A Aug. 20-22 CNN/USA Today/Gallup Poll shows Kerry and Bush tied at 46%-46% among registered voters. with 8% still undecided. This represents an improvement in Kerry’s standing from July 19-22 when Bush lead among registered voters 49%-45%
TDS Strategy Memos
Latest Research from:
By Ed Kilgore
There’s been a lot of buzzing lately about the president’s job approval ratings, so I gave them a close look at New York:
Midterm elections almost always serve as referenda on the perceived job performance of the president. It’s no accident that just prior to the two recent midterms (1998 and 2002) in which the party controlling the White House gained House seats, which is rare, the president’s job approval ratings were in the 60s. On three other occasions (1962, 1986, and 1990), presidents with job-approval ratings of at least 58 percent kept midterm losses for their party in the single digits.
With Democrats currently holding a five-seat margin in the House, President Joe Biden would probably need one of those very-high-job-approval moments to save his party’s control of the House in the 2022 midterms. Unfortunately, in this era of partisan polarization, a Biden boom in popularity is improbable. According to the FiveThirtyEight polling averages, the president’s approval rating peaked at 55.1 percent on March 22. It’s currently at 42.1 percent, and it is “underwater” (below his disapproval rating) by 9.8 percent. If Biden can turn the fairly steady downward drift in his approval ratings around, he can certainly help Democrats mitigate midterm House losses. On two occasions, presidents with pre-midterm approval ratings in the 40s lost only a handful of seats: Carter’s party lost 11 seats in 1978, with a Gallup rating of 49 percent, and Obama’s party lost 13 seats in 2014, with a Gallup rating of 44 percent. On the other hand, Democrats lost 53 House seats when Obama was at 45 percent in 2010 and 63 House seats when Clinton was at 46 percent in 1994. So Democrats can almost certainly kiss their trifecta goodbye, barring something unforeseeable.
The Senate is invariably another matter, since the particular “class” of senators up for reelection in any midterm has a huge impact on partisan performance. In 2018, to cite the most recent example, Republicans managed to gain two net Senate seats despite losing 40 net House seats, mostly because Democrats had an incredible 26-9 disadvantage in seats they had to defend. It was a terrible Senate landscape for Democrats, who, by contrast, face a marginally favorable Senate landscape in 2022. Still, a pretty reliable Senate model devised by Sean Trende projects that Democrats are likely to lose control of the upper chamber in 2022 if Biden’s job-approval ratings haven’t rebounded to the high 40s by Election Day.
In the longer term, there is a definite silver lining to Biden’s current popularity woes, reflected in this ostensibly very negative headline at FiveThirtyEight: “One Year in, Biden Has the Second-Lowest Approval Rating of Any President.” Guess who had the very lowest at this point in his presidency? That’s right: Donald J. Trump, whose approval rating was at 39 percent a year after his inauguration. If, as is currently more likely than not, the 2024 presidential election is a Biden-Trump rematch, you have to like Biden’s odds for a second term if he can improve his public standing even slightly between now and then. Yes, ex-presidents usually become more popular after they leave office. But they generally don’t insist on making the low point of their presidencies a signature moment, and defense of it a litmus test for their party, as Trump is doing with the Capitol riot and his broader campaign to steal the presidency. It’s as if George W. Bush had tried a 2012 comeback (and hadn’t been term-limited) based on the eternal righteousness of the Iraq War.
With better skill than he has showed lately and a bit of luck, Biden can make the midterms a nick rather than a grievous wound to his party. Then, like his friend and former boss Barack Obama, he could win reelection two years after being underestimated.