Here’s a chart. I’ve obscured all of the labels and titles so that all you can see is the raw data, but that shouldn’t be a problem. Can you, with your amazing, highly-evolved pattern-seeking neocortex, notice the trend?

Of course you can: almost all of these numbers are in the red. And if you were trying to analyze this data, the most obvious hypothesis you could propose is that some underlying factor makes it likely that these numbers will always be red. This is such a simple point that I suspect when you started reading this you were anticipating something more profound or complicated, because of course there’s a persistent reason why all of these numbers are red; you barely even have to think about it to make that point. It’s the kind of inference that even tapeworms are capable of.

With that in mind, here’s the data we’re looking at:

Now, if we were being rational about this, revealing the labels wouldn’t change anything. You would look at these vote share numbers, and you would assume that some persistent force in US politics, demonstrable over a 68 year data set, guarantees that the president’s party always has a major disadvantage during midterm elections. A more than seven point disadvantage on average that quite often jumps into the double digits. Proceeding with that hypothesis, you would then look for some unchanging factor that governs this trend, and it would be a central consideration in your analysis of our current politics and your predictions for the future.

Yet if you look at how our media analysis of electoral politics in the US proceeds, this is absolutely not how it works. Here’s a typical example of how it actually works:

First we get some polling (here, on Roe v. Wade) that seems to suggest that one party has an advantage on some particular issue. And then the kicker:

The question now is whether this…will be enough to energize otherwise unenthusiastic Democrats and alter the course of an election that had been trending in the GOP’s direction.

There’s already a vast catalogue of articles out there asking this exact same question about Ukraine, inflation, student loans, wokeness, defunding the police, and so on. And they’re only going to continue to pile up over the next couple of months.

The premise in all of them: that these are proximate causes of the outcome. But while this may seem plausible at first glance, the pattern noted above creates a serious problem. Is it simply a massive coincidence that all of these diverse and seemingly unrelated factors contributing to all of these election outcomes have, for nearly seventy years, happened to favor the opposition party?

Let’s step back to the abortion article’s question. Will Roe alter the course of this election? Here’s my prediction: no. Democrats are going to lose vote share in the House, just as the White House’s party has in 16 of the last 18 midterms.

I have to stress that I am not making this prediction given any kind of significant analysis of voter sentiment or likely turnout or how big either party’s war chest is or anything like that. All I’m doing here is looking at the black box of elections in the modern US and noticing an extremely reliable pattern in what’s coming out of the box regardless of what’s going on inside. This may seem unsophisticated, but it’s really just the first step of trying to understand elections in a rational way. If it turns out that this approach makes better predictions than the fine grained analysis of massive datasets, then so much for the data.

That said, on top of the prediction, I’ll also propose an explanation. The reason that neither abortion nor inflation nor Ukraine nor anything else that comes up is likely to break the change the outcome is that Americans do not see voting as a real opportunity to address particular political problems. They may have strong feeling about all of these issues, but since they don’t think voting is likely to do anything about them, none of these proximate “causes” end up driving election outcomes.

This explanation is particularly consistent with the well known drop off in turnout during midterm election years. Since the mid-twentieth century, nothing has convinced more than half of all eligible voters to show up at the polls. One could imagine that this is simply because nothing has happened in a 70 year span that voters care about, but this is plainly at odds with the polls. The more likely explanation is just that over the past 70 years, Americans just haven’t seen voting as a plausible solution to anything. If this is true, then issue-based analyses like the abortion article above are mostly irrelevant to what will happen in November, and the real question these articles should be asking is something like have Americans decided that voting matters yet?

This attitude towards government also, I propose, has a second crucial effect: anti-establishment contrarianism. Some Americans express their frustration and lack of confidence in our unresponsive democracy by simply declining to vote. But others, I suspect, react to it by simply voting against whoever seems to have the power at any given moment.

If there’s anything to this theory it would go a long way in explaining the midterm opposition trend. On one hand, frustration with the government predicts that a significant number of voters will cast contrarian protest votes in a way that corresponds exactly with that pattern. On the other, meanwhile, it also prevents other potential political concerns from derailing this trend, since Americans do not see voting as a way to address them. The exact same attitude towards the government that drives up contrarian voting also drives down the sort of issue-based voting that could drown it out.

One compelling point of evidence in favor of this theory can be found in a 2015 NBER paper by Brian Knight that modeled various potential explanations for the midterm pattern. Its conclusion:

…the bulk of the evidence points towards the Presidential penalty hypothesis playing a stronger role than surge and decline and a reversion to the mean in voter ideology.

Translated into English, what this means is that you can get really far in understanding the midterm voting data by just assuming that some people reflexively vote against whoever’s in the White House. Farther, in fact, than you can with any other theory. And that includes theories about ideological centrism among voters: they aren’t just fluctuating this way as a way to maintain some kind of “mean in voter ideology.” It’s just a penalty against whoever is in power.

This framework I have laid out here is not in its broadest outlines all that controversial among political scientists. It’s generally understood that voter disengagement has uncoupled popular concerns from electoral politics, and it’s also generally understood that even a small number of contrarian voters can have an outsized effect in low-turnout elections. But what our media in particular consistently ignores is just how overwhelming a role this dynamic plays in determining election outcomes. And I think there’s a reason for that, too.

Political life in the United States has, particularly since the 70s, been dominated by neoliberalism’s assault on state power. This has always been the core of neoliberalism: rich people who don’t want to pay taxes, don’t want to be subject to regulations, don’t want the government doing things they could be extracting a profit from, and (above all) don’t want to worry about the prospect of socialism. Neoliberalism come to mean all kinds of bizarre things in the discourse these days, but as I argued recently: if you don’t understand capital’s drive to destroy state power, you have missed one of the foundational forces in contemporary American politics.

This campaign has had two related ideological effects. First, insofar as the rich have succeeded in destroying our government and impeding even its most basic functions, the American people have understandably lost all confidence in its ability to improve their lives. And second, the rich have long engaged in quite deliberate and conscious efforts to spin the state’s ruin as self-inflicted, the inevitable consequence of its fundamental malevolence and dysfunction. In tandem, these efforts have successfully embedded neoliberal ideology deep within the American psyche. Everyone just takes for granted now that this is just how government is and how it always has to be.

The consequences for voting are obvious. Neoliberalism teaches Americans that there is no point in voting, even when the issues are serious and when the government could hypothetically do something to address them, since in practice the government can never do anything to help anyone. And even those who do vote often find themselves engaged in contrarian voting — either as a gesture of empty protest, or in a kind of nihlistic effort to bring even more chaos into our politics. It is tempting to imagine one party or the other as the party of neoliberalism, and to conclude that as a systematic and ideological force in our politics, neoliberalism simply works to place one or the other in power. But that is the opposite of what neoliberalism wants, because a party with power is always in danger of becoming a party with responsibilities. What neoliberalism wants is exactly what we have: a politics that is always moving back towards an equilibrium of partisan gridlock.

And that’s why the media talks about our elections the way that it does: as an affair that’s almost entirely contingent on various issues-of-the-moment. Can you imagine if every article on the midterms that you come across in The New York Times or Vox just read something like this?

Republicans are set to gain ground in Congress in the 2022 midterms because of a systematic feature of US elections that has guaranteed this outcome for nearly three quarters of a century. Historians describe this as “basically a scheme set up by the rich, it’s complicated”; as long as it remains in place, they add, nothing that happens between now and November is likely to change the outcome.

This is the sort of problem that, if it were taken seriously, would probably end in civil unrest and political changes far more radical than anyone financing our media would tolerate. It’s one thing for the Times to occasionally run a Jimmy Carter op-ed suggesting that maybe our democracy isn’t working all that well, or to occasionally bury in an avalanche of content a wonky post or passing aside that gives a nod to the midterm trend. It would be quite another for our media to talk about neoliberalism in proportion to its role in our politics.