Jim Barksdale and August Polling
As I’ve mentioned, I think that electing Jim Barksdale in November is important for the future of Georgia Democrats. We finally have some polling on the Senate race, and I’ve taken a look at the two* major polls (WXIA and AJC).
Here’s a brief summary:
There are some stark differences between the polls. The most obvious is the +4 to Trump in the WXIA poll vs. the +3 to Clinton in the AJC poll. Similarly, Barksdale’s change from -9 to -6 between the two. There are some logical reasons as to why this was the case.
First, the polls are of different structures. The WXIA poll used likely voters while the AJC poll used registered voters. This is seen most obviously in the party ID splits. The WXIA poll definitely leans more to the Republican side of the scale (33 vs. 26) from the start and becomes more pronounced as the poll is screened for likely voters. A couple caveats: first, we know that party ID is not very effective in weighting a poll, so it really shouldn’t be used as a strong measure; second, the AJC poll has about 15% of those who did not respond to the party ID question. These two points could explain away the difference, and I’ll say right now that I’m not advocating for the “unweighting” of polls or anything of the like. However, the difference in party ID does explain some of the difference in the two polls. I don’t necessarily think either is right or wrong, but it’s interesting to note especially considering the share of Republicans and Democrats sticking with their respective candidates does not change drastically from poll to poll.
Keeping with the structure issue, there is a difference between likely and registered voters. Obviously, not all registered voters will actually head to the polls, so the likely voter screen is intended to correct for this issue. This is another potential reason for the Clinton and Barskdale bump in the AJC poll: the registered Democrats in Georgia are less likely to actually vote. With Clinton money starting to come into the state, the GOTV on the election day should be strong, and the actual results may trend more towards the registered voter spectrum rather than the traditional likely voters of Georgia.
Second, the timing of these polls could be significant. Trump seemingly spiraled out of control all of last week, so the dates of the two polls will play a role in the results. The WXIA poll was taken immediately after the DNC, but before most of Trump’s crazy week. (Here’s a little refresher of the timeline.) Everything started spiraling out of control on August 1st. Of importance for us here is that the WXIA poll missed all of that week while the AJC poll was taken during the midst. This, combined with the likely vs. registered voter point, explains much of the 8-point bump for Clinton. Barksdale doesn’t get the full eight points but still sees some benefit.
What This Means for Barksdale’s Chances
So what does this all mean for Jim Barksdale? Well the first thing to remember is that we’re only in August. There’s still plenty of time before November 8th. Barksdale hasn’t done much campaigning and hasn’t introduced himself to the voters. Once he does, then the polls will tell us more. Even so, there are a few bright spots for me. I’m focusing on the positives here because we still don’t know enough about Barksdale to know why he might be doing poorly with certain demographics.
The first place to look here is at the undecideds. In both polls, there are more undecided Independents and Democrats than Republicans. This makes sense given the way the campaign has played out. Isakson is an incumbent Republican. The Republicans are going to stick by him and have less to choose. However, Democrats and Independents need to decide if they want the incumbent who has been part of Georgia politics for quite some time or the unknown candidate the leadership had to dig up. Until we know more about Barksdale these people will stay undecided. WXIA has 9% of Clinton voters undecided while 6% of Clinton voters are voting for Isakson. 6% isn’t bad, but we have to be sure the 9% undecided Democrats don’t add to that total. Even if these undecideds break at a similar rate to those with their mind made up, then we will see positive results for Barksdale. Not a plurality at this point, but within the margin of error and keeping Isakson under the 50% threshold.
The most interesting point for me is with the AJC’s Vision of America question. They asked the respondents which of these two statements they agreed with more:
“America is in decline and the issues facing our country today threaten our way of life and demand a president who will establish law and order.” (Trump proxy)
“America is the greatest nation on Earth and is stronger when all citizens are valued and work together to solve our toughest problems.” (Clinton proxy)
Overall, 53% agreed with the Clinton message compared to 40% for the Trump message with 17% declining to answer. Here are the crosstabs on that question for the Senate race:
Just looking on the face of this question, we see a similar race: Barksdale down 5 points. But I see an opportunity in the undecideds and the voters who believe in the Clinton vision yet are still with Isakson. This is where Barksdale can win over voters. Barksdale can show that he is the candidate for Clinton’s vision of America and Isakson is the candidate for Trump’s vision. We have a natural advantage when looking at the scope of the race this way. If Barksdale can define the race on these terms, then he has an opportunity to make this competitive on election day.
There are interesting points in both polls where Barksdale is doing worse than Clinton (females, college educated, etc.), but that’s partly to be expected. Isakson is the incumbent and will pull some weight from that status. Clinton will also take some votes from Republicans who don’t want Trump. This causes a larger gap between her and Barksdale. With all that in mind, the most important takeaway is the potential for Barksdale to show who he is to the Democrats in Georgia. Barksdale hasn’t been defined for anyone yet, so now is the time to start. The debates (if they happen) will be a chance for Barksdale to prove himself to the undecideds and those who are currently leaning towards the incumbent.
Isakson’s floor has to be at least the percentage of Trump voters. Very few of those voting for Trump will crossover to vote for Barksdale. It’s not impossible, and I can understand the argument for those who would crossover. However, the number is too small to matter for Barksdale. The key here is for Barksdale to prove himself to those questioning his abilities as a candidate and Senator. With Isakson failing to separate himself from Trump, he can easily be paired to the top of the ticket. Showing that Barksdale’s vision for America is not the Trump/Isakson vision for America is imperative.
*I chose the AJC/ABT SRBI and the WXIA/SurveyUSA polls because of the quality of these polls compared to the JMC Analytics poll.