Pollster.com

Articles and Analysis

 

About That Generic Presidential Vote

Topics: 2008 , The 2008 Race

That "generic" presidential vote question included on last week's Diageo/Hotline poll inspired quite a bit of discussion and speculation over the last few days. The result showed the unnamed "Democratic candidate" defeating the unnamed "Republican candidate" by a 47% to 29% margin. Meanwhile, most national general election head-to-head questions pitting Rudy Giuliani or John McCain against Hillary Clinton, Barack Obama or John Edwards typically show either a very close race or the Republican a few points ahead.

The discussion started, appropriately enough, when the Hotline's On Call blog posted this excerpt of an email from an unnamed Democratic consultant:

Senators clinton and obama are more than 20 points less desirable to voters than an imaginary Democrat. our "top tier" candidates may be the only Democrats in existence who can't win in 08."

Then Politico's Ben Smith argued that the numbers really indicated something different, namely that Giuliani and McCain are "very strong general election candidates, in part for the same reasons that they're flawed primary candidates." He noted that Edwards fares a little better than Clinton or Obama: "Edwards does show an edge on Clinton and Obama, but nowhere near enough to explain that 18-point gap." Yesterday, MyDD's Chris Bowers speculated that the result showed that Republican leaning voters "have to be prompted before they support Republicans, while voters are ready to support Democrats whether or not they are even being asked." And when Tim Russert chimed in on Meet the Press that he found the results "quite striking," NBC's new political director Chuck Todd (formerly Hotline Editor in Chief) said the result "does show a weakness in the Republican Party."

Two reactions:

First, we are putting quite a bit of emphasis on one result from one survey, especially given that the question came near the end of a long interview and followed immediately after another question that "reminded" respondents about a recent controversy not exactly helpful to the Republican brand:

Congress is currently conducting an investigation into Attorney General Alberto Gonzalez's handling of the dismissal of eight federal prosecutors. In your opinion...

Early in the same survey, Gonzales received a rating of 17% favorable, 35% unfavorable. The seven questions asked before the Gonzales item concerned the job performance of President Bush, the new Democratic Congress and the issues facing both. So altogether, the generic presidential vote followed a set of questions that "primed" a particular focus on issues personalities and issues that may have limited enthusiasm about a Republican vote in 2008.

Second, as Ben Smith suggests, the big difference between this particular generic result and other recent polls that tested actual candidates is the support registered for the Republicans. The table below shows the regression estimates of each candidates support (from Professor Franklin's charts, which are essentially averages of recent polls). The three best known Democrats each get about 43% of the vote (give or take a point or two), a result slightly lower than the 47% for the generic Democrat on the Hotline poll. However, things are much different for the Republicans. The Hotline poll's "generic" Republican wins only 29% of the vote compared to anywhere from 41% to 47% for McCain or Giuliani.

04-09%20generic%20vs%20named.png

Now consider how head-to-head results compare when tabulated by party. The two charts below show how the Hotline's generic results by party identification compare to the results on six head-to-head questions asked by a recent Newsweek poll. First, consider the support for Democrats. While clicking on the image below will display a larger version, you need not bother. The six thin lines generally bunch together, as each named Democratic candidate gets roughly the same support from Republicans, independents and Democrats against each named Republican. The generic Democrat on the Hotline poll (the bold blue line) does slightly worse with independents, slightly better with partisans, but that may be due to differences in the way each poll asks about party identification.

04-09%20dems-sml.png

Now look at the Republicans. Here the chart tells a very different story. The bold red line representing the generic Republican is much lower in each category than the lines for Giuliani and McCain.

04-09%20reps-sml.png

Obviously, pundits will interpret these results differently, but for me they largely highlight the weakness of "generic" questions. Typically, such questions tell us what respondents think of the political parties, and those images are shaped mostly by perceptions of the incumbent president. We know the Republican "brand" is at a low ebb, with a recent Pew Research Center report showing "leaned" Republican identification at a 17-year low of 35% compared to 50% for the Democrats. In this case, the order of questions may have also primed issues or controversies, further reducing the appeal of a Republican vote.

Either way, McCain and Giuliani are not perceived as "generic" Republicans, at least not as of today. Moreover, when confronted with a difference between a "generic" measure and one using the real names of reasonably well known candidates, this pollster will trust real over generic every time.

PS: Just to be clear, I do not mean to jump on the Diageo-Hotline poll. They asked a provocative question that obviously has us all talking. I just wish it had been a little earlier in the survey.

 

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR