Groundhog Day Poll

Americans' faith in their local groundhogs is under assault by national television, according to a survey researcher at the University of Virginia. With all the TV attention given in recent years to Punxatawney Phil, surprising numbers of people have come to see the famous Pennsylvania woodchuck as a national weather prophet, whose predictions could actually override the climatic readings of groundhogs closer to home.

In a 1996 survey designed by sociology professor Thomas M. Guterbock, a random sample of voters in the Charlottesville, VA, metro-area was asked to consider what they would think if it were clear and bright on Groundhog Day in Pennsylvania but cloudy in Virginia. Punxatawney Phil sees his shadow, but the local groundhogs do not. "If you believed in the Groundhog Day story and you wanted to know when Spring would be coming," the survey asked, "which groundhog ought you to rely upon?" Forty percent of those answering chose Punxatawney Phil, 45% chose the local groundhog, and 16% could not decide between them.

Guterbock finds this to be evidence of a disturbing trend he calls "Punxatawney Philistinism." The survey results point to the national media as the cause. Among people who watch the national news on TV every day, Phil outpolls the local groundhog 49% to 34%. But among those who rarely watch TV news, only 22% favor Phil, with an overwhelming 62% choosing to believe the local groundhog.

"TV networks like to show the Punxatawney groundhog because it adds an eye-catching visual' to the Groundhog Day story," Guterbock explains. "The local groundhog who doesn't have a press agent and a web site is simply overshadowed." The sociologist urges responsible local reporters and editors to remind the public that, according to tradition, the unique climatic conditions of each region can be properly read only by ground hogs on the local scene.

The random-dial telephone poll of 260 registered voters was conducted in October 1996 by sociology majors at the University of Virginia as part of an instructional project in survey methods. The poll has a margin of error of plus or minus five percentage points at the 90% level of confidence.