The Washington Post was very excited to report on Feb. 5 that President Obama has finally achieved “the edge” over Mitt Romney in a “general election matchup” poll. The Post was pleased to note Obama was “boosted by improved public confidence” and that he now led Romney by over 50%. Well, he does if you don’t poll actual voters, anyway and therein lies the major problem with the Post’s polling.
The flaw in the Post’s poll is that they seem to have polled “adults” instead of “likely voters” and this fact calls into question the claim in the headline that “Obama holds edge over Romney in general election matchup.” You see, you have to be an actual voter before your opinion in an “election matchup” much matters but the Post apparently did not make sure that its respondents were actual voters before declaring that Obama is now winning over more voters.
But the bigger problem is the fact that the Post has decided it no longer needs to include the partisan breakdown of its respondents for readers to assess. The Post did not include the percentages of Democrats, Republicans, and Independents in its polling data so there is no way to know if the poll included a fair representation of all parties or if the whole poll was weighted heavy with Democrats.
The Post has had troubling polls before. Ed Morrissey notes for instance that a WaPo poll from April of 2011 had 22% Republicans overpowered by 33% Democrats and 38% purported independents. If the Post is shorting Republican representation, no wonder the Obamessiah seems to be surging!
By excluding in reports its partisan breakdown, the Post risks having its results easily dismissed by serious readers. It makes the poll practically worthless. Of course, the problem is that the average reader won’t realize that things are askew with the polling and will accept the claims of Obama’s popularity at face value. But maybe that’s why the Post won’t include its partisan breakdown in its reports? As Morrissey says, “it’s easy to assume that the reason that the Post has ended its sample transparency is because they have something to hide.”
And with quotes from the Post story like, “Overall, 55 percent of those who are closely following the campaign say they disapprove of what the GOP candidates have been saying,” one has to wonder if those respondents scoffing at the Republican message were actual voters that the GOP should pay attention to, or partisan Democrats whom they won’t be able to reach anyway, or even disinterested “adults” that aren’t voting in the first place? Unfortunately, with this poll we have no way to assess the answers to those questions.
Still, the Post assures us that, “Meanwhile, the president’s recent remarks are better reviewed.” How do we know? Well, we don’t. We just have to take the Post’s word for it if we are going to believe it.
Essentially, what we have with these Washington Post polls is simple cheerleading for the president instead of legitimate analysis of the current sentiments of voters.