As midterm elections draw closer, voters will inevitably be exposed to more and more polling data. The data will be cited by various publications as an indication of this or that. It will be brandished about by one party or another, as proof positive that everything is going according to plan. But, as many people already know, all polls and all polling firms do not carry an equal amount of weight. How can you tell which polls are which? Here are some useful facts to help you sift through the propaganda.
The difference between polling firms and PR firms is important.
Once upon a time in America there were just a few polling companies. Their job was to determine public opinion and report on the results. That is not the case in 2014.
Today there are hundreds of organizations and corporations that are involved in polling. Some of the newer companies are market research firms, while others are clearly out to sway public opinion. While it may seem that a polling firm has the singular task of predicting the outcomes of an election, many of the firms that publish data today are PR firms. The job of those firms is not to accurately assess public opinion, but to actively sway it.
Check out this company for example. Wakefield Research is a public relations firm. They specialize in “crafting custom PR surveys that return top-tier media coverage.”
Looking a little further, Wakefield promises to write the polling questionnaire for their customers.
“By relying on our knowledge of public opinion, we write the questions most likely to return newsworthy results. This means that you don’t waste omnibus survey questions on concepts unlikely to return useful data. We combine this with our knowledge of the media landscape to make sure that your data isn’t just interesting, but also newsworthy.”
As one customer so aptly put it; “Wakefield knows that the quality and relevance of the data you get out of a study is only as good as the questions asked to elicit responses.”
So when you get a call or are asked to fill out a questionnaire and the questions are very one sided, you are dealing with a firm that is interested in PR, rather than in determining public opinion.
The problem is that the media often cites PR polling firms. A quick way to educate yourself is to visit this link and click any one of the ‘media’ sources at the bottom. Doing this will give you an idea of just how often PR polls are cited by major media outlets across the country.
How to approach polling firms.
There are three important things to do when you are presented with polling data. If the polling firm cited is not one that you’ve heard of before, look the company up online. Find their ‘about us’ page, and read the description. You might be surprised at just how many polling firms, cited by mainstream media sources, turn out to be PR firms.
Next, look at the methodology used to get the polling results. Who was surveyed? How was the survey conducted? How were respondents selected? What is the margin of error?
Finally, look at the polling questions. Do they seem balanced or are they slanted to get a certain result? Keep in mind that polling questions which are created by PR firms are often designed with the intention of ‘eliciting a certain response’.
How to weigh polling data from non PR sources.
The 2012 elections brought us the ‘unskewed polls.’ The claim was that the ‘liberal media’ and pollsters were oversampling democratic voters, in order to ‘skew the polls’ in favor of President Obama. The antidote, of course, was to skew the polls in favor of republicans, by, in some cases, as much as a 12 point margin, all in the name of ‘unskewing.’
The reality is that of the 23 major polling firms in the United States, 19 showed a Republican bias. These were firms that did not use the ‘unskewed’ polls method, but relied on their own traditional polling methods.
Gallup, one of the most well known polling firms in the US, had the worst accuracy level out of 23 polling firms ranked by Nate Silver. Silver’s list contains all firms that conducted at least 5 polls, leading up to the 2012 election. Gallup had a 7.2 point margin of error, with a republican bias that also measured 7.2 points.
Also at the bottom of the list were Mason-Dixon, American Research Group and Rasmussen polling firms. These firms demonstrated a Republican bias of between 4.2 and 5.4 points.
On the other end of the spectrum, the most accurate polling firm during the 2012 election season was IBD/TIPP. The firm had the lowest margin of error, at 0.9 points. It also had a very small Republican bias, at 0.1 percent.
Among the four polling firms that showed a Democratic bias during the 2012 elections was Pharos Research. Among those four firm, Pharos had the worst margin of error at 4 points. The firm showed a bias toward Democrats of 2.5 points.
The other three firms that favored Democrats included We Ask America, which had a very small bias of 0.1 points. Quinnipiac, which had a Democratic bias of 0.3 points, and Rand Corporation, which had the second-highest left leaning bias of 1.5 points. All of the other polling firms were Republican biased.
Even PPP, a firm which is often accused of being a ‘left leaning’ polling firm, had a republican bias of 1.6 points and a margin of error of 2.7 points.
Nate Silver published a simple chart, which ranked the accuracy of the various polling firms, following the 2012 elections. You can access the full chart here.
Where to find accurate polling information.
Nate Silver now publishes polling information at fivethirtyeight.com. This site will most likely prove to be one of the best places on the web to get accurate information leading up to the 2014 midterm elections.
If the polls don’t look promising at the moment, don’t blame Silver or his analysis of the polling data. The correct response from Democrats is not to come up with a leftist version of ‘unskewed polls.’ Instead, it’s to get involved with grass roots efforts to educate voters and get them to the polls.
Democrats are notorious for staying home during midterms, a fault that could have devastating consequences in 2015. This is not the year to stay home. Just imagine the historic consequences of a Republican-controlled house and senate.
There’s no doubt that Republicans would move to impeach the president, if they thought they could get away with that kind of illegal action. Without strong Democratic leadership in both houses, it’s entirely plausible that they could get away with it. Right wing bloggers are already giddy at the thought of it.
Republicans are counting on the left to follow the usual pattern of turning out in presidential election years, only to stop paying attention by midterms, two years later. There’s too much at stake not to have a historic turnout in 2014. If Democrats get themselves to the polls, on the other hand, we will hold the senate and there’s a possibility that we could also take back the House, something that has to happen, if we are to see the country move forward.
So pay attention to the polls, but know which polls to watch. In the meantime, get involved, get registered and get out the vote.