Temperature and judicial decisions

The title is “Temperature and Decisions: Evidence from 207,000 Court Cases,” the authors are Anthony Heyes and Soodeh Saberian, and here is the abstract:

We analyze the impact of outdoor temperature on high-stakes decisions (immigration adjudications) made by professional decision-makers (US immigration judges). In our preferred specification, which includes spatial, temporal, and judge fixed effects, and controls for various potential confounders, a 10°F degree increase in case-day temperature reduces decisions favorable to the applicant by 6.55 percent. This is despite judgements being made indoors, “protected” by climate control. Results are consistent with established links from temperature to mood and risk appetite and have important implications for evaluating the influence of climate on “cognitive output.”

Here is the (gated) link to American Economic Journal: Applied Economics.  Here are ungated versions.

Comments

I'm gonna go ahead and say this ain't gonna replicate.

Don't be so sure.

From 2011: Judges are more lenient after taking a break, study finds: Prisoners are more likely to be granted parole early in the day or after a break such as lunch, according to researchers

That in itself is a very doubtful study, see http://nautil.us/blog/impossibly-hungry-judges

Next: study shows judges named Lenny are more lenient than other judges.

File under: replication crisis

Judges named Thurgood have airports named after them at a disproportionately high rate.

In any endeavor with multiple actors and multiple events it would be impossible to not have some statically identifiable correlations. The real story or take away from this study is that anyone thought it was significant. Believing that this study has any significance should exempt them from any future serious consideration.

Agreed. Put 207,000 observations into a regression and slice the data a few ways, you're bound to get statistical significance. I didn't read the above posted Nautilus article, but I'd buy that a judges hunger and blood sugar would affect their decisions. What I don't get from the paper on outside temp and judges decision-making is that why would outside temp affect decisions that are being made in air-conditioned surroundings. Here are the authors:

"While we don’t observe their precise movements nor the particularities of the indoor conditions in which they work, we can say that these professionals work in good quality, climate-controlled environments. Also, presumably, they travel to work and move around their cities in a manner consistent with better off professional workers (have air conditioning in their cars, etc.). In other words, the subjects that we study are offered a level of protection against weather variations that most people, even office-based professionals, would find quite comprehensive. That despite this we still observe substantial and robust effects of ambient temperatures outdoors to how these individuals are going about their business indoors, causes us to be sceptical of claims that climate control is likely to be fully-effective in ameliorating climate impacts."

I guess the authors never considered that their findings are just noise. If you have a "theory" that says temperature will affect decision-making, and you find patterns in surrogate measures that don't really have much to do with your theory (i.e. they measure outside temp not inside temp), then you might just be finding all noise and no signal.

The hungry judges soon the sentence sign, and wretches hang that jurymen may dine.

-Alexander Pope, a few hundred years ago.

I'm going to go ahead and say it will replicate, though it doesn't need to.

Well, there is also this paper showing warmer years have lower exam scores from NBER: https://scholar.harvard.edu/files/joshuagoodman/files/w24639.pdf

News article about that one at https://www.bbc.com/news/business-44288982

They took cases from 42 different cities. I read through the appendix, but there isn't much on if they tried to account for city location.

I imagine that in the warmer locations (think Arizona, Texas, So. California) immigration judges may get a different mix of cases than they get in NY or in Michigan.

If their result holds if you compare the same judge in the same location around the same type of year (to avoid seasonal migration patterns) on different temperature days, then it would be more convincing to me.

Global warming makes criminals of us all.

Underrated comment.

Not to get sidetracked on the gender wars, but the gender of the judge makes a difference, with temperature-sensitivity particularly pronounced among females, confirming earlier research (page 15). One has to wonder if temperature is a proxy for other factors. For example, crime increases in warm months. Are judges more likely to be affected in their decision-making during warm months because they are more conscious of criminal activity during those months?

Awesome new addition for the spurious correlations web site.

http://www.tylervigen.com/spurious-correlations

"American Economic Journal" explains it all. News: AEJ has not driven out of business "Mad" magazine, yet.

Tyler is trolling us again ...

Did they control for the temperature(s) when they conducted and wrote the study?

My meta analysis says that the quality of the study is inversely proportional to the pleasantness of the day. 50 of the 1,000 studies in the cohort showed a p.05 level of significance.

I'd be more interested to see if there is any actual evidence that judicial decisions are anything but random and arbitrary. Operation Greylord and the New Jersey child detention kickback scheme I suspect are more representative of how courts work than they are outliers.

There’s a reason why big disputes, involving sophisticated parties on both sides, almost always settle rather than seek a decision from a judge or jury. A negotiated settlement is something that both sides agree to. I might not like paragraph 174 and you might not like paragraph 238, but if we’re both willing to sign it that shows it’s mutually beneficial, all considered. Judges and juries are kind of random. Judges will make decisions based on how gentlemanly an expert is. Jurors will invoke all sorts of biases. It’s a mess. The prevalence of settlement does result in a lot of frivolous plaintiffs getting some payout rather than a 90% chance of zero payout and a 10% chance of a windfall, and there is a cottage industry of lawyers who live off that.

Funny enough, most litigants uses the courts for a thorough discovery and disclosure process. That part of the American system works quite well. If you’ve ever dealt with European or Asian systems you’d agree that the contrast is marked.

It's called the hazards of litigation. A common dispute (one that's becoming more common) is the enforceability of restrictive covenants, especially geographic covenants. Even though it's a common dispute, one won't find many court judgments or opinions that address the subject. There are several reasons (the requirement to post bond being one of them and the inability of the person subject to the covenant to get a job being another), but the unpredictability of the result is the primary reason that the parties settle. As one might expect, some judges are favorable to enforcement (those would be the freedom of contract types) while others aren't (those would be the freedom to work types). Freedom can be in the eyes of the beholder.

My nephew will start law school this year. He chose law when he discovered but "you could get paid to argue." Then he discovered that "if you're a judge you win every argument." That is his current goal.

Don't doubt that a lot of judges are those kids grown up.

TBH, you’re mostly paid to have flexible weekend plans, endure tedium, and generally be a human stress ball for your clients and partners to squeeze as needed and/or frivolously.
In all seriousness, people who like to argue (i.e., “win” arguments) generally make terrible lawyers. Litigation is 50% listening attentively, 48% accepting that you’re screwing up something, but you don’t know what.

What IS the optimal thermal range, in which case, for the practice of "rationality"?

How much is the practice of rationality influenced by humidity concurrently?

How are rationality and decision-making influenced by solar heat as distinct from geo-thermal heat?

What causes perspiration: thermal conditions themselves or the challenges of making decisions?

Seen on Twitter:

"Probably the most eye-opening realization I've had as an adult was that most of what I learned as an Economics major wasn't really true. People are actually pretty irrational and markets aren't efficient. Wish we'd talked more about that stuff."

So .. temperature and decisions are part of that, and threatening to anyone who still holds the ancient view of rational agents and efficient markets.

Unsurprising to anyone with a more cynical and biological view that we are pants wearing monkeys, and lucky to behave semi-rationally, part of the time.

And yet they are same people most likely to fancy themselves as choice architects who think that they can nudge and shove other people into making more "rational" choices.

That's not actually inconsistent.

1) Recognize that we make better and worse decisions.

2) Try to make more of the good ones.

Maybe this is cynicism with a dash of optimism.

Who nudges the nudgers?

+1. Quis custodiet ipsos custodes?

We're supposed to trust the bureaucrats because their technocratic expertise makes them largely impervious to such errors. Just look at the wonderful jobs they do!

Every nudge I have ever heard about was based on publicly available research, and explained as it was applied, with the ever present option to opt out.

That paranoia would develop about something that open and that optional is quite mind boggling.

"OMG, they told me retirement savings was good for me, and the default, but I didn't have to!

Help. Help. I am being oppressed."

In that case, turn up the heat!

The authors understand how to get funding:

"How various factors affect judicial decision-making" == No Funding

How GLOBAL WARMING affects judicial decision-making" == Big Funding!

The underlying assumption that judges, at least federal judges, make these sorts of decisions on the spot (rather than a day, if not several days before) is kinda ... questionable.

This study will go along side the study of how judges make more lenient decisions after lunch.

https://www.theguardian.com/law/2011/apr/11/judges-lenient-break

I am waiting for a follow up study on how judges make decisions after a bathroom break.

Dragnet 2019...Friday: It was hotter than hell in Los Angeles on that day, and judges were convicting defendants left and right. Several were sentenced to death involving non-capital crimes.

One way of summarising this could be: "People get grumpy when it's hot outside; immigration judges are people, too."

In the late dialogues of Plato, the inability to attain certain knowledge that Socrates discusses is a kind of symptom of the quest of trying to achieve epistemic certainty. Nonetheless, epistemic certainty seems to be the goal. It is later, with the development of Academic skepticism (lead by Arcesilaus and later Carneades) that Socrates comes to symbolize a teleology, an actual goal of maintaining epistemic modesty in philosophical discourse. This uncertainty becomes seen as more persistent and a part of the nature of human thought, and the dialogue/debate form is meant to rigorously figure out what is more probable or convincing despite the fact that nothing is fully certain. Cicero's Academica is an underrated text, and the Academics are underrated in their contribution to the way we conceive of Socrates now.

The study is probably right that there's a correlation. The problem is that the researchers assume that temperature has a direct effect on judges, rather than the more plausible hypothesis that it has to do with effects on the parties who likely come in out of the weather not long before their hearing -- sweaty immigrant is less appealing than one who appears dry and calm. That might be wrong also; the problem is that there are multiple plausible reasons for the correlation, so no basis to make a causal inference.

Comments for this post are closed