The goal of prisons
Prisons are widely criticized, and for good reason. Crime is rife within prisons, recidivism is high (roughly half of former inmates return to prison), and incarcerating inmates may, at the margin, increase the number of committed crimes. The root of the problem? The prison system doesn’t satisfy good goals.
Selecting a good goal for a system is hard. The goal that a few people have suggested to me is “Prisons should minimize crime”. This goal is bad. It implies that prisons might spend a million dollars to prevent the theft of a candy bar. Good goals account for benefits and costs, and allow an action if and only if it is in society’s interests.
One such goal might be to maximize the total societal contribution of any given set of inmates within the limits of the law (limits such as “Don’t restrict the freedom of inmates after their release”).
Let’s see what kind of prison system satisfies this goal.
The proposed prison system
Prisons enter a second-price auction for each convict’s contract: The highest-bidding prison incarcerates the convict and pays the second-highest bid. After the auction, whenever the convict pays tax, the government passes it on to the prison; whenever he uses a government service, the government passes the cost on to the prison; whenever he commits a crime, the prison must pay the government the “equivalent compensation” of the crime (e.g. a former inmate committing a murder might cost his former prison whatever the value of a human life is).
Note that second-price auctions allow negative bids, which will be necessary to sell the contracts of inmates that are likely to reoffend.
If a former inmate is reincarcerated, their new contract goes up for bid, and the sale price goes to his former prison (rather than the government). If the sale price is negative, the old prison pays the new prison.
Once a convict leaves prison, the prison has no control over the him. He abides by the same tax rules as any non-convict. He can access the same government services. He would not be affected at all if the government suddenly revoked his contract after his release (i.e. the government stopped paying the prison his taxes, and so on). This system affects the convict only by changing what occurs during his incarceration.
Defining “future societal contribution”
Given we want to maximize the societal contribution of all inmates and we can’t change contributions from the past, we want to maximize the future contribution of all inmates. So how do we define “future societal contribution”?
Bob has been convicted of a crime. His future societal contribution () can be split into his future monetary and nonmonetary contributions. His future monetary contribution can be split into
- the government’s cost of funding his incarceration (),
- his tax incidence (),
- the amount of welfare and government subsidies he is granted (),
- the monetary effects of Bob’s future crimes (if any), such as cost of replacement and repair.
We will define equivalent compensation as the payment amount that would make us as well off as we would have been had the crime not occurred. Equivalent compensation for the monetary effects of crimes always exists because those effects are, by definition, a monetary amount. Bob’s future nonmonetary contribution includes the nonmonetary effects of the crimes he will commit. Suppose the equivalent compensation of a crime is equal to the maximum amount that we (as a society) would spend to prevent it. Thus, since we do not have infinite money, equivalent compensation for the nonmonetary effects of a crime must always exist. Therefore the monetary and nonmonetary effects of his future crimes can be summed into a single value ().
Let us ignore Bob’s other nonmonetary contributions for simplicity.
Under these assumptions, Bob’s contribution to society can be represented as a dollar amount:
For our proposed system (with the price of Bob’s contract being ),
The prison’s profit
Suppose that the profit a prison would generate from a given convict is independent from the profit it would generate from its other convicts. Then, if Bob is incarcerated by prison , the prison’s profit from Bob () under any system is the funding it receives for him minus its cost in imprisoning him ():
Under our proposed system, the prison’s profit is found by substituting our previously defined level of funding:
(Not true in the case of bankruptcy: A company can’t make unbounded negative profits. To prevent prisons bidding on negative-sale contracts, declaring bankruptcy, and running away with the money, we probably want capital-rich intermediates to underwrite the prisons.)
The profit-maximizing bid of each prison
Given our assumptions, if prison wants to maximize its total profit (), it must maximize the profit it generates from each of its convicts (all ) simultaneously:
(This is, strictly speaking, not true. The error in the argument is the “profit independence” assumption. But it doesn’t affect anything. Prisons want to maximize their profit over all inmates, and they’ll select a set of inmates that does so, which will tend towards maximizing the total societal contribution of the given set of inmates. We can probably avoid the assumption by having a continuous simultaneous auction of all inmate contracts.)
Let’s suppose that prisons do maximize profits. Thus, each prison bids its profit-maximizing bid (), which, in a second-price auction with non-colluding bidders, is the bid () that would make its profit zero if it paid its bid:
Then, the profit-maximizing bid can be found by making the relevant substitutions:
Note that each prison’s profit-maximizing bid is independent of the other prisons’ bids. Also note that the bid is equal to the sum of society’s and the prison’s benefit:
Sufficient criteria to maximize the contribution
Each inmate’s future contribution to society is maximized if
1) each inmate is allocated to the prison that can create the greatest for that inmate, and
2) all prisons, to their best of their ability, maximize of each of their inmates, and
3) the incarcerating prison’s profit (i.e. its share of ) is zero.
The proposed system satisfies the sufficient criteria
Criterion 1: The inmate is allocated to his best prison
Prison imprisons Bob if and only if its bid, its profit-maximizing bid, is the highest.
The prisons reveal which prison is best just by submitting their bids. Since the profit-maximizing bids happen to equal , the highest bidding prison is the best prison, and Bob gets allocated to it.
I.e. criterion 1 is satisfied.
Criterion 2: Prisons maximize the total contribution
Recall that the winning bidder pays the second-highest bid:
Recall that, barring collusion, a profit-maximizing prison will bid independently of the bids of any other prison: The winning prison cannot affect the second-highest bid, despite that bid being a factor in the prison’s profit. Therefore, the winning prison maximizes its profits by maximizing the sum of the other factors of its profit function, which happen to equal :
I.e. criterion 2 is satisfied.
Criterion 3: The prison’s profit is zero
The winning prison’s profit equals the difference between the total benefit it would provide and that of the second-best prison:
If the best two prisons are equally capable, the profit is zero. I.e. criterion 3 is satisfied.
(Well-run companies can have zero economic profit. I should have included the opportunity costs of prisons and society at the beginning of the proof.)
Poorly estimating equivalent compensation doesn’t always create inefficiencies
We want to rehabilitate if and only if the benefits exceed the costs. The costs are a dollar figure, so when choosing whether to rehabilitate, we need to put a dollar value on the benefits: Estimating each crime’s equivalent compensation is necessary under any well-functioning prison system. Since the estimates are ultimately a normative decision, it’s impossible to solve using empirical data alone. Not that that’s stopped anyone from trying.
Fortunately, our valuations of equivalent compensation do not need to be accurate. To understand this, there are three figures to keep in mind:
- The true value of the crime’s equivalent compensation,
- Our valuation of the crime’s equivalent compensation, and
- The prison’s cost of rehabilitation to prevent that crime.
If the cost of rehabilitation is lower than the true value, we want the rehabilitation to occur. And it does as long as the cost of rehabilitation is lower than both the true value and our valuation. And since prisons earn zero profit, the cost to society is the cost of rehabilitation.
If the cost of rehabilitation is higher than the true value, we don’t want the rehabilitation to occur. And it does not as long as the cost of rehabilitation is higher than both the true value and our valuation.
Inefficiencies occur only when cost of rehabilitation falls between the other two values: Either the rehabilitation will occur when it shouldn’t, or it won’t occur when it should.
Measuring tax contributions and usage of government services
We don’t need to measure everything, only the things that are significant and can be changed by the prisons. For example, we don’t need to measure the past (since it cannot be changed). We don’t need to calculate the wear-and-tear a convict causes to a road’s surface (since it’s not significant). We only want to measure the big things, such as welfare payments, public housing, education and healthcare subsidies, and tax incidence. The amount of welfare a person is paid is relatively easy to get (I’ve written code to retrieve that value from government databases). I don’t know how hard it is to measure the other values, but calculating “Net tax contribution” is essential to solving many policy issues.
Immigration should use a similar mechanism
Same strokes, different folks
When I tell people that prisons and immigration should use a similar mechanism, they sometimes give me a look of concern. This concern is based on a misconception: Using the same mechanism does not imply goods are related in any meaningful way. The transaction mechanism for computers and peanut butter is the same, but that’s the only thing those goods have in common.
Our mechanism applies to any transaction between two parties where either one party doesn’t have a choice in the matter (e.g. inmates), or one side of the transaction is always identical, so one party doesn’t care who is on the other side of the transaction. The latter is true for immigrants: It doesn’t matter who rubber stamps an immigration form.
A decentralized immigration system
Applying our system to immigration is essentially a prediction market for whether any given prospective immigrant will be a positive contributor to society. For our (non-refugee) immigration system, we can copy and paste our prison funding mechanism, with two main differences: We're not imprisoning immigrants, and negative bids are not allowed.
The rest of the system is as you’d expect. Companies enter a second-price auction for the contract of each immigrating family unit. If bids are submitted, the winning bidder pays the second highest bid and the family is allowed into the country. After the auction, whenever the family pays tax, the government passes it on to the company; whenever the family uses a government service, the government passes the cost on to the company; whenever the family commits a crime, the company pays the government the equivalent compensation of the crime.
I'm not sure about the following modifications.
- If the family commits a crime, their contract goes up for auction again, no negative bids allowed. If no one bids on their contract, the family is deported. If there are bidders, the transaction amount (i.e. the second-highest bid) is transferred to the previous holder of the contract (the previous contract holder is allowed to bid, in which case no transaction occurs between companies). This modification reduces the potential downside of allowing an immigrant into the country, which results in more people becoming viable immigrants. It results in bids being higher, which means more money to the government. It also deters immigrants from committing crime.
- If the government wanted to control the number of immigrants, it could demand a fixed payment from the winning bidder of each immigrant’s contract. This would reduce the number of immigrants that companies would be willing to accept.
- We might also want to allow public donations to act as a decentralized valuation of a potential immigrant’s desirable nonmonetary contributions to society.
EDIT: A cheap(ish) test
The data for income tax paid, welfare used, and crime data already exists. This data, along with valuations of crimes, could be linked to the inmates’ attributes, which would allow the prisons to have good reference classes from which they could estimate their profit-maximizing bids.
1. Randomly select enough inmates to fill US prisons. Randomly split that population in half: one half in the control group, the other half in the experiment group.
2. Allow any prison to bid on the contracts of the experiment group. The participating prisons must bid on all of those inmates’ contracts.
3. Select prisons from the group of bidders in such a way that those prisons are filled, and that the sum of the bids is maximized. Because all prisons are bidding, you get a thick(ish) market, and the assumptions of my argument should hold (you may have to compensate prisons who make reasonable bids to incite participation).
4. Fill a random selection of remaining prisons with the inmates in the control group. The selection bias introduced is intended: In my proposal, the best prisons self-select by making higher bids (criterion 1).
5. Measure the outcomes.
Note: The images in your post are broken, likely for everyone but you:
Google Drive generally doesn't work very well as an image hosting service. I recommend using imgur, or any of the other dozens of image hosting services out there.
One level up, I apologize for it not being easy to just upload images to the EA Forum and LessWrong. We have a new editor in beta that I expect to go live on the EA Forum soon that would allow you to just upload images without having to worry about image-hosting services.
As an offtopic aside, I'm never sure how to vote on comments like this. I'm glad the comment was made and want to encourage people to make comments like this in future. But, having served its purpose, it's not useful for future readers, so I don't want to sort it to the top of the conversation.
Oh okay, thanks for the advice. I'll see if I can get it to work.
First, two rather technical notes: in your graphics you have areas represent both positive (assuming the future contribution might be positive) and negative figures, which is confusing IMO. If the second figure is to represent net contribution, maybe state that explicitly. Also, I don't understand one of the formulae, namely
I thought ΠBob=MBob+NBob and I don't see why NBob=−FundingBob−CrimesBob or did I miss something?
Then, though I get your approach's theoretical appeal, I don't see why it should be be superior compared with other possible interventions. I don't whether we can predict very well whether a former immate will commit new crimes after their release and how that probability can be reduced. Since private prisons won't be able to operate very well without that knowledge base, why not prioritize research into those topics, before undertaking a difficult structural reform? This research and implementing its finding directly would also be much easier to sell to politicians, particularly on the left.
The graphs show what is encapsulated by what. The area to which a label corresponds is the smallest convex shape that encapsulates the label. For example, M is the whole lower-left quadrant, which also encapsulates the monetary effect of crimes (which is why the monetary effect of crimes is not explicitly included in the formulas). M doesn't stand for all monetary factors. It stands for every monetary factor except Funding.
If the convict pays tax, that's a good thing for society (all else being equal). Π should increase. And it does, since more tax means a higher M. If the convict has to use welfare, that's bad. Π should decrease. And it does, since you get a lower M. If the convict's incarceration requires more funding, that's bad (all else being equal). Π should decrease. And it does, since funding is subtracted. And so on. There is part of the graph that is not included in Π: The nonmonetary effects that are not crimes. One of my simplifying assumptions was to ignore this section ("Let us ignore Bob’s other nonmonetary contributions for simplicity.")
This system has an advantage over public prisons in that it provides a mechanism to choose which research should be pursed. Should we trial inmates wearing pink uniforms? Is that worth the cost of research or not? I don't know. But there are people who are informed enough to be willing to make a bet on the matter. The people who believe strongly that they can get good outcomes will make those bets. If they're wrong, they lose money and leave the market. If they're right, they make money and gain a greater share of control.
One thing I want to note: I'm not saying "Implement the system as I've described by next month". I think the system is something to carefully work towards.
Thank you for the clarifications.
According to Peter's comment, there already seem to be many informed people around working both inside and outside the prison system. Maybe it would be sufficient incentivize them better to make those bets, by introducing premiums for prisons that reduce the number of reconvictions of their previous inmates, taking into account some priors how likely they were to recidivate based on what their crime was and socio-economic background or so. One could also try to increase their agency if needed, I mean, letting public officials make decisions without having to worry too much about protocol or having to obtain permission from elected superiors who might want to take a tough stance against crime, letting researchers pursue any research they think is promising. Given the number of national and sub-national prison systems a lot of different insights would potentially result from that, which then could be shared and produce large benefits - especially if you pay the sharers.
Maybe the private system would still be more effective, but I am unsure by how much and in that case still my case holds up, I think, that you could make political progress much faster when you pursue a less radical idea - and the potential downsides would be lower.
There is a lot of research done in forensic psychology/psychiatry as to which offenders have which rates of reoffending (and how that rate can be reduced). There are instruments like the HCR20, the Psychopathy Checklist, the SVR20, etc. In Germany there are nice statistics about which released groups do commit the same/different crimes with which rates of recidivism. I am pretty sure, other countries have comparable stats. The rates are well below 50%, but we only have some 80 persons out of a hundred thousand behind bars, in the united states that number is nearly ten times higher, so people get faster into those (mostly private?/ profit orientated?) prisons. My guess would be: get fewer people into prisons (than in the united states), get them therapy if needed (for example reasoning and rehabilitation aka r&r) and most important: give them good aftercare (possibilities).
Ah, I think we've both made the same mistake (believing recidivism rates were similar across countries). It appears recidivism has quite a large range.
"For all reported outcomes, a 2-year follow-up period was the most commonly used. The 2-year rearrest rates ranged from 26% (Singapore) to 60% (USA), two-year reconviction rates ranged from 20% (Norway) to 63% (Denmark), and two-year reimprisonment rates ranged from 14% (USA – Oregon) to 43% (Canada – Quebec, New Zealand) (see Table 3 for 2-year rates from included countries)."
In any case, my argument doesn't hinge on what the true statistics are.
My instinctive emotional reaction to this post is that it worries me, because it feels a bit like "purchasing a person", or purchasing their membership in civil society. I think that a common reaction to this kind of idea would be that it contributes to, or at least continues, the commodification and dehumanization of prison inmates, the reduction of people to their financial worth / bottom line (indeed, parts of your analysis explicitly ignore non-monetary aspects of people's interactions with society and the state; as far as I can tell, all of it ignores the benefits to the inmate of different treatment by different prisons).
This is in a context where the prison system is already seen by some as effectively legal slavery by the back door, where people believe that (for example) black people have been deliberately criminalized by the war on drugs as part of an explicit effort to exploit and suppress them, and where popular rhetoric around criminals has always been eager to treat them as less than human.
This perspective explicitly doesn't address the specific content of your proposal, but I think there's a few reasons why it's important to pay attention to it:
I'm posting this as a "frame challenge", but I think I also have some critiques of your model on its own terms, which I'll post as a separate comment.
No one is going to run a prison for free--there has to be some money exchanged (even in public prisons, you must pay the employees). Whether that exchange is moral or not, depends on whether it is facilitated by a system that has good consequences. I think a worthy goal is maximizing the societal contribution of any given set of inmates without restricting their freedom after release. This goal is achieved by the system I proposed (a claim supported by my argument in the post). Under this system, I think prisons will treat their inmates far better than they currently do: allowing inmates to get raped probably doesn't help maximize societal contribution. "Commodification" and "dehumanization" don't mean anything unless you can point to their concrete effects. If I've missed some avoidable concrete effect, I will concede it as a good criticism.
Not every desirable thing needs to be explicitly stated in the goal of the system: Good consequences can be implied. As I mentioned, inmates will probably be treated much better under my system. Another good implicit consequence of satisfying stated goal, is that prisons will only pursue a rehabilitative measure if and if it is in the interests of society (again, you wouldn't want to prevent the theft of a candy bar for a million dollars).
I account for the nonmonetary aspects of the crimes. But yes, the rest is ignored. If this ignored amount correlates with the measured factors, this is not really an issue.
In the predominant popular consciousness, this is not sufficient for the exchange to be moral. Buying a slave and treating them well is not moral, even if they end up with a happier life than they otherwise would have had. Personally, I'm consequentialist, so in some sense I agree with you, but even then, "consequences" includes all consequences, including those on societal norms, perceptions, and attitudes, so in practice framing effects and philosophical objections do still have relevance.
Of course there has to be an exchange of money, but it's still very relevant what, conceptually or practically, that money buys. We have concepts like "criminal law" and "human rights" because we see benefits to not permitting everything to be bought or sold or contracted, so it's worth considering whether something like this crosses one of those lines.
I agree that seems likely, but in my mind it's not the main reason to prevent it, and treating it as an afterthought or a happy coincidence is a serious omission. If your prison system's foundational goal doesn't recognize what (IMO) may be the most serious negative consequence of prison as it exists today, then your goal is inadequate. Indirect effects can't patch that.
As a concrete example, there are people that you might predict are likely to die in prison (e.g. they have a terminal illness with a prognosis shorter than their remaining sentence). Their expected future tax revenue is roughly zero. Preventing their torture is still important, but your system won't view it as such.
Now that I'm thinking about it, I'm more convinced that this is exactly the kind of thing people are concerned about when they are concerned about commodification and dehumanization. Your system attempts to quantify the good consequences of rehabilitation, but entirely omits the benefits for the person being rehabilitated. You measure them only by what they can do for others – how they can be used. That seems textbook dehumanization to me, and the concrete consequence is that when they can't be used they are worthless, and need not be protected or cared for.
No, this consequence was one of my intentions. It was not an afterthought. Not every goal needs to be stated, they can be implied.
...by the convict's own free will. And just because that's the only thing being measured, doesn't mean I'm disregarding everything else. Societal contribution and a person's value are different things: A person who lives separately from society has value. But I don't know how to construct a system that incorporates that value.
This is a misunderstanding of the policy. Crimes that occur within prison must be paid for, so the prisons want to protect their inmates.
This is a good point. Maybe they should be put in a public prison.
Possibly a tangent, but I think it's maybe relevant that QALYs do not have that problem.
That's a good point. You could set up the system so that it's "societal contribution" + funding - price (which is what it is at the moment) + "Convict's QALYs in dollars" (maybe plus some other stuff too). The fact that you have to value a murder means that you should already have the numbers to do the dollar conversion of the QALYs.
I'm hesitant to make that change though. The change would allow prisons to trade off societal benefit for the inmate's benefit, who, as some people say, "owes a debt to society". Allowing this trade-off would also reduce the deterrence effect of prisons on would-be offenders, so denying the trade-off is not necessarily an anti-utilitarian stance.
And denying the trade-off doesn't mean the inmate is not looked after either. There's a kind of... "Laffer Curve" equivalent where decreasing inmate wellbeing beyond a certain point necessarily means a reduction in societal contribution (destroying an inmate's mind is not good for their future societal contribution). So inmate wellbeing is not minimized by the system I've described (it's not maximized either).
I'm not 100 percent set on the exact funding function. I might change my mind in the future.
Thanks for your engagement!
Agreed, but at least in theory, a model that takes into account inmate's welfare at the proper level will, all else being equal, do better under utilitarian lights than a model that does not take into account inmate welfare.
This may be an obvious point, but I've made this same mistake ~4 years ago when discussing a different topic (animal testing), so I think it's worth flagging explicitly.
Please feel free to edit the post if you do! I worry that many posts (my own included) on the internet are stale, and we don't currently have a protocol in place for declaring things to be outdated.
What if the laws forced prisons to treat inmates in a particular way, and the legal treatment of inmates coincided with putting each inmate's wellbeing at the right level? Then the funding function could completely ignore the inmate's wellbeing, and the prisons' bids would drop to account for any extra cost to support the inmate's wellbeing or loss to societal contribution. That's what I was trying to do by saying the goal was to "maximize the total societal contribution of any given set of inmates within the limits of the law". There definitely should be limits on how a prison can treat its inmates, even if it were to serve the rest of society's interests.
But the more I think about it, the more I like the idea of having the inmate's welfare as part of the funding function. It would avoid having to go through the process of developing the right laws to make the prison system function as intended, and it's better at self-correcting when compared to laws (i.e. the prisons that are better at supporting inmate welfare will outcompete the prisons that are bad at it). And it would probably reduce the number of people who think that supporters of this policy change don't care about what happens to inmates, which is nice.
I appreciate the effort that went into this post, and the use of actual math to describe the benefits of the proposed system. I also really like that you took feedback from the comments and edited the post -- not enough people do that!
That said, this reads to me as a proposal for massive structural change with no discussion of moderate/feasible reforms that could get us closer to a system with some of these benefits.
As a moderator, I think posts like this are fine. As a reader of the Forum, I prefer that political discussion on the Forum either:
a) Note clearly that a post is meant to be speculative (e.g. MacAskill's Age-Weighted Voting), or
b) Engage with current political systems by discussing proposals that have existing support, or that at least have some chance of being implemented piece-by-piece without an enormous overhaul of an entire societal institution (I'd say the same had this piece been about education, military spending, etc.)
I think Eliezer Yudkowsky's recent piece on police reform does this well; some of his ideas are more realistic than others, but quite a few could reasonably be proposed as legislation within weeks if a congressperson were to write a bill. And while he doesn't cite the people who side with him on some of these proposals, I'm aware that those people exist.
I'd have been more convinced by the post if it referred to any existing policy which mirrors any aspect of these proposals. Does any country in the world create estimates of individual citizens' cost to social services? Does any country in the world have a system where companies can bid for the right to collect individuals' future tax revenue? Has anyone else (politicians, researchers, etc.) ever argued for a system resembling this one?
(I wouldn't be surprised if the answer to that last question were "yes," but I don't know who's made those arguments or whether they ever made any legislative progress.)
In one comment, you note:
I'd have loved to hear thoughts in the post on how we might "carefully work towards" a system that works so differently from any that (AFAIK) exist in the world today. What intermediate steps get us closer to this system without creating a full transition?
I'm not aware of any "net tax contribution" measurement, but I haven't done an extensive search either. I'm not aware of anyone arguing for anything close to the system I proposed. The closest (but still far away) system that I've heard of is social impact bonds, which have been implemented in Australia to some degree. In the implementations of them that I've seen, they give prisons bonuses for reaching a low level of recidivism.
There's several weaknesses of that model. Maybe one prison happens to get allocated inmates that are not likely to reoffend, in which case, they get paid for no reason (bidding on inmate contracts stops this). A reoffending murderer is given the same weight as a reoffending thief, so the prison is indifferent to who they rehabilitate (valuing "equivalent compensation" of crimes stops this). And the government is not cost minimizing (again, bidding stops this). It also doesn't incentivize prisons to increase the net tax contribution of the convict (whereas mine does).
I have some ideas, but no strong theoretical reasons for believing they're ideal.
You could implement a subset of the system: Just the income tax paid and welfare used. That data already exists. This data could be linked to the inmate's attributes, which would allow the prisons to have good reference classes from which they could estimate their profit-maximizing bids. The primary difference between this half-measure and the full proposal is that the prisons don't pay the government when a crime is committed (secondarily, income tax paid is different from tax incidence).
When you have the right metrics for the full system (valuations of equivalent compensation for all crimes, and tax incidence) if, for some reason, tax incidence data can't be generated for the past, you could keep the half-measure system in place for a few decades. But during that time, you do the measurements and store the data. When it's time to implement the full system, again, prisons have good reference classes from which they can estimate their optimal bids.
If politicians are particularly skittish on the idea, they could say to the prisons "We'll pay you 80% of what we would under the current scheme, and pay you 20% of what we would under the new scheme." I wouldn't recommend this, because some viable rehabilitative measures won't be taken.
I think the best argument for conservativism is "Out of all possible sets of institutions, the one we're in is pretty good". But this doesn't mean "never implement radical change" (if you never do so, you'll get stuck on a local optimum). It just means that when you're implementing radical change, you have to have good evidence. I wouldn't implement this system based off the mathematical justification I've given here. But I think the problems with the argument could be solved with minimal changes to the actual system. And if one of the assumptions of the argument fails when the system is implemented, we can always change back to the existing system. But if it is successful, it has implications for other policy areas (such as immigration).
An example of radical, evidence-supported policy is the kidney-exchange reform. Rather than exchange kidneys between pairs of people, they now chain groups of people together. The new system facilitates far more exchanges between willing donors and people in need. Countries that waited for the "evidence" (by which they mean "empirical evidence") let people die because they didn't value the logical argument.
The hospital-intern matching algorithm is another example (it was also applied to student-to-school allocation).
The most prominent system that's supported on a theoretical-level is free markets. It is supported by the general equilibrium model. When the assumptions that model stray too far from reality, the system breaks down. When they're close enough, it works very well.
So the question is "Are the assumptions of the prison-policy argument close enough to reality". I think they mostly are, and where they're not, the error is balanced by a countervailing error.
Robin Hanson has written about similar ideas:
I emailed Robin Hanson about my immigration idea in 2018. His post was in 2019. But to be fair, he came up with futarchy well before I started working on policy.
Doing things in proportion (rather than selling the full value) undervalues the impact of good forecasts. Since making forecasts has a cost, proportional payment (where the proportionality constant is not equal to 1) would generate inefficient outcomes: Imagine the contribution of the immigrant is $100 and it costs $80 to make the forecast, then paying forecasters anything less than 80% will cause poor outcomes.
A good post with interesting ideas.
I think it is however worth flagging to the readers that this is a presumably a US centric post. My understanding is that in the UK at least our private prisons system performs well (and out-performs the public prisons).
I expect a lot could be learned by looking at countries that do private prisons well.
That's really interesting - do you have any recommended reading on the UK system?
I do not have a lot of information on it. Maybe start with: https://www.instituteforgovernment.org.uk/our-work/policy-making/government-outsourcing
I think it mostly comes down to having good contracts in place between the prisons and the state so that the prisons have the correct incentives. I do not have a good knowledge of how the contracts work.
I think the contracts in place are temporary and need regular renewal. If a contract for a private prison is not renewed after x years then the building and management will revert to state ownership.
I believe there are both state-run and private run prisons in the UK. They are compared and in some sense this acts as a check and balance because of one system is working much worse than the other it drives change.
I note that prisons are private but that parole services are state-run (or at least reverting to be state run as they did not work privatised).
Hope that helps
Thanks for the kind comment.
My guess is that the US would be the best place to start (a thick "market", poor outcomes), but I'm talking about prison systems in general.
I'm not familiar with the UK system, but I haven't heard of any prison system with a solid theoretical grounding. Theory is required because we want to compare a proposed system to all other possible systems and conclude that our proposal is the best. You want theoretical reasons to believe that your system will perform well, and that good performance will endure.
Most systems can work well in the short run, but that doesn't mean they're good. For example, if I were ruled by a good king, I still wouldn't want to have a monarchy in place. If the UK system currently works well, I suspect that you have good regulators who are manually handling the shortcomings of the underlying system.
THEORY AND EVIDENCE
Correct me if I am wrong but you seem to be implying that the "theoretical reasons" why a policy idea will work are necessary and more important than empirical evidence that a system has worked in some case (which may be misleading due to confounding factors like good people).
If so I strongly disagree:
I have not considered all the details but I do think you have a decent policy idea here. I would be interested to see it tried. I would make the following, hopefully constructive, suggestions to you.
Focus on countries where the prison system is actually broken
There is a lot of failings in policy and limited capacity to address them all. I do think "if it is not broke don’t fix it" is often a good maxim in policy and countries with working systems should not be the first to shift to the system you describe.
Be wary of the risk of motivated reasoning
Nothing you said substantiates this claim and from what I know about the UK system (which is admittedly minimal) I don’t think this is the case. Now this claim might be true and you might have good evidence for it that you didn’t state, but it did raise a red flag in my mind when I red it.
Don’t under-value evidence, you might miss things. An underplayed strength of your case for fixing private prisons is that the solution you suggest is testable. A single pilot prison could be run and data collected and lessons learned. To some degree this could even be done by an committed entrepreneur with minimal government support.
Look at what can be learned from systems that work elsewhere. Eg a feature of the UK system is that there are both state-run and private prisons. These can and have been compared and can and have acted as a check on each other. If one is clearly failing it motivates change in the other. This learning can make you case for trailing auction based prisons stronger as you can highlight how different systems running in parallel act as a check on each other. Yet at the same time this learning also makes the case for running 100% private auction based prisons weaker as maybe some amount of state-run prisons can provide a useful check on the system.
Hope that helps.
Good theoretical evidence is "actual evidence". No amount of empirical evidence is up to the task of proving there are an infinite number of primes. Our theoretical argument showing there are an infinite number of primes is the strongest form of evidence that can be given.
That's not to say I think my argument is airtight, however. My argument could probably be made with more realistic assumptions (alternatively, more realistic assumptions might show my proposed system is fundamentally mistaken). My model also just describes an end state, rather than provide any help on how to build this market from scratch (implementing the system overnight would produce a market that's too thin and inefficient).
Theory can go wrong if the assumptions are false or the inferences are invalid. Both of these errors happen all the time of course, and in subtle ways too, so I agree that empirical evidence is important. But even with data, you need a model to properly interpret it. Data without a model only tells you what was measured, which usually isn't that informative. No matter what the numbers are, no one can say the UK prison system is better than that of the US without some assumptions. And no one can get data on a system that hasn't been tried before (depending on the reference classes available to you).
Can you show me a theoretical model of school building that would convince me that it would work when it would, in fact, fail? I don't think I would be convinced. (How can you be sure the teachers are good? How can you be sure good teachers will stay? How can you be sure students will show up?) You can't bundle all theory in the same basket (otherwise I could point to mathematics and say theory is always sound). Whether a theory is good evidence hinges on whether the assumptions are close enough to reality.
The process of building the argument changed what I believed the system should be. I have no dog in this fight. I claimed "If the UK system currently works well, I suspect that you have good regulators who are manually handling the shortcomings of the underlying system" for several reasons:
A pilot prison wouldn't work because it wouldn't have competitive bidding. I did mention in Aaron Gertler's comment what data should be collated prior to implementation. I'm all for looking at the relevant data.
That doesn't seem like a good system.
The bidding process and the actualisation of losses (tied to real social interests) keep the prisons in check.
I strongly disagree. Additional checks and balances that prevent serious problems occurring are good. You have already said your system could go wrong (you said "more realistic assumptions might show my proposed system is fundamentally mistaken") and maybe it could go wrong in subtle ways that take years to manifest as companies learn how they can twist the rules.
You should be in favour of checks and balances, and might want to explore what additional systems of checks would work best for your proposal. Options include: A few prisons running on a different system (eg state-run). A regulator for your auction based prisons. Transparency. The prisons being on 10 year loans from the state with contacts the need regular renewing so they would default to state ownership. Human rights laws. Etc. Maybe all of the above are things to have.
As an example, one thing that could go wrong (although it looks like you have touched on this elsewhere in the comments) is prisons may not have a strong incentive to care about the welfare of the prisoners whilst they are in the prison.
Ah yes, I have mentioned in other comments about regulation keeping private prisons in check too. I should have restated it here. I am in favour of checks and balances, which is why my goal for the system contains "...within the limits of the law". I agree with almost everything you say here (I'd keep some public prisons until confident that the market is mature enough to handle all cases better than the public system, but I wouldn't implement your 10-year loan).
Yep, I'm all for that. One thing that people are missing is that the goal of a system kind of... "compresses" the space of things you want to happen. That compression is lossy. You want that goal to lose as few things as possible, but you will lose some things. To fix that, you will need some regulation to make sure the system works in the important edge cases.
This is incorrect. They do have a strong incentive, since the contract comes into effect immediately after the auction: If a crime happens in their prison, the prison has to pay the government. The resulting problem is that prisons have an incentive to hide these crimes. So I recommended that prisons be outfitted with cameras and microphones that are externally audited.
Looks like we are mostly on the same page. We both recognise the need for theoretical data and empirical data to play a role and we both think that you have a good idea for prison reform.
I still get the impression that you undervalue empirical evidence of existent systems compared to theoretical evidence and may under invest in understanding evidence that goes against the theory or could improve the model. (Or may be I am being too harsh and we agree here too, hard to judge from a short exchange like this.) I am not sure I can persuade you to change much on this but I go into detail on a few points below.
Anyway even if you are not persuaded I expect (well hope) that you would need to gather the empirical evidence before any senior policy makers look to implement this so either way that seems like a good next step. Good luck :-)
TO ADDRESS THE POINTS RAISED:
Firstly, apologies. I am not sure I explained things very well. Was late and I minced my words a bit. By "actual evidence" I was trying to just encompass the case of a similar policy already being in place and working. Eg we know tobacco tax works well at achieving the policy aim of reducing smoking because we can see it working. Sorry for any confusion caused.
A better example from development is microcredit (mirofinace). Basically everyone was convinced by the theory of small loans to those too poor to receive finance. The guy who came up with the idea got a freking Nobel Prize. Super-skeptics GiveWell used to have a page on the best microcredit charity. But turns out (from multiple meta-analyses) that there was basically no way to make it work in practice (not for the worlds poorest).
Blanket statements like this – suggesting your idea or similar is the ONLY way prisons can work still concerns me and makes me think that you value theoretical data too highly compared to empirical data. I don’t know much about prisons systems but I would be shocked if there was NO other good way to have a well managed prison system.
I still think it could help the case to think about how a pilot prison could be made to produce useful data. Could the prison bid against the state somehow? Or could it work with two prisons? Or one prison and two price streams?
I was against microfinance, but I also don't know how they justified the idea. I think empirical evidence should be used to undermine certain assumptions of models, such as "People will only take out a loan if it's in their own interests". Empirically, that's clearly not always the case (e.g. people go bankrupt from credit cards), and any model that relies on that statement being true may fail because of it. A theoretical argument with bad assumptions is a bad argument.
That's not what I said. I said "Most systems can work well in the short run", but systems that don't internalize externalities are brittle ("For a system to work well and be robust"). Internalizing externalities has a very strong evidence base (theoretical and empirical). If anyone can show me a system that internalizes externalities and doesn't look like my proposal, I will concede.
I think we agree that you want "cheap and informative tests". Some of the data you're talking about already exists (income tax data, welfare payments, healthcare subsidies), which is cheap (because it already exists) and informative (because it's close to the data prisons would need to operate effectively).
Social impact bonds for prisons are already in place, and that system is similar to mine in some respects (though it has poor, ad hoc justification, and so the system should fail in predictable ways). You're right about me not being interested those systems. Social impact bonds are probably the best reference class we have for my proposal. But if they failed, I wouldn't update my beliefs very much since the theory for the impact bonds is so poor.
Hmm. Actually, you're right, you can make a small trial work.
1. Randomly select enough inmates to fill n US prisons. Randomly put 50 percent of those inmates in set A, and the other inmates in set B.
2. Allow any prison to bid on the contracts of set B. The participating prisons have to bid on all of those inmates' contracts.
3. Select the n/2 prisons in such a way that those prisons are filled, and that the sum of the bids is maximized. Because all prisons are bidding, you get a thick market, and the assumptions of my argument should hold (you may have to compensate prisons who make reasonable bids).
4. Select n/2 random prisons to take on set A. Does this introduce selection bias? Yes, and that's exactly the point. In my proposal, the best prisons self-select by making higher bids (criterion 1).
I'm interested to hear what you think.
Unfortunately I don’t have much useful to contribute on this. I don’t have experience running trials and pilots. I would think through the various scenarios by which a pilot could get started and then adapt to that. Eg what if you had the senior management of one prison that was keen. What about a single state. What about a few prisons. Also worth recognising that data might take years.
I used to know someone who worked on prison data collection and assessing success of prisons, if I see her at some point I could raise this and message you.
Of course, prisons also serve an important deterrent function, which is not well-addressed by this model.
That's a good point to bring up. There are a few ends that other people assign to prisons that come to mind: rehabilitation, deterrence, punishment, and removing the criminal from the population (protecting innocents). However, some of these goals can be achieved by other systems. The death penalty is completely compatible with the system I proposed: Though you may disagree with killing criminals for other reasons, it is (at least on the face of it) a deterrent, and it doesn't need to be carried out by prisons. The law could specify ways in which the prisons must treat their inmates. For example, it could forbid prisons from providing computer access.
If the punishments are not dictated by law, they are the ad hoc decisions of the prison warden (or the decisions of the other inmates).
I'm very interested in this topic, but found the framing here difficult to follow. (I think in part, as Johann pointed out, because positive and negative impacts are combined into one graph.) As a person who hasn't spent much time reading about math and related fields, equations like "ΠBob=MBob−FundingBob−CrimesBob" don't make a lot of sense to me and require a bunch of scrolling around to remind myself what "ΠBob" is.
It might be easier to understand if a plainer-language summary were included under each equation.
Sorry about the confusion. I hope the new notation makes it easier. (I've removed the graphs.)
I'll suggest that some people's concerns are due to an accurate intuition that your proposal will make it harder to hide the resemblance between prisons and immigration restrictions. Preventing immigration looks to me to be fairly similar to imprisoning them in their current country.
No, I think they imagined I wanted to imprison all immigrants (which, as you can see, is not what I'm suggesting). To be clear, I'm not talking about preventing anyone from leaving your country; I'm talking about how to select which non-citizens can become permanent residents in your country.
As for non-open borders being equal to imprisonment, that's incorrect. The fact that I can't live in North Korea does not mean that I'm imprisoned in my country. By this definition of imprisonment, everyone is the world is imprisoned.
I believe this system would allow for more immigrants than there are currently. Government is slow at determining which people can enter (at least in my country). This might fix that. And knowing that every immigrant family is a positive contribution to the government's balance sheet (and unlikely to commit crimes) will probably help society see immigration as a good thing, which may help gain political support for more immigration.
Sorry, this is just a general comment. And it is only an opinion. I don‘t like the idea of profit-orientated prisons. The aim of a reform might be „good prisons“, „good treatment“, but that‘s not where the money is/get‘s generated? Low reoffending rates, espacially for crimes like murder, sexual assault, etc. are „producing“ the profit. I am afraid, people will find cheaper (but not necessarily better ways) to make the profit. For example: if some of my clients ever stands in court again: I will pay them a really good lawyer: I loose money, if they get convicted... / I will make a really good group-therapy, on how not to get caught and how to act with the police. I‘d have an incentive to work (bribe) the police/ the jury ... I‘d have an incentive to kill my high-risk-people shortly after releasing them, espacially, if their situation worsens... („incentive“ might easily mean millions).
I‘m sure one could get past some of those objections... but not past all of them.
I‘d rather like a public prison system, good statistics and a competition for being a good prison, without going to those extremes. That would mean less competition, sure.
Still, thx for your post, I am extremely interested to more comments and your replies.
Yep, those perverse incentives that you identified are all good criticisms. If there's a theoretical model that says why a system will work, the real-world failure points of that system will be the assumptions of its model. The assumptions can be made to be true with the right regulations. My model assumes that prisons will act lawfully, which I think they will under the right punishments (since there's always a possibility of being caught).
I knew about the prison's incentive to murder high-risk inmates, but I didn't consider the others you mentioned. Maybe some activities should be illegal, such as providing inmates with lawyers, but I'd wait and see how that plays out in the real world before banning it. There's one big problem that you missed: under-reporting of crime (e.g. drug use, rape) within prison (remember prisons have to pay the government for each crime after the auction). To prevent under-reporting, I'd consider mandating that each prison puts microphones and cameras in every room. The recordings could be accessed by government auditors at any time.
I think you'd agree that the main dangers lie with high-risk inmates. To avoid that issue (at least until you have more data on how the system actually functions), you could prevent negative bids from going over a certain size (i.e. you can't bid less than −x dollars). The remaining inmates, whose contracts aren't bid on, would go to public prisons. The bid restrictions could be loosened as we gain a better understanding of the system and impose better regulations.
Public systems have common problems. It's hard to overthrow poorly performing incumbents: If I think I can run prisons better than the existing government, I have to overthrow the entire government in an election. The people in charge of prisons don't have the right incentives: If they could prevent a murder for 2 million dollars, they don't have access to that capital. And sure, they could run tests to see which rehabilitation measures work best, but can they make good decisions on which rehabilitation theories to test, especially when the payoffs for the prison don't exist?
The payoffs for the prison don't exist, but that might be fixed - at least to some extent - by introducing premiums and there are payoffs for the state. Although states are not as constrainted for funding as private companies, the costs of imprisonments have not gone unnoticed.
Another perverse incentive: campaign for making more offenses punishable by prison sentences.
No, I don't think this is a problem. The prisons are competing against each other, not acting as a single, unified block. Why would a prison spend money on making something illegal (through lobbying) when they still have to outbid their opponents? Not only that, prisons would also have an additional liability to pay for their existing prisoners who might commit these new crimes after their release.
This will predict that organized corporate lobbying efforts to a first approximation do not exist, except for entrenching intrasector monopolies against direct competitors.
I think this is sometimes untrue in practice.
You mean the first part? (I.e. Why pay for lobbying when you share the "benefits" with your competitors and still have to compete?) Yeah, when a company becomes large enough, the benefits of a rule change can outweigh the cost of lobbying.
But, for this particular system, if a prison is large enough to lobby, then they're going to have a lot of liabilities from all of their former and current inmates. If they lobby for longer sentences or try to make more behaviours illegal, and one of their former inmates is caught doing one of these new crimes, the prison has to pay.
One way prisons could avoid this is by paying someone else to take on these liabilities. But, in the contract, this person could ensure the prison pays for compensation for any lobbying that damages them.
So a lobbying prison (1) benefits from more inmates in the future, (2) has to pay the cost of lobbying, and (3) has to pay more for the additional liabilities of their past and current inmates (not for their future inmates though, because the liabilities will be offset by a lower initial price for those inmate contracts). Points 1 and 2 are the same under the current prison system. Point 3 is new, and it should push in the direction of less lobbying, at least once the system has existed for a while.
I would assume that for a private prison that has become good at its business the benefits of more inmates would outweigh the liabilities and that at some point it would (in principle, ignoring the free rider problem for a moment) become easier to increase the profits by increasing the revenue by making more things illegal than trying to reduce the reoffending rate. Also do administrators profit from more crimes in a public system? It of course increases the demand for administrators, but I don't see how it would increase the salary of a significant number of them.
Does insurance contracts typically contain clauses for future "products"? I would have assumed that the insurance of the prison would only cover the damage of the point in time the contract was firmed.
Ignoring the free-rider problem ("problem" being from the perspective of the prison), as the prison gets more and more current/former inmates, it becomes harder for that cost-benefit calculation to make sense. With no change in the law or the performance of the prison, the prison's liabilities will grow until the point at which the current/former inmates who die are are as numerous incoming inmates. So for lobbying to make financial sense, it would probably have to occur soon after the prison is started or soon after the system is implemented. But that time is also when the prison has the least information about their own competence (in terms of rehabilitation and auction pricing).
Not really, but that's besides the point. The point is that they don't benefit from rehabilitating their inmates. They don't benefit from firing abusive guards. They don't benefit from reading the latest literate on inmate rehabilitation and creating policies that reduce the chance of their inmates re-offending.
I don’t know much about insurance, but I think you can write pretty much whatever contract you like, as long as no laws are broken.
Actually, I was referring to a point you made in an earlier comment:
So do we both agree that (1) does not hold in the current system?
As my other comment promised, here's a couple of criticisms of your model on its own terms:
That's correct. Profit=Revenue−Costs. The profit that most people think about is the accounting profit. Accounting profit ignores opportunity costs, which is what you give up by doing what you're doing (bear with me a moment). Economic profit, on the other hand, includes these opportunity costs in the calculation. For example, let's say Tom Cruise quits acting and decides to bake cakes for a living. Even if his cake shop earns him $1M in accounting profit, he's giving up all the money he could earn acting instead. So his economic profit is actually negative.
I think you could actually just fix this in the model and still reach the same conclusion (though you'd need extra assumptions to make it work). I really just wanted to introduce my idea for the prison system, rather than make an airtight argument to justify it.
It is very difficult, but that's exactly what the financial markets do.
Yep. If someone is great at running prisons, you want them to do so, regardless of how good they are at predicting the future. Ideally, you would have a system that allows any good expert to thrive, even if they know little about anything outside of their expertise. But companies deal with this all the time. When they're developing a new product, they have to predict which research ventures will be fruitful and which won't be. They have to predict how well products will sell. They have to predict product breakage rates. They have to predict what advertising will work the best. All these things are hard, which is why companies fail. But they are replaced by ones who better succeed at solving all the issues.
Well, yeah. That's why I say to not measure those things. Only measure the big things. The reason why I mention that later in my post, rather than including it in the core argument, is because you need to "smooth things out" with simplifying assumptions to make logical arguments work.
You could actually use my proposal as a secondary, opt-in public education system as well.
Sure. But I don't see why we can't fix those systems as well. (Just to clarify, ideally salaries are paid based on marginal contribution, not the total contribution of the industry--which is why we don't pay farmers an infinite amount. But I agree that not everyone is paid their marginal contribution.)
Yes, that's right! But it is a solvable problem. A taxation system that financially compensates people for rule changes would mitigate this. In effect, the prisons would be paid as if the taxation system were fixed at the time the inmate contract was made.
I'm not sure what you're saying here. People still get to vote. The government has simply exchanged their taxation stream for its present value. Are also you saying private companies shouldn't be allowed to buy government bonds?
Thanks very much for writing this! I have always thought this was a great idea, and a huge mistake that reformists focus on abolishing private prisons, rather than using them. With privatisation you get what you pay for, and at the moment we pay for volume.
Minor note: you define CrimesBob as
However, might it be better to define it as the minimum amount we would have to be paid in order to release someone? If someone is expected to cause spectacularly large amounts of disutility, we should want them to stay in prison even if we can't spend more than total GDP on it.
You still can!
Yeah, me too. I've told people that "I have an idea for a private prison system" and they think it's a bad idea before they've heard any details. I think the government has probably done a better job than the private sector with prisons, so it's a bit of hard sell.
Correct! The performance of the private sector depends on what the system maximizes. The prison's current profit-maximizing behaviour is to minimize the prison's cost per inmate and make sure that inmates are "return customers".
No, how long someone stays in prison is in the domain of the laws, rather than the prison system. The question is "Would this prison system prevent an ideal set of laws from being implemented?" I can't see any reason why they shouldn't work together. Someone who has caused great harm, and is likely to cause more great harm, should not be allowed out. But that's for the judge to decide.
I would, but I'm working on another post.