Polling in 2012

MEMORANDUM

TO: INTERESTED PARTIES
RE: POLLING AND THE 2012 ELECTION
DATE: NOVEMBER 12, 2012

There are a number of challenges facing the survey research industry, and every election offers the opportunity to continue to learn, change, and adapt. As a leader in the industry, we take every election, whether there are good election outcomes like 2010 or bad election outcomes like 2012, as a learning opportunity to do our job better.

There is no question that the 2012 election was challenging for pollsters. The number of assumptions that must be made has increased significantly: including enough cell phone interviews, ensuring that younger voters are represented in a sample like they are in the electorate, projecting the white and minority percents of the electorate, and having the right partisan mix are all tough calls.

Recognizing that those challenges have been upon the industry for a number of years, Public Opinion Strategies has worked hard to stay at the forefront of innovation, including cellphone interviewing, quotas by age to make sure that voters age 55+ are not over-represented in samples, Spanish language interviewing, and ensuring the right blend of minority and white voters.

We conducted over 1,000 surveys in October, and approximately twenty or so were made public. Campaigns own the data and have every right to release data at a high point in the campaign, but given that late deciders voted Democratic on the generic ballot by double digits, (41% GOP/56% Democrat on the generic ballot), polls released prior to the election do not necessarily capture the final movement or turnout trends in a campaign. This is especially true as millions of dollars were still being spent on those campaigns even as some polls were released with two-three weeks left in the race.

As a part of the Republican polling community, our prescription includes doing at least one-third of the interviews with cell phone respondents going forward, adjusting as required, ensuring that we include enough younger voters in our sampling, and (in many cases) polling until the final weekend of the campaign. This is going to cost campaigns and organizations more money on polling, but it is necessary to have a more accurate representation of the electorate.

The Obama campaign does deserve credit for helping drive younger, African American, and Latino turnout in a way that replicated 2008 turnout. The data is clear – very few voters split their ticket from the Presidential to Statewide to Congressional levels, and even down to the state legislative levels. This straight ticket voting had a significant impact at the congressional level. Not many candidates ran significantly ahead of Mitt Romney, underscoring the straight ticket voting that characterized this election year. Those candidates who ran far behind Romney did so because they were, in some instances, flawed candidates, or they were running against well-regarded Democrats.

Across multiple campaigns with which we worked, our research did what it is designed to do – provide strategic counsel to campaigns about key target groups and messages designed to help them win. That’s how we were able to help the Republican party gain the one Senate seat it did pick up (Senator-elect Deb Fischer in Nebraska), and poll for two of the four GOP Congressional challenger winners, as well as ten new Members of Congress overall.

In the new normal – Presidential year turnout of significantly more Democrats than Republicans, as well as more minorities and more younger voters than ever before – Public Opinion Strategies is dedicated to facing head-on the difficult challenges the industry faces in reaching and accurately measuring our changing nation.