The interpretation of the long-standing fiscal illusion hypothesis presented here is that the illusion is the result of attempts by utility maximizing public managers (politicians or bureaucrats) to direct the resources of the community toward themselves. In general, they will succeed if the costs of information about or enforcement of managerial behavior are nonzero, and political competition is imperfect. The consequences of this application of the Williamson theory of the firm, in the context of a median voter model, include the following: 1. There will be an optimal level of illusion for the public manager and median voter simultaneously; and this is consistent with the rationality of all agents. 2. Underestimation of tax-prices by the median voter at this point by itself does not imply a continually growing public sector. 3. Misperception of taxes or expenditures alone does not imply anything about the affect of misperception on the size of the budget. For the same reason, a change in information costs is ambiguous (in the present model) in its effects on budget size, because this will influence both tax and expenditure illusions. 4. The importance of tax or expenditure structure manipulation in determining the size of the budget is an empirical question. It may be that the fiscal illusion hypothesis is empirically relevant, or it may not. Whether illusion collapses over time depends on such factors as the age structure of the electorate and the extent of political competition. It is interesting to note, incidentally, that the real 'hidden persuaders' may well be in the public sector if competition is more imperfect in political markets than in private. Yet, those who complain about the artificial creation of wants by private business call at the same time for greater government intervention.