The Commissioner's instructions to the panel open with a general paragraph calling for an appraisal of Bureau of Reclamation procedures, and for our own recommended basis for procedure. The main subject with which these procedures deal – secondary benefits and costs arising from water-use projects – has been defined to us as consisting of differences, not covered by primary benefits and costs, in the total national output of goods and services with the project as compared to similar output as it would be without the project. We understand that it is as measures of these differences that we are asked to appraise existing or proposed procedures.
Instruction (1) concerns Bureau of Reclamation procedures. After careful consideration, it is our conclusion that these procedures for evaluating secondary or indirect benefits and costs include some elements for which quantitative estimates are warranted, and further elements for which an arguable (but not conclusive) case could be made for limited application of quantitative estimates; but that the applications actually made by the Bureau go far beyond what can be soundly identified as quantitatively measurable secondary benefits (in the above sense) attributable to public water-use projects. More broadly, we believe in the importance of secondary benefits, but find them so ramifying, involved and conjectural that the attempt to compute them as a national total, in dollar terms, by the methods of the Manual or any other methods that appear at present available, cannot properly be regarded as "measurement," though computations of pertinent items may be useful as guides to judgment in rating the importance of these benefits.
Accordingly, we are able to "set forth a recommended basis for the evaluation of secondary benefits and costs" as directed in instruction (2) only on the assumption that "evaluation" can include, for important parts of these benefits and costs, ratings by the exercise of judgment which are not precise enough to justify regarding them as quantitative measurement. As to what secondary benefits and costs should be taken into monetary account, these appear to include the following.
(i) Increased employment during construction, as a non-recurring benefit or offset to cost of construction, dependent on expected economic conditions during the construction period, and subject to offset, in conditions of high-level employment, for the non-measurable impacts of probable inflationary pressures and/or seriously high taxes, including business taxes.
(ii) Locally-accruing gains as a basis for securing reimbursement. The propriety of including these in an appraisal of aggregate national benefits appears to be an arguable question, difficult to settle in favor of either affirmative or negative.
(iii) Local "public" benefits.
(iv) Increased (or decreased) utilization of identifiable existing facilities.
As to the more general question of an aggregate increase in the national product attributable to a project (excess of product with the project over product without it) during the period of operation, as distinct from that of construction, we believe there may be such effects of three general sorts: (a) in the case of irrigation, increase (in the above sense) in productivity of mobile factors of agricultural production resulting from increase in the available supply of arable land, (b) expansion of the field of employment of non-agricultural sorts from a net increase (in the above sense) in the amount of primary agricultural products flowing into the economy, and (c) the furnishing of other basic facilities for production or community life by more economical methods than the cheapest available alternative, the saving representing an amount of productive factors freed for other uses in the general economy. As to (a) above, we think it possible (though far from certain) that current investigations may in future yield results by which the magnitude of the increased marginal productivity might be approximately indicated, but we know of no methods now available which we could conscientiously recommend as accomplishing this. As to secondary increases of product in general, the main methods used by the Bureau are estimates of benefits "stemming from" the disposal and processing of the products grown on a project, and benefits "induced by" by the spending of the primary income received from the sale of the project's raw products. Both of these methods appear to contain some probable validity; but as to how much (i.e., how large an estimate on these scores is warranted) we think that is a matter of judgment and the best check on arbitrary estimate is a survey of the forms in which such an increase in the national product must be embodied (as developed in Part III, Section 3, below).
This being the nature of our judgment, we are hardly in a position to recommend an alternative formula purporting to measure these secondary benefits. We recognize that both the majority and the minority of the Subcommittee on Benefits and Costs have accepted the "stemming" and "induced" formulas as parts of their (conflicting) procedures; and accordingly our specific comments include, in addition to critical examination of these formulas, conditional suggestions that if formulas of these general sorts are to continue in use, certain kinds of limitations should be imposed on them, tending to prevent them from being built up to totals that appear clearly unwarranted. Since the Commissioner's eight particular questions presuppose the use of procedures of these general sorts, and get their operative meaning from this setting, our answers should be understood in the same sense, as what would be applicable if and to the extent that such procedures are to continue to be used, but not as implied by accepting the idea that these procedures constitute quantitative measurement of the aggregate difference between national product with a project and national product without it.
Some of our specific recommendations bear on matters that do not stand or fall with this major question of basic procedure, and the significance that should be attached to it.
With this explanatory comment, we present our answers to the eight questions put by the Commissioner.
1. Basic assumptions and procedures in the Bureau of Reclamation Manual.
These assumptions and procedures appear sound to the extent that they follow what we conceive to be the controlling principle: that of the "with and without" comparison. They appear sound, as against the majority report, in rejecting double deduction of costs of secondary activities, which an unsigned memorandum describes as "netting benefits already net." However, they do not appear sound in adding benefits "stemming from" and benefits "induced by" a project nor in carrying the computation of "stemming" benefits through to the final consumer and giving the project credit for the whole. Our report will weigh alternative proposals for dealing with this difficulty and will make recommendations, the reasoning on which they rest being developed in the appropriate sections.
2. (a) Should it be assumed that needs met by a project would be met if the project were not constructed?
The answer appears to be: "yes, if it appears reasonably clear that this is what will happen – for example, where power will be produced by a steam plant if not by a hydroelectric project." There it is merely a question of what uses private enterprise will make of resources as a result of a project not being undertaken, the probable similarity of the needs met should be examined. In that case it is our judgment, that any assumption of difference in needs met which would carry an assumption of smaller secondary benefits without the project than with it, should bear a definite burden of proof, which burden will not be easy to sustain.
2. (b) "By what procedure should the analysis account for stimulating expansion of the nation's productive capacity?"
We believe this factor to be highly important, but have found no procedure that can properly claim to measure it quantitatively. The choice appears to lie between two courses: (1) treating this as one of the factors which can be described and given weight by judgment; and (2) using procedural rules of thumb, presented as such, without claim that they can truly "measure" this quantity.
3. "Can the procedures for primary benefits and the general principles of the Subcommittee's May 1950 report be logically expanded to provide adequate evaluation of secondary benefits?"
The report of May 1950 commands high respect for its understanding and general soundness, and is an excellent but incomplete guide. For purposes of expansion of primary-benefit procedures to secondary benefits, its statement on page 9 needs clarification on one key point as to the use of costs to represent benefits foregone. This lack of clarity permits a major disagreement as to whether secondary benefits should be counted on both sides of the "with and without" comparison. The same passage in the May 1950 report needs to be qualified or clarified by the proviso that the significant alternatives are not limited to uses of the identical resources, but include any uses of resources that will be made in the economy in the absence of the project. This qualification figures in one of our suggestions on procedure. As to procedure for determining the scope of a proposed project, the report appears to fall into an error, not of basic importance, which can be rectified without difficulty of principle.
As to the question asked about expanding primary-benefit procedures to cover secondary benefits, it appears that a simple "yes" or a simple "no" answer would be about equally misleading. There are common elements, but the more complex features of ramifying and non-marketable secondary impacts require additional analysis. It is notable that "induced" and "stemming" benefits are barely mentioned in the May 1950 report, and are not analyzed, thus avoiding the most thorny differences between primary and secondary procedures.
4. (a). "Should project costs, associated costs, and secondary costs, in terms of market value, be considered an adequate measure of benefits foregone from alternative uses?"
This is a far-reaching question of central importance. Briefly, ordinary accounting costs should not be so considered; full economic costs, including market rates of return on investment and compensation for entrepreneurial services, may be taken as adequately representing the direct and marketable benefits foregone from alternative uses, but not secondary for which private producers cannot collect market compensation. If these secondary benefits are reckoned on one side of the "with and without" comparison, they should properly be reckoned or reflected on the other side also, in one way or another.
4. (b). "Should the effects of alternative uses be compared with or deducted from the benefits of project uses?"
Unless one notes the context out of which this question arises, one is tempted to reply: "What difference does it make?" The question appears to arise, in connection with the dispute over double deduction of costs in reckoning secondary benefits. And the answer appears to be that a single deduction of this sort is warranted, but not a double deduction. The insistence on comparing rather than deducting seems to be intended as a way of avoiding the second deduction made in the example given in the majority report, top of page 9.
5. "Should the same basic assumptions govern the analysis of primary and secondary benefits?"
This appears to have been dealt with in the answers to Questions 3 and 4(a) above.
6. (a). How can measurements of benefits from a local viewpoint be converted to represent the national public viewpoint?"
In principle, the answer appears to be: by making them one side only of a comprehensive national "with and without" comparison, and deducting corresponding benefits that would accrue elsewhere without the project. However, in practice, local benefits are identifiable, and less uncertain than unidentified nation-wide alternative benefits. Where it seems warrantable to assume that there will be a net national balance of benefits larger than the sum of benefits accruing within the locality, it may be warrantable to count the locally-accruing benefits as part of the national sum, provided the rest of the national sum computed is sufficiently limited to safeguard against exceeding the probable national total.
6. (b). "If direct irrigation benefits (increases in net farm income) represent a national as well as a local viewpoint, should increases in net income in secondary activities represent both viewpoints?"
This question appears to be controlled by the answer just given, and also by the answer to Question 4(a) above, in which the economic meaning of "net income" hinges on the kind of costs that are deducted in arriving at it. One answer would be: Neither primary nor secondary net income should be accepted as representing national benefits unless the costs used are such as represent national benefits foregone, and secondary benefits foregone are taken into the reckoning.
7. "Do secondary benefits vary substantially for different types of commodities and different projects?'
There is a strong presumption that national benefits "induced by" spending bear a substantially equal ratio to the spending, or to the income from which the spending is derived, not differing in measurable ways between different commodities or projects. This kind of benefit is generated only by a net increase of spending, and is most clearly applicable to cost of construction during the construction period. Where locally-accruing benefits, "induced" or "stemming", 'are computed during the period of operation, they might be expected to differ between products and projects; but these differences should not affect the factors used to set limits on any national total which may be used as representative of secondary benefits.
National totals of benefits "stemming from" a project may differ slightly for different kinds of products and projects; but we do not believe they differ to anything like the extent, or in the way or for the reasons that appear in the method now accepted. We believe this method seriously distorts the relative benefit-standing of different kinds of products. We believe it is more faithful to the probable facts to adopt the prima facie assumption that these benefits follow the same proportions for different products in the same general class of project, and the, same for products with a project as without it, putting the burden of proof on any claim of quantitative difference. As to differences of view between the water-use agencies, the most controversial bearing of this question is on the disputed method of computation culminating at page 9 of the majority report. The method of the Manual has an effect equivalent to assuming that secondary benefits with the project exceed those without it in the ratio which of "value added" bears to costs in the "stemming" processes. In our judgment, this is too large a difference for the purpose in hand, exaggerating the net total of "stemming' benefits. The costs used in this analysis appear not to afford valid evidence for an answer to this difficult and far-reaching question.
8. (a) "Can an identical procedure be used to evaluate secondary benefits from irrigation, power, municipal and industrial water supply, and other purposes?"
The problems appear sufficiently different to call for variants of specific procedure. It should be possible to place these on a common basis of principle.
8. (b). "Should savings to power consumers from lower power rates be considered a primary or secondary benefit?"
The economic analysis of proposed hydroelectric power developments should be viewed primarily as a comparison of the relative economy of power generation from different possible sources. It is reasonable to assume that power demand in the United States will continue to grow and that if a proposed hydroelectric plant is not built, the demand will be met by power from the most economical alternate source, usually steam. Hence the total benefit figure for comparison with the project costs added by the inclusion of power features in a river basin development should be the cost of equivalent power from the most economical alternate source. The use of such a benefit figure will fully reflect savings to the public resulting from the decision to choose hydro power rather than steam power for generating purposes. This topic is discussed at greater length in Part V of our report. In Part V, we recommend that the total power benefit, namely, the cost of power from the most economical alternate source, be viewed entirely as a primary benefit.