Do Unto Others...

John Quiggin and Hal Varian are discussing the Stern Review on Global Climate Change and the pure rate of time discount in my email inbox:

Hal Varian: Stern is advocating levying a $70 tax on today's generation to transfer $940 to a future, much wealthier, generation. This is exactly the opposite of the experiment you suggestion. Given the absence of time travel, the relevant transfers are all of from the present to the future.

John Quiggin: At the margin, it's the same question. If you oppose paying $70 now to transfer $940 to wealthy people in the future, you should support taxing wealthy people $940 now to pay $70 to poor people, unless you specifically discount the utility of future generations simply because they are in the future. I'm not saying that this is an implausible preference, just that it indicates the kinds of trade-offs that are involved.

Hal Varian: Here is my problem with 0 social rate of time discount. Suppose I believe that it is ethically neutral to transfer $1 from Brad DeLong to John Quiggin. Furthermore, John and Brad are indifferent between $1 now and $1.10 next year. Should it not be the case that it is ethically neutral to take $1 from Brad DeLong now to give John Quiggin $1.10 in the future? In general, if I discount my own future consumption and I should "treat others as myself" doesn't that mean that I should discount other people's future consumption (just as I do my own)?

My view--which I admit may well be wrong--of this knotty problem is that we are impatient in the sense of valuing the present and near-future much more than we value the distant future, but that we shouldn't do so. Our habits of mind of preferring a bird in the hand to two in the bush come from a time when life was nastier, more brutish, and definitely shorter, and that if you went into the bush to try to get the two birds you might well not come out.

The fact that we want to apply a pure rate of time discount much greater than the risk of extinction to problems of planning for the far future is, I think, a flaw in our reasoning.

So I come down with John rather than with Hal on this one.

Comments

[John Emerson]

This question was discussed by Cobb and Daley in "For the Common Good". I don't have the book here, now, but as I remember they concluded that discounting the future is only good for planning our own portfolios, and is just the wrong way to deal with the questions about our obligation to future generations. At any reasonable discount rate, the value of the future quickly approaches zero.

It's also wrong to assume that people of the future will be richer than we are. Maybe they will be, maybe they won't. Disaster is possible. So is stagnation.

When Varian grudgingly assumes for the sake of argument that a transfer from DeLong to Quiggin today is justified, that doesn't mean that in most cases he will continue to be generally opposed to what he calls transfer payments, and will continue to define any social policy as a transfer payment if it interferes in the market at all and benefits any individual while harming any other individual.

For me this kind of analysis is destructive in other areas than just discounting the future, since it makes the default position a mythically perfect free market world, and requires all other positions to be justified in terms of that.

Cobb and Daley also cited an anonymous economist wisely reflecting, "You know, I can't see that we have **any** obligation to future generations". And really, how could we? What have future generations ever done for use? They're like unicorns, they don't even exist.

And of course, there's another problem -- suppose we make big sacrifices for future generations right now, but the human race goes extinct anyway? We will have reduced our present income and wealth for absolutely nothing. Prove that that won't happen! I'm outraged at this possibility!

As I keep saying, sometimes it's important to think like an economist, and sometimes it's at least equally important **not**
to think like an economist. But I'm a crank, and economists rule the world.

[Robert Waldmann]

I go with John too. I think that we can be unfair to our future selves. In fact, I think those of us who are addicted to nicotine have definitely been unfair to our present selves.

I think our pure rate of time preference is morally wrong. Now I am a liberal and don't believe that we should generally force people to act in the interests of their future selves. However, my liberalism is not based on the belief that people are the best possible judges of their own interests . People clearly aren't.

It is a morally wrong interference in the sovereignty of the self to force people to give equal weight to their future selves. This is because they have rights to harm their future selves (not unlimited I am not in favor of legalizing crack).

This argument does not apply to future generations. We should treat others as we should treat ourselves. We should force each other to treat future generations as they should treat themselves. We should allow them to mistreat their future selves a little because impatience is better than paternalism.

[dcbob]

I think that no matter what ethical position you take, consistency requires that you stay with the opportunity-cost approach, which implies that you should compare present and future values in a way that is consistent with the ability to convert present income into future income (in the form either of environmental benefits or of compensating income). I think that implies that you want to incorporate expected returns into your estimate of opportunity cost. In doing so, however, you want to adjust for uncertainty both in environmental damages and in long-term returns. Doing that correctly implies that low-return, high-damage scenarios get increasingly greater weight, the further out your projection, and greatly increases your estimate of the present value of future damages. You don’t get a zero discount rate, but you get a pretty low one, and one that is consistent with the opportunity cost approach.

[John Emerson]

I'm not willing to allow a foolish consistency to require me to do any such thing.

It seems to me that you're trying to say reasonable things about the issue at hand while not stepping outside economic theory, You can do this by jacking up estimated future costs until you get a result that seems OK, but why bother?

In particular, not all kinds environmental degradation have the catastrophic possibilities that make that kind of discounting trick possible. In this case, even a Biblical Christian analysis in terms of stewardship would seem far preferable to the economic analysis. Something like "leave the world in as good a condition as you found it, or better."

[jml]

I don't understand how unborn generations are waiting impatiently for future consumption to occur sooner rather than later in the same way those currently alive are. So I don't understand Varians or Quiggins analogy. Isn't there also an implicit assumption that the tax rate and future wealth are unrelated, but if any subject required a general equilibrium approach to wealth formation, global warmng would seem to be it. And the discount rate and risk of extinction are also treated as if unrelated, and all three of these variables can be set independently of the others?

If global warming increases the risk of extinction, and unborn generations have no disutility of waiting at all until they are born, since they are unborn yet, what is the relevance of living peoples' discount rate for living people's delayed consumption to them, if their risk of extinction goes up, or tax policy now to reduce global warming might have a significant effect on their wealth? Exactly what is the cost of waiting for our delayed consumption for them.

So, other than mathematical problems of a zero discount rate, I dont see the problem.

but I am no expert in this field, so if I am confused or wrong or missing something basic on this, please correct or clarify.

[Bruce Wilder]

The whole discussion seems bizarre to me.

We are assuming that future generations are going to be vastly richer than we are, while discussing the necessity of undertaking a policy to prevent the impoverishment, or even extinction, of those future generations.

Where global warming is concerned, there's a discontinuity out there, a point of no return. We can have a merry ol' time, driving our SUV's and increasing our standard of living (and admittedly increasing the standard of living of a lot of people, who are today living materially miserable lives), and if we do that, global temperatures may rise by 5 degrees celsius this century, which will result, in all probability, in the extinction of at least half of all species and, (over the course of 300 years or so), a rise in sea level of 80 feet, displacing at least 50 million in the U.S., 250 million in China, 150 million in India. (See James Hansen, http://www.nybooks.com/articles/19131 )

And, you guys are worrying about the correct discount rate to use, in a policy, which you "assume" will result in the transfer of value from current generations to future generations, "assumed" to be much richer. Are you kidding me?

[Matt]

Mathematics needs a past, to be continuous. Psychologically we need to measure the past to appreciate the future.

The net present liability for global warming is determined by the net present cost we are paying for yesterday's global warming.

Hence, we need some liability judgements for the current damage, which is measurable. Once we start collecting payments for current damage, future liability is better determined.

Brad is extending the previous post he had on sandwiches for sale. The problem Max posed suffers the same malformation, things to not start today, rather today reflects past actions.

[Bruce Wilder]

Just for the record: a zero discount rate is absurd.

There used to be a Microsoft guru, who would get a big laugh from a line he used in talks: he would review the path of computing advances, and then suggest that with some particular, computationally-intensive problem, the decision was whether it would be better to start computing now, in the expectation of finishing the computation in 25 years, or to start computation 25 years from now, in the expectation of finishing in two weeks or so.

There is an urgency in doing some things now.

[CapitalistImperialistPig]

Wow! I sometimes wonder if economists are operating in a world where Darwin was never born. "You know, I can't see that we have **any** obligation to future generations". And really, how could we? What have future generations ever done for use?

If future generations are really likely to be much richer than us, then it is indeed pretty silly to lavish much of our wealth on them. If our actions now produce a substantial risk that they will be catastrophically poorer than us, the question is different.

What will it have profited our progeny if one hundred years from now the GDP is a gigabuck per capita, all of which must be consumed for bare survival in the hell we have created.

[Will]

"we are impatient in the sense of valuing the present and near-future much more than we value the distant future, but that we shouldn't do so."

There has to be some time preference that leans towards preferring good things to happen sooner rather than later. Surely its a good thing that the Internet was invented 30 years ago instead of last year; that the Bill of Rights is over 200 years old instead of 20; that we emerged from the African savanna 100 milinea ago instead of 1000 years ago. Surely its better that direct mind to computer interfaces will be developed in 10 years rather than 100.

(On the other hand, if you're arguing against hyperbolic discounting, then I agree. But Professor Varian isn't suggesting hyperbolic discounting...)

[dale]

There is a problem in dividing the flow and continuity of human society into the conpetual box of generations.

[jml]

Bruce Wilder thinks we need a real options analysis. Anyone done that?

[kharris]

"Yes" to John E. This is one of those situations, I think, in which we comfortably revert to a powerful, familiar tool to think about the future, without giving much thought to the appropriateness of the tool. Equilibrating this consumption stream to that consumption stream is all we know of ethics? Discounting is the way to think about obligations? This is sad.

Brad's weblog has devoted a good bit of electronic ink to the notion of happiness as it relates to relative wealth, to income and the like. Have we not, in all of this, seen far enough into human nature to realize that we are consuming in an unprecedented manner merely because we are programmed to do so? Consume more in case of famine. Acquire more in case it renders additional reproductive opportunity. Most of us fight off reproduction nearly every time we have sex (if memory serves, anyhow), and worry that we are getting fat anytime we really enjoy a meal. Do we really want to do a discounting exercise to weight "more, more, more" against the risk of fire and ice for our descendants?

[walkingtheline]

This is only an interesting question if perfect markets are a gross misrepresentation of reality.

Which is to say that it is an interesting question... why all this fuss about markets?

[Taggert J. Brooks]

I don't get it. Maybe I'm missing something, but how can you not have a positive and non-zero rate of discount? I think Varian's analogy is spot on. I would only trade a dollar of consumption today for 1.15 next year. Am I worried about getting killed while off in the bush trying to find food? Nay, but I am worried about getting hit by a bus when I walk across the street. Brad, maybe you havea point if we all knew exactly how long we were goinf to live and where and when we would die. But alas, that is still a mystery, and thus it is at least one reason for the rate of time preference.

[Bruce Webb]

Of course some of those future generations are right here right now. If we take the conventional period of 30 years we can say there are portions of four different generations,sharing this planet. Now I may not owe anything to generations more than four out, I do owe something to the generation before me and I want to deliver something to my nephew and niece, who in turn have small children.

Dale's point is acute. If humans simply died off in masse at regular intervals to be replaced by a new generation of hatchlings then some of this theorizing would make sense. But my great-nephew Spencer is on the planet as we speak and so vicariously I am sharing his horizon. And if I am lucky enough to reach 80 years old I will get a chance to share a new horizon with his children.

This is so obvious as to be banal. Your time horizon is simply a combination of all the horizons of people you do or will care about.

Now this horizon extension should not go on forever, one of the stupidist innovations ever was to introduce Infinite Future Horizons into the Social Security discussion. Trying to project hard numbers a hundred or a thousand years in the future is more voodoo than rational economics. On the other hand the traditional 75 year window is a reasonable compromise, it serves to capture the entire work history and retirement for everyone currently paying into the system.

Global warming is enough of a clear danger to Spencer to justify taking action now. If you are childless and friendless your mileage may vary. But then you have more problems than calculating discounts.

[Bernard Yomtov]

I share the bewilderment at this approach expressed by several commenters. Consider the statement,

"Stern is advocating levying a $70 tax on today's generation to transfer $940 to a future, much wealthier, generation. "

What is this ridiculously precise $940? Can we stop pretending we are measuring the orbits of planets and admit that this is more a wild guess at a not-very-good measure of future welfare than the result of careful and well-grounded calculations?

[doug]

I actually think a smart discounter would choose a Beta greater than one -- not less. This is another behavioral anomaly, of course, that our utility is determined by changes in the present value of the future consumption path and in increases in realized consumption -- not the overall level. The problem with a Beta greater than one, however, is that our workhorse economic models wouldn't yield analytical solutions...

[Jeffrey Miller]

I side with John Q.

I am skeptical of the claim that people two or three or one hundred generations hence will, without a good deal of concerted effort on the part of people who are now living, be wildly richer than Americans are in 2006. Given the likely rates of extinction of species in the coming years, and the ever increasing demands on the biosphere, it's not implausible that'll we'll see a collapse, a la Jared Diamond. It's that eventuality that we should be attempting to forestall and if we don't, future generations are likely to be very poor indeed.

[wkwillis]

Look at it both ways.
We save more than we need for our future needs because we are uncertain of our future needs and need to be sure of meeting them.
The surplus (or deficit) belongs to our children because they are the ones around after we die.

[wkwillis]

For nuclear winter, it's whether there will be enough food for the physically weak elderly people (like us) to get enough to survive after the harvests are ruined by July frosts.

[wkwillis]

For global warming, it's whether the half of America that lives in a county on the coast can maintain* the* value* of* their* real* estate* in the presence of category five hurricanes coming ashore in Nova Scotia.

[wkwillis]

Sorry about the four posts and asterisks. Comment spam filter.

[gordon]

The debate continues here
http://johnquiggin.com/index.php/archives/2006/12/19/reviewing-the-stern-review-again/
with a link to a draft paper on Stern by Prof Q.

[gordon]

OK, I don't know how to get a hyperlink up on this blog, but there is a new post by Prof. Quiggin linking to a draft paper of his, on his blog dated 19 Dec., for those who already have the address.

[Richard Tol]

Going back to Brad's initial point: Yes, we should evolve and become less impatient. But, this does not only for decision about climate change. It holds for any long-term issue.

Furthermore, although there are good reasons why the social discount rate, and therefore the government discount rate should be lower than the average individual discount rate -- there are also good reasons why government policy should not stray too far from the will of the people.

[garhaneg]

No need to get all twisted around with this stuff about future generations. The thing is happening right now. Here on the west coast of Canada we have had a december to remember, and I do not refer to romance. We need to do what Stern calls for for our own sakes, starting yesterday.