There are reasonable concerns about the practicability of my suggestions, and I aim in the last section of this chapter to assuage some of these. Certainly, it is not easy to conduct a CBA. They cost money, and the more expensive the CBA is, the farther the project’s benefits will need to outweigh its costs in order for the project to pass muster – assuming that the cost of doing the CBA is counted as a cost of the project, and that the project will be approved only in the event that its benefits outweigh its costs. Working to ensure that people are well-informed and rational will may be a very cumbersome task for the analyst—more cumbersome, in any case, than conducting a CBA without worrying at all about the informedness or rationality of those individuals whose preferences are being gauged. This does not change the fact that the less of this information the individual – even the sincere one – specifying a preference has, the more grains of salt, as it were, the specification should be taken with, the less authoritative the specification should be considered to be, and the less reason we have for counting the satisfaction of the preference as a policy level benefit. But it does raise a number of important questions.
Though I do not take myself yet to have said anything directly about what at the outset of the project was dubbed Claim 3 – that is, though nothing I have said yet has directly supported or opposed the claim that all preferences can accurately be expressed in terms of WTP – there is a point regarding money that is clearly relevant. Providing people with information and helping them to understand and process it before making an assessment of costs and benefits is going to require time and money. The question then arises: How much time and money is it worth devoting to this purpose? I do not have an answer. Clearly, there will be a point at which too much money and time are spent on this task. One proposal for figuring out how much time and money to spend on the task is to perform a cost-benefit analysis. But then we need to know what to count as a cost and what to count as a benefit. It was wanting to know precisely this that led in the first place to the suggestion that people be provided with information and helped to understand and process it. It is clear, then, that this proposal leads to a regress. The question of how much time and money to devote to informing people must not be answered by CBA. However, I am optimistic that cost-benefit analysts will be able, without too much difficulty, to help people become well informed and rational, and for this reason I think analysts will generally be able to satisfy my proposal without devoting to the task an excessive amount of time or money. I will try to explain some of the thoughts I have that ground my optimism.
To be sure, the requirement that the analyst reasonably ensures that individuals be well informed before registering their preferences is not fulfilled just by generally ensuring that the relevant information is available somehow, somewhere, at least to those willing to comb through every book, booklet, textbook, guidebook, handbook, brochure, catalog, circular, dictionary, encyclopedia, flier, folder, folio, journal, leaflet, magazine, manual,
pamphlet, program, tome, tract, and treatise. There must be a commitment to making relevant information readily accessible to people. Taking this commitment seriously may well add significantly to the analyst’s task. It obviously takes time and energy to compile this information and put it into a form in which it can be apprehended as easily as possible. It takes time and energy for the individuals administering contingent valuation surveys to share this information with respondents. For real analysts to fill this role well, a great deal of skill would be required, which means that a great deal of costly and time-consuming training would be required, and the analysts’ work out in the field may well also be quite time consuming and expensive. This need not mean that there is no place for CBA, or for CV, but may be enough to suggest that some changes are in order regarding how they are done.
For one, public debate can and should play a bigger role in the policy process, even if it going to still terminate in surveys asking people to specify their preferences. Public debate is one of the ways of helping people to gain complete as possible information. It can also help people organize and process the empirical information. It can reveal to someone that her preferences are inconsistent with one another. Public debate has the ability to lead people to think harder about their priorities, the life plans, their ultimate goals. It is healthy to subject an individual’s preferences to public debate, to challenge her to defend her preferences against competing views regarding what kind of preferences, and indeed what kind of life plans, are most worthy, and to embrace her goals, plans, and particular preferences only after careful consideration. These are all incredibly important functions of public debate. As I have said, an individual’s preferences – even those that are well informed and consistent with the individual’s other preferences – are not so sacred that it is wrong to put the individual into a context in which the topic regarding which she has preferences is
publicly debated and scrutinized. Not are preferences so fundamental or primitive that it is purposeless to do so.
Consider first an issue that (so far as I know) has no policy implications whatsoever. I start with it because despite its simplicity it is instructive. Suppose that a friend of mine prefers listening to a wide variety of other British Invasion bands over listening to The Who. I find this hard to believe! I engage him in discussion to see what he really knows about The Who. Does he know the band’s history, their influences, what the many songs they have recorded sounds like? Even if I am convinced that he does, I could still say something like this: “You, friend, should nonetheless rethink your preference. Listen more carefully to the drumming of The Who’s Keith Moon. Surely upon doing so you would acknowledge that he was more skilled drummer, and brought much more energy to his playing, than Ringo Starr (from The Beatles) or Charlie Watts (of The Rolling Stones).” I could even present him with reviews written about Moon by various people who my friend would concede are authorities on rock and roll drumming. I might continue: “You surely know that The Who pioneered the rock opera, but I think you discount how truly special a fact about them this is. The rock opera gave rock music a kind of substance that it had never before and allowed for a fuller development of a theme over the course of an album. What did the bands that you prefer over The Who do to match this? Is it not also noteworthy that The Who’s inversion of the typical structure of having the bassist lay down the rhythm and the guitarist play the melody makes them different in a neat way. That surely counts as one of their great virtues, and one that should earn them a more favorable place in your catalog of great bands.”
The same thing goes when the preference is one regarding a policy issue. Public debate can help someone gain, organize, and reason well about information pertinent to a
policy decision. Even once an individual has a basic grasp of the information, it is good for her to hear about how and why these facts matter to others, what preference other people form in response to the empirical information, and why. Subjecting a policy issue – and in a sense then, people’s preferences regarding that policy proposal – to public debate is a way of ensuring as far as possible that each person is comfortable upon reflection with the preference she expresses. This is all work that the analyst could do out in the field as she approaches individuals one by one with the survey device. But the analyst’s task is much less cumbersome when the registration of a preference is preceded by public debate. CBA suffers, and my general recommendations are much more unwieldy, to the extent that an analysis attempts to bypass public debate on the way to making policy recommendations. Certainly, CBA is not incompatible with public debate. It is possible for there to be debate – extensive debate even – on an issue, and only after this for preferences to be solicited via the revealed and/or stated preference method(s). But this is typically not how it works. CBA is typically used as a way to get a policy decision without having to go through the “messy,” time consuming process of public debate. Sometimes this is understandable. There is not enough time, space (be it in newspapers, television programming, town halls, etc.), or energy for public debate on every policy issue. Still, proponents of CBA make a mistake if they fail to acknowledge the value of debate. For this reason, the more important the policy issue, the more important it should be to have public debate, and the more skeptical we should be of policy recommendations that are made in the absence of such debate. Oftentimes, the real blame here rests with the elected officials who solicit CBAs in the first place.
Experts have an important role to play in uncovering the information that is passed on to citizens, in helping to convey this information to people, and in helping people to
understand and process this information. But their role is not one that renders the preferences of lay people, of non-experts, unimportant. Ideally, the experts are not to say what is in people’s interests, and are not to by themselves decide policy. One may well wonder why, if the information the experts have is so important, we should not just let them make the decision? Again, doing so is not consistent with respecting people’s freedom, and moreover, the experts are simply not always, despite their access to such good empirical information, in a position to say what will promote people’s well-being, since, as I explained earlier, the informed-preference account of well-being is not a convergence account. Two people reasoning equally well and with the same complete information will not necessarily respond to the information with the same preferences. The experts are (essentially by definition) possessed of superior empirical information, but the normative information – how good or bad certain outcomes would be – must be supplied by the individuals affected by the policy.
The internet also has a significant role to play. Obviously not everyone has internet access or knowledge of how to use it. Moreover, simply tucking a piece of information in some obscure “corner” of some obscure website is not going to suffice to satisfy the demands of the provision of information. But it possible to house relevant information on well- publicized websites. Consider, for example, the following information provided by the U.S. Environmental Protection Agency:
In 1984 a deadly cloud of methyl isocyanate killed thousands of people in Bhopal, India. Shortly thereafter, there was a serious chemical release at a sister plant in West Virginia. These incidents underscored demands by industrial workers and
communities in several states for information on hazardous materials. Public interest and environmental organizations around the country accelerated demands for
information on toxic chemicals being released “beyond the fence line” – outside of the facility. Against this background, the Emergency Planning and Community Right-to-Know Act (EPCRA) was enacted in 1986.
EPCRA’s primary purpose is to inform communities and citizens of chemical hazards in their areas. Sections 311 and 312 of EPCRA require businesses to report the locations and quantities of chemicals stored on-site to state and local governments in order to help communities prepare to respond to chemical spills and similar
emergencies. EPCRA Section 313 requires EPA and the States to annually collect data on releases and transfers of certain toxic chemicals from industrial facilities, and make the data available to the public in the Toxics Release Inventory (TRI). In 1990 Congress passed the Pollution Prevention Act which required that additional data on waste management and source reduction activities be reported under TRI. The goal of TRI is to empower citizens, through information, to hold companies and local governments accountable in terms of how toxic chemicals are managed.120
The TRI is available online (http://www.scorecard.org) and is indeed a fantastic resource for those seeking the kinds of information it makes available. It provides information that is relevant – though not all such information – for certain regulatory decisions. The question for policy technicians is how to ensure that all those individuals whose preferences are being gauged are made aware of this information.
This leads to what is perhaps my most important response to the concern about practicability. I can imagine a kind of approach to making public investment and regulatory decisions that blends elements of traditional CBA with elements of the legal system’s use of a jury. The idea is that a group of individuals – not just specialists, policy wonks, scientists, etc., but the kinds of ordinary people who make up a standard jury – would come together to consider a policy issue in great detail. These individuals would be selected at random. Perhaps there would then be some more nuanced selection process from among these individuals. Perhaps participating when one is called would be touted as one’s “civil duty” but not legally required. Perhaps one would – absent special circumstances – be legally required to participate like when one is currently called for jury duty. I leave these questions unanswered. In any case, it would be with these individuals’ preferences that the costs and
benefits of the proposed policy would be identified. But of course it would not simply be the preferences that these people enter the process with that would matter. Rather, it would be those they express after they have been exposed to and deliberated over a robust amount of information relevant to the policy. Consistent with my desire throughout this project to stay neutral on CBA’s theory of right action, I do not have a view on whether the members on this kind of “jury” should be required to come to a unanimous decision (like a normal jury must) or whether each individual should, after all the discussion has taken place, simply specify whether she counts the policy as a cost or benefit (and to what extent), with these costs and benefits then being added up in some particular way.
The fact that it is only a relatively small group of people whose preferences will determine the costs and benefits of a policy that will affect many people, including people not on the jury, should not cause concern. Contingent valuation studies typically approach only a random sample of people within the target population. There are methods of doing this that generate reliable results. Of course, it is possible that someone who would have had a very extreme preference – and incredibly high WTP, say, or an infinite WTA (on the supposition that these measures are employed within this jury) – and then the policy decision is affected by whether or not this person is included. Suppose this person’s incredibly strong (informed and rational) preference favors the policy being considered. She could throw off the generalization. If she is on the jury, will the total benefits of the policy be exaggerated? If she is not on the jury, will the total benefits of the policy be underestimated. This is a complicated question, but the problem that underlies the question is no more severe for the jury model of decision making that I have proposed than for the standard model of
approaching a randomly selected group of people with a survey and then generalizing from these findings.
There is, however, much to recommend the jury model. Again, for the average policy, the empirical information will be quite extensive and complex. With respect to the average policy, I cannot imagine that those administering CV surveys will be able to get each respondent to the point where she counts appropriately well informed and rational. It is in this context, but perhaps only in this context, that ordinary citizens can be sufficiently well- informed about a policy for their preferences to tell us all that we want them to tell. Suppose, however, that we become convinced that even this method – more generally, that no method – does a satisfactory job of supplying people even relatively accurate and complete policy- relevant information and helping them develop reasonably rational preferences in response to this information. Again, I am confident that this problem can be avoided; but if it does happen that people are clearly ill-informed or not reasoning well, the proper response is not to just give up on informing people and to proceed, via the revealed-preference or the stated- preference methods, recording people’s actual preferences altogether ignoring the fact that they are uninformed or irrational. These uninformed or irrational preferences should in this case be recorded and given some weight in the policy decision, but input could at this point also come directly from experts. The view is not the experts always know better. Nonetheless, in cases where people are irredeemably ill-informed or irrational, the next best alternative to creating actual informed and rational preferences is to give a select group of experts a significant role in crafting the policy. Ultimately, decision makers would then have to balance these inputs, and more generally, the various aims that I discussed in so much detail in the previous chapter.
Two points mentioned earlier bear repeating before the chapter closes. First, it is possible that a person simply does not want to be so involved in the policy process. She just wants the state to take care of her, that is, to make decisions on her behalf and with what the state takes to be her best interests in mind. I do not want to deny that there are people of this