(FORTUNE Magazine) – Remember the good old days? Airplanes were roomy and pleasant, sometimes even fitted with amenities like Polynesian bars and pianos. Flight attendants were always cheerful. Pilots never went on strike. Reservations clerks all went to charm school. The Clampetts either stayed home or took the bus, leaving the skies to the gilded few.

In fact, these are the good old days. Since 1978, when the Civil Aeronautics Board gave up regulating who could fly where and when and for how much--not to mention the price of every drink and headset--air transportation has improved. Fares have fallen, after inflation, by almost 40%. Tens of millions more people are flying, and more cities have more service. Employment in the industry is up; death in the air is down. It's true that flying has lost a sense of graciousness, as aviation became a mass endeavor rather than an elite one. But against the benefits, that is a small price to pay.

That said, air travel does have big problems, as every battle-weary frequent flier knows. Planes are more crowded than ever. Delays hit an all-time high in 2000: One in four flights was late, diverted, or canceled altogether. Bumping went up. The number of regularly scheduled flights that regularly failed to fly quintupled. Congealed clods of food continued to pass as in-flight cuisine. No wonder consumer satisfaction is at an all-time low. And while flying remains remarkably safe, it's not good news that the number of dangerous runway incidents rose by a third in 2000. The system ain't quite broke, but it's not working smoothly either.

"Aviation is in three parts," notes Secretary of Transportation Norman Mineta. "Air-traffic control, airports, and airlines. If these three elements are the stars, the sun, and the earth, they're not in alignment." Actually, they've been out of whack for decades. Though routes and fares were deregulated in 1978, airports and air traffic were not. In important ways, the system is still shielded from the "destructive competition" that used to horrify the Civil Aeronautics Board. The problem, then, is not that deregulation began, but that it was never finished. So the system is clogged with practices that are the unmistakable products of the 1940s and 1950s. Here's how things have gone wrong--and a few modest proposals to drag aviation into the 21st century.

The airline-airport relationship has long been unhealthily co-dependent. Starting shortly after World War II and continuing to the present, communities--reluctant to raise taxes or sell bonds to pay for airports--often arranged for the airlines themselves to underwrite development. In return, airlines got a high degree of control over operations, including the right to veto airport expansion that could bring in new competitors. Airlines also influenced decisions to shut down or restrict the use of nearby facilities. When Denver's flashy new airport opened in 1995, for example, its smaller but perfectly serviceable predecessor was torn down. At Love Field in Dallas, where Southwest Airlines is based, carrier operations are sharply limited by law, with most long-haul traffic reserved for the Dallas-Fort Worth airport. Restricting competition and shutting down capacity was profoundly stupid in an industry where poor service and congestion are the most common complaints. Congestion is likely to get worse: The Federal Aviation Administration predicts one billion emplanements (people boarding planes) in 2010, a 50% increase over 2000.

For years, airports have also commonly granted airlines long-term leases at gates, often of 20 years or more. This guarantees a flow of revenue to the airport, but it also can leave gates idle, even when traffic is frantic or when new airlines are knocking at the door. In a number of hubs, including Charlotte, N.C., Cincinnati, Detroit, Minneapolis, Newark, N.J., and Pittsburgh, the dominant carrier has exclusive access to two-thirds or more of gate capacity.

These policies have two unpleasant effects: They make it hard for new airlines to compete, and they make congestion worse. When it comes to competition, there is no doubt that many operators--Altair, say, or Golden West--started flying on little more than two wings and a prayer and failed because of their own inadequacies. People Express died of hubris and overexpansion. But it is also clear that the game is rigged in favor of the existing players, grounding many promising newcomers. "The real threat to deregulation," a Department of Transportation report argued last year, "is the use by dominant airlines of their market power to end competition." Overwrought? No. Only about 5% of passenger trips today are on low-fare airlines founded after 1978. And in the past five years, fewer new airlines have launched than in any five-year period since deregulation.

If it's hard for a domestic operator to dip a wing in the U.S. sky, it's harder for a foreign one. Carriers from countries that have signed "open skies" agreements with the U.S. are allowed to land and take off, but they cannot stop in New York, say, on their way to Chicago. Nor are they allowed to own a majority interest in U.S. airlines. Even if they own 49%, U.S. law prevents them from voting more than 25% of their shares.

This matters because competition drives down prices and forces service to improve. And in practice, it is not the presence of two or more major carriers in a market that seems to lower prices--the majors have their own spheres of influence and tend to leave one another alone--but the threat from low-fare carriers. In 1999 the DOT found that short-haul routes (less than 750 miles) with competition from a low-fare airline like Southwest, Vanguard, or AirTran had a nominal average fare of $84; for routes without such competition, the fare was $175. The National Business Travel Association notes that short-haul routes dominated by one or two carriers cost more than 90 cents a mile; those in which a low-fare carrier had at least 10% of the market cost less than 40 cents a mile. Finally, at hubs where a single airline dominates and there is no low-fare competitor, such as Charlotte (U.S. Airways) or Cincinnati (Delta), average fares tend to be higher than at hubs like Atlanta that have low-fare competition (AirTran).

All this is not an argument against hubs, which have helped to expand service. Nor is it an argument to favor the plucky underdog. It is an argument to finish deregulation by ensuring fair access to the aviation infrastructure. One step in the right direction: Federal legislation known as AIR-21, which passed last year, requires airports to submit competition plans. They can lose their airport-improvement grants if the DOT finds they operate in an anticompetitive manner.

Now for the problem of congestion. Most of America's 546 commercial airports have no trouble coping with traffic; some are begging for it. But 70% of passengers travel through just 31 airports. Of these, 12 account for more than half of all delays, and eight--LaGuardia, Newark, Chicago, San Francisco, Boston, Philadelphia, Kennedy, and Atlanta--are considered the worst.

"Build runways. That's the answer" to congestion, declares Deborah McElroy, president of the Regional Airline Association. She's correct--up to a point. In 1990 the wonks at the Federal Aviation Administration projected that by 2000 there would be 678 million emplanements, an increase of almost 50%. They got it almost exactly right. During the same period, though, just six new major runways were built and one major airport. Eighteen new runways are under construction, but demand is still expected to outstrip supply at the 25 biggest airports.

Building our way out of delay is not going to happen. For one thing, resistance to new runways and airports is intense: All six candidates for mayor of Los Angeles have ruled out further expansion at LAX, even as the region runs out of capacity. For another, the regulatory process could only have been invented in hell. It typically takes longer to build a runway than it took the Apollo program to send a man to the moon. San Francisco needs approvals from 31 agencies just to reconfigure a runway, much less build one from scratch. Orlando spent five years to get its Environmental Impact Statement for a new runway, and another eight to get permits and zoning approvals. The runway is scheduled to open in 2003, 18 years after it was first proposed.

The flip side of adding capacity is better using what we have. Of the eight most congested airports, six have no room or plans for new runways--but demand is expected to grow substantially at all of them. One essential fix: Reform the "slot" system. Since 1968 the FAA has managed demand at four of the most heavily trafficked airports--LaGuardia, O'Hare, Kennedy, and Reagan National in Washington, D.C.--through slots, or the right to a certain number of flights in a given hour. Slots were distributed free to the major airlines then. The FAA has made it clear that slots are not property rights, but they are rarely revoked, and airlines can buy, sell, and lease them--even use them as collateral for loans. Just as with exclusive gate leases, airlines that have slots have no interest in allowing new competitors to get near them. In 1998 only 17 of 286 slots were in the hands of airlines founded after 1978 (last year the New York-based startup JetBlue got 75 at Kennedy after lobbying the FAA). Slots are thus another barrier to entry bequeathed to the new, more or less liberalized aviation system by the old, rigidly regulated one.

Perhaps with that in mind, Congress last year abruptly scrapped the slot system at LaGuardia. The airlines promptly filed plans for 608 more flights, a 60% increase. Result: gridlock. In response, airport authorities instituted a slot lottery to regulate demand. That calmed things down a bit. This "slottery" is due to expire in September, though, and under AIR-21, slots will be phased out everywhere except Washington, D.C., by 2007.

Given that some form of capacity management is necessary at airports where demand exceeds supply, something else will have to be done. One possibility is for airports to impose their own caps (slots by any other name) on the number of flights and periodically auction the rights to them. This would be fairer than the current system because it would be open to all, and airlines that had slots and didn't use them would pay a price.

Another idea, mentioned by both Mineta and FAA Administrator Jane Garvey, is "congestion pricing." Currently, landing fees are set by weight--a relic of the era when airports were trying to draw in as much traffic as possible. But now some airports have more de-mand for traffic than they can cope with; pricing by weight does not reflect the economic value of access to a runway. When a service is underpriced, says Economics 101, it will be overused. That is what is happening.

Under congestion pricing, airports would impose variable landing fees--higher at peak hours when the skies are crammed, lower at less popular times. The idea is to divert enough traffic from the peaks to ease the strain. Planes that choose not to pay the higher fees would skip the busy times, thus smoothing out peaks and valleys in the schedule, or fly instead to nearby, cheaper airports with capacity to spare. Right now, landing fees tend to be so trivial--averaging out at just a few dollars per passenger--that they are not an economic factor. The idea of congestion pricing is to tip the balance, so that airlines would have an incentive to save money on landing fees by using bigger aircraft on busy routes. It makes no sense that regional jets, which have a capacity of no more than 100 passengers, are plying the packed Boston-Washington corridor in the middle of the day. Bill DeCota, director of aviation at the Port Authority, which manages Kennedy, Newark, and LaGuardia airports, estimates that the average plane at LaGuardia--which has a rate of delay twice as high as any other airport--carries just 60 passengers.

Airlines love regional jets because by using them, they can add more flights to the schedule: Frequency is a competitive advantage. But RJs use up almost as much runway space as bigger planes and demand just as much air-traffic-control expertise. In terms of moving more people, more reliably, the priority should be to fly bigger airplanes. However, the average aircraft size in the American fleet has actually shrunk, from 170 passengers in 1990 to 158 in 2000, according to the trade publication Air Transport World. San Francisco airport is considering requiring airlines to use larger aircraft on heavily trafficked corridors (like San Francisco-Los Angeles). Congestion pricing would accomplish the same thing.

Critics contend that congestion pricing is an idea only an economist could love. "We don't think that gets to the heart of the problem," says Michael Wascom of the Air Transport Association, a trade group for the major airlines. "It is an admission of failure, like putting a CLOSED FOR BUSINESS sign at the airport." McElroy, whose group represents small carriers, agrees: "We shouldn't be looking to restrict demand." Let's grow up: At LaGuardia, which covers an area smaller than Central Park, there simply is no place to put another runway. Such realities cannot be wished away. And the airlines' hostility to differential pricing looks bizarre, given that no two seats on the same flight seem to have the same price.

Air-traffic control is another part of the answer to congestion--and also a problem in itself. The FAA uses a computer system whose software dates to the 1960s, and controllers still track aircraft using paper strips. "We have to acknowledge that in the 1980s we lost time" when several major modernization programs fizzled, admits Garvey. Actually, the FAA lost time in the 1990s too. Congress noticed and, as part of AIR-21, proposed spinning off the agency's air-traffic function into a separate entity. But more than a year later a chief operating officer has yet to be hired; it's difficult to find an executive willing to step into the FAA snakepit for a government salary. Many critics of the agency, including at least one former administrator, have called for more radical reform, such as transferring air-traffic control to the private sector--as Australia, Britain, Canada, Germany, New Zealand, and Switzerland have done. That is not going to happen: The Clinton Administration repeatedly sent proposals to Congress to do just that, and each died with scarcely a whimper. Mineta, too, has ruled out the idea.

Happily, some things are looking up. A transition to "free flight," which would allow pilots to choose the fastest route to their destination rather than having to fly on a set route, is running on time and on budget. And lots of small changes are already working. For example, new agreements with Canada and the U.S. military allow use of their airspace, permitting aircraft to fly around bad weather. On a recent day that arrangement meant 300 planes flew that would have been grounded a year ago. And an initiative to ease the chokepoints in the Bermuda triangle of American air traffic--Chicago-Washington-Boston--has already improved traffic in and out of New York considerably.

The Holy Grail of air-traffic control is global positioning system technology. Explains Steve Brown, associate administrator for air-traffic services at the FAA, "Ultimately, what can happen is the same volume of airspace can have a few more airplanes than today and, logically, shorter lines." GPS is at least a few years away. But even if better air-traffic control transforms the sky into a fast-moving 12-lane highway, the planes still have to land somewhere. At most, better ATC could add perhaps 15% to capacity, figures Garvey.

So that's the agenda for change: Clear out the regulatory underbrush left from the 1970s, promote competition, experiment with market mechanisms to manage demand--and keep your fingers crossed that the FAA will shape up. Sounds sensible, simple, and straightforward, doesn't it? It is anything but. Every player wants to keep the bits of the system that benefit it: Small communities (and their congressmen) want their subsidies; private pilots want access to even the busiest airports; no one wants to live under a flight path. Every reform gores someone's treasured ox.

Still, now that airports have their own financing (through the "passenger facility charge" of up to $4.50 per ticket), they are not so dependent on the airlines. "The facilities issues are not as bad as they used to be, particularly gates," says David Neeleman, president and CEO of JetBlue. And Mineta, who has a baseball bat in his office signed by the world's two most prolific home-run hitters, Sadarahu Oh and Hank Aaron, is widely respected. If he is willing to take a bat to anticompetitive practices, flying the friendly skies could become more than a joke. Someone is going to try congestion pricing somewhere, and it will work. Congress remains resistant to foreign competition, but it is hard to imagine that this state of affairs will last for another 60 years. Just because other countries are unwise to ban competition, the U.S. should not punish its consumers by following their poor example. And imagine the kick in the pants competition from, say, Singapore Air or Virgin could provide airlines and staff that treat their customers like enemies.

In short, there's enough going on in terms of runways, policies, and technology that things will certainly improve a little. With luck and political will, they could improve a lot. But none of this will happen fast. So grit your teeth and grab an inflatable pillow: Air travel is going to get worse before it gets better.

REPORTER ASSOCIATE Alynda Wheat

FEEDBACK: cmurphy@fortunemail.com