I’ve never been a big believer in the wisdom of voters. Indeed, I’ve devoted much of my academic career to writing about the dangers of widespread political ignorance, going all the way back to my first academic article. It was published in 1998, at a time when most experts tended to be relatively optimistic about voter competence. Since then, I published a book on the subject—Democracy and Political Ignorance – and many other articles exploring various dimensions of the problem, its implications for legal and political theory, and possible solutions.
In these works, I explained how most voters often don’t know even basic facts about the political system and government policy, and those that know more (the “political fans”), often tend to evaluate political information in a highly biased way. I also argued that information shortcuts and “miracles of aggregation” largely fail to offset ignorance and bias, and sometimes even make thing worse. Moreover, this sad state of affairs is not the result of stupidity or lack of information, but of generally rational behavior on the part of most voters: a combination of “rational ignorance” (lack of incentive to seek out political information) and “rational irrationality” (lack of incentive to engage in unbiased evaluation).
Since the rise of Trump and similar politicians in other countries, academics and political commentators have become more aware of the dangers of public ignorance. I wish I could say my own take on the subject has been vindicated. But, in one crucial respect, the Trump era has shown I wasn’t pessimistic enough.
Though I have long argued that voter ignorance and bias are serious dangers, and that information shortcuts are overrated, I also asserted that shortcuts actually work well in one important way: democratic electorates will punish politicians who cause great harm in clear and obvious ways. For example, I cited economist Amartya Sen’s famous finding that mass famines never or almost never occur under democracies, while they are all too common under dictatorship. Even ignorant and biased voters will notice a famine is going on, blame incumbent politicians for it, and punish them at the ballot box. Knowing this, democratic political leaders have strong incentives to avoid famines and other obvious disasters. And they generally do just that, at least when they have the necessary knowledge and resources (disasters can still happen if avoiding them is difficult).
“Retrospective voting”—rewarding and punishing incumbents for things that happen on their watch—often works poorly in less extreme and less clearcut cases. As explained in Chapter 4 of my book, voters often reward or punish office-holders for things they didn’t cause (most notably short-term economic trends; but also things like droughts and even sports-team victories), while ignoring some that they are in fact responsible for. But retrospective voting is a great mechanism for punishing politicians for obvious large-scale awfulness, one that works very well.
Or so I thought, along with many other scholars. But Trump proved me at least partially wrong. I was too optimistic.
Trump’s effort to use force and fraud to overturn the 2020 election was exactly the sort of obvious and blatant awfulness that retrospective voting theory predicts the electorate should decisively repudiate. Peaceful transitions of power are fundamental to democracy, and Trump’s 2020 activities struck at the very heart of this norm. Had he succeeded, it would have severely damaged the basic structure of our liberal democratic institutions. Yet a large majority of GOP voters renominated Trump again this year. And he has roughly an even chance to win the general election this year. If he goes on to lose, it will probably be by a very narrow margin, not the kind of overwhelming repudiation that would vindicate the theory.
Some people who would otherwise vote GOP are punishing Trump for his 2020 behavior by voting for Harris, or at least abstaining. Mike Pence and former GOP Rep. Liz Cheney are not alone. Thanks in part to these defectors, Trump is doing worse than a Republican nominee untainted by 2020 probably would be. But the number of such voters is much smaller than optimistic versions of retrospective voting theory would predict.
Ignorance and bias are playing a huge role in Trump’s relative success. Polls consistently show that a third or more of Americans—including a large majority of Republicans—believe Trump’s lies about the 2020 election, despite the overwhelming evidence against them, including numerous court decisions rejecting Trumpian claims of voter fraud (including some written by conservative judges appointed by Trump himself). Ignorance and partisan bias are great enough that many millions of GOP base voters reject fairly obvious facts here. If you believe the 2020 election was “stolen” from Trump, then his reaction may well seem justified, or at least excusable.
But this isn’t the full story. If Trump only had the support of voters who actually believe his lies about the 2020 election, he could still have won the 2024 GOP nomination. But he would be losing the general election in a landslide of about 60-40 or even more. He remains competitive with Kamala Harris because there are many voters (probably around 10-15% or so of the electorate) who reject his take on 2020, but prioritize other issues, such as the economy or immigration.
Here, more conventional political ignorance is playing a role. Most polls that the economy is the highest priority for voters, including swing voters, and many are angry about the inflation and price increases that took place in 2021-23. Here, there is a fairly standard political ignorance story. Swing voters blame incumbent Democrats for the inflation and price increases, even though actually both parties supported the policies that caused them (primarily massive Covid-era spending). Even worse, they tend to think Trump will bring down prices, even though his agenda of massive tariff increases and immigration restrictions would predictably raise them.
It’s not unusual for voters to misallocate blame for ordinary bad developments or to misunderstand the impact of policies. But, for a large bloc of swing voters, this relatively conventional ignorance about price increases and the policies that cause them is enough to outweigh concerns about what Trump did in 2020. Bad conventional retrospective voting forestalls beneficial retrospective voting against Trump’s extraordinary 2020 awfulness and the danger failing to punish it poses to the constitutional system.
What is true of price increases also applies immigration. Increased immigration is actually beneficial, not harmful, and the best way to deal with disorder at the border is to make legal migration easier, not harder (as Trump proposes to do). But even if you’re more of a border hawk, it’s hard to show that problems caused by migration are as pressing as threats to the constitutional order. At the very least, GOP primary voters could have picked one of several available highly restrictionist candidates who weren’t involved in Trump’s efforts to overturn the election. The belief that immigration is not just a policy problem but an “invasion” amounting to a huge crisis, is itself heavily linked to ignorance.
One possible way to reconcile optimistic retrospective voting theory with recent developments is to say what happened in 2020-21 wasn’t really that bad, because Trump’s plan to overturn the election failed and the “guardrails” held; thus, we need not worry too much about it. It’s not clear if any significant number of voters continue to support Trump because of these sorts of considerations. But, if they do, it’s very bad reasoning. Libertarian political philosopher Michael Huemer explains:
Let me tell you how I view this [argument]. Say you’re on a bus ride on a winding mountain road. You see the driver suddenly swing the wheel to the right, trying to send the bus over the cliff. Fortunately, the guard rail on the side of the road holds, and the bus bounces back onto the road. The bus driver does this repeatedly during the drive, but every time, the guard rail holds the bus back.
When you finally get off the bus, one of your fellow passengers declares that this was an excellent bus driver. He proposes hiring this driver to drive the same group to another city.
“What are you, out of your f—ing mind?” you reply. “He tried to drive us off a cliff!”
“Oh that,” says the other passenger. “The guard rail held, so what’s the big deal? Don’t worry, this next drive won’t go by a cliff. Since the rest of his driving performance was fine, we should hire him…”
Do I have to spell it out…? Driving off a cliff is not the only bad thing a bus driver can do. There is an indefinite number of disasters a crazy person can cause. Anyone who would try to drive a bus off a cliff can never be trusted with a bus, or indeed anything else, and if you think he’s an acceptable driver, you’re as crazy as he is.
I would add that a driver who tried to drive off a cliff once could do so again. And even a small chance of the guardrails failing is an enormous danger when the stakes are the future of constitutional democracy. Moreover, failing to punish politicians who seek to overturn elections by force and fraud incentivizes more such behavior. And some of those who attempt it in the future might be more successful than Trump was.
This isn’t the first time large numbers of people failed to retrospectively penalize truly awful policies and candidates because of a combination of belief in lies and flawed ordinary retrospective voting. The horrific calamity of World War I should have led Europeans to repudiate the expansionist nationalism that caused it. Some did. But many Germans actually doubled down on nationalism and imperialism because of the “stab in the back” myth that held that Germany only lost the war because of betrayal by Jews, leftists, and others.
Later, the combination of the stab-in-the-back myth and conventional retrospective voting against the Weimar Republic government that presided over the Great Depression helped bring the Nazis to power. In the US, the political consequences of the Depression were less bad. But ignorance did lead voters to embrace a range of harmful policies that actually made the crisis worse.
The Great Depression, at least, was a horrendous crisis that caused truly enormous suffering. Today’s price increases and border problems pale by comparison. If even the latter can lead many voters to forego punishing truly awful political leaders, that means retrospective voting is much less effective than I and others gave it credit for.
Recent developments don’t prove that retrospective voting is totally useless. Amartya Sen is, I think, still right about democracy and famines! Democracy is still better than dictatorship. But the threshold for reliable and accurate retrospective political punishment is higher than I and some others previously believed. A mass famine may be enough. But a blatant threat to the foundations of liberal democracy doesn’t necessarily cut it. All too many people are easily persuaded that the threat was actually justified, or that it is at least outweighed by relatively ordinary policy issues.
Voter ignorance and bias are far from limited to the right side of the political spectrum. I’ve previously written about left-wing examples (e.g.—here). But the Trump situation is the most dramatic proof that the problem is worse than even relative voter-knowledge pessimists—like me—previously thought.
The election could yet invalidate my new more pessimistic view. If, contrary to what polls indicate, Trump loses by a large margin, that would indicate he may be paying a higher political price for 2020 than I currently expect. But if he wins, or only loses narrowly, then the increased pessimism is warranted.
There is no easy way to “fix” political ignorance. I assess a range of possible options in a recent article on “Top-Down and Bottom-Up Solutions to the Problem of Political Ignorance, and in my book Democracy and Political Ignorance. I believe the best approach is to make fewer decisions at the ballot box and more by “voting with your feet,” where incentives to seek out information and use it wisely are better. But I admit that any effective approach will take time, and there may be no one fix that is sufficient by itself. We may need a combination of several strategies.
Be that as it may, recent developments strongly suggest the problem is even worse than I previously believed. That makes the need for solutions even more pressing.
