tion managers in the form of new production orders. Similarly, changes in technology or in resource availability are usually quickly recorded by the market through altered relative prices.

This mechanism of conveying information about changes in tastes, technologies, and resource availabilities throughout the rest of the economy is a simple one. If parties with altered demands place additional orders for a commodity, producers will be faced with decreasing inventories or a backlog of unfilled orders. Viewing these as opportunities for increasing sales and profits, producers will increase their output. If, in the process of generating these increased outputs, profits rise, an added inducement to producers to increase production will be provided. Resources are thus diverted from producing other things to the production of the commodity with an increased demand. This response is accomplished without central guidance.

Such output responses occur fairly quickly and will usually yield outputs conforming to the patterns of demand at minimum cost. Naturally, the facility with which such responses are made is dependent upon quick communication of market information and the push of effective competition.

It is on the basis of this reasoning that we conclude that an economy which has well-functioning markets as the mechanism to allocate resources produces maximum national output (and income) for the society.

The economic efficiency characteristic of a smoothly functioning market economy has implications for public investment policy in the natural resource area. It suggests that the outputs of public investments should be evaluated by the actual or simulated market demands of users in terms of their willingness to pay, that is, to forego other products, and that the costs must measure the value of the opportunities foregone by diverting inputs to the public investment from other uses. Were public sector investments to be chosen on other grounds, they would be employing resources which could be producing a greater value in other uses. Such public undertakings, by diverting resources from a higher to a lower valued use, would cause a decrease in society's economic well-being. Indeed, if it is assumed that the market system is operating smoothly, it follows that the federal government should not consider a proposed project as adding to the nation's economic well-being unless the observed or simulated willingness to pay for the output exceeds the social value of the resources required to produce the output. These concepts of benefits and costs correspond to what are commonly known as primary benefits and primary costs. We therefore conclude as a second basic principle that:

II. Unless there are serious market failures and obsta-
cles to the smooth functioning of the market system,
total national economic benefits equal the real out-
puts of a public project valued at observed or simu-
lated market prices and total national economic
costs equal the real inputs employed in a project
valued at observed or simulated market prices.

Having set forth a definition of relevant economic benefits and costs for a market economy without serious imperfections, we would do well to state the conditions which must prevail if Proposition II above is to be logically valid. The conditions which must be met in the economy are:

1. Reasonably full employment of labor and capital;

2. Labor and capital mobility, i.e., the ability of units of labor and capital to shift to new jobs and uses;

3. No significant economies from large-scale production of pertinent commodities; and

4. Generally competitive conditions.

If these conditions are met to a reasonable approximation, then any secondary beneficial impacts by a project on market-related activities are simply local or regional in nature with offsetting effects occurring elsewhere in the economy. If funds are diverted from the private sector for purposes of public investment, not only are primary impacts of the foreclosed private investment foregone, but so are any net secondary impacts. If we wish to credit public investments with their own secondary benefits, we must also take into account the net secondary impacts which would have been experienced through the foregone private spending. There is no more reason to anticipate positive net national gains from secondary impacts than to expect negative net changes.

Suppose, for example, that a federal irrigation scheme producing water for various crops is constructed. Čertain regional industries will expand, both to supply fertilizer and machinery to the irrigation scheme, and to process and merchandise the crops. If the economy exhibits efficient operation characterized by the abovestated conditions, competition and mobility will provide additional labor and capital to these industries which will eliminate whatever temporarily higher profit rates they experience from the expansion. After the influx of labor and capital stimulated by the temporarily higher profit rates, profitability will fall to a normal rate in short order, and before long, this capital will be earning approximately what it had earned elsewhere prior to the project.

The people newly employed in these activities were bid away from other jobs, implying the existence of higher wages and incomes,

but again, competition among mobile workers will tend to reduce and ultimately eliminate wage differentials. Therefore, in the absence of significant departures from the four above conditions, secondary gains, if significant at all, will be temporary.

In the context of a national accounting stance and assuming that the economy can be characterized as a smoothly functioning market economy, relevant national secondary impacts will be negligible on both the benefit and cost side. This leads us to a third principle: III. If the conditions for a smoothly functioning market economy prevail, there is no justification from a national point of view for the recording of secondary benefits which would accrue to the region of project location, nor for the recording of secondary costs which are experienced elsewhere in the economy due to the financing of the public project.

The logic of the two preceding principles also generates the conclusion that where serious market imperfections are present, there may be secondary effects which do entail changes in the nation's net income and which will require either the measurement of secondary benefits and costs or adjustments to the observed values of primary benefits and costs. For example, when serious and intractable regional unemployment exists or when the nation as a whole is confronted with unemployed resources, or where a region with immobile labor and capital is confronted by a loss of some vital resource base, then a natural resources development project may result in secondary local income gains which are also net national gains. Adjustments to observed market prices of project inputs and outputs may also be required.

As an example of how secondary effects which accrue to a region may represent changes in national income, consider the case of immobile labor and capital. If labor and capital cannot (or do not) move quickly out of industries which are forced to contract as a result of the construction of a natural resources project (e.g. the displacement of non-irrigation agriculture by irrigated acreage), they will experience a decrease in their net incomes. This decrease in net incomes represents a reduction in national output over the period of the unemployment of these resources. Because a loss in the nation's income is experienced as a result of this immobility, federal planners should legitimately account for the loss in estimating project costs.

A second example of how secondary effects may lead to national increases in income relates to the existence of increasing efficiencies of large-scale production in some pertinent production sectors.

Assume that because of a natural resources development project, certain industries expand while others contract. If the expanding industries experience decreasing unit costs while the industries experiencing contraction have constant product costs, and if the degrees of expansion and contraction are approximately equal, then a similar volume of output will be produced at a smaller total input cost. The reduction in input cost constitutes a net national benefit and, like the above examples, represents a case in which real secondary economic gains occur for the nation.

An example of how an imperfectly working economy may require an adjustment of observed market values of project inputs and outputs can be cited in terms of the impact of, say, a dam building project whose construction period coincides with substantial unemployment of labor. If the on-site labor used would otherwise have been unemployed, its true social cost will lie below the apparent market cost. If off-site procurement requires production which utilizes otherwise unemployed labor, even in far-away regions, the cost of those off-site procurements must be reduced. Appropriate techniques for these adjustments have been developed.

From these examples, a further basic principle of benefit-cost measurement is derived. This principle, which is a corollary of the previous principle, can be stated as follows:

IV. When the economy is characterized by unemploy-
ment, resource immobility, decreasing costs, or a
lack of competition, it is appropriate to investigate
net national benefits and costs which derive from
secondary effects. Such benefits, when found and
quantified, should enter the benefit-cost analysis.

Having offered this principle, however, we would also emphasize a few caveats which relate to it. First, it should be pointed out that the unemployment relevant to the existence of real national secondary benefits generated by project construction must be longterm, structural unemployment, and not just that from a temporary recession. The planning-construction period and the operating life of natural resources projects each will exceed the duration of cyclical unemployment.

Second, we would also emphasize that labor and capital immobility should not be presumed to be a permanent feature of the social landscape. It is often on the basis of such immobility that "rescue operations" are proposed to bring water to established agricultural areas to replace exhausted ground water supplies. Because business complexes which specialize in agriculturally related activities as well as agriculture itself would be left idle if the area were forced to revert to dry farming or to abandon farming altogether, the

existence of substantial net secondary benefits for such investment has often been claimed. Surely over the period when capital and labor would otherwise have remained idle, the newly generated capital and labor income should be counted as a net national gain, as should any difference in land rent. It must be pointed out, however, that units of capital and labor will be immobile for a far shorter period than the life of the project. For this reason the incomes from avoided unemployment should be attributed as benefits to the water supply project only over the appropriate periods of immobility. The fact that technological and market changes would be inducing changes in employment and capital structure independent of matters of water supply makes it doubly difficult to apply the "with-without" criterion.

A third caveat pertains to longer term growth which might be induced by the project and the relationship of this growth to national gains. Clearly, the question of the contribution which such growth makes to national economic gain hinges on the possible advantages which exist for, say, processing primary products in that region relative to other areas. It is not warranted to assume that any particular project will automatically generate such related investments or that the incomes generated by such investments represent net additions to the national income. What is needed is a careful analysis of the extent to which the project creates in the region a comparative advantage relative to other regions in terms of basic raw materials, power, process water, or transportation. Moreover, if it is concluded that project-induced investment is likely to occur in the project area, the portion of the incomes created by this new activity representing net additions to the national income and the portion representing transfers from other areas must be determined. Only the net additions are countable as benefits.

Any such analysis of project-induced investment should be approached with care in the case of presently depressed areas, since it is reasonable to presume that the conditions which have resulted in a declining area's depressed economic condition will continue to inhibit further investment. It is unlikely, for example, that the provision of flood-free land or an improved water supply will suffice to make private investments profitable.

Finally, if the federal government is interested in inducing development in a particular region or set of regions, it should not be restricted in its choice of instruments to water resource or, more generally, natural resource development investments. Indeed, there is no presumption whatsoever that natural resource investments are more likely to be significant employment or investment generators than labor training programs, housing programs, recreation programs funded by the federal government, or federally subsidized private investments in the region, or federal investment in programs to