Earl Ofari Hutchinson's take on the politics of the day
The Hutchinson Report
American Urban Radio Network
Al Sharpton Show
Monday 10:00-11:00 AM PST 2:00 to 3:00 EST
Streamed on http://tunein.com/radio/WURD-900-s23419
KPFK Radio Los Angeles 90.7 FM
Saturdays Noon to 1:00 PM PST
Streamed on http://www.kpfk.org/programs/181-hutchinson-report.html
In an exclusive interview on The Hutchinson Report Newsmaker Hour with host Earl Ofari Hutchinson on KTYM 1460 AM Los Angeles on February 24, with Peter A. Brown, Assistant Director of the Quinnipiac University Polling Institute
EOH: Why are there so many different variations in percentages in polls?
PAB: All polls are not created equal. There are different kinds of polls Number one is who is polled? You have to come up with a random sample and you do that through random digit dial, where a polling operation buys blocks of exchange phone numbers, area codes and the first three digits. Then we randomly computer generate the next four numbers who are called. Number two, when you make your phone calls, what kind of phones do you call? Almost 1 in 3 adults have cell phones only. If you just call land lines, you’re only going to get about 70% of the electorate. Cell phone only people tend to be younger, a little less affluent, so if you don’t call cell phones you are missing a broad swath of the electorate. Quinnipiac calls cell phones, as does the New York Times CBS Poll, NBC Wall Street Journal, ABC Washington Post, Gallop and Pew. A lot of other polling organizations do not because the law prohibits computer generated phone calls to cell phones. We can dial cell phone numbers individually because we use human beings to ask the questions. There are other polls that are computer generated where computers make the phone calls, so those polls don’t have cell phones.
EOH: That could somewhat skew the results of a poll.
PAB: That’s correct. We noticed in 2008 that young people disproportionately voted for Barack Obama. If you didn’t get cell phones, you probably had a smaller margin for the president. The universe of people called influences the results. The next question is, when you call a household, how do you know who you get and who do you want? You want a random sample of the people in that household. When we call a household, we ask to talk to the person with the next birthday, therefore we randomize who you are going to talk to and ask the questions of. Another issue on polling is sample size. The larger the sample size, the smaller the margin of error. Most national polls do up to 1000-1200 interviews. We go larger. We had a national poll released this week that had over 2000 interviews.
EOH: What are your thoughts on the criticisms of polls?
PAB: All polls are not the same. Polling is a business and there are all kinds of pollsters. In the political polling business, there are three kinds of pollsters; the ones who only work for Republicans, the ones who only work for Democrats, and the ones who are completely independent. Quinnipiac is an independent poll. We take no money from anybody and we don’t do survey’s for anybody. There are some other universities that do polling in their own states, and quality may vary. We take a great deal of time to make sure our questions are not biased and we weigh the language very carefully. When you’re in a business like that, everybody complains. The best way to measure a poll’s worth is to look back after the end of a political campaign and see how accurate they were to the actual results.
EOH: Like polls such as Dewey and Truman in 1948 …
PAB: Those polls were almost non-existent. My understanding is that they stopped polling a week or two out because they thought it was a landslide. They were obviously wrong. But again a much smaller percentage of the country had telephones, and that’s important. It was a different environment.
EOH: Do politicians pay close attention to the polls?
PAB: We don’t work for politicians, so whatever I would say is observational. Polling is important to them because it tells them where they are in the election, and gives them a sense of where everybody else is. Good polling asks questions about issues, and that certainly provides answers that they may find important. Major politicians have their own pollsters. They will look at the public polls, but the campaign pollsters allow a measurement of how the candidates’ message is playing with voters.
EOH: Have you seen politicians shift their views because of polls?
PAB: I’d be naive to think that it doesn’t happen, but it’s hard to pinpoint because you have to be inside to understand what’s going on.
I don’t doubt that politician’s read polls done by the general media. The question is, how it affects their behavior. Good poll numbers help a candidates fund raising; bad numbers hurt a candidates fund raising. That is another reason why they are very focused on them. Something to consider about this Republican campaign is that everybody wants the momentum. Everybody thinks if you’re the hot candidate you’re gonna win.
EOH: Has it been your observation that polls can influence voting by the general public?
PAB: Certainly there are instances where one candidate or another is up in the polls by a large amount and you see voters who earlier had been indicating they might go for someone else deciding either not to vote or vote for the front runner. Barack Obama got 53% of the vote in 2008. My guess is if we had taken a poll 4 months later and asked people who they voted for, you would find a much larger number than 53% would say Barack Obama. It’s human nature to want to be with the winner.
If any of your listeners want to take a look at our polls, go to www.quinnipiac.edu and look for the polling institute prompts on the main web page. We are completely transparent. We disclose our results and our questionnaires.