For a sense of how research is being misrepresented in the health-care debate, look no farther than the campaign for a government-run health plan.
Without such a "public option," it's claimed, some states will have too few choices among private health insurers. The popularity of some health plans is offered as evidence of monopoly and high premiums -- rather than evidence of superior performance in keeping customers satisfied.
Wall Street Journal columnist David Wessel writes, "There are plenty of places where one or two insurance companies dominate the market, much to the consternation of employers and consumers. 'Most Americans live in markets dominated by a small number of insurers, and most markets are becoming more concentrated over time,' Northwestern University economist Leemore Dafny and two colleagues reported in a recent National Bureau of Economic Research working paper. 'Increases in concentration do raise premiums,' they said."
Actually, the authors of that paper wrote, "We do not find evidence that premiums are rising more quickly in markets that are becoming more concentrated." On the contrary: "Premiums are not rising more quickly in markets experiencing the greatest increases in concentration" [emphasis in the original].
What part of the word "not" is difficult to understand?
Wessel's selected quote ("increases in concentration do raise premiums") refers to a single unique merger. Out of many mergers among health-insurance companies and HMOs, the study found "only one" -- the 1999 merger of Aetna and Prudential Healthcare -- that significantly increased the merged firms' share of a few local markets. They contend that this briefly raised premiums, which is the basis of Wessel's quote.
Yet the authors also note that the merger had "no lingering effect on market concentration after 2002." Enrollment in the newly combined firm reached 21 million in 1999, but "fell rapidly thereafter [to] 13 million in 2002." If the combined Aetna-Prudential plans actually "dominated" many markets, how could they possibly lose 8 million (38 percent) of their enrollees in just three years? Companies without serious competition can't easily lose business to competitors.
Wessel is alluding to a new Dafny paper, with Mark Duggan and Subramaniam Ramanarayanan, which claims, "Markets are becoming more concentrated over time." Specifically, the trio say: "The four-firm concentration ratio for the nation as a whole... increased from an impressive 58 percent in 1998 to 79 percent in 2006."
In fact, the Census Bureau survey "Establishment and Firm Size" reports the four-firm concentration for health and medical insurance carriers was only 23.5 percent in 2002. Dafny & Co. inflate the figure to 79 percent by defining the market in a uniquely imaginative way.
As Dafny herself noted in another paper earlier this year, 82 percent of large firms don't pay premiums for employee health insurance, but instead "self-insure" by paying benefits themselves (while still contracting with an insurance company to handle the actual claims process), though many also offer their employees the option of joining an HMO.
Yet Dafny excluded self-insurance, so that HMOs accounted for more than 90 percent of the plans in the study -- even though only 20 percent of those with employer-sponsored health insurance choose an HMO, according to the 2009 Kaiser Family Foundation survey.
It is clearly nonsensical to define competition in health insurance as if HMOs accounted for more than 90 percent of "the market."
In any case, it's absurd for Wessel to cite these studies of large-company health policies -- because that market is totally different from the one for non-group customers that the "public option" is supposed to serve.
Then, too, each of the Dafny studies "does not examine the relationship between market structures and premium levels." In other words, she didn't claim premiums are higher in markets in which a few major insurers capture most of the business, nor did she offer evidence of Wessel's alleged "consternation of employers and consumers."
Those who fulminate about "market concentration" in health insurance -- including Leemore Dafny, the American Medical Association and economist James C. Robinson (who holds the Kaiser Permanente chair at UC Berkeley) -- haven't demonstrated any connection between premiums and their antiquated measures of market dominance.
Premiums are not higher in states where one or two health plans are most popular, and (as the latest Dafny paper said) "premiums are not rising more quickly in markets experiencing the greatest increases in concentration." Columnists and editorial writers who suggest otherwise are abusing bad economics to promote bad policy.