This political season will no doubt be recorded as one of the most historic in American history. The United States elected its first African-American president, Barack Obama, who defeated Republican nominee John McCain.
If McCain would have defeated Obama, the year of political firsts would still have been in play, as the Arizona Senator would have been the oldest first-term president and his running mate, Alaska Governor Sarah Palin, would have been the first female vice president.
With the election now in the rearview mirror, pundits and the public will break down every facet of the 2008 race, including the political polling and survey landscape. Were Americans truthful about their answers prior to the election? What did the numbers mean? Were they right on target or way off base?
Political polling has been a lightening rod for controversy ever since the first one was taken in 1824 by a Pennsylvania newspaper. The Harrisburg Pennsylvanian is widely acknowledged to have conducted and then to have published a presidential preference poll. Oddly enough, the poll was conducted not in Pennsylvania but in Delaware.
The newspaper surveyed Wilmington citizens during July 1824, asking about their presidential favorites. The poll results referred to the open seat election after the retirement of James Monroe. The straw poll revealed Andrew Jackson with a commanding 70 percent lead (335 votes) over Secretary of State John Quincy Adams (23 percent), Speaker of the House Henry Clay (four percent) and William Crawford (two percent).
The newspaper’s forecast of Jackson was accurate, as the military hero did indeed win the popular vote, beating his closest opponent, Adams, by 77,000 votes. But as Al Gore can attest, being most popular doesn’t mean you get to be president. None of the candidates won a majority of electoral votes; hence the election was thrown into the House of Representatives – where Jackson eventually lost to Adams.
Today, polling has become both art and science, and highly accurate in the majority of cases. A study by Fordham University’s political science department shows that the polls in the recent national presidential election were pretty close in predicting the outcomes. The average of 23 polls predicted an Obama win of 7.52%. The final outcome was an Obama win of a little over 6.1%. Overall, it was a good night for the pollsters.
But political pollsters do suffer setbacks and a blow to their credibility came during the New Hampshire democratic primary election, which was predicted to end in Obama’s favor by an 8 percent margin, but went to Hillary Clinton, who emerged from the primary with a small victory margin.
“I think that illustrated more clearly than anything in recent history that political polls are a fickle beast,” Sherman Willow of Chicago-based Pollsters Club said in a recent interview.
The challenges faced by pollsters have changed as technologies and attitudes have changed.
For example, telephone polling, once the standard-bearer, has been tripped up by caller ID and the cell phone. The first major assumption political polling organizations used to make is that everyone uses a landline phone. Many younger Americans, however, are moving to only having cellular phones.
Despite its critics, Internet polling is gaining plenty of steam as a viable alternative, and is well on its way to becoming as accurate as telephone polling. Internet polling tends to feature a pool of respondents that are more likely than the average person to be passionate about their candidate or issue, and, in essence, more likely to vote. This, of course, means that results from internet polling are more likely to reflect actual election outcomes.
It’s well-known that the wording of the questions and the order in which they are asked can have a profound impact on a poll. This can also, however, be the result of legitimately conflicted feelings or evolving attitudes, rather than a poorly worded survey.
When faced with intangibles that may alter results, pollsters control these issues through several means, including split-sampling, which is having two different versions of the questions. They also control effects by asking enough questions or analyzing the results with techniques which synthesize the answers into a few reliable scores and detect ineffective questions.
“Political polling can be very tricky, especially when it comes to phrasing the questions, “said Dr. Jan West, founder and CEO of the Dallas-based National Business Research Institute, a political polling and survey company. “It’s critical to accurately understand the attitudes, opinions, and beliefs driving human behavior. NBRI political research surveys produce the actionable information needed to guide policy and make informed decisions. Staying informed of the political climate is imperative whether it’s for a candidate, a cause, or a business.”