Many pundits and policymakers have concluded that monopoly power is a huge economic problem resulting from government influence, digital network effects or the natural tendencies of the capitalist system. Some economists, however, have urged caution. It's possible, they argue, that big dominant companies gain market share not because the system is rigged or unstable but because some companies are simply a lot more productive than others. If those companies are more profitable, this story goes, it's because they put out better products or find clever ways to hold down costs.

There's also a third possibility that falls somewhere between these extremes. Though economists typically think only about how technology affects productivity, it's also possible that new technologies have altered the ways companies compete.

The internet makes it easy for companies to gain market knowledge, promote their products, advertise and manage supply chains across great distances. Whereas three decades ago expanding a store or brand to a new location required lots of boots on the ground, now much of the work can be done remotely. This is even truer of online services companies, which have little physical presence apart from their main offices. Meanwhile, the internet also allows rapid development of new products, letting companies easily enter one another's lines of business.

The internet thus puts companies in greater competition with one another than ever before. Businesses for which vast distances or niche specialization once provided a refuge are now forced to battle multiple rivals on a suddenly level playing field. And the companies that win in one place tend to also win everywhere because they have some edge in terms of quality, efficiency or innovation. So industrial concentration rises nationwide.

That's the theory, anyway, but there is some evidence to back it up. Several recent papers have found that even as national concentration has risen, local concentration has gone down. That suggests that a few juggernaut companies are competing away the hometown heroes in towns across the country.

Meanwhile, a paper by economists German Gutiérrez and Thomas Philippon found that the country's most efficient companies have experienced slower productivity growth since 2000 even as their size has increased. This could be because the top companies secured more intellectual property protections or government favors. Or it could be because technology helps them expand into more markets as their initial productivity edge goes from being a slight advantage to an overwhelming one.

Now, economists Philippe Aghion, Antonin Bergeaud, Timo Boppart, Peter Klenow and Huiyu Li have a theory attempting to explain how this process works. Their research suggests new technology lets companies expand more easily into new product lines within an industry. The companies that are more efficient win the competition. Their initial efforts to expand produce a burst of productivity growth as they strive to develop new products. But after they win in those markets, they slow down and become complacent superdominant incumbents. That's when monopoly rot sets in and productivity growth drops.

That story fits disturbingly well with the estimates by Gutiérrez and Philippon. They found that in the 1990s, superstars really did behave like superstars, making outsized contributions to productivity. But the effect reversed after 2000. And though Aghion and his colleagues didn't model geographic competition, the math probably works the same in many ways.

So it's possible that the superstar companies dominate American capitalism simply because the internet has made dominance easier. If that's true, antitrust crusaders and regulators need to rethink their approach. Breaking up behemoth companies by region, as was done with Standard Oil Co. and Bell Telephone Co., might not be as effective with modern superstars because it would be so easy for them to expand back into their old markets. Breakups by product line, such as a proposed separation of Facebook and Instagram, might suffer similar rapid reversals.

But if companies can't be dismantled by region or product line, can they even be broken up at all — or is regulation the only hope for taming them? Economists and legal scholars who are thinking hard about antitrust need to have an answer to this question. If information technology has fundamentally changed the way companies compete, then regulation needs to evolve to keep up.

Noah Smith is a Bloomberg Opinion columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.

Editor's note: Write for Business Forum. Submissions should run 750 to 900 words and avoid promoting individual businesses. Submissions will be edited for publication and run at the discretion of the editor. Submissions can be sent to doug.iverson@startribune.com.