The Problem with Polls
The problem with polls (or maybe it’s the reporting of poll results) is that they ask (focus on) too many closed-loop questions. For instance: “Do you approve of the way the President is handling health care reform?”
The answers to such questions lend themselves to neat and tidy conclusions — perfect blog fodder — such as Chris Cillizza’s suggestion today that the President is losing the support of Independents on health care reform.
However, what such data — and/or its parsing — does not explain is “why” Independents are abandoning the President on this issue. Is it the projected cost of reform? Is it the threatened taxes to pay for reform? Or is it something else all together, e.g., Independents don’t think the President is doing enough to shape the debate; that he’s deferring too much to Congress on policy details?
It’s an important question because the “why” on this issue — and many others — can make a critical difference in how policymakers respond to the “what.”
Case in point: If most Independents are dissatisfied with the level of Obama’s involvement in shaping policy, his team’s response would likely be to get him more involved (as they appear to be doing). In contrast, if most Independents are dissatisfied over the projected costs, then Team Obama’s response might focus not on the details of policy, but on all-out efforts to lower the expense of said policy.
Granted, the end-game might be the same, but the paths for getting there could be remarkably divergent.
8 Responses to “The Problem with Polls”
Just review good sites like Pew (below) or Polling Report (to look at polls _and_the_questions_) and put it all into context, or proper place. Note that polls shouldn't be attacked and derided (or Questioned [tm]) simply because they are revealing what is obvious to anyone observing the public as well as Obama and the Congressional Dems. The more reckless over-spending and debt, the more stupid rushing to have the federal government take over and interfere in more and more, the more even the lesser-intellectually-endowed Dem-Wit [tm] voters will second-guess their otherwise-still-unflagging support for the Dems. (The Bush people and the Republicans have become so insignificant they no longer are handy objects for diversion, Red herrings.)
Notably:
http://people-press.org/report/517/political-va…
What is so strange is that this problem with the polls is such a recent phenomenon.
It seems like just one administration ago that polls were so incredibly accurate that they could show the combined judgment of the entire populous on a daily basis. Even with incredibly leading and bias questions, the percentage of the country opposing anything Bush did was known within minutes of the occurrence.
Who are these foolish people who respond to pollster questions with answers that don’t reflect the good Obama is trying to do for the country? Why do the organizations doing the polling feel the need to report the numbers without adjusting them to reflect what a more enlightened public would say? Is it possible that with the use of sampling we can adjust the population numbers to better reflect correct thinking?
I’m certain these and other questions are being discussed right now between the Obama administration and the news media. Of course I trust whatever they decide to tell us will be best for the country.
J. West. What you describe are also confusing things going back to the 2000 elections, as well as with, say, the Census and hiring ACORN to Properly Adjust [tm] census figures (using creative “sampling” methods in place of Dem officials' openly doing this). Obviously the media, who know better than we do (as Pew noted, journalists had a much higher view of the public's judgment when Clinton was President than when Bush was President) along with their fellow Dems in Washington, can save us all time and trouble. Why bother to continue with elections when these fine folks already know what's best for us all?
I think you guys are missing the point that Pete is making, namely that the poll questions are too closed-ended to get a handle on the underlying rationale for the answers. A follow-up question would have been helpful, ie if you said you do not think Obama is doing a good job what do you think is the biggest problem with the plan.
Really though the same phenomenon was evident in the polls that merely asked “would you support nationalized health insurance” without mentioning the potential costs or changes that might be necessary (a few polls did try to get into more depth). Those polls were trumpeted as a reason to proceed with the public health option.
Polls seem to have too much power for as imperfect as they seem to be.
DaGoat is right about the point Pete is making.
Also, while the headline may be that Obama is losing support, he's still way ahead of the Republicans.
DaGoat,
I don’t seem to remember Pete’s articles about how the polls needed follow-up questions during the Bush administration.
If the pollsters gave a misleading, biased and obviously leading question that produced the correct answer for the liberals, no follow-up was ever necessary.
I don’t seem to remember Pete’s articles about how the polls needed follow-up questions during the Bush administration.
jwest I didn't get the impression Pete was making a partisan issue out of this so I elected not to as well.