How to Make Deregulation Work
Alfred E. Kahn, the father of airline deregulation, firmly defends it in an interview with IEEE Spectrum--but is less sanguine about the effect on electricity and communications
ARTWORK: PETE MCARTHUR
This is part of IEEE Spectrum's special report: Critical Challenges 2002: Technology Takes On
All happy families resemble one another, but each unhappy family is unhappy in its own way." So runs the famous first sentence of Anna Karenina. With but a little poetic license, one might say that every regulated industry manifests the same rules, while each deregulated industry is unruly or even dysfunctional in ways unique to it. Figuring out how to fix each one's problems is in its own right a huge technological challenge--technological in the economist's sense of referring to all aspects of business organization.
In a regulated industry, services are provided by a single or a fixed few companies, monitored by some quasi-governmental authority set up to protect the public interest. In exchange for its accountability, the industry is assured of a modest yet still satisfactory rate of return, often made more attractive to investors with tax breaks. Though the authority is supposed to have an arms-length relationship to the industry, in point of fact, after working hand-in-glove over decades, regulators and regulated hardly ever see things differently. Innovation may atrophy, but everybody is happy, even most consumers, who scarcely can imagine things being any different, having never experienced an alternative.
Remove that regulatory body and plunge the industry into the bracing cold waters of competition, and good things are supposed to happen: new entrants appear, prices come down, and consumers are delighted with products and services previously unimagined. Thus, deregulation was the dominant trend in economic policy throughout most of the industrial world during the 1980s and 1990s, and initial results were often impressive. But recently, signs of trouble have been appearing all around, with some deregulated industries beginning to resemble Leo Tolstoy's unhappy families.
In electricity, even before California's debacle, reliability problems and price spikes were on the rise throughout the United States, coinciding uncomfortably with the period in which the U.S. Federal Energy Regulatory Commission (FERC) ordered the national transmission networks opened to competition [see figure, above]. Critics complained that infrastructure was not being maintained because newly competing companies seeking to cut costs were uncertain how investments would be repaid in the long run.
Even in the UK, the pioneer of electricity restructuring, there were repeated assertions that the two big generators were rigging prices. Hardly anywhere in the world did homeowners find they were getting power at much more attractive prices than before. As for product and service differentiation, well, there were "green" electrons to be had from companies generating power from renewable sources.
The U.S. airline industry, even before 11 September, was suffering severe declines in profits [see figure, above], which were bound to lead to a new round of consolidation, leaving it perhaps even more concentrated than when it was deregulated in the late 1970s.
After deregulation, fares did plummet for a while, with start-up airlines vying for passengers. But now, with congestion fouling all the busier airports, passengers are bemoaning intolerable delays, widening disparities in ticket prices, and a general sense of things going too far out of control. True, air travel is within reach of many more people than ever, but not just the rich are wondering about the price that has been paid in terms of convenience and comfort.
In communications, the U.S. public profited enormously in the 1980s and 1990s from competition in long-distance telephony, as rates plunged and new competitors offered all manner of special services. But when the 1996 Telecommunications Act tried to bring competition to local and regional markets, benefits were unexpectedly slow in appearing [see figure, above]. A host of new companies were founded to provide digital subscriber line (DSL) connections, but after a short while they went under in droves, as the Regional Bell Operating Companies (RBOCs) reasserted their strangleholds on local markets.
Meanwhile, the seven so-called Baby Bells have merged into just four huge companies, while, worse yet, AT&T--the queen of all communications companies just 20 years ago--teeters on the brink of bankruptcy. What's going on? To gain some perspective, IEEE Spectrum sought out Alfred E. Kahn, the father of U.S. airline deregulation, at his offices in Ithaca, N.Y. Now an emeritus professor of economics at Cornell University, Kahn first came to national and even international prominence in 1977, when President Jimmy Carter picked him to head the Civil Aeronautics Board (CAB), the agency that regulated the U.S. air industry. Kahn's mission there was to abolish his own job, which he did to general acclaim.
Previously, Kahn had served as chairman of the New York State Public Service Commission, having pretty much written the book with his two-volume work, The Economics of Regulation (1970, 1971). Now 84 years old, Kahn continues to publish frequently on all aspects of regulation and to give testimony when his views are solicited, which they frequently are. In the interview that follows, Kahn expresses skepticism about introducing consumer choice in electricity, mounts a robust defense of airline deregulation, and expresses some optimism about the prospects of making local telephone systems more competitive.
SPECTRUM: What's happened in electricity deregulation? I don't see a decrease in my rates, and at least in New York , if I leave the incumbent utility for someone else, they're under absolutely no obligation to take me back.
Kahn: I've always been a skeptic about retail consumer choice in electricity. I'm not crazy about the choices I have to make among telephone service providers either; but we all have clearly benefited enormously from the deregulation of long-distance service, whose prices were in any case grossly inflated by regulation in order to subsidize basic local service. We clearly have to include the cost to consumers of making intelligent decisions when we reckon up the costs and potential benefits of introducing competition. I'm skeptical about how that balance will turn out in electric power at the retail level; the power comes out of a totally interconnected network, and it seems to me the possibilities of competition at the local distribution level are limited, it's such a narrow-margin business anyhow. In any event, the complications of the whole thing seem to me to make it essential that you first get wholesale competition right, and that means ensuring the inescapably essential integration between transmission and generation.
So California moved too fast?
California was a wonderful example of misguided populism exhibited by both political parties. The people in California--or their political leaders--were determined to reap the benefits of deregulation for small consumers, first. But unregulated electric markets would be subject to wide fluctuations, as the balance between demand and supply shifts hour by hour, day by day, and season by season. The only way competition could work, then, would be if prices to consumers reflected that constantly changing balance, so purchasers would vary their consumption accordingly--for example, by putting off their use of electric dryers, dishwashers, and hot water from day to night at times of shortage. But that requires real-time metering, and the overwhelming majority of customers in California are still not real-time metered.
In addition, the designers of California's deregulatory framework felt they had to micromanage the whole thing, and so they required that all transactions had to be through a spot market--a spot market that would be subject to huge fluctuations, although, of course, we didn't adequately perceive how serious the imbalance between capacity and peak demands was or would prove to be.
The other mistake that I'm just beginning to understand is that if you're going to put caps on energy prices to protect people from spikes (which may or may not involve monopolistic manipulation), then you also must retain requirements that the load-serving entities maintain some stipulated reserve capacity, so that those spikes don't get completely out of hand. In New York, under regulation, we required generators to maintain 18 percent reserve capacity. California didn't retain such a requirement, and when peak demands soared, there wasn't a guaranteed margin of surplus capacity to keep prices from exploding. And since consumers were never confronted with those price spikes--they were rolled into their monthly bills--they had no means of protecting themselves by curtailing their purchases at those particular times and places.
You caused quite a stir last year by signing on to a letter endorsing wholesale price caps in California. You argued that given extreme elasticities of supply and demand, aggravated by absence of metering, markets were not working very well. Couldn't those arguments in favor of price caps also be made for a windfall profits tax, the proceeds being fed into energy investments or just given back to consumers?
I think that's right--I feel stupid it never occurred to me. Had it occurred to me, however, I would almost certainly have rejected it, because it would have required legislation; and the last thing we wanted was to encourage legislators to get into the process. Such price controls as they would almost certainly have wanted to impose could well have been counterproductive, discouraging the investment in the expanded generation capacity that was the fundamental remedy. We concentrated on FERC because it could impose the caps on its own authority.
In fact, I complained that the first draft of that letter advocating caps had the words "just" and "reasonable" at least 10 times, referring to the existing statutory authority. But just offhand, I would think that a carefully designed windfall profits tax would be preferable because it wouldn't interfere with the pricing system and the proceeds could be handed back to consumers in a lump sum, so as not to discourage efficient conservation.
At a Department of Energy hearing in San Francisco two years ago, a representative of Dynegy Inc., of Houston, suggested creating something almost like the Federal Reserve System, with sort of a federated structure of boards setting reserve requirements and enforcing them. Is that a good idea?
It would have sounded like it to me until about two weeks ago, when I read the latest manuscript by Sally Hunt [of National Economic Research Associates], in which she argues you don't need reserve requirements. It can all be done by energy markets, provided you have sufficient metering.
There has always been a popular tendency to underestimate the elasticity of demand. Even in California, without metering, there's been an extraordinary amount of recent demand response. And of course a lot of demand is industrial demand, where it doesn't take very much of a price increase for companies to be better off shutting down. Remember, it's not the majority of customers--but only the major portion of the total load--that you have to have real-time metered, because most of the load is accounted for by large business customers.
If real-time metering is a good idea in electricity, what about local telephony, where it has proved so difficult to introduce real competition?
I certainly remember pressing for local measured service [LMS] as a natural application of the same principle, and that was in the 1970s, when I was chairing the New York State Public Service Commission. I think a lot of the wind went out of the sails when Bridger Mitchell did a study for RAND concluding that the costs of administering such a system would far outweigh the benefits.
I have measured service on my phone...
Oh, yes, in New York City, we required it. I was testifying in a state telephone case way back in the 1980s, and was arguing in favor of local measured service, and the presiding judge said, "Dr. Kahn, do you have LMS?" And I said, "Your Honor, I'm served by an independent telephone company, and I barely have S."
When you were the New York commission chairman, that included telephony, electricity, gas, and water, but not transportation?
Couldn't that also account for so much of your success when you went to the Civil Aeronautics Board (CAB): you went in with a clean slate, with no preconceptions?
You're absolutely right. I discovered I was very good at making use of the public meetings law ["sunshine" legislation enacted in the 1970s requiring meetings where public business is transacted to be opened to the public] as a means of educating both myself and the public. First, it scared me, but I found that I could use it to force the staff to argue with me and educate me, and in this way demonstrate publicly that we were being very careful and responsible.
Given the problems in air travel of late, can one still refer to what you did at the CAB as a poster child for deregulation?
Regulation has an inherent tendency to be protective of the people being regulated. The airline industry was a case in which regulation had become increasingly protectionist. In the entire history of the CAB, from 1938 to 1978--with one minor exception--not one single new airline had been certificated. It was a real cartel. So prices were originally set at such a level--fares for transcontinental travel were held particularly high to subsidize shorter flights--that the airlines could break even with only 35 percent of their seats booked.
So in the decade before deregulation, the average load factor was 52 percent. And you know in recent years it's been well over 70. And how did that come about? Price competition, taking the form primarily of discounts to fill seats that would otherwise be empty--that's good economics. As a result of those discounts and the higher efficiency with which airlines are operating, travelers are saving something like US $19 billion a year, according to Brookings Institution estimates.
Hasn't the hub-and-spoke system that emerged as a result of deregulation caused enormous congestion?
In terms of efficiency, fuller use of craft, and offering travelers a much richer choice of destinations, the development of hub-and-spoke has been a very good thing: for most purposes, it has proved to be the most efficient way of operating. Yet we didn't predict it under deregulation--it was the deregulated markets that demonstrated its superiority.
Once Don Carty of American Airlines showed me that if you have five airplanes on the East Coast and five on the West Coast at any given time and you don't have a hub, you can serve 10 transcontinental routes at a time. Put a hub in the middle and the number of potential offerings goes up to 70. When you take nine planes from the East and West, instead of the 18 flights possible without a hub, you put a hub in the middle and it goes to 198.
But what about the congestion and overscheduling of flights?
The airlines have no choice but to schedule flights when and where people want them. The problem is failure of the FAA [Federal Aviation Administration] to supply the needed infrastructure and to price it correctly. As long as a government agency runs our airports and air traffic control [ATC] systems, it must have the ability to raise the capital needed and to charge [fees] in such a way as to expand ATC capacity and airports.
The basic problem is the way airports and ATC charge airlines for their services. Let me crib from Bill Vickery, the Columbia University economist who won the 1996 Nobel Prize. What do you think would happen to the market for beef, he asked, if you priced everything that came out of the cow at the same price per pound? Yet that's what airports do; they charge per pound of the plane. Which means that a tiny private plane with just a few passengers may hold up a plane with 300 people--imposing, what's more, an added cost because you have to have wider spacing. You charge those few people only a tiny fraction of what you charge the 300 people that they hold up.
It would be like charging the same price per pound for soup bones, for hamburger, and for steak--you'd run out of steak. You'd have long lines; you'd have riots...
Let the transcript convey that you said that, not I. Anyway, as to the failure of FAA, I think it's very largely the failure of the way it's organized as a government agency. What's needed is privatization or at least corporatization--an independent agency, able to raise its own capital, not limited by the congressional budget process or civil service requirements. An agency, cooperatively owned by industry, with real incentives to meet demand, and expand capacity, and price intelligently.
You know that former President Clinton introduced a proposal for corporatization, and Secretary of Transportation Norman Mineta is supporting it, too. [Mineta, a Democrat, was chairman of the House Aviation Subcommittee in 1978 and helped shepherd the deregulation bill through Congress.]
What about all the consolidation in the industry? Aren't there fewer airlines now than when you deregulated?
I think that's misleading because we have a large number, a constantly changing fringe of new entrants, low-cost, low-fare. I remember in the period from about '93 to '97 or '98, we had some 30 or so new airlines established. Some of them have been driven out of business, to be sure, and we have a very considerable concern about predatory tactics. But there are some very successful, very low-fare carriers, like JetBlue. Every time I turn around, I hear about another one operating successfully--yes, and another closing up shop.
So the industry is more consolidated, in the sense that the top three or four carriers have a larger share of total traffic nationwide than previously. I don't think there's any question of that. I wouldn't be surprised if--and here I'm really guessing--that the top four may have had 60 percent of the total in the late 1970s, and now it's 85 percent. But the average number of carriers serving individual routes, which is where competition does or does not occur, has gone up, not down, as carriers have exercised their new freedom to extend the scope of their operations.
Is that concentration an element in the widening disparities between fares, especially the high ticket prices paid by the business traveler?
The spread between discount fares and full fares has widened enormously under deregulation, and that is a source of dismay--to some extent, in my opinion, legitimately. Average fares, adjusted for inflation, have declined about 40 percent. Full fares, paid on fewer than 5 percent of all miles traveled, have risen about 70 percent. So this enormous differentiation of fares is what most people point to as an instance of monopoly exploitation.
I do worry about the adequacy of competition to protect that business traveler, but a very large portion of the differentiation is not discriminatory at all. It costs more to have frequent scheduling, which business travelers prize. It costs less to serve people who won't fly at all normally but will fly if you pack them in planes in narrower and narrower seating, and they're willing to wait to go once a day, and so on. It costs more to fly into congested than noncongested hubs, but the hubs make possible the convenient scheduling to multiple destinations that business travelers particularly need.
Anyway, although those unrestricted business fares are also discriminatory, they are certainly not producing monopoly profits in the aggregate. On the contrary, the greater danger would seem to be that the combination of heavy capital investment, sharply fluctuating demand, the almost zero marginal costs of taking on additional passengers on flights that are going out anyway--that all this will have a systematic tendency to generate really destructive competition; and in fact the airlines lost a fantastic amount of money in the early 1990s because they had engaged in a binge of new capacity acquisitions in the middle of the late 1980s and the growth of demand sharply decelerated around 1990. But they learned from that experience, and in the 1990s, as demand recovered, they did not overinvest so severely, and found themselves with extremely profitable load factors, in excess of 70 percent.
What did you think of the 11 September bailout package?
As I said at the time, I thought the circumstances justified some assistance--not to offset the effect of the recession, which need have been nothing more than a repetition of the 1990-2000 experience. But clearly the incremental effects of September 11 were not an ordinary commercial event. They promised to impose on the airlines the enormous additional burden of doing police and military work. And finally, there was the consideration that the airline industry is a kind of backbone industry, on which the productivity of so much of the rest of our economy depends.
So for all those reasons I could not oppose the bailout. But what I have opposed, most vociferously, is the criteria developed by the Office of Management and Budget, which placed the principal emphasis on the development of a strong financial plan that would maximize the likelihood of airlines' being able to repay. In effect, instead of stepping in to assist the victims most threatened with eradication, if you adopt this as your criterion, it amounts to descending to the battlefield and shooting the wounded.
They should have adopted criteria--I have no idea about their administrability--intended, as closely as possible, to restore the conditions as they would have been if September 11 had not occurred, that is, to keep the structure of the industry as it would have been but for September 11. You figure out how to do that.
Discussing concentration in the airline industry, you've sometimes mentioned passengers liking to stick with one carrier for a trip, not having to worry about losing their baggage. One-stop shopping and seamless networking would seem a big selling point and a significant economy of scale. Yet AT&T CEO Michael Armstrong bet the farm on that strategy in telecommunications and seems to be losing.
He was wrong, and so were most of us. Obviously we overestimated the technological benefits and marketing economies of offering a full complement of telecommunications services. I was involved in about six cases on behalf of what was then Bell Atlantic in the 1990s, contesting in various federal district courts a [now defunct] congressional prohibition on telephone companies' offering cable TV services. We won all six cases on behalf of Bell Atlantic, but I don't think they ever made it work commercially, even with the added incentives provided by the 1996 Telecommunications Act.
What do you think of the act?
It seems to me a reasonable compromise between the interests of the RBOCs, in getting into the long-distance business, and AT&T and others, to offer a full complement of services, local and long distance. And conceptually, the notion is very attractive--that if you could dissolve the monopoly of local networks, you would have come very far in the direction of genuine deregulation, and this in an environment of rapidly developing technology, huge investments, and enormous technological uncertainty. So I don't have a real quarrel with the act.
What about that premise, breaking up the local monopoly? That is beginning to look tougher than the act presupposed.
That's probably right. Even though I've worked for RBOCs quite a lot, I am simply incapable of sifting through and appraising the persistent assertions of their obstruction of local competition.
I do have to emphasize, however, that you're never going to get ubiquitous competition as long as you don't allow prices to reflect costs. All of the competition from 1959 on was in long distance because that's where the rates were designedly held far above costs, including access charges. Those were regulated to generate billions and billions of dollars to flow to the local companies to hold down local residential charges. So every major city in the United States has had a competitive alternative local-access provider, using fiber-optic rings to originate and complete long-distance calls. Teleport originally was just an access provider. So was MFS. And those independents began very rapidly to compete intensely and successfully with local telephone companies, using their high-capacity facilities to offer service to business customers, whose rates, just as long-distance rates, had been held outrageously above cost, in order to subsidize residential service. And, of course, you had to require the local telephone companies to interconnect with them on demand. And they grew extremely rapidly.
I was in Minnesota talking to the legislature, and I remember Ameritech, at that time Northwest Bell, pointing out that their regulators were coming out with estimates of costs of the unbundled access network of $15. Their local rates to residential consumers were $16 to $18. But the rates for the identical service to businesses were $50 a month. Is it any wonder, then, that you had fiber-optic rings being built in metropolitan areas all over the country, taking over a rapidly growing share of the business market? Or that no one took an interest in building competing local networks out to farms or other sparsely settled areas, to serve people, some of them very rich like [the actor and film director] Robert Redford, at local rates that almost certainly are a tiny fraction of costs?
But what about the small and disadvantaged residential telephone consumers those local rate subsidies are supposed to protect?
Those protections were put in place originally to make universal telephone ownership possible. But today, with phone service within the reach of almost everybody anyway, a multibillion-dollar subsidy is a dreadfully inefficient way of ensuring universality of service. We don't regulate the price of food in order to see that poor people get enough to eat.
Point to Ponder
TELECOMMUNICATIONS STOCKS AS LOSS LEADER Between spring 2000 and summer 2001, the plunge in their prices accounted for 90 percent of the overall drop in the stock market's value, according to an in-depth analysis that appeared in The Wall Street Journal.
BEFORE THE FALL An Enron share was US $0.26 when the firm declared bankruptcy in December, down from its $90.56 peak in August 2000. One large investor in the Houston energy trader that capitalized on deregulation saw holdings go from $2 billion to $6 million
Still, how realistic was the premise of the act that the monopoly stranglehold of the RBOCs could, in fact, be broken?
It has never had a fair test, because the regulated retail rates are so far out of line with costs. The act set up several means of entry: one was by leasing unbundled network elements, another [by] purchasing network elements from the incumbent telephone companies or, alternatively, purchasing their retail services at a prescribed local discount many times the costs they saved the local companies, for resale in competition with them. In other words, competitors could offer some services with their own facilities, some with the leased facilities, and still others by direct wholesale purchases from the incumbents. AT&T came in at one point and argued they should not be required to acquire network elements one by one, but should also be able to acquire a "platform," as they called it, which is all the network elements necessary to provide a service. So in effect what they were demanding--and got--was an alternative way of reselling.
The studies that I've seen found that where the platform became available, there you had no construction of facilities--unsurprisingly. If a would-be competitor can acquire all the network elements it wants, prebundled, at prices prescribed by the FCC to reflect the minimum costs the most efficient new entry writing on a blank slate could achieve--why would anybody assume the risks of making those multibillion-dollar investments itself? I can't conceive of any method of easing entry that would be more fatal to encouraging construction of competitive facilities.
You've dubbed that FCC rule TELRIC-BS, supposedly for TELRIC-Blank Slate. Verizon has challenged that rule in a proceeding now before the U.S. Supreme Court. You're saying that if consumer telephone rates were set to reflect true costs, then you would begin to see real competition in the local markets?
dere02That's the argument. What I can say is this: where rates were clearly excessive, we have seen intense and effective competition develop. Where rates were demonstrably subsidized, there you have no competition. And third, where a would-be entrant could purchase elements or whole systems at rates actually below the total and even the incremental costs of the incumbents, you get little or no facilities-based competition.
But whether in the end, if the local companies were free to set rates on the basis of their own costs, local telephony might turn out to be a natural monopoly after all, I don't know.
It's your local company that installs all your local equipment, so if you go, say, to AT&T for local service, when that equipment breaks down, you will have to go back to the local company to get it working again. That makes you very hesitant to look elsewhere.
There are two more hopeful cases. One, broadband service. Right now DSL-equipped copper loops are way behind coaxial cable of the cable companies--they have only about half of their market share, and the two major competitors together serve only about 10 percent of all residential subscribers. And then there is the prospect of fixed wireless, and satellite and cellular--they are all beginning to offer those kinds of services, the broadband Internet access. These are services that don't depend upon the copper loop. And so I think that's the principal hope.
But, to return to where we began, AT&T's Armstrong bet on those hopes and appears to have lost.
It's only the extreme uncertainty of the technology and about what will prove attractive to customers and what will not that did Michael Armstrong in. But that's an argument by my standards for deregulation, with a requirement only that all competitors interconnect with one another.
You think that's where it could go, that the ultimate deregulatory state for the communications industry would be just an interconnection rule and not much else?
I think so. I hope so!
To Probe Further
For additional insight into some deregulation problems, see Mark N. Cooper's Reconsidering Electricity Restructuring: Do Market Problems Indicate a Short Circuit or a Total Blackout?, published by the Consumer Federation of America, Washington, D.C., 30 November 2000.
To learn more about Alfred E. Kahn's views, check out his Whom the Gods Would Destroy, or How Not to Deregulate, published by the AEI-Brookings Joint Center for Regulatory Studies, Washington, D.C., 2001.
Highlights of the work done by the Committee on Broadband Last Mile Technology are available in the National Research Council's Broadband: Bringing Home the Bits, published by the National Academy Press, Washington, D.C., 2001.
For critiques of U.S. telecommunications regulation, from radically different ideological perspectives, see "Telechasm," by George Gilder, The American Spectator, September-October 2001, pp. 12-14; "Disconnect: How Bush and Michael Powell are Killing the New Economy," by Karen Kornbluh, The Washington Monthly, October 2001, pp. 22-28; and "Remote Control: The Federal Communications Commission Has the Power to Protect Consumers ... but the Telecom Industry is in the Driver's Seat," by Brendan I. Koerner, Mother Jones, September-October 2001, pp. 41-45, 90-92.