In yet another close election season, pundits have been grasping at information to predict the course of the race. As a result, polling has become a media obsession; a daily narrative about who is ahead and where. Unfortunately, however, these ever cited polls have flaws- flaws especially apparent in this historic election- that are badly misconstruing the current state of the race.
One of the most pressing problems with polls is that they are conducted almost entirely on landlines, excluding those who use primarily or only use cell phones. Two of the most Democratic voting groups in this election- minorities, and even more importantly the youth- fall strongly into this category. According to electoral projection website, fivethrityeight.com, “fully half of all adults under the age of 30 fall into the cellphone-only or cellphone-mostly buckets, and the number is growing every day. About a third of adults aged 30-44 are cellphone-only or cellphone-mostly” . Their exclusion from polling has a noticeable effect on the results, according to electoral analyst Nate Silver. “Including a cellphone sample improves Obama’s numbers by 2.8 points”, he writes, and with polls so close, it is a significant difference which could easily knock some swing states into Obama’s column.
Cell phone exclusion is but one problem with current polling procedure, however. Another large issue has to do with the category of ‘likely voters’ pollers use, which is currently weighted towards older, working class whites. This year in the democratic primaries, however, youth voting grew by 52.4%, black voting by 7.8%, and latino voting by 41.9% from 2004 levels. While the enthusiasm and probability of voting by these groups has increased dramatically, however, their weight as likely voters has not. As a result, they are constantly underrepresented in polling samples, despite all indicators showing there will be record turnout among these groups.
To compound these problems, pollsters and pundits have overlooked probably the biggest issue in the entire campaign: voter registration. One can gather an idea about its likely impact in 2008 by looking at the data in a few battleground states. In Ohio, over 700,000 new voters have been registered since December 1st 2007; 40% in counties heavily favored by Kerry in 2004 to only 27% favored by Bush. The results are familiar in Florida; nearly 290,000 new democrats have been registered since the start of the primary compared to only 90,000 republicans. Similarly, the Obama campaign registered 49,000 new voters in Virginia in August (on top of 36,500 in July and 28,000 in June), and is well on track to meeting its goal of 150,000 new voters by October 5th . 40% of these new registrations are under the age of 25; suggesting they will overwhelmingly vote Democratic.
Let’s do a hypothetical example of what all this could mean. Take Virginia for instance. Bush beat Kerry in Virginia by 262,217 votes. Current polling between Obama and McCain gives Obama ~70% of youth, ~94% of blacks, and ~65% of latinos. Based on recent growth in turn out, one can also reasonably assume the youth will turn out an extra 40%, blacks an extra 20%, and latinos an extra 30% in the general election. If we add these changes to Bush and Kerry’s totals in 2004, Obama has a 54,200 vote win, and that’s still excluding the likely 150,000 more voters his ground operators will have added to the registration list, as well as the increased dissatisfaction nationally with the Republican Party.
The media and the pollsters have largely ignored this increase in the Democratic base and the increased enthusiasm of the youth and minorities in this election. The exclusion of cellphone users, the underexagerration of youth and minority turnout, and the disproportionate increase in Democratic to Republican voter registration is making the perceived race and the real race two very different things. While Democrats should be more optimistic, however, it doesn’t mean they should get complacent. In 2004 the Democrats out-registered Republicans in Florida by 60,000 votes, but had lagging turnout rates compared to the Republicans and lost the state. Democrats need to keep up on their ground game and make sure they can follow up their new registrations with high turnout. After all, in the end it’s not the spin or the polls or even the enthusiasm that counts. It’s the votes.
Home 