What follows merely represents some ideas that have proven to be of use in practice for many managers, policy makers, and other decision makers faced with the task of developing economic values for decision making. They are guidelines and not “rules” to follow in all cases.
Ultimately, economic value measures are only one, often small, input into decisions regarding forest use. Further, there are many different ways in which to go about deriving and using such values. As emphasized throughout this discussion, no one rule exists. Much depends on the defined context within which decisions will be made. Thus, the first group of guidelines relates to defining context and developing a set of tradeoff criteria for looking at proposed changes in forest use.
In making decisions about changes in forest use, the manager or administrator needs to consider the tradeoffs in values associated with the proposed change. This means comparing the costs and benefits with and without the proposed change, i.e., comparing a proposed change to the status quo. In the case where the manager has an array of changes to consider, the comparison also will be between the alternate changes.
The fundamental need at this stage is to define the basic type of decision being made and reach consensus on the tradeoff criteria that will be used in making the decision. The guidelines that follow derive from the earlier discussion in this paper:
Define the policy context as clearly as possible
In defining both the type of decision and the tradeoff criteria to be used, it is essential that some early consideration be given to defining the policy context in which decisions will be made (see chapter 2).
Policy may limit the amount of effort that needs to be devoted to economic valuation. The most extreme case is where national policy dictates that decisions will be made on criteria other than economic ones. In such cases, very little if any effort needs to be devoted to economic valuation of costs and benefits associated with the proposed change(s).
Define the administrative context in terms of what decision criteria are acceptable
A second need is to define the administrative context or environment in which decisions will be made, since that will influence the valuation approaches that should be used (see chapter 2). Remember that, even if economic values are not needed to make decisions, administrative agencies need financial data (strictly based on market prices and actual dollars in and out of the agency) for budgeting purposes.
Limit the number of different interested groups considered
Generally, the policy and administrative contexts in which a decision will be made define the relevant interested groups to be considered and how they will be weighted in terms of decisions. Often, the public sector takes a national accounting stance, where the costs and benefits accruing to all interested groups, valued by wtp measures, are aggregated to arrive at a net benefit figure. In such cases, members of all important interested groups (important in terms of the costs and benefits accruing to them) are in theory being considered, although distinctions among their value perspectives are not made.
The fact is that managers and administrators seldom have the time or expertise available to deal with very complex valuation and decision frameworks. Therefore, the general guideline put forth here is to first spend time understanding the different interested groups or constituencies, and then pick for the decision context those groups that would be affected in a major way by the proposed change (see chapter 3), adding other groups into the assessment as time and resources permit.
Make sure to define the decision context broadly enough to include the main externalities involved.
One major reason market prices often fail to provide true measures of economic value for environmental services or damages is the existence of negative or positive “externalities.” As discussed earlier, these are effects that occur outside the decision framework that establishes the values used. Thus, the price of fertilizer most often does not incorporate the negative off-site pollution effects; and the value of production of some crop that leads to heavy erosion on upland areas generally does not incorporate the negative downstream costs of the increased erosion. Similarly, the value of production of crops from nitrogen fixing plants often does not internalize the benefits from the nitrogen fixation.
A major reason for shadow pricing, or adjusting market prices to reflect true wtp, is to allow for externalities. By considering early the broader context of decisions, it makes it easier later to define and allow for externalities in the valuation process.
Define and get agreement on economic trade-off criteria as early as possible in making decisions.
Although valuation is quite distinct from decision making, it is important to get consensus on the tradeoff criteria at an early enough stage so they can be considered in the valuation process. To take an extreme example discussed earlier (see box 2.2), when considering the alternatives of putting a forest into extractive reserve status vs. permitting continued slash and burn agriculture, it is important to determine the extent to which the tradeoffs between outputs will be considered per unit of forest, per unit of labor input, per unit of capital invested, etc. These three criteria could produce quite different results.
Similarly, tradeoff criteria might partly be in physical terms, e.g., tons of soil loss prevented by one or another change in forest management. Sometimes, such physical tradeoff criteria make sense, since attaching economic values becomes a game in futility and the resulting information is so uncertain that decision makers would be better off just concentrating on the physical/biological data available for the different alternatives.
Finally, often the relevant tradeoff criteria are mainly political or social. In such cases, economic tradeoff criteria-may be irrelevant.
Apply the with and without principle
As was pointed out in chapter 1, values are determined by both supply and demand conditions with and without the activity or project being considered. The example of the watershed management project that prevented sediment buildup in a dam reservoir (box 1.3) indicated the errors that can be made if both supply and demand are not considered with and without the proposed change(s). Also, remember that the with and without condition is not the same as the before and after condition. The example of an erosion prevention project illustrated that point (box 1.2).
Do not attempt to value everything in economic terms
It has been pointed out often that serious errors can be made by trying to apply economic values to goods and particularly services for which there is little basis for deriving such values. In such cases, it is better to stick to measures of physical-biological changes than to apply erroneous value measures to them. Furthermore, there are certain conditions that cannot be adequately addressed in economic terms, e.g., “irreversibilities.”
Similarly, economic values often are put forth as a comprehensive measure of the value of a forest, when, in fact, those values do not include many significant benefits that might significantly increase the aggregate value of the forest. The reason they were not included is that the manager did not feel comfortable putting values on them. However, the manager might be willing to do so in another case. Obviously, the two situations would not be comparable in economic terms. Some clear agreement within a given decision-making unit (or where comparisons among alternatives will be made) needs to be reached on what will and will not be considered in economic terms.
Another point here is that generating quantitative value information usually costs money. We only want to incur that cost if we think that the benefits of having the information are worth the costs. Sometimes, we might spend a great deal of time, effort, and money generating data that would have little impact on a decision. We obviously want to avoid that.
Here we can put forth several practical guidelines based on experience. The suggestions below provide in a sense a summary of points made in earlier chapters.
Start by measuring and estimating the easiest-to-measure, important benefits
Since in an operational sense values are only needed as an aid to making decisions, it follows that we only need to generate the amount of value information needed to make the decision. For example, suppose we know the costs of some proposed change. We can start to generate values for those benefits that are easiest to measure. If they are large enough compared to costs to justify the change, based on whatever criteria employed by the decision maker, then we will not need to measure additional, more complicated benefits.
If the easiest to measure benefits do not reach a high enough level to meet the criteria for change, then we need to value additional, increasingly difficult to measure, benefits until we either have exhausted the benefits that we have identified and can value, or until we reach a point where aggregate benefits are at a level where they meet the criteria for acceptance of the proposed change. The basic point of this approach is that we can avoid wasting a lot of resources and time attempting to value all benefits, when the more easily valued ones are large enough to justify the proposed change.
When benefits are variable, the same advice does not hold for the cost side of the equation, since the same type of decision rule is not available as for benefit estimation. A reasonable attempt should be made to estimate the values of all significant costs, since they are cumulative and must ultimately be justified as a whole in an economic analysis.
Use market prices where such exist (see exceptions)
As a general rule, use market prices where they exist. Warnings that market prices may not reflect wtp at the margin and that further work is needed include cases where:
there are effective minimum prices or price ceilings in effect (and particularly if a black market exists for the good or service being valued);
the market is thin, i.e., few items are sold and demand is likely to fall off sharply with increased production and sale, indicating that existing market prices may be high if production is expanded and that existing market prices may be low if production is curtailed; and
obvious, known externalities are involved in the production and/or consumption of the good or service in question.
Use policy prices that reflect the existing and likely future policy context
A practical rule of thumb is to use policy prices (see box 1.4) unless there is an explicit, overriding reason not to do so.
We pointed out in chapter 2 that the policy context defines to some degree the boundaries that need to be considered in assigning values to outputs and inputs. If a forest is designated by national law as an area of land that will be used only for forest, then, using policy prices, we would estimate the opportunity cost of land to be equal to the next best forest-based use. If we ignore policy pricing rules, then we might consider any use of the land in deriving a measure of opportunity cost. Given the existence of the policy, the decision could be wrong.
Take into account capital value changes as well as flow values
This guideline supports the recent interest in use of natural resources accounting values in overall assessments of forest use (see chapter 4).
Present use of the forest for logs, minerals, and other products conflicts directly with the use of the forest for wilderness or aesthetic values, but may complement management of the forest for future wildlife production. The types of relationships or interactions are well-known and generally considered in discussions of forest values. More complex and often subtle interactions also exist. Thus, present use of the forest for wildlife products affects our ability to use the forest for such products in the future.
This concept relates to what we call stock, capital, or asset and flow values. If properly managed, a forest (as a stock or capital asset) can produce flows of outputs (goods and services) over time. Thus, total value of the forest today depends not only on which uses we choose, but also on the pattern of those uses over time. The essence of good forest management is understanding the different potential combinations of value flows over time that can be derived from multiple uses of forests as a stock resource. Also required is a good understanding of the concepts of time value and discounting.
Take the value of time into account
Other things being equal, if a subsistence-level forest dweller had to choose between (1) cutting down a tree for firewood to keep warm today, and (2) freezing now and cutting down the same sized tree two years from now to keep warm, the forest dweller most likely would choose to cut the tree now. This relates to the common concept of “time value” in economics. We have to keep in mind this generally applicable principle: the further into the future a given forest use will occur, the less will be the value of that use when compared to the same use in today's terms. The principle is dealt with in economics using an interest rate to “discount” future values back to the present, or compound present values into the future. Values that occur in various times then can be compared on an equal basis. Those not familiar with the concept and practice of discounting and compounding can get an overview from a companion document (Gregersen and Contreras 1992).
Because forests take long periods to grow and change naturally, and because some uses today affect the potential for other uses in the future, a time dimension has to be attached to all measures of forest value. Good decision making depends on understanding society's preferences for consumption today over consumption sometime in the future, i.e., the discount rate or time value that society applies to future vs. present consumption.
A concern of decision makers dealing with forests is the uncertainty surrounding various estimates of value available to them. Two points are particularly relevant to this discussion. The first relates to the poor state of knowledge about the physical input-output information associated with forest change; the second relates to uncertainty about future values.
Adjust for poor input-output information.
Valuation is hampered by uncertainty over basic input-output relationships and yields for tropical forests. Information on such relationships is an essential ingredient in the valuation and decision-making processes. Information on the productivity, dynamics, and other basic characteristics of tropical forest systems is weak for most forest areas. Very little is known about spatial patterns of forest products production, and even less about how such production affects the ecology of the species and ecosystems involved. Also, because of the heterogeneity of composition of most forests, studies of the composition and values of a particular location are very site-specific, and the results often cannot be extrapolated usefully over larger areas to arrive at total values for a forest.
A specific example of uncertainty is the impacts of forest change on climate change. The burning of large areas of cleared forest releases substantial amounts of carbon dioxide into the atmosphere. But it is not clear to what extent this additional CO2 is absorbed, for example, by new vegetation on previously burned land. Nor is the likely impact of rising carbon dioxide levels on different parts of the world clearly understood yet. Similarly, it is known that transpiration from tropical forests accounts for a substantial part of recycling moisture back into the atmosphere; but empirical evidence as to the impact of disruption of this flow through forest removal is limited and inconclusive.
The decision maker can take certain actions to deal with data-poor situations and high levels of uncertainty surrounding available data and information. The various options include the following:
use information from similar decision situations elsewhere, e.g., from forest areas elsewhere with similar conditions, but better data;
apply various extrapolation methods to the available data, or reorganize existing data in different ways to bring out more clearly the relationships being sought;
use proxy values or measures;
use indicators (such as species indicators in looking at change in forest over time);
use additional, low cost data collection methods to improve existing data, e.g., rapid appraisal methods; and
use sensitivity analysis.[4] (Many economics textbooks will suggest use of sophisticated probability models to analyze risk elements in decisions. Such approaches involve estimating the probabilities that certain events - both quantity and price variations around the best estimate - will take place. While useful as sources of information, such analyses seldom provide the bottom line information needed to judge alternatives according to common economic decision criteria.)
Further, there are some things that can be done to improve the situation over time. These include making data and information needs clearly known to researchers and developing new data generation methods that can be applied over time to improve the data situation and reduce the uncertainty surrounding estimates.
Adjust for uncertainty over future values
Uncertainty over future values is a problem that affects all valuation approaches, but in different ways. Thus, it is an important contextual issue to consider before choosing valuation methods. For example, uncertainty surrounding the dynamics of use, and thus the value of the products of tropical forests, is great. Gathering and sale of forest products is usually just one of several income generating options available to those engaged in them. Profit margins and returns to labor are typically very narrow, so that economies based on these activities can be very fragile. The emergence or decline of alternatives, changes in labor availability, and fluctuations in forest (or crop) prices, are among the factors that can trigger rapid shifts into or out of forest-based activities. Thus, present values and magnitudes of involvement in the forest-based sector provide only limited guidance as to the future values of forest products.
The history of agriculture, and of industries based on agricultural outputs, suggests that the path of emergence of major commercially important forest-based products would follow that of rubber, oil palm, coffee, and the many other originally wild forest species domesticated as plantation (or smallholder) crops cultivated outside the forest. Growth in commercial demand for a forest product could thus cause the value of production from the natural forest resource to decline rather than increase, as synthetics or plantation grown outputs take over. Such a transition would, of course, affect the values associated with the natural forest outputs.
Recent studies have pointed out that even subsistence supplies tend in practice to come from semimanaged or domesticated sources rather than from the unmanaged forest. In locations studied in the forest zone in West Africa most forest foods and saleable produce were gathered or hunted in secondary wooded formations such as bush fallow and farm woodland, with only selected products coming from the forest itself (Davies and Richards 1991, Falconer 1991). To some extent this is because of proximity and convenience. But to a considerable extent it reflects manipulation over time of the forest structure in favor of species and products that either do not occur in the natural forest or which can be produced more intensively in a fallow or managed system. As pressures on land reduce the area under secondary tree cover, farmers widely move to planting particular species of value to them. The distinction between production from the forest and from domesticated sources is therefore not always a clear one, further complicating the task of deciding what part of the values derived from tree outputs is actually attributable to the continuing existence of the natural forest.
Some well-tested ways of dealing with the implications of uncertainty surrounding future forest values are as follows:
Use sensitivity analysis (see Gregersen and Contreras 1992). In situations where values are highly uncertain, it is common to carry out a sensitivity analysis to see how sensitive the various measures of worth calculated using the value estimates are to alternate assumptions concerning various values. If the outcome is not sensitive to changes in value assumptions, then one has to worry less about uncertainty surrounding the assumptions. On the other hand, a highly sensitive outcome may show that more effort should be devoted to reducing the level of uncertainty.
Focus on irreversibilities. Other things being equal, more time and effort should be devoted to reducing uncertainty surrounding irreversible events than reversible ones (see previous discussion on policy context). By irreversible, we mean those events that trigger irreversible change. For example, clearcutting a given old growth forest precludes forever the pristine wilderness benefits from that forest. The reverse is not so: one can opt for the wilderness benefits and always reverse the decision and cut sometime in the future.
Closely related to the previous point, we should avoid unsustainable positive activities. We can develop early warning signs that an activity or change is becoming unsustainable.
Develop contingency plans that build on alternate sets of forest values so if uncertain events move in one direction or another we will be prepared.
Ultimately, however, we cannot escape the need to make assumptions and guesses about the future.
The guidelines provided above are not presented as rules to be followed, but as suggestions on how the manager's job might be made easier and the results made more reliable. Ultimately, if a relatively sophisticated economic valuation exercise is needed, the manager will have to bring in economics expertise. Effective economic valuation - particularly of the comprehensive benefits to be derived from forests - is a complicated and complex task. Doing it right can provide much useful information for decision making. Doing it wrong can be worse for decision making than not doing it at all.
|
[4] In a sensitivity
analysis, we vary the values of key factors and relationships to see what impact
such variations would have on the measures of economic worth, i.e., we look at
the sensitivity of the measures to changes in values. |
