Over the course of only a few years, Americans have seen a revolution in wireless technology. We have moved from a world in which mobile devices were about making phone calls to one in which they are about having a portable computer with you wherever you go.

This tremendous innovation is a result of fringe competition in the marketplace. It was outsiders like Apple and Google and fringe competitors like T-Mobile, not Verizon or AT&T, that drove this revolution. AT&T's proposed merger with T- Mobile, however, has the potential to fundamentally alter the wireless marketplace in a way that threatens innovation.

The wireless market today is concentrated. AT&T and Verizon function as a duopoly with fringe competition from Sprint, T-Mobile and a few local firms.

Economic theory has trouble predicting how duopolies will behave. Sometimes they work in parallel and act like monopolies. In other cases, they compete vigorously. But economics does teach us that fringe competition is critical to deciding whether duopolists will compete or collude. Fringe competition does not only influence price; it can – and does – drive innovation.

Take T-Mobile, for example. At a time when the market seemed to be settling into a "one-smartphone-per-provider" routine, the low-cost, high-customer satisfaction provider moved us in a different direction by being the first company willing to sign onto a new alternative platform, Google's Android. T-Mobile brought a new platform and a new provider, an opportunity that only came to fruition because of its existing national wireless network. And in doing so it exploded the idea that if I want phone X, I must sign up with wireless carrier Y.

We need innovation in our applications, devices and platforms. The history of innovation – not just in the telecommunications industry but more generally – suggests that it is competition, not market dominance, that's going to drive that new innovation.

Incumbent wireless carriers have no incentive to allow "tethering" apps that permit you to access the Internet from your laptop via your wireless connection.

Nor do they want voice-over-Internet competitors like Skype to flourish. In general, established incumbents have less incentive to allow these new uses where they compete with existing businesses. Without some sort of outlet, innovative companies have less incentive to invest in these kinds of transformative new technologies if they know that there's not a platform in which they can place that technology. But smaller, hungrier fringe competitors like T-Mobile just might allow those things.

And in doing so they may force their bigger rivals to compete.

Allowing the merger of AT&T and T-Mobile effectively means we've given up on the competition experiment in the wireless marketplace. Sprint may not be able to survive as an effective fringe competitor in the combined AT&T/T-Mobile world. The lesson of history is not only that concentrated markets produce higher prices, but that they produce less disruptive innovation.

If we do give up on competition and allow this merger, moving back toward the Ma Bell days, maybe it's time to start regulating again. Without small and nimble competitors that can adapt, encourage and utilize new technologies, the industry and the government will be forced to turn their attention to regulating the dominant carriers that remain.

We may need rules that permit the connection of any device or app to a network, for instance, just as the Federal Communications Commission had to require the old Ma Bell to allow interconnection. We would need a wireless network neutrality regime to permit companies from doing exclusive deals with particular apps and shutting out others.

Network neutrality has proven controversial and politically challenging in the land-line world, and it will likely be the same with wireless. Regulation is certainly not the best option. We would be better off with competition in the wireless market and the innovation it fosters. But if we allow AT&T to buy out the competition, regulation is our only other choice.

© Copyright The Sacramento Bee. All rights reserved.

Mark A. Lemley is the William H. Neukom Professor at Stanford Law School, where he teaches intellectual property, antitrust and Internet law.

• Read more articles by Mark A. Lemley