| U.Va. Center for Politics Director Larry J. Sabato is contributing a regular column to Politico Magazine. This week, he examines polling averages, finding that while they are a very useful source of information, they occasionally misfire. — The Editors |
“Do you ever get the feeling that the only reason we have elections is to find out if the polls were right?” — Comedy writer Robert Orben, who once penned speeches for President Ford
America has a love-hate relationship with polls. If you’ve clicked on this column, you likely cull through the latest survey numbers every chance you get. Activists thrill to the polling that shows their party’s candidates ahead, and they tout these favorable numbers as though they were handed down to Moses on the Mount.
But if a survey has the opponent leading, many partisans not only feel a punch to the gut, they dig into its “internals” (racial demographics, party ID of respondents, etc.) to find fault. “Too many Republicans are in the sample!” “No way will Democratic minorities be so large a proportion of voters on election day!” There are always defects that supposedly twist the real state of the race, as perceived by activists.
Mostly, these are defense mechanisms to keep people from becoming discouraged. It’s also true there are plenty of bad polls floating around — party internal polls that find a way to skew the results in their nominees’ favor, poorly constructed surveys and samples that distort the numbers, and flash polls done on the fly.
To read the rest of the column, please click here.