Here are a few trends that have been going on in the United States over the past several decades: The share of income going to labor has declined. Corporate profits as a share of the economy have increased. And net investment (outside of the construction industry) has stagnated. Explaining all three of these facts in one coherent story about what’s happened to the economy since the late 1970s doesn’t necessarily fit with a few well-known hypotheses about the relationship between labor and capital.

First, take the research of Loukas Karabarbounis and Brent Neiman, both of the University of Chicago. In their view, the share of income has declined because capital has become less expensive due to technological progress. Computers have become much cheaper, so companies have responded by buying more computers and displacing workers. In economic models, this would require the elasticity of substitution between capital and labor to be greater than 1. In other words, companies being able to readily switch between capital and labor.

Karabarbounis and Neiman’s hypothesis makes sense when accounting for the decline in the labor share. But if that share were going down because firms were buying more capital equipment such as computers, then investment would be increasing as a share of the economy. But it’s not. We know that the price of investment goods is declining, so it looks like firms are just keeping the savings from the decline in the prices instead of investing more. That would explain the increasing profit share as well.

While the mechanics of how the labor share decreased are different, Research by Thomas Piketty of the Paris School of Economics also depends upon the elasticity of substitution between capital and labor to be greater than 1 to show that the labor share of income is declining, but he posits different reasons for this happening. His mechanism is that the rich increasingly save their income and the resulting deepening in capital results in a declining labor share. In his research with Gabriel Zucman of the London School of Economics, Piketty finds an elasticity greater than 1, but like Karabarbounis and Neiman, this result comes from looking at macroeconomic aggregates.

When the elasticity is estimated at a micro level, however, it looks like the elasticity is lower than 1, meaning that switching capital investments for human labor is not as productive. Ezra Oberfield of Princeton University and Devesh Raval of the Federal Trade Commission looked at how individual firms substitute between labor and capital and built that up to an aggregate elasticity that is less than 1. Yet they also found that the labor share has declined.

What could explain a declining labor share with an elasticity lower than 1? A look at the assets of corporations reveals a potential answer. Economist Robin Hanson at George Mason University reports on data that shows that 5/6ths of market valuation of the S&P 500 companies are now “intangible assets” such as patents and trademarks, intellectual property and, importantly in this case, brand names and business methodologies.

To understand this concept, imagine that you want to replicate a company. Let’s say it’s Facebook. You go out and buy all the assets of the company: servers, real estate, and the firm’s employees. But Facebook’s large amount of “intangible assets” means that if you purchase all these physical assets you’d only have a firm that was less than the original firm. The original Facebook obviously has something intangible that can’t be replicated. For the S&P 500 overall, the replicated firm would only be worth 1/6 of the original firm.

Hanson runs through potential explanations for why these intangibles have increased in importance. They include patents, branding, and monopoly power. His colleague at George Mason University, Tyler Cowen, does warn that part of this increase in intangible assets could be from measurement error, but still it seems that any one of these reasons would mesh with the three theoretical explanations above about why capital investments are low yet the labor share of income is still declining (with an elasticity below 1).

Especially when it comes to monopoly power, increased profits and stagnant investments are the sign of monopolies. In his widely read research on the labor share, Massachusetts Institute of Technology PhD. student Matt Rognlie points to a potential increase in market power among companies.

If market power is a significant reason for the increase in intangibles, then this means economists need to focus on understanding how market power and monopolies work in the modern age. Can we do anything about the natural monopolies that arise from social networks companies? Would we want to? Those and many more questions would have to be dealt with.