Externalities: A bridge too far
CPW sent me a link to the following blog post on Econlog. In the post Bryan Caplan mentions an economist from Princeton (Roland Benabou), who argues that externalities provide a bridge between an economists conception of the world, and non-economists concepts. Although this may be a tad over the top (as non-economists place more value in normative statements than economists would ideally), I believe this is an important point insofar as it allows us to generalise our models, to take into account more possible states of the world.
Bryan Caplan puts forward four points of difference that he believes will still exist between economists and non-economists, however I think they were a touch over-cooked, here’s why:
Note: I will paraphrase the points, so they are in my words. If you think I messed it up, and that the author meant something else, tell me.
- The concept relies on economists idea of ‘willingness to pay’, where non-economists believe in fairness, intrinsic value, equality etc
Surely the economists conception of willingness to pay can also be related to ‘fairness’ and ‘intrinsic value’. As peoples willingness to pay is determined by the value they receive from that specific outcome, I can’t see how it is actually separate from the concept of fairness etc. If people actually value these things, they will be willing to pay to preserve/create them. This makes sense if people gain value from ‘doing the right thing’.
However, you may argue that if people have an intrinsic value for the environment, and the individuals contribution does very little they may free-ride. You would be right! In this case we require some institutional struction (eg government, moral code) to enforce payment, and solve the potential prisoners dilemma problem.
- We solve externalities with a tax or subsidy. Regulation is inappropriate, however the public will still want it.
Not true. As we have discussed often on this blog, banning something is like setting the price infinitely high (or possibly lower in the face of a black market!), and so is like a great big tax, (just with no revenues from removing money from the economy). If it is optimal for society to consume very little or none of a good, it may be administratively cheaper to ban it than it is to tax it, making that preferable.
It is likely that their is a bias towards regulation in society, however we can not help society move past this bias without understanding why they view regulation as appropriate. The truth is that many people in society see the ‘costs’ associated with goods, before they see any benefit – as a result, they believe more good fit into the category where regulation is more efficient than really do.
- Externalities are non-excludable costs and benefits, as a result there are a lot of issues that do not fit into the externality frame.
If we include ‘intrinsic values’ into a persons willingness to pay, then we can make anything an externality. Of course this all depends on value judgments. However, stating that such social values do not exist involves making a value judgment – namely that the value of such things is zero. Sometimes economists are too happy to make this judgment without positioning it as either a ‘simplifying assumption’ (this is what I like to do) or a full scale normative statement.
- Externalities can effect government as well, and have a cost to solve. As a result, not all externalities are worth ‘solving’.
Yes. However, people in society may not understand the full cost associated with a given policy. If economists can frame an issue in general externality associated terms, and then do a cost-benefit analysis, we can derive whether a policy is worth-while. The fact that some externalities may not be efficient does not mean we are wrong to ask about them (eg fat tax)
Ultimately, the purpose of this blog is too explore issues such as externalities. Any concept that ‘generalises’ economic models is useful insofar as it gives us more scope to turn around and work out what assumptions we have to make in order to find a certain result. Although talking about externalities may not make economists and non-economists say exactly the same things, it may help us see eye to eye on more issues.
I think the Caplan’s first point is crucial. Non-economists find it very hard to deal with issues, especially emotive ones, in a monetary framework. I suspect a question like: “can you put a value on human life?” would generate a huge split between economists and non-economists.
I share his concern that externalities just becomes a lazy excuse to justify intervention. Cigarettes are a good example where an honest externality framework could well lead to the conclusion that cigarette taxes should be lower (at least in terms of the public sector cost externality).
“I think the Caplan’s first point is crucial. Non-economists find it very hard to deal with issues, especially emotive ones, in a monetary framework”
I agree that the first point is the crux of the issue. Furthermore, I am certain that non-economists and economists would view costs and benefits differently, that is why I said “(as non-economists place more value in normative statements than economists would ideally)”.
“I share his concern that externalities just becomes a lazy excuse to justify intervention.”
My concern is that economists ignore externalities in order to state that their policy is objective, ignoring a cost is just the same as assuming that it equals zero – it is a normative judgment.
“Cigarettes are a good example where an honest externality framework could well lead to the conclusion that cigarette taxes should be lower”
Definitely, we have all talked about that here before: http://tvhe.wordpress.com/2007/09/03/cigarette-taxes/
I think this fits into the case when “many people in society see the ‘costs’ associated with goods, before they see any benefit – as a result, they believe more good fit into the category where regulation is more efficient than really do” – this bias will exist, however it can be countered by information.
“My concern is that economists ignore externalities in order to state that their policy is objective”
Examples? To be fair, I don’t think I’ve ever seen a politician or advocacy group fully account for the costs of their policies, so it can’t just be economists.
“To be fair, I don’t think I’ve ever seen a politician or advocacy group fully account for the costs of their policies, so it can’t just be economists.”
Interest groups have an incentive to ignore costs if they think they can get away with it, economists shouldn’t.
“Examples?”
Anytime an economist comes to a conclusion without a (implicit) general model or stating the set of simplifying assumptions. Economists do this constantly (well at least I do), and I think it is just as concerning as people ‘over-cooking’ an externality and ignoring government failure (as it is implicitly the same sort of thing). An example from the same blog would be:
http://econlog.econlib.org/archives/2008/02/dogs_and_the_en.html
I realise that this was just a throw away type post, and that it is impossible to state all your assumptions etc. But there are some obvious social costs that should be involved in analysis before we try to make any sort of policy prescription. Dumping these social issues is fine if you want to just look at technical efficiency, however it is useless if you want to make any sort of normative claim.
I’m afraid I’m just not as confident about the Coase Theorem always coming to save the day as the guys at GMU are.
I don’t know if what you want is possible most of the time. Most policy issues are pretty micro, it’s hard to cover the total social implications (if that’s what you mean by a general model) because the analyst has no say over what the opportunity costs ultimately will be in terms of foregone policies. Externality taxes are a good example of what I mean – they may not stack up from a cost benefit view in isolation, but they may still be more efficient that some other form of tax. The furor over the Stern report also showed that there is no particular consensus on how to measure the social costs once we throw in a time dimension.
Maybe I’m biased, but I think economists do normally state their assumptions – or maybe the word economist should just automatically connote a framework that implies utility maximization and demand and supply curves that slope in standard directions. Perhaps for non-economists it is hard to determine whether economists are making objective or normative statements. That argument about tax cuts and productivity the other day struck me as an example of this.
The post you cite strikes me as a good example of the Caplan’s point on externalities. The typical response of someone worried about global warming (the externality) is to ban or regulate something like SUVs, not to ban or regulate dogs. The externality becomes an excuse to exercise existing preferences, not a starting point for finding an efficient solution. The “Extermination of dogs” line is throwaway, but less dogs does seem like a likely consequence of full carbon pricing.
GMU’s defining characteristic is intervention skepticism, not market faith.
I agree that a completely general model may be impractical (potentially even useless), however I feel that there has to be some slack both ways. I got the feeling from the Caplan post that he was saying things that went a bit far.
Firstly, and most importantly, the intrinsic value of things to society does matter for normative analysis – and so we should recognise it if we move from our objective seat to our prescriptive mode. Saying that an economic externality does can not include these things seems inappropriate too me. Saying that we should exclude these things during the objective part of our analysis is fine.
Secondly, the economists bias against regulation. Yes regulation is often inappropriate, but there is a transaction cost argument for it that is valid.
Nothing to add for the third thing, and in the fourth thing he is right, the government also causes failures. However, I tend to think that a lot of non-economists also get this, and if they don’t using externality arguments would help!
The language of externalities generalises our economic models, and I expect we could fit in a lot of the ‘non-economists’ point of views into this frame by assuming some value judgments.
Ultimately, I think us economists have things to learn from non-economists and vice versa – I don’t see it as the one way street Caplan does. I do think our methods are the best, but the rest of the world outside economics does have useful insights.
I spend 2 hours lecturing on externalities in my Econ and Current Policy issues class. I still don’t think that gives anywhere near adequate coverage to the topic. It takes at least two hours to hammer home that policy relevant externalities aren’t just any externalities, they’re ones that are technological (as opposed to pecuniary), Pareto-relevant, and haven’t already been internalized via some bit of contracting. Screaming baby on a plane? Not an externality: I have a contract with the airline for a certain class of service, and I haven’t paid them to guarantee me a no-screaming-baby plane. Smelly guy on a bus? Same thing.
Again: two hours to define what kinds of externalities can potentially be improved upon via Pigovean taxes and subsidies. The five minute version leaves folks WAY too happy to give an econ justification for taxing any damned thing they happen not to like.
Hi Eric
“It takes at least two hours to hammer home that policy relevant externalities aren’t just any externalities, they’re ones that are technological (as opposed to pecuniary), Pareto-relevant, and haven’t already been internalized via some bit of contracting.”
I’m not disagreeing that there are plenty of externalities that the government shouldn’t touch – as it would just make things worse (either through institutional mismanagement, perverse incentives, or the fact that the cost of the scheme exceeds the benefit).
However, I do think that the short definition of externalities (when there is a cost or benefit to some third party not involved in a transaction) is extremely useful. I agree that we can paint most things as externalities when we do this – but doesn’t that mean that we can use externalities to generalise our models, and thereby define what types of normative assumptions people must be placing on our logical framework in order to come up with their conclusions.
If this is the case, the language of externalities is useful, insofar as it allows us to see what sort of normative judgments people in society are making, vs the implicit assumptions we make.
I am not saying we should just do what every interest group is saying because it could be defending by an externality argument. I am saying we can clarify the differences between two positions by appealing to externalities – thereby making it clearer to everyone whats going on.
“The five minute version leaves folks WAY too happy to give an econ justification for taxing any damned thing they happen not to like”
Indeed it does – but this does not make the externality argument wrong. People are often happy to moan about the cost associated with an externality – but they don’t think about the cost associated with solving it. If their externality argument is unjustifiable we should be able to make a case that states that the costs of solving the problem are greater than the benefits from solving it – I think this is preferable to ignoring the externality altogether.
Hi, nice post. I couldn’t understand some parts of the article but it sounds interesting..
Continue writing…