The Day Before Super Tuesday: Looking At The Polls
March 5, 2012 - 4:30 am
In my day job, I deal with a lot of looking for patterns in things that cross my work PC (sitting right next to my laptop — working from home is great) every day. So, I thought that, with Super Tuesday coming up, we’d take a look at the polls from five states having primaries, to see if there’s any indications of a pattern emerging.
Before I begin, however, I’ve been a confirmed poll skeptic for quite a while, so I’m going to offer the following disclaimer:
Take all political polls with a large grain of salt, especially polls where you do not know the demographic breakdown of respondents. It’s common these days for polls to heavily oversample Democrats (sometimes by double-digits) and undersample Republicans. It’s also easy to insert bias based on how the questions are worded. Please do not assume that any given poll is unbiased, or that the sample matches reality, without digging into the details, and maybe not even after digging.
Okay, with that said, let’s look at the states. I’m using the Real Clear Politics (RCP) average where possible because averaging different polls is helpful in smoothing out some of the wild results that polls come up with from time to time. Averaging doesn’t really do much to overcome the bias and sampling errors I mentioned before, because if you average polls that all oversampled Democrats, the average is still going to be more Democrat-heavy than the real world is.
The RCP average has always had Gingrich in the lead in Georgia… sometimes by as many as 53 percentage points, but one poll showed Newt ahead by a mere 1 point. Right now the RCP average has Gingrich leading with 39.3%, Romney second with 23.8%, Santorum third at 20.5%, and Paul trailing with 6.0%. This gives Newt a 15.5 point edge (the difference between first and second), so I think it’s pretty safe to say that Gingrich is likely to win Georgia.
The Sooner State doesn’t have very many polls, almost not enough for RCP to create an average, but they managed it. One poll from last year had Gingrich up by 19, but now the average has Santorum leading with 41.0%, Romney second with 20.5%, Gingrich third with 20.0%, and Paul trailing the field with 7.5%. On paper, this gives Santorum a 20.5% advantage, but the most recent of these polls was taken two weeks before Super Tuesday (21 Feb), and a lot can happen in 2 weeks, especially in a primary season as volatile as this one. Bottom line, Santorum has a good shot to win Oklahoma, but it’s not in the bag yet.
Ohio has been back and forth and back again between October and today. All three front-runners have had double-digit leads at one point in that time frame, and right now the leader only has a 2.7% lead in the RCP average, so it’s definitely not in the bag yet for anyone. The RCP average as of right now has Santorum in the lead with 34.0%, Romney a close second with 31.3%, Gingrich a distant third with 15.7%, and Paul in last with 12.0%. With first and second separated by less than 3% (the margin of error in many polls is around 3%), this is an extremely difficult race to call, so I’m not going to. Ohio is a toss-up state on Tuesday.
These last two states didn’t have enough polls to form an RCP average, but we’ll do the best we can.
Yes, that Massachusetts, one of the deepest blue of the blue states, and current home state of Mitt Romney. Is it any wonder that all three of the Massachusetts polls that RCP lists have such strong leads for Romney that it’s almost not worth polling there? The latest of these polls has Romney way out in front with 64%, Santorum a very distant second with 16%, Paul and Gingrich bringing up the rear with 7 and 6% respectively. This looks like an easy Romney victory, but beware sampling errors and bias in polls with such a huge gap between first and second.
Another northeastern state, this should be pie for Romney, and the more recent of the two polls listed does have Romney ahead with 34%, Santorum second with 27%, Paul in third with 14%, and Gingrich in last with 10%. However, this poll was nearly 2 weeks ago (22 Feb), so things could have changed by now. I’m calling this one a tentative Romney victory.
So, the final results, 1 easy win for Gingrich, 1 easy win for Santorum, 1 easy win and 1 maybe win for Romney, and one too close to call. Ohio is definitely the state to watch Tuesday nite.