The New Health Dialogue

The latest study released by the insurance industry, while better than the one that came before it, is riddled with flaws. The recent report produced by Oliver Wyman (an management consulting firm with an actuarial services arm) for the Blue Cross Blue Shield Association (BCBSA) is more reasonable than the PriceWaterhouseCoopers study issued by America's Health Insurance Plans (AHIP), but it, too, is designed to scare Americans into accepting the status quo rather than comprehensive health care reform.

Lets' be clear here: some of the best plans in the country are Blue plans, and we need all of them to get better post-reform. It's worth noting that the two things the Blues have been lobbying about most vigorously are the same positions that Wyman's analysis supports: (1) prevent the melding of the small group and individual markets, and (2) keep the insurance exchange as small as possible.

The report gets one thing right: reformers must make sure insurance coverage is affordable and that the mandate to buy insurance is enforceable so that most Americans get coverage. Otherwise, requiring insurers to sell to any customer who wants to buy (guaranteed issue) will risk the stability of the risk pool in the exchange. On this much we (and many others) agree.

What I cannot agree with is the attempt to sway the reform debate with data that cannot be checked and assumptions that seem to be designed to produce the results BCBSA wants us to believe. On many points Wyman and BCBSA could be correct. But since their data are not public, it is impossible to check.

These are the report's weaknesses:

The Wyman report ignores potential savings from the exchange or delivery system reforms. The Congressional Budget Office (CBO) found that exchanges would reduce the administrative costs of non-group plans. Clearly, health plans' administrative costs would be reduced with guaranteed issue, an end to health status rating, streamlined risk adjustment, and the ease of marketing within an exchange context. Indeed, these new regulations will reduce or even eliminate the profitability of traditional risk selection strategies. In addition, the CBO found that the Finance Committee bill would slow the rate of health care cost growth and begin to "bend the curve." Neither of these findings is reflected in the Wyman report.

The Wyman report uses proprietary Blue Cross Blue Shield plan data. It is impossible to know whether the data are representative because they are not publicly available and cannot be verified independently. This is unlike the Urban Institute's data in its HIPSM model. The Wyman report states that its model is based on:

... a database of actual claims, premium and underwriting information from over 375,000 small groups, representing 4.2 million covered lives, and 1.24 million individual policies, representing 1.6 million covered lives. The database includes blinded information on approximately 1-in-10 purchasers in the individual and small employer markets today. These data are representative of states across the country and reflect the varying rating rules that are used today. This allows the model to provide insight into the impact of reform at the state level.

The model uses 10 percent of Blue Cross Blue Shield data. Does this mean 10 percent of plans? Of states? Of enrollees? If only a few states are represented, how can we be sure that the market environments are representative of either today or what a reformed insurance market would be like?

The Wyman report overstates the impact of insurance reform on the non-group market. The Wyman report concludes that insurance market reform will raise claims costs in the non-group market. They reach this conclusion by assuming that average medical claims for the uninsured are 20 percent higher than claims in the current individual market. This appears to be pure assertion, unless the Blue Cross Blue Shield database somehow maintains data on the uninsured.

Misinformation about the non-group market has always been a problem. A few years ago, for precisely this reason, Health Affairs asked Mark Pauly and me to do a study of the non-group market. Health Affairs thought that since Mark had advised many Republicans over the years and I had advised mostly Democrats, if we could agree on the implications of the non-group market data that people on both sides of the aisle might believe our conclusions and stop arguing unproductively about unsubstantiated assertions. (This process is of course an option for people in Washington, DC, but I digress).

In our paper we concluded that one of the main reasons the non-group market fails to work well is because a majority of the relatively healthy uninsured have no effective way of reporting just how very healthy they are. This is why we don't see premium offers nearly as low as their far lower than average expected claims costs.

Consider this: data show that the uninsured use about half as much health care as the insured. Certainly some newly insured will be sick. But on average, when you include the mostly healthy uninsured that will be in the exchanges post-reform, it is very unlikely that the newly insured will cost 20 percent more than the currently insured. The report mentions the high cost enrollees who will come from high risk pools. There are fewer than 150,000 high risk pool enrollees nationwide. CBO estimates 25-30 million people will be in the exchange, a large percentage of whom are currently uninsured. Therefore, it is highly implausible that health costs will be 20 percent higher than today, even with current high risk pool enrollees.

The Wyman report conflicts with CBO's complete assessment of the individual mandate. The Wyman report makes an important point about the need to get as many people as possible into the risk pool. I certainly think that we should make coverage affordable and the penalties for not buying such that they are rarely used -- we would much rather people opt for coverage. The report, however, reaches a very different conclusion about the stability of the risk pool than does CBO, which considered the entire bill and set of policies.

Perhaps one difference in perspective is that consulting actuaries are naturally accustomed to advising clients to bid high or just exit when regulations get tighter. CBO knows that the individual and small group markets will largely reside in the exchange, that competition could be quite considerable within them, and that "bidding high when in doubt" in that context could lead to significant loss of market share.

Again, it is the totality of conditions -- underlying demand for health insurance, subsidies, penalties, market incentive for insurers -- that affects the effectiveness of the mandate. I am by no means saying I would not be more comfortable with more generous subsidies and stronger penalties. What I am saying is that CBO's analysis of the specific legislation in question is more complete than Oliver Wyman's, by its own admission.

The Wyman report overestimates the number of insured individuals in the individual market who have plans with an actuarial value below the reform threshold. Wyman asserts that 50 percent of the policies in the individual market fall beneath the actuarial value standard put forth by the Finance Committee bill. Remember, the Finance Committee package allows existing policies to be "grandfathered"-- if you like your coverage you can keep it. Even still, more general survey data suggests that around 25 percent of individual market enrollees have plans below the 65 percent actuarial value threshold.

The Wyman report does not consider subsidies when analyzing the impact of age rating. Wyman asserts that narrow age rating (2:1 or pure community) would be an unmitigated disaster for young people. We know that some insurers are generally supportive of large age rating bands (5:1), because they think young people will drop coverage if their premiums go up. Linda Blumberg of the Urban Institute just released a competing report (using representative data, methods that are explicitly described, and non-profit foundation funding). She found that if you account for the fact that many young individuals will be subsidized to purchase coverage, even pure community rating does not increase premium costs much for the vast majority of young workers or add to the aggregate cost of subsidies. Like PriceWaterhouseCoopers, Wyman computes the impact of age rating without accounting for subsidies. What is the point of that?

Wyman should be commended for improving on the PriceWaterhouseCoopers attempt. But this study is still sufficiently flawed and/or impossible to verify. It should not sway informed opinions about the implications of reform.

The Wyman report does strike a legitimate cautionary note about the importance of affordability of insurance and the enforceability of an individual requirement to purchase coverage. The Finance Committee bill can be strengthened. But there is no reason to panic. Pessimism and fear are not evidence.