by Alan Abramowitz
Alben W. Barkley Professor of Political Science
Emory University
The South is the most conservative and most Republican region of the country. In both the 2000 and 2004 presidential elections, the Democratic presidential candidate failed to carry a single state of the old Confederacy, although Al Gore probably did win a majority of the intended votes of Floridians. And even though Democrats made modest gains in the South in the 2006 midterm elections, Republicans continue to hold the large majority of the region’s Senate and House seats.
Looking at the bleak Democratic landscape in the South, Tom Schaller argues in Whistling Past Dixie that not only should Democratic presidential candidates write off the South, they should actively campaign against southern values in order to maximize their electoral prospects in the rest of the country. What Schaller is advocating is not just a non-southern strategy for Democrats, but an anti-southern strategy.
The assumption underlying Schaller’s argument is that not only is the South more conservative than the rest of the nation, but that southern values are now so antithetical to those of voters outside of the region that trying to appeal to southerners will only reduce a candidate’s appeal outside of the region.
But is it true that a candidate who appeals to voters in the South will reduce his appeal in the rest of the country? Based on an examination of the evidence from the past six presidential elections, the answer to this question is a loud and clear no. In fact, the evidence supports the opposite conclusion: the better a presidential candidate does in the South, the better that candidate will do in the rest of the country and, especially, in the key battleground states that determine the outcomes of presidential elections.
In order to test the viability of Schaller’s anti-southern strategy, I examined the correlations among Democratic presidential candidates’ vote margins (Democratic percentage minus Republican percentage) in five states across the last six presidential elections. The five states that I chose included two southern states, Georgia from the Deep South, and North Carolina from the Rim South, and three battleground states, Pennsylvania from the Northeast, Ohio from the Midwest, and Colorado from the Mountain West. The results are displayed in Table 1.
Not only are all of the correlations positive, all of them are very strongly positive—a correlation of 1.0 indicates a perfect relationship between two variables, and most of these correlations are very close to 1.0. It is clear that over the last six presidential elections, the better the Democratic candidate did in Georgia and North Carolina, the better that candidate did in Ohio, Pennsylvania, and Colorado.
There is no reason to believe that the positive relationship between a presidential candidate’s appeal in the South and that candidate’s appeal in the rest of the nation, including the key battleground states, will change in the future. The better the Democratic (or Republican) candidate does in the South in 2008, the better that candidate will do in states like Ohio, Pennsylvania, and Colorado that are critical to winning the presidency. That is because southern voters respond in the same way to the candidates and issues as voters in the rest of the country.
No matter whom the Democrats and Republicans nominate for president in 2008, the South will almost certainly be the most difficult region for the Democratic candidate. But is also almost certain that no matter whom the Democrats and Republicans nominate for president in 2008, the better the Democratic candidate does in the South, the better that candidate will do in the rest of the country including the key battleground states and the better that candidate’s chances will be of winning the presidency.
Latest Research from:
tds-wwc-roundtable
Someone made a comment that Democrats should try to win over Hispanics, in say Colorado, rather than “Southerners” who once were Democrats.
I think the Democratic Party should be very careful about trying to “win over” Hispanics. Hispanics are, and probably will remain, a swing vote. The reason is that there is no monolithic Hispanic group.
Generally, Puerto Ricans vote Dem, Cubans Americans vote Rep., and Mexican Americans split.
And on issues such as immigration, large percentages of Hispanics vote against the stand taken by Democrats in general (pro-immigant rights?).
There are just no “Latino” issues to exploit. My personal view is that trying to win back the South will be easier than getting a core of loyal Latino voters in the West. No matter how hard we try to create this monolithic Hispanic group, the reality will always be that it does not exist. And how can you “win over” something that does not exist.
For example, I am an historian, and I know that generally Mexicans do not see themselves as one group of people; instead, they have regional identities. But somehow, incorrectly, we in politics somehow equate Latinos as a group with similar histories, as is true with African Americans who all have links to slavery, racism, and legal segregation.
Instead, and I think most people here seem to agree, winning the South, or at least winning back two or three states in the South makes more sense. Consequently, the issues needed to win in the South will make a candidate viable anywhere in the nation.
As I said in another post, many people in the South who vote Republican are transplants from elsewhere (including many Latinos!). A Democrat who can win these votes while keeping those of loyal Southern constituencies such as African Americans, workers, and just plain old Yellow Dogs will take the White House every time!
David Pye
The south is not a place to write off. Sure there are plenty of die hard republicans. There also are lots of people who are on to the lying and criminal actions that the republicans have been up to for years. Many women especially will want to remove the republicans from office. There are a lots of Democrats in these southern states that want to turn our states BLUE.
We Dems need to be fighting hard in the South.
What many people forget is that a major chunk of our African-American base vote resides in this region. If properly motivated can force the GOP to burn up resources fighting to hold Georgia, North Carolina, South Carolina, Florida, Virginia and Louisiana.
Consistent wins by Democrats in this region will jump start the process of peeps who are ROC’s (republicans of convenienece) to switch to the Democratic party
Texas is ripe for a return to its progressive roots after being mismanaged by Republicans since 1994. The Txea Democratic Party needs to be rebuilt in order to make that happen.
There are 50 states in this country and our party needs to be competitive in all regions, including the South.
Senator Mitch McConnell should be ripe for the picking if he keeps leading his fellow GOP off the cliff.
Just his support for the drug company’s proifts alone should do him in. Sen. Schumer should taget this turkey early and let the people of his state know who’s pocket McConnell is in. Women should be a prime go-to group to knock off with. Now if he recruits a female candidate, they could be making history as well.
This is an interesting argument. It does underscore the fact that we must nominate a candidate who can truly appeal to all of America. In my opinion, the only candidate we have that has a combination of the experience needed to be President, and the appeal needed to win in the northeast, west, AND the south is General Wes Clark. If/when he jumps in to the race it will change the dynamic. He also gives us the only option for a candidate who can take the national security issue completely away from anyone the GOP nominates. I honestly believe that Wes Clark gives us our most genuine chance for a landslide victory in November of 08′.
Dr. Abramowitz’s thesis would be much more powerful if it took place in a country with relatively static demographics. But that clearly is not the case in the USA in the 21st century.
Let’s all take a deep breath and remember that the power of “the socially traditionalist working class” is waning everywhere except in the South where it is waning more slowly.
To build a party that truly represents working Americans, we must focus on the increasingly critical Latino and Under 30 voters. And not even the ones who vote, but the ones who don’t.
We can convince a non-voting Hispanic in Colorado to vote Democratic a great deal easier than we can convince a Reagan Democrat(?) in Georgia to switch back to the Democratic Party.
Sure, we’d love to appeal to both, but I want to use our resources most effectively now and in the future. It’s like the effort it takes to seduce someone who knows you slightly is a great deal less than convincing an ex-girlfriend to take you back.
Schaller here. Let me begin by saying Alan and I are friends and colleagues, and he’s a great econometrician and somebody rightfully respected in our discipline. (We’re even working on a piece together right now on a different subject.) I know his post here is intended to offer some provocative ideas and a bit of data to back them up.
But I’m not sure he has much of a result here, and not merely because of the choice of the states or the quality of the correlate analyses. Rather, it’s a somewhat superficial result, a discussion of slopes when it is the intercepts that matter.
For the math-challenged, think of the following scenario: One group of people is tall, another group is short, and then for some reason or set of reasons (e.g., the introduction of fatty, drive-thru foods and growth of suburbs which reduce daily walking rates) both groups gain weight. (Actually, this is happening in America, which is quite sad.) The weight increases for the two groups will likely be positively correlated and statistically significant. Is the taller group now, somehow, the same height as the other group? Of course not. Taller people are, on average, heavier than shorter people; when both groups gain, the taller remains heavier and, obviously, taller.
So, yes, it may well be that in some states the partisan gains are significantly and positively related, and perhaps that means that Democratic messages are working equally everywhere. But I’m not so sure: Go look at the Clinton-Gore gains between 1992 and 1996, which are anything but uniformly 6.3% in every state, which was their national improvement. States like MA, NH, ME, RI and such gained more than 10%; elsewhere, in all but FL and LA among the former Confederate states Clinton-Gore improved by less than the national benchmark, including several states where Clinton and Gore only improved 2-3%.
Finally, a thought experiment: Imagine a scenario in which Democrats improve by 4 percent in 2008 over their 2004 performance in EVERY SINGLE state. How many states would flip from red to blue? Answer: Four non-southern states–OH, IA, NM and NV, for a total of 37 precious electors, enough for an electoral college majority. None of the southern states would move into the Democratic column, because Bush won all 11 by at least 5 percent.
Aha, so the slopes could be all the same–a perfect correlation of 1.000–and because the intercepts (starting points) vary so much, the Democratic nominee would still will not a single southern state.
I realize Alan’s point is to refute the idea that rates of improvement outside the South will be different than inside the South. That would be a good sign that voters respond equally, and everywhere, to themes and candidates. Given the demographic heterogeneity of states, this to me to be very unlikely.
Righting off the South is a very bad idea. With the large number of African American voters in the region, not appealing to the South will only serve to ignore one of the Democratic Party’s most loyal constituencies.
Moreover, the South is growing leaps and bounds as people from other regions move there. These people live in the suburbs that ring southern cities. The “dirty little secret” is that these people help keep the republican Party strong in the South. Without the infusion of new blood from the transplants contests in the South–particularly on the state and local levels–would be more competitive.
A national candidate can do well in the southern suburbs, will win the Presidency. Ignoring or running against the South will only leave to defeat.
And last, doing so will leave African Americans stranded and will send a message that the Democratic Party does not stand with its friends!
I have the highest respect for Dr. Abramowitz, but I am unconvinced by these data. It stands to reason that the better a candidate does in one state, the better he or she will do in another state. I would be wiling to bet that there are similar correlations between the vote margins in swing states and vote margins in the Northeast; however, despite this probable link, Republicans have effectively exploited an “anti-northeastern liberal values” sentiment in the past. These data are meaningless, in my opinion, without something to contrast with: correlations between states in the Northeast. If the vote margins in Northeastern states are less correlated with sthe vote margins in swing states than are the vote margins in southern states (what a mouthful!), then I will buy Dr. Abramowitz’s argument.
Democrats cannot afford to write off the South. Our party must appeal to socially traditionalist voters in order to build a long-term governing majority.
It is imperative that our party win back the socially traditionalist working class and these voters are by no means limited to the South. As Dr. Abromowitz points out, such voters are quite numerous in states like Ohio and Pennsylvania.