Hostname: page-component-59f8fd8595-klmmf Total loading time: 0 Render date: 2023-03-22T23:40:31.376Z Has data issue: true Feature Flags: { "useRatesEcommerce": false } hasContentIssue true

Why The Government Budget Is Too Small in a Democracy

Published online by Cambridge University Press:  18 July 2011

Anthony Downs
University of Chicago
Get access


In a democratic society, the division of resources between the public and private sectors is roughly determined by the desires of the electorate. But because it is such a complex and time-consuming task to acquire adequate political information, the electorate is chronically ignorant of the costs and benefits of many actual and potential government policies. It is my belief that this ignorance causes governments to enact budgets smaller than the ones they would enact if the electorate possessed complete information. Yet these undersized budgets stem from rational behavior by both the government and the electorate; hence they are extremely difficult to remedy. Furthermore, the resulting misallocation of resources becomes more and more serious as the economy grows more complex.

Research Article
Copyright © Trustees of Princeton University 1960

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)


1 Downs, Anthony, An Economic Theory of Democracy, New York, 1957.Google Scholar

2 For a complete explanation of this theory, see ibid. The government budget is discussed in chap. 4.

3 The remainder of this article assumes a two-party system. Its conclusions are also applicable to multi-party systems, but the corresponding proofs are too complicated to be presented in an article of this length.

4 This explanation of the budget process ignores the effect of government administrative bureaus upon the budget's final size. If self-aggrandizing bureaus were included in the model, each would try to maximize its own income, power, and prestige within the government. Hence it would submit a maximum estimate of its needs to the central budgeting agency (i.e., the directors of the governing party). The bureau might even enlarge this estimate beyond its real needs in anticipation of the budgeting agency's desire to minimize expenditures. Its inflated requests would be bolstered by assertions that all of its spending would pay oft well in votes. Since this process would distort the budgeting agency's information about what expenditures would in fact gain votes, the actual budget would tend to be larger than if bureaus were not self-aggrandizing. However, the central budgeting agency would be aware of the bureaus' inflationary tendencies and would develop outside checks against each bureau's vote-gain estimates. If the governing party failed to make such direct checks with the voters, it would be vulnerable to defeat by more alert opponents. Therefore, the information distortion caused by government bureaus could not be expected to offset the basic tendency for government budgets to be too small.

Another possible impact of administrative bureaus upon die model is their tendency to create situations in which their services are needed; e.g., by building missiles, the defense establishment of country A causes rival country B to counter with better missiles, thereby increasing the need for country A to spend even more money on missiles, etc. Merton, Robert K. describes this process in “The Self-Fulfilling Prophecy” (chap. 7 of Social Theory and Social Structure, Glencoe, Ill., 1949).Google Scholar Since this characteristic of bureaus raises a whole set of fundamental problems beyond the scope of my model, I have made no attempt to account for it in this study. However, a model is under development which contains government administrative bureaus as a set of actors in addition to parties and voters. It is hoped that this model will shed further light on the effects of government bureaucracy.

5 Some readers of this argument may object that spending for such categories as national defense cannot be evaluated in terms of votes but must be decided largely on technical grounds. I do not agree. For example, the United States government chose to abandon maintenance of strong conventional forces and stake the nation's entire defense upon the use of nuclear weapons. This decision was made against the technical advice of Army planners. From statements made by leading government officials at the time, it is clear that the decision was designed primarily to avoid asking the electorate to pay for both nuclear and conventional forces. In spite of the fact that every subsequent Army Chief of Staff has bitterly opposed this policy, the governing party has maintained it because the cost of its alternative is politically unpalatable. Thus in the real world, even regarding national defense, major budgetary questions are usually decided by vote possibilities.

6 In the United States, defense expenditures by the federal government constitute over 40 per cent of the total spending by all federal, state, county, local, and other government units. This figure aoplies to 1954 and is taken from U.S. Department of Commerce, Statistical Abstract of the United States: 1956, p. 401.Google Scholar

7 Where the government does stand or fall on every issue, as in the French Fourth Republic, it can function successfully only if strong consensus exists among the majority. Otherwise it is continually defeated by “Arrow problems.” See Downs, op.cit., chap. 4.

8 This sentence appears to contradict the one preceding it, but in reality they are perfectly consistent. To illustrate, assume that an urban citizen pays $500 per year in taxes toward a government budget which is spent entirely for farm subsidies. Because he has no interest in farm subsidies, he thinks the budget is too large in relation to what he is getting out of it. However, he strongly desires urban renewal, and would be happy to pay $1,000 per year in taxes if the government budget were spent entirely on urban renewal. Thus, in his eyes, the actual budget is simultaneously too large and too small, depending on what alternative it is compared with. It is too large compared with a budget in which those expenditures he dislikes have been eliminated and all others remain the same. Yet it is too small in comparison with a budget in which the expenditure pattern has been changed to what he regards as optimum.

9 If competition is not perfect, he bears some of the cost himself. This fact strengthens the argument that citizens are relatively aware of the sacrifices imposed upon them by taxes, even indirect ones.

10 This is not to deny that there is a great deal of actual waste in government which justly deserves censure. However, many political charges of “waste” are really attacks on production of genuine—but remote—benefits. These attacks are designed to capitalize on rational ignorance for political gain at the expense of the actual benefit of the citizenry.

11 In my ooinion, the elected officials of a democratic government are not significantly motivated to maximize expenditures. Their primary rewards are the perquisites of holding an elective office, and their attention and energies are focused upon overcoming the difficulty of remaining in that office in spite of challenges in every election. However, permanent bureaucratic functionaries in large governments do not have their energies absorbed by the problem of retaining their jobs. Hence they can concentrate on increasing their significance through expanding the size and influence of the departments under them, which usually involves increasing the amount of resources they control. Thus whether the expenditure-maximizing assumption enters a model of democratic government depends upon whether government in the model is simply a team of elected officials, or is a team of elected officials plus a set of permanent bureaucrats. The impact of the latter assumption has already been discussed in footnote 4.

12 However, this is not the only distortion caused by preponderant ignorance. It also encompasses the previously described tendency to create budgets that are too small because voters are ignorant of remote government benefits. The net effect of these two opposing forces is discussed in Section VIII.

13 Many citizens sell their time and labor rather than an objective product. They are therefore interested in policies which affect both (1) the sale of their labor and (2) the sale of the particular products their labor is used to create.

14 The classic example of this asymmetry is the tariff. A few producers manage to get government to set protective tariffs at the expense of millions of consumers, even though politicians seek to maximize votes. This is possible because producers are much more intensely interested in their income than consumers are in the individual prices that face them. See Downs, , op.cit., pp. 253–57.Google Scholar

15 This conclusion and many of the ideas in Section VII were developed in discussion with Gordon Tullock, to whom I am much indebted.

16 This is not the sole function of government, but it is one of the most central.

17 During periods of rising national income, government receipts will increase without any change in tax rates. Assuming that such increases are not accompanied by an inflation which destroys their real value, the government will have greater purchasing power available to it. It might therefore appear that government could increase its spending beyond the “correct” amount without the voters knowing about it. However, this argument ignores two facts. First, the opposition party serves as a “watchdog” ready to call voters' attention to such tendencies. If the governing party tried to increase spending covertly with these funds, the opposition might defeat it by uncovering the added receipts and promising to return them to voters via tax cuts. Second, voters will realize that their absolute taxes are rising, even if meir incomes are also rising. For both reasons, voters will be aware of rising government receipts. Since the governing party has no vested interest in maximizing its spending anyway, it cannot afford to risk antagonizing voters by trying to hide such increments. Hence it will evaluate them by weighing votes, as with any other receipts, and either return them to voters via tax cuts or spend them so as to gain further support. Thus whether the model is conceived of as static or dynamic is irrelevant to its major conclusions.