Additivity and Payments for Ecosystem Services, Take Two
Awhile back I wrote a couple vague posts about PES and additivity (also, here). I am going to try to clarify my thinking a little. All of this comes in response to a great workshop on payments for ecosystem services, sometimes called payments for environmental services, that was held in Chapel Hill a couple weeks ago. The workshop was hosted by the Property and Environment Research Center, a Libertarian think tank based in Bozeman, MT. PERC is headed by Terry Anderson who has long espoused "free market environmentalism." His broad thesis is that, with well-defined property rights, private markets can solve environmental problems more efficiently and effectively than government regulation.(**)
While PERC has what some might consider fringe Libertarian underpinnings, for this workshop they corralled a broad group of interesting participants with varied (or non-) ideological leanings. It was also relatively small, with lots of time set aside for discussion. As James Salzman said at the meeting, PES is an interesting concept that can bring together people with very different ideological leanings.
My role was a small one: I just served as a discussant for one of the papers. I used my five minutes to present my basic thinking on the so-called problem of making payments for ecosystem services "additive," which really underlies the paper I discussed, and several others. My simple diagrammatic analysis is below, followed by a brief description.
In my view, attempts to make PES additional--that is, pay only for ecosystem services that would not have been provided otherwise--are really attempts at price discrimination. Now, the term "price discrimination" has derisive overtones that it really shouldn't. We see price discrimination everywhere and there are some really positive aspects to it. Consider how much plane ticket prices vary across passengers on a typical flight. Price discrimination often allows people to enjoy a good or service that otherwise wouldn't, especially in monopolistically competitive markets. But price discrimination is the term we're stuck with, because that's what's in all the textbooks.
The challenge with price discrimination, in all contexts, is that it's basically impossible to do perfectly. In the context of PES, perfect price discrimination would have all payments for provision of ecosystem services just equal landowners' opportunity cost for participating, not a penny more.
Now, since you cannot discriminate perfectly, there are going to be inefficiencies. The trick is to figure out how to minimize those inefficiencies, which takes things to the beautifully intricate world of contract theory and mechanism design. Paul Ferraro has a couple nice papers summarizing this theory in the context of PES.
To make all of this more concrete, let's take a real world example, the Conservation Reserve Program, which is arguably the world's largest PES program. CRP pays farmers annual rental payments to establish conservation practices on land formerly planted with crops. Last time I checked, annual payments totaled around $1.8 billion to retire some 34 million acres (about the size of North Carolina).
How does CRP price discriminate? Several ways. First, farmers cannot enroll land in CRP unless it has been cropped for three of the previous five years (or has already been enrolled in CRP). In effect, owners of uncropped land have zero opportunity cost for providing much of the environmental services associated with that land, so CRP excludes uncropped land since it costs those landowners nothing to provide environmental services, even though there is positive marginal social value to the services provided.
Second, while CRP uses a competitive auction-like environment for enrolling land, landowners opportunity costs vary widely across participants, so the government price discriminates by putting a county and soil-specific maximums on the on rental rates that farmers can request. Getting these maximums right can be tricky, as I've noted previously. But the essential idea here is price discrimination.
Third, the competitive nature of the CRP bidding mechanism encourages landowners to make offers that are somewhat closer to their opportunity costs. However, these incentives are limited, since landowners will still tend to make offers that tend toward the expected equilibrium price, not their opportunity costs for participating. I'm in the midst of putting the final touches on a long overdue paper that considers new procurement auction mechanisms that might price discriminate more effectively than standard auctions.
The broader point is that addititivity, or lackthereof, really has no direct bearing on the efficiency of PES programs, but rather the distribution of rents that derive from them. The more price discriminating the PES program, the more "buyers" will gain and the less sellers (typically landowners) gain. What's really interesting, however, is that from the buyers' point of view, price discrimination is really necessary. Otherwise, they may find themselves paying largely for environmental services that would have been provided anyway. And while social surplus may grow, the implicit transfer from buyers to sellers is larger than the total increase in surplus.
While I think this illustration of the problem does a lot to clarify some difficult tensions, it certainly doesn't provide any clear answers. It's always hard to figure out how to split a pie. And, in this case, there are almost certainly going to be tradeoffs between the size of the pie and how it is split.(##) But this is not going to be the usual inequity verses efficiency tradeoff one usually envisions. For example, some poorer countries with lots of rain forest could gain the most from a relatively more competitive (and less additive) market for carbon offsets.
Clearly, the largest problem involved with developing a true market for ecosystem services is corralling the broad, diverse and typically non-excludable interests that make up the "buyers" or consumers of ecosystem services. The large transactions costs associated with corralling these interests is the obvious challenge to Coasian solutions that do not involve government. In my view, this standard challenge is exacerbated further by the difficult tensions between efficiency and rent distribution. Because, if buyers could corral their broad and diverse interests and establish new trade for ecosystem services, they may see that such efforts would ultimately be self-defeating.
(**) This idea follows a classic paper by Coase. Where PERC may be more fringe than mainstream economists concerns where "property rights" ultimately come from. Many economists (myself included) think that government often needs to play a larger role in defining and protecting property rights. Terry Andersen and PERC seem to have a relatively narrow view of property rights, one in which the law and government might be involved with enforcing private contracts, but very little beyond that. But where individual economists draw the box around the government role in defining property rights seems to me like mushy territory.
(##) Economists like to fantasize about lump sum transfers as a way to have the largest possible pie and then split it, but in reality that almost never happens.
While PERC has what some might consider fringe Libertarian underpinnings, for this workshop they corralled a broad group of interesting participants with varied (or non-) ideological leanings. It was also relatively small, with lots of time set aside for discussion. As James Salzman said at the meeting, PES is an interesting concept that can bring together people with very different ideological leanings.
My role was a small one: I just served as a discussant for one of the papers. I used my five minutes to present my basic thinking on the so-called problem of making payments for ecosystem services "additive," which really underlies the paper I discussed, and several others. My simple diagrammatic analysis is below, followed by a brief description.
SLIDES HERE:
In my view, attempts to make PES additional--that is, pay only for ecosystem services that would not have been provided otherwise--are really attempts at price discrimination. Now, the term "price discrimination" has derisive overtones that it really shouldn't. We see price discrimination everywhere and there are some really positive aspects to it. Consider how much plane ticket prices vary across passengers on a typical flight. Price discrimination often allows people to enjoy a good or service that otherwise wouldn't, especially in monopolistically competitive markets. But price discrimination is the term we're stuck with, because that's what's in all the textbooks.
The challenge with price discrimination, in all contexts, is that it's basically impossible to do perfectly. In the context of PES, perfect price discrimination would have all payments for provision of ecosystem services just equal landowners' opportunity cost for participating, not a penny more.
Now, since you cannot discriminate perfectly, there are going to be inefficiencies. The trick is to figure out how to minimize those inefficiencies, which takes things to the beautifully intricate world of contract theory and mechanism design. Paul Ferraro has a couple nice papers summarizing this theory in the context of PES.
To make all of this more concrete, let's take a real world example, the Conservation Reserve Program, which is arguably the world's largest PES program. CRP pays farmers annual rental payments to establish conservation practices on land formerly planted with crops. Last time I checked, annual payments totaled around $1.8 billion to retire some 34 million acres (about the size of North Carolina).
How does CRP price discriminate? Several ways. First, farmers cannot enroll land in CRP unless it has been cropped for three of the previous five years (or has already been enrolled in CRP). In effect, owners of uncropped land have zero opportunity cost for providing much of the environmental services associated with that land, so CRP excludes uncropped land since it costs those landowners nothing to provide environmental services, even though there is positive marginal social value to the services provided.
Second, while CRP uses a competitive auction-like environment for enrolling land, landowners opportunity costs vary widely across participants, so the government price discriminates by putting a county and soil-specific maximums on the on rental rates that farmers can request. Getting these maximums right can be tricky, as I've noted previously. But the essential idea here is price discrimination.
Third, the competitive nature of the CRP bidding mechanism encourages landowners to make offers that are somewhat closer to their opportunity costs. However, these incentives are limited, since landowners will still tend to make offers that tend toward the expected equilibrium price, not their opportunity costs for participating. I'm in the midst of putting the final touches on a long overdue paper that considers new procurement auction mechanisms that might price discriminate more effectively than standard auctions.
The broader point is that addititivity, or lackthereof, really has no direct bearing on the efficiency of PES programs, but rather the distribution of rents that derive from them. The more price discriminating the PES program, the more "buyers" will gain and the less sellers (typically landowners) gain. What's really interesting, however, is that from the buyers' point of view, price discrimination is really necessary. Otherwise, they may find themselves paying largely for environmental services that would have been provided anyway. And while social surplus may grow, the implicit transfer from buyers to sellers is larger than the total increase in surplus.
While I think this illustration of the problem does a lot to clarify some difficult tensions, it certainly doesn't provide any clear answers. It's always hard to figure out how to split a pie. And, in this case, there are almost certainly going to be tradeoffs between the size of the pie and how it is split.(##) But this is not going to be the usual inequity verses efficiency tradeoff one usually envisions. For example, some poorer countries with lots of rain forest could gain the most from a relatively more competitive (and less additive) market for carbon offsets.
Clearly, the largest problem involved with developing a true market for ecosystem services is corralling the broad, diverse and typically non-excludable interests that make up the "buyers" or consumers of ecosystem services. The large transactions costs associated with corralling these interests is the obvious challenge to Coasian solutions that do not involve government. In my view, this standard challenge is exacerbated further by the difficult tensions between efficiency and rent distribution. Because, if buyers could corral their broad and diverse interests and establish new trade for ecosystem services, they may see that such efforts would ultimately be self-defeating.
(**) This idea follows a classic paper by Coase. Where PERC may be more fringe than mainstream economists concerns where "property rights" ultimately come from. Many economists (myself included) think that government often needs to play a larger role in defining and protecting property rights. Terry Andersen and PERC seem to have a relatively narrow view of property rights, one in which the law and government might be involved with enforcing private contracts, but very little beyond that. But where individual economists draw the box around the government role in defining property rights seems to me like mushy territory.
(##) Economists like to fantasize about lump sum transfers as a way to have the largest possible pie and then split it, but in reality that almost never happens.
Hi Michael,
ReplyDeleteYour figures are very useful. Thanks for those. However, I think that you way that you've drawn them is misleading. Both the point where environmental services begin to have a positive marginal cost, Q0, and the elasticity of the marginal benefit curve make it appear as though there may be significant loss of consumer surplus with competitive pricing. Although it's an empirical question, I'm not convinced that the supply of zero-marginal cost environmental services is that large. The "lost" consumer surplus may actually be much smaller than the gain from competitive pricing. More to the point, is there some reason why consumer surplus is valued more than landowners' surplus? Otherwise the competitive pricing outcome, even without price discrimination, is the one that maximizes total surplus.
Regardless of distributional quibbles, I think that your main point is correct. Designing PES policy is inherently second-best - i.e., it is not possible to perfectly price discriminate. A few months ago, I was thinking along the same lines in a more specialized setting (see here: http://www.canadianagrifood.ca/?p=76).
One final note. Coase's property right arguments may have a role to play in environmental services. In Canada, environmental groups and forestry companies recently agreed to a very Coasian arrangement (see here for links: http://www.canadianagrifood.ca/?p=40 ).
Brandon