Not the most productive evening, what with tornado sirens periodically blaring (the man, the dog, and I hunkered down in the basement for a while) and nifty interactive maps at the NYTimes to keep refreshing.

The Tennessee county-level results currently show that, although Clinton won the statewide race by a 23 13% margin over Obama, Obama decisively carried three of five major metropolitan areas (Memphis, Nashville, and Chattanooga). (To be honest, I was startled to see Clarksville on that map, but after checking population figures, I can't include Chattanooga and leave out Clarksville. To give you a sense of scale: Chattanooga, Clarksville, and Knoxville are smaller cities with population below 200,000; Memphis and Nashville have over 500,000.) Looking closer at the demographics:

Memphis: 61% black, 34% white; Obama wins 71% of the vote and four surrounding counties
Nashville: 79% white, 15% black; Obama wins 59% of the vote and one neighboring county
Clarksville: 68% white, 23% black; Clinton wins 52% of the vote
Chattanooga: 60% white, 36% black; Obama wins 52% of the vote
Knoxville: 80% white, 16% black; Clinton wins 49% of the vote

A quick rollover across non-urban counties shows Clinton posting wins in the 65%-80% range.

Major disclaimer: my population figures and percentages are kludged from multiple sources from different years (i.e., whatever I found on the 'net within the past five minutes). I'd laugh myself sick if an analyst tried to hand it to me as solid statistical evidence of anything.

That said, the patterns here loosely, anecdotally support my overall perception of how the state's regions characteristically pan out: Shelby County (Memphis) and Davidson County (Nashville) historically skew much further left than the rest of the state. Middle Tennessee (which includes Nashville) is sometimes characterized as significantly different than East or West Tennessee, but I can tell you that if I drive forty-five minutes in any direction, I will end up in a county that's typically regarded as a Republican stronghold. I'm slightly surprised that Obama did well in traditionally conservative Williamson County (to the south of Nashville) -- 49% to Clinton's 47%; it would be interesting to find out if he's perceived as the more pro-business candidate, since my impression is that half of Franklin works in Nashville but lives south of town for the bigger yards and better public schools. (Again, this is anecdotal, not serious sociological observation, and the anecdotes are skewed by a certain demographic among my friends being thirty- and forty-something executives and consultants with young children.)

The Wiki article on TN's 5th Congressional District is pretty much on target; this is the district I live in, and the one represented by Jim Cooper (one of the subjects of David Brooks's much-forwarded op-ed piece). [I read Brooks with my salt shaker within reach, but I've eaten barbecue with Cooper, I like him, and he responds cordially to my e-mails; his 2002 primary race is also the reason I'm no longer a member of EMILY's LIST (the negative ads they produced on behalf of a female competitor convinced me that I could not in good conscience continue to give them money). So I am openly biased here.]

The Republican results are harder to parse. A good chunk of Middle Tennessee went to Romney, but the only county he truly dominated was Marshall (77% of the vote, as opposed to 30-39% in the others. I'm looking at Marshall County's industry figures and I don't have a clue. McCain won a chunk of East Tennessee (around Knoxville), some counties along the Kentucky border (Ft. Campbell is near Clarksville), and three others in Middle Tennessee. The rest of the state technically went to Huckabee...

...but wait. Let's look at the percentages. Statewide, it's 34% Huckabee, 31% McCain, 24% Romney. In the border counties, McCain won in the 38%-41% range; in the Knoxville area, the range is 34%-41%, and Coffee County shows both Huckabee and McCain at 33%. Elsewhere -- there may be conclusions to be drawn from why Huckabee won by 20% plus in some counties and less than 5% in others, but I don't know enough to make 'em. That said, I'm looking at the TN GOP delegate distribution rules, and here's the thing: only the top two finishers in each congressional district count. There are nine districts in the state: in each district, the guy in first place gets two delegates, and the guy in second place gets one. Which means that while the statewide totals reflect only a 10% spread among the three candidates, the delegate distribution is not going to reflect that balance, because the congressional districts don't map out that way: at a glance, Huckabee is going to score 2 delegates for every 1 awarded to McCain, everywhere except District 5 (Nashville -- 2 for Romney, 1 for McCain) and possibly Districts 2 or 3.

This is where practice trumps principle, I think -- because, if it were a matter of theory alone, I'd prefer the Republican method of choosing delegates via public election rather than a separate caucus. I'm cynical, though, about how many folks do much research when presented with really long ballots (it'd be interesting to compare % of voters voting in the presidential race vs. % of voters who voted for one or two specific delegates vs. % of voters who voted for fifteen), and whether the results really end up all that different than party loyalists convening to vote among the faithful. More important, that 2:1 x 9 formula isn't sitting right with me; it beats winner take all, to be sure, but the way it will likely transform a 34% share of the vote into 52% of the district-based delegation is troubling...

...except then you add in the 12 at-large slots on today's ballot, the 13 delegates to be picked by the party's Executive Committee on March 1, and that's how you end up with percentages that currently look more like this.

Holy pachyderms and p-series. No wonder Tennessee Democrats claim their process is more efficient. I'm sure there's decades of reasons and/or reasoning that led to the current composion of the delegation, but wow.

AND WAIT: Overseas ballots have until February 15 to arrive in Tennessee to be counted, due to 46 counties were late in mailing them out. Huh. I guess the effect will be negligible, since the votes will likely be spread out across districts. But that's another wrinkle...

The TDP delegate selection plan is fascinating as well, but it's well past my bedtime and the sirens have stopped going off every fifteen minutes. The Green Papers projections offer food for thought, and I didn't know until tonight that the plan includes gender-, ethnicity-, and age-based recommendations (not quotas). For example, from page 28:

...the Tennessee Democratic Party has set a goal for 10% of our delegation to be young people aged 35 years or younger which translates into 9 delegates. TheAffirmative Action Committee is charged with the responsibility of fundraising to help with expenses for these young people, as well as other delegates, where needed.

The 10% reflects an estimate that the youth vote accounted for 10% of votes cast in the 2006 primary. The goals of 24 African American delegates, 2 GLBT delegates, and 3 "Hispanic/Other ethnic minority"delegates were similarly derived.

This state has a such a long, long way to go when it comes to GLBT issues, but I have to say, seeing any gesture of inclusiveness in the delegate plan -- I wasn't expecting that. (Just the other day, a friend and I were discussing how certain leaders in Nashville still regard open support of GLBT-related measures as political suicide, and there's a rabidly homophobic ex-councilwoman currently angling for interim superintendent.) I'd love for Tennessee to reach a point where we can take representation of diversity for granted, rather than having to spell it out in charts and PDFs. But we're not there yet, so I rejoice in hopeful signs like these.

[And as long as I'm on this issue: An analysis of Obama and Clinton's stances on DOMA. That right there may account for some of the Obama yard signs in my 'hood; some of the comments I've read in the past 48 hours have cited DOMA and "Don't Ask, Don't Tell" as the reasons the posters are now bitterly, fiercely anti-Clinton.]