(FORTUNE Magazine) – MOST Americans don't think they need an economist to tell them the nation has been underinvesting in its infrastructure. They can feel the evidence when they bounce through a pothole and see it in the litter that blights highways and streets. Says Alcoa CEO Paul O'Neill: ''Every time I come back to the U.S. and drive from LaGuardia or Kennedy into New York City, it's like I'm arriving in a Third World country.'' That sinking feeling is rapidly giving rise, at least in some quarters, to a new conventional wisdom: One sure-fire way to help cure what ails the U.S. economy would be to dramatically boost public investment, both in the nation's physical capital -- things like roads, bridges, airports, and sewers -- and in its human capital, through higher spending on education and training. President Bush proudly hails as his own the six-year, $151 billion highway bill that Congress enacted last year and promises to spend at least $2 billion a year more on training. The heart of Bill Clinton's economic recovery package is a four-year, $220 billion investment plan that includes major programs for training and college aid, and $80 billion in new spending on the physical infrastructure. Claims Clinton adviser and Harvard political economist Robert Reich: ''The big scandal of the federal budget is not the deficit per se, but the fact there's such a big deficit and so little public investment.'' Trouble is, it ain't necessarily so. Remember the supply-siders? Those ardent tax cutters of the early 1980s, you will recall, also had a valid point -- taxes were too high. But they overstated the problem, oversold their prescriptions, and by swelling the budget deficit, made a bad thing worse. Without some radical rethinking, the emerging consensus on infrastructure could easily lead the country down the same sorry path. Let's start with how much we're spending. If you define infrastructure broadly to include not just bridges and highways but education and training as well, real annual public spending in the U.S., far from falling, has been climbing steadily; this year it will hit an all-time high of $425 billion. (For more on investing in human capital, see the story on jobs). Focus on physical capital alone, and the picture is more mixed. In 1992 total federal, state, and local government spending on infrastructure will amount to roughly $77 billion, or 1.29% of GDP. That's still a constant dollar record. But as a percent of GDP, it's below the peak of 1.74% reached in 1964, though up from the trough of the early 1980s (see chart). What was the impact of this modest decline? The most serious charge, levied by economist David Aschauer of Bates College in Lewiston, Maine, is that U.S. productivity growth -- the engine that lifts any country's living standards -- would have been 50% higher had infrastructure investment not fallen off during the 1970s and 1980s. Even if Aschauer overstates his case, the bridge- and-tunnel crowd argue, he has surely fingered the right culprit. After all, they say, look at Japan and Germany, the perennial champs of productivity growth among the major industrial powers. During the 1980s Germany's annual rate of public investment in new buildings, highways, and railroads averaged 3.7% of GDP. The comparable rate in Japan? An astonishing 5.7%. How could differences of that magnitude not explain a big chunk of America's productivity problem? Here's how. In a report last year, the Congressional Budget Office -- a research arm of the legislature and thus no hotbed of doctrinaire conservatism -- pointed to a slew of studies that contend productivity growth may be the cause of increased public investment rather than a consequence of it. That's because flush times make more taxpayer money available for all kinds of good works. The CBO's findings were echoed in a 1991 study by economists Robert Ford and Pierre Poret of the Organization for Economic Cooperation and Development (OECD) in Paris. They examined the link between infrastructure spending and productivity growth in 12 countries and found that while infrastructure spending had slowed in every one during the 1970s, productivity growth had decelerated in only half. Nor is a look at gross spending numbers from the past any guide to how much public money a nation should be investing. During the late 1950s and early 1960s, the rapid rise in U.S. infrastructure spending was fueled by a massive one-time injection to build the 40,000-mile interstate highway system. Says Brookings Institution economist Alice Rivlin: ''No one's going to suggest we do that again.'' What's more, say the OECD's Poret and Ford, the U.S. may not need to go back to the level of spending that prevailed 20 years ago because its economy is shifting away from manufacturing industries, which depend heavily on public infrastructure, to service industries that don't. Finally, it's always wise to hold on to your wallet when policymakers simply assert the need to spend large sums of money without justifying, on a case by case basis, where that money should go. Last year's highway bill, for instance, contained $6.5 billion for over 500 so-called special projects picked not because of their merit but because of their location in strategic congressional districts. The pell-mell Reagan defense buildup of the 1980s was rife with similar excesses. Fear of a giant civilian pork barrel in the making is one reason only 17% of CEOs surveyed in a recent FORTUNE poll named increasing public investment spending as a priority for government action. THAT SAID, it remains incontestable that a big chunk of the nation's physical infrastructure urgently needs repair or upgrading, as disasters ranging from the incessant closure of New York City bridges to last spring's flood in Chicago, caused by a collapsing river retaining wall, attest. But at a time when finances at all levels of government are seriously strained, what we should be debating isn't the level of spending required but which additional investments promise the greatest returns. Politicians naturally prefer grandiose new building projects. But an honest reckoning of costs and benefits suggests that thinking small would generally be far more beautiful. Example: According to the CBO, the U.S. would achieve the highest real annual return -- up to a stunning 40% -- by investing additional public money in the mundane task of filling in potholes, or $ ''pavement deficiencies'' as they're euphemized in some government analyses. Such repairs, it figures, would cost roughly $4 billion a year more than what's currently budgeted for highway spending. Sizable real returns could also be realized from new spending targeted at relieving congestion. The Department of Transportation estimates that traffic jams cost Americans $9 billion a year in time and gas, while air-traffic delays impose $5 billion in annual costs. Anyone who's had to endure Los Angeles freeways that turn ten-mile drives into two-hour commutes or the taxiway at Dallas-Fort Worth Airport, which at 6 P.M. resembles a Boeing parking lot, would say amen to that finding. But floating a raft of construction bonds is no solution. As urban planners have learned to their horror, open a new stretch of roadway or a couple of new lanes, and commuters and truckers inevitably switch from less preferable routes and times, filling roads to capacity again far sooner than anticipated. At the same time, the cost of building new capacity is increasingly prohibitive. Boston, for instance, is spending $5.8 billion just to bury a portion of downtown highway that at rush hour is one of the nation's busiest and to build an adjoining tunnel under Boston Harbor to Logan Airport. The neighbors may object too. In Chicago, plans for a third airport have stalled mainly over South Side residents' opposition to having their neighborhoods paved over for runways and terminals. ONE WAY OUT of this bind would be to stop granting commuters free use of America's roads. Argues Brookings Institution economist Clifford Winston: ''Make oranges cost nothing, and people will always want more of them. You can say the same thing about the demand for roads at peak hours.'' Instead, Winston proposes establishing a system of tolls on congested highways, with the highest charges levied at peak hours. Trucks, the main source of road wear, would be charged to compensate for the damage their axle- weight inflicts. That would lessen the country's maintenance burden by creating incentives for haulers to place less weight on each axle or to have them shift from trucks to trains for long-haul transportation. The idea of charging directly for road use, instead of paying for it indirectly through mechanisms like gasoline taxes, has been kicking around since the 1950s. What makes it practical now are technological breakthroughs that eliminate the need for toll booths. For instance, sensors can be placed underneath a car's license plate and motorists recorded as they whizz by, later receiving a bill in the mail. Makers of the new systems include Amtech, an electronics company in Dallas, and a partnership between AT&T and Mark IV Industries of Amherst, New York. This technology already is in use on the North Dallas Tollway and is being tested in New York. A similar market-based strategy would reduce delays at airports. Planes landing during the morning and evening crunch could be charged higher fees than those arriving at other times. (They now pay only according to their weight.) This would encourage non-commercial planes carrying few passengers -- the enthusiast in his Cessna, for example -- to switch to off-peak hours. This kind of innovative management of the infrastructure in which the U.S. already has roughly $1.5 trillion invested could deliver a lot of bang for very few bucks. Winston estimates that combining congestion pricing with new spending of just $2.7 billion a year to thicken road pavements and build new airport runways would yield a return of $25 billion a year. That figure reflects both the value to travelers of time they would no longer spend in congestion and savings from reduced highway maintenance. The political objection to charging for use of roads that now are free is that it would be an additional burden for poor and middle-class commuters. Counters Winston: ''The returns would be so great that we should be able to compensate those least able to pay and still have all of us come out ahead.'' Toll revenues might be used to offset equally regressive gas taxes or to subsidize mass-transit systems that are now underused. Could the kinds of technology that make Winston's proposals practicable be used to embark upon a wholesale privatization of America's infrastructure? Plans for private toll roads are moving ahead in Virginia and California. William Weld, the Republican governor of Massachusetts, says his cash-strapped state could raise $1.4 billion by selling off the Massachusetts Turnpike. President Bush earlier this year sought to make privatization easier for state and local governments by issuing an executive order that allows them to sell off assets built with federal funds without fully reimbursing Washington. But Steven Steckler, a senior manager with Price Waterhouse, points out that not many roads would yield enough toll or concession revenues to make them attractive to the private sector. ^ Where the private sector ought to take the lead role is in building what could prove the 21st-century equivalent of the interstate highway system -- a national information network based on fiber-optic lines. Apple CEO John Sculley insists that such a network, which could cost as much as $400 billion to install, should be in place by the turn of the century. What the system's potential creators -- the Bell operating companies and others who might link up with them -- mainly need from government is a strong, competent hand that would set standards and clear regulatory obstacles. In fixing America's infrastructure, as with so many other complex policy problems, it's a lot easier and more satisfying to issue calls to arms and promise expensive new programs than it is to fiddle with pricing systems, fill in potholes, and remove regulatory hurdles. But those are the kinds of challenges that the next President must take up. For a nation faced with the need to make painful choices to restrain its runaway budget deficits, there's simply no other option.

BOX: INSIGHTS

-- Higher spending on infrastructure won't automatically lift productivity growth. -- Smaller projects, such as filling potholes, offer the biggest bang for taxpayers' bucks. -- Reducing demand by imposing tolls and fees on congested roads and airports makes more sense than building new ones. -- Privatization is no panacea. But in a few cases -- a new fiber-optic network is one -- business, not government, can do the job.

CHART: NOT AVAILABLE CREDIT: FORTUNE CHART/SOURCE: CONGRESSIONAL BUDGET OFFICE CAPTION: INFRASTRUCTURE