« Should libertarians side with the Democrats? | Main | Is procrastination rational? »
Discount rates, again
Arnold Kling discusses Dasgupta, Stern, and de Long. Here is his bottom line:
My concern is with Stern, Dasgupta, or DeLong playing social engineer and picking a social discount rate that deviates from market interest rates. I think you get unreliable conclusions any time you do that.
I'll recap my views as follows:
1. For resources which will be reinvested, use the rate of return on capital, adjusting for taxes, risk, and the like.
2. For resources which will otherwise be consumed within the current generation, use the market rate of interest, again with adjustments. That rate reflects time preference within a life.
3. If we are comparing different consumption units across the generations, there is no time preference in the economically meaningful sense. (Prior to 1962, I was not impatiently waiting to have been born.) Use zero, noting this is an ethical rather than economic choice.
Depending on the configuration of the variables, the correct "net" discount rate usually will be above zero but below the current market rate.
Note that the future consumption/investment ratio is not a current fact about the world, but rather a matter of choice. That means the correct discount rate is also not a "current fact about the world," but rather will depend on other choices we make. This is a common confusion.
Arnold also writes:
Even if the Stern report had nothing to do with global warming, its assumption for the social discount rate has radical policy implications. Implicitly, it argues for an all-out effort to reduce private-sector and public-sector consumption and to increase investment instead.
I am less worried. Not all or even most current consumption reductions will much benefit the next generation. (Does anyone know what percentage of a marginal increment of savings ends up in bequests?) The prospect of current consumption is also a (the?) driving force behind innovation and technological improvement, which most definitely benefits future generations. We should invest more in the future, but the intergenerational zero rate view, correctly interpreted, does not require that we should limit consumption as much as possible.
Addendum: Here is commentary from The Economist.
Posted by Tyler Cowen on December 5, 2006 at 06:57 AM in Economics | Permalink
Comments
"If we are comparing different consumption units across the generations, there is no time preference in the economically meaningful sense. (Prior to 1962, I was not impatiently waiting to have been born.) Use zero, noting this is an ethical rather than economic choice."
I think most people would agree that we should weight the preferences of individuals across different generations equally. However if we look at it as an finitely repeated game (with unknown ending time), we still might want to adjust the (zero) discount rate. For instance, shouldn't we discount for the probability that the world would end this (and each subsequent) year?
On a more a-la-Parfit question: Tyler, do you think the discount rate used to compare preferences across generations should also be adjusted for the amount of population of each generation? In other words, if we expected population in 50 years to be 10 times the population today, do you think ethically we should apply a higher discount rate than if population was to remain the same? From an ethical perspective, what other factors should adjust the discount rate, (if any)?
Posted by: Economister at Dec 5, 2006 10:49:36 AM
Shall we turn an ethical decision into a biological one? After all, it is too hard to fight a war on pure ethical ground. How about taking the average rate of individual gene pool decay for the discount rate across generations? Let us assume on average an individual produces an offspring at the age 30, i.e., diluting his/her genes into half, the equivalent discount rate is roughly 2.3%.
Posted by: Yan Li at Dec 5, 2006 11:03:54 AM
Tyler,
I am repeating myself, but I guess I am curious
regarding what your view actually is, as you
do not address it. Do you oppose the Chichilnisky
"green golden rule" argument that says use
higher rates, one more tied to market rates for
shorter time horizons, but then move to lower
rates (but still positive) for longer time
horizons, especially when contemplating enviro
issues?
Of course the obvious reason to oppose this is
time inconsistency, exactly the problem that
shows up with procrastination, which is usually
modeled as a time inconsistency/hyperbolic
discounting issue, although in that case, the
short term discount rates are well above market
rates. If you do not go with Chichilnisky, then
one gets into these sterile debates over market
rates, that totally shaft the future generations,
zero rates that totally shaft current generations.,
or the murkiness of that magic rate that is in
between and "just right."
Posted by: Barkley Rosser at Dec 5, 2006 2:42:37 PM
I'm am very pleased to read this - I was thinking it depends what sort of good / choice is in hand (having recently been taught the Ramsey model, and that Ramsey thought the only defensible inter-generational discount rate was 0) but wasn't sure enough to weigh into the debate. Seems to me that in the context of a report on a persistent phenomenon like global warming, a sub-market discount rate is entirely defensible.
It's a great shame that Arnold Kling chose to use such intemperate language about Stern (and he still appears not to believe that such a rate could have been chosen thoughtfully, as opposed to being either sign of incompetence or mendacity) and an equal shame that Chris Bertram chose to call him a hack for doing so. Both should be about to write about people who they think are mistaken with a bit more respect.
Posted by: luis_enrique at Dec 5, 2006 5:12:58 PM
"If we expect the population in 50 years hence to by 10 times that of today".
Of course most developed countries now are seeing sinking populations, so another scenario might be that in the future the number of humans is vastly less. They could easily be richer as well, and possessed of magical technologies such that solving global warming to them is trivial.
We can have a debate on this, but really it is pointless, we have no real way of estimating anything this far in the future. I think this is the problem with this whole debate, it is being framed as an investment type problem (invest a sum of money, get a return) but since we really have no idea whatsoever about what the payoff from this investment will be in the future we can’t use investment analysis to give us an answer. The return is indeterminate; we can’t even agree an upper or lower bound on it. Investment analysis is about choosing between alternatives, so if we frame the problem as an investment decision we should automatically reject the investment case for investing in prevention of global warming. This is true even if you can crack the problem about how me (as the investor) values a return which goes to my great-grandson - first we have to agree whether there is a return first.
So it is totally an ethical problem – we have to agree that there is an ethical imperative that says we have we have to invest whenever we can to avoid a potential problem for future generations. This speaks to some deep instinct in some people, probably something to do with genetic survival, but it is not generally practical or historically wise for us to do this. Today’s concerns are rarely tomorrow’s. Take the hypothetical case of people in the 19C, who might have decided that they would reduce the use of coal to leave some for future generations (a real concern to the Victorians), that investment would have cost them dear, but would have benefited us only trivially.
Another problem of course is that any solution will involve classic IV or V spending.
http://www.epochtimesnews.net/index.php?content/view/689/47/
Posted by: ChrisA at Dec 6, 2006 5:49:24 AM
penetrazioni vaginalibook escort polacche ragazzine nmani foto passere calde uomini pelosi nudi foto porno orge pompini sborrate pornostar giapponesi diciottenni porno porno fumetti fiche vogliose asiatiche in calore erotiche porno nonne zoccole diciottenni sborrate piercing estremo foto leccare leccate cazzo innocenti giovani sexy fige
Posted by: www at Dec 6, 2006 8:02:03 AM
ChrisA wrote: "So it is totally an ethical problem – we have to agree that there is an ethical imperative that says we have we have to invest whenever we can to avoid a potential problem for future generations."
We still have to choose between different types of "investments" to avoid a potential problem for future generations. The question is how do we choose. Do we pick randomly? Or try -with all the limitations- to make an informed guesstimate of the costs and benefits of each?
In other words your statement is incomplete as (most) a moral guide. "...there is an ethical imperative that says we have we have to invest [how much] whenever we can to avoid a potential problem [of some probability] for future generations" Or simply how much should we invest?
Saying that its "totally an ethical problem" does not solve the issue about which potential problems for future generations we must address and which specific policies we must adopt.
Posted by: Economister at Dec 6, 2006 4:58:00 PM
Economister: I don't think the Parfit dilemma appears here. The problem Parfit posed is relates to actions which change the number of people who are born. If we expect there to be 10 times the people, this doesn't really come into play. I would say the more people there will be, independent of our saving decisions, we should save more now because of diminishing marginal utility - wealth/income produces more happiness the more people it is spread over, all things equal.
On the other hand, it is possible that our saving could affect the level of population. For example, investment in a cure for diseases could result in higher populations in the future than otherwise (although it could easily do the opposite as disease eradication leads to development, which eventually leads to a decline in birth rates). In that case, we would be in a Parfit dilemma.
Posted by: catquas at Dec 6, 2006 10:20:45 PM
Economister - "totally an ethical problem" does not solve the issue about which potential problems for future generations we must address and which specific policies we must adopt.
I agree 100%, there is no "rational" way to allocate resources on an ethical basis, because there is no rational way to prioritise one ethical concern above another. That is the point of ethics right? It is a personal decision, the starving baby in Africa versus the possibility of a flood in NY in 2050.
My point was that the Stern report, in particular,
presented their analysis about what to do about global warming as an investment
analysis (like the Govt deciding to build a road)which it cannot be for all
sorts of reasons - not least of which is that the investment return cannot be calculated.
Posted by: ChrisA at Dec 7, 2006 3:24:53 AM
For my thoughts, go here.
Posted by: Eric Rasmusen at Dec 7, 2006 4:47:09 PM
Weight loss is easier than you think!
Posted by: adipex at Jan 28, 2009 7:10:11 PM
EVE ISK
EVE Online ISK
Buy EVE ISK
Cheap EVE Online ISK
Posted by: aion kina at Mar 20, 2009 2:36:52 AM
