Against Externalities

I have previously written quite a bit against claims about externalities. This article collects those points in one place and adds a bit more info.

Older Points

Additional Points

I propose a rule of thumb in economic policy: The error of an analytical approach is measurable by the share of external effects it identifies.

This is a useful heuristic based on my deeper assertion that externalities don’t actually exist. Apparent externalities are manifestations of analytical error. I argue that externalities don’t exist based on the older articles and a few additional points:

  1. Empirically, low quality studies including obviously slanted CBAs typically identify large external effects to motivate intervention. High quality studies find small external effects, include comparative analysis, find that external effects are smaller than transaction costs, or find that external effects were internalized all along.
    1. Go see a CBAs delivered to any US State Department of Transportation, for example. Opportunity costs, such as benefits from privatization, are generally excluded. Inflation is often ignored. Project costs and maintenance costs are understated, the completion schedules are optimistic, congestion benefits to the public and revenue (in the case of toll roads) are overstated as would be expected in a public choice perspective.
    2. David Friedman vs Climate Scientists on externalities from global warming are a great example. “The sky is falling!” vs “Actually, this could be good for the economy.”
  2. Logically, there is no distinction between a positive and a negative externality other than the opinion of the analyst. Free riders and hold out problems are examples of negative externalities, right? Wrong. A free rider is a slur for someone who enjoys a positive externality. The hold out problem is a scape goat for utility-optimizing investment/gambling based on subjective value.
  3. The best argument against me is perhaps Caplan’s externality from murder argument.
    1. His example is as follows: Person A hires person B to kill person C. A and B are better off, but clearly C is worse off right? I grant this point, but I see several problems in saying this demonstrates an externality problem on the market:
      1. This is not how normal external cost analysis proceeds. Normal external cost anlysis would ask whether society is better or worse off, not whether some person is better or worse off. If society is better off with C dead we would not see an externality under a standard analysis.
        1. In fact, one could argue society is worse off with C alive. Maybe C is a mass murder. But I argue neither. I argue a well-formed market would have externalities in neither direction because the concept is erroneous.
      2. It is not clear that C was not involved in this transaction why would A want C dead? It seems to me the analysis ignores some prior provocation. It seems to me that the inclusion of this provocation would eliminate the existance of an externality.
        1. Consider historical cases as in point 4 for evidence to my point. I consider this like a Newton’s First Law of Economics. Caplan might say “some people are crazy enough to pay something to kill someone for no reason.” I doubt it. A small reason? Yes. No reason? No. And it’s the efficiency of the market definition of “small” which makes the market > government here.
      3. There is no comparative inefficiency if the market handles the situation better than the government would. Caplan is famous for his employment of this rule, but he dodges his own standard here. The alternative to deregulated assassination is government regulated assassination. Typical public choice arguments afflict the latter.
      4. In a situation of futarchy, an assassination market would be created by “predicting” the death of A on some date. It becomes a motive for assassination because a predictor can endogenously gain by causing the death of A on said date. This historically occurred in an underground Bitcoin-denominated assassination market. Bernanke had a hit on him for something like 1.2 Million USD at some point in time.
        1. Why call this inefficient? Immoral, I agree, but in a GDP, GDP growth, or Kaldor-Hicks sense the deregulation of this market is a clear gain to efficiency and welfare.
        2. In a situation of futarchy, if C was innocent then the market would protect him. The assassination would never be profitable. Imo, this is the final blow to the Caplan defense of murder regulation on economic grounds; I admit if he is going to make a moral argument then a whole series of other considerations must be made and he is in fact likely to win. If C is assassinated under futarchy he is guilty by the standard of the most efficient calculation machine; the market. Far better than a typical jury system, etc.
          1. Given my theory that moral and economic efficiency are in harmony, how can I reconcile that Caplan has a moral point but not an economic point? My typical response in the case of gay marriage, abortion, drug use, and all other sort of objectionable activity from my perspective: People should be allowed to do it, they just shouldn’t actually do it. (Eg hire hitmen in this case.)
        3. Permissionless innovation tho.
    2. Interestingly, Caplan and Hanson succumbed to this argument and suggested regulatory futarchy to neutralize this effect. “Register the bettors, and/or cap the bets.”

Maybe it would clear things a bit to restate what I have already said in the fashion I said it on Facebook:

1) I feel like low quality studies find large externalities and high quality studies often notice that the market was internalizing optimally all along.

2) I think a credible argument can be made that there are no real things called externalities, both in theory and in reality. Instead, apparent externalities arise from incorrectly identifying contracts and products.

Interesting extra info: Caplan’s “externality from murder” is perhaps the most clear cut case against me, but there I see a missing presupposition. Why would anyone kill another? There is likely some prior provocation being ignored. A Newton’s First Law of Economics if you will. Similarly, I think an innocent would be protected on an assassination market.

  • 1

Leave a Comment