One of the arguments that the reactionary movement makes is that in a democracy, politicians will have to do what voters want, and voters are often wrong about policy. For example, rent control is a bad policy and yet a majority of voters support it. When you add up all the ways that voter opinion deviates from an objective standard of good policy, the loss in terms of economic output is very large. This inevitably leads to worse outcomes than if a rational dictator implemented good policies and then simply took a cut off the top. 

I won't argue that voters always pick good policies, or whether a dictator would be likely to do better. ^[1] However, I think voters are more aware of their own lack of policy expertise than some assume, and they respond to this ignorance more rationally than one might appreciate at first.

A Simple Model

Imagine you are a voter who doesn't understand anything about politics or policy. Not only do you not understand what policies are good, you don't even know what aspects of life are controllable by politics. That is, suppose you are having trouble finding a romantic partner. Is there anything that politicians could do differently that would prevent this? You have no idea. 

You also have no idea whom to trust when considering whether to defer to peers or experts. For any relevant political issue, you can find a lot of people with various credentials on both sides. Sure, the credentials are different--one side may have a lot of medical doctors and another may have a lot of chiropractors--but you have no way of knowing which credentials signal knowledge, if any. Perhaps you know that some people are very convincing and will be able to make you believe anything about policy, regardless of its true value.

If you are this voter, your optimal strategy is to vote for the incumbents whenever your life is going well and vote out the incumbents whenever it is going poorly. Of course, you need to define a threshold for how well life has to be going for the incumbent to get your vote. One simple option is to ask the question: are you better off today than you were four years ago? That is, you vote for the incumbent to another n-year term if the past n years were better than the n years before that.

The benefit of this strategy is that when in office, a politician cannot do better than sincerely trying to make your life better. If every voter voted like this, a politician in office could not do better than making at least a majority of people's lives better. You will still end up voting for a lot of lucky bad incumbents and against a lot of unlucky good incumbents, but it should average out in the end (barring the issue in the next paragraph).

The obvious downside of this strategy is that politicians can make decisions that affect you after they're out of office, so you would incentivize politicians to do things that make your life better in the short run and worse once they're out of office. This is a real and serious downside, and it's not clear whether it turns the whole strategy from rational to irrational. Regardless of whether this is a fatal flaw of the simple model strategy, I think a lot of voters are implicitly following this strategy. ^[2] This partly solves the problem of voters not understanding the effects of policies, while introducing the new problem of politicians kicking cans down the road. ^[3]

Reality

I think there is decent evidence that a politically significant subset of voters follows something like this strategy in the US. This explains why macroeconomic conditions like inflation are so politically important, regardless of whether or not economists would attribute blame for the inflation to the incumbent party. It explains why it's so hard to do anything about long-term issues like the national debt and pandemic prevention. It explains why a lot of (potentially) bad policies that have broad public support--like widespread rent control, high minimum wages, high corporate taxes, wealth taxes, etc--have limited (though still some) adoption within the US.

Probably a lot of this is obvious to everyone. I just hadn't seen this kind of model explicitly written down and justified according to a sort of bounded rationality, and I think it explains both the surprising resistance to immediately bad ideas and high susceptibility to bad ideas that only play out in the long run of modern democracies.

1. ^{^}

   I think the answer is clearly no

2. ^{^}

   At least in the US, the only country I'm familiar with

3. ^{^}

   Of course, politicians would do this anyway as long as they could get away with it, but in a situation where voters intentionally avoid even trying to understand issues and only vote myopically, the problem is worsened