The testimony I have prepared can be put in full in the record. I will summarize that briefly. (The statement appears at p. 806.) Senator DOMENICI. Fine.

Mr. HANKE. The policy conclusions that I have derived that are presented in the written testimony and that I will discuss briefly today were primarily based on a study that Dr. Robert K. Davis and I did for the National Water Commission in 1971. That particular report was entitled, "Pricing and Efficiency in Water Resource Management."

Let me limit my remarks today to briefly what I think to be the important rationale for user fees on inland waterways and then some specifics in regard to the particular types of user fees and the level for those user fees that I would recommend.

The first element that you have alluded to already this morning with regard to a rationale for user fees deals with the concept of economic efficiency. Without belaboring that point, let me just say in order to attain economic efficiency and derive the most welfare or the greatest net benefit from the use of resources, one has to price the resources at their marginal cost, their relevant marginal cost.

If the price is too low, you tend to overuse the resources and overinvest in it. If it is too high, you tend to under-use it and underinvest in it.

The second efficiency argument that has not been talked about, this morning at any rate, deals with the political incentives associated with not charging a price for output that is equated to the marginal cost of that output. In the political marketplace, if the price for output is below its marginal cost, there is a tendency to put pressure on political representatives to overestimate the benefits to be derived from the output in question.

On the other hand, if the price happens to be too high, there is pressure to underestimate the benefits from those projects. Of course, with regard to cost, if the price is low and zero in the case of waterways, there are political incentives put on to underestimate the costs of these projects.

So that by charging the price that is equated with the marginal cost, not only the pure economic incentives in the market itself for these resources would be improved but also the political marketplace would be improved and the biases that have crept into that particular process would be eliminated or reduced.

The third aspect deals with the bureaucratic incentives, and the bureaus that are engaged in the construction and improvement operation of these waterways are, of course, in the business of doing just that.

Even if the political market was more perfect that it is, there would still be the tendency for the bureaucratic market itself to overestimate the benefits from these projects and underestimate the

cost.

We have a long history of very shoddy benefit/cost analysis, not only in the waterway business, but throughout the public establishment. The main reason for this is due to the fact that there are indeed great incentives for bureaucrats to bias the benefit/cost analysis.

The only check on these studies is a system of user fees in which the price charged for the output was equated to the marginal cost of that output.

This to my way of thinking is one of the more important reasons for tying the locks and dam 26 to the user fee proposal. You have to have a policy of user fees that is consistent with the assumptions of benefit/cost analysis to have a reliable benefit/cost mechanism to make projections into the future.

The only way that you can do that with the standard benefit/ cost techniques that are used now is to have a user fee that equates price of output to marginal cost of providing that output. So those are three elements in terms of the rationale for user fees.

Just what types of user fees would I recommend? The first element that would reflect the marginal cost of operating and maintaining the waterways would be segment tolls in which the segment toll was equated to the relevant marginal cost of maintaining and operating the particular segments in question.

This would be based on a ton-mile basis per segment.

This would be employed, since this decision has to be made on all waterways, whether they are existing waterways or whether they would be new waterways, or improvements in waterways. So if you have the testimony that I prepared in front of you, on the next to the last page, page 11, there is a little table here with the elements of the particular user fee system that I would recommend.

The segment tolls would apply, therefore, on all waterways and be equated to the annual operating and maintenance cost per ton mile on a particular segment involved. Basically, this would give planners information about whether to continue to maintain on an annual basis these waterways or to phase these waterways out.

I might add that the differential in these segment tolls would be rather great. They could be up to 350 times to 400 times different in terms of magnitude; but, of course, the lower Mississippi being extremely low and segments like the Kentucky Waterway being extremely high.

The second component deals with the marginal capital cost associated with either improving a waterway or building a new waterway. This particular component would be captured by an annual license fee again on a segment basis for those waterway users who wanted to use a particular waterway. They would have to have a license. They would pay an annual license for going on the waterways they wanted to operate on. This charge would not be imposed obviously on existing waterways, because there is no marginal or incremental investment. There is no capital cost associated with those. It would only apply to the new waterways or improvements in old waterways.

This would give us more accurate information as to the benefits or willingness of shippers to pay for improvements or new extensions to the waterway system. The third component that is also a relevant marginal cost deals with congestion tolls, which have not been dealt with, this morning at any rate. These costs are not outof-pocket costs in the sense that operating and maintenance and investment in new segments or improvements in existing segments

of waterways are. They are costs that the users impose upon themselves in the sense that once a particular user comes into a lock, he not only has to wait himself, but he creates waiting costs for everybody else.

These waiting costs he imposes upon other people are not taken into account in his calculation on whether to enter a lock area at a particular time. But they do in total add up to a social cost that is relevant and should be charged for.

This, to my way of thinking, is a cost that could be reduced or eliminated by congestion tolls at the locks, and it would increase the value of our inland waterway system because the highest valued users of the waterway would be through the locks before the lower valued users of the waterway. Just a simple numerical sample: If you have one barge operator who has waiting costs of $100 an hour and another barge operator with $600 an hour waiting cost, and the $100 operator goes through before the $600 operator, you have lost $500 an hour by having this current existing allocation process that we now have. If you had a pricing process that charged congestion tolls to eliminate the queues, the $600 or higher valued man would go through first, and you would have a net gain of $500.

By the way, the congestion fee should be charged on existing as well as new waterways. Those, in brief, summarize the three elements that I would see as appropriate to attaining economic efficiency in the operation of waterways. The items are consistent with those items listed in the bills in terms of the structure of the system I propose, and the one that could be accommodated within the bills. However, the level of charges is constrained in the bills in the 50-percent provision, and in fact with the system that I am recommending, not only would you collect 100 percent of your out-ofpocket costs, but you would collect revenues in addition to that with a collection of congestion tolls. So it would be more than a 100 percent of out-of-pocket costs that would actually be collected in this kind of system. I might add that not only will the congestion tolls increase the efficiency of the existing waterways, but that revenue could be used to improve the locks and expand the locks where the congestion actually occurred, so you would have there, again, a good investment signal sin terms of willingness to pay.

That concludes my testimony.

Senator DOMENICI. Thank you very much.

Rather than question you, let's have both of you testify and then we will ask questions of both of you.

STATEMENT OF JOHN C. SPYCHALSKI, PROFESSOR OF BUSINESS LOGISTICS, COLLEGE OF BUSINESS ADMINISTRATION, PENNSYLVANIA STATE UNIVERSITY

Mr. SPYCHALSKI. Mr. Chairman and members of the subcommittee, my name is John Spychalski. I am a professor of business logistics in the College of Business Administration of the Pennsylvania State University. My duties include teaching and research in transport economics and policy, an area in which the subject of user charges on publicly provided facilities obviously plays a significant role. I

would like to emphasize that the views expressed in my testimony are my own. They bear no relationship to either the Pennsylvania State University, or any other faculty member within that university. I think we have already heard much about the general case in support of waterway user charges from the standpoint of principles of economics. I will not repeat those principles here. There is some reference to them in the prepared statement which I filed with the subcommittee. However, I do want to reiterate that these principles are valid and acceptable. They have been presented clearly and impeccably in more than a score of reports issued by both governmental and impartial private study groups since the early 1930's. Indeed, as I listened during this morning's parade of witnesses, I could not avoid the thought that one would have heard a scenario much the same if one had the ability to turn back the clock to the 1950's and 1940's. Let me now focus on various specific elements within the proposed S. 790.

Like some of the other witnesses, I think that the proposed bill's 50-percent recovery standard is infinitely better than the present recovery rate of zero. However, the 50-percent limit still leaves open a wide range for the underpricing of inland water transport services in comparison with the services of other transport media.

The potential for continued transport resource misallocation which stems from this condition prompts me to respectfully suggest that the subcommittee consider replacing the 50-percent limit with a provision that would produce a somewhat greater measure of equality between user charge-generated revenues and navigation-related expenditures.

I would like to endorse the proposed authorization of the application of congestion charges for navigation facilities. We have already heard some testimony concerning the conceptual or theoretical rationale for such charges. To demonstrate their practicality, I would like to focus on some existing sets of conditions that, in effect, already accomplish the same thing, or are directed toward similar objectives, in other areas of transport activity. I would first like to mention the pattern of pricing that is now going on and has for many years in the pricing of services of the so-called exempt inland water carriers. Because of their exclusion from economic regulation under the terms of section 303 (b) of the Interstate Commerce Act, such services can and do carry freight at rates which move freely upward and downward, minute by minute, day by day, week by week, in close response to changes in the demand for and supply of inland water carriers' service capacity.

During periods of relatively heavy water transport demand, you will find in existence relatively high water freight rates that cause some shippers who value the water transport service at amounts less than at peak period rate levels to seek out and utilize alternatives which yield higher net gains to them.

Such shippers will either use other forms of transport or hold commodities until peak water freight rate periods pass and fall to levels equal to or less than those which they-the shippers-are able and willing to pay. This is nothing more than the adjustment of transport relative to other elements in the logistics system. You

88-866 - 77 - 32

can adjust your inventory holding policy, your storage location policies, et cetera. You may even reconfigure your marketing policies in various ways to accommodate changes in the real economic cost of transport as compared with other cost elements in a logistics system.

Additional support for moving ahead with the application of congestion charges on the inland waterway system can be drawn from the fact that the Congress has, in effect, already endorsed the principles underlying such charges. It did so when it enacted certain provisions contained in the Railroad Revitalization and Regulatory Reform Act of 1976.

Those provsions are designed to encourage the development of new railway ratemaking patterns based on seasonal, regional and other forms of peak and off-peak fluctuations in demand for rail transport service. If we can presently ration exempt barge carriers' capacity on the basis of prices that change readily in response to changes in demand, and if we can move forward with the applica tion of idntical pricing methods in the allocation of rail car capacity, then I suggest it is time to consider applying the same pricing approach to the inland waterway infrastructure.

I also endorse the concept of segment charges. Professor Hanke has already ably articulated the basic points that favor that concept. The concept of segment charges, interestingly, also parallels some things that were endorsed by the Congress in the Railroad Revitalization Regulatory Reform Act of 1976.

For example, that law, the so-called Quad or 4-R Act, encourages the filing of separate rates for distinct rail services, and the pricing of such services in accordance with the various cash outlavs for such services and the demand therefor. Furthermore, the 4-R Act directs the Interstate Commerce Commission to develop a new cost and revenue accounting and reporting system for railways which is supposed to assure, and I quote:

That the most accurate cost and revenue data can be obtained with respect to light density lines, main line operations, and factors relevant in establishing fair and reasonable rates.

I think that one of the merits of the segment charge approach is that it will focus additional attention on the relative differences in costs and benefits of different segments of the inland waterway systems. It will frankly subject some of the relatively high cost/ low traffic density sectors of the inland waterway system to closer scrutiny.

I might note that we have had some similar developments in the area of the railway branch line question. That is to say, there is a parallel between low density waterways and the so-called railway branch line problem; in other words, segments of railways do not generate revenues sufficient to cover the cost of their continued operation. In some case public support has been provided for the continuance of money-losing rail branch lines. But the general direction, even in the recent Northeast restructuring, continues to be that of letting only those facilities which can support themselves in the longer run continue in existence. Economic logic suggests that the same policy approach should be applied to waterway segments which