Thursday, February 27, 2014

Newly Released Kaiser and CBS/NY Times Polls

Two polls -- Kaiser Foundation and CBS/New York Times -- dealing (in whole or in part) with health-insurance issues were released in recent days. Two points grabbed my attention.

First, the two polls showed virtually identical numbers on Americans' preferences for retaining, retooling, or repealing the Affordable Care Act. In the Kaiser poll, 8% want Congress to "Keep the law as it is," whereas another 48% want it to "Keep the law in place and work to improve it." Combined, these two categories suggest that 56% of Americans want the essential features of the ACA retained. Twelve percent want the law repealed and replaced "with a Republican-sponsored alternative" and another 19% want Congress to "repeal the law and not replace it." 

The CBS/NYT poll asked respondents: "Which comes closest to your view about the 2010 health care law? The law is working well and should be kept in place as is. There are some good things in the law, but some changes are needed to make it work better. OR, The law has so much wrong with it that it needs to be repealed entirely." Again, a combined 56% of respondents fell into one of the two categories favorable to retention of the ACA, although the component percentages were slightly different (6% keep as is, 50% make changes to improve). Forty-two percent called for repeal.


Second, although the retention numbers seem pretty good for Obamacare, basic "favorable/unfavorable" numbers do not. They were 35% and 47%, respectively, in the February Kaiser poll. Opposition seems to have really exploded among the uninsured, who might be expected to embrace the law. As shown in this Kaiser trend diagram, favorable responses to the ACA tended to exceed or match unfavorable ones among the uninsured for nearly four years of monthly polling. In the most recent Kaiser poll, however, unfavorable dominates favorable 56% to 22%.

As acknowledged by Kaiser (and highlighted by HuffPost/Pollster), it may not be opinions shifting so much as the composition of the uninsured. Those who were uninsured before the October 1, 2013 launching of the online exchanges presumably fell into at least two groups: those who were embracing the opportunity to sign-up for free or reduced-cost health insurance (and ultimately did so); and those who, for ideological (or other) reasons, were opposed to participating in Obamacare (and who decided not to).

As roughly 10 million Americans have now signed up for coverage, philosophical opponents of the ACA will comprise an ever-growing share of the uninsured. (As I discussed in another context, the uninsured include some well-off people who choose to pay directly for their medical treatment.) Kaiser summarizes the situation thusly: "As more Americans gain coverage under the law, we can expect the group who remain uninsured to change over time, and some changes in opinion may be attributable to changes in who remains uninsured, rather than a shift in opinion among individuals."

Also, because the uninsured comprise a relatively small share of the U.S. population (around 16%, according to Gallup research), their numbers among Kaiser's overall 1501-person sample would likely be small, thus increasing the margin of error for the uninsured-specific findings. In fact, Kaiser provides a nice methodology report, which states that 137 uninsured individuals were present in the "unweighted" sample (i.e., before statistical adjustments for any under- or over-representation of demographic subgroups to match their percentages of the nation's population). Associated with this subsample size, the margin of error for analyses focusing on the uninsured is plus/minus 9 percentage points (much larger than the typical +/- 3 for samples of around 1,000).

Wednesday, February 12, 2014

Gallup Shows Decline in % of Uninsured Americans

HuffPost/Pollster provides a round-up of expert opinion on Gallup's latest finding of a reduction in the percent of Americans reporting themselves to lack health insurance.

Sunday, February 9, 2014

Hurt by Health Care Reform?

Some of the major polling outfits ask questions in their health care surveys on whether respondents feel the Affordable Care Act has helped or hurt them personally. Variations on this question might ask about the respondent and his/her family collectively, or even whether respondents feel the ACA has hurt Americans at large. There are some interesting things to notice in Gallup's most recent poll (which uses the "you and your family" wording).

As Greg Sargent discusses, of the 19% of Gallup respondents claiming they or their family have been harmed by Obamacare, "Sixty percent of those who say the law has hurt them or their families are Republicans or GOP-leaning independents. Only 23 percent are Dems or Dem-leaning independents, and another 15 percent are non-leaning independents."

Sargent adds that, "It’s always possible Republicans and conservatives are disproportionately impacted by the law, but it’s also possible that some who already dislike Obamacare are more likely to tell pollsters they’ve been negatively impacted by it, perhaps because they’re inclined to blame problems they’re experiencing with the health system in general on the law. Of course, this could work the other way around: Dems and Dem-leaners could be less inclined to say they are being hurt by it."

Another aspect of this issue that jumps out at me is how stable the trend lines for "help" and "hurt" appear to be in Gallup's historical polling on the matter. In February 2012, when relatively few Obamacare provisions had gone into effect, 16% said they or their family had been hurt by the law and 12% said they or their family had been helped. Unfortunately, Gallup went a long time without polling on this question, but its three most recent percentages for "hurt" have been 19, 19, and 19 (from polls taken November 23-24, 2013; January 3-4, 2014; and January 31-February 1, 2014). The respective percentages for "helped" were 9, 10, and 13 in these polls.

Considering the turmoil with the malfunctioning federal-exchange website from its debut on October 1 until the re-launch on December 1, and the controversy over many individual-market plans being cancelled (which was heavily in the news from late October into November), it seems remarkable that the percentages of respondents claiming harm to themselves or their families rose only modestly from the 16% in Feb. 2012 to 19% in late 2013 and early 2014.

Two hypotheses come to mind (although additional plausible ones could probably be generated). One is that a fairly sizable share (greater than the 3% aggregate increment) of Americans who had not felt harm in Feb. 2012 did so in late 2013/early 2014, but that others who were claiming harm in Feb. 2012 (or would have, had they been polled) were no longer feeling this way by late 2013, thus limiting the net increase in felt harm to the aforementioned 3%. The other hypothesis (a la Sargent) is that the help/harm survey questions are being answered heavily through an ideological lens, with little basis in how people really are tangibly affected. This latter explanation is cynical, to be sure, but would seem to explain the consistency of the results from early 2012 to late 2013 and beyond.