Alcohol and addiction: part II

We posted recently criticising the method in which the Law Commission has decided to approach its revamp of alcohol regulation. In the comments there was some vociferous criticism of the BERL report on which the Law Commissions relied for its estimates of the harm of alcohol usage. We were also lucky enough to have a personal reply from the author of the report, explaining why BERL did things the way that they did. I thought this would be a good time to clear up any confusion about where I stand on the issue.

I think the Law Commission is approaching law-making the wrong way if it seeks to minimise the harm from alcohol usage. I think that law should seek to maximise welfare rather than minimise harm. However, they have been very explicit about the approach that they have taken and I would not accuse them of being deceptive or misleading in the way they’ve presented their ideas.

I have no problem with the fact that BERL’s report did not include the benefits of drinking. That was outside of the scope of the work that they were commissioned to do and there is no reason why they would consider it. Developers of the law should consider the benefits, but that is not BERL’s job.

BERL have also been criticised by Paul Walker and Eric Crampton for not clarifying the scope of their report. While I was initially sympathetic to this criticism, I have since thought better of that stance. I don’t feel that BERL’s report has been misrepresented and commenting on the following work of the Law Commission is unnecessary.

Regarding the content of BERL’s report, Adrian has generously taken the time in his comment to clarify what they did and his comment’s well worth reading. He points out that a lot of the costs of drinking are not internalised due to transaction and monitoring costs. For example, being drunk at work might ideally be reflected in your pay, but in reality the costs of changing pay and monitoring that sort of impairment often outweigh the benefits. In that sort of situation, the cost is not internalised and is a net social cost.

He also elucidates BERL’s approach to addiction. Essentially they reject the idea of alcohol addiction being a rational choice. I think it’s fair to say that there’s still dispute amongst academics about how addiction should be modelled. However, once BERL chose to model addiction as sub-optimal behaviour, there are costs to society and to the drinker from their drinking. As I understand Adrian’s reply, these costs from ‘harmful’ drinking make up the majority of the costs outlined in BERL’s report.

If we believe that there are net costs to drinkers and those around them as a result of their addictive behaviour then there is likely to be a place for government intervention. What that intervention might be I’ll leave to discuss another time.

Hopefully this clarifies my earlier post: the issue is the Law Commission’s use of the report, not the report itself.

  • What would Hayek say

    The Law Commission appears to be following the path of the former LTSA to create a figure of what the social cost of an activity is and use that as the basis to argue to regulate an activity (i.e driving), however neither LTSA and in this instance the LC have considered the benefits of an activity.

    A further slight of hand that happens sometime is in cost/benefit analysis is carried out, where the cost of regulation only considers the financial cost of setting up a regulator e.g. $2m for new commission/law and then this is put against the benefits of the regulator which have a wider scope e.g. reduced social harm. In this case you then have a loaded CBA which favours regulation as the cost to society from the loss of an activity is never considered.

    The next slight of hand is to fiddle with the Discount rate. I’ll shut up now because discussing that leads to accusations of being a climate change denier.

    In the end its all a variation of lies, damned lies and statistics and shows the risks to decision makers in formulating good policy.

  • What would Hayek say :
    The next slight of hand is to fiddle with the Discount rate. I’ll shut up now because discussing that leads to accusations of being a climate change denier.

    Hehehe.

  • @What would Hayek say
    I just don’t really think that there’s any CBA going on here: Geoffrey Palmer said that he wanted to minimise the harm done by alcohol consumption. I see no sleight of hand, just an approach I disagree with.

  • Rauparaha: you’re too conciliatory. Slack says that they take no position on the benefits of drinking to heavy drinkers ’cause it’s too hard to reckon, but that’s not true. He defines those benefits as being equal to zero. There’s a big difference between the two.

    Note also what it takes to be counted as a harmful drinker. 40 grams of alcohol per day for men. Seem obscure to you? I tend to figure that when something seems obscure for no reason, it’s cause it’s trying to hide something ridiculous. Best I can reckon, 40 grams of alcohol is 36 ounces of a normal 5% beer. Or, just over 2 pints. Go just over 2 pints, and you get zero gross benefits from drinking. Hmm.

    The sleight of hand is in defining as costs and harms things that are not. The cost of producing alcohol isn’t a harm unless you define the benefits as being equal to zero. They define 50% of alcohol as being harmfully consumed. 513,000 New Zealanders fall into the harmful drinking category. I don’t buy any of it.
    http://offsettingbehaviour.blogspot.com/2009/04/berl-redux.html

  • What would Hayek say
  • OK, I tried reading the main body of the report to figure out who to believe, but now I’m totally confused. BERL estimates ‘net social cost’ of drinking, but without including the benefits of consumption. They say it’s normal to do an economic cost study without including benefits, and it’s not a CBA. But then they say that they’re estimating net, not gross, costs. So there must be something subtracted off, just not the benefits.

    After reading the first dozen or so pages, I still have no idea what BERL actually ended up estimating and how to interpret it. I guess I should stay away from commenting on things I don’t understand 🙁

  • ben

    Long post, sorry, but there is so much to talk about with this thing.

    However, they have been very explicit about the approach that they have taken and I would not accuse them of being deceptive or misleading in the way they’ve presented their ideas.

    BERL have also been criticised by Paul Walker and Eric Crampton for not clarifying the scope of their report. While I was initially sympathetic to this criticism, I have since thought better of that stance. I don’t feel that BERL’s report has been misrepresented and commenting on the following work of the Law Commission is unnecessary.

    I think you are bending over backwards in your generosity, Rauparaha.

    I have no problem with the fact that BERL’s report did not include the benefits of drinking. That was outside of the scope of the work that they were commissioned to do and there is no reason why they would consider it. Developers of the law should consider the benefits, but that is not BERL’s job.

    The problem I have is this: what is a report on costs but not benefits good for? BERL must have anticipated that a study of costs but not benefits would be mis-used by the Ministry and probably others to lobby for policy changes. So where is the warning up front saying that’s not what this report is good for, and now that its happened where is the public statement saying that’s not what it’s good for? The more I think about it, the more Slack’s comment that once his work is in the public domain he has no obligation at all to correct misuse of his work by non-experts is close to guilt by association. How is public policy or the public welfare, or your own reputation for that matter, assisted by not doing so?

    He points out that a lot of the costs of drinking are not internalised due to transaction and monitoring costs. For example, being drunk at work might ideally be reflected in your pay, but in reality the costs of changing pay and monitoring that sort of impairment often outweigh the benefits. In that sort of situation, the cost is not internalised and is a net social cost.

    Sorry, Rauparaha, what on earth makes you think firms would have so much trouble detecting bad performance for any reason including being drunk? Don’t firms exist precisely because they reduce these sorts of transactions costs? And don’t firms face strong incentives to detect shirking employees? Remember, BERL is not saying some firms are better than others at this. BERL has assumed NO firms detect any of the lost productivity. And even if that were true there’s still the small matter of showing that this makes lost productivity an externality, which is far from obvious.

    He also elucidates BERL’s approach to addiction. Essentially they reject the idea of alcohol addiction being a rational choice.

    That isn’t quite right. First BERL has set an absurdly low threshold for addiction: slightly more than two standard drinks per day. How this qualifies for addiction I don’t know.

    But then BERL doesn’t something that doesn’t make sense. Irrationality simply means a consumption choice where costs exceed benefits. So does BERL take the conservative route and assume addicted people enjoy benefits = costs – epsilon

    No. BERL does the exact opposite. They assume zero benefits to anybody who consumes harmfully. That is an assumption that produces billions of dollars in apparent costs. I don’t believe that assumption has any basis in economics and I doubt there is any support for it at all in any scientific literature either. I’m not aware of any evidence for utility function having step changes like this.

    In short, BERL appears to have simply made a collection of assumptions designed to confirm the arbitrary of the client. What is the purpose of BERL’s 181 page report except to give credence for paper thin assumptions and a methodology guaranteed from the start to produce a big number for the client. At an absolute minimum, this is a total waste of taxpayers money, in the sense that little or nothing was added to the client’s (or BERL’s) priors. This appears to me to be an exercise in deception.

  • @ben
    1) BERL’s responsibility: A CBA of alcohol use in NZ would be an enormous undertaking. BERL weren’t asked to do it so why would they? They’re not a charity and there is a large opportunity cost to their researchers’ time.

    You have made a value judgment that costs should be weighed up against benefits when judging the harm from alcohol usage. The Law Commission apparently feels differently. BERL have not taken a stance either way, and why would they? They’re not policy makers.

    2) The report: I haven’t really understood what’s going on in the report so I don’t feel the need to comment either way. I can see that there is substance in the concepts, but I don’t know enough to evaluate them.

    If you read the report you’ll find that they draw their threshold for harmful use from epidemiological studies. That seems reasonable to me, taking experts in that field at their word when you have no expertise.

    I’m really surprised that so many people who haven’t read the report feel qualified to criticise it in such detail. It’s written by very qualified, very smart economists so it’s unlikely they didn’t think of these critiques when they wrote it. Sure, there will be valid objections that can be raised to their methods — as there always are — but to call it outright deception is simply hyperbole.

  • Eric Crampton

    Raupahara: the only things netted out are the excise tax revenues paid to government on alcohol, and that the non-harmful half of alcohol consumption is deemed to be costless (so any costs therefrom aren’t added in).

    If something seems unduly obscure, sometimes there’s a reason for it. Look for the monster hiding beneath the sheet….

    Ben: I disagree with you but only very slightly. A pint is more than a standard drink. It’s my standard drink, but more than an official standard drink. http://en.wikipedia.org/wiki/Standard_drink . A half litre is two standard drinks, and a pint is a little more than a half litre, at least in the parts of the world that measure these things correctly.

  • Eric Crampton

    BERL HAS taken a stand by deeming benefits to be zero. Like Ben said, the conservative approach would be to say benefits are slightly less than costs. Instead, they’ve fixed it at zero, which means that a whole lot gets to be counted as costs in a way sure to be misunderstood by folks looking at the conclusions of the report.

    I can buy that the med lit starts finding harmful effects if you’re drinking a bit more than 2 pints a day. But is it anywhere near plausible that drinking at that level means you get zero enjoyment from your drink? Hell no. Instead, that level gives a plausible point for starting to knock back the amount of consumer surplus! I’d believe that. Setting it to zero once you hit 4 ounces more than 2 pints beggars belief.

  • ben

    @rauparaha

    1) BERL’s responsibility: A CBA of alcohol use in NZ would be an enormous undertaking. BERL weren’t asked to do it so why would they? They’re not a charity and there is a large opportunity cost to their researchers’ time.

    You misunderstand me. I am not saying BERL should have done something they weren’t asked to do, and I am certainly not saying they should donate their time to do it. I think my position on this, explained above, is pretty clear.

    You have made a value judgment that costs should be weighed up against benefits when judging the harm from alcohol usage. The Law Commission apparently feels differently. BERL have not taken a stance either way, and why would they? They’re not policy makers.

    I didn’t say that because I don’t believe it. The way to count harm is to define what it means and then go count it. BERL did that, albeit very badly. The LC’s mistake is to think this has policy implications. That is not supported in economics if welfare maximisation is your objective. Pointing out that the welfare effects of a policy cannot be tested using a gross cost analysis is not a value judgment, it’s a fact. Both BERL and LC have made that mistake, BERL explicitly and LC, so far as I have seen, implicitly. And I didn’t say BERL has taken a stance. In fact they are careful in their report not to advocate policy. But theirs is not the methodology of a dispassionate observer. They have produced an analysis using a mehtodology that could not find anything except what it found. The time and effort to write that report was not an information discovery and learning process. It was an exercise in giving credibility to one or two bad assumptions and give the huge numbers resulting from it the veneeer of credibility.

    Here’s a value judgment: whether it is them or their client that ordered those assumptions is uncertain, but I can tell you as a former professional economist I would resign before putting my name to that analysis. I choose not to be in the business of misleading people.

    The fact is that others have misunderstood what their methodology is capable of detecting and BERL has not come out and put them straight, either up front of after the fact. They are apparently content to see non-economists and policymakers deceived by what they have written.

  • Pingback: Did you see the one about . . . « Homepaddock()

  • What would Hayek say

    I agree with Ben and Eric. Loking at the BERL report I’m left wondering why it wasn’t shortened to simply say “The costs to society of alcohol consumption currently exceed the benefits received by society, targetting the generators of those social costs through through the imposition of taxes or restrictions on their access to alcohol would improve overall social well being whilst allowing society to maintain the benefits received from moderate alcohol consumption”. The rest of the 180 pages is almost meaningless as analysis.

    This maybe harsh on BERL and they have probably walked a tricky rope balancing the scope of the brief from the Law Commission and providing a robust evaluation, but when your having those problems you should talk to your client (or put a professional caveat on the report) about how there are problems with the result due to the narrow scope. This has been done on a number of reports I’ve seen.

    Again its like Arthur Anderson siging off the Enron accounts, by looking at the corner of the room we can confirm there is no dancing elephants in the conrner of the room…

  • Adrian cites the BERL report as externally refereed. Apparently that means something different in the consulting world than in the academic world. In the academic world, if my paper built on the seminal model by Professor X, the editor would be sure that Professor X isn’t the only referee. In this report, Collins and Lapsley are the only listed referees (as noted in the acknowledgments) and provide the main methodology employed by the report. Hmm.

  • Pingback: Here are the Long-Term Effects of Alcoholism | Self Help Blog()

  • Pingback: Compulsive Gambling Habits | Self Help Blog()

  • Pingback: Some Long-Term effects of Alcoholism | Self Help Blog()

  • Pingback: Alcohol and Pregnancy - The Dangers | Bady Parenting Blog()