Boy oh boy, did I choose the wrong day to travel.
The generic ballot polls to come out today are just absolutely mind-blowing. 21 points in CNN? 23 points in Gallup? Granted, the Gallup survey has real problems with its likely voter screen. It tends to measure partisan voter excitement, and so it will swing heavily with the latest news in any election. I don't think that is a very good way to measure the electorate, but it does show just how severely the Foley cover-up has hit the Republican base. This scandal has the potential to develop into a long-term voter retrenchment problem for the white, conservative, evangelical Republican base. A lot of them may just go back to not participating in politics altogether.
But now, for some perspective.
The seven congressional ballots conducted entirely post-Foley produce a mean of Democrats 53.7%--38.1% Republicans (seven-poll mean. Sounds pretty big, right? That is because it is. Here are some
recent national vote totals for all House elections:
- 1994: Republicans 52.4%--45.2% Democrats. Republicans won 230 seats.
- 1998: Republicans 49.1%--47.7% Democrats. Republicans won 223 seats.
- 2000: Republicans 48.3%--47.9% Democrats. Republicans won 221 seats.
- 2002: Republicans 51.0%--45.9% Democrats. Republicans won 228 seats.
- 2004: Republicans 50.1%--47.5% Democrats. Republicans won 232 seats.
It doesn't take a genius to see that the current Democratic lead of 15.6% is far larger than anything Republicans managed when they consistently won about 230 seats. However, as amazing as this all seems, it is time for some perspective on our current lead:
- 1. Republicans still control the maps. As we can see form the totals above, 2004 congressional maps were much more favorable to Republicans than were the 2002 maps, thanks to Texas redistricting. Also, the 2002 maps were much more favorable to Republicans than were the maps in the 1990's, thanks to Democrats losing many state legislatures and governorships, and thanks to improved redistricting software. Since we are facing new and improved Republican maps (including the new ones in Georgia) there is no telling just how many seats a popular vote win like this would translate into. This is not to mention that without the Gallup poll, which this far out is unreliable, the Democratic lead is actually "only" 14.3%, not 15.6%.
- 2. The Republican brand is trashed, but individual Republicans might not be. In a recent Democracy Corps generic ballot of 45 Republican-held districts (PDF), Democrats led by six on the generic ballot, but trailed by three on the named ballot. Remember that in an environment like this, where the Republican Party is itself tarnished, individual, named, incumbent Republicans will typically fare better than the generic version of the party. We will have the Incumbent Rule to help cancel out this effect, but right now we should not assume that a big generic lead means a big lead for individual Democrats facing well-known incumbents.
- 3. Democrats have consistently performed better in the generic ballot than in the final vote. Disturbingly, in all but three of the last thirty congressional elections (going back to 1946), Democrats have actually performed better in the final generic ballot than they performed in the final national vote total. While that is not a good omen, it does not mean that it is a lock that Democrats will perform worse from the final generic ballot to the final vote totals. Then again, I still firmly believe that our excellent recruiting this year will help reverse that trend (that, and I'd also wager large sums of money that Democrats are going to turn out at higher rates than Republicans this year).
- 4. Doing well in the popular vote total helps, but does not guarantee big gains in the House. This was possibly the most disturbing lesson of all to take from these two posts. It seems pretty clear that doing better in the national popular vote results in taking more seats (duh), but it is not a guarantee that even a very large popular vote win (up to 6%) will result in a large majority in the House of Representatives. The gains are going to have to come where they count the most in order for Democrats to take over. Fortunately, the only information we have on the most competitive races indicates that Democrats are indeed making significant gains where they count the most (see also the Democracy Corps poll previously linked).
Democrats look to be headed for control of the House, but these polls cannot give us a reliable estimate of how many seats Democrats will win next month. It could still be very narrow control. Further, with this much time left, there is still an opportunity for Republicans to recover and maintain narrow control. I wanted to write this lest people start discussing Democratic landslides that approach 50-60 seats, which are the sort of expectations that can only produce huge letdowns and bad post-election media narratives. Perhaps even more worrisome, over-reaching expectations can cause voter and activist retrenchment if Republicans start to close the gap in the final few weeks, and our unrealistic hopes no longer seem achievable.
Unless something truly remarkable happens in individual district polling, I will never forecast a Democratic gain anywhere near 50 seats. No one should. However, even with all of the caveats I listed above, polls like these do suggest that it is now likely that not only will Democrats win Congress, but that they could very well do so with a majority surpassing anything Republicans have put together since 1994. To pull that off, Democrats would need a 30-seat gain, making their total number of seats 233. Once we pass 218, hitting 233 seems like a pretty darn good target to me. Anything after 233 is just, well, kind of hard to believe. A thirty-seat gain would be our largest in thirty-two years, so that is already shooting pretty high. Let's not get our hopes up beyond that, because I don't see either what good it would do us, or how it is all that probable.
