On Balance: Value of Improved Information about Environmental Protection Values: Toward a Benefit–Cost Analysis of Public-Good Valuation Studies
Environmental valuation has over the last 40 years grown into a major field within environmental and resource economics. Sizable resources are every year put into environmental valuation work, and an entire industry of analysts is devoted to it. There is however little discussion of benefits versus costs of these studies. A small part of them are innovative and part of fundamental research, and should clearly be funded, and published. But by far most valuation studies are much more practical and aim to assess particular goods or policies with less general interest to the broader public. Their usefulness should therefore be scrutinized.
This paper develops a methodology for analyzing the value of environmental valuation studies, and to uncover the benefits of the information added by such studies, versus their costs. It can be claimed to launch a new branch of welfare economics: the “benefit-cost analysis (BCA) of public-good valuation work”, and thereby establish principles for how public-goods valuation activities can themselves be assessed.
Public-good valuation studies are designed to inform decisions about whether to provide or not provide particular environmental or other public goods or services, or to protect or not protect and maintain particular natural objects, including forests, lakes, rivers, parks, and landscapes. Our paper studies the welfare gain by making public decision-making processes related to public goods more precise. Consider, for example, the decision to enact or not enact an environmental policy, or protect or not protect an object with environmental or natural resource of significance. The example used in our paper is the protection of a rainforest (or part of one). Given perfect information about both its protection value and its opportunity (or exploitation) value (for example by cutting down the forest and converting it into agricultural land), and a socially optimal decision process, no mistakes will be made: the rainforest will be saved when its protection value is greater than its opportunity or exploitation value; and it will be converted (cut down) when the exploitation value is greater.
In practice there is however always uncertainty in such decision processes, usually mostly about the forest’s protection value. One can then make two types of mistakes under uncertainty: 1) fail to save the forest when it ought to be saved; and 2) save it when it is welfare-enhancing to convert it (when its true use value exceeds its protection value). The valuation study or set of studies makes the protection value more precise, and reduces or eliminates these mistakes, thus increasing social welfare. The key question: is this welfare gain greater than the cost of doing the study? If that is the case, the study ought to be performed.
The paper itself is highly mathematical and I will not go into those technical details in this blog. It is more useful for readers to focus concretely on the rainforest example. Consider the Amazon, or rather a part of it being valued. Our (or “a given”) value estimate is used as a basis for saving or not saving this part of the rainforest. Our question is: do we want to carry out more valuation studies of the Amazon rainforest, to make this decision more precise?
A valuation study can be shown to have particularly high value when the resource to be valued is “highly contested”. By this we mean that the exploitation value is known to be “close” to the protection value; but it is not clear which is higher. This is intuitive: when the protection value is known to be much larger (smaller) than the exploitation value, we are quite sure that the forest ought to be protected (not protected). Information gained from a new valuation study even if the new study provides an entirely correct and quite different valuation outcome than what we initially thought, will not change our initial assessment, nor our decision. The valuation study is then of no value for this decision. When instead our initial assessment is that the two values are quite similar, it is much more likely that our relative assessment, and then also our decision, will be changed by the new study. The study can then have great impacts, and high value.
The valuation study can also have great social value when the true value of the public good is particularly large, even when the probability that the decision can be changed is relatively small.
The paper provides a numerical example based in part on data and in part on “educated guesses” about the distributions of both exploitation and protection values for the Amazon. One assumption we make is that 10% of the rainforest is likely to be “threatened”, and we assume that protection and exploitation values are similar for this part of the rainforest. This analysis finds extremely high value of a valuation study given that it significantly improves the basis for our decision to save or not save this part of the Amazon. We consider a hypothetical case where the valuation study (or set of new such studies) removes all uncertainty; this is unrealistic but analytically useful as it provides an upper bound on the value that can be achieved from new studies. Assuming then also that information is used optimally as basis for the decision to deforest or not, the value of the study (with plausible uncertainties for both protection and exploitation values) is in the range 5-8 % of the total net benefit of protection. In reality, a set of new studies may remove perhaps half of the initial uncertainty about protection values; they will then provide at least half of this total value. Given a (very conservative) protection value of the Amazon rainforest of $5000 per hectare, the value of new valuation studies that eliminate half of this decision uncertainty is a very high number, about 3-5 % of the rainforest’s total protection value. Even if the magnitude of “contested” rainforest is far smaller (say, only 1 million hectares, approximately equal to the annual deforestation level in the Amazon in recent years), this value is still $150-250 million.
Our analysis thus shows that a very large amount of valuation work may be efficient to carry out, to make the value of the Amazon rainforest, and the decision to save or not save it, more precise. This is not terribly surprising; but it is good to have such a conclusion verified in a rigorous way. Similar conclusions for other natural resources or environmental policies are however not equally obvious. Here, we have at least developed a robust procedure for investigating the value of such valuation work.