“We hurt the ones we love because we can” is a cliche, though perhaps I should be attributing it to a specific writer. Its truth is something that I find extends beyond our close familial, platonic, and romantic relationships. The mechanism behind misdirected aggression is simple: we are exposed to a source of stress that we are unable to affect, and the innocent bystanders most proximate to us become collateral damage specifically because we can affect them. The anger inside us needs to go somewhere and, in a parable of true irony, your mutual affection becomes the channel through which you express anger and frustration that has nothing to do with them.

There are a lot economists, writers, pundits, public intellectuals whose work I consume. Often I agree, sometimes I don’t, but I keep reading them because I consistently learn from them. Lately I’ve found myself becoming more frustrated with a greater share of their writing, often because they’re not being hard enough on the Trump administration, attempts to dismantle core insitutions, or the indiscriminate cruelty behind the rampant incompetence. I want them to be meaner and angrier and more direct. I want them to have an affect that I can’t. To be clear, I’ve attributed more power and influence to them than they actually have, but I think that’s not the real problem.

The real problem is that I know that no one in the Trump administration cares that they are cruel or incompetent. You can, at best, embarrass these people briefly, but you can’t shame them. They only internalize consequences and they’ve yet to experience any. They have coalesced around the singular belief that has served as the North Star for Trump’s entire life: there are no rules. Rules are fake. An illusion. A mass delusion. There are no rules and you can do whatever you want in the moment that serves your ambitions and ego and then move on to the next thing.

What do you do when your entire mechanism for affecting and contributing to world is the written word, criticism, the speaking of evidence-based truth to power and that power doesn’t care? What I find myself tempted to do, and what a lot commentators our there (especially on bluesky) are doing, is attacking the people who might and do actually listen with an undeserved fury. The criticism is often valid, but it’s just 30% meaner than it needs to be. More personal. More cruel.

I care about AI. I care about energy subsidies. I care about crime and education and health. But, if I’m being honest, there are times every day when I don’t. I’m a professor, I care about and contribute to bleeding edge research, but the moment we are living through isn’t about PhD level questions. These are 5th grade social studies times. Democracy. Rule of law. Citizenship. All men are created equal. Basic human dignity. That’s the reality and it’s not hyperbole.

I hope everyone will keep doing their research and commentary about the nitty gritty of day to day science. I also hope that everyone will take the time to grant just a bit more space emphasizing the basics, to leave no doubt about where they stand. Becaue no matter how someone might identify politically, in this moment it’s mostly irrelevant. Liberal, conservative, libertarian, classical liberal, neoliberal, new liberal, social democrat. The differences are trivialities. There are only two groups that matter: those who want to keep the basic institutions intact and those that want to burn it to ground. That’s it.

So just keep that in mind when you’re mad about someone online, about what they wrote, what you think they believe. Are they trying to hold the world together while bandits are stripping the walls for copper and carving out chunks of marble from every load-bearing pillar? If the answer is yes then they deserve grace. I’m trying and I hope you’ll do the same for me.

I’m not just doing this to plug my own event. It’s also about the only thing on my mind after spending the week leading and moderating this timely discussion.

If you like to read and discuss with smart people, then you can make a free account in the Liberty Fund Portal. If you listen to this podcast over the weekend: Marc Andreessen on Why AI Will Save the World  (2023) you will be up to speed for our asynchronous virtual debate room on Monday May 12.

Keeping in mind the stark contrast between this and the doomers we discussed in the past week, here is Marc’s argument in a nutshell:

“The reason I’m so optimistic is because we know for a fact–as sort of one of the most subtle conclusions in all of science–we know for a fact that in human affairs, intelligence makes everything better. And, by “everything,” I mean basically every outcome of human welfare and life quality that essentially we can measure.”

When it’s put that way, it’s hard to disagree. Who would want less intelligence?

See more details on all readings and the final Zoom meeting in my previous post.

Another interesting bit by Marc:

“By the way, look: there’s lots of work happening that’s not being published in papers. And so, the other part of what we do is to actually talk to the practitioners.”

Even though it might seem strange to look to podcasts instead of published books and papers for cutting edge information, it really does seem like the story was told in human voices for the past 3 years. Dwarkesh was probably the best, but Tyler and Russ deserve credit as well for bringing these conversations out of the closed rooms and into the public domain.

As of 10:30am this morning Berkeshire Hathaway is down 5.6% on the news that Warren Buffet is retiring at the end of the year. At first blush, this makes sense. Buffet is an irreplaceable input into their production function. However, the man is 94 years old, a full 24 years after nearly everyone retires, so this was not exactly an unforseeable event. Why wasn’t more of this already baked into the price? Further, this would appear a far better outcome – announcing retirement more than 6 months in advance- than a more sudden and unfortunate event, such as the passing of a man in his mid 90s. It’s not unreasonable to suggest that both event possibilities would be baked into the price and, with his retirement beingthe better outcome, thus the price could have even gone up.

To me, this is a reminder that there limits to how much Knightian uncertainty can be baked into a price. Put another way, it is a reminder of the costs that uncertainty (nearly?) always imposes on markets. We would all, voters and legislaters, be wise to remember that as the current Presidential administration continues to inject seeming daily boluses of constitutional, existential, and economic uncertainty into our lives.

Can Enhanced Street Lighting Improve Public Safety at Scale?  MacDonald, John and Chalfin, Aaron and Moritz, Maya and Wade, Brian and Mendlein, Alyssa and Braga, Anthony and South, Eugenia C.

Add to the growing evidence that few things deter crime so well as being observable by others.

I’ve got a new working paper circulating.

“Better Stealing Than Dealing: How do Felony Theft Thresholds Impact Crime?” by Stephen Billings, Michael Makowsky, Kevin Schnepel, and Adam Soliman.

The abstract:

“From 2005 to 2019, forty US states raised the dollar value threshold delineating misdemeanor and felony theft, reducing the expected punishment for a subset of property crimes. Using an event study framework, we observe significant and growing increases in theft after a state reform is passed. We then show that reduced sanctions for theft have broader effects in the market for illegal activity. Consistent with a mechanism of substitution across income-generating crimes, we find decreases in both drug distribution crimes and the probability that a released offender previously convicted of drug distribution is reincarcerated for a new drug conviction.”

For those interested in a bit more of the nitty-gritty, we analyze both arrest and recidivism data within a stacked event study because we are dealing with staggered (diffent years) and fully-absorbing treatments (i.e. once they raise it they never lower it back). States raise their felony theft thresholds for a portfolio of stated and unstated reasons, but the reality is that the value of the marginal stolen good is often deteriorated by decades of inflation only to be doubled or tripled by a single act of legislation. This makes for an excellent before/after experimental setting to test the effect on crime.

We’re going to look at two things broadly: arrests and recidivism. The importance of arrests is straightforward: they give us a sense of the rates of crime across populations. Recidvism is more subtle. More on that in a bit.

In the quarters leading up to a threshold change (above) we see flat pre-trend with a coefficient of zero i.e. nothing happening. Nothing happening is good, it means that neither law enforcement nor criminals exhibit any sign of anticipating the change. Once a given state makes the change, we see an uptick in rates of theft within 6 months that persists for three years. Speculating beyond that is dangerous – too many other things happening in the world. But criminals seem to be responding.

We don’t see any effect on Burglary or Robbery, however (below). This is also a sign of rational criminals since these thresholds don’t apply (i.e. they are always a felony, regardless of property value). In other words, we don’t see an effect on all property crime, just on those crimes for which expected punishment is reduced.

We do, however, see an interested effect on drug distribution (below). In the quarters after a theft threshold reduction, we see a significant and persisting reduction in drug arrests. Yes, we include controlling covariates for medical and recreational marijuana legalization. There’s something else going on here. Are people exiting one income-generating crime for another?

This is where recidivism comes in. Using detailed, restricted-access, prisoner records, we track when prisoners are released and if/when they are returned to prison. By stratifying the analysis by the crime types they were previously incarcerated for, we can separately estimate the effects of felony threshold changes on individuals with human and social capital in the drug distribution business from those who do not. What we observed is both striking and subtle.

For indidividuals previously incarcerated for drug distribution (top left), their rate of return for future drug convictions is immediately lower with a reduction in the felony threshold. For those who were never in the drug trade, there is no effect (bottom left). Reducing the expected punishment for theft is pulling individuals out of the drug business.

Now let’s look at the return rate for felony larceny. For most prisoners (bottom right), there is a massive reduction in the rate of return for larceny. This makes complete sense – if more theft is classified as a misdemeanor, you are much less likely to be re-incarcerated with a new sentence for it. When we look at prisoners previously incarcerated for drug distribution, however, there is no observed effect (apologies for the changing y axis scales, there’s no good way to keep them constant). What does this mean? We interpret this as evidence that the reduction in punishment for theft is canceled out by the shift into theft as a preferred way of earning income. The labor substitution effect cancels out the effect of reduced punishment.

There’s obviously a lot more in the paper. No, there is not an effect on violent crime (Table 2). No, there is not an observed effect on officer enforcement intensity (Appendix Table A3). No, we can’t do a regression discontinuity at the threshold values (too much bunching, see Appendix Figure A7). The conclusions are both obvious and subtle, but the most important may simply be the reminder that all policies have tradeoffs and spillovers, no matter how narrow they might seem.

TLDR; When states increase the property value threshold delineating misdemeanor from felony theft, prospective criminals respond by a) committing more theft and b) substituting out of drug distribution and into theft. This pattern of substitution in the criminal labor market is more evidence that criminals are not only rational and respond to deterrence incentives, but are also selecting across criminal options, which means we should expect spillovers across crimes when policies create differential changes in expected punishments, enforcement, and returns.

In no particular order:

Caetano, Gregorio, and Vikram Maheshri. “Identifying dynamic spillovers of crime with a causal approach to model selection.” Quantitative Economics 9.1 (2018): 343-394.

The “broken windows” theory of crime (i.e small crimes lead to bigger crimes) continues continues to find very little support.

Cabral, Marika, and Marcus Dillender. Air pollution, wildfire smoke, and worker health. No. w32232. National Bureau of Economic Research, 2024.

Air quality remains an underrated public good.

McBride, Michael, and Garret Ridinger. “Beliefs also make social-norm preferences social.” Journal of Economic Behavior & Organization 191 (2021): 765-784.

It’s conditional cooperation all the way down.

Literature on Recent Advances in Applied Micro Methods

Your one-stop-shop for an updating list of the papers currently advancing causal identification in social science

What, you think I’m going to pretend anyone is paying attention to anything but the trainwreck on Wall Street? As of 10:15AM this morning, the market is down 8% in 5 days, almost 20% off it’s peak, and is still falling. It’s entirely attributable to a unfathomably stupid trade war that has been forecast for months, if not years. This is the kind of probabalistic event that is usually internalized within the market in advance, which suggests that either very few people thought Trump a) was telling the truth, b) would be able to execute, or c) other forces within government would be able to stop him.

The legislative branch has largely ceded power to the executive, with only the judicial hanging on as some check against power. The open question, then, is at what level of damage will the legislative branch find incentive to reassert itself against an executive that (probably) doesn’t have the constraint of a future electoral victory to pursue? Will the destruction of great swaths of the US and global economy warrant reclaiming of power or impeachment of an executive?

I’m not optimistic.