Today Stan Greenberg and Matt Hogan of Democracy Corps released a strategy memo for 2006 built around the results of a post-election survey (pdf) from the 2005 Virginia Governor's race. Dem Corps interviewed 2,300 Virginians last fall, including 600 registered voters who did not turn out to vote. Here are the key findings:

* Failure to mobilize the Republican base doomed Kilgore. The demoralization of Bush voters and lack of enthusiasm for Kilgore seriously hindered the Republican candidate's chances, underscoring the dangers of taking the base for granted.

* Non-voters were disillusioned with Bush, unimpressed by Kilgore. Those who voted in 2004 but not in 2005 overwhelmingly supported Bush over Kerry, but mounting frustration with Bush and a lack of fondness for Kilgore prevented them from turning out. While many incumbents will be better received by voters than Kilgore, the impact of the disillusionment with Bush highlights why it is so important for progressives to tie incumbents to Bush.

* Positive agenda was crucial to winning over swing voters. While Kilgore alienated voters with his attacks on Kaine's position on the death penalty, Kaine reaped the benefits by focusing on education, an issue that was of particular importance to both his base and swing voters.

* Republicans voter outreach program is not to be underestimated. Although Kilgore's turnout effort came up short, his campaign was much more effective at contacting both base and swing voters, as well as those who were still undecided in the final days of the campaign.

* Essential to have sufficient resources for the final few weeks of the campaign. Nearly one in five voters did not to decide who to vote for until the last few days of the campaign and 40 percent held off until October.

Regarding this last point, see Perry Bacon, Jr. in Time.com. Democrats clearly have the edge in enthusiasm and the political wind at their backs, but Republicans have a slight edge in money and a generally superior ground game. Whether that will be enough to save the GOP majority or simply mitigate the size of the Dem wave in November remains to be seen. As I said the other day, the second half of the game doesn't even start until next Tuesday.

UPDATE: More fodder on battle for the House from Robert Novak (via Drudge):

If The Election Were Held Today: To date, we have discussed this election in terms of what the final outcome will look like in November. We have also mentioned Republican fears that, as one House committee chairman has said privately, Republicans will lose 25 seats -- or as we were told that national internal polls suggested, they could lose as many as 26 seats.

From here in, now that primary season has approached its end, we will resist such broad prognostication, particularly since we have not yet seen evidence that such huge losses are imminent when looking at the races as we always have in past cycles -- on a district-by-district basis. As we noted last week, "it is still at least challenging to construct a scenario of a 15-seat Democratic gain without positing some improbable upsets."

The emphasis, by the way, is in the original.