5 ms·

There is so much more to this experiment than drunk driving! It shows that modest, swift, carefully metered punishment is a lot more effective than the more tr

by DenisM 11y ago

There is so much more to this experiment than drunk driving! It shows that modest, swift, carefully metered punishment is a lot more effective than the more traditional slow, but heavy punishment. This is a big deal, because it shows a promising path to dramatically reduce the US prison population. Don't mind the drunk driving, think about all of the crimes out there, and all of the people who got a hammer dropped on them.

aidenn0 11y ago
As a parent of a difficult child, this is not at all surprising. What little evidenced-based parenting advice is out there[1] suggests the same thing: taken individually, immediate minor punishments are as effective as major punishments. The big advantage comes with the fact that the parent can be more consistent as minor punishments are typically less work than large ones. Consistent annoyances are better at provoking change than occasional calamity. 1: If you thought the story for evidence-based medicine was bad, just skim through the parenting books at a book store.
- DenisM 11y ago
  It's not the novelty of this approach that excites me, it's the evidence of it working so well. We can take this data to a legislature and lobby for new treatment plans. A parent's hunch will not carry the same weight.
  - joshuapants 11y ago
    Evidence-based politics is even less popular than evidence-based medicine.
    - DenisM 11y ago
      I think your point is too blunt in its cynicism, it's not too useful this way. Evidence is currency, and so are demagoguery, posturing, campaign money, endorsements, promises, actual political platform, public sentiments, etc. Sadly, evidence does not trump all else. Luckily, even a modest amount of evidence currency can tip the scales of the outcome. I would like to draw your attention to a media campaign waged in recent month by forces unknown to reduce harshness of the US criminal justice system. It's pretty much every week that I get to read an article about in the New York Times, and they all refer to some sort of research being released, or a small investigation into a particular injustice. Now we get this article from the Wall Street Journal. You know it's on when both the NYT and the WSJ converge on something. The tide is turning rapidly away from "Tough On Crime" to "Measured Response", and this evidence will be a contributing factor in this sea-change.
      - smt88 11y ago
        Actually, you're wrong about evidence. People become more steadfast in their opinions in the light of evidence.[1] 1. http://youarenotsosmart.com/2011/06/10/the-backfire-effect/ http://youarenotsosmart.com/2011/06/10/the-backfire-effect/
- toddsiegel 11y ago
  Do you have a link or book about this to share for a new parent?
  - aidenn0 11y ago
    Most of my research has been for dealing with kids that have trauma backgrounds, and there are different things you do there. I will say that I've found the Love & Logic series less terrible than most of the mainstream stuff, and you'll likely be able to find a local class, which (assuming you're not a single parent) can be huge on getting you on the same page with your significant-other. That made a big difference for me, as my wife and I were raised very differently. A few snippets of advice: * Operant Conditioning[2] works both ways: your child's behavior will affect you, and you will subconsciously avoid things that are distressing to you. Not taking this into account can cause a lot of problems. * One thing at a time. e.g. don't potty train and teach table manners at the same time. Partly because it's harder on the kids, but mainly to prevent burnout for the parents. * Focus primarily on having a positive relationship (versus a lot of training & discipline) with kids under about 2. It seems likely that Martha Sears[1] went overboard, but there is definitely some phenomena like "attachment." (Also, Attachment Parenting, Attachment therapy and Attachment-based therapy are all different things, and the terminology is not always agreed upon) 1: https://en.wikipedia.org/wiki/Attachment_parenting https://en.wikipedia.org/wiki/Attachment_parenting 2: https://en.wikipedia.org/wiki/Operant_conditioning https://en.wikipedia.org/wiki/Operant_conditioning
    - danieltillett 11y ago
      Yes children are experts at manipulating their parents. When you think about it they are on average as smart as their parents, but they have the advantage that their full time job is getting their parent to do what they want. Parents really don’t stand a chance.
      - bitJericho 11y ago
        Parents ar so much more experienced though. Reverse psychology works well even into the early teen years. How many parents fall for blatant reverse psychology. That's just one example of many techniques. Others include conditioning, reinforcement, outright education.
- icelancer 11y ago
  >1: If you thought the story for evidence-based medicine was bad, just skim through the parenting books at a book store. Yes. All the ridiculous advice my partner and I get about our children is unbelievable. And I'm someone who had to self-diagnose medical issues after years of being ignored by doctors and other "experts" who wouldn't look at evidence-based approaches. I thought it couldn't get worse. But it does. Oh, it does... so much worse.
- dlss 11y ago
  The result here isn't what you said: > taken individually, immediate minor punishments are as effective as major punishments The result here is that frequent minor punishments are 72% more effective than infrequent major punishments (see the article on Hawaii's 5 year study of this approach linked to by https://news.ycombinator.com/item?id=10064425 https://news.ycombinator.com/item?id=10064425).
  - jeorgun 11y ago
    If I'm understanding right, grandparent isn't saying "effectiveness(minor, frequent) == effectiveness(major, infrequent)"; they're saying "effectiveness(minor) == effectiveness(major) && effectiveness(frequent) > effectiveness(infrequent)".
    - aidenn0 11y ago
      That's exactly what I was saying, though this (and my original comment) was a simplification.
  - aidenn0 11y ago
    That seems completely consistent with my original quote of "Consistent annoyances are better at provoking change than occasional calamity." My point was a single minor punishment is about as effective as a single major punishment. It's more complicated than that of course, (sometimes a single major punishment is more effective, sometimes a major punishment is actually counterproductive), but in any event, frequency seems to be much more important than severity.
- proactivesvcs 11y ago
  This explains why my friends and family who use the "Naughty Step" seem to have discipline down to a T. I have even heard a few stories of kids putting themselves on the Naughty Step. Even as someone who's not a parent, when I have done babysitting for said parents, I'm able to use the Naughty Step without fuss.
- stcredzero 11y ago
  Perhaps this also applies to industries? There seem to be a lot of 50 year old nuclear plants. There also seems to be greater impetus to upgrade natural gas turbines. Maybe we need to favor machines/processes/designs that fail early and fail often, as opposed to ones that rarely fail but in catastrophic ways?
samstave 11y ago
Timeouts are better than grounding.
chimeracoder 11y ago
> It shows that modest, swift, carefully metered punishment is a lot more effective than the more traditional slow, but heavy punishment. I'd be very nervous about extending this beyond drunk driving. With DUIs, breathalyzers are essentially "proof" of the crime (to the extent that there is ever incontrovertible proof for any crime). And if you refuse a breathalyzer, you've violated the implied consent law, so there's no ambiguity there either. You may be able to get off on a technicality, but it's rather difficult to abuse this, short of a police officer falsifying breathalyzer results outright. But for most other crimes, there's a lot of ambiguity, which is why we have a separation between LEOs and the court system. I'm already uncomfortable with the level of unchecked power that LEOs (and non-LEOs, like the TSA) have today. I really don't want them to have any more leeway with detention of civilians.
- DenisM 11y ago
  You seem to be under the impression that this method abdicates judicial review? That's not the case, the metered punishment is delivered in pieces after the court order. That is rather than send a person to jail upon conviction, a judge would put them on electronic probation, where each violation results into a small, immediate penalty. This will help in two way - prevent harsh sentences for simple missteps, and actual train people to follow the rules of the law by shortening feedback loop of the events.
- hackeraccount 11y ago
  In a perfect world if you extended this beyond DUI the punishments given out would be reduced. Especially, the punishments that happen after you serve your time. One of my biggest problems with criminal justice system is the way the ex-cons are dogged by their mistakes for the rest of their life. I understand the interest an employer (or in the worst of cases a neighbor) might have in knowing that about a person but... somehow it seems deeply unfair.
  - TeMPOraL 11y ago
    > I understand the interest an employer (or in the worst of cases a neighbor) might have in knowing that about a person but... somehow it seems deeply unfair. It is unfair if you believe that the jail is supposed to rehabilitate people. It seems to me that people believe it in theory but don't believe it in practice.
unabst 11y ago
To immediately think to not do something is often a reflex based on memory, and not a rational decision. Training reflexes is all about applying how the brain works. Taking a hint from dog training, if you punish a dog too late, they forget what they're being punished for. Also the severity also doesn't seem to matter. Bad is bad no matter how loud or how painful, so just keep it to a minimum and have it for convenience. Beyond that rewards work much better for training brains.
enraged_camel 11y ago
>>This is a big deal, because it shows a promising path to dramatically reduce the US prison population. Don't mind the drunk driving, think about all of the crimes out there, and all of the people who got a hammer dropped on them. I hate to be that guy, but the US prison population is large because certain powerful interests want it to be large. This is unlikely to change anytime soon. And it definitely won't change as a result of a study about drunk driving. That said, it's still an interesting study. Like many people, I've lost a friend in an accident involving a drunk driver, so the subject is important to me.
- kbenson 11y ago
  This is an important point. It's hard to enact change when it threatens the status quo. In California, Prison Correctional officers are backed by a powerful lobby[1]. 1: https://en.wikipedia.org/wiki/California_Correctional_Peace_Officers_Association#Lobbying https://en.wikipedia.org/wiki/California_Correctional_Peace_...
- lotu 11y ago
  Something to rember about the "powerful interests" they only remain powerful because vast majority of voting Americans either agree with them on that don't care about that about that issue or care so little about the issue that it dosen't matter. When this changes policy changes. Look at Marijuana legalization, where are the powerful interests benefiting from it?
sliverstorm 11y ago
It shows that modest, swift, carefully metered punishment is a lot more effective than the more traditional slow, but heavy punishment. I thought we basically already knew that? I thought we had already shown in studies that swiftness & certainty of punishment are more effective deterrents than severity.
- dlss 11y ago
  It's not obvious to me that I wouldn't be seeing a comment just like yours on a study finding the opposite. "We already know people just do what they want if the consequences are minor" Consider: > Hindsight will lead us to systematically undervalue the surprisingness of scientific findings, especially the discoveries we understand—the ones that seem real to us, the ones we can retrofit into our models of the world. http://lesswrong.com/lw/im/hindsight_devalues_science/ http://lesswrong.com/lw/im/hindsight_devalues_science/
- TeMPOraL 11y ago
  Indeed, this is something we've known since time immemorial. It's even in the Bible: Because the sentence against an evil deed is not executed quickly, therefore the hearts of the sons of men among them are given fully to do evil. - (Ecclesiastes 8:11)
tehchromic 11y ago
what boggles my personal mind is that this has to be demonstrated at all, and yet that's a truth that spans the width and depth of our institution mediated existence.
slavik81 11y ago
That is not a new idea. It's long been known that punishments are dramatically more effective if they are swift and certain. But, if the solution to crime is simply to catch every criminal immediately after they commit their crime, the question is how will you do that? The real interesting part of this was the sobriety program. They found a way to identify bad habits at a relatively low cost. How to extend that model to other crimes is unclear. Parole basically exists for that purpose already, so the question is how it can be made more effective. I just don't really see a clear path to general crime reduction here.
- crusso 11y ago
  I think that the key generalization you could make is that once someone is shown to be engaging in criminal behavior, it's better to let that person stay free but have frequent checks than to put that person in a prison cell to rot. It probably does work better for something habitual and detectable like alcohol use, but I could see how it could be applied to gang membership and other forms of thuggery. Maybe spending our resources on light consistent touch programs would help more than the significant resources needed for full-time incarceration.

by DenisM 11y ago

aidenn0 11y ago

As a parent of a difficult child, this is not at all surprising. What little evidenced-based parenting advice is out there[1] suggests the same thing: taken individually, immediate minor punishments are as effective as major punishments. The big advantage comes with the fact that the parent can be more consistent as minor punishments are typically less work than large ones. Consistent annoyances are better at provoking change than occasional calamity. 1: If you thought the story for evidence-based medicine was bad, just skim through the parenting books at a book store.

DenisM 11y ago
It's not the novelty of this approach that excites me, it's the evidence of it working so well. We can take this data to a legislature and lobby for new treatment plans. A parent's hunch will not carry the same weight.
- joshuapants 11y ago
  Evidence-based politics is even less popular than evidence-based medicine.
  - DenisM 11y ago
    I think your point is too blunt in its cynicism, it's not too useful this way. Evidence is currency, and so are demagoguery, posturing, campaign money, endorsements, promises, actual political platform, public sentiments, etc. Sadly, evidence does not trump all else. Luckily, even a modest amount of evidence currency can tip the scales of the outcome. I would like to draw your attention to a media campaign waged in recent month by forces unknown to reduce harshness of the US criminal justice system. It's pretty much every week that I get to read an article about in the New York Times, and they all refer to some sort of research being released, or a small investigation into a particular injustice. Now we get this article from the Wall Street Journal. You know it's on when both the NYT and the WSJ converge on something. The tide is turning rapidly away from "Tough On Crime" to "Measured Response", and this evidence will be a contributing factor in this sea-change.
    - smt88 11y ago
      Actually, you're wrong about evidence. People become more steadfast in their opinions in the light of evidence.[1] 1. http://youarenotsosmart.com/2011/06/10/the-backfire-effect/ http://youarenotsosmart.com/2011/06/10/the-backfire-effect/
toddsiegel 11y ago
Do you have a link or book about this to share for a new parent?
- aidenn0 11y ago
  Most of my research has been for dealing with kids that have trauma backgrounds, and there are different things you do there. I will say that I've found the Love & Logic series less terrible than most of the mainstream stuff, and you'll likely be able to find a local class, which (assuming you're not a single parent) can be huge on getting you on the same page with your significant-other. That made a big difference for me, as my wife and I were raised very differently. A few snippets of advice: * Operant Conditioning[2] works both ways: your child's behavior will affect you, and you will subconsciously avoid things that are distressing to you. Not taking this into account can cause a lot of problems. * One thing at a time. e.g. don't potty train and teach table manners at the same time. Partly because it's harder on the kids, but mainly to prevent burnout for the parents. * Focus primarily on having a positive relationship (versus a lot of training & discipline) with kids under about 2. It seems likely that Martha Sears[1] went overboard, but there is definitely some phenomena like "attachment." (Also, Attachment Parenting, Attachment therapy and Attachment-based therapy are all different things, and the terminology is not always agreed upon) 1: https://en.wikipedia.org/wiki/Attachment_parenting https://en.wikipedia.org/wiki/Attachment_parenting 2: https://en.wikipedia.org/wiki/Operant_conditioning https://en.wikipedia.org/wiki/Operant_conditioning
  - danieltillett 11y ago
    Yes children are experts at manipulating their parents. When you think about it they are on average as smart as their parents, but they have the advantage that their full time job is getting their parent to do what they want. Parents really don’t stand a chance.
    - bitJericho 11y ago
      Parents ar so much more experienced though. Reverse psychology works well even into the early teen years. How many parents fall for blatant reverse psychology. That's just one example of many techniques. Others include conditioning, reinforcement, outright education.
icelancer 11y ago
>1: If you thought the story for evidence-based medicine was bad, just skim through the parenting books at a book store. Yes. All the ridiculous advice my partner and I get about our children is unbelievable. And I'm someone who had to self-diagnose medical issues after years of being ignored by doctors and other "experts" who wouldn't look at evidence-based approaches. I thought it couldn't get worse. But it does. Oh, it does... so much worse.
dlss 11y ago
The result here isn't what you said: > taken individually, immediate minor punishments are as effective as major punishments The result here is that frequent minor punishments are 72% more effective than infrequent major punishments (see the article on Hawaii's 5 year study of this approach linked to by https://news.ycombinator.com/item?id=10064425 https://news.ycombinator.com/item?id=10064425).
- jeorgun 11y ago
  If I'm understanding right, grandparent isn't saying "effectiveness(minor, frequent) == effectiveness(major, infrequent)"; they're saying "effectiveness(minor) == effectiveness(major) && effectiveness(frequent) > effectiveness(infrequent)".
  - aidenn0 11y ago
    That's exactly what I was saying, though this (and my original comment) was a simplification.
- aidenn0 11y ago
  That seems completely consistent with my original quote of "Consistent annoyances are better at provoking change than occasional calamity." My point was a single minor punishment is about as effective as a single major punishment. It's more complicated than that of course, (sometimes a single major punishment is more effective, sometimes a major punishment is actually counterproductive), but in any event, frequency seems to be much more important than severity.
proactivesvcs 11y ago
This explains why my friends and family who use the "Naughty Step" seem to have discipline down to a T. I have even heard a few stories of kids putting themselves on the Naughty Step. Even as someone who's not a parent, when I have done babysitting for said parents, I'm able to use the Naughty Step without fuss.
stcredzero 11y ago
Perhaps this also applies to industries? There seem to be a lot of 50 year old nuclear plants. There also seems to be greater impetus to upgrade natural gas turbines. Maybe we need to favor machines/processes/designs that fail early and fail often, as opposed to ones that rarely fail but in catastrophic ways?

samstave 11y ago

Timeouts are better than grounding.

chimeracoder 11y ago

> It shows that modest, swift, carefully metered punishment is a lot more effective than the more traditional slow, but heavy punishment. I'd be very nervous about extending this beyond drunk driving. With DUIs, breathalyzers are essentially "proof" of the crime (to the extent that there is ever incontrovertible proof for any crime). And if you refuse a breathalyzer, you've violated the implied consent law, so there's no ambiguity there either. You may be able to get off on a technicality, but it's rather difficult to abuse this, short of a police officer falsifying breathalyzer results outright. But for most other crimes, there's a lot of ambiguity, which is why we have a separation between LEOs and the court system. I'm already uncomfortable with the level of unchecked power that LEOs (and non-LEOs, like the TSA) have today. I really don't want them to have any more leeway with detention of civilians.

DenisM 11y ago
You seem to be under the impression that this method abdicates judicial review? That's not the case, the metered punishment is delivered in pieces after the court order. That is rather than send a person to jail upon conviction, a judge would put them on electronic probation, where each violation results into a small, immediate penalty. This will help in two way - prevent harsh sentences for simple missteps, and actual train people to follow the rules of the law by shortening feedback loop of the events.
hackeraccount 11y ago
In a perfect world if you extended this beyond DUI the punishments given out would be reduced. Especially, the punishments that happen after you serve your time. One of my biggest problems with criminal justice system is the way the ex-cons are dogged by their mistakes for the rest of their life. I understand the interest an employer (or in the worst of cases a neighbor) might have in knowing that about a person but... somehow it seems deeply unfair.
- TeMPOraL 11y ago
  > I understand the interest an employer (or in the worst of cases a neighbor) might have in knowing that about a person but... somehow it seems deeply unfair. It is unfair if you believe that the jail is supposed to rehabilitate people. It seems to me that people believe it in theory but don't believe it in practice.

unabst 11y ago

To immediately think to not do something is often a reflex based on memory, and not a rational decision. Training reflexes is all about applying how the brain works. Taking a hint from dog training, if you punish a dog too late, they forget what they're being punished for. Also the severity also doesn't seem to matter. Bad is bad no matter how loud or how painful, so just keep it to a minimum and have it for convenience. Beyond that rewards work much better for training brains.

enraged_camel 11y ago

>>This is a big deal, because it shows a promising path to dramatically reduce the US prison population. Don't mind the drunk driving, think about all of the crimes out there, and all of the people who got a hammer dropped on them. I hate to be that guy, but the US prison population is large because certain powerful interests want it to be large. This is unlikely to change anytime soon. And it definitely won't change as a result of a study about drunk driving. That said, it's still an interesting study. Like many people, I've lost a friend in an accident involving a drunk driver, so the subject is important to me.

kbenson 11y ago
This is an important point. It's hard to enact change when it threatens the status quo. In California, Prison Correctional officers are backed by a powerful lobby[1]. 1: https://en.wikipedia.org/wiki/California_Correctional_Peace_Officers_Association#Lobbying https://en.wikipedia.org/wiki/California_Correctional_Peace_...
lotu 11y ago
Something to rember about the "powerful interests" they only remain powerful because vast majority of voting Americans either agree with them on that don't care about that about that issue or care so little about the issue that it dosen't matter. When this changes policy changes. Look at Marijuana legalization, where are the powerful interests benefiting from it?

sliverstorm 11y ago

It shows that modest, swift, carefully metered punishment is a lot more effective than the more traditional slow, but heavy punishment. I thought we basically already knew that? I thought we had already shown in studies that swiftness & certainty of punishment are more effective deterrents than severity.

dlss 11y ago
It's not obvious to me that I wouldn't be seeing a comment just like yours on a study finding the opposite. "We already know people just do what they want if the consequences are minor" Consider: > Hindsight will lead us to systematically undervalue the surprisingness of scientific findings, especially the discoveries we understand—the ones that seem real to us, the ones we can retrofit into our models of the world. http://lesswrong.com/lw/im/hindsight_devalues_science/ http://lesswrong.com/lw/im/hindsight_devalues_science/
TeMPOraL 11y ago
Indeed, this is something we've known since time immemorial. It's even in the Bible: Because the sentence against an evil deed is not executed quickly, therefore the hearts of the sons of men among them are given fully to do evil. - (Ecclesiastes 8:11)

tehchromic 11y ago

what boggles my personal mind is that this has to be demonstrated at all, and yet that's a truth that spans the width and depth of our institution mediated existence.

slavik81 11y ago

That is not a new idea. It's long been known that punishments are dramatically more effective if they are swift and certain. But, if the solution to crime is simply to catch every criminal immediately after they commit their crime, the question is how will you do that? The real interesting part of this was the sobriety program. They found a way to identify bad habits at a relatively low cost. How to extend that model to other crimes is unclear. Parole basically exists for that purpose already, so the question is how it can be made more effective. I just don't really see a clear path to general crime reduction here.

crusso 11y ago
I think that the key generalization you could make is that once someone is shown to be engaging in criminal behavior, it's better to let that person stay free but have frequent checks than to put that person in a prison cell to rot. It probably does work better for something habitual and detectable like alcohol use, but I could see how it could be applied to gang membership and other forms of thuggery. Maybe spending our resources on light consistent touch programs would help more than the significant resources needed for full-time incarceration.