Editor's note: This commentary is part of a collaboration between the University of Arkansas at Fort Smith and The City Wire to deliver an ongoing series of political-based essays and reports. Dr. Williams Yamkam is an assistant professor of Political Science at the University of Arkansas at Fort Smith where he teaches multiple political science courses including a course on campaigns and elections. Besides the various professional trainings that he has received in campaign operations, he is a graduate of American University’s Campaign Management Institute in Washington, D.C.
Opinions, commentary and other essays posted in this space are wholly the view of the author(s). They may not represent the opinion of the owners of The City Wire.
When conceiving this column a few months ago, we planned to make the space allotted us on this website a place where we explain the strategic motivations behind the actions of prominent political candidates and those of prominent political leaders.
For the past few weeks, we have devoted our time exploring the strategies and the status of the U.S. Senate race between U.S. Sen. Mark Pryor, D-Ark., and U.S. Rep. Tom Cotton, R-Dardanelle. This week, we will follow the same trend and explain the latest polling developments in two of the closely watched statewide races in Arkansas: the U.S. Senate race and the race for Arkansas Governor.
LATEST POLLING DEVELOPMENTS
Per realclearpolitics.com, a site that aggregates and averages the most reliable public polls on political races, the U.S. Senate race in Arkansas is slightly trending Pryor’s way.
By averaging all the reliable public polls that have been released since the beginning of this race, Pryor edges Cotton 45.8% to 41% – a difference of 4.8%. Of the most recent polls, put Pryor up by double digits, including the latest NBC/Marist poll that puts Pryor up by an astonishing 11 points.
As for the Arkansas gubernatorial race, Realclearpolitics.com averages the public polls that have been released since the beginning of this race and finds that the newly minted GOP nominee Asa Hutchinson is leading the newly minted Democratic nominee Mike Ross 44.8% to 41% – a difference of 3.8 %. Of the most recent polls, put Hutchinson up by 8 points and 7 points respectively!
At first glance, the results of these polls might seem counterintuitive and generate some legitimate questions: Why and/or how is it that despite the bad political environment for Democrats and President Obama in Arkansas, and despite Pryor’s vote for the unpopular Affordable Care Act (a.k.a. Obamacare), Pryor could be leading Cotton? Why and/or how is it that Ross is trailing Hutchinson, despite the fact that by most account Ross is a better retail politician than Hutchinson and Ross cannot easily be associated with liberal national democrats that are unpopular in Arkansas? Why are Cotton and Ross apparently not up to par with the expectations that many political handicappers had for these two candidates? These questions would mostly find an answer in the following two factors:
• The nature of the target demographics used in the above-mentioned polls; and
• The level of name recognition that each candidate has among the electorate
Before the operationalization of a public survey, pollsters have to determine, among other things, what the target group – technically called population – of their study would be. For example, if pollsters wanted to assess the standing of say Gov. Beebe among adult Arkansans, the target population of this particular survey would be all Arkansans aged 18 and older. But if pollsters wanted to assess Gov. Beebe’s standing among registered voters, the target population of this survey would be all the registered voters of Arkansas. Last but not least, if pollsters wanted to probe likely voters, on these latter’s feelings toward Gov. Beebe, only those who are likely to vote would interest the pollsters in this survey.
Because it would be extremely tedious and cumbersome to interview all the individuals that make up a target population, pollsters randomly select a representative group out of the target population so as to more or less infer the feelings of the whole target population. For example, can you imagine how tedious and cumbersome it would be to individually interview all the 1,624,186 Arkansans registered to vote? It thus becomes imperative for pollsters to randomly select a representative group – technically called the sample – out of the target population. The ideal would be for each individual in the target population to have a known and preferably equal probability to being selected as part of the sample.
With this in mind, one could better understand why the results of polls pitting Ross against Hutchinson, or Cotton against Pryor would yield different results depending on whether the pollsters survey adult Arkansans, registered Arkansas voters, or likely Arkansas voters. Of the seven reliable public polls released this year on the Ross-Hutchinson race, five polls used the registered voters model. In those five polls of registered voters, Ross is either tied with or leads (+6) Hutchinson in 2 earlier polls, while Hutchinson leads Ross in 3 recent polls (+ 1, +8, and +7 respectively).
Of the nine reliable public polls released this year on the Cotton-Pryor race, five polls used the registered voters model. In those five polls of registered voters, Cotton leads Pryor in one earlier poll (+4) and Pryor leads Cotton in 4 recent polls (+10, +10, +1, and +11).
The main reason why pollsters would want to use the registered voters model is to gauge the ultimate standing of the candidates among the bigger pool of all the registered voters. Though all polls are just snapshots of the mood of the electorate at one given moment and may not predict the outcome of elections, polls that use the registered voters are even more problematic in predicting electoral outcomes. If the election were to be held tomorrow, the only way for Hutchinson to win by 8 points – the latest NBC News/Marist poll puts him up by 8 points – and the only way for Pryor to win by 11points – the latest NBC News/Marist poll puts him up by 11 – would be for these two candidates to have a humongous Get-Out-The-Vote operation that would get almost all registered Arkansas voters to the poll. For that to happen, Hutchinson and Pryor would have to defy historical trends and get close to 100% of the registered voters to the poll, which would be quite a feat.
Over the past 14 years, off-year elections (2002, 2006, and 2010) in Arkansas have seen an average of 55.12% voter turnout, with the highest turnout occurring in 2010 with a 66.7 % voter turnout. So, registered voters polling models though insightful do not quite capture the candidates’ standing among the voters who would most likely show up at the polls. Hutchinson and Pryor appear to have a considerable lead over their respective opponents, mostly because at this stage of the campaign registered voters can identify these two candidates better than their respective opponents.
Because Hutchinson and Pryor have been on statewide ballots for the past few decades, their names are familiar to Arkansas voters – many of whom are (still) part of the pool of registered voters who may or may not be politically engaged. The main reason why Ross and Cotton appear to be trailing in polls of registered voters is that Ross and Cotton are mostly known only in their native 4th Congressional district and have not yet aggressively campaigned in the rest of the state.
They did better in some earlier registered voters polling models mostly because of the buzz that surrounded the launching of their respective campaigns and generated some positive media coverage. Ross and Cotton appear to consistently fare better in polls of likely voters who tend to be more in tune with the political zeitgeist and less influenced by a candidate’s last name.
ON TO THE GENERAL ELECTION
Stating the above should not take anything away from Hutchinson’s and Pryor’s respective performances so far. Both have campaigned hard to win the support of Arkansas voters. However, it is now that the primaries are over and the candidates have secured their respective parties’ nomination that we will begin to get a better sense of where the two political races stand.
Overall, our sense is that these two political races are (still) in a statistical dead heat, with each candidate being within the margin of error from his main opponent. Before the primaries, the candidates were somewhat restricted in how they could campaign, because they had to avoid looking pretentious by refraining from acting and campaigning like nominees without officially securing their respective parties’ nomination. These four candidates devoted the past few months mostly to raising funds and building campaign infrastructures.
The next few months would be mostly geared toward establishing a contrast with one’s main opponent, fine-tuning a solid grassroots organization, and crisscrossing the state to better connect with Arkansans. The general election has just officially begun.