Until this morning, whenever I asked myself that question, the answer was an uncontested, "no possible way". But then, while perusing the latest Newsweek poll on Bush versus various Democrats (Clark, Dean, Kerry, etc.,) I noticed a startling result, reproduced in the table below (specific data from PollingReport.com):
"Suppose the next general election for president were being held TODAY and you had to choose between George W. Bush, the Republican, and [see below], the Democrat -- who would you vote for?" If "Other" or "Undecided": "As of TODAY, do you LEAN more toward Bush, the Republican, or [see below], the Democrat?"
George W. Bush Wesley Clark Other/
UndecidedALL 47 43 10 Men 52 41 7 Women 42 45 13 Republicans 82 11 7 Democrats 17 76 7 Independents 42 44 14 South 45 47 8 Non-South 47 41 12
While hailing from Arkansas, General Wesley Clark was leading pseudo-Southerner George Bush on his home turf, undermining the accepted premise that the South was in fact a solid bastion of Republican support. And in contrast, Bush's numbers against Clark outside of the South more closely mirrored his overall total.
I wanted to try and discover if there was any underlying reason which could possibly account for these surprising results, and I figured that the unemployment numbers I'd been crunching yesterday were a good place to start. I decided to compare stats from the South with those from another region of generally strong Republican support, the Plains states.
I first looked at the overall unemployment rates for both areas: Southern states averaged a rate of 5.6% in July 2003, while the Plains stood over a point lower, at 4.5%. But I wasn't fully convinced this was the deciding factor, as both areas had experienced a similar increase since January 2001: The South, 2.0%, the Plains, 1.8%, on average.
So instead, I looked at the actual number of jobs lost or gained since Bush took office for both regions. The results are listed in the table below:
Region/State | Total net employment loss/gain (1,000s) | Manufacturing jobs lost (1,000s) | % of total manufacturing |
| South: | |||
| Alabama | -67.3 | -59.6 | 16.9 |
| Arkansas | -18.2 | -42.0 | 16.9 |
| Florida | +85.4 | -94.7 | 19.6 |
| Georgia | -88.0 | -126.4 | 21.8 |
| Kentucky | -79.1 | -48.8 | 15.3 |
| Louisiana | -58.3 | -26.6 | 14.5 |
| Mississippi | -22.0 | -46.4 | 20.5 |
| North Carolina | -127.4 | -162.3 | 21.1 |
| South Carolina | -115.0 | -69.2 | 20.1 |
| Tennessee | -80.3 | -88.5 | 17.7 |
| Texas | -143.7 | -172.6 | 15.9 |
| Virginia | -25.9 | -80.0 | 20.7 |
| West Virginia | -8.7 | -14.0 | 17.6 |
| Total South | -740.4 | -1031.1 | |
| Plains: | |||
| Iowa | +173.5 | -40.0 | 15.4 |
| Kansas | +146.5 | -34.9 | 16.6 |
| Minnesota | +295.4 | -95.7 | 21.8 |
| Missouri | +251.5 | -82.4 | 20.7 |
| Nebraska | +84.1 | -15.9 | 13.3 |
| North Dakota | +32.6 | -0.8 | 3.2 |
| Oklahoma | +230.6 | -37.4 | 20.7 |
| South Dakota | +55.5 | -11.0 | 22.4 |
| Total Plains | +1269.7 | -287.1 |
[Source: BLS.gov]
Also listed in the table are the total of jobs lost in each state in the manufacturing sector since January, 2001, as well as what percentage of the total manufacturing base that number represents.
It's readily apparent that the significant job loss in the South, more than 740,000 net jobs lost, and over a million in manufacturing, are undermining support for Bush in the region. And while the Plains states have also experienced a large decrement in good paying manufacturing jobs, other sectors of the economy have made up for those losses, at least on paper, and thus the "pain factor" is most likely not as immediately transferred in anger towards those viewed by the electorate as responsible for the state of the economy.
I wouldn't be surprised if there are additional factors at play here as well. The large numbers of military families based in the South are growing increasingly wary of the Administration's plans for their loved ones in Iraq and Afghanistan, and it's likely that will eventually be reflected in such polls, if it isn't already. The growing retiree population in the region is concerned with Bush's plans for Social Security and Medicare, and veterans see governmental policies abound which run contrary to their best interests.
But overall, I think that the economic situation as a whole, and jobs in particular, are key to a scenario where the Democratic nominee could make tremendous inroads into areas previously ceded to Republicans. Of course, much of that depends upon the nominee. Obviously, Clark has shown his appeal to this region, but will he be hobbled in the future by his relative inexperience in economic matters? While I'm admittedly biased, I can't help but offer Senator Edwards as the best candidate at this point to tap into this economically-based anti-Bush sentiment. He not only can project sincere empathy, that "son of a mill worker" theme so many cynical pundits complain of, but he has detailed policies aimed at turning around the crisis plaguing rural and urban areas alike.
It's an opportunity no Democratic candidate can afford to throw away. Here's hoping no one does.
Posted by MB Williams at September 21, 2003 01:21 PM | TrackBackThanks for the analysis -- interesting as always.
Since so little is known about Clark by most of the electorate as yet, I wonder if his being a general is carrying considerable weight with the generally pro-military South. That might help explain why Clark in particular does so well there -- after all, many more poll respondants will know about his military background than about his economic plans.
Also, Fox has not yet had time to hammer away at Clark the way they have on the other main democratic candidates.
Posted by: Brandonimac at September 21, 2003 02:28 PMI'm not convinced Clark's Southern support is based completely on his military credentials, as I'd suspect that he'd have less of a gender gap if that were the case (unless you're arguing that it's Southern women who are finding the military background appealing.)
I'm sure it will all play out - I'd love to see Newsweek actually ask the same questions for another Southerner - Edwards or Graham, and see if the trend holds.
Posted by: MB at September 21, 2003 04:01 PMI think Fox had better be careful with their treatment of Clark. With his early gaffes in mind, it seems that they will attempt to characterize him similarly to Gore: socially maladroit vs. the off-the-cuff, shoot-from-the-hip approach of the president. If they begin to question his record in Kosovo, they open a Pandora's box with questions about Iraq and Afghanistan. As in 2000, the right wing will want voters to select their president on personality vs. issues. Voters had that luxury in 2000. There will be a different sense of urgency in 2004.
Posted by: Daniel at September 24, 2003 10:48 AMThe results are not valid statistically/methodologically (as much as it would give me great joy if they were). According to the poll-takers, the total sample size was 855 registered voters NATIONWIDE. Thus, one can reasonably draw the appropriate conclusions, but ONLY at the national level. Since "the South" is a subset of the national poll, the largest potential sample size is much smaller (muliply 855 by the proportion of the total US population comprised by the states of the South, and you'll get a sample size of a few hundred at most). Thus, the margin of error for the South only is not +-4 points, but could be as much as +-10 points (with a sample size of 250 in the South).
Unless there was oversampline of the South, one cannot validly draw the conclusions that you have.