Contact InformationMain office:
3rd Floor, main building
Meet the department:
The CCRI Department of Marketing and Communications is the hub for all internal and external college communications. If you have news you want people to know about, let us know!
Web Content manager
Pollster Joe Fleming ’74 analyzes public opinion during Election 2012
Nov. 19, 2012
From the Myrth York vs. Lincoln Almond campaign to Sheldon Whitehouse vs. Lincoln Chafee, to David Cicilline vs. Brendan Doherty, Rhode Island pollster Joe Fleming ’74 has been predicting the state’s landmark elections for almost 30 years.
A 1974 Community College of Rhode Island graduate, Fleming is the president and co-founder of the polling firm Fleming & Associates and a political analyst for WPRI Channel 12. If you paid attention to local political polls this past election season, or even contributed to one yourself by taking a telephone survey, Fleming was likely the man behind the effort. Fleming worked on his first political campaign, a Town Council race in his native Pawtucket, in 1969 and fell in love with politics. In 1984 he opened the polling firm Wolf, Fleming and Associates with his business partner Scott Wolf. Wolf left shortly after to run for Congress, and Fleming took over the business himself.
Fleming & Associates used to be a sideline business and a way for Fleming to indulge his interest in politics, but it has become more of a full-time job since he retired as assistant principal of Samuel Slater Junior High School in Pawtucket in 2009.
“I love doing it,” Fleming said. “I just love politics ... It’s always very intriguing finding out how voters feel about issues and candidates, and the job is different every two years.”
Americans were inundated by polling data over the last few months, but how were these polls conducted? Fleming’s method is similar to that of pollsters nationwide: He commissions many, many telephone calls. Participants are called at random and asked a series of questions that differ depending on the topic of the poll.
Simply asking a large number of people for their opinions does not guarantee an accurate survey, though. The Literary Digest magazine found that out in 1936 during one of the first comprehensive nationwide polls ever conducted, which was about the Franklin Delano Roosevelt vs. Alfred Landon presidential election. Thousands of the magazine’s subscribers were asked about their presidential preference and most selected Landon, causing the magazine to predict a Landon landslide. But they neglected to consider the fact that their subscribers were overwhelmingly of the upper classes and a skewed sample resulted in a skewed survey.
To prevent a mishap like this, caused by a phenomenon called sample bias, Fleming and other modern pollsters have to worry about demographics. In any given sample size, the percentages of men and women, age groups, socio- economic classes and races among the respondents have to correspond to the population being represented. It can be hard to reach members of demographics such as young men (men answer the phone less often and young people frequently don’t have landline telephones) so a poll sample of, for example, 500 people actually can take thousands of calls to complete.
Another of Fleming’s worries is biased questions. “It’s important that the questions are unbiased,” he said. “I could do a poll showing that you are going to be elected governor by the way I ask the questions.” Finally, in the name of accuracy, pollsters have to be sure they are surveying people who will actually vote – people who stay home on Election Day can’t influence an election, after all. Pollsters determine that respondents are so-called “likely voters” by asking whether they are registered to vote, voted in the last election, know where their polling place is and how likely they are to vote. In the closing days before an election, pollsters do not bother to survey people unless they say they are definitely voting. Earlier on, though, their opinions are still sought as potential voters. Including or excluding uncommitted voters is just one example of the ways poll results can change over time.
“A poll is nothing more than a snapshot of voter opinion at that time,” Fleming said. “It doesn’t mean they’re going to feel that way in three weeks.”
Politicians commission polls to gauge their popularity before deciding to run and then during their campaigns. Polls also are a great way to find out what issues voters care about most. Candidates can see how popular they are with different demographics and test out messages that will appeal to, for example, Latino voters or young women. “If you’re a politician, you’re using whatever groups you can move,” Fleming said. Poll results also have value as news stories, showing where the electorate is leaning and framing the stakes in a race. Fleming said that the sheer number of polls can be extreme, speaking specifically about the daily updates on the recent presidential election, but they have greater news value in Rhode Island because they are conducted less often.
Public opinion can change mightily be- tween polls. Last February, for example, Fleming’s surveys showed Brendan Doherty far ahead of incumbent U.S. Rep. David Cicilline. Cicilline took the lead in the summer and early fall, and then it was a statistical dead heat by Election Day.
Fleming believes that the methodology of polling is going to change in the near future. He expects to see greater use of the Internet, with online and email polls, and predicts more calls to mobile phones as people abandon landlines.
For all of his work in politics, Fleming has never held or sought elected office himself, except as a student senator and then student government vice president during his time at CCRI. He ran with two close friends, Tom Coffey and Jeff Sanders, as a ticket called “The Slate.” He met frequently with former CCRI President William Flanagan and called the student government “a good learning experience.
“I had a great two years at CCRI,” he added. “It really helped me get focused in life on what I want to do. My professors got me focused on attending a four-year college [Rhode Island College for a bachelor’s degree and Providence College for a master’s degree, both in education] and they seemed to really care what I did.”
Fleming still wears his CCRI class ring, embossed with a gavel as a token of his time in the student government. He is the president-elect of the CCRI Alumni Association.