The Problem With Iowa Polling

As mentioned in a previous post, Iowa pre-caucus polls are notoriously unpredictive, as not every voter's preference will drive him or her to action. For instance, someone who supports Romney and plans to vote for him might find himself quickly uninterested in voting if it were to rain on Thursday night. Ron Paul supporters, however, are expected to vote even if a hailstorm were to overtake the entire state of Iowa.

Another issue may play a part. Recent polls have Gingrich plummeting. Politico reported a poll which puts Gingrich at 14%, after achieving a high of 32% just two weeks ago. However, most polls seem to have Paul, Romney, and Gingirch in similar polling territory.

One problem I have with this sudden precipitious drop for Gingrich numbers are the way polls are taken. If a polling service polls 1000 households, but only 700 households answer the phone, the pollster will keep calling until he reaches 1000 actual people, which may take over 1500 actual calls. This, however, is an imperfect way of obtaining true results. Perhaps voters who don't answer the phone or simply aren't home are more likely to endorse a certain candidate, say, Newt Gingrich?

A significant, reliable stable of Gingrich supporters are said to be of the elderly crowd. It's no secret that elderly people have a spotty history at answering the phone. My grandmother is not home from 8 am until 5 pm every day, and she doesn't even have a job. What's she doing all day? Elderly things. Going to banks, supermarkets, haggling over prices in different stores, and spending time in the library. If a pollster were to call her, said pollster has about a 45 minute window during the actual day to catch my grandmother and record her probable Gingrich support.

Your question now might be, "Well, if pollsters are missing all of the Gingrich support, how do you explain his current drop to 14% from 32%? Why didn't they miss it when he was at 32%?"

The answer is simple- polls are much more frequent now, lending the way toward higher margins of error. There are still a number of polls which have Gingrich in the mid 20's percentage. While Newt may have dipped a bit, the precipitous nature in which the media has portrayed it seems to be a bit overdone.