Vice President Joe Biden has few doubts about the outcome of this fall’s midterm elections. Biden had this to say last week to Democratic party leaders in St. Louis -- according to The New York Times:
“The reports of the death of the Democratic Party have been greatly exaggerated,” Mr. Biden said, paraphrasing Mark Twain as he addressed party leaders here. “The day after the election, there will be a Democratic majority in the House and a Democratic majority in the Senate. If it weren’t illegal, I’d make book on it.”
To be sure, Biden sets the bar fairly low here. He didn’t opine that the Democrats would gain seats or keep their current seats; he more cautiously prophesied that the Democrats would not lose control of the House or the Senate.
Given that the Democrats currently control 255 House seats out of the overall total of 435, it’s clear that the Democrats can lose a lot of seats this fall and still maintain control of the House. Biden’s bet will pay off even with a fairly miserable performance by the Democrats -- as long as that miserable performance is just not quite miserable enough to push Democrats to the turnover point.
Still, the probability that the Republicans can win enough seats to end up controlling the House is well above zero. Biden’s bet is certainly not a sure thing based on all available evidence. Nothing that I see in the data we monitor on a daily and weekly basis here at Gallup appears to indicate a dramatic turnaround in the dynamics of this race. If anything, the trends appear to be at least slightly in the GOP’s favor.
As noted recently by my colleague Jeff Jones, we have now seen four weeks in a row in which the Republicans have held an edge among registered voters on the generic ballot. This is the first time this has happened since we began tracking the generic ballot in March.
Any edge for the GOP among registered voters has unique danger signs for Democrats, because Republicans have a high probability (although not absolute) of turning out in higher numbers than Democrats on Election Day.
That’s exactly what happened in 1994. In the last three polls conducted by Gallup before Election Day, the Republican leads over Democrats among likely voters were higher than the leads among registered voters by margins of 9, 13, and 7 percentage points. To be specific, the registered voter gaps were +3 Republican, + 3 Democrat, and dead-even in Gallup's final three polls in 1994. The likely voter gaps were +12, +10, and +7 Republican. The last such numbers are from Gallup’s final pre-election poll, in which Republicans and Democrats were at 46% each among registered voters, but in which Republicans were ahead by 51% to 44% among likely voters.
One can get a feel for the type of gains Republicans could make this fall by extrapolating these same likely voter jumps to this year's numbers. If Republicans manage to maintain a lead on the generic ballot among registered voters, their margin among likely voters -- again in theory -- could reach double digits. Keep in mind that any likely voter margin for the GOP on the generic ballot over about four points historically would translate into the real possibility of Republicans gaining enough seats to control the House. A 10-point margin among likely voters would, in theory, translate into a 10-point margin in the national vote for Congress (that is, when the vote in all of the districts is combined together) and that, in turn, would translate (forgetting undecideds for the moment) into something like a 55% to 45% GOP popular vote margin, the largest Republican gap in recent American history.
Two things to keep in mind. First, the Democrats moved out ahead of the Republicans for three weeks in our generic ballot tracking earlier this summer, suggesting that it is possible that they could do so again. There is no guarantee that the Republicans will be able to maintain their lead all the way through to Election Day. Second, we don’t know for sure yet if the same type of registered voter to likely voter differential we saw in 1994 will happen again this year. By way of example, in 2006, Gallup’s final pre-election poll showed an 11-point Democratic advantage among registered voters, and a 7-point Democratic advantage among likely voters, a more modest 4-point swing. Still, we don’t have examples in any recent midterm elections in which the Republicans did not gain at least a few points in the transition to likely voters from registered voters.