04 May 2021

Tyler Cowen’s Expansive Definition of Externalities

David R. Henderson, Health economics, Tyler Cowen 24 Comments

Jeff Bezos can spend his money on himself, or he could give it to other people, so they could spend it on themselves. If he opts for the former, this is a clear-cut negative externality. Right?

I think most economists would agree that that is a terrible argument. And yet, it seems to be what Tyler Cowen says, regarding old people spending their own money on themselves before they die, rather than passing it on to others. David R. Henderson shares my confusion.

24 Responses to “Tyler Cowen’s Expansive Definition of Externalities”

  1. Andrew_FL says:

    This is actually the entire implicit premise behind the false notion that we spend “too much” on healthcare in general-since a huge chunk of that spending is old people spending their own money on themselves.

    Folks like Tyler have bought so hard into the conventional wisdom of “too much healthcare spending” that it completely warps their reasoning on the subject.

    • Capt. J Parker says:

      You said: “a huge chunk of that spending is old people spending their own money on themselves.” Are you sure this is the case? My prior is that most if this spending is old people spending other peoples money on themselves either through medicare, or other insurance which transfers costs away from the elderly on onto younger and healthier people.

      • random person says:

        Medicare / insurance that they almost certainly paid into when they were younger. I believe the rules are to qualify for quote-on-quote “free” Medicare Part A, one has to have paid Medicare taxes for at least 10 years — or in other words, it’s not really free, it’s something they simply paid for in the past.

        Certainly, there are many valid criticisms to be made of Medicare and other insurances in the United States, but it’s not as if, as a general rule, elderly people are simply freeloaders spending other people’s money. Except maybe if the accountants manage to trace back that money to forced labor or some other crime against humanity, i.e. if the income they earned wasn’t legitimate income to begin with, but that would still be a case-by-case thing.

        • Capt. J Parker says:

          random person,
          I am not trying to make any moral assessment of Medicare users or Medicare in general. My point is simply that when Medicare recipients make decisions about consuming healthcare they don’t personally bear much of the cos. of those decisions and so they are not as Andrew_FL contends “spending their own money on themselves.” One possible consequence of this is Medicare recipients might consume more healthcare than they otherwise would and it is quite possible that the claim that we spend “too much” on healthcare would be valid.

          • random person says:

            Apologies for the long response time.

            I think it can go both ways — it’s possible that Medicare users might spend more on themselves that they otherwise would (if Medicare will let them), and it’s also possible that Medicare will cut services because they already got the Medicare users’ money when they were younger and now lack incentive to provide a good quality service. It’s also possible, and likely, that many medical providers will charge more than they would if people didn’t have Medicare — a theory supported by travelling to Mexico and finding nearly as good medical care there at much cheaper prices.

            Also, I recently saw Paris Hilton’s new documentary, “This is Paris,” in which she describes, among other things, being tortured at a child torture facility in Utah. Looking at some old news articles about the child torture facility in question, it appears that they receive some funding from Medicaid. In Paris Hilton’s case, it seems her parents paid for her to be tortured, but apparently, in some other cases, Medicaid is paying for children to be tortured. (Disclaimer: the documentary gives Paris Hilton’s parents the benefit of the doubt, and assumes they didn’t know that Paris Hilton was being tortured though, even though the way they apparently took her to the child torture facility was to send men to kidnap her in the middle of the night in front of her parents while she screamed.)

            I am reminded of a documentary, “Pray the Devil Back to Hell,” when they talked about how the murderous raping warmongers were arguing about government positions at some “peace” talk, and the way one of the pacifist women responded to this was by asking, “You want us to pay you for killing us?” Anyway, the pacifist women ended up winning a war, establishing a democracy, and not giving those government positions to the warmongers, so far as I know.

            To modify that sentiment to fit the issue of Medicaid being used to pay for child torture, I’m thinking, “They want us to pay them for torturing children?”

            • Harold says:

              These institutions are apparently called “emotional growth schools”

              There may be some good one.

              • random person says:

                I do not know if there is some good one. Perhaps there are, perhaps that is how the torture facilities get away with it — by using the same marketing as good schools and blending in with them.

                But whether or not there are good ones is not my point. There are child torture facilities, and at least some of them are apparently receiving Medicaid funding. They should not be. They should not be receiving Medicaid funding, and they should not be allowed to continue operating at all, with or without Medicaid funding. And, for that matter, they also owe reparations to the children they’ve tortured.

                I remember in her documentary, Paris Hilton said she wanted to go in there and get those children out, except she couldn’t because they would probably go nuts and call the police. Which is probably a valid concern.

                What does it say about our society that one expects the police to side with child torturers over people who wish to rescue children from torture?

      • Andrew_FL says:

        It’s both. Even without medicare, the elderly would spend a lot on healthcare.

        • Capt. J Parker says:

          Andrew, you said: ” Even without medicare, the elderly would spend a lot on healthcare.” But the point being debated is do we spend “too much” on healthcare because the systems we have in place now value the lives of the elderly more than they themselves would value them. So the question is: without medicare, would the elderly spend less? If they spent less would we now no longer be “spending too much” on healthcare?

  2. gedeBard says:

    I apologise, but, in my opinion, you are not right.

  3. Capt. J Parker says:

    I totally get the libertarian objection to saying my spending of my money on me is some kind of externality. But I have to agree with Dr. Cowen’s thesis that “willingness to pay” measures of the value of a life will diverge markedly from societal “value of a statistical life” measures for low human capital high net worth individuals. This shouldn’t be in a world with perfect human capital markets and rational economic actors unless there is some externality.

    • Harold says:

      My understanding was that VSL was based on WTP calculations. It gives a different answer to QALY based systems.

      The thing about statistical lives is that must remain statistical. Once we start pinning them down to sub groups, the measure does not work anymore.

      VSL correlates well with wealth. If we did use sub-groups, then the Govt would spend more to save a wealthy life than a poor one. As it is, the Govt uses one value for all lives and decides on spending accordingly. We don’t presume to know if the life is wealthy or poor.

      QALY’s on the other hand, are individually based. We need to identify the individual, or a narrow group of similar individuals, to even start to apply the method.

      Since we do know something about who dies with Covid, there is at least some justification for using the QALY type approach, but it is a difficult area.

      If average life expectancy (from now) in US is 40 years, and VSL is $10M, average life expectancy are age of death from Covid is 8 years. We should value each life at say $2M.

      I really don’t know the best way to do this. I have certainly seen both measures used, but my impression is that economists seem to use VSL more.

      • Capt. J Parker says:

        Thanks for the correction. Now that I look at the Hopedale link more closely I see my error. Indeed WTP data is used to generate VSL curves.

        Still, Cowen (and Kaplan) are complaining that VSL is deficient in some respects that makes it place too high a value on older lives compared to younger ones based on their intuitive understanding of what a “correct” valuation scheme would look like. Those “correct” schemes in the case of Caplan look more like QALY as you point out. Cowen seems to me to be saying life valuation should be something akin to NPV of ones human capital and this really should be the same a VSL but as Cowen points out, when it come to avoiding certain death WTP = spend everything I own I don’t see the great error in hypothesizing that this is still rational behavior and the divergence between one’s WTP in this instance and what NPV of human capital says you ought to pay is due to an externality.

        I have to say that I think Caplan overstates the case against Hopedale’s VSL curve. We have plenty of societal support for the idea that the lives of the young are worth less. Consider third world societies today (and the US and Europe say 100 years ago) The norm is to have families with may kids and the expectations that many won’t live to adulthood.

        And then there’s this bit from the movie “A Civil Action” which I always think about when the topic turns to how to value a human life.

        Jan Schlichtmannn (narrating) It’s like this. A dead plaintiff is rarely worth as much as a living, severely-maimed plaintiff. However, if it’s a long slow agonizing death, as opposed to a quick drowning or car wreck, the value can rise considerably. A dead adult in his 20s is generally worth less than one who is middle aged. A dead woman less than a dead man. A single adult less than one who’s married. Black less than white. Poor less than rich. The perfect victim is a white male professional, 40 years old, at the height of his earning power, struck down in his prime. And the most imperfect? Well, in the calculus of personal injury law, a dead child is worth the least of all.

        • Harold says:

          I have thought about this quite bit, but have no answers as yet. here are some of my thoughts.

          The key thing is that we are talking about policy. We are not asking what a life is actually worth, but how much a Govt or agency out to spend to prevent someone dying. Hence we get comments like “you can’t put a value on human life”, which is sort of right in one context, but Govts still have to allocate money somehow.

          I think Govts generally take the entire country as a whole, so we get VSL for America, or Britian, or Bangladesh, which are quite different. The thinking is I suppose that the Govt cannot predict which of the citizens will die.

          Americans on average spend enough on risk reduction to calculate $10M for the VSL. If we were to divide the population into sectors, we would get very different VSL,a as we do for Bangladesh. VSL correaltes well with wealth. Very poor people would have a low VSL and rich people a high VS because they spend less on risk reduction. The Govt “should” therefore spend less on a policy that was targeted as saving the lives of poor people compared to rich people. There is some resistance to this politically, so generally no such sub division is made.

          For Covid, there seems not much evidence that older people spend very much less then younger people to reduce risk, so the VSL should be the “standard” one, whether we are thinking about sub-dividing the population or not.

          Which brings us to QALY’s. The QALY principle treats everyone’s healthy life year as equal regardless of wealth. The Govt or whomever must decide how much to spend per QALY. This would result in less spending to save old people.

          Some argue for a reduced value from the standard VSL because we do know the age profile of the expected deaths, but this ignores the calculation of VSL for that demographic. They use the value of a statistical life year (VSLY) which simply chops up the VSL into years of life expected. This seems an odd hybrid method to me, with no real justification for distorting the WTP principle used to derive VSL in this particular manner.

          We can justify a wide range of values to use, which means the discussion will continue. We can put limits of not more than VSL and not less than QALY based on say $100,000 per QALY, which gives us roughly between $1M and $10M per life for USA.

          • random person says:

            Interesting that you should mention a part of the world that was, in the past, subject to genocidal levels of taxation at the hands of the East India Company, as an example for the argument that, “The Govt ‘should’ therefore spend less on a policy that was targeted as saving the lives of poor people compared to rich people.”

            In Bengal, which if I understand my geography right covered much of modern day Bangladesh, was, under East India Company rule, subject to levels of taxation so severe, that it rose to the level of genocide.

            (Please note that I disagree with the way the author of that article worded the headline. There have been so many genocides throughout history, that it is presumptuous to label whichever genocide one happens to be studying as the “worst” one in history.)

            Incidentally, the fact that the East India Company was responsible for this proves a point I’ve made a number of times already that there is no fundamental difference between a company and a government. The East India Company wielded taxation as a tool of capitalist exploitation and got rich off of committing genocide just as easily as many of history’s brutal governments could have.

            But, from studying the history of Bengal / Bangladesh, we can see that it is not right that government or corporations should automatically spend less on poor people, especially in those situations where they are responsible for making those people poor to begin with. Rather, the amount the East India Company should have returned as much as they had looted from the Bengal people, plus interest. But, of course, that would have defeated their goal, which wasn’t to do the right thing, but rather, to make a load of profit.

            • random person says:

              As a disclaimer, the argument I am making does not apply to when a government taxes enslavers or other violent criminals (including those who commit their violence by means of command responsibility).

              So, for example, if a government taxed Isaac Franklin, a notoriously evil slave trader in the 1820s and 1830s, Isaac Franklin would have no moral right to expect any of that money back from the government in question. Rather, the government in question should spend that money on paying reparations to Isaac Franklin’s victims.


            • random person says:

              I wrote a disclaimer that got stuck in the moderation queue, but to summarize it:

              As a disclaimer, the argument I am making does not apply to when a government taxes people who subject others to forced labor or other violent criminals (including those who commit their violence by means of command responsibility).

              So, for example, if a government taxed Isaac Franklin, a notoriously evil human trafficker in the 1820s and 1830s, Isaac Franklin would have no moral right to expect any of that money back from the government in question. Rather, the government in question should spend that money on paying reparations to Isaac Franklin’s victims.

  4. Tel says:

    As soon as Tyler Cowan says, “And make capital markets perfect” … it’s obvious he is joking around with you guys and testing to see who picks up on it.

    Imagine some guy is diagnosed with cancer and tells you, “Loan me half a million because after I’m cured, in the next 20 years I will produce a great architectural masterpiece and pay you back at 5% interest”. Are you going to lend to a terminal cancer patient because he says he will to something brilliant, some time or other? Would a bank lend this?

    Even if he does produce that masterpiece and then dies before he pays you back, what’s the chance you actually recover your money?

  5. Major_Freedom says:

    “The lives with more human capital will be worth more, at least according to economic standards.”

    This isn’t a statement of “standard economics”. It’s a personal value judgment of the writer to replace that which is in fact the “standard”, namely, human life itself. Cowen is inverting the value if LIFE with the value of LABOR.

    One person’s LABOR may be more valuable than another person’s LABOR on the open market, but nowhere in economic science does this permit an objective assessment of the value of the person’s life itself as being more or less valuable than another.

    If Cowen claims by some reference to “standard economics” that the LIFE of person X is less valuable than the LIFE of person Y, it is still the case that person X would not be incorrect to regard their own life is more important than what Cowen subjectively values based upon how Cowen might personally benefit from person X, and they would not be violating any economic law or principle in saying that.

    Human life is the standard of any rational economics. A high net worth retiree’s LIFE is their own highest value, and a low net worth recent college graduate’s LIFE is their own highest value.

    Economics isn’t a science that puts an objective value on people. It’s a tool of and for people.

    Contrary to Cowen’s claim that he isn’t getting into morality, the truth is that it is ONLY by way of morality can anyone be told that “this is standard economics” which underneath the hood is “we ought to value one human life as more than another”.

    I believe nobody is above another. Each life is “worth” exactly the same. I may value one person’s labor more than another, but this is not an assessment of their LIFE. Marxists may equate human life with labor, but I don’t.

    The error of inverting the true standard, i.e. human life, with a subordinate standard, i.e. market price of labor, is what lead Cowen to make the error that old people spending their money on themselves has some sort of a “negative externality” economic outcome.

    In reality all Cowen is saying we have a moral obligation to divide humanity into classes of higher valued PEOPLE and lower valued PEOPLE, based on NOTHING BUT their own labor, and we’re supposed to believe it because that’s what “standard economics” says. There’s a tacit moral claim here that can’t be ignored just because it’s labelled as “standard economics”.

    Well by that “standard economics” Cowen’s argument logically implies that “the economy would grow” if all high net worth retirees were looted for the benefit of younger more productive people.

    Of course, there is a reason INCENTIVES played no role in the article, for if this “standard economics” were put into practise, nobody would save for their retirement and we can all say goodbye to capital accumulation. For who would work hard and save and invest if they expected Cowen’s “standard economics” to guide economic policy?

    Remember the good old days of “boring” economics when the rewards of prosperity in retirement years was an incentive for younger people to save and invest? If only those articles got as many clicks…

  6. random person says:

    Bob Murphy wrote,

    Jeff Bezos can spend his money on himself, or he could give it to other people, so they could spend it on themselves. If he opts for the former, this is a clear-cut negative externality. Right?

    I thought about this for awhile, and I concluded that the “negative externality” occurs when Jeff Bezos does one of two things:

    1. Earns his money in an immoral way. (So if Amazon sells something that was produced by wrongfully arrested Chinese people being forced to work in prison, and he profits off that sale, that money was earned immorally. There are cases of religious dissidents being sent to prison in China and forced to produce things for American consumers, for example. Since Amazon has likely sold some of those products, and since Jeff Bezos probably made some profit off those sales, some portion of the money he has earned in his life would have been immorally earned.)
    2. Spends his money in an immoral way. So, for example, if he bought Christmas lights that were made with forced labor in a Chinese prison, this would be an immoral way to spend money.

Leave a Reply