Yesterday, I explained that the ABC/Washington Post poll finding that support for a public health care plan " dropped sharply, to 37 percent" when “respondents were told that meant some insurers would go out of business” was deeply flawed, because the poll gave respondents an argument against a public plan, but not one for the public plan.

Well, there's another, related, flaw in the poll that I should have realized yesterday, but didn't. Take another look at the questions the poll asked:

21. Would you support or oppose having the government create a new health insurance plan to compete with private health insurance plans? (IF SUPPORT) Would you rather have that plan run by a government agency, or run by an independent organization with government funding and oversight?

21a. (IF SUPPORT) What if having the government create a new health insurance plan made many private health insurers go out of business because they could not compete? In that case would you support or oppose creating a government-run health insurance plan?

As I explained yesterday, question 21 doesn't actually give an argument for a public plan. It doesn't, as the Washington Post's write-up implies, compare a public plan to the “patient-friendly Medicare” program; it doesn't really include any reason why someone might support it. Question 21a, on the other hand, includes a reason why someone might oppose it: the possiblity that many private health insurers would go out of business.

That's a problem with the poll's wording, and it's probably enough of a problem that you should disregard the results. But there's another problem, one with the poll's structure, and it's definitely enough that you should disregard the results.

The ABC/Washington Post poll gave respondents two chances to oppose a public plan, and only one chance to support it. People who answered question 21 by indicating their opposition to a public plan were placed in a “No” pile and left alone. But people who indicated their support for a public plan were then told something bad about such a plan, and given another opportunity to oppose it.

The poll should have included a question 21b, in which those who opposed a public plan in 21 were told something good about a public plan -- say, that it could be cheaper and more effective than private insurance -- and given another chance to support it. That would have balanced things out.

But that isn't what ABC and the Washington Post did. They gave respondents an argument against a public plan, but not an argument for it. And they gave respondents two chances to oppose such a plan, and only one chance to support it. Then the Post hyped their finding of only 37 percent support for such a plan as the result of balanced questioning. It wasn't. The wording of the questions was slanted, and the structure of the poll was rigged. The 37 percent figure is meaningless, and should be disregarded.