Home Section A - News Dr. Don Levy Discusses Poltical Polling

Dr. Don Levy Discusses Poltical Polling

Photo Courtesy of Google Images
Photo Courtesy of Google Images

By Tyler Hofmann-Reardon


On Wednesday, March 11, Dr. Don Levy of the Siena Research Institute in New York, came to Roanoke College to discuss his work as a political pollster and to share his analysis of the voter tendencies becoming more common across the nation. Levy began his talk by describing the work he does as the director of the Siena Research Institute, explaining the process he goes through during every election season in New York. He explained that first the institute conducts a “horse race” poll during which pollsters directly ask the public, a few months before an election, which candidate they would vote for if the election was today. Depending on how far away from the election, these results will reveal varying levels of certainty from the public. The farther away the election, the less certain the public tends to be about which candidate they will vote for. According to Levy, this is caused by multiple factors related to the voter’s political efficacy and interest in the election or the candidates. With just the horserace analysis, it difficult to determine which candidate will win on Election Day because a significant amount of the surveyed voters fall into the undecided or “don’t care” categories. However, Levy and the Siena Research Institute have become experts at conducting complementing cluster analyses of the horse race data by asking a series of pointed issue questions. These questions are used to better put the voters in two camps based on the stances they take on certain issues. A cluster analysis of these numbers is then used as a more accurate predictor of the likely voting tendencies of the public. Using this method, Levy was able to accurately predict fifteen of fifteen elections held in New York last year.

After discussing his polling strategies and successes, Levy began to explain a fairly new concept in political polling; analysis of the “Zone of Ambivalence.” Along with Roanoke’s own Dr. Harry Wilson, Levy has, for the past eight years, analyzed various types of polling data looking for a group of voters who don’t care about issues, elections, and the decisions the government makes regarding the issues and elections which directly affect them. Using cluster analysis along with numerous other statistical procedures that delve deep into voter’s tendencies and what issues spark their voting fires and how much and issue influences a person’s likelihood to vote. After intense and in-depth analysis, Levy was able to notice a steady increase in the Zone of Ambivalence across the nation. The expanding of this zone has led to politicians focusing on a smaller percentage of the electorate, those outside of the zone. Unfortunately, this tendency just increased the expanse of the zone because people have begun to feel that politicians do not focus on their issues and thus voting numbers have continued to drop.

Overall, Levy’s talk was very informative and his analysis of the electorate and the Zone of Ambivalence was groundbreaking. For more information about Levy’s work, talk to Dr. Harry Wilson or visit Siena.edu and search for the Siena Research Institute.