Conclusions and Recommendations
B. Summary of Principal Recommendations
This summary does not include all the minor and detailed recommendations and suggestions which appear in Parts III - VI of our report, but concentrates on those of major importance.
1. Secondary benefits are much less certain and calculable than primary, and more dependent on far-reaching hypotheses. moreover, it appears that usable formulas cannot be based on data that are capable of furnishing complete and accurate comparisons of effects "with and without" a given project. In recognition of this, we recommend that, in benefit-cost ratios, primary and secondary benefits be separately shown; e.g., if the total ratio is 1.8:1, but the primary-benefit ratio is 1.15:1, the result be expressed as (1.15 + .65):1.
2. It appears that Manual procedure for determining the scope of projects could result in carrying one project out on a scale such that the last separable increment would have a separate benefit-cost ratio barely more than one whereas early separable increments of some other projects, whose total benefit-cost ratio was lower than the total benefit-cost ratio of the first, might show a higher ratio than that of the final increment of the first project. We therefore recommend that benefit-cost ratios for separable increments be separately stated, in order that appropriations may be so allocated as to produce the combination of increments yielding the greatest total benefits.
3. We find that benefits "induced" by increased demand and benefits "stemming" from increased supply are of such character that, with minor exceptions, it is not proper to add them in figuring total secondary benefits from a given project at a given time, During construction, secondary benefits are all "induced", while during subsequent operation induced benefits (so far as not covered by "stemming" benefits) appear to be limited to identifiable particular cases of minor magnitude, and the main secondary benefits are of sorts for which the "stemming" procedure (with modifications which we will suggest) affords an appropriate rule-of-thumb approach. We therefore recommend that the calculation of benefits be separated between periods of construction and operation of a project, as, explained further in recommendations 4 and 5.
4. As to induced benefits taking the form of employing otherwise-unemployed resources during construction, we recommend that the Bureau explore with other appropriate agencies, such as the President's Council of Economic Advisors or the Bureau of the Budget, the possibility of setting up a sliding scale of such benefits, varying with the probable state of activity in the economy (the construction industry deserving special weight). Such a sliding scale might be made available as a standing formula for guidance in estimating such benefits. The appropriate time to apply the formula would be the last possible moment before decision on the undertaking of a project, when some reasonable forecast of the probable state of economic activity may be feasible. Any such overall national formula should be supplemented by local or area information which may indicate possibilities of mopping up local pools of unemployment or utilizing unused facilities. For the construction period, estimates varying with the probable state of economic activity, to the extent that forecasting is feasible, appear superior to blanket allowances, e.g., of 5% or 6%.
The result might be treated as an "offset to cost" of construction or as a public benefit. We see advantages and disadvantages for each method, and have no firm preference between them.
5. As to benefits "stemming from" increased supply of project products, we consider benefits of this sort exist, to an extent impossible to measure, but far less than the full amounts computed by the Manual procedures, We would accept a more limited form of this procedure as a rough approximate representative, not only of literal "stemming" effects, but also of more general effects, diffused through the economy and not at present measurable. We have in mind gains from lessening pressure of mobile resources on land areas other than project lands, and otherwise in-creasing productivity or lessening its decline. We propose four limitations on existing "stemming" procedure:
(a) Subtraction of full economic costs of "stemming" activities, or some agreed estimate of them, instead of the costs now subtracted.
(b) Reduction of disparities between commodities by exclusion of major processing operations.
(c) Allowance on the "without" side of the "with and without" comparison for benefits "stemming" from alternative uses of resources represented by project costs.
(d) As indicated in recommendation 3, above, induced benefits during operation be not included except in particular cases where a definite showing is made that their inclusion is justified.
6. Inclusion of "public" benefits is justified. We judge that, in the case of municipal water supply in instances where no alternative source is available, public benefits beyond rates charged for water are warranted and recommend further study of this question. In the case of irrigation, we take. exception to the allowance for "settlement opportunity", and recommend that this be treated as a qualitative, "intangibles" factor.
7. As to power, we find that it is proper to assume that, in the absence of a project, power will be supplied from an alternative source, and therefore propose that the total benefit imputed to project power be the cost of power from the next cheapest alternative source. This excludes all benefits of the "stemming" type, specifically those designated in the Manual as B-2, B-3, and C-3. If the B-2 type of benefits (benefits of private distributors of power purchased from a public project) are nevertheless employed, we find them computed on an exaggerated basis and recommend that they be reduced by deducting full allowance for economic return on all capital in accordance with recommendation 5 (a), above.
8. As to qualitative and intangible benefits and costs our study has led us to look toward diminished reliance on quantitative computation and toward attaching greater relative importance to qualitative effects of the alterations in distribution of population, types of community, etc. We therefore suggest that these matters are worth increased attention and study, including sociological aspects. These are, of course, matters that can be described and appraised only by judgment.