It was 30 years ago that lawyers from the Justice Department first went into court to ask a federal judge to break up AT&T, and much has happened since then. The breakup of the Bell system into long distance and the regional “Baby Bells.” The rush of new competition and the resulting drop in prices. The decline of wire-line service, and the explosion of voice and data over wireless broadband.

The irony is that after all that entrepreneurial energy and technological innovation and fierce competition, the telephone market once again threatens to consolidate back into the hands of two national giants, AT&T and Verizon, which are direct corporate descendents of Ma Bell.

That is the context to keep in mind as the Justice Department and the Federal Communications Commission consider AT&T’s proposed $39 billion acquisition of T-Mobile.

The government has two choices.

It could stick with the competitive, lightly-regulated model and try to make it work by blocking a merger of the No. 2 and No. 4 competitors that will leave 75 percent of the wireless market in the hands of AT&T and Verizon.

Or it could acknowledge that, because of powerful economies of scale and limits on the amount of wireless spectrum that is available, the “telephone” market is a natural oligopoly that can support only a handful of players and limited competition — and, by implication, requires much stronger government regulation.

The best support for my formulation is to be found in the filing by AT&T and T-Mobile seeking FCC approval of their deal. In it, the companies make clear that, with the advent of handheld devices that can be used for downloading videos and exchanging e-mails and browsing the Web, the name of the competitive game is spectrum. The companies that have the most of it, and have the capital and the economies of scale to take the greatest advantage of it, will be the survivors in the coming competition. Only they will be able to offer the nifty new applications on a national basis, and by doing so grab market share from those who do not.

Because it has neither the spectrum nor the willingness to invest the needed capital, T-Mobile is not viable — or so the companies claim — while even AT&T now finds itself running out of spectrum in key markets because of the explosive growth in spectrum-hogging applications. A merger magically solves the problem for them while hastening the rollout of a national, next-generation wireless broadband.

The unstated implication of the AT&T filing is that in a world of iPhones and iPads, the limit on spectrum forces a tradeoff between the number of firms that can compete and the level of service that those companies can provide. With fewer companies and less competition, there will be enough spectrum for all of them to offer 4G-type service across the country. With more firms and more competition, by contrast, there won’t be enough spectrum for many, if any, to offer the latest broadband applications.

AT&T, of course, doesn’t want anyone to draw such implication, because if you take it to its logical conclusion, it means its merger with T-Mobile must be deemed bad for competition. After all, if T-Mobile, as the fourth-largest player in the industry, is too small and has too little spectrum to be viable, then how can it turn around and say that much smaller players such as Leapfrog, Cellular South and LightSquared will provide all the competition that is necessary to keep AT&T honest? The answer, of course, is that before very long they won’t, which is why on standard antitrust grounds, this merger ought to be a non-starter.

A lack of spectrum, however, is not the only thing that disadvantages the small players. Most of them are also regional in scope, which means that when their customers travel to other cities, they must rely on the network of competing firms for service. Without such a “roaming” capability, the smaller firms’ business would quickly dry up.

Until recently, because no company had enough spectrum to build a national network, even the biggest players had to negotiate roaming agreements. They also all had to offer phones that worked on all the networks and all parts of the spectrum.

But since 2001, when a Republican-led FCC lifted caps on the amount of spectrum any one company could control, that has changed, particularly as it relates to next-generation data service. Now, through spectrum swaps, mergers, partnerships and purchases of large blocks of new spectrum at government auction, Verizon and AT&T have each assembled enough concentrated spectrum to be able to offer coast-to-coast service on their own next-generation networks. For them, roaming and equipment interoperability are no longer so important.

Not surprisingly, then, the two industry leaders recently began demanding hefty roaming fees from small competitors for use of their networks, since no reciprocity is needed. The FCC responded with a proposed ruling requiring them to reach roaming agreements at reasonable rates, a rule that both companies vigorously opposed. And after the FCC adopted it anyway, the always-litigious Verizon filed suit in federal court to overturn it.

Equally interesting is that Verizon and AT&T are now offering phones that work only on their networks but not on those of their smaller competitors, which disadvantages the small guys in two ways.

Now, when the smaller companies go to the manufacturers of handsets and chip sets, they can no longer just order the same sets that Verizon and AT&T order and get the advantage of the same low price. Instead, they have to pay significantly higher prices to get the manufacturers to produce different versions of the phones just for them. Some even say that they can’t get phones at any price.

By selling devices that work only on their part of the spectrum, AT&T and Verizon have also created a big disincentive for their customers to switch to one of the smaller providers when service contracts expire. To do so would mean throwing away a perfectly good handset and incurring the expense of buying a new one that works on all networks.

Such behavior by the industry giants can be found in almost every industry, of course, but over the years it has proven to be particularly effective in telecom, where there are huge efficiencies and other advantages that flow from being big. These scale advantages are the reason there has been so much consolidation in the industry over the past 20 years, why AT&T is willing to pay $39 billion to buy what it claims is a struggling competitor, and why, if it succeeds, Verizon will try to regain the top position by buying No. 3 Sprint.

The situation was much the same in 1913 when, in response to the first antitrust suit against AT&T, President Theodore Vail convinced the Justice Department that the phone business was a natural monopoly and that AT&T ought to be it, under government regulation. Today, there is probably competitive headroom for AT&T, Verizon and maybe one other firm in the market for wireless voice and data services, at least until the next technological breakthrough comes along.

And therein lies the central question in all this: Which arrangement — a tightly-regulated oligopoly or a lightly-regulated market with numerous firms of varying size — is most likely to produce the next innovation that improves services while lowering costs? At different times, we’ve had success with both models, but surely the worst outcome would be the unregulated oligopoly that AT&T and Verizon would have us embrace.