A Citizen’s Guide to Early Primary Polls

Recently, a relative asked me about her favorite presidential candidate, concerned he wasn’t doing well in the polls. In the case of this candidate, I explained to her that, from all accounts, he was doing well enough on all points such as fundraising, campaigning, and building his organization, that there was nothing to worry about.

I doubt she’s alone in looking at early polls with a bit of concern and confusion. The media coverage of polling in presidential primaries is extensive, but often not very responsible and lacks context. The media likes polls because it gives their horserace coverage some substance, some idea of who is up and who is down. Yet, polls at this point in the race are volatile and mostly irrelevant, and they’ll remain so for some time.

National primary polls are probably the most useless, mainly because we don’t hold a nationwide primary vote. While voters in Iowa, New Hampshire, and South Carolina are comfortable with the fact they never know when they’ll run into a presidential candidate, for most of the rest of the country, the Presidential election is fifteen months away and not something that most have focused in on. Using these polls for anything is absurd.

You can’t have a debate with sixteen or seventeen people on the stage, but it is absurd for CNN to use averages of national polls to determine which candidate will end up in the top ten debate and which will end up in debating in the bottom six. You might as well do fifty “Man on the Street” interviews showing people pictures of the candidates and asking them which candidates looked like they could be president as a basis for forming the top ten debate. It would be better to have eight candidates in each debate with participants determined by random drawing.

The one useful thing in national polling is candidate favorability. If a candidate with high favorable numbers wins an early primary state, they may have an easier time parlaying that into greater successes in later primaries, while one with lower favorable ratings could inspire people to unite behind a single candidate to stop them. A candidate who is unknown could win an early state only for it to be very easy for the media and his or her opponents to define them or, at the very least, cast doubt upon them through the news reports and Super PACs.

Because of the nature of the Iowa caucus, it’s hard to determine who will go to a Caucus. In the last two Iowa GOP caucuses, the polling has been off by quite a bit. Real Clear Politics reports extensively on polls and also will average out poll results to better control for all the variances. It’s often very accurate in general elections and in later primaries. However, when it comes to the Iowa Caucus, the final polls were way off. In 2008, the average poll had Huckabee winning by 3%. He ended up winning by 9.2%. In 2012, the final Iowa Caucus poll average had Mitt Romney defeating Rick Santorum by 6.5%. In reality, Santorum won by 0.1%.

The polls didn’t predict the right order of candidate finish. All but one poll had Ron Paul in second place, except for the last Public Policy Polling poll which had Ron Paul in first. Rick Santorum won the Iowa Caucuses in 2012 without finishing any higher than third in any public poll taken during the 2012 campaign.

In New Hampshire, the situation would appear more stable on the Republican side-particularly at the end of the primary season. In 2008, the RCP polling average was a McCain victory by 3.6% and he won by 5.5%. In 2012, the RCP average was a 20% Romney victory and the result was 16.4% which is a little bit off, but not horrible.

However, this only tells part of the story. In 2012, New Hampshire was always going to be Mitt Romney’s for the taking, leading to little interest and therefore little volatility. In 2008, while the GOP race played out closely to what the polls had indicated, the Democratic race was another tale. Obama had led all of the final polls in New Hampshire by an average of 8.3% only for Hillary Clinton to win the state by 2.6%. Similarly, John McCain’s eighteen percent margin of victory in the 2000 New Hampshire primary was way beyond anything that the polls predicted while former Senator Bill Bradley’s campaign against Vice-President Al Gore fell short.

New Hampshire is hard to poll when both parties are running primaries because Independent voters vote in large numbers and which primary they’re going to vote in does vary and pollsters have to create polling models that predict exactly what percentage of a party’s primary electorate will be Independents.

There are two ways in which Iowa and New Hampshire polling can be useful. First, is that over time, you can notice the trendline of a given candidate over a lot of polls. Take 2008, when polls showed Huckabee moving from single digits to consistently in the mid-teens in October of 2007 to finally overtaking Romney in the polls in mid-late November. Also, when you get to the end of the process, while it may not give you an idea of who’ll win, it can give you a good idea of who’s going to lose. When the Iowa Caucus results were announced I would have been flabbergasted in 2012 if Rick Perry, Michele Bachmann, or Jon Huntsman actually won as they were the bottom three and well back of the top spot in the polls.

Probably the most important news in polling recently has been Ohio Governor John Kasich finishing in double digits in the last two polls in New Hampshire. This indicates that despite a relatively late announcement, Kasich does have a definite opportunity in New Hampshire, a state that has rewarded mavericky candidates with stances that don’t appeal to many conservatives. It’s not hard to imagine Kasich winning New Hampshire.

A win in New Hampshire could make Kasich an establishment alternative to Bush. Or if Bush wins Iowa, he could take a page from his brother’s playbook, vilify Kasich with conservatives, and try to unite conservatives around himself to stop Kasich. It’s certainly not how Bush would choose to win, but he’s certainly not going to choose to lose.

For the average voter who wants to figure out what’s going on, it’s better to look at how a candidate’s presenting themselves, the strength of their plans for the country, and how good a President they would make. Of course, that’s all very subjective.

The appeal of looking at polls to determine the state of the campaign is that, in theory, it gives you an objective view of how effective each campaign has been performing. However, while figures don’t lie, history tells us that sometimes they don’t know what they’re talking about.