Republicans as the Lesser of Two Evils

Thoughts on operating in a world of limited knowledge, and how to pick a political party

With the midterms right around the corner, I thought it would be a good time to provide some thoughts regarding the question of how a smart person should go about choosing a political party, and why I consider Republicans the lesser of two evils. Of course, most people rely on tribal signals. But you aspire towards something better. How should an intelligent person who wants to improve the world go about deciding whether to vote for, give money to, or otherwise support either the Democrats or Republicans in zero-sum political contests?

The information problem related to this question is very rarely treated with any depth. I think most intelligent observers would say that you should look into the issues and decide which side, on balance, has superior policy ideas. The Wall Street Journal recently ran a poll asking “Which party has a better economic plan to​ make life easier for people like you?”, and found that suburban white women said Republicans, whereas they had supported Democrats two months earlier.

I find this shift amusing, not only because most suburban white women are not studying policy documents, but because deciding whom to vote for based on the specifics of either party’s “plan” – as if such a thing even exists in the first place – isn’t a plausible standard for anyone, even those who follow politics fulltime for a living. But what’s the alternative?

You Need to Pick an Ideology

To get a scope of the information problem facing the voter, consider that an American president is allowed to appoint up to 4,000 officials in the federal government, although many jobs go unfilled at any one time. How many of those appointed by Trump can you name, and of those, how many policies that they implemented are you familiar with? The last president signed 220 Executive Orders throughout his time in office, Biden added around 2,000 pages to the Code of Federal Regulations in 2021, and the current Democratic-controlled Congress has passed 213 laws since the beginning of last year. You probably have never read a single page of the CFR.

The point here is that “become informed about policy” is not a realistic goal. You might focus on gathering facts in one or two areas, but you will remain extremely ignorant about over 99% of what the government is doing. I know civil rights law and foreign policy very well, and have done a lot of research on issues related to the pandemic. But I have limited knowledge on basically everything else, despite still being more informed than all but the tiniest sliver of the population.

I remember in college talking to some kid who said he liked Obama because he seemed so smart and well-read. Let’s grant that Obama was three times more thoughtful and informed about what his appointees were doing than your typical Republican president. That would still have made him ignorant about the vast majority of things that were happening within his administration. Being able to think carefully about 1% of governance instead of 0.33% is not a major improvement.

And this is why we have ideologies. Usually, “ideological thinking” gets a bad rap, implying narrow-minded rigidity. But it is the only way to have opinions in any but the narrowest sense. Those who claim to be “non-ideological” are usually exaggerating the ability of humans – not least themselves – to digest and process information, while nonetheless sneaking their own assumptions into their conclusions. This is why Andrew Yang rubs me the wrong way.

One thing you might try to do is outsource the job of picking a party to science (or “science”). The most interesting study I know of that tried to do something like this used a regression discontinuity design to see which party was better. You can’t just compare places where Democrats are in power to places where Republicans are in power because different kinds of states elect different politicians. A 2007 paper by Andrew Leigh tried to get around this problem by conducting an analysis that took into account the political leanings of states, finding “no evidence of gubernatorial partisan differences in tax rates, welfare generosity, the number of government employees or their salaries, state revenue, incarceration rates, execution rates, pre-tax incomes and inequality, crime rates, suicide rates, and test scores.” Democrat governors were associated with higher post-tax incomes and lower unemployment rates, but both findings show tiny effects that barely meet standards of statistical significance, and even that only under certain specifications. In other words, the study found no grounds for thinking that there were any major differences with regards to state-level outcomes based on the party of the governor.

The null results here don’t prove much, since a Democrat who can win in a Red State or a Republican who can win in a Blue State is likely to be more moderate than the typical member of their own party. So while a positive outcome in Leigh’s study would’ve indicated to me that it was capturing something important, a negative result is not very useful. We don’t know what the benefit is of a state going from one end of the ideological spectrum to the other. Moreover, the parties are further apart ideologically than they were in the time period covered in the 2007 study, so I wonder if the results would still hold today. In fact, a 2017 paper shows that election outcomes increasingly matter for the ideological tilt of state-level policy, although the in-state differences are dwarfed by the variation between states, and this study doesn’t measure ultimate outcomes we care about like crime and economic growth. Looking at the federal level, we simply don’t have enough of a sample size to say much of anything, as there’s only one president at a time and a party might go a decade or more out of power.

So facts are little help in determining one’s politics, and the same is true for science. In the end, one should see partisan politics as a matter of picking the tribe with the right priors. If you’re voting Democrat in one election and Republican in another, or splitting your ticket, you are likely behaving illogically, unless you have some sophisticated theory about the benefits of divided government.

A president will appoint perhaps hundreds of judges and thousands of other government officials and words will be added to or subtracted from the Code of Federal Regulations. The next Congress will either support or reject the president’s nominees, and pass hundreds of laws that you will never even hear of, much less read or understand. You outsource your thinking not to the tribe that thinks the most logically within the confines of its own worldview, but the one that is better in terms of the underlying assumptions that it rarely if ever questions while making decisions.

Pro-Market is the Best and Most Important Heuristic

I’d argue that the most basic difference between the parties here related to how much they trust markets relative to experts. While the idea that the parties disagree on the regulation of markets is far from original, I think a lot of people have either become confused on this point or started to give undue weight to less important issues. A list of all the things that one can potentially consider a “problem” would be infinitely long. What is the threshold for government doing something about each one? And is what government does likely to be beneficial or not? My view is markets have an amazing capability to aggregate information and create incentives for pro-social behavior. We rely on them too little, rather than too much. I haven’t investigated every proposed regulation in every industry, but I’ve looked into a handful, and have come away remarkably unimpressed by the anti-market side almost every time. A recent example is Biden’s war on the gig economy, where the government has seemingly decided that having the choice of working a job with more flexibility in setting one’s hours is a bad thing that workers need to be protected from. Yes, occasionally liberals point to a market failure that should be addressed by policy, but usually their analysis goes no further than suggesting we do something because it superficially sounds good. If people who wanted to intervene in the economy tried to do a cost-benefit analysis each time, that would be one thing, and we could check their math. But they don’t even make a token effort; it was something of a revolutionary project when Cass Sunstein tried to bring cost-benefit analysis into regulatory policy, and he largely failed. Not that I would take any cost-benefit analysis presented to me at face value - I doubt that all the knowledge in the world can lead government to do a better job of regulating, say, the labor market or education system than processes that involve aggregating the choices of millions of individuals, each operating under conditions of scarcity and having to directly live with the consequences of their own actions. But the fact that government doesn’t even go through the motions of trying to prove that the net impact of its actions are beneficial increases my confidence that it should do as little as possible.

And yes, in general, the parties do follow their stated ideologies in governing, with Obama imposing many more burdens on the private sector than Bush, and Trump setting a new standard in reducing the impacts of federal regulations. When Congress negotiates over budgets, it is Democrats who want to spend more and Republicans less.

Of course, there are exceptions to the “government should do nothing” rule. Closing the hole in the ozone layer was good, as were some of the non-libertarian aspects of Operation Warp Speed. The state should prosecute robbery, rape, and murder. My model of government – and the exact numbers are nothing more than a guess – is that 10% of what it does is absolutely essential, maybe 10% is on net good, and 80% is somewhere between wasteful and disastrous for society. That’s at all levels; when the federal government is considered on its own, it’s probably even more useless and counter-productive. And I don’t think any modern party is going to get rid of much of the essential stuff, so our politics is mostly about how many harmful regulations to eliminate or prevent from being enacted.

I think everyone writing about politics should be clear about their priors here. We’re all to a large extent just guessing when we’re talking about things we haven’t done our own intensive research on and siding with one side over the other, and it’s good to be honest about what assumptions we bring to the table. I don’t trust the judgment of anyone who pretends that they arrived at their position that one party is better than the other through researching a significantly large enough portion of the policy space.

The Relevance of Other Issues, and Existential Risk

Republicans do make one major exception to their greater trust of markets, and that’s in the areas of immigration and sometimes trade. Many seem to believe that the laws of economics are different depending on whether one is interacting with someone born in the United States or another country. Those who generally support allowing the market to set labor conditions suddenly take a different view when it comes to hiring Mexicans, just as how people who wish American women would have more babies start talking like Malthusians when the topic of immigration comes up. But I think, on balance, Republicans are still the more pro-market party, and xenophobia might have to be the price one has to pay to get a relatively pro-market party into a position of power in a democracy. As some liberals have pointed out, Republicans run on crime and stopping immigration, but they often in practice end up trying to cut Social Security and Medicare, and hopefully one day they’ll succeed. As for the supposed cultural impacts of immigration, it’s hard to take such concerns seriously in a country already as large and diverse as the United States, one in which our politics today revolve around fundamental moral disagreements between two factions that hate each other.

Regarding social issues, I tend to think that they’re overrated as something to vote on. Culture is important, but, unlike with economic policy, where the state sets the tax rate and decides under what condition one is allowed to make a living, it is uncertain how much governments operating within the Western tradition can do to change things in areas like gender relations or attitudes towards religion. If one party could snap its fingers and make the kids less gay, it would be worth taking that into account, but I think TikTok and chemicals are more important than what’s being taught in public schools (someone should investigate chemicals). I’d say something similar about crime. Democrats are the pro-crime party, but I’m just not sure that electing Republicans, especially at the federal level, makes much of a difference given that we can’t, say, start automatically executing large numbers of criminals upon their second or third felony conviction (which I would support once we got rid of all the victimless crimes). And while I think civil rights law has had disastrous social consequences, its cultural effects were unpredictable when the policies were originally created, and my opposition to the principle of treating anti-discrimination as a government concern is to a large extent just a subcategory of my opposition to central planning and interfering in labor markets more generally.

What about existential risk? On a limited number of issues, it might be worth doing a deep dive, since if certain doomsayers are right, then there are one or two issues we absolutely have to get right. We can therefore go down the list regarding the main dangers humanity faces going forward.

If I was worried about the AI singularity and thought one party had a better plan to deal with it, I could vote on that issue. But I think people who argue against such a thing being on the horizon make some very good points, and even if they didn’t, I would agree with Tyler that there’s probably not much our puny brains can hope to do to restrain a superintelligence, either before or after the fact. As for global pandemics, the priorities here are:

1. Preventing gain of function research. Republicans appear better on this for purely partisan reasons, given their hatred of Fauci and the Chinese. But maybe their lack of enthusiasm for government regulation makes them less likely to act on the issue.

2. Cutting through red tape to create future vaccines as soon as possible. Republicans were more willing to do this during Operation Warp Speed, with Democrats calling for respecting normal CDC and FDA processes. I’m not sure this would hold in a future crisis, as anti-vaxx has made a lot of inroads on the right, which is very disturbing, especially since much of the rhetoric on this issue mirrors the kind of safetyism that made the CDC and FDA so terrible in the first place.

3. Avoiding lockdowns and mask mandates, which will in all but the most extreme cases crush economic growth and the enjoyment of life while not being worth the costs. Republicans are clearly better here. If we had a disease that was deadly enough that such extreme measures were worth it, I think even Republicans would be on board and our entire politics would get scrambled. Covid just wasn’t ever that serious, especially after vaccines were developed and the only rational policy became to stop doing anything else to prevent the spread of the disease.

On avoiding nuclear war, I don’t think there’s that large of a difference between the two parties, and forecasting the geopolitical impacts of a major shift in American foreign policy would be difficult even if there were. Finally, I am very hopeful about the prospects of humanity making use of embryo selection and genetic engineering, and would consider any attempts to stop progress in this area to be another existential threat to our species. But neither party is likely to ban these things in the immediate future, and I just hope they stay too distracted by the culture war to do so. If I was to take a guess, I would say Republicans are more likely to take action here based on religious principles than Democrats doing so on egalitarian grounds, but if liberals can keep teaching LGBT to the kids maybe conservatives will continue focusing on that instead.

All of this means that Republicans are worth supporting because they have better priors with regards to economic regulation, which is most of what government does. The most important area that can be considered a social issue is civil rights law, and again the conservative approach is preferable. On issues that we may consider existential, Republicans are either better on those too, or there’s a great deal of uncertainty about which party is preferable, with attitudes towards reproductive freedom being a major exception in having a likely Democratic advantage. The focus on priors and a handful of existential issue are why I can support Republicans while acknowledging that they are the stupider, and in some cases more immoral, party. No one is smart enough to run the economy or design an “industrial policy” or whatever, and it’s a dangerous delusion to think that you can. Alternatively, a belief in the Democrats as the lesser of two evils would have to make the case that at the margins, most of what government does is good. This would need to assume that there is such low-hanging fruit out there that even a state that forgoes cost-benefit analysis is usually on the right track. Some kinds of scientific and technological investment might fall into this category. Or, even if most of what government does it wasteful or counter-productive, Democrats are at least likely to get things right on a limited number of controversies that are the most important issues. For the reasons given above, I’m skeptical of either argument, and therefore hoping Republicans do well on Tuesday.