To: Utah House of Representatives, Utah Senate,
Subject: MEDICAID EXPANSION UPDATE #7: BYU poll is likely biased due to misinformation
Date: Fri May 30 23:18:38 MDT 2014
You may have seen news reports and analysis of a recent BYU poll asking for the opinions of a sample of Utah voters about Medicaid expansion. The reported results showed majority support for expanding Medicaid. But is the poll accurate? Does it provide complete information for the Medicaid expansion options described, the lack of which could bias the poll results?
We think it provides incomplete information and, for this reason, we recommend you look before you leap in accepting its conclusions regarding the popularity of Medicaid expansion.
A clear problem in the survey is the cost information provided for the various Medicaid expansion choices. In the Medicaid expansion question (question #5/6 in the survey results memo) the survey describes four options for expansion – the Access Utah program, full expansion via traditional Medicaid, no Medicaid expansion, or full expansion via the proposed Healthy Utah Plan. While each description begins by giving poll respondents information about how much “federal Medicaid funds” are either accepted or rejected, only one description displays information about how much the plan will cost state taxpayers.
The description of the Access Utah program says it would cost $35 million in “state money.” But the survey is silent regarding state costs for full expansion via traditional Medicaid, which is projected to cost the state $60 million in fiscal year 2021. It also includes no mention of potential costs for full Medicaid expansion via the Healthy Utah Plan. The cost information provided thus implies that the Access Utah program will cost state taxpayers tens of millions of dollars, while full Medicaid expansion will be completely free to the state – an inaccuracy that would bias the survey results toward full Medicaid expansion.
The omission of state costs for the Healthy Utah Plan could be explained by the fact that state cost projections are not yet available because the details of the plan are still being negotiated with the Obama administration. However, to provide a complete and accurate impression of state costs for respondents in order to generate unbiased poll results, it would seem prudent to at least include a disclaimer on the likely eventual costs of the Healthy Utah Plan, such as this: “While no cost projections for the state exist for this option, under the Affordable Care Act states must eventually pay 10 percent of the costs of Medicaid expansion.”
A second, and much less certain, potential problem with the BYU poll was the wording of the question itself, which set the context through which respondents viewed the four Medicaid options. The question read as follows: “Several proposals for Medicaid funding were debated in the recent legislative session. Given the following options, which state Medicaid funding plan would you prefer?” Note that nowhere in that question does it mention the concept of expanding Utah’s Medicaid program. Given the fact that the poll analysis interpreted the results as a measure of support for Medicaid expansion, the lack of that term in the survey question is both noteworthy and strange. It is also possible that this word choice biased poll respondents’ views of the “Medicaid funding” options they were presented.
Because the options were labeled as “Medicaid funding plan(s),” it is reasonable to think that some respondents may have viewed the various options presented in the context of “do I support funding Medicaid?” There is a difference between that question and the question of “do I support expanding Medicaid?” Many people could logically, reasonably and consistently answer “yes” to the former and “no” to the latter, because the former could plausibly apply to the entire Medicaid program in Utah while the latter would only apply to a specific addition to that program. If this was indeed the case in the poll sample, it could have added to any bias created by the misinformation already discussed.
For example, if any Utah voters taking the survey viewed the Medicaid expansion question as asking about Medicaid funding in general, they might have been less likely to choose the status quo option. They might have viewed it as saying they oppose Medicaid funding generally, rather than just opposing Medicaid expansion – a term that was not given to them in the question. Again, this is only a possibility, and its likelihood is impossible to measure. But considering that most Utahns have a sincere desire to help their neighbors and to address social problems – as measured by its nation-leading rates of charitable giving and volunteerism – this issue could have had a significant influence on the poll results.
Suppose you were told in a survey that Utah policymakers had three options regarding Medicaid in Utah. First, they could spend $35 million in state tax dollars to provide partial health care benefits to 54,000 or fewer poor people. Second, they could spend zero state tax dollars and help no one. And third, they could spend zero state tax dollars and help 110,000 poor people get full health coverage in one of two ways. Which would sound the best to you?
In this light, perhaps the most surprising result from the poll was not that 76 percent of Utah voters polled supported some form of Medicaid expansion when provided misinformation about it, but rather that 24 percent chose something besides giving full health care coverage to Utah’s poor at absolutely no cost to the state.
Taken as a whole, we could reasonably conclude that the BYU poll results were biased by misinformation, and perhaps also by poor question wording. Rather than being a genuine reflection of where Utah voters stand on the issue of Medicaid expansion, the BYU poll results are more likely a reflection of the difficulty of conducting an accurate and informative poll. It should come as no surprise that large majorities of voters indicated support for Medicaid expansion when given incomplete information that creates a false impression of the impacts of Medicaid policy options.
The real question is this: What do Utah voters think when accurately informed of the context and impacts of Medicaid expansion in Utah – whether via traditional Medicaid or the Healthy Utah Plan? And on that question, the BYU poll provides little useful information.
Director of Policy
Director of Public Affairs