The MacIver Institute, a conservative advocacy group which describes itself as a “think tank,” issued a press release claiming to find holes in a Citizen Action of Wisconsin research report which reveals shockingly large disparities in health insurance costs between Minnesota and Wisconsin on the new marketplace created by ObamaCare. The Citizen Action report identifies key policy decisions made by the Walker Administration that are substantially contributing to higher rates in Wisconsin.
Not unlike the right-wing communication infrastructure’s persistent efforts to deny the science behind climate change, we find nothing in MacIver’s critiques that should lessen concerns that Wisconsinites will be paying much higher health insurance rates than Minnesotans. MacIver’s critiques largely nibble at the margins and attempt to cast doubt rather than providing alternative analysis and explanation. Even if the MacIver response is taken at face value, at best it means that the large disparity between Minnesota and Wisconsin is somewhat smaller than we report, but not that it does not exist. Is MacIver saying there is no rate gap, that there is a somewhat smaller rate gap? They do not say. Is MacIver saying that it is impossible to tell? If so, that would be absurd.
We will examine first MacIver’s attempt to refute the causes of the rate disparity our report establishes, and then their attempt to challenge the rate gap itself.
The Citizen Action report found two decisions made by Governor Walker that can partially explain the rate disparity with Minnesota: (1) Wisconsin’s rejection of enhanced federal Medicaid dollars; and (2) Minnesota’s more robust use of rate review as a premium reduction tool.
First, MacIver commits a logical fallacy regarding correlation and causation in attempting to cast doubt on our conclusion that Governor Walker’s rejection of enhanced federal Medicaid dollars can explain 8-10% of the 79% to 99% health insurance rate gap between Minnesota and Wisconsin. Our estimate came from a Rand Corporation study, and is based on the well established research that lower income people tend to have more health conditions, which will increase cost for the health marketplace insurance pool when they are not covered by Medicaid. MacIver counters that some states with relatively low premiums did not accept the enhanced federal Medicaid dollars while others with relatively higher premiums on the new marketplace did. This is a non-sequitur because a variety of other factors influence premiums, including underlying medical costs in a state, the age and health of the overall population (of which Medicaid eligible people are only one slice),and the efficiency of the state’s medical delivery system, impact a state’s overall cost ranking. Medicaid expansion is not the largest factor in premiums, but it is a significant factor. A state with low relative premiums because of other factors could have even lower premiums if it accepts the enhanced Medicaid dollars. A state with higher overall costs would have somewhat lower premiums if it accepted enhanced federal Medicaid dollars, but not enough to cancel out other variables. MacIver offers nothing that can challenge the underlying rationale of the Rand Corporation’s findings that rejecting federal Medicaid dollars increase rates on a state’s health insurance exchange.
Second, MacIver claims that Wisconsin has robust rate review, citing Gov. Walker’s Office of the Insurance Commissioner and a hospital official. Their argument is that Wisconsin is following the letter of the law. This confuses mere legality (which we do not take a position on in our report) with rigor and robustness. MacIver does not challenge our contention that Minnesota’s rate review process is much more aggressive, and has led to tangible reductions in premiums. As the Citizen Action report points out, Gov. Walker’s state regulators have yet to find a single health insurance rate submission excessive, which is shocking evidence of regulatory passivity, and has further not tried once to use their power to block any insurance rates. In an earlier report we questioned whether the Walker Administration really is following the letter and spirit of the law, but that is a separate issue from whether Wisconsin’s process is more passive. MacIver does not answer our report’s finding that Minnesota has rejected a number of insurance industry rate submissions as excessive, and won up to 37% premium reductions as a direct result. MacIver also fails to challenge our finding that Walker’s regulators have yet to find a single health insurance company rate submission to be excessive in cost.
MacIver also attempts to question whether there is actually a rate disparity between Minnesota and Wisconsin on the new health insurance marketplaces.
First, MacIver attempts to question our data, but provides no data of its own challenging our finding that there is a large unexplained disparity in rates between two similar states with similar underlying medical costs. For our report we use the best available data from the federal government which weights rates based on population. Our data comes directly from the federal rate chart which lists the weighted rates of each state. There are other data sets we examined which do not weight rates by population (and therefore under-weight the major cities in each state) which also show dramatically lower rates in Minnesota than Wisconsin. As MacIver neither effectively challenges the existence of a large and unexplained disparity in rates, nor provides its own numbers, it is unable to refute the main finding of our report. Policymakers from all perspectives should be very concerned about a large unexplained disparity which makes Wisconsin health insurance much more expensive than a neighboring state, especially when that state has similar underlying medical costs. At best, MacIver’s criticism could mean the rate is somewhat smaller than our report finds, but the disparity is so large that there is still a large gap which demands attention.
Second, MacIver attempts to argue that our data is out of date, even though the data came directly from a U.S. Department of Health and Human Services (HHS) report released in September. MacIver incorrectly states that the “enrollment” and “premiums” data came from a Rand Corporation report (the same report cited above on Medicaid) which cautioned that they can not have absolute certainty about the precise rates that would be charged. However the HHS report makes clear that the data came directly from actual premiums for 2014 submitted by the states. HHS did use a Rand Corporation model on age distribution within each state to weigh the cost of premiums in each state. This was necessary because rates do increase based on age. As this is an economic model, it is certainly an estimate. This in no way means the information is not useful for policy analysis, as it is fully accepted in the social sciences that economic models are a legitimate and essential policy analysis tool. As in their first critique, MacIver offers no alternative numbers. In addition the order of magnitude in Citizen Action’s numbers is so great (a 99% to 79%) that even if it were 50% off in either direction (which is unlikely) the health insurance cost disparity between Minnesota and Wisconsin would still be highly significant and worthy of attention. The most fundamental flaw here is that MacIver confuses a reasonable level of social scientific uncertainty, which is always present and admitted by qualified researchers such as those who conducted the Rand report, with an absence of evidence or proof.
In conclusion MacIver fails to demonstrate that there is not a large health insurance rate disparity between Minnesota and Wisconsin, and offers no-counter data of its own. In addition it fails to offer any of its own policy explanations. While it is possible to debate both the magnitude of the disparity, and the causes, MacIver adds almost nothing to this important discussion, and does little more than provide cover for policymakers in Madison who want to sweep the problem of higher health insurance costs in Wisconsin under the rug.