Mr. MACDONALD. Where the maintenance is paid for and administered directly by the highway departments, but I would like to make a little statement about that later, to-morrow morning, if it suits your convenience.

Mr. ALMON. Very well.

Mr. MACDONALD. I wish now to present Mr. Allen.

STATEMENT OF T. WARREN ALLEN

Mr. ALLEN. Federal aid, as of February 1, 1926, had been allotted to projects totaling 65,133 miles, of which 52,402 had been completed at an average cost of $18,400 a mile. This completed mileage consisted of 10,492 miles, or 20 per cent, of graded and drained roads at an average cost of $8,200 a mile; 4,775 miles, or 9 per cent, constructed with surface of sand-clay at an average cost of $7,500 a mile; 20,135 miles, or 382 per cent, with gravel at $10,500 a mile; 1,067 miles, or 2 per cent, waterbound macadam at $18,000 a mile; 2,939 miles, or 512 per cent, bituminous macadam at $29,500; 1,427 miles, or 234 per cent, bituminous concrete at $35,300; 10,761 miles, or 2012 per cent, Portland cement concrete at $37,900; 704 miles, or 11⁄2 per cent, constructed with surface of brick at $44,300 a mile, and 102 miles, or one-fourth per cent, in bridges.

As has been heretofore stated, the type selected for construction is dependent upon traffic conditions, which in turn are indicated in a general way by the number and concentration of motor vehicles. How closely practice has adhered to this may be shown by compar ing numbers of motor vehicles with the percentages of types constructed. This results in the following, the States in which comparison is made being grouped by cars per square mile of area:

The cost per mile of a type varies widely between geographical subdivisions. There is a variation between States in the same geographical subdivision; there is a variation between projects in the same State, and the cost of an individual project would be different depending upon whether contracted for construction this year or last year or next year.

This may be better understood if we give some consideration to the operations that are performed during road construction. The major operations are grading, surfacing, and structure, which have covered approximately 20, 57, and 18 per cent, respectively, of completed Federal-aid projects. In the case of graded and drained roads, the item of grading alone for the whole country has averaged

about 70 per cent, ranging from 80 per cent in the Pacific States where the topography is rough to under 60 per cent in the East South Central and West South Central States where the country is generally undulating. When to the larger grading quantities in the Pacific States you add the greater cost of labor, which during 1925 averaged about 52 cents per hour as against 26 cents per hour for East South Central and West South Central States, it may be expected that graded and drained roads cost more per mile in the Pacific States. Our records show for graded and drained roads that the cost per mile in the Pacific States was 55 per cent higher than in the East South and West South Central States. Variations of this character influence the costs of all types. Not only are common labor wages greatly different in the different sections of the country, but they are different from year to year in the same section. For example, labor in Middle Atlantic States varied from about 30 cents an hour in 1917 to 50 cents an hour in 1920, 33 cents an hour in the early part of 1922, and about 45 cents an hour in 1924 and 1925. Following these variations in labor costs we find the costs of all types rise and fall in the individual States from year to year and vary among States in the same year. The costs of earth excavation in the East North Central States were from 40 cents a cubic yard in 1917 to 80 cents a yard in 1920; dropped back to about 45 cents in 1922 and during 1925 has ranged from 45 to 50 cents a cubic yard.

During the month of September, 1925, bid prices received for earth ●xcavation throughout the country averaged 42 cents a cubic yard and ranged from over $1 in New England to below 25 cents in the West South Central States. The nature and quantity of material to be moved are also important influences on unit cost.

The demand for the improvement of roads is yearly becoming more insistent and the bureau, while sensing the justice and force of this demand, at the same time realizes there are limitations to the amount of money which may be raised to satisfy it. Its research work done largely in cooperation with the State highway departments is directed toward answering the questions. What highway transportation service should be facilitated through the improvement of highways, and how we may best provide such improved highways?"

66

In addition to such other measures as may be taken to make our highway funds go as far as possible, it will undoubtedly be acknowledged that good management in the use of the funds available for construction is absolutely essential. Wherever there is poor management there is waste. Everyone has watched a grading job where wagons are standing in line waiting their turn to secure a load, or a steam shovel or grader is idle waiting for wagons. Again, we have seen a concrete paving job on which the concrete mixer is idle awaiting trucks bringing paving materials. A contractor with equipment broken or in need of repair is endangering his profits. But this is not the whole story. The delay which may be noticed is the result of poor management which is reflected in higher-bid prices. One of the bureau's activities in furtherance of better management is a study of unit construction costs.

I will venture to say that for the more than a billion and a half dollars spent annually on road improvement, at least a third more work could be done than is done by the elimination of waste in time and

effort. This is not mere guess-work but is an approximate estimate based on field studies made on a large number of going construction jobs.

The bureau is collecting information to show the results of poor management and of good management. This information is secured by studying the going work of a large number of road contractors. We first studied grading operations and then paving operations. We shall also study structures. We hope to study all units of road construction work in order to secure unit costs on all of them. We are studying the work of individuals and the work of machines. We study work on which the management is good and work on which it is poor, and hope by showing the contractor wherein he falls behind the jobs on which the management is better than the information we give him will enable him to increase his production and lower his unit costs.

There are a great many more items in the lost-time category than would appear possible at first thought. There are times during working hours where for one reason or another no work at all is done, and during the time when work is actually performed there are widely varying results ranging from excellent progress due to good management to slow progress due to poor management.

Studies have been made of the operations of all kinds of roadgrading equipment-of slip scrapers, fresnos, wheel scrapers, elevato graders, steam shovels, and drag-line excavators. For each grading project one or the other of these pieces of equipment is the most economical to use, and for whichever should be used there is one design which will permit of its most economical use. In general it may be stated that grading costs may be reduced by any general modification in design which will reduce the haul distance.

The generally accepted idea of the usefulness of design is that of an instrument for letting contracts and a guide to be followed in construction. That with a given limiting gradient and rate of curvature and practically without change in the quantity of earth moved, the cost of grading with an elevating-grader outfit may be varied by as much as 20 per cent, and yet the road produced in each case will be practically the same is a fact quite generally lost sight of. Again, variation in design may bankrupt a contractor or give him a profit, although in each case his contract unit price may be the same and the completed road equally useful and economical in the service of highway transport. Unit costs as influenced by design is one of the items we are studying.

From the management standpoint the outstanding problem in a construction job is how to obtain high production. The time required to perform each necessary operation and the degree to which unnecessary operations are eliminated are the determining elements.

The difference between good and average management as measured by the results obtained may be illustrated by the following example: A normal fresno job will show an average haul of perhaps 150 feet. On such a job the average output under average management would be about 73 cubic yards per fresno per day of 10 hours. Under good management it would be about 95 cubic yards per fresno per day. With a 10-fresno outfit the difference in output would be 220 cubic yards per day, which at 25 cents per cubic yard, is $55 per day. In other words, on such a job as this a saving of one

tenth of a minute per load means a saving of from $7.50 to $10 a day for the contractor.

In the case of the elevator grader ordinarily used the conclusion drawn from a careful analysis of all losses is that average production can be readily increased from 700 cubic yards a day to at least 1,500 without change in the present design of grader. By suitable modification of the machine a still greater increase would result.

Data of this sort are being obtained in the field by time studies of contractors' operations. With a mass of data of this character it is possible to show contractors who are not doing well wherein their production is below the average, and by suggestions to enable_them to increase output. By eliminating losses, we have also succeeded in raising production on well managed jobs.

A study of power shovel work in grading lately made indicates that a good three-fourths-yard steam shovel in the hands of a good operator can move a dipper load of loamy clay every 15 seconds, assuming that the dipper swings only one-fourth of a circle. If onehalf swing is necessary, a dipper load should be moved every 21 seconds. On this job the productive shovel time does not exceed 63 per cent. It is apparent that contractors could increase this percentage by giving more thought to the causes responsible for this condition. Shovel operators do not usually get full dipper loads except under the most favorable conditions. The studies show the average load is about one-half the rated capacity of the shovel. They indicate that improper handling of wagons and trucks accounts for a large amount of lost time.

These studies indicate that there are few concrete paving projects on which the average daily output can not be increased 25 per cent. On many jobs the output can be increased 50 to 100 per cent, and this can be accomplished by some reduction in both the amount of labor and equipment employed.

A slow, indifferent, or incompetent mixer operator may easily lose $35 a day on a job with a daily payroll of $200.

Loss of time due to mixer trouble averages about 5 per cent and runs as high as 10 per cent. A contractor with a daily pay roll of $200 may lose from $2,000 to $4,000 a season from this cause alone. On a large number of paving operations studied last year it was found that the lost time during working hours was between 40 I and 45 per cent, and of this about 48 per cent was due to rain and 10 per cent to wet subgrade following rains; 10 per cent was due to delay in the preparation of subgrades, 22 per cent to mixer trouble, 62 per cent to lack of material for the mixer, a little over 9 per cent was consumed in moving, 212 per cent due to trouble at the material or batcher plant, 3 per cent to cold weather, and the remainder to miscellaneous sources.

On quite a number of the paving jobs studied an attempt was made to increase production. On one of these stop-watch readings were taken for a week, and the oufit was found to be operating at an efficiency of 37 per cent.

Mr. ROBSION. Did you say that on a test case it was found to be only 37 per cent efficient?

Mr. ALLEN. Yes, sir; on this individual case.

During the month that followed suggestions were made to the contractor and to the men on the job in an effort to eliminate delays

that were cutting down production. The mixer operator was shown the correct way of operating the mixer. The subgrade was prepared in such a way that it would not interfere with mixer operations. Changes were made in the hauling operations. The mixer efficiency was increased to 65 per cent and the hauling efficiency of the available trucks was increased above 90 per cent. The production of this outfit might have been increased still more had the contractor been willing to cooperate to the extent of furnishing a sufficient supply of trucks.

Mr. ROBSION. Do we let contracts under such favorable conditions that any concern can get a contract and operate at 37 per cent efficiency and keep from failing?

Mr. ALLEN. They seem to do it, somehow. They do not always keep from failing.

Mr. ROBSION. Does that not seem to be a very favorable contract, if that could be done?

Mr. ALLEN. It appears that the average contractor does not operate above about 50 per cent efficiency, but the bids made and the contracts entered into are at a sufficiently high figure and enable money to be made.

On this project the labor cost per square yard of pavement was reduced from an average of 45 cents for the month before sugges tions were made to an average of 27 cents during the month that suggestions were being made. During the last week of the study the labor cost was reduced to 20 cents per square yard.

On another project where the management was studied, another attempt was made to increase production. Stop-watch studies of mixer operation were made over a period of 17 days, and showed the mixer operation of this outfit to be 61 per cent efficient. Much time was also being wasted in hauling operations. During the month that followed assistance was given this outfit in eliminating the delays that were retarding production. The mixer operator soon learned the correct method of operating the mixer and no time was lost at this point. All necessary delays on the road were reduced to a minimum. Complete charge was taken of the hauling outfit, and its efficiency was greatly increased. The mechanical condition of the trucks was improved and the trucks were properly handled by the drivers. Many other changes were made which tended to increase production. The paving contractor's labor cost per square yard of pavement that had averaged 20.7 cents for the two months prior to the time that these studies were started to increase production, was reduced to an average of 14 cents per square yard during the month that suggestions were being made and to an average of 12.8 cents during the last week of the study. The hauling contractor's cost for truck drivers per square yard of pavement was reduced for the different lengths of haul. The efficiency of the oufit was gradually increased until it reached 97 per cent during the latter part of the study. The record run for this outfit was 1263 lineal feet of 18-foot pavement laid in 12.5 continuous hours. Many other studies have been made and with similar results.

It is the intention of the bureau to continue these unit cost-construction studies. Every item will be given intensive study. Stand