(FORTUNE Magazine) – Washington will be buzzing in the months ahead with talk about balancing the budget, cutting back government outlays, and pruning federal programs, but here's a prediction: Until Congress dramatically changes the rules for legislating federal spending, it will never make much progress toward any of those laudable goals. The main reason the federal budget is out of control is that legislators' incentives are all wrong. Under the current budgeting system, Senators and representatives--no matter what their party or ideology--have an incentive to spend, spend, and spend some more.

What if you went to a restaurant with 14 friends, and all 15 of you asked for separate checks? Then you wouldn't order a $4.50 margarita unless you valued that margarita at $4.50 or more. Maybe a glass of ice water would do. Similarly, you would value any other items you ordered--appetizer, main course, or dessert--at least as much as the price you had to pay for them. Economists call this efficiency.

But now imagine that you and your 14 pals ask the waiter for one check and agree to each pay 1/15 of the bill. I guarantee that your order will be dramatically different, and entirely in one direction: You will order more expensive items. Why? Consider how your incentives have changed. Whereas separate checks made you pay $4.50 for a drink, paying only 1/15 of the bill means the drink now costs you only 30 cents. Similarly trivial is the cost to you of an expensive appetizer, main course, or dessert. Bon appetit!

You are not alone. Virtually everyone at the table will order items that he or she values much less than the cost--because none of the diners pays the full cost. Of course you all will wind up paying more in total than if you had asked for separate checks and chosen more parsimoniously, but the cost to you of a particular item that you order will be much less. Put it another way: If everyone else pops for the $4.50 margarita, why shouldn't you have one too?

What does this have to do with Congress? Everything. Congress is like that 15-person group in the restaurant. Fifteen committees in the Senate and 17 in the House of Representatives make the federal government's spending decisions. Not surprisingly the members of those committees typically represent districts or interests that want spending to increase. The House Agriculture Committee, for example, is heavily loaded with members from farm districts.

Picture the House Ag committee deciding whether to spend more on farm programs. God knows--or at least agricultural economists know--that most subsidies to farmers are inefficient. Virtually everyone in the country pays for those subsidies, partly through higher taxes and partly through higher food prices. Most of the benefits, however, go to the constituents whom the Ag committee members represent. Result: The Ag committee votes to increase farm spending. That committee behaves like you when trying to decide whether to order a $4.50 drink that costs you only 30 cents. Oh, waiter!

"By spreading responsibility for spending authority among so many committees, Congress has created a 'tragedy of the commons,' " says John Cogan of the Hoover Institution, a former official with the Reagan Administration Office of Management and Budget and an expert on the budget process. He is referring to the principle that land owned in common is treated worse than privately owned land. Just as ranchers have little incentive to keep their cattle from grazing on pastures that are owned in common, so congressional committees have little incentive to cut back the programs under their control.

It wasn't always so. For long periods of U.S. history, Cogan notes, just one House committee and one Senate committee made spending decisions. The committee members knew that a dollar not spent by them was a dollar not spent by the government, period. The result was much more restraint in federal spending. Between the 1790s and 1885--the first period of centralized budgeting--the budget deficit averaged only 0.26% of gross national product. But in the mid-1880s, the House transferred much of the Appropriations Committee's spending authority to other committees. Some legislators warned--correctly--that the change would usher in a period of extravagance. By 1893 federal spending was 60% higher than in 1886. After the Senate decentralized its budget process in 1899, spending shot up another 35% between 1900 and 1916. During the whole period of decentralized budgeting, from 1886 to 1921, the deficit as a percent of GNP more than doubled.

In 1921 both the Senate and the House, seeing the tragedy of the commons that had been created, moved back to centralized budgeting. It worked. Between 1922 and 1931 the government ran a budget surplus, averaging 0.77% of GNP. But Congress decentralized the budget process again during the Great Depression and spread spending authority to more committees in the 1960s and 1970s. The result? Average budget deficits of 3.7% of GNP from 1932 to 1995.

Today many Congresspeople, especially the Republican freshmen, have a strong and admirable commitment to cutting government spending. But unless they recognize the perils of decentralized budgeting, they will find themselves on the losing end of many budget-cutting fights, and often will themselves fight to preserve unwise federal spending. Early this year, for example, the House Ag Committee voted to exempt food stamps from the Republicans' block grant transfers to the states. The block grant program, which would transfer funds to the states and save the federal government billions in the process, is integral to the Republicans' plan to eliminate the deficit over seven years. Carving out food stamps will make that harder to do.

The solution is to centralize the budget process once again. A single appropriations committee in each house could then put together one omnibus spending bill for an up-or-down vote.

Paradoxically the deeper and more comprehensive the cuts in spending, the easier it would be to get the political support for such a reform. If the cuts are deep enough, then most beneficiaries of most government-spending programs could be persuaded to go along. Why? Because government is stunningly inefficient. In almost all its activities, it takes $1 from the private sector, churns it through the government factory, and produces, if we're lucky, something worth 50 cents. So if we cut the budget by, say, $100 billion a year, the gain to the winners--the taxpayers--is $100 billion plus the saved costs of tax collection. The losers' loss is in the neighborhood of $50 billion. Net gain to the society: more than $50 billion. The more programs that are hit with cuts, the vaster the majority of individuals who would be net winners. To make the win even more visible, Congress could combine the spending reductions with a $1 tax cut for every $2 spending reduction.

Would there be net losers? Sure. But the most likely losers are not farmers, Amtrak riders, or the elderly. The sure losers are the lobbyists, government employees, and other Washington insiders who have invested in the political process and honed their skills at the negative-sum government game.

"The state," wrote 19th-century French economic journalist Frederic Bastiat, "is the great fictitious entity by which everyone seeks to live at the expense of everyone else." Although Bastiat may have exaggerated, the federal government is now so big that, today, literally everyone does benefit from some government program. What's gone is the fiction. We all gain from government, but we almost all lose a lot more. Let's cut government down to size and share the gains.

David R. Henderson teaches at the Naval Postgraduate School in Monterey, California, and is an adjunct fellow at the Hoover Institution.

Reporter Associate Lenore Schiff