For years, economists have believed that competition tends to equalize profits across firms, as inefficient firms either learn from better ones or go out of business, and as new firms enter markets and so compete away high profits. Several things, however, suggest that we need to change our mental model, because this basic common sense intuition might be wrong.

A few things I've seen recently point to this. Jason Furman and Peter Orszag say (pdf) "there has been a trend of increased dispersion of returns to capital across firms" in the US. David Autor and colleagues show that there has been a rise (pdf) of "superstar firms" earning very high profits. And Andy Haldane, echoing Bloom and Van Reenen (pdf), has said (pdf):

An upper tail of companies and countries has maintained high and rising levels of productivity. These productivity leaders are pulling ever-further away from the lower tail. Or, put differently, rates of technological diffusion from leaders to laggards have slowed, and perhaps even stalled, recently.

All this is the exact opposite of what we'd expect to see if competition equalized profits. So what's going on? I suspect that here, as everywhere, there are no mono-causal explanations. A few possibilities are:

- Low interest rates and creditors' forbearance mean that inefficient companies aren't being killed off to the same extent they were in the 80s and 90s.

- Credit market imperfections prevent efficient firms starting or expanding (though this should be less of a problem now than a few years ago).

- Bad managers just can't see how to up their game or exploit profit opportunities.

- Tough intellectual property laws impede firms from learning.

- Contrary to what Hayek thought, prices are not sufficient information. High prices don't necessarily encourage firms to enter a market because they don't tell us how long the profit opportunity from such prices will last.

- Many profitable companies operate in niches which are big enough to offer nice profits, but too small or too specialized to attract entrants: each week, the IC describes dozens of such firms.

- New technologies enjoy increasing returns to scale. As Stian Westlake and Jonathan Haskell say: "Intangibles are often very scalable: once you've developed it, the Uber algorithm or the Starbucks brand can be scaled across any number of cities or coffee shops." This creates a winner-take-all effect.

- Network effects generate an incumbency advantage. If I set up a more efficient supermarket than Tesco, I should win enough business to get by, and to force Tesco's prices down. But if I set up a potential rival to Facebook, I face a tougher job. Because the value of a networking site depends upon others using it, Facebook has a more entrenched advantage.

If all of this is right, and these are long-lasting changes - which is a big if - then we need to ditch some old mental models and rediscover some others. We should abandon the idea that competition equalizes profits and restrains monopolies and perhaps return to some older Marxian ideas.

Marx thought capitalism tended to generate monopolies. Some of his followers, most famously Baran and Sweezy, thought this would generate a tendency towards stagnation, as those monopolies generated more profits than they could spend.

For a long time, these ideas were mistaken. However, at a time when we're seeing superstar firms build (pdf) up massive cash piles amid talk of secular stagnation, perhaps we need to rethink their relevance and implications.