The

that should be made according to the old section 21 formula. apportionment of this amount of money according to that formula has, in my opinion, nothing to justify it except tradition. The improvements which have been made through successive applications of this formula have been very much worth while, as we all recognize. That is not to say that the Federal expenditure might not have been productive of even more good if the apportionment formula, set up to meet conditions of 1916, had been changed from time to time to keep pace with changing highway and traffic conditions and needs. Probaby the severest indictment that could be made of the present and future use of this old formula would be to concede its applicability to the 1916 road and traffic conditions which it was framed to meet. If it was sound then, it cannot be sound now.

Factual data now available make it obvious that the factors which make up the section 21 formula (and with a little different weighting, the formula in H. R. 2426) are very poor measures of today's road need. These factors, as you recall, are are, post road mileage and population.

Most people probably recognize that area in itself has no significance even in relation to road mileage, except as the area is densely covered with roads or sparsely covered with roads, which in the end can be measured, if you will, by the road mileage itself. There are, however, some persons who have stoutly maintained that area is a valid factor in measuring road needs. It is interesting, therefore, to take two regions of equal area in order to see what variation they show in respect to other factors which are obviously and admittedly related to road need.

There are 8 States in the Rocky Mountain census area which have about 29 percent of the area of the United States and consequently would get 29 percent of such part of any Federal authorization as is based on area. There are 26 States in the census areas east of the Mississippi River which, by chance, have an area almost_identical with that of the 8 States in the Rocky Mountain region. To the extent that a formula is based on area, these 26 States would receive the same amount in Federal-aid as would the 8 States in the Rocky Mountain region. How do other characteristics of the two regions compare?

Only 4 percent of the motor-vehicle traffic in the United States is developed on the road systems of the eight States. According to the States' own estimates of immediate road needs, the eight States have but 6 percent of the total needs of the country. It is further signifi

1 These estimates were made in July 1943, by the various State highway departments at the request of the American Association of State Highway Officials. They were not based on any suggested uniformity of standards, or of the road systems to be considered in adding up the needs on main or principal highways. However, in groups of the extent used here for comparative purposes, it is probable that they provide a reasonable comparison. It may be noted that such studies of these estimates as we have made indicate that the more urban industrial States have had a tendency to exclude urban highway needs, whereas rural road needs have been adequately included in most States. Obviously the largely rural Rocky Mountain area would not have its road needs greatly increased by the inclusion of urban needs, whereas the road needs of the area east of the Mississippi River would apparently be substantially increased by a more complete inclusion of urban needs. This correction would still further increase the disproportion between road needs and area as a measure of such needs.

In presenting data on the secondary highway systems last week, Mr. F. R. White used another tabulation of road needs which was prepared by Mr. Gentry, of Texas, and which was on a long-time rather than an immediate basis. According to this tabulation, the Rocky Mountain States claim 7 percent of the Nation's road needs, and the States east of the Mississippi claim 60 percent. Here again, however, urban highway needs quite obviously were not estimated on a comparable basis with other needs.

cant that they have been contributing but 4 percent of the motor-vehicle imposts that are drawn from the motorists of the country. On the other hand, the equal area from the Mississippi River east develops 62 percent of the motor-vehicle traffic, claims 59 percent of the highway needs, and provides 64 percent of the Federal motor-vehicle imposts. This utter lack of any relationship between area and factors of admitted importance in consideration of Federal-aid legislation is illustrated in chart 2. Even on the basis of the States' own estimates of their road needs, area is obviously not a measure of such need.

At first glance, road mileage would seem to be a reasonably fair measure of road need. Examination shows that this is not the case, however. According to the section 21 formula-or the slight modification of it proposed by our executive committee-a mile of road is a mile of road. When you come to build or rebuild a mile of Federal-aid

road, however, that is not the case. There are great differences in the cost of a mile of road, as shown in chart 3.

Now what causes these differences? It is not that Massachusetts, and New Jersey and Connecticut are more extravagant than South Dakota and Wyoming and Nebraska. I believe that the reverse is true and that the road standards of the first three States are probably lower than those of the last three, as I shall attempt to point out to you in a moment. Various factors account for the differences in cost. Ruggedness of terrain, severity of climate, unit costs of labor and materials all may play a part. Much the greatest factor in producing the cost differences, however, is the different job that different roads have to do. To illustrate: The annual average daily traffic on an average mile of the New Jersey State highway system as determined by the planning survey was 6,300 vehicles a day. The annual average daily traffic on an average mile of the New Mexico system was, in contrast, approximately 160 vehicles a day. The road that has to be built to carry even the average New Jersey traffic is an entirely different sort of facility than that required to carry less than 200 vehicles a day. The differences are even more extreme in the case of a road that carried an annual average daily traffic of 40,000 vehicles a day, as some miles of New Jersey roads do. And note, I am not talking about city streets. I am talking about part of the highway system, such as in Connecticut, where before you move out of the influence of one city you have passed into the influence of another, although you are traveling in areas outside the corporate limits of the cities.

This explains the difference in Federal-aid per mile costs and reveals the weakness of using road mileage as a fair basis for the apportionment of Federal funds. The differences in cost per mile are illustrated graphically in chart 3.

Population is probably a more reasonable factor for an apportionment formula than area or road mileage, but it is not a direct measure or even a very close measure. Has anyone stopped to consider why population has a place in such a formula? It is my understanding that the alleged justification for using population as a basis for allocating Federal funds is the supposed correspondence between population and the amount of road use. It has already been pointed out that even in comparing two large cities there is no such correspondence. Considering State-wide areas, moreover, there are available much more accurate measures of road use than population. Gasoline consumption is one, and perhaps the most accurate. Automobile registration is another. Chart 4 shows the lack of any consistent relationship between population and gasoline consumption in the different States.

Again considering State-wide areas, road need, which we may define as the potential usefulness of correcting road deficiencies, is probably as well measured by road use as in any other way. Mr. White, of Iowa, has argued road use as a justification for the apportionment of funds, and Mr. Purcell and Mr. Williamson have both stated that in their States, in apportioning State funds to the counties, road use, as measured by automobile registration, has been found a satisfactory basis for apportionment. But population is a very uncertain measure of road use-certainly far from the best available measure, and therefore a poor factor for inclusion in an apportionment formula.

Note: Costs are based on four year average of all federal aid highway projects completed in fiscal years ending in 1938, 1939, 1940 and 1941, (from annual reports of Commissioner of Public Roads Administration).

21 New Hampshire
22 Maine

23 Rhode Island

24 Colorado

25 Illinois

26 Minnesota

27 Nebraska

28 New Mexico

29 South Dakota

30 Missouri

31 Wisconsin

32 Massachusetts

33 Oklahoma

34 Pennsylvania

35 Maryland

36 Virginia

37 New York

38 North Dakota

39 North Carolina

40 Georgia

41 South Carolina

42 Louisiana

43 West Virginia

44 Tennessee

45 Kentucky

46 Arkansas

47 Mississippi

48 Alabama

EXPLANATION

This chart is obtained by dividing the 1940 gasoline consumption of each state by the 1940 population of the state.

Road use is widely rec-
ognized to be most accurately
measured by gasoline consump-
tion. It has been advanced
that road use is approximately
proportional to population,
and therefore that the popu-
lations of the different
states would provide a suit-
able factor for use in a
formula for apportionment of
federal aid funds. This is
not the case, however. The
number of gallons of gasoline
consumed per individual in
each of the states is shown
in this chart and is seen to

vary widely. If gasoline con-
sumption is an accurate measure
of road use, then population
is not.

Source of Data: Public Roads Administration statistical table 6-2, 1940.

ᏟᎻᎪᎡᎢ 4

How do the over-all apportionments of funds to the States provided for in H. R. 4170 compare in equity with those provided for in H. R. 2426?

Since there obviously is a close connection between the extent of road use and the extent of public benefit derived from a road system; and since there would seem to be a close connection between the public benefit from a road system and the legitimate claim of that system to a share of public funds, let us examine the connection between the use obtained from the highway systems of the different States and the apportionments of Federal funds which they would receive from H. R. 4170 on the one hand, and H. R. 2426 on the other. Under which bill will the Nation's needs be best served?