Thursday, October 10, 2013

Is This a Fair Question?


A Twitter post by Greg Bluestein of the AJC caught my eye.
I'm a poll nerd, so I checked it out. Here's the column by the AJC's Jim Galloway that gives the basic lede and some analysis:
Atlanta Mayor Kasim Reed isn't the only one who thinks state Sen. Jason Carter's flirtations with a gubernatorial bid should be taken seriously.

Better Georgia, a left-leaning guerilla (sic) group, commissioned a Public Policy Polling survey of 602 registered Georgia voters this week that found Carter and Gov. Nathan Deal are running nearly neck-and-neck. Executive director Bryan Long, who tweeted #RunJasonRun earlier this week, said he ordered the poll after reading the news that Democrats in Washington in Georgia commissioned a survey to test Carter's popularity.
In the full poll results, there's this question:
Q3 If the candidates for Governor in 2014 were Nathan Deal, the incumbent Republican, or Jason Carter, a Democratic state senator from Atlanta and grandson of Jimmy Carter, how would you vote?

The boldface above is my own because it got me wondering about the effect of identifying Carter as the grandson of former Georgia Gov. (and oh yea, President) Jimmy Carter. For some, it might hurt. For others, it might help. Do the two cancel each other out? Among older respondents who may very well remember Carter's time in office, Deal wins. Among younger respondents, Carter wins. How much this is a case of remembering Carter's presidency and how much is a matter of party identification or the age of the candidates themselves it's impossible to say.

Among men Deal does better, 52-36, which is not unusual for a Republican. Among women Carter, the Democrat, leads 43-36. Again, not unusual. And it hardly bears mentioning that Deal cleans Carter's clock among conservatives while Carter does well among liberals. But among moderates, Carter wins 58-21 percent. That's a wow and must be heartening to those who support the guy.

But back to the question itself. Is it fair to identify him? Would you, if conducting a poll?

It makes sense to do so in that he enjoys little name recognition, but then again we'll see the same Nunn effect in the Georgia U.S. Senate race, so it raises an interesting question -- aren't we giving him name recognition by explaining the family connection, thus influencing the poll results? My gut says yes, but I'd love to see someone conduct a split-ballot poll in which half of the folks are randomly given the Carter family identification and half are not to see if it gets him a bump.







No comments: