Democrats, their allies in the liberal mainstream media,and party activists are abuzz over the results of a CBS News/New York Times Survey that appears to show massive public disapproval of congressional Republicans’ insistence during the debt-ceiling debate that any final agreement actually cut federal spending and not raise taxes.
On the surface, the numbers from the survey do indeed paint a grim picture for the GOPers led by House Speaker John Boehner and Senate Minority Leader Mitch McConnell. Here’s how CBS News’ Political Hotsheet described the results:
“The survey, taken on August 2nd and 3rd – immediately after the deal was reached – found Americans more frustrated with congressional Republicans than their Democratic counterparts when it came to the negotiations.
“That’s not to say congressional Democrats have much to crow about. Sixty-six percent of Americans disapprove of their handling of the debt ceiling debate; just 28 percent approve.
“But Republicans fare worse: Seventy-two percent of Americans disapprove of their performance during the debt ceiling debate, while just 21 percent approve.
“And Republicans get most of the blame for the standoff. Forty-seven percent blame Republicans in Congress, while 29 percent blame President Obama and congressional Democrats; 20 percent say both are to blame.”
But as always seems to be the case with this survey, a close look at the questions asked (and apparently not asked) and the survey sample reveals serious structural flaws that render it useless as a predictor of electoral behavior.
First, the survey is of 960 adults, not of likely voters. That fact alone requires a major footnote because surveys of likely voters are always more reliable predictors of electoral outcomes than those of adults, or registered voters.
Second, the single most important characteristic that determines the credibility of any survey is the randomness of its sample selection process. This particular survey is based on re-interviews with people who previously participated in two separate, earlier surveys, one by the Times and another by CBS. Thus, the people questioned in this survey were not selected randomly.
Third, there is a paucity of vital information about the sample demographics. All CBS News/New York Times discloses is that 44 percent of its sample were Independents, while 32 percent were Democrats and 24 percent were Republicans.
Based on the Gallup Poll’s long-running measure of party identification, these proportions are about right for Democrats, but there is a significant under-representation of Republicans.
But there is an even more troubling problem here than party identification. For some reason, CBS doesn’t provide a break-out of the ideological leanings of the Independents in the survey. Since Independents are by far the biggest group within the sample, it is especially vital that we understand more about these respondents.
Gallup tells us that conservatives (40 percent) outnumber liberals (21 percent) and moderates (35 percent) by healthy margins, but we have no way of knowing from the data provided by CBS News/New York Times how many of these respondents leaned left, how many leaned right and how many resisted any sort of ideological orientation.
Put these factors together and what we end up with is a useless predictor of electoral behavior of registered voters, but a very useful propaganda tool for Democrats and their friends to bash Republicans.
