Excerpts from “Democrats Just Took the Lead In This Poll Average and No One Noticed” by Bill Scher at The Washington Monthly:
In last Tuesday’s newsletter, charting Joe Biden’s clearest Electoral College path, I noted his recent improvement in national and some swing state polling averages from Real Clear Politics and FiveThirtyEight since March.
While Biden has made modest gains, he still doesn’t lead in any of those averages.
However, there is another poll average where, on April 22, Democrats took the lead for the first time in five months:
The Real Clear Politics generic congressional ballot test average.
Generic congressional ballot tests are poll questions that ask which party’s candidate you would choose to represent your congressional district.
Scher asks, “Does this mean Democrats are well-positioned to take back the House?,” and answers:
In the Real Clear Politics generic congressional ballot test average, Republicans led their widest lead of the year, 2.6 points, on March 6, just before Biden’s State of the Union address. As of May 7, Democrats now lead by 1.4 points.
A slightly less dramatic but similar story is told in the FiveThirtyEight average, with a Republican lead around 1 for most of January and February, Democrats tying at the end of February, followed by a series of tiny lead changes. As of May 1, the last reported result, Democrats are up 0.7 points.
Should Democrats feel good at all about such a small lead? Doesn’t gerrymandering favor Republicans so much that Democrats need a big polling lead—and big national popular vote lead on Election Day—to take back the House?
Not so.
Scher continues, “Back in 2021, for the Monthly, I wrote that “When Democrats won control of the House in 2018 and 2020, their share of the popular vote (53.4 percent and 50.3 percent, respectively) was roughly equal to their share of the House seats (54 percent and 51 percent, respectively).” Further,
The 2022 midterm House national popular vote also tracked the House seat share. Republicans won 50.6 percent of the popular vote and 51 percent of the seats.
Of course, a tiny lead within the margin of error six months before Election Day tells us nothing about the final outcome beyond the necessity for determined get-out-the-vote efforts.
But if Democrats maintained a 1 point polling lead, would that be enough to win the House? Or has there been systemic bias among pollsters inflating the Democratic numbers, and therefore, requiring Democrats to build up a large polling lead to ensure at least a narrow Election Day victory?
Scher notes further, “In August 2022, writing for Real Clear Politics, I observed, “In the 10 House elections for which RCP produced a generic congressional ballot average, Democrats outperformed the poll average four times.” Also,
And in the November 2022 midterm, the polling averages were darn close, with slight GOP overperformance. Republicans won the House national popular vote by 2.8 points. The Real Clear Politics generic ballot average was 2.5 and FiveThirtyEight‘s was 1.2.
Wider divergences are possible, but the widest since 2012 was in 2020, when the Democratic House popular vote margin underperformed the final Real Clear Politics average by 3.7 points. More often, the final margin is within 2 points of the poll average.
Scher concludes: “So while a 1-point margin in the generic congressional ballot test average may not be quite enough to instill confidence in the prospect of a Democratic House takeover, it certainly means Democrats are competitive with six months to go.