With only a few exceptions, almost all the close races will be in suburbs—particularly those in Houston, Dallas and Austin. As we wrote in our elections issue, the suburbs are changing drastically. Formerly all white towns are now giving way to large minority communities. The election results will go a long way towards telling us just how engaged those communities are. The Democrats have been saying for weeks that their field operations are better than pundits realize. It’s hard to see major turnout from the left, since its voters are so unenthusiastic this year, but it’s possible. Furthermore, we may not have seen those voters in the polls, since minority voters are less likely to have voted in the past, and therefore often do not get counted among polling’s likely voters.
Which brings me to point two….
Polling is not an exact science. As We’ve been seeing numbers favoring the Republicans by large margins, but, as Nate Silver notes at fivethirtyeight polls can often favor Republicans. Polls like Rasmussen rely on “robocalls” to poll people—effectively asking voters to “press one for white, two for perry.” Such methods tend to benefit Republican candidates.
Secondly, as I’ve written in the past, most of the polls we’ve seen use “likely voter” models. These models help candidates with an enthusiastic base—the Republicans this year. If the Democrats have convinced unexpected people to come out after all, we might see a bunch of “unlikely voters” swing the results.
For a better understand of polling problems, check out our polling survival guide.
Domino Effect. There are a whole lot of “semi-competitive races” this year—basically meaning races we can’t really predict. As precincts report their numbers, we’ll be able to gauge the “semi-competitives” by those that surround them. For instance, the race between state Rep. Diana Maldonado and her Republican opponent Larry Gonzales is probably the closest fight in the Austin area. If we see Maldonado losing by large margins, that’s not just bad for Round Rock Democrats. It also tells us that the predicted Republican wave may well be materializing. In that case, we can then look to the race between Democratic state Rep. Valinda Bolton and GOP challenger Paul Workman. If the Bolton campaign also shows signs of losing, then look to the surrounding Democratic incumbents like—state Reps. Donna Howard and Mark Strama.
On the other hand, if the numbers show Maldonado beating Gonzales at all, it means these other Democratic incumbents are almost certainly safe.