xml:space="preserve">
xml:space="preserve">
Advertisement

Hogan and Brown: tortoise and hare?

Aesop's fable "The Tortoise and the Hare" provides an appropriate allegory to reflect upon polling and Maryland's recent gubernatorial election.

Here's the story: A tortoise challenges a hare to a race. Everyone (pundits and voters), including the hare (and his campaign organization), thinks the hare is going to win. At the beginning of the race, the hare jumps out far in front of the tortoise. He is so ahead (in the polls) that instead of continuing to run the race, he decides to take a nap. The hare awakens (in late October) to find the tortoise nearly at the finish line. The hare runs as fast as he can, but (even with help from the Obamas and Clintons) he is unable to catch up to the tortoise (who now has voters cheering for him at the finish line).

Advertisement

The tortoise/hare metaphor works particularly well in the bright, post-election light. In fact, during his victory speech, Governor-elect Larry Hogan was quick to point out, "The polls and the pundits and the media seem to talk to each other. It's sort of like an echo chamber." In other words, pollsters, pundits and the media were busy talking about how unlikely it would be to beat the hare, instead of focusing on the progression of the tortoise.

Part of Mr. Hogan's observation is absolutely correct: The vast majority of pundits and political observers thought the 2-to-1, Democrat-to-Republican ratio in the state would be too much for Mr. Hogan to overcome, even if he ran the better campaign. However, in regard to polls in the echo chamber, the issue is not that the polls were talking to the media and pundits.

The real issue is that not enough polls were conducted following the debates and leading up to Election Day. If there had been more polls in Maryland during that crucial timeframe, the echo chamber would have sounded different.

Polls are snapshots in time. In a fluid election — particularly where one campaign is stagnant (a napping hare) and the other is steadily pushing to the finish (a resolute, on-message tortoise) — we cannot expect polls taken a month before the election to predict the outcome. The live-caller polls in Maryland provided an accurate representation of the gubernatorial race, during the timeframes in which they were conducted.

Consider when Maryland's publicized polls were administered to likely voters. The poll put out by The Washington Post/University of Maryland was conducted before the first debate, and only one day of the polling by The Baltimore Sun/OpinionWorks crossed into the post-debate timeframe. From here, any number of post-election analyses by leading politicos can explain how Mr. Hogan was able to overcome a polling deficit prior to the debates. In keeping with the analogy, the hare had jumped ahead to a big lead then stalled, and the tortoise steadily gained ground right up until the very end of the race.

The DC-based research firm Wilson Perkins Allen, working on behalf of the Hogan campaign, did correctly predict the outcome of the election — hats off to them because this is never an easy task. However, independent polls — such as The Sun/OpinionWorks or The Post/UMD polls — would have likely gotten it right, too, if they were polling during the days right before the election. The lesson here is that the timeframe of polls and their respective campaign contexts matter a lot when evaluating polls.

I am not advocating for blind faith in public opinion polls. Polls of public sentiment need to be grounded in the time-bound context in which they are conducted. Pollsters are not fortunetellers, and their polls can only tell us what the target population thinks during the timeframe in which respondents are surveyed.

Mileah Kromer is the director of the Sarah T. Hughes Field Politics Center at Goucher College, which conducts the Goucher Poll. She is also an assistant professor of political science. Her email is mileah.kromer@goucher.edu; Twitter: @goucherpoll.

Advertisement
Advertisement
Advertisement