Misguided Perriello victory memes

Chris Cillizza yesterday offered his list of top congressional campaigns for 2008, and it included our Perriello vs. Goode campaign in VA-5. Earlier in the week, our race graced Politico's list of election upsets. While I am elated and thankful that our work is honored by both of these lists, I am a little upset that our victory is explained in facile and erroneous terms. To their defense, these pundits have to look at 435 congressional races, so their "on-the-ground" knowledge of the district is inherently abstract and distant. I feel, however, that if I don't correct some of these misguided Perriello victory memes, later pundits will dutifully regurgitate these misconceptions.

After Cillizza offers a couple particular strengths of our campaign, he concludes by musing, "Would he have been able to pull the narrow -- 745 votes [727 after recount] -- upset without Obama leading the ticket? Probably not." In essence, Tom would not have won without Obama's coattails. I am not a statistics expert, and I am not particularly sure how to statistically measure coattails, as this is a challenging and nebulus argument to prove. On the quick and easy, Perriello outperformed Obama within the district, both in total votes (158.810 v 157,362) and in percentages (50.08% v 48.29%) - keep in mind the inevitable voter drop-off from presidential races to congressional races. If you want to dissect these numbers per locality, we outperformed in total votes in 10 of our 22 localities. Again, total votes includes inevitable voter drop-off, so if you look at vote percentages per locality, we outperformed Obama in 15 of the 22 localities. To conclude that Obama's coattails pulled Tom to victory misjudges the actual data, and at the same time, it belittles the hard work and dedication our staff and volunteers put into this race. Please hear me when I say that the Obama field organization was absolutely top-notched throughout our district, but here, I am arguing against a misguided meme, not the Obama field team or their work.

County/City	Perriello total votes	Perriello % 2008	Obama total votes	Obama % 2008
Albemarle	31827	63.3%	29792	58.4%
Appomattox	2758	36.4%	2641	34.6%
Bedford	7124	38.3%	6298	33.0%
Bedford (City)	1316	49.7%	1208	44.2%
Brunswick	3720	62.3%	3985	63.9%
Buckingham	3446	50.2%	3489	49.9%
Campbell	8837	35.6%	8091	31.3%
Charlotte	2596	44.1%	2705	43.9%
Charlottesville (City)	15909	80.8%	15705	78.4%
Cumberland	2195	48.4%	2255	47.7%
Danville (City)	11487	58.0%	12352	59.1%
Fluvanna	6564	52.3%	6185	48.6%
Franklin	9475	37.7%	9618	37.9%
Greene	3733	46.0%	3174	38.4%
Halifax	7528	46.8%	8126	48.2%
Henry	6846	42.9%	6862	42.6%
Lunenburg	2737	49.5%	2703	47.8%
Martinsville (City)	3974	61.2%	4139	63.5%
Mecklenburg	6454	44.6%	7127	47.3%
Nelson	4562	56.3%	4391	54.0%
Pittsylvania	11025	37.7%	11415	37.5%
Prince Edward	4697	53.8%	5101	54.3%
Total	158810	50.1%	157362	48.3%

Politico's justification is just as problematic, but easier to data-crunch. They said:

When all the ballots were cast, counted and then recounted, Perriello’s strong margin among the progressive-minded university community around Charlottesville helped propel him to victory by less than 1,000 votes.

We absolutely burned the city of Charlottesville; we took names and numbers, both literally and figuratively. When you put this in a district-wide context, however, the claim that Charlottesville/UVA alone pulled us through is weak - let's not squabble over the semantics of the word 'helped' here please. Al Weed, the two-time Democratic congressional candidate, got roughly 40% of the vote in 2006 compared to Goode's 59%. In effect, we needed to gain 10% of the vote in all parts of the district to win the race, which of course, in a zero-sum context would subtract 10% from Goode's performance. Compared to 2006, we made percentage gains in every locality, with Nelson County making the least gains with a 5.4% increase. Importantly, we made 10% gains or more in 13 of the 22 localities, and the gains made in these localities covered the areas where we came up short. Where did we make the most percentage gains? Bedford City 12.1%; Lunenburg 12.2%; Prince Edward 12.4%; Cumberland 12.8%; Henry 13.5%; Danville 14.1%; Martinsville 20.8%! Notice something here geography wizards? That's right, the highest percentage gains were made in the Southside. Sure, at a certain point Charlottesville numbers must hit a ceiling as large percentage gains aren't possible there: 74.6% in 2006 to 80.8% in 2008. The overall argument, however, can easily be made that the Southside carried us to victory, not the Charlottesville area (Martinsville, anyone?).

County/City	Weed % 2006	Perriello % 2008	2006/2008 Net % Gain
Albemarle	54.3%	63.3%	9.0%
Appomattox	29.0%	36.4%	7.4%
Bedford	28.3%	38.3%	10.0%
Bedford (City)	37.6%	49.7%	12.1%
Brunswick	53.3%	62.3%	9.0%
Buckingham	40.0%	50.2%	10.2%
Campbell	27.4%	35.6%	8.2%
Charlotte	33.8%	44.1%	10.3%
Charlottesville (City)	74.6%	80.8%	6.2%
Cumberland	35.6%	48.4%	12.8%
Danville (City)	43.9%	58.0%	14.1%
Fluvanna	41.8%	52.3%	10.5%
Franklin	28.1%	37.7%	9.6%
Greene	34.6%	46.0%	11.4%
Halifax	35.7%	46.8%	11.1%
Henry	29.4%	42.9%	13.5%
Lunenburg	37.3%	49.5%	12.2%
Martinsville (City)	40.4%	61.2%	20.8%
Mecklenburg	35.2%	44.6%	9.4%
Nelson	50.9%	56.3%	5.4%
Pittsylvania	29.5%	37.7%	8.2%
Prince Edward	41.4%	53.8%	12.4%
Total	39.9%	50.1%	10.2%

The truth of the matter is, with a difference of 727 votes, you can't really pinpoint any one thing or one area as the silver bullet in this congressional race. All parts came together in the perfect storm. These stats, to me, show that a national (and even local) understandings of this race contain superficial analysis. With that in mind, and with my in-the-trenches perspective, I can offer a few macro-level observations:
1.) While the 2006 numbers on their face once seemed daunting, Goode's district-wide support was soft. People voted for Goode out of habit, and past candidates could not penetrate that practice. When presented with a candidate that voters deemed viable - when presented with a legitimate choice - past Goode voters flocked to Tom. In the eyes of the voters, Tom wasn't just a regular Democratic candidate, Tom was an electable, worthy-of-office candidate.
2.) Tom got to define himself before Goode could. Usually the incumbent has the ability to define the challenger and the challenger has difficulty shedding that portrait. In this race, however, Tom traveled throughout the district meeting with thousands of voters before Goode took notice. Tom's defined himself first as a hard-working, solutions-oriented man, which deflated Goode's "Charlottesville liberal" and "New York lawyer" characterizations.
3.) Similarly, Tom's positive, solutions-oriented message was comforting, while Goode's negative attacks were off-putting. For example, Tom offered positive solutions to the jobs and economic crisis while Goode scapegoated anchor babies. The contrasting messages were nowhere more evident than in the debates, where several people went in with Goode stickers on their lapels and left with Perriello stickers on their cars. Also, Tom's positive television ads were a tipping-point in this election.
4.) Tom showed up and ceded no voter. Tom traveled the district, covering hundreds of miles every day. He never slept and never took a day off. He met with voters everywhere, whenever possible. Places like Martinsville never saw so much attention by a congressional candidate. He showed up to Goode's hometurf of Franklin County almost weekly. Structurally speaking, Tom focused on building a large district wide grassroots network, with an emphasis on the Southside. Our campaign opened eight district offices, seven of which were located in the Southside. This gave us the ability to reach every voter in every corner of the district.
5.) Tom's exceptional fundraising prowess provided the financial stability to achieve all of the above. Tom raised a cool $1.8 million on this race, compared to the $607K Weed was able to raise in 2006. Money raises the ceiling on the number of voters a campaign can contact, and Perriello was able to go toe-to-toe with Goode.
6.) Tom's staff and volunteers never stopped believing. Even when polls showed us down 34 points three months out, we weren't alarmed; we knew that the polls would close and that we were always in the game. Once the ads went up, the race tightened significantly. And all the while, we faithfully continued our ambitious strategy.