Why Generic Ballot Polls May Not Be as Predictive

National Journal: “What has changed? Polarization—at two levels. Gerrymandering has reached new heights, with a Republican-favoring map set for the 2022 elections that has fewer competitive seats than we’ve seen in decades. That the map generally favors Republicans is not new—Democrats have long needed to maintain an advantage on the generic-ballot question to overcome the reality that more districts favor Republicans than Democrats.”

“The new part is how few competitive races are actually left, as the number of competitive districts has been trending downward for the last 30 years. The Cook Political Report with Amy Walter lists 55 districts in competitive territory for 2022—down from to 75 in 2018. Moreover, in 2018, only five of those 75 districts were Democrat-held, meaning Democrats had a lot more to gain than lose. In 2022, the script is flipped: 41 of the 55 are held by Democrats, making significant Republican gains possible—but not to the degree of the Democratic gains in 2018. As Charlie Cook has noted, predictions of massive Republican gains are unrealistic, given that the GOP already begins the game with 212 seats.”

“The second piece of the polarization equation as it relates to the generic ballot is how survey responses reflect partisanship. Discussions about presidential job approval are instructive here: At first it seemed that chronically low approval ratings were a problem unique to former President Trump, but Joe Biden quickly wore out his honeymoon, and himself settled into Trump-level approvals. It now appears that partisans are simply less likely to admit liking anything about the president of the opposing party than they used to be.”