclusion is that, if bureaucracy is to be modelled, a principal‐agent approach will have more to offer than the neo‐classical maximizing framework adopted by Niskanen in particular. A further conclusion, at odds with those of Niskanen and weakening his case for wholesale privatization, is that bureaucratic dysfunctions may be amenable to reform of the political and bureaucratic rules and systems.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
