Map of Virginia counties and cities
It's election day in a number of places around the country; as is always the case in a year after a presidential election, the main events are the gubernatorial races in Virginia and New Jersey, and the mayoral contest in New York City. However, these races, often the source of drama, and often a harbinger of what happens in the following year's midterm elections, simply aren't on track to be close. New Jersey is poised to re-elect Republican Governor Chris Christie by a margin of at least 20 points, while New York City is poised to elect its first Democratic mayor in decades, Bill de Blasio, by a margin of ... I'm not even sure I can count that high.
Of that trio, the open seat Virginia gubernatorial race is the least set-in-stone, although even it looks like a likely victory for Democratic candidate Terry McAuliffe. Republican opponent Ken Cuccinelli hasn't led a single poll since July, and poll aggregators put McAuliffe's lead in the high single digits. HuffPo Pollster, for instance, puts it the projected margin at 7.4%, and gives McAuliffe a 100% chance of winning. (Why 100%? A 7.4% margin in one poll can be a little dubious, but when it's spread out over literally dozens of polls, the margin of error dwindles, and really the only possible source for an error would be an unprecedented level of across-the-boards methodological fail.)
Nevertheless, let's take a look at the county-level benchmarks for this race, since it's probably the one you'll be watching the closest. In fact, the initially reported results in this race may be deceptive, in that the small rural counties tend to report earlier in Virginia, with the metropolitan areas where most of the Democratic votes are located reporting later. (If you think back to 2012, you'll recall Mitt Romney and George Allen leading for a fairly long period of time, until they collapsed at the very end as Democratic votes were added to the heap.)
By breaking down the results county-by-county, and keeping in mind what percentage McAuliffe needs to get in each major county in order to hit at least 50% overall, you'll have a much better sense of how the race is going than if you just look at the topline numbers. (I'm only including the counties and independent cities that, individually, make up 2% or more of the state's total vote; there's no need to keep track of how the candidates are doing in counties with only a few thousand people in them. Taken together, they account for about half of the state's votes.)
County |
% of 2012
vote in-district |
Obama/Romney
vote share
in 2012 |
What McAuliffe
needs to
hit 50% |
% of 2009 vote |
Deeds/McDonnell
vote share
in 2009 |
What McAuliffe
needs to
hit 50% |
Statewide |
100.0 |
51/47 |
50/48 |
100.0 |
41/59 |
50/50 |
Fairfax |
13.7 |
60/39 |
59/40 |
13.8 |
49/51 |
58/41 |
Virginia Beach |
5.1 |
48/50 |
47/51 |
5.1 |
36/64 |
45/55 |
Prince William |
4.7 |
57/41 |
56/42 |
3.8 |
41/59 |
50/50 |
Chesterfield |
4.4 |
45/53 |
44/54 |
4.5 |
34/66 |
43/57 |
Henrico |
4.2 |
55/43 |
54/44 |
4.4 |
44/56 |
53/47 |
Loudoun |
4.2 |
52/47 |
51/48 |
3.3 |
39/61 |
48/52 |
Arlington |
3.1 |
69/29 |
68/30 |
2.8 |
66/34 |
75/25 |
Chesapeake |
2.9 |
50/49 |
49/50 |
2.7 |
40/60 |
49/51 |
Richmond |
2.5 |
78/21 |
77/22 |
2.3 |
69/31 |
78/22 |
Norfolk |
2.3 |
72/27 |
71/28 |
2.0 |
60/40 |
69/31 |
Newport News |
2.1 |
64/34 |
63/35 |
1.9 |
50/50 |
59/41 |
I've added a wrinkle to this particular set of benchmarks, by using two different election models: a high-turnout model (the 2012 election, where the Obama campaign surprised many observers by keeping pace with the allegedly once-in-a-lifetime turnout from 2008), and a low-turnout model (the terrible 2009 gubernatorial election, where Democratic candidate Creigh Deeds failed to arouse any excitement). As you can see, though, the two models still require very similar percentages in the various counties, with a few notable exceptions: there was little falloff among the reliable Dem voters of affluent Arlington, helping Deeds to overperform there, while further out in the northern Virginia exurbs, Prince William County saw a particularly stark Deeds underpeformance (though it's hard to tell whether that's because of steep falloffs among irregular Dem voters — Prince William has rapidly grown, with most of that increase among Hispanic and Asian voters — or because of a sharp turnaround among swing voters, which Prince William also has a lot of).
There's one other caveat you should be aware of: neither 2012 nor 2009 were characterized by a lot of third-party votes, but Libertarian candidate Robert Sarvis has been polling surprisingly well, sometimes in the low double digits. As a result, McAuliffe won't actually need to hit precisely 50% to win ... but it's hard to predict just how big an effect Sarvis will have. Third-party candidates rarely come anywhere near their poll numbers, and PPP's final poll of the race captured that, finding a late collapse in Sarvis's numbers. So you'll need to make some minor mental adjustments to the benchmarks, based on what percentage Sarvis is putting up in early results (and whether Sarvis's votes seem to be coming out of Cuccinelli's tally or are more evenly spread).
So how did Virginia get to this place, where a formerly red state is poised to elect a Democratic candidate who has "issues," and do it by a significant margin? Part of that is due to the fundamental crappiness of the Cuccinelli product, part of that is due to a well-coordinated Democratic effort ... and part of that is due to the massive demographic shifts occurring in Virginia. Over the fold, we'll look at how these counties that I mentioned have changed...
Although Barack Obama winning Virginia in 2008 was quite the turning point, it wasn't entirely out of the blue. Even as John Kerry was badly losing the state overall in 2004, you could see significant progress for Democrats in the parts of the state that were actually growing, compared with where they were in the 1990s or 1980s. (Obama's analytics people no doubt noticed that, and that probably factored in greatly in the decision to compete in Virginia, something that Democrats hadn't tried in many years.)
To illustrate that more clearly, I've put together a chart of how the presidential vote has evolved in the major counties and independent cities of Virginia over the last three decades. Each year's vote is expressed as both percentages (the Democratic candidate's percentage is on the left of the slash, and the Republican is on the right), and more usefully, as Partisan Voting Index, which calculates that county's deviation from the national average.
County |
2012 % |
2012 PVI |
% of state |
2008 % |
2008 PVI |
% of state |
2004 % |
2004 PVI |
% of state |
Statewide |
51/47 |
R+0 |
100.0 |
53/46 |
R+0 |
100.0 |
45/54 |
R+3 |
100.0 |
Fairfax |
60/39 |
D+8 |
13.7 |
60/39 |
D+7 |
13.9 |
53/46 |
D+5 |
14.4 |
Virginia Beach |
48/50 |
R+3 |
5.1 |
49/50 |
R+4 |
5.4 |
40/59 |
R+8 |
5.5 |
Prince William |
57/41 |
D+6 |
4.7 |
58/42 |
D+4 |
4.4 |
46/53 |
R+2 |
4.1 |
Chesterfield |
45/53 |
R+6 |
4.4 |
46/53 |
R+7 |
4.4 |
37/63 |
R+12 |
4.1 |
Henrico |
55/43 |
D+4 |
4.2 |
56/43 |
D+2 |
4.2 |
46/54 |
R+3 |
4.2 |
Loudoun |
52/47 |
D+0 |
4.2 |
54/45 |
D+0 |
3.7 |
44/56 |
R+5 |
3.4 |
Arlington |
69/29 |
D+18 |
3.1 |
72/27 |
D+19 |
3.0 |
68/31 |
D+20 |
3.0 |
Chesapeake |
50/49 |
R+1 |
2.9 |
50/49 |
R+3 |
2.9 |
42/57 |
R+6 |
2.9 |
Richmond |
78/21 |
D+27 |
2.5 |
79/20 |
D+26 |
2.5 |
70/29 |
D+22 |
2.3 |
Norfolk |
72/27 |
D+21 |
2.3 |
71/28 |
D+18 |
2.4 |
62/37 |
D+13 |
2.2 |
Newport News |
64/34 |
D+13 |
2.1 |
64/35 |
D+11 |
2.2 |
52/47 |
D+4 |
2.1 |
Balance |
45/53 |
R+6 |
50.8 |
47/52 |
R+6 |
51.0 |
41/58 |
R+7 |
51.7 |
County |
2000 % |
2000 PVI |
% of state |
1996 % |
1996 PVI |
% of state |
1992 % |
1992 PVI |
% of state |
Statewide |
44/52 |
R+4 |
100.0 |
45/47 |
R+6 |
100.0 |
41/45 |
R+6 |
100.0 |
Fairfax |
47/49 |
R+1 |
15.1 |
47/48 |
R+6 |
15.1 |
42/44 |
R+5 |
15.1 |
Virginia Beach |
42/56 |
R+7 |
5.5 |
41/51 |
R+10 |
5.2 |
32/50 |
R+14 |
5.4 |
Prince William |
45/53 |
R+4 |
3.7 |
43/50 |
R+9 |
3.2 |
35/47 |
R+11 |
3.0 |
Chesterfield |
35/63 |
R+15 |
4.1 |
32/61 |
R+20 |
3.9 |
27/56 |
R+20 |
4.0 |
Henrico |
43/55 |
R+7 |
4.2 |
40/53 |
R+12 |
4.2 |
34/52 |
R+14 |
4.3 |
Loudoun |
41/56 |
R+8 |
2.8 |
40/52 |
R+11 |
2.0 |
35/46 |
R+11 |
1.6 |
Arlington |
60/34 |
D+14 |
3.1 |
60/35 |
D+9 |
3.1 |
58/32 |
D+11 |
3.2 |
Chesapeake |
45/53 |
R+4 |
2.7 |
46/47 |
R+5 |
2.6 |
38/47 |
R+9 |
2.4 |
Richmond |
65/31 |
D+18 |
2.4 |
63/31 |
D+12 |
2.8 |
60/31 |
D+13 |
3.1 |
Norfolk |
62/35 |
D+13 |
2.3 |
63/31 |
D+12 |
2.5 |
54/32 |
D+9 |
2.7 |
Newport News |
51/47 |
D+2 |
2.1 |
51/42 |
R+0 |
2.2 |
42/44 |
R+4 |
2.4 |
Balance |
42/55 |
R+7 |
52.0 |
44/47 |
R+7 |
53.2 |
40/45 |
R+6 |
52.8 |
County |
1988 % |
1988 PVI |
% of state |
1984 % |
1984 PVI |
% of state |
Statewide |
39/60 |
R+6 |
100.0 |
37/62 |
R+4 |
100.0 |
Fairfax |
38/61 |
R+8 |
15.0 |
37/63 |
R+4 |
13.6 |
Virginia Beach |
30/64 |
R+15 |
5.1 |
25/74 |
R+15 |
4.5 |
Prince William |
32/67 |
R+13 |
2.7 |
31/69 |
R+10 |
2.4 |
Chesterfield |
24/75 |
R+22 |
3.6 |
20/80 |
R+21 |
3.2 |
Henrico |
30/69 |
R+16 |
4.1 |
25/75 |
R+16 |
4.0 |
Loudoun |
33/66 |
R+13 |
1.4 |
31/68 |
R+9 |
1.2 |
Arlington |
53/45 |
D+8 |
3.4 |
51/48 |
D+11 |
3.4 |
Chesapeake |
39/61 |
R+7 |
2.2 |
37/62 |
R+3 |
2.1 |
Richmond |
56/42 |
D+11 |
3.4 |
56/44 |
D+15 |
4.1 |
Norfolk |
55/44 |
D+9 |
3.1 |
52/48 |
D+11 |
3.5 |
Newport News |
39/60 |
R+6 |
2.5 |
39/60 |
R+1 |
2.6 |
Balance |
39/59 |
R+6 |
53.5 |
37/62 |
R+4 |
55.4 |
Before compiling these charts, I would have guessed that the big story would be substantial growth in the suburbs and exurbs of northern Virginia; however, that's not really the case, as each county stayed pretty consistently at the same percentage of the state's total votes. For instance, the state's most populous county, Fairfax, represented 13.6% of the state's votes in 1984, and 13.7% of the state's votes in 2012.
The only places where there was significant movement was Prince William and Loudoun Counties, the more exurban counties further outside of Washington DC; Prince William went from 2.4% of the state's vote in 1984 to 4.7% in 2012, and Loudoun went from 1.2% to 4.2%. The flipside of that is what happened to the "balance," i.e. "Real Virginia," as the McCain campaign so charmingly put it... in other words, all the other counties and cities that individually make up less than 2% of the state. The balance does contain some medium-sized locales that either have large African-American populations (Portsmouth, Petersburg) or large concentrations of highly educated voters (Alexandria, Charlottesville), but for the most part, the balance is more rural, whiter, and more downscale than the large cities and counties. And it's also declining as an overall share of the state: from 55.4% of the state's votes in 1984, down to 50.8% of the state in 2012.
Instead, the main story is how much the large cities and counties moved in the Democratic direction as they diversified. For instance, Fairfax County, one of the nation's most archetypal affluent, well-educated suburbs, went from R+8 in 1988 to D+8 in 2012. The changes are even more profound further out, as Prince William County went from R+13 in 1988 to D+6 in 2012 — and also in Richmond's suburbs, with Henrico County going from R+16 in 1988 to D+4 in 2012 and Chesterfield County moving from a beet-red R+22 to an almost-swingy R+6. In fact, every one of the 11 large jurisdictions moved sharply in the Democratic direction (with the always-reliably-Democratic Arlington — just across the river from Washington and full of federal employees — and reddish Chesapeake — the most suburban part of the Hampton Roads metro area — moving the least).
Compare that with the "balance," which has stayed put in the same narrow band over the decades: R+4 in 1984 to R+6 in 2012. Almost all of Virginia's progress, instead, occurred in the large jurisdictions ... which, not coincidentally, are becoming a larger and larger percentage of the state's votes.
I mentioned earlier that those 11 large jurisdictions were also the parts of the state that are rapidly becoming more diverse, so here's one more table that puts that into sharper relief, showing the evolution of those counties during the 1990, 2000, and 2010 censuses. The "Races" number reflects racial composition, showing Non-Hispanic White/African-American/Asian/Hispanic percentages; the "Educ." number is the percentage of persons 25 or more with a bachelor's degree or more.
County |
2010 Races |
2010 Educ. |
% of state |
2000 Races |
2000 Educ. |
% of state |
1990 Races |
1990 Educ. |
% of state |
Statewide |
65/19/5/8 |
33.8 |
100 |
70/19/4/5 |
29.5 |
100 |
76/19/2/3 |
24.5 |
100 |
Fairfax |
55/9/17/16 |
58.0 |
13.6 |
64/8/13/11 |
54.7 |
13.7 |
77/8/8/6 |
49.0 |
13.2 |
Virginia Beach |
64/19/6/7 |
31.9 |
5.5 |
69/19/5/4 |
28.1 |
6.0 |
79/14/4/3 |
25.5 |
6.4 |
Prince William |
49/20/7/20 |
37.5 |
5.0 |
65/18/4/10 |
31.5 |
4.0 |
81/11/3/4 |
27.6 |
3.4 |
Chesterfield |
65/22/3/7 |
35.7 |
4.0 |
75/18/2/3 |
32.6 |
3.7 |
84/13/2/1 |
29.2 |
3.4 |
Henrico |
57/29/7/5 |
38.9 |
3.8 |
68/25/4/2 |
34.9 |
3.7 |
77/20/2/1 |
28.0 |
3.5 |
Loudoun |
62/7/15/12 |
57.2 |
3.9 |
80/7/5/6 |
47.2 |
2.4 |
88/7/2/3 |
32.7 |
1.4 |
Arlington |
64/8/10/15 |
70.1 |
2.6 |
60/9/9/19 |
60.2 |
2.7 |
69/10/7/14 |
52.3 |
2.8 |
Chesapeake |
60/29/3/4 |
27.9 |
2.8 |
66/28/2/2 |
24.7 |
2.8 |
70/27/1/1 |
16.9 |
2.5 |
Richmond |
39/50/2/6 |
32.6 |
2.6 |
38/57/1/3 |
29.4 |
2.8 |
43/55/1/1 |
24.2 |
3.3 |
Norfolk |
44/42/3/7 |
23.7 |
3.0 |
47/44/3/4 |
19.6 |
3.3 |
56/39/2/3 |
16.8 |
4.2 |
Newport News |
46/40/3/8 |
23.5 |
2.3 |
52/39/2/4 |
19.9 |
2.5 |
61/33/2/3 |
18.4 |
2.7 |
Balance |
74/18/2/5 |
24.6 |
51.0 |
77/17/1/2 |
20.9 |
50.3 |
79/18/1/1 |
16.8 |
53.2 |
The state in general has become more diverse, with a similar composition and pace as the United States as a whole, going from 76% white in 1990 to 65% white in 2010 (though the black population has stayed the same at 19%; the gains all came among the Asian and Hispanic populations). But notice how the large jurisdictions got much less white much faster; Fairfax, for instance, went from 77% white to 55% white in the same period. It's even faster in Prince William (81% in 1990, 49% in 2010) and Loudoun (88% in 1990, 62% in 2010). (The least movement was in Arlington, which, in fact, got whiter in the period between 2000 and 2010, perhaps in large part because it also got much more expensive.)
Contrast that with the "balance," which did become more diverse over the decades, but at a rate significantly slower than the state as a whole: it went from 79% white in 1990, to 77% white in 2000, to 74% white in 2010. And notice how the places that became more diverse more quickly also were the ones that became more Democratic more quickly (especially Prince William and Loudoun Counties in the DC exurbs, and Henrico and Chesterfield Counties in Richmond's suburbs), while the places that became diverse more slowly were also the ones that moved the least, politically (Arlington, Chesapeake, and, of course, the "balance" in the rest of the state).
The other part of the puzzle is that Virginia became significantly better-educated over the decades. The major cities and counties are much better-educated than the "balance" of the state (with the exception of some of the more blue-collar parts of the Hampton Roads area, like Norfolk and Newport News), but the rate of change in education is fairly proportionate throughout the state.
For instance, statewide, the trend was from 24.5% college-educated in 1990 to 33.8% in 2010, while both Fairfax (from 49.0% to 58.0%) and the "balance" (from 16.8% to 24.6%) evolved at a similar pace. The exception is the more upscale parts of northern Virginia, which became even more heavily-concentrated with the highly-educated; Arlington went from 52.3% in 1990 to 70.1% in 2010, while Loudoun County went from 32.7% to 57.2%.
Demographics isn't entirely destiny. Campaign quality still matters, and being a good ideological fit for the place where you're running is essential. Considering how ham-handed the Cuccinelli campaign has been, it's quite possible Terry McAuliffe would be winning handily even if Virginia weren't dramatically evolving. However, where the population is more diverse and better-educated, it's more likely that Democrats are going to win — and Virginia is heading particularly rapidly in both those directions. Certainly it helps that the GOP offered up a broadly unacceptable candidate, but the groundwork for the likely Democratic victory in Virginia has been building for decades.