Where is the externality here?

There have been wild debates surrounding the BERL report into the social costs of alcohol.  I haven’t read the report, I haven’t read the replies, I have to admit I have been busy.

However, in one of Eric Crampton’s many posts on the issue I see that generally an externality from lost output, excessive unemployment, and forgone wages has been assumed in the discussions.  I’m sorry but what?

The labour market is a market, how can we have an externality when there is a market with a market price (wages).  Yes, alcoholics produce less, less of them are employed, and they tend to have lower wages – but this isn’t an externality it is part of the market process.  They are paid less because their marginal product is lower, and they are willing to be paid less because the benefit they receive from consuming alcohol is sufficient compensation – this is a completely internalised decision for the drinker isn’t it, so where is the social cost.

And don’t say it is too the firm – the firm can set a lower wage because of the fact that the marginal product of this worker type is lower.

And if we are going to look at it in terms of society as a whole (which involves moving away from externality logic), sure having a lot of alcoholics lowers our “capacity to produce”, but given that this is the result of a maximising choice by individuals we can say that the benefit of drinking exceeds the cost of this lost production – the fact that people are doing all this drinking illustrates that the drinking is more highly prized among society as a whole than the output they could have produced.

As the ultimate goal of “production” is to lead to create outcomes that satisfy peoples preferences through consumption we can ultimately say that the forgone production is being consumed in an optimal way everytime an alcoholic has a drink – excellent, go alcoholics!

The only market failure I can think of stems from asymmetric information.  A firm hires someone without knowing they are an alcoholic and agrees to pay a wage, through selection we could end up with an adverse selection problem.  But I don’t really buy it, given that these attributes are partially observable, and future wage increases do help us push towards the market clearing price for alcoholic labour.

So convince me that there is an externality here …

  • Incomplete information and incomplete contracts mean that the cost to the firm isn’t efficiently dealt with. I can accept that there’s an efficiency loss here, I just struggle to believe that it’s relevant to policy-making.

  • @rauparaha

    I can definitely only picture it as an asymmetric information problem for contract making. And even then I am struggling to see how policy can help at all.

    Is this one of those rare occasions where we completely agree rauparaha 😛

  • Matt, we were trying to be generous to BERL in this one, believe it or not. BERL was arguing that the costs of harmful drinking, in terms of output forgone, is all of the accumulated lost wages times 1.87: the ratio of GDP to total wages. In other words: the total product. My preferred approach would be to count only lost wages, and those entirely as an internal cost, but we decided to be generous. Some measures of the “value added” for labour put it at 1.05-1.1; we counted that as an external cost when the labour is lost. I don’t like it, but at least it’s a number based on some statistical evidence and broadly within the approach that BERL established. We were trying to stick within their framework and fix their numbers. Counting only forgone wages, and those as internal, would have had BERL just replying that Crampton was being ideological. Oops, they’ve already done that.

    The real external costs from forgone wages that come in from our numbers aren’t from that 1.1 multiplier though: they’re from wages forgone when a drunk driver hits and kills/injures somebody external to the drunk driver’s vehicle. I can buy those as external costs.

  • @Eric Crampton

    Hi Eric,

    I can understand that – I definitely wasn’t aiming to attack your post in anyway as I know you didn’t initially make the argument.

    I was just struggling with the idea of an externality in a place where there is a defined market price between the agents involved in the transaction 😛

  • I do like folks coming in telling me that I’ve overestimated the costs of alcohol though. Please continue.

  • Adrian Slack seriously believes that you impose an externality on your employer if you get drunk at the company Christmas party and show up hung-over and less productive the next day. At least that’s what he told me at the NZAEs. If that’s his starting point, you can see why they’re so strenuously objecting to my using bog-standard economics in trying to correct their numbers.

  • @Eric Crampton

    I get the feeling Adrian is probably talking about moral hazard in this case – once someone has signed a contract with an employer there is the chance for them to increase their drinking and reduce their productivity.

    I can see how that would be an externality, although personally I imagine that this would be captured in future pay rises and could be fully internalised.

  • In any case, the vast bulk of BERL’s forgone wage costs come from premature deaths, not from absenteeism. In that case we need really only consider transitional hiring costs. If the worker can be replaced, then having his total counterfactual product as a cost is a bit much.

  • @Eric Crampton

    “In any case, the vast bulk of BERL’s forgone wage costs come from premature deaths, not from absenteeism”

    I see I see – that is a different issue.

    As you said earlier though, if it comes from a drunk driver that is fine, but if it comes from someone killing themselves I am a bit uncertain about painting it as an externality.

  • I don’t think they’d have considered forgone wages as an external cost; they’d count them as an internal cost that get to count as a social cost by virtue of there being zero benefits from harmful alcohol consumption. The multiplier part they’d probably consider external, but nowhere do they really make any note of the difference between external and internal (except for the note at page 173 that describes why they count internal costs as social costs).

  • @Eric Crampton

    I suppose by definition social cost = internal cost + external cost. Of course, in such a case it is important to remember what is policy relevant or else we would end up regulating anything with a cost 😛

    I also have to point out here that I am not attacking the BERL report as I haven’t read it, or much about it. I’m still busily working on post forecast things.

    I just wanted to make sure I wasn’t getting confused when I felt that people treating alcoholic workers as an externality is a bit weird. I like to think that my level of alcoholism is nicely incorporated in my wage 🙂

  • Robbie

    Big change of tack, but won’t somebody think of the children? Surely the people that are seriously disadvantaged by alcoholism are the children of alcoholics.

    I assert that not only are drunk people worse at parenting(anyone is less willing and indeed able to help with reading after a bottle of bourbon) but more inclined to commit crimes against their children (abuse). This is traumatic for the children at the time, and also has repercussions on their learning, development, and outlook on life.

  • @Robbie

    Indeed. However in so far as thinking about the regulation of alcohol, this matters insofar as we think the average alcoholic parent does not value their child relative to a non-alcoholic parent. This has to be the policy relevant point, and I’m not sure that a tax would be the best way to fix that.

    The parent child relationship is always a tough one because of the asymmetry in power, both legally and socially. Furthermore, we should recognise that it matters for development. However, quantifying the specific cost of alcoholic parents, and figuring out the policy relevant way of dealing with it, is a difficult question.

    Furthermore, any focus on that question without the realisation that non-alcoholic parents can do similar things, and without an understanding of the relationship between alcohol and other facts, would likely lead to a greater than optimal amount of state intervention.

    We may see cases where we wish the state did intervene at the moment, but is the benefit of intervention in these cases worth the cost of interference in other healthy parental relationships?

  • was nice article.
    thank you.

  • very nice post. dude, this blog is powerful.. 😉

    im write here:

    http://www.mert2407.com/

  • I advance that not alone are bashed humans worse at parenting(anyone is beneath accommodating and absolutely able to advice with account afterwards a canteen of bourbon) but added absorbed to accomplish crimes adjoin their accouchement (abuse). This is alarming for the accouchement at the time, and aswell has repercussions on their learning, development, and angle on life. Big change of tack, but won’t somebody anticipate of the children? Surely the humans that are actively disadvantaged by alcoholism are the accouchement of alcoholics.

  • I just wanted to make sure I wasn’t getting confused when I felt that people treating alcoholic workers as an externality is a bit weird.

  • Incomplete information and incomplete contracts mean that the cost to the firm isn’t efficiently dealt with. I can accept that there’s an efficiency loss here, I just struggle to believe that it’s relevant to policy-making.

    I generally agree with this statement. Efficiency loss in evident in this market.

  • Where is the social cost??? I’m sorry if I get a tad bit high pitched here, but are you kidding me. Look around. It’s not just a work thing. Yes employers stand to lose quit a bit if they have drunk employees. I did a few times. It goes much deeper then this though. It’s not only in the work place we lose because of excessive drinking. Wake up and just look out our county jails and prisons. They are overflowing with drug and alcohol offenders. Alcohol kills more people each year then all of the other drugs combined. And to top it off, us non drinking tax payers have to dip into our pockets every year to fund state run rehabilitation clinics so that some drunk can get sober only to go back out on the street and start the process all over again. If society is to blame should really be blamed for being so tolerant of alcoholism as a whole. Yes there is a high social cost of alcohol. It is our tolerance of it…

  • “It’s not only in the work place we lose because of excessive drinking. Wake up and just look out our county jails and prisons. They are overflowing with drug and alcohol offenders … etc”

    I was not saying there is no external cost from alcohol – far from it, we have accepted an external cost in the past, and discussed how to deal with it.

    I was saying that I don’t think THIS SPECIFIC external cost exists. Such as when I said this at the start:

    “I see that generally an externality from lost output, excessive unemployment, and forgone wages has been assumed in the discussions.”

  • Pingback: Curious Cat Investing and Economics Carnival #3 at Curious Cat Investing and Economics Blog()