Pollster.com

Articles and Analysis

 

A Likely Story

Topics: Gallup , Likely Voters

My NationalJournal.com column for the week is now online. It revisits the nearly two-week old USA Today Gallup poll that showed a big difference between registered voters and those selected as "likely voters" with a focus on the age of the likely voter pool.

After you read the column, the following data may be of interest. First, notice that while the most recent, conducted in late July, showed a net shift of seven points between registered and likely voters, no such gap existed in the poll conducted just a month before. In mid-June, Obama led by six percentage points among both registered and likely voters.

08-07 gallup last2.png

What makes that difference interesting is the additional data generously provided by Jeff Jones of the Gallup organization showing how respondents in different age groups answered the four questions used to identify likely voters. As noted in the column, younger voters tend to score lower on all four questions. Notice that the percentage of 18-29-year-olds who said they had given "quite a lot of thought" to the election plummeted from June (60%) to July (45%). Similarly, the percentage who rated their chances of voting as a 9 or 10 on a 1-10 scale dropped ten points (from 69% to 59%).


08-06 Gallup likely questions.png

Thoughts anyone?

Update:  Nate Silver has additional thoughts.  Note that the method he describes as "the most logical way to handle" the likely voter problem is, in essence, the way the CBS/New York Times poll will model likely voters in October.  Their most recent release provides results for registered voters, but not likely voters.

Also, see the related comments we just posted from Time/SRBI pollster Mark Schulman. 

 

Comments
nimh:

Young people are more fleetingly interested, and are thus more likely to be mobilised only when the race is big?

Right now the elections feature less prominently than they did at the end of the primaries, or than they will in November. The commitment of those most fleetingly interested will peel off first in months like these, but that doesnt mean they wont mobilise back again once the race heats up.

Alternatively, the youth buzz over Obama has peaked, and young people are returning to their standard perspective that they're all the same and it doesn't matter anyway.

But I want to point to something else.

Look at the numbers again, and don't just look at the much feted 18-29 year olds. Look at the 65+ group.

The weakening of interest and attention among the youth that you flag so prominently in this post is no smaller among the elderly.

Sure about voting this year? The percentage among young voters is down 10; but among the 65+, it's down 16!

If you add the numbers in the four columns given, the total for young voters is down 31, the total for elderly voters is down 29.

How does that figure in?

____________________

brambster:

I made some comments about this in the original thread concerning both the young and the urban voter, and how the likely voter model that you described being used was pre-dispositioned to under-count these groups.

Having voted in the last election is a qualifier, and it is unlikely that most

On the other hand, McCain is strongest with older voters who are more likely to have voted before and are likely to vote again, as well as likely to have paid more attention to the election as they are the biggest demographic in news.

Gallup just simply threw out results if they scored 1 or less of the three, and greatly discounted the 2's, and as a result over-counted the 3's. This might have worked out well in a previous poll due to sampling error among a much smaller sub-group, but I would suggest that the science of their likely voter model is flawed. You really must count some votes from every group, even the ones that say they aren't likely to vote. These groups are so skewed from end to end that something must be done to take the age-related bias out of the weighting.

____________________

brambster:

Let me fix the text that was left out somehow...

Having voted in the last election was a qualifier and it is impossible that anyone under the age of 22 (by November) had voted in a presidential election. That right there put the young group, Obama's best age demographic, at no better than 2 on their scale of 3, and accordingly discounted their votes. Also, urban populations of lower socioeconomic status do not follow campaigns generally as close to surburban middle and upper class professionals, and the urban populations tend to be the heaviest cell-phone users and harder to reach due to odd work hours and general city life. So for Obama's best geographic group, they are also being too heavily discounted and also possibly under-polled.

____________________

jsh1120:

I have to say that I was flabbergasted to see just how the Gallup LV model is applied. I had always assumed that a probalistic weight was applied to responses to take into account the LV effect. To see that it's simply a binary (and arbitrary) cutoff leaves me slackjawed.

I can only assume that this results either from their experience in applying the model late in the election cycle in the past or (as I unfortunately suspect) limitations in the calculation software and an unwillingness to modify it and risk other unexpected effects.

I don't mean to disrespect the Gallup organization, but I've been in the sofware consulting business long enough and worked in the polling industry long enough to be suspicious that failure to adjust models appropriately as often stems from a reluctance to touch ancient COBOL code as from a considered analysis of the value of a change.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

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