What Can we Learn from the 2022 Mid-Term Elections?
Things are not as they seem to the Punditocracy
The 2022 midterm elections were expected to be a referendum on the Democrats, Joe Biden, and a supposedly struggling US economy. Historically, this is how the first midterm election of any presidency has gone, with the result that the president’s party loses control of at least one house of Congress. That is how most pundits and analysts evaluated the race, and polling that showed things differently was often dismissed or ignored.
The baseline for the partisan makeup of the polls was assumed to be R+4 or greater in line with historical precedent, and many mainstream pollsters took this as gospel, as well as the pundits. Many mainstream polls, however, looked at the makeup of the special elections earlier this year, particularly in NY-18, and assumed a slightly higher Democratic turnout. These polls would prove to be correct. This contrasted with partisan, Republican-affiliated polling outlets that assumed a R+5 turnout advantage or greater in polls that were released to skew the averages and buttress a narrative of a “Red Wave”.
The most important polls are elections. If we look at the elections that had been held up until this midterm, we could assume a better-than-average Dem turnout that, while not necessarily delivering a house majority, would at least keep things close. Pat Ryan, the Democratic nominee in the NY-18 special election, would outperform the polls by about 3 percentage points and win.
This was the best election to base a mid-term projection since the bump from the Dobbs decision had begun to fade and inflation was still at the front of voters’ minds. It was reasonable to think of this election as a dry run of the midterms, with the assumption that the fading intensity of Dobbs would be counteracted by inflation declining due to interest rate increases.
Based on this information, I decided in early October to predict the results of the mid-term elections for Congressional offices. I took the polling of mainstream races and shifted it three percentage points towards the Democrats across the board to reflect the NY-18 special election results. Then, assuming that the general election turnout would be slightly more Republican (since Dobbs would have faded as an issue) I shifted the races one point back towards the GOP. Doing this for the 32 House races seen as genuine toss-ups, I came up with the Republicans winning a two-seat majority in the House.
I changed my methodology for the Senate races. My prediction for these races was based on how they have historically run with the House, rather than extrapolating directly from shifting the polling data. For example, a 3-point shift to the Democrats in Ohio House races from the polling baseline would historically have delivered an Ohio Senate race to the Democrats. The result was a prediction of 3 senate pickups by the Democrats, in Ohio, Wisconsin, and Pennsylvania. It was the one instance where I went off my gut and not the data, and I ended up being wrong.
It appears that the GOP will have 222 seats in the House, so I was not far off the mark. But I was off by two seats in the Senate, where Wisconsin senator Ron Johnson narrowly bested Democratic nominee Mandela Barnes, and Tim Ryan was trounced by JD Vance in Ohio. But, in principle, I correctly projected a better-than-expected performance by the Democratic party, one that was largely helped by the decision of President Joe Biden to rally his sprawling coalition of disaffected Never Trump Republicans, Democrats, and independents around the defense of democracy, necessary both politically and for a nation swarmed with candidates who denied the 2020 election results.
This was a case of the pundits ignoring the data and opting for historical analysis that has held since the Reagan era. However. they missed several key events that completely upended the historical environment. One was the breathtaking Supreme Court decision in Dobbs vs. the State of Mississippi to completely overturn the Roe v. Wade and Planned Parenthood vs. Casey decisions.
This in itself might not have generated the enthusiasm it did for the Democrats if not for the politically suicidal decision of many red state legislatures to institute trigger bans with no exceptions for rape, incest, and the life of the mother. Not even the most radical pro-lifers anticipated such restrictive abortion regimes. Senator Lindsey Graham did no favors in proposing a national abortion ban, completely contradicting conservative’s arguments about abortion and state’s rights.
The other key historical anomaly was the “Stop the Steal” movement and the anti-democratic efforts to capture key election offices by candidates who denied the legitimacy of the 2020 election. Not only was this detrimental to the country, but it also portrayed the Republicans as out-of-touch radicals who could not be trusted to tackle the economic issues voters prioritized. Yes, inflation is important, but that does not mean electing crazy people to deal with it.
This did create an awkward messaging issue for the Democrats since it seems hypocritical to be told that voting for one party in a democratic election undermines democracy. But if that opposing party harbors candidates who would overturn election results on false claims of widespread fraud, the messaging is consistent. It’s a message that helped on the margins mainly with Republicans who left the GOP specifically because of Donald Trump and his authoritarian tendencies.
Voters may not have been actively moved by the president’s clarion call on democracy, but it did not hurt. And, from the perspective of political coalition building, identifying a common enemy of an unwieldy base that is only a fraction of the opposition is smart. Biden even neutralized the criticism of demonizing his opponents by clearly stating that he respected “conservative Republicans” regardless of political differences.
Concerns over inflation, crime, and the border were major concerns in voters’ minds. But why did it not translate into the so-called “red wave”? For one, the inflation rate had declined dramatically in the few months leading upto the election. Voters feel the inflation happening right in front of them, not so much inflation that had already occurred. This is where reporting inflation like we do GDP might have been more valuable. Inflation in Q3 was a total of 0.5%, corresponding to a 2% annual rate that was targeted by the Federal Reserve. The year-over-year rate may have been 8.2 through October, but that is because the months dominated by high inflation were rolled into the average. So, inflation was a concern, but not as acute of one as the reporting could leave you to believe.
What about crime? Again, the increase in crime happened over a year ago. Violent crime rates skyrocketed in 2020, by up to 500% in New York City. But since this high-water mark, crime rates have been slowly declining and experienced a mild drop in 2021. So, crime was still unacceptably high, but voters were not “feeling” it to the extent pundits believed. It was certainly not enough to see an exodus of voters from the Democrats to the GOP. More than likely, it hurt on the margins with independents and froze some Republicans in place who were otherwise skittish about the direction of their party.
What about immigration? In general, it’s not as big of an issue for Democrats and functions more as a wedge issue to drive turnout among Republicans. Republican governors such as Ron DeSantis and Greg Abbott invested resources in bussing migrants across the country. GOP voters saw it as a sign of seriousness about immigration, and an endorsement of their belief that drawing attention to the issue would lead to Democrats addressing it. While this issue might drive Republican turnout on the margins, it also generated opposition among Democratic voters and alienated independents who winced at the cruelty of these Republican governor’s actions.
So, while these issues taken in total would indicate a favorable electoral environment for Republicans on the surface, it is not an obvious advantage when you dig in and understand the opposition that it generates from Democrats. For this reason, I hypothesized that turnout would slightly improve for Republicans compared to the NY-19 special election, but not by much. Around 1 percentage point was my guess. This aligned with a 2-percentage point improvement over mainstream non-partisan polls.
What can we take away from these results? The most important result is the wholesale rejection of the Ultra-MAGA Republicans and their continued election denial and general anti-democratic orientation. Voters want inflation fixed, lower crime, and control of the border. They do not trust the party out of power to take care of these issues if they act in a fundamentally unserious and chaotic way. Senate Republican leader Mitch McConnell said as much in a recent press conference, stating that his party had become associated with chaos and negativity.
But it was not a decisive break from the Republicans either. Political gravity still applies, meaning that the president’s party will always struggle to make up ground on the opposition. The focus on democracy and calling out the MAGA movement by name is not politically hurtful and even helps, as the Democrats found a theme that unites an unwieldy coalition of Democrats, independents, and disaffected Republicans. If the Democrats stick to their guns on this unifying issue as the Republicans spiral into disarray, and if the economic fundamentals improve, they will be very tough to beat in 2024.