The Los Angeles Times reports today that their polling demonstrates that more Americans agree with the Democrats than the Republicans on how to reform health care. A slender majority want government mandates for people to carry health insurance, and a larger majority supported a mandate for employers to offer it. However, the polling sample has much more to do with the results than the Times acknowledges:
Two of the main proposals advanced by Democrats received majority support in the poll.
Sixty-two percent said they supported requiring large employers to help pay for coverage whereas 31% opposed it. And 51% said they favored a mandate that individuals purchase health insurance, much as drivers are required to carry auto coverage; 39% disagreed.
Tax breaks to make insurance more affordable — a leading Republican idea — more closely divided the public, with 44% backing that approach and 45% opposing it.
In one of the most politically significant results, the poll finds that independents and moderates were generally lining up with Democrats in the healthcare debate.
Just as with the CBS polls, the LAT/Bloomberg poll delivers results that favor Democrats — because the pollsters oversampled them. A sample of 1039 registered voters contained a split of party identity of 46% Democrats, 36% Republicans, and 18% independents. Newsweek’s last survey of party identification puts the split at 34%/30%/36%, which gives a much different perspective on the questions.
Yesterday, I warned about this at Heading Right and predicted that policy questions would give necessarily skewed results. It isn’t hard to imagine that a poll with 46% Democrats would come up with 51% supporting the Democratic Party’s policies. It might be a little more surprising that Republicans could pick up eight points for its own approach to health care with so few independents represented. In fact, the 8-to-5 point pickup might indicate that the Republican policy has more appeal to independents.
The intraparty results certainly have some value, but any questions regarding the overall policy preferences in such a poor sample are simply unreliable. I also note that the PDF of the polling done by the Times does not include the questions for the health care analysis which they report today. The questions are critical to know how the pollsters framed the issue and whether they pushed for a certain response. The lack of this data in their methodology makes the results even more suspect.