Pollsters disagree over the importance of the so-called “generic ballot” question, which asks, “Will you vote for the Democratic or Republican candidate in your district?” The conventional wisdom is that the generic ballot heavily favors the Democrats, but the superior GOTV and general closing strategy of the Republicans has led the GOP to overperform its generic ballot result. What’s more, gerrymandering and name recognition tends to favor the incumbent.
Well, Chris Bowers at MyDD has averaged the generic ballot results for each of the last four elections and concludes that the CW is only partially right. In 2004, Dems led the generic ballot in late October by 2 points. In 2002, they led by half a point. In 2000, they led by 3.3 points (a year when they pulled about even in the Senate). And in 1998, the Republicans led by 0.3 percent (even though Democrats picked up seats that year.) All of these numbers were quite close, so GOTV may have swayed the electorate in the end.
How about this year? Democrats lead by 15.7%. Even if you assume that Republicans will shave off six points due to better GOTV, the Democrats would be poised for a historic victory along the lines of 54-46 percent. Assuming the actual seats reflect those numbers, (contrary to CW, gerrymandering tends to even itself out such that House allocations tend to match overall vote totals), the Democrats would end up with 235 seats, or a 33-seat pickup. The nonpartisan prognosticators often mention that a 30-seat pickup is very possible.
But all of that assumes the GOP will continue to out-GOTV the Dems and shave off the generic lead as in the past three elections (but not 1998). With Independents breaking strongly for the Democrats (see the post below), this GOTV shave-off is possible. Indies will break to the Dems but Republicans will still bring their voters to the polls better than the Dems.
However, if the Dems are as motivated to vote as they claim, they will show up in greater numbers than the Republicans. And this would mean that Dems would meet or exceed their generic total. How would it look if the Dems won the House with a 15 point margin? At 57.5 to 42.5, the House would have 250 Democratic seats and 185 for the Republicans. It would be a 48 seat pickup, still less than the Republicans in 1994 but the best in recent history for the Dems.
Keep these numbers in mind as we get close to the election. Political analysts love to malign the generic ballot test, but often unfairly. As pollsters adjust their likely voter models close to the election, generic ballot numbers start to look more like the actual outcome. If the 15-point margin holds up, there will be a massive shift in Washington.