People walk by an AT&T retail store, Monday, Oct. 24, 2016, in New York. AT&T plans to buy Time Warner for $85.4 billion.

Concerns about business consolidation and increasing market power have entered the center of the U.S. policy conversation. Consider the reaction to the proposed merger of AT&T Inc. and Time Warner Inc. where both presidential campaigns have raised concerns about the consolidation of the two large companies. The traditional argument for mergers and acquisitions is that the joining of two companies will lead to efficiency gains that will help customers and the economy overall. But some researchers and policymakers are increasingly skeptical that increasing consolidation will lead to such efficiency gains. A newly released research paper will give them reason to retain their skepticism.

The new research by economists Bruce A. Blonigen of the University of Oregon and Justin R. Pierce of the Federal Reserve Board was released yesterday as a National Bureau of Economic Research working paper. The two economists look at how mergers and acquisitions affected productivity and market power in the manufacturing sector from 1998 to 2006. The authors specifically look at how productivity (both labor productivity and total factor productivity) and markups (the difference between the price charged by a business and the marginal cost of producing a good) change in individual factories that are part of companies that merge or are acquired. They compare the changes to a number of other groups of plants to tease out the effects. As Noah Smith, writing about the paper at Bloomberg View puts it, they “compare factories that get acquired in a merger to similar factories, and to factories for which a merger has been announced but not yet completed.”

What the two economists find is that the efficiency gains that were supposed to appear didn’t in fact happen. This study found that increased consolidation doesn’t have any significant effect on plant-level productivity. What it does have an effect on is prices (and profits): markups increased due to the consolidation of manufacturing firms. The increase in markups appears quite significant: 15 percent to 50 percent of the average markup in the data. In other words, the price of manufactured goods increased while there were no productivity gains. The increasing amount of rents in the manufacturing sector could very well have contributed to higher income inequality, either intrafirm or interfirm, depending on how the gains were shared within the company.

Interesting, Blonigen and Pierce are able to look at the effects of different kinds of mergers and acquisitions. They looked at how these effects on productivity and markups might change if the merger is horizontal—between two companies in the same industry—or if they were consolidation across industries. They find the productivity effects are negative when mergers are horizontal, but there appear to be some productivity gains when the mergers are non-horizontal.

This new research gives policymakers reason to be wary of mergers. Of course this paper looks only at the manufacturing sector and researchers would be well served to try to extend its approach to other parts of the economy. Interest in the effects of market power isn’t going away anytime soon.