People who commit crimes are already very likely to be young (under 30 for large majority) and have significantly higher than average time discounting. You are proposing a punishment that won't take effect for between 3 to 4 decades (if ever). It is not particularly likely to work at any reasonable amount.
Agreed. We’re talking about people who by definition struggle with tradeoffs and are impulsive. Imposing a penalty with a 40 year delay simply won’t compute
This policy seems like a good way to achieve both higher crime and higher senior poverty
Yes, supposedly the effectiveness of prohibitions depends on the likelihood of detection and the speed of reaction more than the magnitude of the punishment. Delaying punishment by 50 years on a group with high time preference seems doomed to fail. Jail sentences, as ineffective as they may be, at least get violent persons away from potential victims. Arguably, such segregation is their main benefit.
Yes. The US criminal justice system operates mainly on incapacitation, not deterrence. Incapacitation isn't ineffective at all at reducing crime, because most crime is committed by a small group. Once you finally bust a bad guy, give him a long sentence so that he hopefully ages out of his prime felonious years, and crime will be reduced by a lot.
The main disadvantage of incapacitation is that it's costly to society, and -- to the degree you care about them (in my case, very little) -- the criminals. You have to maintain lots of prisons. A lot more life-years are spent locked away than would be necessary with effective deterrence. But in the US, especially after the Warren Court, it's very difficult to deter most kinds of crime because it's difficult to secure convictions compared to many other countries, even in Europe.
"because it's difficult to secure convictions compared to many other countries, even in Europe"
Yes because otherwise how could we imprison more people than the world combined /s. I mean imagine if it were any easier, they only people left not in prison would be those related to the legal system.
Nor is the US based on incapacitation, it's based on spite and harm. We don't jail first time 55 year old offender's five years for reckless felony littering (yes a real thing) to prevent future crime, we do it to harm them to make ourselves feel better because we can't take it out on the guy they can't catch who keyed your car.
Also we don't do convictions anymore. You claim it may be hard to secure convictions but we don't. We (judges ) tell people"You can take a plea deal of a $500 fine with six months probation OR I will give you ten years no parole and FYI the state has a 100% twenty year conviction on this statute". Oh and then send the guy to prison ten years anyways because the next day we have his probation officer intentionally construe a scenario that gets him busted on a technical violation.
>I mean imagine if it were any easier, they only people left not in prison would be those related to the legal system.
This assumes there are not countervailing forces limiting the size of the prison population, but clearly there are or else the prison population would already be unlimited in size.
The size of the US prison system is, I think, mostly a function of voters' tolerance for crime. We expand the prisons until crime is low enough that we're OK with it and we stop prioritizing law and order in our elections. We can't figure out any policies for reducing crime besides increasing sentences and policing, so that is what politicians end up relying on.
Instead, imagine if the prison population were held constant but convictions were easier to secure. In this case, there would be more convictions but shorter sentences, deterrence would be more effective, and crime would be lower, based on everything we know about human psychology: deterrence is far more effective when the punishment is swifter and more probable. Alternatively, we could shrink the prison population and hold crime constant.
This, by the way, is a trade-off: I'm not saying that deterrence is pareto optimal. Better deterrence generally requires a greater likelihood of innocent people being punished (though there are still better and worse ways to design systems).
>Nor is the US based on incapacitation, it's based on spite and harm.
The US system is based on incapacitation, even though not every aspect of the system reflects this, because incapacitation and deterrence are the only effective ways for punishment to reduce crime, and the US is more or less uniquely ineffective at deterrence, at least for a rich country.
But incapacitation and deterrence are effects, and not always consciously pursued ones. As you allude to, retribution is a very real motive that factors heavily into our system, for political reasons. A desire for retribution may happen to enable incapacitation and/or deterrence, but in other cases it may be useless for preventing crime. My go-to example of a useless desire for retribution would be applying child endangerment laws in cases that were clear accidents: no useful deterrence or incapacitation will be accomplished by punishing someone for letting a child die that they very much didn't want to die. While felony littering prosecutions may at times be ill-advised, I don't think they're necessarily useless -- the crime may not warrant the sentence, and there may be better ways to accomplish it, but if you want to reduce littering, then applying harsher penalties to litterers will have that effect.
>Also we don't do convictions anymore. You claim it may be hard to secure convictions but we don't.
1. My sense here is that you're applying an abstract ideal here, instead of looking at the real-world examples of Europe or the pre-Warren-Court US.
2. You're focusing on cases where the prosecutor succeeded instead of cases where he failed, and the far larger number of cases in which he didn't even bother to press charges because he didn't think he could succeed, or didn't have the bandwidth to succeed. An environment where it's difficult to secure convictions may manifest as prosecutors taking less-severe cases that represent easy convictions instead of harsher cases. To use your felony littering example, perhaps you target that guy, whom you have dead to rights and who can't afford a lawyer, instead of a much shakier case against an organized-crime-connected-murderer with a good lawyer.
Europe generally has lower violent crime rates than the US, despite lower prison sentences as punishment. One likely explanation is that many European countries have much higher rates of police officers per capita, and the likelihood of conviction is much, much higher than in many US cities (Chicago has a less than 25% murder clearance rate, for example).
If we increased the likelihood of punishment significantly, we could both reduce the total amount of crime and the sentences we impose on most criminals (a few groups, like child sex offenders, with very high rates of recidivism probably should have long sentences for incapacitation reasons).
Yes, this is the point I was trying to make. Except I don’t think the point about police per capita is true, unless you think there’s a problem with this data (EDIT: my link from Wikipedia isn't working for some reason, but search for "police per capita Wikipedia").
US is on the higher end of large rich democracies according to the table.
Also, to be sure, differences in criminal justice systems aren’t the only driver of differences in criminality between countries; the populations and cultures also make a difference.
That was exactly my thought as well. Further, you couldn’t easily target current benefits either, for as Caplan points out most crime is committed by young men but young men receive relatively little redistribution, so there isn’t much to take away.
The better solution is you're forced to work off the cost of the damage you did. People don't like this because it rhymes with slavery, but it really isn't the same. The other problem is poor work quality, for this there could be independent bodies that assess work quality and if they think you aren't doing a good job, your sentence increases until you do a good job.
That might have been true for some time at the height of the drug war, and may still be true for the federal prison system. But in state prisons, where the vast majority of prisoners are, most people have either committed a violent crime or significant property crime, or both. That may be in addition to drug offenses, or fueled by addiction or motivated by drug sales, but broadly speaking states don't have that many exclusively drug using/possessing criminals in prison anymore.
I'm willing to have my mind changed by evidence, but my intuition is that cutting the social security benefits of people with high time preferences is unlikely to be an effective form of deterrence.
I think this might actually serve as an incentive to commit more crime. "All that'll happen if I steal this cool stuff I want right now is that I don't get some hypothetical future money that I don't care about? Sign me up!"
No, current money as well. He says: "You plan to live fast and die young? Then take away current benefits — SNAP or Medicaid — instead of future benefits."
1) Criminals would be deterred by the loss of benefits decades in the future. Since criminals are known for their forethought and lack of impulsivity, this is bound to work. /s
2) Do-gooders won't persuade the government to restore benefits to criminals when those criminals are old and gray. "Boohoo look at this haggard old man who's forced to work because of a mistake he made when he was young!"
Only people who've lost all contact with the criminal class and life in the street would think of this proposal. It has zero deterrence. Criminals will game the sysytem for any benefits they may want.
An intriguing idea, but I agree with other commenters about the time preference problem: people committing crimes are already low-IQ, high present reward preferring people. Immediate punishments are more in sync with their mentality. Corporal punishment would actually be preferable. I'd recommend some kind of marking too, but criminals do that to themselves already: witness tattoos.
Removing corporal punishment from general practice has been an unmitigated disaster. Both in its vast increase of general lawlessness and the corresponding vast increase of incarceration rates and duration. Given how both cruel and expensive prisons are compared to such things as flogging, and the sheer futility of rehabilitation every time it's been attempted, no honest observer can say it's been worth the moral tradeoff.
There are two different forms of corporal punishment: Punishment which inflicts temporary pain and immediate embarrassment, and punishment which inflicts permanent or at least long-lasting injury. One is too lenient and other too harsh for the intended purpose.
Historically speaking until last week the metric “too harsh” wasn’t a factor in sentencing. For instance the traditional punishment for sheep-stealing was death.
This is total wishcasting on your part. If this were remotely true, why has it been a successful staple of punishment for rulebreaking in every single human society that ever existed until, historically speaking, last week?
In Switzerland, many fines are income based. I've heard that when the CEO of Nestle was fined for speeding on his way to work, the fine covered a huge portion of the canton budget for the next year. Avoiding the higher fine for a repeat justified hiring a professional driver for his commute.
Most criminals are neither savers nor especially hard workers. They are proven by their revealed preferences to an easily earned dollar today to hard work and ten dollars tomorrow.
Cutting current benefits seems more likely to work, but still unlikely. Many criminals do not receive much if anything in the way of entitlements; instead, they leech off of others’ welfare (parents, grandparents, girlfriends, baby mamas). Furthermore, many entitlemente are not granted for the benefit of the adult but their children if any and taking food (stamps) from babes to punish their miscreant fathers is politically indigestible.
A nonstarter across the board but still creative; instead, I suggest that prisons should more or less be converted to labor camps, each inmate’s output measured, and each sentence be made equivalent to some fine. For all but violent criminals (or crimes committed with threat of violence), you can pay a fine or serve your term, in any combination, with your labor output converted to payment towards the fine/time.
This is almost satire of "economist brain"; have you ever met a criminal?! The idea that this would have any deterrence value whatsoever is nonsensical.
And for the habitual offender: instead of starting when you're 172 your pension will start when you're 183. Don't habitual offenders commit a disproportionate percentage of crime?
Interesting idea. But, I would want to take away the welfare state from almost everybody. Including good, well-behaved people. There goes the carrot to take away! 😛
It kinda reminds me of Pierre Poilievre's plan to incentivize municipalities to allow more homes to be built in Canada. If they don't increase issuance of building permits by 15%, federal grants to the municipality get slashed by 15%. But if I was PM, municipalities wouldn't get federal grants to begin with!
Cutting medicaid won't do much to impact poor young men. Either they don't really use it or if they do they will just go to the hospital and get "free" treatment.
“Why isn’t austerian deterrence already widely practiced? The only credible answer is selective squeamishness.” Surely a major reason is that the prison employee unions represent a strong special interest.
"You plan to live fast and die young? Then take away current benefits — SNAP or Medicaid — instead of future benefits."
Most criminals are young men. If they're willing to risk losing years of their young adulthood to prison, or potentially their lives, by engaging in criminal activities, I doubt they'll be overly concerned about losing their food stamps (if they even get any). Nor their medicaid. I just saw an interview with an older con who said most young criminals aren't even aware they have health/dental issues until they go to prison and get the free, mandatory health care.
People who commit crimes are already very likely to be young (under 30 for large majority) and have significantly higher than average time discounting. You are proposing a punishment that won't take effect for between 3 to 4 decades (if ever). It is not particularly likely to work at any reasonable amount.
Agreed. We’re talking about people who by definition struggle with tradeoffs and are impulsive. Imposing a penalty with a 40 year delay simply won’t compute
This policy seems like a good way to achieve both higher crime and higher senior poverty
Yes, supposedly the effectiveness of prohibitions depends on the likelihood of detection and the speed of reaction more than the magnitude of the punishment. Delaying punishment by 50 years on a group with high time preference seems doomed to fail. Jail sentences, as ineffective as they may be, at least get violent persons away from potential victims. Arguably, such segregation is their main benefit.
Yes. The US criminal justice system operates mainly on incapacitation, not deterrence. Incapacitation isn't ineffective at all at reducing crime, because most crime is committed by a small group. Once you finally bust a bad guy, give him a long sentence so that he hopefully ages out of his prime felonious years, and crime will be reduced by a lot.
The main disadvantage of incapacitation is that it's costly to society, and -- to the degree you care about them (in my case, very little) -- the criminals. You have to maintain lots of prisons. A lot more life-years are spent locked away than would be necessary with effective deterrence. But in the US, especially after the Warren Court, it's very difficult to deter most kinds of crime because it's difficult to secure convictions compared to many other countries, even in Europe.
"because it's difficult to secure convictions compared to many other countries, even in Europe"
Yes because otherwise how could we imprison more people than the world combined /s. I mean imagine if it were any easier, they only people left not in prison would be those related to the legal system.
Nor is the US based on incapacitation, it's based on spite and harm. We don't jail first time 55 year old offender's five years for reckless felony littering (yes a real thing) to prevent future crime, we do it to harm them to make ourselves feel better because we can't take it out on the guy they can't catch who keyed your car.
Also we don't do convictions anymore. You claim it may be hard to secure convictions but we don't. We (judges ) tell people"You can take a plea deal of a $500 fine with six months probation OR I will give you ten years no parole and FYI the state has a 100% twenty year conviction on this statute". Oh and then send the guy to prison ten years anyways because the next day we have his probation officer intentionally construe a scenario that gets him busted on a technical violation.
>I mean imagine if it were any easier, they only people left not in prison would be those related to the legal system.
This assumes there are not countervailing forces limiting the size of the prison population, but clearly there are or else the prison population would already be unlimited in size.
The size of the US prison system is, I think, mostly a function of voters' tolerance for crime. We expand the prisons until crime is low enough that we're OK with it and we stop prioritizing law and order in our elections. We can't figure out any policies for reducing crime besides increasing sentences and policing, so that is what politicians end up relying on.
Instead, imagine if the prison population were held constant but convictions were easier to secure. In this case, there would be more convictions but shorter sentences, deterrence would be more effective, and crime would be lower, based on everything we know about human psychology: deterrence is far more effective when the punishment is swifter and more probable. Alternatively, we could shrink the prison population and hold crime constant.
This, by the way, is a trade-off: I'm not saying that deterrence is pareto optimal. Better deterrence generally requires a greater likelihood of innocent people being punished (though there are still better and worse ways to design systems).
>Nor is the US based on incapacitation, it's based on spite and harm.
The US system is based on incapacitation, even though not every aspect of the system reflects this, because incapacitation and deterrence are the only effective ways for punishment to reduce crime, and the US is more or less uniquely ineffective at deterrence, at least for a rich country.
But incapacitation and deterrence are effects, and not always consciously pursued ones. As you allude to, retribution is a very real motive that factors heavily into our system, for political reasons. A desire for retribution may happen to enable incapacitation and/or deterrence, but in other cases it may be useless for preventing crime. My go-to example of a useless desire for retribution would be applying child endangerment laws in cases that were clear accidents: no useful deterrence or incapacitation will be accomplished by punishing someone for letting a child die that they very much didn't want to die. While felony littering prosecutions may at times be ill-advised, I don't think they're necessarily useless -- the crime may not warrant the sentence, and there may be better ways to accomplish it, but if you want to reduce littering, then applying harsher penalties to litterers will have that effect.
>Also we don't do convictions anymore. You claim it may be hard to secure convictions but we don't.
1. My sense here is that you're applying an abstract ideal here, instead of looking at the real-world examples of Europe or the pre-Warren-Court US.
2. You're focusing on cases where the prosecutor succeeded instead of cases where he failed, and the far larger number of cases in which he didn't even bother to press charges because he didn't think he could succeed, or didn't have the bandwidth to succeed. An environment where it's difficult to secure convictions may manifest as prosecutors taking less-severe cases that represent easy convictions instead of harsher cases. To use your felony littering example, perhaps you target that guy, whom you have dead to rights and who can't afford a lawyer, instead of a much shakier case against an organized-crime-connected-murderer with a good lawyer.
Europe generally has lower violent crime rates than the US, despite lower prison sentences as punishment. One likely explanation is that many European countries have much higher rates of police officers per capita, and the likelihood of conviction is much, much higher than in many US cities (Chicago has a less than 25% murder clearance rate, for example).
If we increased the likelihood of punishment significantly, we could both reduce the total amount of crime and the sentences we impose on most criminals (a few groups, like child sex offenders, with very high rates of recidivism probably should have long sentences for incapacitation reasons).
Yes, this is the point I was trying to make. Except I don’t think the point about police per capita is true, unless you think there’s a problem with this data (EDIT: my link from Wikipedia isn't working for some reason, but search for "police per capita Wikipedia").
US is on the higher end of large rich democracies according to the table.
Also, to be sure, differences in criminal justice systems aren’t the only driver of differences in criminality between countries; the populations and cultures also make a difference.
That was exactly my thought as well. Further, you couldn’t easily target current benefits either, for as Caplan points out most crime is committed by young men but young men receive relatively little redistribution, so there isn’t much to take away.
That's why he said: "You plan to live fast and die young? Then take away current benefits — SNAP or Medicaid — instead of future benefits."
I see this more as a cost saving measure than a deterrent
The better solution is you're forced to work off the cost of the damage you did. People don't like this because it rhymes with slavery, but it really isn't the same. The other problem is poor work quality, for this there could be independent bodies that assess work quality and if they think you aren't doing a good job, your sentence increases until you do a good job.
The problem is in most cases there was no harm or damage. You forget most people in prison are there for harmless crimes.
That might have been true for some time at the height of the drug war, and may still be true for the federal prison system. But in state prisons, where the vast majority of prisoners are, most people have either committed a violent crime or significant property crime, or both. That may be in addition to drug offenses, or fueled by addiction or motivated by drug sales, but broadly speaking states don't have that many exclusively drug using/possessing criminals in prison anymore.
I'm willing to have my mind changed by evidence, but my intuition is that cutting the social security benefits of people with high time preferences is unlikely to be an effective form of deterrence.
Good point! I think a lot of criminals are not very forward thinking.
He already addressed this. He says:
"You plan to live fast and die young? Then take away current benefits — SNAP or Medicaid — instead of future benefits."
I think this might actually serve as an incentive to commit more crime. "All that'll happen if I steal this cool stuff I want right now is that I don't get some hypothetical future money that I don't care about? Sign me up!"
No, current money as well. He says: "You plan to live fast and die young? Then take away current benefits — SNAP or Medicaid — instead of future benefits."
This proposal assumes that:
1) Criminals would be deterred by the loss of benefits decades in the future. Since criminals are known for their forethought and lack of impulsivity, this is bound to work. /s
2) Do-gooders won't persuade the government to restore benefits to criminals when those criminals are old and gray. "Boohoo look at this haggard old man who's forced to work because of a mistake he made when he was young!"
Put criminals in jail and keep them there.
Like yourself? You are aware every American has committed numerous felonies in their life.
Only people who've lost all contact with the criminal class and life in the street would think of this proposal. It has zero deterrence. Criminals will game the sysytem for any benefits they may want.
My first thought as well.
This is an ivory tower idea.
An intriguing idea, but I agree with other commenters about the time preference problem: people committing crimes are already low-IQ, high present reward preferring people. Immediate punishments are more in sync with their mentality. Corporal punishment would actually be preferable. I'd recommend some kind of marking too, but criminals do that to themselves already: witness tattoos.
Removing corporal punishment from general practice has been an unmitigated disaster. Both in its vast increase of general lawlessness and the corresponding vast increase of incarceration rates and duration. Given how both cruel and expensive prisons are compared to such things as flogging, and the sheer futility of rehabilitation every time it's been attempted, no honest observer can say it's been worth the moral tradeoff.
There are two different forms of corporal punishment: Punishment which inflicts temporary pain and immediate embarrassment, and punishment which inflicts permanent or at least long-lasting injury. One is too lenient and other too harsh for the intended purpose.
Historically speaking until last week the metric “too harsh” wasn’t a factor in sentencing. For instance the traditional punishment for sheep-stealing was death.
This is total wishcasting on your part. If this were remotely true, why has it been a successful staple of punishment for rulebreaking in every single human society that ever existed until, historically speaking, last week?
In Switzerland, many fines are income based. I've heard that when the CEO of Nestle was fined for speeding on his way to work, the fine covered a huge portion of the canton budget for the next year. Avoiding the higher fine for a repeat justified hiring a professional driver for his commute.
This is already more or less proven not to work.
Most criminals are neither savers nor especially hard workers. They are proven by their revealed preferences to an easily earned dollar today to hard work and ten dollars tomorrow.
Cutting current benefits seems more likely to work, but still unlikely. Many criminals do not receive much if anything in the way of entitlements; instead, they leech off of others’ welfare (parents, grandparents, girlfriends, baby mamas). Furthermore, many entitlemente are not granted for the benefit of the adult but their children if any and taking food (stamps) from babes to punish their miscreant fathers is politically indigestible.
A nonstarter across the board but still creative; instead, I suggest that prisons should more or less be converted to labor camps, each inmate’s output measured, and each sentence be made equivalent to some fine. For all but violent criminals (or crimes committed with threat of violence), you can pay a fine or serve your term, in any combination, with your labor output converted to payment towards the fine/time.
Take a proportion of their stock portfolio.
This is almost satire of "economist brain"; have you ever met a criminal?! The idea that this would have any deterrence value whatsoever is nonsensical.
And for the habitual offender: instead of starting when you're 172 your pension will start when you're 183. Don't habitual offenders commit a disproportionate percentage of crime?
Interesting idea. But, I would want to take away the welfare state from almost everybody. Including good, well-behaved people. There goes the carrot to take away! 😛
It kinda reminds me of Pierre Poilievre's plan to incentivize municipalities to allow more homes to be built in Canada. If they don't increase issuance of building permits by 15%, federal grants to the municipality get slashed by 15%. But if I was PM, municipalities wouldn't get federal grants to begin with!
Cutting medicaid won't do much to impact poor young men. Either they don't really use it or if they do they will just go to the hospital and get "free" treatment.
“Why isn’t austerian deterrence already widely practiced? The only credible answer is selective squeamishness.” Surely a major reason is that the prison employee unions represent a strong special interest.
Your colleague Robin Hanson already came up with the solution to fining judgement-proof defendants: https://www.overcomingbias.com/p/requirelegalliabilityinsurancehtml
"You plan to live fast and die young? Then take away current benefits — SNAP or Medicaid — instead of future benefits."
Most criminals are young men. If they're willing to risk losing years of their young adulthood to prison, or potentially their lives, by engaging in criminal activities, I doubt they'll be overly concerned about losing their food stamps (if they even get any). Nor their medicaid. I just saw an interview with an older con who said most young criminals aren't even aware they have health/dental issues until they go to prison and get the free, mandatory health care.