It is election time. And, like most election seasons, we get online and look at polls. Those that show our candidate leading, we cheer and say are legitimate. On the other hand, we look at polls that state the opposite, and claim that the polling firm is sympathetic to the opposition.
In a political era in which we do rely on polling data (mostly for our own personal amusement), we have changed from having a few, reliable polling firms to everyone and their brother conducting polls. With all of these polling firms out there, it is hard to get a true and accurate account of what is happening in the election.
Lets look at recent polling conducted in Florida in the 22nd Congressional District. In one poll conducted by Democratic polling firm Anzalone Liszt Research , Ron Klein, the current Democratic congressman, is leading Alan West, mostly of “I have a higher security clearance than President Obama” fame, 48% to 40%. In another poll conducted just a week later, this time by Republican polling firm ccAdvertising, West leads Klein 43% to 37%.
With both of these polling firms conducting big margins of victory for their respective party candidates, nobody really has any idea where the district really lies. Actually we do, and Klein should win the district. Still, even so-called respectable polling firms, like Rasmussen Reports, tends to favor Republican candidates by larger-than-expected margin.
Even with the differences in these polls, nobody can really predict what will happen come November. And while these polling firms are supporting their candidates, one also has to ask if the respondents to the polls are either truthfully answering the questions or misunderstanding the questions in general.
Lets look at the current generic polls taking place for the balance of power in the House of Representatives. Most of the time, if not all of the time, people are asked who they will support in the general election. With their generic answer being Democrat or Republican, the firm gets all their data together and publishes it to the public.
But what if the respondents didn’t understand this.
I worked for a very short time at a polling firm years ago in which we asked this question. We asked specifically about the House and the Senate and which one the respondent would support. But many times, the respondents thought they were answering “who do you think will win” in the general election. Could this error still be prevalent in our polling data today? Asking people to answer polls truthfully (either through deception or through no fault of their own) can never be checked.
In addition to not understanding questions, nearly all respondents say that they are going to vote. Even though the respondents to a poll are not public record, most of them want to think that they “will be voting this November.” And if you look at the numbers of likely voters in the polls compared to actual voters during the election, the results can be quite drastic, especially in mid-term elections.
Also, another point to keep in mind is that media outlets, which is where we get most of our polls, usually conduct the cheapest polls possible with the largest margin of error so that they have something that their readers can chew on. Rarely, if ever, do media outlets spend big bucks to publish an accurate poll. That is mostly for the campaigns.
Therefore, we see all the polls. We see that the Democrats are going to get killed this November. And while I feel the Democrats will retain both the House and Senate, I feel than any polling data in regards to this current election is, well, crap!