May 14, 2012 at 12:45 pm
Austin FraktI’m pleased to announce that Aaron and I will be posting several times a month on the AcademyHealth blog. My first post appears today. It’s a lot about what insurers and a little about what hospitals do and don’t do with market power. It summarizes some recent publications and includes a must-see chart. Go read it!
Comments closed
by Steve on May 14th, 2012 at 14:49
No matter what shape heajth care reform takes in the future, i think we have pretty good evidence that we should eliminate tbe individual market. Exchanges look to be the most acceptable venue.
Steve
by Lisa Simpson on May 14th, 2012 at 14:57
Austin and Aaron, delighted to be partnering with you to help make evidence accessible and informative for the tough challenges facing us today!
by Jonathan H on May 16th, 2012 at 07:17
I left a comment on the other site, but wanted to engage here as well. The hospital data and conclusions are well-established, but your conclusion about insurer market power is not nearly so well-supported. Why did you look at the rate of insured, rather than the thing you want to measure? That would be either the amount of premium, or the profit margin of insurers, or both.
It is possible to have an insurer with 80% market share in a state, with 85% MLR, 3% net income (typical numbers) and yet a higher than average rate of uninsured because of social, demographic and political conditions in the state.
I can think of many reasons why states with market dominance by a single insurer would have higher rates of uninsurance. A state with a single dominant insurer is almost always a rural state, and that insurer is a Blues plan, and it is almost always a “red” state (consistent with your point that the effect you see occurs in low regulation states). But we know these states have more poverty, fewer unions, less highly educated workforces and loads of other factors that correlate with a smaller percent of companies offering substantial coverage, and a smaller percentage of employees taking them up on the offer if there is one.
by Austin Frakt on May 16th, 2012 at 07:40
Did you read the paper? Did you see the list of control variables? The state and year fixed effects? The IV approach? You may still not be convinced, and I’m not saying you should be. All studies have flaws. But please don’t presume none of your concerns were considered by the authors. Please tell me you at least looked at the paper. There is an ungated version, after all.
As for why I wrote about the papers I did: they’re all very new and I thought the world should know about them.
by Jonathan H on May 16th, 2012 at 21:05
Properly shamed, I did go and read the Bates et al paper. They did a better job separating employer sponsored from individual insurance than I assumed, but other than that your summary encapsulated the key points, and I still don’t agree with the approach. Or maybe it’s that I don’t get it.
If you want to look at whether market power brings insurers to raise their premiums more than insurers in more competitive environments, then study that, not a secondary effect on the rate of people with insurance, which introduces more complexity. Do health insurers in states with a high HHI have higher average profit margins (weighted by market share) than insurers in states with low HHI? What about the correlation between HHI and AER (administrative expense ratio) as a proxy for efficiency of operations?
And I do want to repeat the point about Blues plans. Is there even one state with an HHI over 2,000 that doesn’t have a Blues plan as the dominant insurer with over 50% market share? Most of these Blues are still non-profits and have fairly low net income, and most of them are in rural or semi-rural states. If they are failing people, it is primarily in failing to negotiate more aggressively on prices with providers and more generally failing to control the growth in health spending. It isn’t in padding their net income or letting their AER get bloated compared to more competitive markets (though there may also be effects here that are worth bringing out).
By the way, not so much excuse as explanation: I used to practically bathe in this data about market share and cost trends a few years ago when I was in the strategic planning department of an insurer. That’s why I felt comfortable firing off some objections without reading the article (and didn’t even notice it was ungated). I don’t have the data at my fingertips any more, sadly, or would have cited some figures.
by Rex on May 17th, 2012 at 06:56
Jonathan, thank you for your comments on my paper with Laurie Bates and Jim Hilliard. If the data were available we might have looked at the profit and expense preference behavior consequences of health-insurer market concentration in another paper. However, in that paper we were more concerned with the welfare implications and not the private efficiency consequences of health-insurance market power.
by Jonathan H on May 19th, 2012 at 23:16
Thanks for responding. Comprehensive data may not be available for free, but data definitely is available. Most health plans, even nonprofits, provide financials in their annual reports. Also there are companies like Sherlock and HealthLeaders-InsterStudy that have relationships with insurers and provide this data for a fee. Perhaps as an academic researcher you could get this for free or a reduced price.
by Jonathan H on May 19th, 2012 at 23:57
These are only two data points, but the examples of two most dominant insurers by state support my reservations. The blues in North Dakota and Alabama control roughly 90% of the insured market in those states. These states have mid to high rates of uninsured, so looks to support your thesis. Yet, the net income for both these insurers has been around 1% or lower for years. And both plans have lower than average administrative expense ratios. For Alabama, the MLR is about 90%. https://www.bcbsal.org/about/facts.cfm