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.
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.
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?
EDIT: 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_...:
> 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."
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.
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.
Prison ships to Australia were originally paid per passenger that boarded in Britain. The death rates were horrible, in the double-digit percentages. When the transport ships were instead paid per passenger that _arrived_ the death rates went down phenomenally.
And someone like you might say: "omg these prison ships gamed the system by saving lives!!!" Yes, that's what we want to happen.
(edit: not sure what the downvote is for, but I want to be clear that I'm not for gaming the rates in this way. I'm merely saying we don't want to incentivize lowering recividism rates at the expense of everything else)
It’s a fantasy to think you can simple “hire people that care about the public good”. If that were the case the government wouldn’t be filled with incompetent red tape slingers vying for nothing more than a good retirement.
Yep. Thinking that you can just design a good incentive structure and be done with is hopelessly naive.
> 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.
Thinking that you can just hire people that care about the public good is just as naive as thinking that you can just devise the right incentive structure.
EVERY SINGLE SYSTEM GETS GAMED. NO MATTER WHAT YOU MEASURE.
Every metric you could possible think of can be manipulated in ways you can't imagine. But give private companies the incentive and someone in them will figure it out.
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.