With the traditional general election bell-ringer of Labor Day on tap Monday, for the most part it’s now past time for observers to speak in some sort of future tense about what the two parties might do to position themselves for November. Yes, eight states do still have pending primaries, along with one that still has a runoff; these contests will significantly affect at least two Senate and three gubernatorial races. Yes, national decisions still have to be made about the precise deployment of financial resources in particular races, and there are a scattering of individual candidate decisions that could fundamentally change the landscape in particular races (most notably in Colorado, where doomed-looking Republican gubernatorial nominee Don Maes is under increasing pressure to drop out). And there are, of course, mistakes that can be made on the campaign trail or in debates that could move lots of voters. In most cases, however, campaigns will focus on getting out their vote, and on appealing to a very narrow range of swing voters. This is likely to revolve around negative attacks on opponents, since that is one of the few devices that can serve as both a motivator and a persuader.

To a considerable extent, much of the “political news” you are going to hear between now and November will have as much to do with the measurement of public opinion as with efforts to influence it. This is when most pollsters switch from surveying registered to “likely” voters (though some, notably Rasmussen and Survey USA, have long been deploying “likely voter” screens, which helps account for the relative strong Republican performance in their polls). And this is when various ambitious models for predicting the results will be announced and refined.

So what should the discriminating political junkie watch for during the next two months?

With respect to the fight for control of the House, the most important objective data point will probably be the final Gallup poll generic ballot results in October. Back in 2002, political scientist Alan Abramowitz developed a very reliable model for predicting the overall results from the final Gallup generic. Just yesterday Pollster.com’s Harry Enten released a refinement of the Abramowitz model, which suggests, for example, that a five-percent GOP advantage among likely voters in the final Gallup generic ballot poll would translate into 225 Republican seats.

Senate races are a much tougher nut to crack, since they are less susceptible to national “waves.” The most influential model, by FiveThirtyEight’s Nate Silver, currently shows Democrats likely to sustain a net loss of six to seven seats, but with a 20 percent chance of losing ten or more, which would give Republican control of the upper chamber. Nate’s system also supplies a probability rating for any given Senate seat “flipping” from one party to the other; right now the range is from a 100 percent probability in North Dakota and Arkansas to zero percent probability in eleven states.

And governor’s races are even more complex, since non-national issues (including specific economic and fiscal conditions) can be a major factor, and responsibility for an unpopular status quo doesn’t always reside with Democrats. So given the national dynamics and the slow but steady trend towards partisan polarization in state as well as federal politics, it makes sense to watch closely those “red state” gubernatorial contests where Democratic candidates have managed to remain competitive so far—particularly Texas (Bill White), Florida (Alex Sink) and Georgia (Roy Barnes), three states where a Democratic win could have a major impact on congressional redistricting.

Speaking of redistricting, there are also vicious battles being fought for control of state legislative chambers. I won’t go into this in detail today (it will be the subject of a future Political Memo), but the basic situation is that Republicans have significant advantages in the political landscape while Democrats appear to be a bit better focused and financed.

Finally, it’s important to keep in mind that the results of any given election cycle are usually over-emphasized. Obviously the Democratic “wave” elections of 2006 and 2008 have not, at least in the short-term, led to any sort of realignment of our political system. It works both ways, though: after the last redistricting cycle, it was generally thought that Republicans had won a “lock” on control of the House until 2012; it didn’t turn out to be much of a lock. But Democrats do need to relearn the basic problem that our constitutional system builds a conservative bias into the composition of the U.S. Senate, and the balance of control among governorships and state legislatures. But by the same token, Republicans need to understand that the gains they make this year need to be contextualized by their terrible performance in 2006 and 2008, and that the shape of the electorate in the presidential cycle of 2012 will be very different, and much less friendly to the GOP cause.

Ed Kilgore’s PPI Political Memo runs every Tuesday and Friday.

Photo Credit: OZinOH’s Photostream