Implications of the Harvard Pilgrim-Tufts Merger

Share on Facebook
Share on Twitter
Share on

Today’s news brings word of a potential merger between Tufts and Harvard Pilgrim. For this observer, it raises the question of how much more concentrated the health insurance market will become if the #2 and #3 insurers are merged?

The standard measure of market concentration is the Herfindahl-Hirschman Index, which is the total of the squares of each participant’s market share. In a pure monopoly, the 100% market share translates to an index score of 10,000. In an almost total atomized market (100 participants each with 1% shares), you get an index score of 100.

To give a sense of perspective, the Federal Trade Commission considers a market with an HHI of between 1500 and 2500 to be moderately concentrated. And an HHI of over 2500 is highly concentrated. In concentrated markets, the FTC notes that any mergers that raise the index by 100 or more will receive significant scrutiny.

If you look at the existing market (see slide 7), its already pretty concentrated, with BCBS at 41%, HP at 14% and Tufts at 13%. Taking all the competitors into account, the market has an HHI score of 2164, meeting the FTC’s definition of moderately concentrated. The merger will result in a revised score of 2509, a jump of 345 that moves the market into the ‘highly concentrated’ category.

To be fair, most state health insurance markets are pretty highly concentrated. The AMA does a regular concentration analysis of each state’s market. Their numbers differ from mine (see explainer below) but should be consistent internally. In some states, like Maine and Alabama, their health insurance market is a practical monopoly with a single insurer holding close to 80% of the policies in the state. Massachusetts still won’t be close to that, even after a merger.

My point is not to defend or criticize the merger, rather its to note that a concentrated industry will become more concentrated, although still at lower levels than many other states. Implications for cost control and consumers? Don’t know.

Explainer: My HHI numbers are lower than the AMA’s and I suspect it is because they roll up the small insurers under a certain threshhold. When I roll up the insurers below BCBS, HP, and Tufts, then recalculate the HHI index, I get a 2010 HHI total for Massachusetts of 3102, which is consistent with the AMA’s 2007 figure of 3128. Using this assumption, the merger drives the HHI total to 3446.