If you were running a U.S. airline in the mid-1990s, the future looked bright. Both the economy and demand for air travel were growing, fuel prices were falling and airlines were making their best profits since 1978, the year the industry was deregulated.

Each of the seven "legacy" carriers – the large, formerly regulated airlines that had survived the 1980s (American, Continental, Delta, Northwest, TWA, United and US Airways) – had staked out one or more hub airports where it had cornered most of the traffic. They could charge high prices for travel to or from these "fortress" hubs, and faced little risk that another airline would try to muscle onto their turf. Many in the industry thought they had finally overcome the turmoil of the transition to free markets and were now on a glide path to stable profits.

Or not. With strong demand and high profits, the last half of the 1990s did meet expectations. But airlines responded to good times by adding record numbers of aircraft to their fleets – just in time to be slammed by the harsh realities of the new millennium. First came the recession of 2000-2001; then 9/11, which shut down the industry for days and, more important, curtailed demand for years. The airlines, it seems, had gotten a lucky respite in the 1990s rather than finding a profitable modus operandi. Indeed, the domestic carriers lost more money in the first decade of this century than they had made in the previous 22 years. Even counting the past few years, in which the industry has returned to the black, net returns since 1978 are negative.

In 2000, the U.S. airline industry consisted of the seven legacies plus Southwest, America West and a handful of small fry. By 2010, all seven and America West had declared bankruptcy at least once. United, American and Delta are now the last legacies standing after swallowing Continental, USAir/America West and Northwest, respectively. (American had already absorbed TWA in 2001.) Southwest also did a bit of shopping, grabbing AirTran, the discount airline that had long fought Delta in Atlanta. The four surviving mega-carriers (including the indentured regional airlines that fill out their low-density routes) now serve 71 percent of all domestic U.S. traffic. By comparison, the market share of the top four had fluctuated around 55 percent from 1980 until as recently as 2009.

Monopoly Power in Sight?

From a consumer's perspective, this concentration looks worrisome. But the airlines beg to reassure. They note the national market share of low-cost carriers (LCCs) – the upstarts that have substantially lower operating costs and usually charge lower prices (and force all on their routes to do so) – has risen steadily from 4 percent in 1980 to 19 percent in 2000 to 32 percent in 2012. Southwest, the prototype for successful LCCs, has for decades been expanding its market share based on its low-fare model, while remaining the only carrier that has reported profits every year. In 2012, 65 percent of all domestic traffic flew on a route where a LCC competed, up from 43 percent in 2000. And that's not only Southwest's doing. In the past decade, JetBlue has become a viable airline and Virgin America has gained a toehold; Spirit, Allegiant and Frontier are hanging in there, too.

Air fare trends have also been favorable to consumers since the turn of the century. Domestic fares, adjusted for inflation and average trip distance and including all taxes and fees, have fallen, while airline costs – operating costs per available seat-mile – have not. The airlines managed this feat by filling an ever-higher percentage of available seats. The average U.S. flight now leaves the ground about 83 percent full; this figure has climbed steadily from about 50 percent at the time of deregulation.

Most airline costs – fuel, crew, aircraft depreciation, landing fees – are fixed if the plane takes off, so filling more seats is pure gravy. On the other side of the ledger, of course, is the passengers' reality that flying is less fun when you are fighting for the armrest. But most travelers seem to value low fares more than comfort.

Strikingly, the biggest winners in the last few decades have been premium-fare travelers. The lowest fares have fallen a bit across the period, but the highest fares have dropped more. The 80th percentile fare on a route is 85 percent higher than the 20th percentile fare today, down from about 141 percent higher in 2000.

High-end (generally business) travelers had been paying particularly high prices at hub airports dominated by single airlines. Concentration at those airports has dipped somewhat, but the "hub premium" has fallen much more significantly. Both fare inequality and the hub premium peaked in 1996, and declined by 42 percent since then. In 1996, there were 10 airports among the top 50 where passengers paid average prices at least 20 percent above national average. By 2012, there was only one.

The Sad History of Airline Finances

The airlines have shaken off their troubles and returned to profitability only in the past few years. Why can't they seem to make money for long?

Actually, they have at times, such as the mid- and late-1990s, when they were profitable even on domestic routes. (International routes have almost always been more profitable because they are less competitive.) That was a time of strong demand growth and low jet fuel prices, bottoming out at around 60 cents a gallon in 1998.

But airline profits are necessarily volatile because the two major drivers – demand and fuel prices – are subject to large shocks. Demand fluctuates more in the airline business than in nearly any other major industry. Meanwhile, jet fuel costs, which are closely linked to the price of diesel fuel, have represented as little as 10 percent of airline costs and as much as 40 percent.

Volatile demand and costs don't necessarily mean big losses, but that has been the result through much of the era of deregulation. In part, this is because airlines haven't been able to resist expansion when times were good. There have been two aircraft buying sprees since deregulation – in the last half of the 1980s and the last half of the 1990s. Each ended abruptly when demand soured; the surplus aircraft spent years parked in the desert. The industry may be in the midst of another boom now, though it seems to be a more restrained one.

Why is what airline analysts call "capacity discipline" so elusive? After all, there are many industries with large fixed costs in which firms recognize the need to invest prudently so they won't get stuck with idle capacity when demand weakens. The difference in the airline industry is the perceived – and actual – competitive advantage an airline gains from expanding its route network.

The advantage doesn't come from driving down costs – there's no convincing evidence that larger route networks lead to lower costs. In fact, the carriers with the smallest networks have the lowest costs per seat-mile. Instead, the advantage is in product offerings and "loyalty" programs.

A larger network allows an airline to offer passengers a way to fly to more destinations without having to change airlines. Back in the 1980s, this didn't matter as much. Airlines played nice together in ticketing: it was possible to get a reasonably priced ticket with different flight segments on different carriers, an ability to mix and match that economists call product compatibility. By the end of the 1980s, however, some of the largest airlines had decided that compatibility wasn't in their interest. Accordingly, it became almost impossible to snag a bargain fare unless you bought the whole trip on one carrier.

No more going out on United and back on American. From 1984 to 1997, the share of round-trip nonstop tickets with different airlines providing the service in the each direction went from 13 percent to a mere 1.5 percent. And no more changing planes from Delta to Northwest in order to fly from Atlanta to Seattle. The share of one-direction trips requiring a plane change that involved more than one major airline fell from 6 percent to 1 percent. This artificial incompatibility put smaller airlines at a disadvantage and increased the value of network expansion. It also created a potent incentive to invest in new airplanes and routes now, and worry later.

Similar forces play out with frequent-flier programs. They encourage travelers to concentrate their miles on one carrier, as much for the perks that come with being a high-­status customer as for the free trips. By the same token, an airline needs a broad network to make those perks valuable to travelers.

Frequent-flier programs, incidentally, are also encouraged by tax policy. When your employer pays for the trip but you keep the miles, you are getting untaxed compensation. The IRS recognizes this – nowhere is it written that frequent flier miles are freebies. But for logistical and political reasons, the taxman has so far decided to look the other way.

Less well known, but equally important to the competitive landscape of the airlines, are loyalty programs for businesses, known as corporate discount programs, which can be tailored to leverage relative network size even more effectively than frequent-flier programs. American Airlines knows this when its sales people tell a Dallas-based corporation that they need to get 90 percent of the firm's travel to and from Dallas in order to provide the best discount. Note, moreover, a subtle difference that adds an extra competitive bite to corporate discount programs: most FFPs are based on the miles of travel on a carrier, while the corporate discounts are designed to target the share of the company's travel the airline gets.

Network leveraging through ticket incompatibility and loyalty programs led to alliances among otherwise competing, or potentially competing, airlines. These partnerships originally consisted of carriers from different countries, which made up for the fact that neither could fly domestically in the other's country. That coordination made a certain amount of sense to passengers as well as airlines. So did the carriers' partnerships with commuter carriers. The commuter carriers make the short hops from hubs to hundreds of regional airports that major carriers cannot serve cost-effectively.

But by the late 1990s, when the first domestic alliances among legacy carriers were created, the genuine benefits gained in terms of access to an otherwise inaccessible market (as opposed to the ersatz benefits from ticket compatibility and loyalty programs) were hard to find. Alliances are often a substitute – and a poor one at that – for a carrier having a larger network.

So, when the economic expansion of the mid-1990s came, airlines had big incentives to increase capacity and broaden their networks. That helped keep air fares falling despite strong demand. And it also created the preconditions for the industry's free fall a few years later. After reporting positive earnings for eight straight years, the industry saw its collective profits plummet in 2001 and stay negative for five years. Then, just as the airlines were starting to look profitable in 2006, fuel prices soared and, shortly thereafter, the economy tanked. U.S. carriers didn't make money again on domestic operations until 2010.

In late 2009, after the airlines had spent years in the red, Secretary of Transportation Ray LaHood formed the Future of Aviation Advisory Committee, a group of 20 that included airline CEOs, labor leaders, airport operators, aircraft and parts manufacturers, industry analysts, a consumer advocate and two academics (one of them, me). At the committee meetings, the carriers openly fretted about the sustainability of the U.S. airline industry. Some argued that excessive taxation and insufficient support from the public coffers, particularly for airport services and a new airline navigation system, were fundamentally undermining their viability.

In truth, it's unclear whether the high taxes on airlines and air travelers under-compensate or over-compensate for the public services provided – airports, security, air traffic control and immigration services, among others. But even if taxes exceed the benefits returned, they aren't going to destroy the industry. Rather, they are going to affect its size. Higher taxes will lead to a somewhat smaller industry and lower taxes will do the opposite.

Fuel costs were also called out as a threat, which they no doubt have been to the shareholders of U.S. carriers and some workers. But as we have seen since, after shedding some capacity and reorganizing, the industry can adapt to higher fuel prices. The problem and solutions, after all, are hardly unique to airlines. The oil production industry itself wouldn't be "nonviable" at $40 a barrel, but it would be smaller than it is today. Gasoline at $8 a gallon (where it stands in much of Europe) wouldn't kill the U.S. auto industry, but auto use (and sales) would shrink.

Will Competition Contain Fares?

The bust of the 2000s was extraordinary because the biggest demand drop in the industry's deregulated history was followed by the biggest fuel price increase and then the second-largest demand decline. Before 2001, the deregulated U.S. airline fleet had never shrunk year over year, but it did in many of the early years of the 2000s, and shrank overall between 2000 and 2009. This downsizing could be viewed as both a response to massive losses and an expression of determination on the part of the carriers to stop the vicious cycle. As a consequence, we've experienced a fundamental reorganization of the industry into fewer, larger airlines.

Now that the market has merged down to the four mega-carriers, airline executives and analysts are asking whether the industry will be able to hold the line on capacity. For their part, antitrust economists and lawyers are asking whether the industry is going to become too profitable – that is, whether the surviving behemoths will have the market power to raise fares above their long-term costs.

Two things stand in the way of the market-power scenario: competition among the existing carriers and the entry (or threat of entry) of new airlines hungry for market share. No one knows whether they'll be enough.

There is a widespread view that competition among the legacy airlines isn't very effective. When United proposed merging with Continental in 2010, the consultants whom the two carriers hired to persuade the regulators to allow the combination argued that the loss of a legacy carrier wasn't a big deal. That's because two legacy carriers on a route generally charge prices that are nearly as high as either one would if it had a monopoly.

History lends some support to the idea that the majors aren't really into competing with one another. In the early 1990s, the Justice Department charged that the airlines had been coordinating their price changes by announcing them in advance and then signaling agreement through a complex system of notes circulated through the industry's computer reservation systems. These notes were readable by other airlines, but not by customers. The Justice Department settled that antitrust suit against the legacy carriers (the LCCs Southwest and America West were not accused) in 1994 with an injunction ending pre-announcements of proposed price changes and the use of the reservation systems for coded communication.

The airlines responded with real-time announcements of price increases as required by the settlement, but made the announcements on weekends when fewer tickets (particularly business tickets) are booked. If other airlines in the relevant markets didn't match the increase by Sunday night, the price increases were rolled back. It's a more awkward mechanism system for collusion, but quite possibly as effective.

Note, too, that when the Justice Department settled its lawsuit with the legacy airlines in 1994, there were seven big carriers. Now there are three – and that surely reduces the communication needed to act in unison.

Besides, there seems to be less need for coordination because the incentive for price cutting among the three has fallen. The point of a price cut is to steal competitors' customers, gaining more revenue from the extra fliers than you lose from lowering prices on the ones you would have gotten anyway. Since the 1990s, however, loyalty programs have grown larger and more sophisticated, reducing customers' inclination to switch airlines in response to price changes.

My own research shows that on a route with multiple carriers, customers show a strong bias toward flying on the airline that dominates their home airport, even after controlling for price, schedule convenience and airline amenities. The price premiums at hubs have fallen, but this home-carrier bias has remained and even strengthened slightly over the years, particularly on business routes. Price wars may thus be a thing of the past.

That sounds like bad news for consumers, but there is some good news here as well. Even the studies that don't find much impact of competition among the legacy carriers on fares still report a substantial decline in prices when Southwest, JetBlue, Virgin America or one of the other LCCs is present. And with LCCs now competing for nearly two-thirds of all domestic traffic, the competition is very real. Indeed, the expansion of LCCs is probably the single most important factor that has kept fares falling even when demand was strong and oil prices were on the rise.

Can the existing LCCs continue to expand and remain profitable? Will they keep their costs down and their prices low? And will new entrants still be able to get a foothold against the large network carriers? Industry leaders and consumer advocates often assert that they know the answers to these questions, but no one really does. Here's what we do know:

• Without the LCCs, consumers would be in big trouble.
• The primary advantage of the legacies is their networks. That advantage is here to stay.
• An LCC entering a new route is trading off a big cost advantage against a big marketing disadvantage. In the last decade, however, the cost advantage has shrunk as legacies emerged from bankruptcy reorganizations with lower wages and streamlined work rules. Moreover, the marketing disadvantage has widened as legacy mergers have expanded networks – and their value to customers. So the rate of growth shown by LCCs over the last decade is unlikely to continue, at least at the pace we have seen.
• New airlines will surely continue to be formed, if only because there always seems to be a maverick entrepreneur who believes he or she can come up with a new business model that will be profitable in spite of the incumbents' advantage. But Virgin America, which first flew in 2007, is the last new large-jet carrier to get a foothold. Strictly by count of startups, new entry is on the decline. The real test will be the next few years, as demand expands with the economy.

Good News, Bad News

For most of the deregulation period, legacy airlines have been trying to use network advantages, real and artificial, to offset their cost disadvantage vis-à-vis the LCCs. Until the last few years, the legacy strategy had been failing, resulting in massive losses and ubiquitous bankruptcies. But the cost-cutting associated with Chapter 11 reorganizations and the mergers that extended their networks are turning that around.

The legacies are better positioned than ever before to handle the demand volatility and oil price shocks that will surely continue. And they are also better positioned to compete with LCCs. Airlines will almost certainly perform better in the next decade than they did in the last, and customers will almost certainly see higher prices.

In my view, that's OK. For consumers, the 2000s were a honeymoon that couldn't last. We are now seeing the pendulum swing back to more normal prices and more normal profits for the airlines. The $64 billion question is whether the increased concentration and the growing strength of loyalty programs will push the pendulum past a competitive balance between consumer and producer benefits. Stay tuned.