The Coup of the Century? Another Look at the MCI-WorldCom Merger
"Our goal is to build a new kind of communications company. One that behaves very differently from the monopolies that have developed over the past hundred years ... We have created the first company since the break up of AT&T to bundle together local and long distance services carried over international end-to-end fiber networks controlled by a single company ... [And] we are among a handful of players to own 100 percent of our infrastructure. We think that is a major competitive advantage that sets us apart from our competitors."-- WorldCom CEO Bernard Ebbers, 1996 annual reportHow is it that so little attention is being paid to the proposed merger between MCI and WorldCom? The Internet is a fashionable topic. The sums involved ($37 billion) are impressive. Computers and telecommunications occupy an increasingly important part in the daily life of both businesses and consumers. The future role of data communications in transforming the economy is uncontested.Yet until very recently there has been near silence on a merger that will give the resultant company the same dominant role in the Internet that Microsoft has carved out for itself in software.Just as generals fight old wars, policymakers tend to see problems in terms defined by past circumstances. No one at the Justice Department's antitrust division ever said, "Bill Gates seems like a nice fellow with an interesting product. Let's make him the richest man in the world and give him an effective monopoly over the computer software business." But Justice's antitrust folks did something that had the same net effect: they allowed themselves to grow so fixated on the computing industry's past -- embodied by IBM and its mainframes -- that they failed to see how ownership of the basic operating system for the personal computer could be used to leverage a dominant position in that industry's future.As a result, Bill Gates is now worth $39 billion, and Microsoft has effective control of a key segment of the American economy. This would never have been permitted to occur as the object of policy, and had policymakers been able to peer into the future and seen what Microsoft would become, they would almost certainly have taken steps to thwart it. This is not the same thing as saying that Microsoft ought to be broken up, as AT&T was, or that it should be closely regulated, as utilities are. It is not to say that Microsoft is a "bad" influence on its sector of the economy or that Bill Gates deserves to be a dollar poorer. It is simply to observe that, outside the universe of Microsoft shareholders, no one would ever have wished this state of affairs into being.That point alone should be enough to concentrate the minds of the Federal Communications Commission and the Justice trust busters. If they fail to subject the MCI/WorldCom merger to serious and close examination, the FCC and Justice will have given their imprimatur to what a leading Internet magazine (Boardwatch) is already calling "the coup of the century -- total control and ownership of the Internet."To put it more bluntly, the FCC and the Justice Department will have dodged the single best opportunity they will ever have to avert the consolidation of huge portions of the Internet under the control of a single corporation.The old Internet and the newThere are two common arguments for waving off concerns about the MCI-WorldCom merger.The first is that the Internet is not a thing but an idea, and no one company is capable of owning or even controlling it. No matter how large a combined MCI-WorldCom might be, it could never hope to duplicate the dominant role that Microsoft has established for itself in computer software. Not every industry is going to look like software, and not every company is going to look like Microsoft. The second argument is more subtle: Maybe technology is somehow different from other industries. Maybe it lives by different rules. Perhaps there is not only necessity but virtue in the "natural" monopolies that seem to be emerging in software, semiconductors, and backbone bandwidth.These two arguments might even be merged: "It ain't gonna happen and if it does, well, it might have happened anyway, and is probably for the good." This cheery fatalism has a certain appeal. It encourages us to dismiss concerns about MCI-WorldCom as paranoid fantasies. And it absolves policy makers of their responsibilities by suggesting that their efforts would be futile or misplaced. None of us want to be tagged a paranoid, and no policymaker wants to sign up for projects that will ultimately be judged a waste. But neither argument is ultimately persuasive. Let's look at each in turn.The first argument (no one can "own the Internet" so why fuss?) was patently true of the old Internet, the accidental by-product of government's intervention in university telecommunications. The old Internet was anarchic and subversive. Access was free or relatively cheap, content was uncontrolled, and technical standards, where they existed at all, varied wildly. Profit was initially irrelevant and often elusive. The only consistent principle was openness. And openness is the mortal enemy of monopolistic ambition.But that Internet is fast disappearing. The demand for bandwidth is doubling every 3.6 months. At that rate, in three years it will be more than a thousand times what it is today. More than 35 million people now use the Internet (up from nine million just two years ago). Some analysts predict that by the year 2001, the Internet will have more than half a billion users. Another forecast shows revenue skyrocketing from $1 billion in 1995 to $23 billion in the year 2000.The number of Internet service providers (ISPs) is about to crater -- from 1800 (or more) to less than 200. Those ISPs that survive this brutal shake-out will bear three common traits: low-cost bandwidth (and ready access to it), close relationships with telecom and Internet companies (the better to bundle new products), and access to large sums of capital (wars are always more expensive than you expect them to be). MCI-WorldCom will be as strong a force in this market as any company in the world. If it manages to acquire America Online -- a hot rumor in Internet circles, perhaps fueled by AOL founder Steve Case's recent decision to join WorldCom's board -- it would easily be the strongest and biggest ISP of all.The number of companies in the Internet backbone market is also collapsing. ("Backbone" refers to the Internet transmission lines that smaller ISPs must use to give their customers Internet access.) Already, there are only three companies with more than a small fraction of this market -- MCI, WorldCom and Sprint. If WorldCom swallows up MCI, it will have 63 percent of the ISP market (or more precisely, 63 percent of all ISPs will be connected to its network). Sprint would be a distant rival, with less than half that. AGIS and BBN together would have about 13 percent and all other backbone companies would have trivial shares. This matters because the old Internet was highly fragmented. No single ISP or backbone company was strong enough to resist a fairly open and free system of reciprocal connections, or "peering," in the local lingo. But all that's changing. Since the privatization of the Internet backbone in the early Nineties, there's been no requirement that backbone networks connect with each other. Perhaps to avoid government intervention, private network providers have employed voluntary "peering arrangements" to allow all ISPs to freely exchange traffic across regional and national networks. Then someone had the bright idea to scratch out the word "free."As a result, large backbone companies have begun to charge small ones for both transit (the right to use the companies' lines to reach a given address) and access (the ability to send a message to a customer whose address lies within their domain). "Peers" are being remanufactured into "customers" and some peers are being dropped from the realm -- through the abrupt cancellation of peering agreements.No company has been more outspoken about its desire to turn a profit on the net than WorldCom. When its main backbone company, UUNET, led the charge on peering fees, other large companies quickly followed suit. UUNET has also required ISPs to sign nondisclosure agreements regarding peering arrangements, increasing its ability to charge a noncompetitive price for backbone access.If it's big enough, a backbone provider could refuse to connect smaller backbones and ISPs. That would squeeze out the competition and further concentrate the market. It could set prices to favor its own ISPs. It could even deliberately run down the quality of service it provided to smaller ISPs in an effort to encourage their customers to make a switch.In a fully competitive market, this couldn't happen. But size matters. And MCI-WorldCom would easily be large enough to upset the conventional forces of market competition. (In addition, the Internet is free of the conventional forms of regulatory oversight -- in contrast to basic telephony, it's deemed an "enhanced service" -- so there's no effective prohibition on price-setting, discriminatory pricing, predatory pricing, or even refusing to make ISP connections.) WorldCom's frenzied acquisition of niche players in the Internet industry, combined with the large number of strategic alliances it has struck with other telecom companies, domestic and foreign, has already made its ambition to establish a Microsoft-like dominance over the Internet industry painfully obvious:It has joined with Telefonica de Espana, S.A., the dominant telephone company in Spain and Latin America's largest communications provider, to offer expanded services and new products to multinational customers. (The as-yet-unmerged MCI is also a part of this alliance.) This will increase the scope of its European operations, and position MCI-WorldCom to dominate data traffic between the United States and Latin America and between Europe and Latin America.It has signed a distribution agreement with Polycom, a leader in the teleconferencing industry, which gives it access to state-of-the-art communications technology for both video and dataconferencing. This gives WorldCom a further boost in the race to capture the lucrative high-end of the data and telecommunications industries.Through its subsidiary, UUNET Technologies, it has formed an alliance with 3Com, Ascend, and Whistle Communications to offer Internet hardware for resale to WorldCom customers, thus combining Internet access and hardware in a single package. This exemplifies WorldCom's increasing level of interaction with other Internet hardware providers, as well as its ability to exert influence over the type of products those companies develop for the Internet market.Through its subsidiary, ANS Communications, it has formed an alliance with Check Point Software Technologies, the leading provider of security technologies for ISPs and data and telecommunications companies. This gives WorldCom the ability to provide its customers with the most advanced security technologies currently available. This will make it easier to draw major data traffickers onto its own networks and improve its ability to bundle integrated services for corporate customers.The economics of cybercomOpenness, the defining characteristic of the old Internet, is slowly giving way to the demands of commerce. There's nothing wrong with this -- wiring the world is expensive. Companies have every right to try to look for ways to make the new data networks pay for themselves. If they couldn't do this, they wouldn't be able to spend the billions they are spending on new technologies that will integrate data, voice and image in entirely new ways. But it has certain consequences: Some companies are going make the wrong bets on the emerging technologies. Others will arrive on the scene too late or just be unlucky. That meansSome very unhappy shareholders will be forced to write off billions of dollars, and some companies -- big and small, new and old -- will disappear. Shareholders will put enormous pressure on managers to show consistent earnings growth, and managers will do all they can to increase gross margins. In other words, cybercom will be like every other industry, but maybe a little more so. Meanwhile, The best technology will be reserved for the best-paying customers -- invariably corporate clients and affluent individuals. Packet-switched data networks will replace public switched telephone networks as the infrastructure for this new cybercom, merging voice, data and video in a dizzying array of applications. The least profitable segments of the new Internet will be then ghetto-ized in a patchwork of obsolete technologies and aging facilities. (Sort of like FedEx and Evian versus the Postal Service and tapwater. In this, at least, the new Internet will resemble the rest of the economy.)Then there are costs. They will increase. A common defense of Microsoft is that software prices have been falling for years, at least when performance improvements are factored in -- not the result you'd expect from a monopoly. Nothing of the sort will happen on the net in the foreseeable future. With backbone providers charging for access, those $19.95 a month flat rate access plans are marked for death. Users might soon be forced to pay for how long they remain online, or how much data they sent or receive, or even how quickly (or efficiently) that data circumnavigates the net.Which brings us back to MCI and WorldCom. From their point of view, the merger makes great sense. (And to be quite fair, Wall Streeters seem to be rooting, as well.) With MCI under its belt, WorldCom would be free to "re-brand" its acquisitions of the past several years -- IDB, WillTel, MFS (including UUNET), Brooks Fiber, parts of CompuServe, and AOL's ANS -- with the MCI logo. This would dramatically boost their value in a scattered world marketplace, and increase the new company's ability to compete for international business. Think of this as the Coca-Cola or Disney approach to the Internet. It may take ten years to build up WorldCom to that level of brand recognition. In the meantime, it has the next best thing: the MCI label. Very helpful if you're racing to transform yourself from a run-of-the-mill long-distance provider out of Jackson, Mississippi to the new Dark Lord of the Internet.With the cash generated by a surge in international sales and MCI's (publicly-sheltered) long-distance business, MCI-WorldCom would be in a position to accelerate completion of its own free standing data networks -- long before competitors could hope to achieve comparable results. As CEO Bernie Ebbers observed even before the bid to buy MCI, WorldCom is one of the very few companies to own "100 percent of our infrastructure. It is a major competitive advantage that sets us apart from our competitors."To understand this, it helps to remember some of the economics of the telecom/Internet business: More than half of a company's costs are represented by lease payments to other companies for use of their facilities. If a company can bring its customers "on-net," where it has no need to pay other companies to service its clients' local or long-distance needs, it can increase cash-flow margins by an eye-boggling amount. (WorldCom itself estimates 70%.) Even before its acquisition of MCI was proposed, Morgan Stanley estimated that WorldCom was achieving "cash margins in the low 30% range, approximately 10 percentage points higher than stand-alone Internet providers." The other, more elusive matter is "synergies" -- a euphemism for costs (hard and soft) that change when two companies are combined. MCI and WorldCom have published their own projections, which show synergy levels reaching $5.6 billion per year after five years. Their investment bankers have estimated that the net present value of future cost savings will be $31 to $49 billion. (Cynical analysts have suggested this "synergy" is nothing more than the savings the combined company would realize by dumping its local residential business -- as would happen if it proved to be a drag on earnings growth or stock price, as it surely would.) But the really startling aspect of this projection is that it reflects only the cost savings the two companies would incur through combination. Nothing is said -- not a word -- about the increased revenues that would result from the merger. This is odd because the two companies "fit together" so well -- there is virtually no between their existing operations. Yet there are so many opportunities for "cross-selling" products and services to the expanded client base that a revenue bump of 15% to 20% from the merger alone would hardly be surprising. Which raises the subject of "Bundling." Cross-selling is the most overused buzzword in telecommunications. But cliches, like great ideas, have to start somewhere. In a competitive industry, there are relatively few ways to gain an advantage: lower your cost structure, increase the stability of your client base, improve the quality of your product. Cross-selling responds to this dilemma. It represents an effort to get ones existing client base to say "yes" to a greater multitude of products without really budging from their chairs. (A customer's attention span is precious.) This solves the problem of attracting new clients -- or worse, losing clients you already have. And it improves margins, since it scarcely affects a firm's cost structure.One form of cross-selling is "bundling"-- wrapping several different products or services together in a single package. Bundling can be a good deal for the consumer. If you go to McDonald's or Burger King, you can buy a bundled product -- a "value meal" or some such thing -- for less than the same items might cost if purchased separately. But it can also be bad for consumers, forcing you to buy products you don't really want in order to get the products you do want. If the seller responds, Well, because of the bundled package, you didn't have to pay full price for these additional products, you're not appeased because any price is, from your standpoint, a gouge. (And if you get the added product for free, well, that's a whole different can of worms -- one we don't have to deal with here.) So if the only way you can buy the product you want is through a bundled purchase, that's hard cheese on you, isn't it? But for the seller, this is a very good deal.In tomorrow's world of cybercom, bundling will be a growing fact-of-life, particularly for firms that find it hard to compete on the basis of price or volume. And it has an important flip side: For firms that already enjoy low cost structures and high network volumes, bundling will be a stake -- or if you prefer, a branch of fiber-optic cable -- that they can drive through their competitors' hearts. Deconstructing BernieThat brings us to WorldCom's sense of its own destiny. Sometimes the best way to figure out what a company plans to do is simply to read its annual report. (Surely Warren Buffet isn't the only one out there telling it straight.) And, sure enough, the WorldCom 1996 annual report is fairly clear: "Our goal is to build a new kind of communications company. One that behaves very differently from the monopolies that have developed over the past hundred years."CEO Bernie Ebbers puts it this way: "We have created the first company since the break-up of AT&T to bundle together local and long distance services carried over international end-to-end fiber networks controlled by a single company." Let's try deconstructing Ebbers. Over the past several years, he's done a very good job buying up niche players in the telecom industry. Now he's in a position to knit them together. Fold in MCI's lines, MCI's customer base, and MCI's Internet presence and it's an awesome sight -- even without an integrated browser. With effective control of the Internet backbone, he'll be in a position to pay out less to others even as others are forced to pay out more to him. He'll not only have the largest combined ISP in the world. He'll have the only one with free access to the world's most extensive backbone network. This gives him higher margins and a steady revenue stream from the more than 50 long-distance companies that must lease his lines. (This is good, because in the past the only way Bernie could find to increase revenues and keep WorldCom's stock price climbing was to buy another company.) Then, on that day when he is actually able to begin charging a toll for Internet access, the coins will start to fall ceaselessly. More than half the U.S. traffic on the Net -- and a goodly share of international traffic -- will pass over his lines. At that point, if Bernie Ebbers and MCI-WorldCom don't actually "own" the Internet, they will have the next best thing -- the effective monopoly of a long-term lease.Bernie Ebbers started out, implausibly enough, as a high-school gym teacher. Perhaps this is not irrelevant: Ebbers must know that MCI-WorldCom's domination of the Internet backbone industry would allow it to set the rules of competition for any company that wanted to operate on MCI-WorldCom's turf. This would create an enormous advantage for MCI-WorldCom, and when combined with its low cost-structure, high revenues, and greater margins, allow it to dominate progressively more elements of the data and telecommunications industries. To understand what this entails, imagine a professional football team that owns its own stadium. Quite apart from the revenues from ticket sales and skyboxes, it can set the terms for its concessionaires -- leading to those $7 beers and $5 Cokes. Then imagine that it finds a way to enjoy even more from licensing fees than the other teams in the league ("calling Jimmy Jones"). Stipulate a large and loyal bunch of local fans in a major metropolitan area and a long-term national television contract. (If the team owner also owns a television network, so much the better.) Assume an owner with deep pockets who won't lose his star players to free agency and a general manager ruthless enough to dump even the most beloved player when his usefulness is gone. And finally, bless the team with a roster of young, healthy players the equal of any in the game.With the usual apologies, that's sort of what Coach Ebbers sees when he looks at the company that MCI-WorldCom would become. And he's not far wrong.It should be enough to worry fans of another game -- competition.The economics of accommodationInevitably, MCI-WorldCom would enjoy profit from the special accommodation it would receive from other industry players, an accommodation that would occur through no necessary effort on MCI-WorldCom's part, but no less without it. This accommodation -- the direct result of market size, market share, and market penetration -- would allow MCI-WorldCom a continuing competitive advantage over all rivals (or those not similarly accommodated). Come again? A unified MCI-WorldCom wouldn't be the AT&T of old. And there would still be extensive competition throughout the telecommunication industry. All true. But the character of that competition would be fundamentally changed. Think about it this way. Today's best computer programmers are "free" to spend years devising "killer apps" for Macintosh. But they don't. Similarly, tomorrow's ISPs and packet technology manufacturers would be "free" to design products that didn't have the effect of making life easier for the new industry giant, MCI-WorldCom. But they wouldn't. And that's the point. (Or more precisely, it's half the point; the other half is this: for a corporation, an easier life has a tangible economic value.) Computer manufacturers, chip makers, software developers -- all the corporate creatures of cybercom -- will be irresistibly pulled in the direction of whatever technical standards WorldCom-MCI employs. The competitive environment of the "new Internet" will be distorted by the sheer size of MCI-WorldCom. It will dominate any market it elects to enter. It could even dominate those markets it was rumored to be "thinking" about entering. (What's Steve Case's presence on WorldCom's board likely to do to the price of AOL stock?) Angry software developers in Silicon Valley dubbed Bill Gates the "Viscount of Vaporware" -- a reference to Microsoft's lordly ability to squash prospective competition by merely announcing that it was considering the development of some new software application. A combined MCI-WorldCom would (rightly) engender the same response.Accommodation has two faces. The positive face is that it sorts out the messy, wasteful struggle over whose technical standards are to be employed by a large, multifaceted industry. It lets the rest of us get on with our business more rapidly. We may wind up using standards that are less than ideal, but we're not spinning our wheels waiting for someone else's war to end. If these standards are less than perfect, at the end of the day it just doesn't matter all that much (unless, of course, you're the patent holder of BetaMax rather than VHS, or Steve Jobs rather than Bill Gates). Letting us get on with things -- including, perhaps, discoveries and innovations that could in their turn spawn whole new industries -- is worth the price.Whether this price -- acquiescing to de facto monopoly -- is worth paying has an economic answer and a political answer. And they overlap. If the alternative to a little bit of monopoly is, say, an endless squabble between lawyers over what is or isn't the proper component of an integrated operating system -- an issue that might better be left to technologists and consumers -- it is tempting to avert one's eyes and cry "Stop!" After all, what has Bob Dole to contribute to the sum of human knowledge on this matter? Yet he is being paid handsomely, by Netscape and others, to help bend the solution in their favor. This is the worst of all worlds. And it is the positive face of accommodation.The dark side to accommodation is the creation of an economic environment that works to the continuing enrichment of the monopolist irrespective of his actions. Size has catapulted the firm into an alternative economic reality where simply to exist is to find oneself surrounded by other firms operating in ways that work to the economic benefit of the monopolist. (But, you say, competition is still there. If any one of these firms had the opportunity to supplant the monopolist, they would rush to do so. That's right -- but it's a different point.) Market size, market share, and market penetration all work together to the monopolist's advantage. The easy life -- the freedom to not worry about certain forms of competition, or the power to deter them -- has a positive economic value that's impossible to quantify but very real. And it persists for some version of a lifetime, if not forever.The purchasing decisions client managers face in cybercom are more technically daunting and financially perilous than any other decisions they make. Corporate spending on telecom will grow from $90 billion in 1996 to $150 billion in 2001. (And that's just in the United States.) Upgrading data networks will cost $40 billion a year. This is serious money.There is an old cliche: No one ever got fired for going with IBM. (A newer version would flog Microsoft or FedEx.) As cross-selling spreads, and bundling increases, the role played by brand recognition among nontechnical executives and directors will inevitably grow. This would create a self-reinforcing product selection advantage for MCI-WorldCom -- as would the growth of data networks generally, since corporate users increasingly make all-or-nothing buys, and that tends to reward size. And no one would be bigger than MCI-WorldCom.The economics of inevitability?To a considerable degree, the economic consequences of this merger are obscured by the competitive and technological upheaval that now consumes the telecom industry. They are obscured -- but they are not made less real. Policymakers are still obliged to weigh them as best they can.That brings us to the second argument in favor of laissez-faire: that the technology industry is somehow different from all other industries (with "network effects" and "increasing returns"), and that monopolies in technology are not only natural but beneficial (promoting efficiency through regularized standards). It is tempting to blow-off such claims as nothing more than the Nineties equivalent of supply-side economics -- a trendy bit of panglossian exceptionalism out of California. Unfortunately, increasing returns are real. They are not novel or unique to technology, but they are real. But so what? It's hard to point to a single instance where increasing returns played a decisive role in the development of industry -- in either this century or the last. As MIT economist Paul Krugman observes, it's easy to tell good stories about the importance of increasing returns. But "it's amazingly hard to identify a critical mass or a tipping point for an actual industry, even after the fact.... Indeed, increasing returns have traditionally been used as arguments against free markets, for government intervention. You may not believe that such intervention will work in practice, but that's a judgment about the rules of politics, not economics."And that brings us back, at long last, to those reinvented trustbusters at the Justice Department. Joel I. Klein, the assistant attorney general in charge of the antitrust division, recently told the New York Times, "We want to make sure innovation thrives because innovation is the greatest stimulant to economic development in these industries." And he described the Department's goal like this: "You want an environment where entrepreneurs believe that they can be the next Microsoft."This is an admirable way to put it. But it raises a question. In the software industry, Microsoft already is the next Microsoft. Maybe not forever, but for a very long time, integrated browser or not. And Justice has its hands full. No such sorting out has yet occurred in the Internet. But if the MCI/WorldCom merger is approved -- without any real examination or consideration of ameliorating actions that might be taken to lessen its impact -- the consequence will be to foreclose, to an entire industry of entrepreneurs, the opportunity to believe that they are working in an environment where they can be the next Microsoft. Because in that environment, at least, WorldCom-MCI will already be the next Microsoft. From the standpoint of politics, this might be an accommodating decision. But it would be the wrong one.Michael Barker and Randy Barber are both economists for the Center for Economic Organising and have advised the Communications Workers of America on certain aspects of the proposed MCI-WorldCom transaction.