Henderson on Armchair

David Henderson reviews the new revised Armchair Economist.

Click here to comment or read others’ comments.

Share/Save

34 Responses to “Henderson on Armchair”


  1. 1 1 Mark Draughn

    Well? Don’t keep us in suspense. Do you think he’s right about your drug legalization mistake?

  2. 2 2 nobody.really

    I keep stubbing my toe on this:

    Consider the choice between building a parking lot and leaving the land as wilderness. Environmentalists argue that the decision to pave is irrevocable. Landsburg grants that point but points out that the decision not to pave is also irrevocable in an important way: people’s opportunity to park there today, if the parking lot isn’t built, is lost. “The ability to park in a more distant future,” he notes, “might be a quite inadequate substitute for that lost opportunity..”

    I don’t get this. Oil is an exhaustible resource. Allowing drilling on federal land does not change the supply of oil and it does not change the demand for oil, so how can it change the price of oil?.

    We need to maintain defense spending because (among other reasons) it “creates jobs”. If Mitt Romney does not understand that the creation of jobs — i.e. the consumption of resources that could be valuably employed elsewhere — is the downside of defense spending, then he has no better chance of “fixing the economy” than a blue-bottomed monkey.

    [C]oncerns about “the consumption of resources that could be valuably employed elsewhere” are more potent when, in fact, there is an “elsewhere” that is actively seeking to employ those assets [i.e., labor hours]. Right now a lot of those assets are on the couch [i.e., perishing].

    What’s the optimal way to price admission to a parking lot if you can’t measure how long people will stay or have stayed? If I wasn’t sufficiently motivated to pursue an affair with that classmate before now, does the fact that she’s about to graduate and leave change my calculus?

    I don’t have a good mental framework for evaluating the distinctions between the durable and the perishable. If I know that my opportunity to choose a good/service is diminishing by the day, how should I model the price I’m willing to pay? If I know that my opportunity to sell a good or service is diminishing by the day, how should I model the price I demand?

  3. 3 3 Jeffrey

    From the review:

    “And here’s where I can extend Landsburg’s thinking in a way that I think he would approve of. … So the big deadweight loss from taxes is spending.”

    That seems like an odd argument. Suppose I want a $500 computer, and suppose the government passes a “buy people a computer tax.” Suppose my share of the tax comes out to $500 and I receive the exact computer that I would have purchased.

    I’m still going to want to evade the tax, because if I evade, I get the computer without the tax. As long as we add the extremely plausible assumption that the law doesn’t require an individual’s tax to equal their benefits, we still have deadweight losses, even with an omniscient government and with zero collection costs.

  4. 4 4 David R. Henderson

    @Jeffrey,
    Here’s the whole quote from my review:

    If, he explained, the government took $10,000 in taxes from you and spent it exactly the way you would have spent it, and if collection costs were zero, taxes wouldn’t matter. Also, you wouldn’t try to avoid them. So the big deadweight loss from taxes is—-spending.

    I was constrained by space, but here’s the idea. Let’s say you want to buy the $500 computer. But instead of you buying it, the government says “We will tax you, Jeffrey, $500, to buy the exact same computer you wanted to buy.” Then there’s no deadweight loss. You’re assuming that it’s a generalized “computer tax.” No, it’s a tax specifically on you to buy what you would have bought.

  5. 5 5 Tristan

    I’ve read Armchair Economist through several times (but never the new one incidentally) and I still don’t quite understand/agree with the irrevocable loss of parking space thing.

    I’ve always assumed the mathematical translation of the environmentalist’s argument hinged on uncertainty. They suppose that the value we would derive from a forest or a parking lot is uncertain fifty years from now (call the current day period 1 and fifty years from now period 2). If, in period 2, we get to observe the actual value of those alternatives and it turns out the forest was really valuable, we’ll be very glad that we preserved the forest. However, if it turns out the parking lot is better to have then we wouldn’t be all that sorry to have preserved the forest since we can just chop it down in period 2.

    In other words, eliminating a forest is a risk because it might turn out to be very valuable, but holding off on building a parking lot is not as big a risk because we can always build it when we realize our mistake. As we all know, risky behavior is avoided to some extent.

    Of course, this is all moderated by the fact that our current consumption also matters. But efficiency requires that future value also be considered. My conclusion has always been that the precautionary principle should hold some non-zero weight, not nothing as Landsburg seems to suggest. Criticism of my argument appreciated.

  6. 6 6 Henri Hein

    Tristan, the value of a parking lot is not the same in the future as it is now. Presumably, somebody wants to build a parking lot because there is a demand for parking. That demand may fall in the future, reducing the value of the parking lot. Even if it doesnt, 50 years later you would have lost the value of 50 years of parking (minus construction costs and the cost of the lost forest — but assuming that equation comes out positive, that value is forever lost.)

  7. 7 7 Tristan

    Henri-
    I’m not denying the possibility that the parking lot will be better than the forests, even in the long run. My point is that a forest is fundamentally different than a parking lot because it can’t be restored (technically it can but you know what I mean) once we learn new information about its value and the parking lot’s value, so it is a big risk to eliminate it but it’s not as big a risk to eliminate the parking lot (as you are effectively doing when you forgo building it).

    If this weren’t a cumbersome text box I would lay out a two period model to try and make this more precise, but hopefully you can fill in the gaps.

  8. 8 8 Ken B

    @Tristan: re 8, it depends on the value of the parking right now. I could make up a scenario but it’s easier to make an analogy. Damming a river is similarly hard to revoke, but the present or near term value — power, flood control, recreation — can be considerable. Just saying ‘irrevocable’ without discussing the costs and benefits isn’t very helpful. Eradicating smallpox is ‘irrevocable’ too.

  9. 9 9 Tristan

    I think we’re all in agreement. My gripe was that Landsburg was implying that it doesn’t matter at all if a resource is non-replaceable (like a forest as opposed to a parking lot). True, they are both irreplaceable in the time between period 1 and period 2, but only the forest is irreplaceable after that, which makes a difference because we often don’t learn critical information about resources until later!

  10. 10 10 Ken B

    @Tristan: ” Landsburg was implying that it doesn’t matter at all if a resource is non-replaceable ..” Nah, I don’t think that’s what he said or implied. He was pointing out that the ‘irrevocable’ cuts both ways. He’s attacking a sort of knee-jerk brandishing of ‘irrevocable’, not minimizing the high cost of error.

  11. 11 11 Tristan

    Ken-

    A knee jerk reaction that the forest is different than the parking lot is justified. To put it in the mathematical terms of one possible model that I mentioned above, one of these can be restored at period 2 and the other can’t, which makes them fundamentally different. The environmentalists may be wrong about a lot of things, but they are right that a forest deserves some non-zero weight due to its “irreplaceability”.

  12. 12 12 Neil

    David R. Henderson

    Suppose the government is buying everyone a computer and everyone wants one. It finances the program with a tax on labor. The tax causes everyone to reduce labor supply trying to shift the tax burden to the other guy causing DWL.

  13. 13 13 Martin

    I’ve always disagreed with the parking lot example and wondered if I was missing something. If you decide to keep the forest, but a year later reverse your decision, then you can simply pave the forest at that point. On the other hand, if you pave the forest, and a year later you decide it was a bad idea, it might take hundreds of years for the forest to regrow, and still you could have wiped out species or whole ecosystems that will never return. One of the alternatives is obviously much more revocable than the other. Steve, what are your thoughts?

  14. 14 14 Dmitry Kolyakov

    Professor Henderson,

    May I suggest a small clarification – If we say that taxes do not matter only if the state uses _every taxpayer’s_ money as per their respective exact wishes, we can avoid the confusion that some commenters sense here. Otherwise, indeed if only my tax is spent exactly as I would spend it, but I can game the system to use others’ money to pay for my needs, taxes still do create a DWL, as Jeffrey and Neil noted above.

  15. 15 15 Mike H

    In my town, the local hospital is building a parking lot, having cleared some forested area to do so.

    “In other words, eliminating a forest is a risk because it might turn out to be very valuable, but holding off on building a parking lot is not as big a risk because we can always build it when we realize our mistake”

    Nonetheless, if we hold off on building it, we lose the opportunity to park there in period A. In Period B, it’s too late to restore that.

    In period A, we have uncertainty. We don’t know the value of the forest in period B, and we don’t know how much period B people will value the fact that we could park in period A.

    Reporter : For our turn-of-the-century program, we’re interviewing Jeeve Stobs, whose invention of a practical space elevator saved billions of lives from last year’s asteroid impact. Tell us, Jeeve, how the idea came to you.

    Jeeve : Well, when I was 21, I had an attack of acute appendicitis, and my girlfriend drove me to hospital. There was a new car park, just built on what used to be a forest. I think the trees were put in a museum or something. Anyway, we caught the elevator to the emergency ward. It was while I was slipping in and out of consciousness, in the elevator, that the idea came to me of just the right pattern of twisted carbon nanofibres.

    Reporter : so, without the car park and elevator, you might never have thought of it?

    Jeeve : hey, without the car park, I probably wouldn’t have survived, and my humanity-saving invention would have gone to an early grave with me.

    Irrevocability cuts both ways. Saying “the risk is less” doesn’t change that fact.

  16. 16 16 Harold

    Irrevocable means cannot be undone or changed in the future. Things in the past cannot be undone or changed, but this is different from the normal linguistic use of “irrevocable”.

    The use of the land as a forest is revocable – we can change it at any time. The use of a car park is not as we cannot change it back to a forest at any time. Semantically, irrevocable is the wrong word to use for the loss of the car park today.

    However, it is possible that the loss of the car park today is of greater value than the maintenance of the opportunity to have a forest in perpetuity.

    For example, we could clean up a disused landfill site and put a car park on it. This is irrevocable, because we cannot ever get back the landfill. However, we anticipate very little future value in the landfill, so the balance is in favor of the current value of the parking. Just because something is irrevocable does not mean we should not do it, but it is incorrect to say that the loss of use of the car park today is irrrevocable.

    It is about the knotty problem of how much we value future people, as well as the value of things like forests. We might disagree about this, but without some regulation, I think we would end up with a very low de-facto valuation. This is because each choice is irrevocable, so if you happen to have a seller and a developer who both place little value on future people (or on forests), they will do a deal. This is like having a ratchet on a see-saw – it can only go one way. No deal ever makes the see-saw tip back the other way, because these choices are not irrevocable.

    This allows us to see that the irrevocable nature of the choice is important.

  17. 17 17 Tristan

    I’m with you Mike H until you said “In period A, we have uncertainty. We don’t know the value of the forest in period B, and we don’t know how much period B people will value the fact that we could park in period A.”

    Using this model, then yes they are both risky, and one or the other is more risky based on the ratio of mean utility to be gained in period 1 to period 2.

    But the way I set up my model was that in period 1 we know for sure the value of the forest and of the parking lot in period 1. What we don’t know in period 1 is the value of those two things after period 2 arrives.

  18. 18 18 Ken

    The use of a car park is not as we cannot change it back to a forest at any time.

    This statement is patently false.

    Can someone please explain to me how a parking lot can forever after NEVER again be a forest? Am I missing something? Did the jackhammer technology somehow disappear and the ability to plant and grow trees and other plants fall out from the body of human knowledge? Humans have the knowledge and, more importantly, the capability to transform pretty much any landscape into whatever we want from whatever is there now, including transforming a parking lot into a forest.

    The point of what I think Landsburg is getting at is that a parking lot and forest cannot exist at the same time in the same place. Thus the loss of a forest today is forever (irrevocably) lost wherever there is a parking lot and similarly, the loss of a parking lot today is forever (irrevocably) lost wherever there is a forest. That there is a loss in both cases is undeniable, just as there is always a gain is similarly undeniable. The main questions is who losses and gains and by how much.

  19. 19 19 iceman

    Irrevocable: “Don’t it always seem to go, that you don’t know what you got til it’s gone.”

    I think it’s fair to say it takes longer to convert a parking lot into a forest than vice-versa. But in general to me the takeaway from this and many other Landsburgian points / posts is that a preconceived notion isn’t necessarily wrong, but if it’s right it’s for more subtle reasons than we tended reflexively to think, and there is educational value in that beyond mere “semantics”.

    However on this specific point:

    Harold — “if you happen to have a seller and a developer who both place little value on future people (or on forests), they will do a deal”

    I think one Ivory Tower response involves “revealed preferences” – the developer might not value future whatever, but if other people really do (not just say they do), doesn’t this suggest someone, group, coalition etc. can bid more to preserve a plot in its alleged ‘highest and best use’? Of course in the real world there are coordination and free rider problems. On the other hand people often just want something for free. I’ve seen petitions to use the govt to block development, but they never ask for donations to buy the land. However my neighborhood did recently raise private money to re-build tennis courts on a *public* park, and I was kinda proud of that.

  20. 20 20 Al V.

    I don’t follow the argument that, “… a helmet law could actually raise insurance rates for those who would wear them anyway.” South Carolina doesn’t have a helmet law, but some people wear helmets anyway. However, insurance companies don’t have any way of knowing who wears a helmet and who does not.

    The average annual insurance cost for motorcycle insurance is about twice as high in South Carolina as in North Carolina, and it seems reasonable to assume that most, or a least some, of the difference is due to the difference in helmet laws.

    Having lived in South Carolina for many years, my observation is that the wearing of helmets is inversely correlated to risk-taking behavior. The people not wearing helmets are largely 50+ men riding Harleys at slow speeds, while the people wearing helmets are young men riding Yamahas and Kawasakis at high speeds, weaving in and out of traffic.

  21. 21 21 Harold

    Iceman: in principle there is nothing to stop concerned citizens stepping in and outbidding the developer to preserve the forest. However, any time this fails due to lack of information etc, the ratchet clicks round one more gear. Since “mistakes” are inevitable, and the process can only go one way, we will surely end up with undervalued forests, and ultimately fewer forests than we would like. I am not entirely sure about this, but it seems a reasonable argument to me at the moment. The argument seems similar to the Brownian ratchet that Feynman popularised – itself an example of Maxwell’s demon.

  22. 22 22 Mike H

    @Ken

    Can someone please explain to me how a parking lot can forever after NEVER again be a forest?

    Because of this cat.

    In any case, it is theoretically possible that a particular parking lot might forever after never again be a forest, just as it’s theoretically possible that a particular forest might forever after never again be a parking lot.

  23. 23 23 Tristan

    Mike H, don’t you think that given the long time it takes to re grow a forest (I’m not talking about the trees, I’m talking about the eco-system centuries if not longer in the making) one of those possibilities is more realistic than the other?

  24. 24 24 Paul T

    Mark Draughn: “Do you think he’s right about your drug legalization mistake?”

    Yes, but his reasoning is flawed. Mr. Henderson claims the suppliers would no longer be compensated for the risk of their business, which has something to do with gains and losses.

    That misses the point. It’s a simple matter of supply and demand – lower costs raises the supply, and more stoners get their fix; that’s what ‘lower price’ means. Econ 101: the supply curve shifts to the left.

    I don’t see that the Indifference Principle pertains.

  25. 25 25 Martin-2

    Tristan (23) – see Ken (18).

    Paul T – So do you think drug dealers will be better or worse off?

  26. 26 26 Ken

    Tristan,

    don’t you think that given the long time it takes to re grow a forest (I’m not talking about the trees, I’m talking about the eco-system centuries if not longer in the making)

    No. And it doesn’t take centuries, it takes only a few years. Go to middle America and you will see over grown farms that were active just a decade or two ago that is now teaming with non-human life and a thriving eco system. When I lived in BFE Missouri for a year and a half as a kid, I routinely ran across machinery (like cars, tractors, engines, etc) in the middle of the forest. These were obviously part of what used to be a farm, but now stood lonely and rusted in the middle of a forest that only took a decade or two to wipe out almost all evidence of human design.

    Weisman describes just how quickly nearly all evidence of humanity would be wiped out. Very little would be left in 100 years. Nearly everything would be overgrown in that time.

    Environmentalists routinely over estimate what humans can do and underestimate the power of non-human life. This is why they are so consistently wrong about all of their predictions.

  27. 27 27 Al V.

    Martin-2, define “worse off”. Costs will go down, competition will increase, and prices will decrease. Presumably, sales will increase, but I would question whether it would increase as much as we would expect for other commodities. Thus, I would expect that profits would shrink somewhat. However, the risk of going to jail, or of dying in a gang war, would also be reduced or eliminated, which I think the drug dealers would view as a good thing.

    I think the biggest question would be how much market share could existing dealers hold onto as legitimate businesses move into distribution. For example, would Philip Morris move quickly into marijuana packaging and distribution? Probably.

    How would we expect the market to evolve? Much like other markets. Existing drug dealers would hold an initial advantage, but new competetion would move in rapidly, and we would see many new brands and distribution channels. Over time, businesses would consolidate, and we would reach a stable market with a limited set of large manufacturers. Some drug dealers would lose their businesses, but others would probably thrive in the legitimate market, and even be able to lead the new businesses.

  28. 28 28 Tristan

    I assumed that the loss of the forest was permanent only for simplicity’s sake. In reality it is technically replaceable, albeit on a long time horizon (remember, it’s not enough to replant the trees, there is an ecosystem etc. that takes dozens if not hundreds of years if not longer to recreate)

  29. 29 29 Drew

    “And here’s where I can extend Landsburg’s thinking in a way that I think he would approve of. … So the big deadweight loss from taxes is spending.”

    Hunh? I thought the big deadweight loss from taxes was… deadweight loss? I.e. that people will do wasteful things to avoid taxation, like not undertake mutually beneficial exchanges.

    Spending was going to happen anyhow, whether with income taxed or not taxed. The only difference is who gets to decide how it’s spent.

  30. 30 30 Ken

    In reality it is technically replaceable, albeit on a long time horizon (remember, it’s not enough to replant the trees, there is an ecosystem etc. that takes dozens if not hundreds of years if not longer to recreate)

    And again this doesn’t jive with observations. Ecosystems bounce back incredibly fast. After the Exxon Valdez spill, the ecosystems in the areas left untouched (just to see how bad the spill really was) came back to over 95% in five years. At most dozens and nothing close to “hundreds of years”.

  31. 31 31 Al V.

    Drew, I think both are right. Certainly we do a lot of things to avoid paying taxes that we would not choose to do, if we didn’t have to pay taxes. I buy TurboTax every year, and pay for electronic filing. What’s the benefit I receive? Lower taxes. H&R Block is a complete deadweight loss from that perspective – which I’m sure is not H&R Block’s view.

    However, some portion of what my taxes go to is also deadweight loss, isn’t it? For example, I pay into Social Security and Medicare, and my parents take out of both – even though they have a lot more money than I do. If I had a choice, I wouldn’t be subsidizing my parents’ income, considering their net worth is ten times mine.

    And I also would not choose to subsidize corn farmers in Iowa or Exxon Mobil, if I was given the choice. I might choose to continue to fund PBS and NPR, but many readers of this blog would not.

    So what’s the loss here? Well, my parents are getting more disposable income, and I’m getting less, even though I’m the one doing the work. Presumably, they are spending their excess income differently than I would choose to (they spend on trips to France, I might use the money on a new car). Similarly, Exxon Mobil is getting additional tax benefits, which should translate into consumers purchasing more gasoline than they would otherwise purchase.

  32. 32 32 Paul T

    Martin-2: “So do you think drug dealers will be better or worse off?”

    Don’t know, depends on the supply and demand flexibility, I suppose.

    Let’s see – new dealers will enter, they’re better off, it’s like a reverse deadweight loss; a deadweight gain? Existing dealers will surely see profits shrink, but will avoid risk of jail. Is that a net gain? Dunno, you’d have to ask them -

  33. 33 33 Paul T

    OK, David Henderson monitors this board. As Mr. Spock used to say:
    fascinating. Therefore, I will offer a criticism of his review.

    Mr. Henderson overlooks an error in “How statistics lie”, involving unemployment numbers.

    In that discussion, Mr. Landsburg claims that random sampling among the unemployed, to estimate statistics of length of unemployment,
    is unreliable, because those longer idle will have greater chance – more days to be called – than the shorter group. This is a fallacy of statistical reasoning.

    In the overall book’s context, it’s a very minor error. More troubling, is the possibility that it represents the consensus of the economics profession. Egregious, if so.

    Caveta: I haven’t seen the new edition of the book, I refer to the first. If this has been revised in the new one, well then, slap my
    wrist-

  34. 34 34 Mike H

    @Tristan (#23)

    I don’t think anyone disputes that it’s harder to build a forest than a parking lot. If “irrevocable” were measurable on a sliding scale, you could say that the loss of today’s parking loss to environmental concerns is “less irrevocable” than the loss of today’s forest to parking concerns.

    However, “irrevocability” is not a relative quality, but an absolute one. Or, let’s assume so for the sake of argument.

    Then, Landsburg’s argument is, I think, that when people say “don’t build the parking lot! Cutting down the forest is irrevocable!” the argument actually cuts both ways. Delaying the parking lot is also irrevocable – once you delay it, it stays delayed forever.

    You may be able to build the carpark later, but you can never undo the delay. Your decision to delay was irrevocable. The correct argument for forests over carparks is not one of irrevocability, but of expected costs of lost forests over lost opportunities to park today.

Comments are currently closed.