In 2006 and 2008, Democrats rode a wave of voter dissatisfaction to two electoral victories — and to large leads in party affiliation.  By 2010, Republicans had the momentum of voter dissatisfaction, but not the realignment seen in the previous two cycles.  According to a new survey by Rasmussen Reports, the realignment seems to have arrived in 2012:

After falling for two straight months, the number of Americans who consider themselves Republicans jumped nearly three points in August.

During August, 37.6% of Americans considered themselves Republicans. That’s up from 34.9% in July  and 35.4% in June. It’s also the largest number of Republicans ever recorded by Rasmussen Report since monthly tracking began in November 2002. The previous peak for the GOP was 37.3% in September 2004. See History of Party Trends.

The number of Democrats slipped to 33.3% in August from 34.0% in June and July. Those two months marked the Democrats’ best showing in 2012. In February, just 32.4% described themselves as Democrats, the lowest level ever measured by Rasmussen Reports.

The number of voters not affiliated with either major political party fell to 29.2% last month. That’s down two points from 31.1% in July and the smallest number of unaffiliated voters since 2009.

Rasmussen is one of only two significant pollsters that track partisan affiliation on a regular basis.  Gallup also surveys on this question, and we’ll get to their measure in a moment.  This metric has significant bearing on poll analysis, as Hot Air readers know, but it’s not the end-all for that, either.  Pollsters have to model the likely-voter turnout, and while party affiliation measures in the general population inform that modeling process, the models aren’t a copy of these surveys, either.  This series, unlike most of Rasmussen’s surveys, poll general-population adults rather than likely voters.

Galllup’s series has not reflected the same trend Rasmussen sees, though.  In fact, there’s been a slight decline for the GOP in that series since spring among affiliated voters, although that evens out when leaners are added:

This then becomes a matter of which pollster would be more reliable on this measure.  Bear in mind that Rasmussen’s tracking on this question did a pretty good job of showing at least the direction of the turnout in 2010’s midterm elections, which nationally had a D/R/I of 35/35/30 in the tracking polls.  In this graphic, the columns are R/D/I:

Gallup missed the surge in that election, however, and had the GOP’s affiliation numbers tracking the other direction (columns are R/I/D, then R+leaners/D+leaners):

By the week of the election, the D/R/I of Gallup was 31/26/41, while Rasmussen had it almost precisely accurate at 35/36/29.  That’s something to keep in mind, and it’s been my experience with Rasmussen polling that it tends to identify trends earlier than almost any other pollster.

In this case, it shouldn’t be a big surprise to see voters shifting more Republican after three-plus years of a flatline recovery, with job creation stalled, and Democrats refusing to talk about either or talk about their own plans to fix either, for that matter.