Let's have some numbers

[andrewducker]

Is there any evidence that raising the sentence for a particular crime lowers its incidence?

We change sentencing recommendations often enough (around the world, anyway) that it must be possible to tell whether doubling the sentence for a particular crime actually has any effect on the number of people carrying it out.

Anyone know of any studies?


My personal intuition is that 99% of people who carry out crimes don't think about how big the sentence is at all. But I'm very happy to be proven wrong with actual numbers.

Comments

[khoth.livejournal.com]

I seem to remember reading somewhere that the sentence doesn't have a huge effect on crime rate, but the probably of being caught does. I don't have a source, though.

[nickys.livejournal.com]

Yes, likewise.
It's the detection rate, not the punishment that reduces crime.

[sigmonster.livejournal.com]

The way I like to put it is that humans are at least* as smart as pigeons, and if you want to train a pigeon you wouldn't ever use negative reinforcements applied totally nconsistently...


*sometimes I feel this is optimistic, but mostly it's true

Re:

[nickys.livejournal.com]

Yup, and in child-rearing you use mild sanctions not severe ones to train kids to behave well.
Screaming at them and hitting them all the time just makes them jumpy and more likely to freak out and have destructive tantrums.

[stillcarl.livejournal.com]

Which suggests that lowering the lengths of sentences and putting the money saved by that into increasing the detection rate is the way to reduce crime.

[nickys.livejournal.com]

Well, yes, that would be the sensible approach.

... if only our leader would stop and think for something over ten seconds...

[nickys.livejournal.com]

A search for "crime rate reduction" on Google Scholar is pulling up lots of results, so if you know anyone with access to academic databases you could get them to look at the articles.

[heron61.livejournal.com]

The only data I've seen is that capital punishment does absolutely nothing to deter serious crime, and it seems exceptionally likely that this result can be generalized.

[ciphergoth.livejournal.com]

There's data on this that goes both ways. Capital punishment was the subject Lord, Ross and Lepper used in their landmark 1979 experiment demonstrating "biased assimilation".

[heron61.livejournal.com]

I've seen data far more recent that 1979 (I remember seeing a bunch of studies from the late 90s) that seems fairly conclusive, since we've had a large sample of states in the US both with and w/o the death penalty. I don't have any of that data on hand, but I did find this comparison, which is admittedly far from perfect.

[innerbrat]

But ANDREW, the lengths of the sentence have nothing to do with reducing the rates of crime and everything to do with making the victims feel better.

Because revenge is totally a recovery technique. All the psychiatrists say so.

[andrewducker]

That's the thing. If deterrent actually has an effect, then I can see an argument for it, and much though I dislike it, such is life.

If, on the other hand, it has bugger all effect, then that makes me feel better at protesting about it.

[innerbrat]

Through the sarcasm, what I was trying to say - and I agree with you, by the way - that most people don't seem to care about sentencing as a deterrent so much as a revenge tactic. This is why the media seem to think that the victims of crimes are more important to consult about sentencing measures than crime scientists.

[andrewducker]

Yup, absolutely.

There are three parts to sentencing, so far as I can see:
1) Reform - turning the person into someone who is able to fit back into society in a more successful manner than they did before.
2) Protection - making sure that people who can't be trusted aren't roaming the streets (obviously connected to reform).
3) Deterrence - if, of course, this effect exists.
4) Punishment - which is basically just revenge.

I'm in favour of (1), think that (2) is justified for as long as someone is a danger to others, would like some proof of (3) and find (4) abhorrent.

[innerbrat]

Me too on 4 and 2.

I find 1 and 3 very strongly related. I will accept "the prison system reduces crime" as general evidence, which can then be refined into "the prison system prevents re-offence" and the "the prison system prevents first offences".

Mind you, I suspect that 1 has more room for improvement, while 3 is an either/or situation

[nickys.livejournal.com]

> 4) Punishment - which is basically just revenge.

It is important to teach people that their actions have consequences.

However, certainly with child-rearing, a mild punishment (such as 5 minutes time out) has been shown to be more effective than a severe one (such as a smack).

[andrewducker]

It is important to teach people that their actions have consequences.

That translates to me as "It is important to teach people that if they do things you don't like then you will do unpleasant things to them."

Which doesn't sound like something I'd support.

If that's not what you mean, then please correct me.

[nickys.livejournal.com]

Not what I mean at all.


People learn stuff by being taught it.

One of the very important aspects of parenting is the setting of boundaries, partly for the sanity of the parents, but also, substantially for the welfare of the children.
Kids who aren't given proper boundaries play up and throw tantrums, which leads to escalation and conflict, and children who are unhappy because nobody likes them.

Society also sets boundaries - no hurting other people, no stealing other people's stuff, etc.
Most people see the sense in following those rules, but where somebody isn't following the rules the best (and kindest) thing to do is to help them to learn the difference between right and wrong.
Learning that there are consequences to socially inappropriate actions is part of rehabilitation, so long as those consequences are proportional to the offence and applied in a consistent manner.

[andrewducker]

You seem to be combining "Children need to learn that there are boundaries and rules in life" and "Children need to be punished by me if they break boundaries." into one thing.

I'm not sure why you feel that punishment and education are the same thing, but I don't agree at all.

[nickys.livejournal.com]

Most people understand the need for appropriate behaviour by having it explained to them with words.

For those who don't / can't understand it that way it is necessary to use other methods of teaching, and one of the classic methods (which lots of research has shown to be effective) is to use a system of rewards and (mild) punishments.

[andrewducker]

I'm rarely convinced by the "can't understand it" except in people with mental problems (or pre-verbal children).

I mostly think that treating people badly (which is what pretty much all punishment boils down to) makes people behave worse, rather than better.

You might need to restrict people's behaviour in order to prevent them doing harm, but beyond that more punishment just seems to push people away and make them believe that their behaviour is justified by the fact that people are treating them badly.

[nickys.livejournal.com]

I think we're using different definitions.

I'd regard restricting people's behaviour as a type of punishment, not a different thing.

[andrewducker]

Whether something is punishment or not is down to intent.

If you're restraining someone because every time they go outside they mug an old lady then you're not doing it to punish them, you're doing it because you're preventing harm. If you're doing it because they've been bad, and therefore something bad has to happen to them "to teach them a lesson" then that's punishment.

[nickys.livejournal.com]

> "to teach them a lesson"

Depends what lesson you're teaching.

Remember humans (adults as well as children) are curious apes, and to a certain extent the natural response to being told "no" is "why not?". There's a bit in a Terry Pratchet book about a door with a sign on it which says "Don't open this door - not even to see what happens" - which the fully grown adult wizards promptly open.


Actions tend to have a natural consequence. And if the natural consequence is unpleasant but mild then you can let people learn things for themselves. However, if the natural consequence is severe then it's better to have an imposed mild consequence, a punishment, as a way of teaching the lesson.

Say for example if the lesson is 'hot things will hurt you' then in the case of a candle it might not be unreasonable to let a child discover why they've been told to stay away from it on their own by experimentation.
However, if it's an oven then there's the possibility of serious injury if they fall into or against it while exploring and then a responsible carer will (for example) set a punishment which is imposed if they go too close to the oven.


Speed limits are another one. We generally know that the lower speed limits are the fewer serious accidents there will be. However, for any individual, the probability of an accident is low, and some people basically ignore it, because they think it'll never happn to them.
So, an imposed mild consequence (a fine) and a high probability of detection (speed cameras) are used in combination as a way of making the roads safer.

[andrewducker]

However, if it's an oven then there's the possibility of serious injury if they fall into or against it while exploring and then a responsible carer will (for example) set a punishment which is imposed if they go too close to the oven.

Or, as I've seen in the case of my nephew recently, take him close to the oven, explain it to him, get him to hold his hand close enough for it to reach _almost_ painful, so that he understand what the problem is, and learns not to do it

Although I do agree that your speed camera example is a good one.

[nickys.livejournal.com]

Actually, it occurs to me: somebody (I think you) once explained that people with Aspergers communicate with words, whereas the neurotypical communicate with words and actions.

A punishment is an action which is intended to communicate information.

[andrewducker]

In my experience punishment is too vague and emotionally loaded to communicate any more information than "I want you to be unhappy." People who are punished read vast amounts of other things into it, and it's a hindrance to communication rather than a help.

You might need, for instance, a time-out in order for people's tempers to reduce to the point where communication is possible again, but time-outs to punish people just lead to resentment and then resolving to avoid doing things _in front of you_.

[nickys.livejournal.com]

Arbitrary and disproportionate punishments do that, but where both sides know there need to be rules (which kids mostly do, really) fair punishments don't cause resentment.

There's any amount of child psychology literature on this.
The ideal place is the middle ground - the extremes of excessive punishment and excessive talk without action both lead to kids who don't understand or respect authority.

[andrewducker]

I wasn't punished as a child. I was occasionally told off for doing things I shouldn't, but that was about it. Julie seems to have been similar.

I understand authority, and find it to be a revolting concept*. Empathy is a far better one.

*except in extremis. Most of the time there's absolutely no need for someone to be on top and someone else to be underneath, except when there's fairly severe danger if something goes wrong. So, when you're learning how to fence, for instance, you need to be paying attention to the person who knows how not to stab themselves with the pointy end. But that's when people give others authority, not when people take it.

Children, of course, are also a special case, but as soon as they're willing to take some responsibility they should be getting it. And they should understand why they're doing things. If they're doing it only because otherwise they'd be punished then they don't really understand, and they cannot be trusted when they're not being watched. Training people that "You don't do this because otherwise you will be punished." strikes me as pretty reprehensible - that's not the reason to do (or not do) something.

[naath.livejournal.com]

I don't think it's possible to use prison effectively as part of a scheme of reward/punishment to train people out of doing things. It's too infrequent and too disconnected from the immediate environment of the crime (also too low-probability). In order to understand the crime --> prison connection you are either going to have to go through it a LOT of times or understand it from an explanation rather than from experience (which is a different sort of learning).

Personally I think training and education are very different things; I'm not sure why people persist in training children that are old enough to be educated, I certainly don't think training techniques should be applied to grown adults.

[nickys.livejournal.com]

It can work, in the same way that an emergency psychiatric intervention of locking somebody up can work.

Trouble is there isn't the funding to do constructive things with people while they're in prison as often as there should be.
Saughton Prison, though, has cut re-offending rates with a pioneering library programme, so it is possible for good to come of it.

[naath.livejournal.com]

Emergency psychiatric intervention by locking someone up "works" in the sense that it gets them to stop doing whatever-it-was that was causing a problem (for instance if they are assaulting people).

It can also "work" in that it is sometimes possible to drug people into compliance (or at least into passivity) with whatever rules you have decided they should abide by. Personally I think that drugging people because they have done something you don't like is utterly disgusting, and a gross breach of their rights (personally I believe that the right to refuse interventions is pretty fundamental, indeed more fundamental than the right to liberty - I'm happy to take away the liberty of people who pose a danger to others, I'm not happy to force them to accept treatment they do not wish to accept).

It should go without saying that treatment requested by people is a good thing, and they should get it.

[don-fitch.livejournal.com]

The Traditional Knowledge/study (which doesn't seem to have been disproven convincingly) is that the by-far-major deterrent is geting caught. The severity of the punishment, within reason, doesn't seem to matter much. And yeah, I agree that "revenge" is a shameful motivation in jurisprudence.

[cartesiandaemon.livejournal.com]

I also take it for granted that it has little effect, except when it raises awareness and enforcement, but I don't actually have figures.

(But on a larger scale, if country A treats something as a small-fine offense and country B as a long-prison-sentence offense, people will react differently. But for most crimes people talk about, I imagine people know there IS a penalty, but rely on not getting caught)

[andrewducker]

But on a larger scale, if country A treats something as a small-fine offense and country B as a long-prison-sentence offense, people will react differently.

Prove it.

Edit: To expand slightly - that's exactly the kind of thing I want _proved_ - if we have multiple countries with different sentences for particular crimes then it should be possible to show a correlation between the sentences and the frequency with which they are committed. But a simple statement that it happens doesn't feel terribly useful.

[cartesiandaemon.livejournal.com]

Thanks for edit. OK, let me retract. I agree with you. However, I agree with you in that in the vast majority of cases when someone talks about "increasing sentences" it's a matter of one or two orders of magnitude. And I knew if I said that, I would leave myself open to a reducto ad absurdum response that said "do you really believe that if the penalty were increased from a small on-the-spot fine to the death penalty NO-ONE would be deterred, even if enforcement weren't increased?" So I wanted to pre-emptively NOT say that.

But -- inevitably for me -- by agreeing with someone but hedging, it sounds like a totally disagree with them. Especially as I dropped a casual comment when you were really looking for facts.

It seems to me obvious that SOME people commit crimes with a reasonable cost-benefit analysis in mind, even if most people don't. And my assumption was that people don't tend to distinguish between "small fine" and "large fine" or "small sentence" and "large sentence" and "little torture" and "lots of torture" because they don't know the difference, but if a penalty is increased/decreased so drastically as to move several categories, then some people will notice. However, that was just a guess I thought was relevant, I didn't have any stats, and admit that some things people think are self-evident ARE false, so I admit that's always likely.

I think clean stats for something like that are unlikely, since people often -- say -- double sentences, but they rarely increase the sentence from "small fine" to "death" without a corresponding increase in enforcement, and usually social pressure.

If there are stats, I'd expect to see stats not in a change, but in a big societal difference, one where the difference is so blatant you can see by looking. For instance, I wonder if anyone ever chooses to rob a bank without using a gun, because the penalties for armed robbery are higher (and the chance of being caught inherently similar). I expect so, but I don't have stats. Are there "crimes" where in one country enforcement is high, but the de-facto punishment is a small bribe, and most people just go ahead and do it, even if they don't in a country where enforcement is less consistent but there's a real penalty?

But I retract it as a claim. If it's NOT obvious, then I certainly don't have anything in support of it.

[andrewducker]

Yeah - I dunno. But I know there are crimes where theft is punished with corporal punishment, and I'd love to know if they have lower levels of theft than places where it's generally a fine or community service.

[nickys.livejournal.com]

Juries are less likely to find somebody guilty if the punishment is perceived as too severe, apparently.

[octopoid-horror.livejournal.com]

If you commit a crime (and are thinking about it), you only need to worry about the sentence if 1] you get caught, b] you actually get convicted, c] you get sentenced for the crime you committed and not a lesser version of it for whatever reason and d] you are likely to actually serve the full sentence.

Increasing the sentence is unlikely to do anything if you don't increase detection/conviction rates.