Filed Under (Health Care) by Nolan Miller on Oct 29, 2009
This week Jeff wrote about competition in the health insurance industry, wondering if the health insurance industry, while not competitive, might be “competitive enough,” and whether, if it isn’t, whether the government will properly price the public option it hopes will increase competition.
I don’t disagree with any of Jeff’s points. But, I thought it might be useful as a follow up to write about how competitive the health insurance industry is. It just so happens that a new National Bureau of Economic Research Working Paper by Leemore Dafny, Mark Duggan, and Subramaniam Ramanarayanan entitled “Paying a Premium on Your Premium? Consolidation in the U.S. Health Insurance Industry” provides new evidence on exactly this point.
But, first, a bit of background. Economists, and particularly the U.S. Department of Justice, measure concentration in a market by summing the squared market shares of all of the participants. We call this number the HHI. DOJ calls a market highly concentrated if the HHI is over 1800. This would be the case if, for example, there were four firms in the market, each of which has a 25% market share (25*25 = 625, 4*625 = 2500). In highly concentrated markets, the DOJ has “significant competitive concerns” about any merger that raises the HHI by more than 50 points. The AMA’s 2007 study on “Competition in U.S. Health Insurance” found that between 96 and 100 percent of health insurance markets were highly concentrated, depending on the market definition. In terms of market shares, between 64 and 76 percent of health insurance markets have at least one insurer with market share greater than 50 percent. In around 10 percent of markets, a single insurer has over 90 percent of the market share.
The AMA data suggests that, by any measure, health insurance is dominated by a small number of players. This does not, in itself, suggest that there is a lack of competition in health insurance. After all, if there are just two insurers who compete in a cut-throat fashion, this will quickly drive prices to competitive levels. Or, if there is a threat of entry into health insurance, then potential competition might discipline prices even though those firms do not actually enter the market. But, the AMA numbers certainly raise suspicion.
The real question regarding concentration in health insurance is whether the high concentration levels in health insurance lead to higher premiums. The Dafny-Duggan-Ramanarayanan paper looks at this question. They use the 1999 merger between Aetna and Prudential, two of the largest health insurers in the country, to investigate it. They find that the Aetna-Prudential merger did increase premiums and they estimate that, overall, the wave of insurer consolidations that occurred between 1998 and 2006 increased premiums by about 2 percentage points on average. In addition, they found that increases in concentration increased the bargaining power of insurance companies with respect to healthcare workers, resulting in lower employment, and lower earnings for physicians.
So, the evidence suggests that the health insurance industry is highly concentrated and that high concentration is probably leading to higher prices. But, there are several things we still don’t know. First, do those higher prices translate into abnormally high profits? Maybe the reason why prices are high and there are not a lot of firms operating in any region is that additional firms cannot profitably enter the market. In this case, the high levels of concentration and prices would be evidence of market forces working, rather than a sign that further regulation is needed. Second, if the insurers are earning abnormally high profits, why don’t other firms come along and enter the market, driving down prices? If the lack of entry is due to practices that are prohibited by current law, then greater enforcement might improve competition. If, on the other hand, they are due to features of the competitive playing field such as the ability to risk-rate premiums, deny coverage based on pre-existing conditions, difficulties in switching insurers do to the connection between insurance and employment, etc., then addressing these issues (as is done to one extent or another in all of the reform bills currently under consideration) might greatly improve matters.
Jeff’s overall point is still one I agree with. The government’s success in designing insurance schemes has been mixed at best. If the government doesn’t know why our current system fails to be competitive (if, indeed, it does), then it is unclear how introducing a government plan, in and of itself, is going to fix things. In the absence of a clear theory of what is wrong with the market and how a public option solves the problem (rather than vague statements about how it will lead to greater competition) I strongly suspect that we’re more likely to successfully improve health insurance and health care by adjusting the rules of the game and improving the government’s role as referee, rather than by introducing the government as a player.