Polls That Matter And Polls That Don’t
This article about Democratic poll numbers is producing a lot of buzz:
“Republicans are in jeopardy of losing their grip on Congress in November. With less than four months to the midterm elections, the latest Associated Press-Ipsos poll found that Americans by an almost 3-to-1 margin hold the GOP-controlled Congress in low regard and profess a desire to see Democrats wrest control after a dozen years of Republican rule.
Further complicating the GOP outlook to turn things around is a solid percentage of liberals, moderates and even conservatives who say they’ll vote Democratic. The party out of power also holds the edge among persuadable voters, a prospect that doesn’t bode well for the Republicans.
…The AP-Ipsos survey asked 789 registered voters if the election for the House were held today, would they vote for the Democratic or Republican candidate in their district. Democrats were favored 51 percent to 40 percent.
Not surprisingly, 81 percent of self-described liberals said they would vote for the Democrat. Among moderates, though, 56 percent backed a Democrat in their district and almost a quarter of conservatives – 24 percent – said they will vote Democratic.
Democrats also held the advantage among persuadable voters – those who are undecided or wouldn’t say whom they prefer. A total of 51 percent said they were leaning Democrat, while 41 percent were leaning Republican.”
Here’s the thing: None of this means very much. People don’t vote Democratic and Republican slates of candidates. They don’t choose between generic Republicans and generic Democrats. They choose between real people with flaws, quirks, and different positions on key issues that matter to the voters.
With no names mentioned, things do not look good for the Republican incumbents in these three districts.
So, if you just looked at the generic poll number for these three districts, you’d think that those incumbent Republicans congressman are in a lot of trouble. They may well be. But, look at how the numbers switch when the respondents are given the names of the candidates.
Linda Slender-D 41
Mike Ferguson-R 50
Jo Sestak-D 41
Curt Weldon-R 51
John Cranley-D 43
Steve Shabot-R 52
So, if these polls don’t matter, what does?
Approval ratings and multiple head to head match-ups of the candidates in a race based on registered, or even better, likely voters, not adults.
A single poll can often be wrong. The demographics used to compile it can be slanted and the margin of error can distort what’s happening as well. However, if you can take a look at the results of several head to head polls, in the period before the election, and see who’s ahead in a race and which way the numbers are trending, you can usually get a pretty good idea of what’s going on. But, generic data? Of course, you’d always rather have your guys running ahead rather than running behind, but whatever the case may be, you can look at those numbers all day long and it won’t give you an accurate picture of what’s going to happen at election time.