Samuel Little, a man who confessed killing 93 women over four decades, died in a California prison in late December” (Hannah Knowles, “Deadliest Serial Killer in American History Dies at 80, with Police Still Searching for his Victims,” Washington Post, December 30, 2020). He illustrated in a horrible way what Nobel-winning economist Gary Becker taught us: criminals are rational in the sense that they respond to incentives; those who are not rational don’t stay long on the market.
Becker was awarded the 1992 Nobel Prize in economics for “having extended the domain of economic theory to aspects of human behavior which had previously been dealt with—if at all—by other social science disciplines such as sociology, demography and criminology.” The higher the cost for a given benefit, the fewer crimes will be committed. (Recall that benefits are subjective.)
Ex ante, the expected cost of a crime from the criminal’s viewpoint is given by his likely punishment times the probability of being caught and condemned (assuming he is risk neutral). Of course, he will do anything that can be done at low cost to reduce the second factor in his expected punishment. For Little, an efficient criminal, this meant killing only women who were less likely to be missed by somebody and whose deaths were thus less likely to be successfully investigated. His victims were prostitutes, drug addicts, homeless women… He has not been the only criminal pervert in history to do so: Jack the Ripper comes to mind. The Post writes:
He boasted to investigators of killing with impunity and avoiding “people who would be immediately missed.”
In a terrible sentence, which also reminds us of the obstacles (including occupational licensure and zoning) that American governments have raised against black flourishing in America, Little, who is himself a black man, explained to the Washington Post:
I’m not going to go over there into the White neighborhood and pick out a little teenage girl.
READER COMMENTS
MarkW
Jan 7 2021 at 7:48am
First of all, thanks to the WP for <i>not</i> saying ‘the most prolific serial killer’ — man I hate that every time I see it. At least <i>somebody</i> in the press now apparently knows what ‘prolific’ means.
Also, one of the good things to happen recently is that becoming a ‘successful’ new serial killer in the U.S. is now nearly impossible. Even if their own DNA is not in a database, some of their relatives’ DNA will be and the police will be able zero in on the killer fairly quickly. Decades old cold-cases are being solved on a regular basis now using this approach. In the 21st century, the probability of a serial killer being caught is approaching 100%.
MarkW
Jan 7 2021 at 7:50am
(Some day I’m going to remember that on econlib — unlike almost every other site — you can’t put the html tags in directly)
Michael Stack
Jan 8 2021 at 7:50am
The ability to triangulate these serial killers through family DNA is such an amazing innovation. Imagine being one of these killers who has escaped justice for 20 – 30 years, thinking for SURE you will never be caught.
I’m sure there are many criminals who lay awake at night wondering when their turn might come.
Phil H
Jan 7 2021 at 11:09am
Hmm. I’m not sure that there’s actually much evidence here of what PL suggests. The claim is that the incentives change the behaviour of criminals. In order to establish that claim, we’d have to see how criminals behaved before, and how they behaved afterwards, and then add some causal reasoning. Here we’ve just got an example of a person who happened to be “successful” because the way he committed crime fit certain criteria. There’s not much evidence that he changed. He claims to have been responding to incentives, but I’m not sure I’m willing to take the story of a serial killer at face value.
Jon Murphy
Jan 7 2021 at 12:19pm
Let’s reverse the question: why wouldn’t criminals alter behavior based on their incentives? Why do criminals act differently in the marketplace then in the criminal sphere?
Phil H
Jan 7 2021 at 2:05pm
Absolutely. If you assume the answer that you want, then you’ll be able to reason yourself round to your favoured answer every time.
That’s not empirical science, though.
Jon Murphy
Jan 7 2021 at 2:16pm
Sure. But that’s not what’s going on here. We’re asking the same question, but in two different ways.
Neither of us are assuming a conclusion. By asking the question multiple ways, you get more robust answers. A robustness test, if you will.
So, the answer to your implied question “why do criminals respond to incentives,” is “because they respond that way in every other facet of their lives.” Basic analytical egalitarianism says we must apply the same reasoning to criminal behavior, at least until we can come with an explanation for its deviation.
So, why do increased security systems deter crime? Why do increased lighting deter crime? Why do certain actions deter crime? If it is not “incentive,” what explains their behavior?
Jon Murphy
Jan 7 2021 at 12:20pm
As an empirical note, there is tons of evidence that criminals do respond to incentives. Gary Becker has written a lot on the matter (as Pierre notes), as have many other economists, sociologists, legal scholars, criminologists, and political scientists.
MarkW
Jan 7 2021 at 7:05pm
Now that DNA can be traced by relatives rather than by the criminal directly, will the nature of crime change? What will prospective serial killers do who realize that they aren’t going to be able to get away with more than one or two victims before being caught? I remember watching a fascinating ‘American Experience’ episode about poison and the beginnings of forensic science (which is available here). Poisoning was far more prevalent before the science was available to detect the poisons than it was afterward. Another case of prospective criminals responding to incentives.
AMT
Jan 8 2021 at 5:04pm
TLDR: The evidence shows that criminals are usually not “rational” because they show only minimal response to the magnitude of the sanction.
This is only correct if you accept the highly implausible assumption that criminals are efficient (rational). The evidence is that the vast majority of criminals are irrational (or they are “rational” but have a high discount rate (different “preferences”), though Becker’s argument would still be incomplete). I think it’s very likely that most crime is also committed by irrational criminals, though this would take some calculation of the numbers (technically we need to know how many rational criminals there are and how many more crimes they commit than the irrational ones). “He has not been the only criminal pervert in history to do so,” doesn’t exactly inspire confidence in the rationality of criminals overall. MarkW’s comments also support the argument that rational criminals aren’t going to evade detection too much more effectively than irrational ones.
Also, we may not want to use the extremely heavy sanctions (which would be logical to efficiently deter crime according to Becker’s framework), because of marginal deterrence (incentive not to commit additional or more serious crime), since Becker only used a single-crime model.
See Lee, Savid S. and McCrary, Justin (2005) Crime, Punishment and Myopia
Abstract:
So, from a policy perspective, “the marginal criminal justice dollar is more effectively spent on raising the probability of imprisonment rather than the incarceration length.”
See also Polinsky and Shavell (1979), who say that Becker’s conclusion does not hold if violators are risk averse. They also say that Becker’s preference for monetary sanctions (fines) does not hold where violators are potentially insolvent…which I would say is a pretty vital consideration.
Daniel Kian Mc Kiernan
Jan 7 2021 at 10:42pm
For what it’s worth, Alex Tabarrok at MarginalRevolution.com attempt to refute Becker, but Tabarrok’s objection assumed that economic rationality required a linear utility function.
AMT
Jan 8 2021 at 5:09pm
Yeah, my discount rate is a trillion percent. I’m fully “rational” just with different preferences!!!! lol
Daniel Kian Mc Kiernan
Jan 8 2021 at 10:47pm
Sneering at the concept of economic rationality won’t change the facts that what Becker said was correct, and that you grossly misunderstood not only him but the model at the base of mainstream economics’ treatment of uncertainty.
It is perverse that someone should be writing for an economics ‘blog named “Marginal Revolution” yet think that economic rationality requires constant marginal utility (for that is what linear utility implies).
AMT
Jan 9 2021 at 4:27am
Alex never said (nor did I) that economic rationality requires a PERFECTLY linear utility function. But the obvious point to anyone with half a brain, is that people with ridiculously large discount rates are not rational, as I pointed out. I’m not sneering at the concept of economic rationality, I’m sneering at the unbelievable naivete it takes for anyone to believe ALL people are rational, and especially the violent criminals that commit crimes without any economic gain, as Alex highlighted. I mean just look at the evidence from the articles I cited. Becker has been proven wrong.
Daniel Kian Mc Kiernan
Jan 9 2021 at 6:38am
As everyone who has taken even just an introductory course in economics has been told, the term “economic rationality” and coordinate terms refers exactly and only to ordering properties of preferences. Becker was no more than claiming that a model with those features fitted criminal behavior. Substituting a different meaning of “rationality” in the interpretation of Becker is either incompetent or sophistical.
AMT
Jan 10 2021 at 3:51am
Dan,
No. Economic rationality isn’t “exactly and only” “ordering properties of preferences.” It actually is about an individual maximizing their utility.
It’s really funny that you call others incompetent and try to make yourself sound smart, even though you have no idea what you are talking about. Thanks for the laughs!
Daniel Kian Mc Kiernan
Jan 10 2021 at 6:57am
Evidently, you don’t understand the relationship between preferences and utility.
In modern, mainstream economics, given that weak preferability (the union of strict preference with indifference) is represented by ‘≽’, a utility function is no more or less than a real-valued function u(X), where X is a state-of-the-world, such that
([u(X₁) ≥ u(X₂)] ⇔ [X₁ ≽ X₂]) ∀(X₁,X₂)
In other words, utility is simply a quantitative proxy for preferences, and to say that an agent maximizes utility is no more than to say that the agent acts according to his or her preferences and that a quantitative proxy exists for these preferences. Such proxies will exist if and only if strict preference is asymmetric, weak preference are transitive, and preferences are complete. (Indeed, uncountably many such proxies will then exist.) The first two conditions are rationality constraints; the third is treated by mainstream economics as a rationality constraint.
The original and most general sense of “utility” is that of usefulness, and at the greatest level of generality this might not be quantified, but the economic conception is that the usefulness of a thing to an agent is in addressing the preferences of that agent, so then to say that an agent maximizes utility is no more than to say that the agent acts according to his or her preferences and that these have a maximum (defined as an ordering position).
To simply say that an agent acts according to his or her preferences is essentially vacuous. To claim that they (or a utility function which proxies them) have a maximum is to declare them to have ordering properties under which there is a maximum.
(Turning back to the conception of mainstream economics: To be a utility maximizer is not intrinsically the same thing as to be a maximizer of expected utility, though Becker certainly treated people (including criminals) as such. In case in which agents are expected utility maximizers, there will remain uncountably many utility functions conforming to the preferences of that agent, though each will be an affine transformation of any other.)
AMT
Jan 11 2021 at 2:27am
Keep the laughs coming!
I would agree that utility is essentially the quantification of preferences, so I love how you contradict yourself by explicitly excluding that from your prior definition by saying it was only ordering…preferences (note my emphasis of your mistake with bold font in the earlier post). Nothing else you say is a relevant criticism of what I said or Alex, so I won’t waste time responding, but feel free to keep digging yourself a deeper hole by erroneously attempting to explain things you clearly don’t understand!
Daniel Kian Mc Kiernan
Jan 11 2021 at 4:01am
Well, either you aren’t good at mathematics, or you hope that the audience is not.
Because, as I explained, rationality constraints on the ordering of preferences are both necessary and sufficient for the existence of a utility function or otherwise for maximization of utility, to say that an economically rational agent maximizes utility is perfectly equivalent to saying that the preferences of the agent conform to rules of ordering. There was no self-contradiction in my acknowledgment that economically rational agents would maximize utility; there is just mathematical obliviousness in your thinking that utility offers more than do the ordering properties that I listed.
Actual decision theorists (which include me) will define economic rationality in terms of the ordering properties, because these are primary.
Your sneering declaration of victory is a poor rhetorical strategy.
nobody.really
Jan 8 2021 at 12:04pm
1: I sense Lemieux wants to sound provocative—but realized that the mere suggestion that people respond to incentive, and the criminals count as people, won’t provoke much of a reaction. Thus he tosses in this:
Seems like a stretch, but I guess Lemieux argues that various government policies—policing, licensure, zoning—have disparate impacts by race. Ok, fair enough.
2: But if we want to provoke, ponder this: How SHOULD we design public policy regarding policing in this matter?
Recall, Lemieux observes:
Further recall that, to defend private interests we have created tort law and property law; in contrast, we create criminal law to vindicate societal interests. Further recall that society has finite resources, and thus must choose among the crimes it investigates, prosecutes, and punishes.
Reviewing the incentives created by society, Little picked targets he thinks will not be missed. Does Lemieux propose rejiggering these incentives so that murders will have equal incentives to kill people would WOULD be missed as people who would not? How would that advance society’s interests?
Let the provocation begin…!
Pierre Lemieux
Jan 8 2021 at 2:24pm
@nobody.really: Incidentally—or perhaps not so incidentally—can you explain to me (1) what “societal” means and (2) what “societal interests” mean? I’ll give you two hints: https://www.econlib.org/the-scientific-look-and-feel-of-public-health/ and https://www.econlib.org/library/Columns/y2014/Lemieuxwe.html.
nobody.really
Jan 8 2021 at 2:53pm
I use “societal” here to distinguish from private. I use “societal interests” to refer to interests that government pursues and does not simply leave to private action.
Police and criminal courts pursue a collective interest in deterring crime generally. Clearly, not every member of society might support this objective–criminals, for example. But if the police decline to investigate or prosecute a crime against me, I don’t know what legal recourse I would have against the police. In short, I don’t have a general property right in law enforcement’s exercise of prosecutorial discretion.
Libertarianism often overlaps with anarchy. Do you mean to argue that, because you dispute the idea of a collective interest, you oppose the establishment of police and criminal law?
Comments are closed.