Top Ten Who Lost: #10
Every election has winners and losers of course, but some wins or losses are more significant than others. Without additional ado, the Political Parlor presents the first on a highly subjective list about the Alabama elections, the Top Ten Who Lost.
#10. The Birmingham News
The Birmingham News election coverage was rounded out by a poll published on Sunday two days before the election, in much the same way that a bridegroom’s attire is rounded out by the toilet paper trailing from his shoe: no matter how nicely put together everything else is, all some people remember is the big mistake. In this case, the paper’s poll missed badly, and insiders with their own polling numbers knew it immediately.
Jim Folsom’s campaign tried to blunt potential damage by circulating a release challenging the accuracy of the numbers and even the motives of the News for publishing it. Actual results of the elections two days later indicate that Folsom’s campaign was on to something; the News missed the numbers for the Democrats in the five races by an average of 11 points.
| Race R/D |
News Poll #’s (%) R/D |
Election Results (%) R/D |
|---|---|---|
| Governor Riley/Baxley |
59/31 | 58/42 |
| Lt. Governor Strange/Folsom |
50/39 | 49/51 |
| Attorney General King/Tyson |
51/33 | 53/47 |
| Sec. of State Chapman/Worley |
42/32 | 58/42 |
| Chief Justice Nabers/Cobb |
43/44 | 48/52 |
The fact that others recognized and challenged the mistakes immediately minimizes the likelihood that the poll was accurate at the time and that the election results reflected later shifts.
The potential damage for such a mistake is large. Such a story risks becoming a self-fulfilling prophecy. Instead of merely reporting the news, the story can create or at least influence the news. The potential for published poll results to influence an election is exactly why France does not allow poll results to be published in the week before an election. If accurate poll results are capable of influencing election results, think of the injustice made possible by the influence of a wildly inaccurate poll.
A commenter here correctly pointed out that The Birmingham News used the same pollster in 2002, and that he missed by double digits then too. On Sunday, Nov. 3, 2002, again just two days before the election, a Birmingham News story reported poll results for only the Governor’s race:
The survey of 600 registered voters who say they are likely to participate in Tuesday’s election shows Riley leading Siegelman 47 percent to 39 percent. Eleven percent said they were undecided, and 3 percent said they would vote for Libertarian candidate John Sophocleus. The margin of error was 4 percentage points.
The final election results two days later actually had Siegelman 10 percentage points higher as he and Riley finished in a virtual tie at 49% each.
I have heard an unconfirmed rumor that as a result Birmingham News editor Tom Scarritt is now telling people that The Birmingham News is out of the polling business.
Right on, Bham News needs to apologize publicy for this
Comment by Anonymous — November 16, 2006 @ 2:42 pm