3 ms·

If you look at my comments throughout this thread, you'll see that I am aware of the potential for incentives to have undesirable higher-order effects. Many peo

by chroma 7y ago

If you look at my comments throughout this thread, you'll see that I am aware of the potential for incentives to have undesirable higher-order effects. Many people in this thread have suggested certain concrete undesirable consequences, and I've replied to them. In contrast, you are putting forth a fully general counterargument to any incentive structure. But clearly financial incentives work in many cases. It's what sports teams use for their athletes and coaches. It's what businesses use for their leadership teams. When we need the best people possible to perform at their best, we tend to pay them money and pay them more money if they meet certain goals. > IMO the only way to run things like prisons is to hire people that strongly care about the public good to run them. Hire people who want to run a good organization that accomplishes the organization's goals. Which for prisons should be rehabilitation and safety. Yes, ideally we'd have every job staffed people who are wise, kind, competent, and hard working. But we don't have nearly enough of such people and we judge traits inaccurately. So instead we have to make do with less effective but more practical methods. Financial incentives are one of those options.

wwweston 7y ago
> a fully general counterargument to any incentive structure. A general counterargument is worth considering. Any incentive sufficiently powerful to draw forth extra resourcefulness in its legitimate pursuit is quite likely to also draw forth extra resourcefulness in its illegitimate pursuit. > It's what businesses use for their leadership teams. This is may help your case considerably less than you think if you look closely. Various forms of the principal-agent problem pop up all the time. Pink and Kohn have also forwarded some research suggesting that extrinsic incentives have diminishing returns or run counterproductive in some cases. And Goodhart's law is a proverb in the business world for a reason. Aligning incentives is still important, but getting optimal performance is rarely as simple as tying rewards to a goal. Hell, half the time defining optimal performance is a project. And when the underlying issues involve social safety and significantly curtailing someone's liberties via imprisonment (and the already fine tension between them), it's an especially tricky topic.
- ShorsHammer 7y ago
  How exactly can a private prison "illegitimately pursue" having a lower recidivism rate? Simply handwaving it away with incentives being gamed seems to complete ignore that you are suggesting that there will be widespread systemic corruption of the entire nation's judicial system. Perhaps the general counterargument could really take a moment to focus on the details. Prisoners not being convicted of more crimes down the line seems to be a no-brainer in societal benefit, or does the line of generalised argument presented here think that's not the case with some more handwaving about optimal outcomes?
  - tlb 7y ago
    For instance, they could teach prisoners how not to get caught. EDIT: what does it mean that I love thinking about ways of gaming systems like this?
    - chroma 7y ago
      > what does it mean that I love thinking about ways of gaming systems like this? That you might be a good security researcher? From https://www.schneier.com/blog/archives/2008/03/the_security_mi_1.html https://www.schneier.com/blog/archives/2008/03/the_security_...: > Uncle Milton Industries has been selling ant farms to children since 1956. Some years ago, I remember opening one up with a friend. There were no actual ants included in the box. Instead, there was a card that you filled in with your address, and the company would mail you some ants. My friend expressed surprise that you could get ants sent to you in the mail. > I replied: "What's really interesting is that these people will send a tube of live ants to anyone you tell them to."
  - wwweston 7y ago
    > How exactly can a private prison "illegitimately pursue" having a lower recidivism rate? Picking prisoners would be one obvious way. Now, you can say "well, we'll just make sure they can't do that." That's fine; the overarching point here is that you can't simply set up an incentive system without thinking about the ways it can be gamed, among other unintended consequences. Though: do you know that you can make sure prisons can't do that? Is it legal to forbid them a choice? Can you set up a system for distributing prisoners that's resistant to capture or influence? It's a little bit like saying "we'll just make sure our software is secure." > you are suggesting that there will be widespread systemic corruption of the entire nation's judicial system. I'm far from the only person suggesting that private prisons corrupt the criminal justice system, but yes, that's part of the argument. Some people concerned about the issue believe that industry lobbyists work to back criminal code changes requiring more prison sentencing. Others believe in milder phenomena like judges internalizing lower costs of imprisonment and therefore using more of it in sentencing. The former could be described as corruption without being even slightly illegal. The latter might not even be recognized as a form of corruption, but it'd be a systemic issue unless you think imprisonment is underused in the US.
tomlockwood 7y ago
> In contrast, you are putting forth a fully general counterargument to any incentive structure. But clearly financial incentives work in many cases. Yeah but I think the big question here is - in what cases? There's a lot of evidence to suggest financial incentives are in some cases, bad. Take for example, the financial incentives given to people for running a private prison. Bam.

by chroma 7y ago

wwweston 7y ago

> a fully general counterargument to any incentive structure. A general counterargument is worth considering. Any incentive sufficiently powerful to draw forth extra resourcefulness in its legitimate pursuit is quite likely to also draw forth extra resourcefulness in its illegitimate pursuit. > It's what businesses use for their leadership teams. This is may help your case considerably less than you think if you look closely. Various forms of the principal-agent problem pop up all the time. Pink and Kohn have also forwarded some research suggesting that extrinsic incentives have diminishing returns or run counterproductive in some cases. And Goodhart's law is a proverb in the business world for a reason. Aligning incentives is still important, but getting optimal performance is rarely as simple as tying rewards to a goal. Hell, half the time defining optimal performance is a project. And when the underlying issues involve social safety and significantly curtailing someone's liberties via imprisonment (and the already fine tension between them), it's an especially tricky topic.

ShorsHammer 7y ago
How exactly can a private prison "illegitimately pursue" having a lower recidivism rate? Simply handwaving it away with incentives being gamed seems to complete ignore that you are suggesting that there will be widespread systemic corruption of the entire nation's judicial system. Perhaps the general counterargument could really take a moment to focus on the details. Prisoners not being convicted of more crimes down the line seems to be a no-brainer in societal benefit, or does the line of generalised argument presented here think that's not the case with some more handwaving about optimal outcomes?
- tlb 7y ago
  For instance, they could teach prisoners how not to get caught. EDIT: what does it mean that I love thinking about ways of gaming systems like this?
  - chroma 7y ago
    > what does it mean that I love thinking about ways of gaming systems like this? That you might be a good security researcher? From https://www.schneier.com/blog/archives/2008/03/the_security_mi_1.html https://www.schneier.com/blog/archives/2008/03/the_security_...: > Uncle Milton Industries has been selling ant farms to children since 1956. Some years ago, I remember opening one up with a friend. There were no actual ants included in the box. Instead, there was a card that you filled in with your address, and the company would mail you some ants. My friend expressed surprise that you could get ants sent to you in the mail. > I replied: "What's really interesting is that these people will send a tube of live ants to anyone you tell them to."
- wwweston 7y ago
  > How exactly can a private prison "illegitimately pursue" having a lower recidivism rate? Picking prisoners would be one obvious way. Now, you can say "well, we'll just make sure they can't do that." That's fine; the overarching point here is that you can't simply set up an incentive system without thinking about the ways it can be gamed, among other unintended consequences. Though: do you know that you can make sure prisons can't do that? Is it legal to forbid them a choice? Can you set up a system for distributing prisoners that's resistant to capture or influence? It's a little bit like saying "we'll just make sure our software is secure." > you are suggesting that there will be widespread systemic corruption of the entire nation's judicial system. I'm far from the only person suggesting that private prisons corrupt the criminal justice system, but yes, that's part of the argument. Some people concerned about the issue believe that industry lobbyists work to back criminal code changes requiring more prison sentencing. Others believe in milder phenomena like judges internalizing lower costs of imprisonment and therefore using more of it in sentencing. The former could be described as corruption without being even slightly illegal. The latter might not even be recognized as a form of corruption, but it'd be a systemic issue unless you think imprisonment is underused in the US.

tomlockwood 7y ago

> In contrast, you are putting forth a fully general counterargument to any incentive structure. But clearly financial incentives work in many cases. Yeah but I think the big question here is - in what cases? There's a lot of evidence to suggest financial incentives are in some cases, bad. Take for example, the financial incentives given to people for running a private prison. Bam.