---

CHAPTER VIII
PROBLEMS OF ROUTING

The general acceptance, both in practice and theory, of the principle that distance is a relatively unimportant element in rate making[249] is significant at this time, in connection with the recent amendment of the Act to Regulate Commerce. It is important also because of the marked tendency toward the adoption by various state legislatures of the extreme opposite principle of a rigid distance tariff. The old problem of effecting a compromise between these two extreme theories by some form of long and short haul clause—the original section 4 of the act of 1887 having been emasculated by judicial interpretation—is again brought to the front. For these reasons it may be worth while to consider certain results which inevitably follow the widespread acceptance of this principle of the blanket rate. Its benefits are indeed certain; namely, an enlargement of the field of competition, and an equalization of prices over large areas, and that too at the level of the lowest or most efficient production. But these advantages entail certain consequences—of minor importance, perhaps, but none the less deserving of notice.

The subordination of distance to other factors in rate making is a logical derivation from the theory of joint cost. This theory justifies the classification of freight, namely, a wide range of rates nicely adjusted to what the traffic in each particular commodity will bear, while always allowing each to contribute something toward fixed and joint expenses. In the same way it explains a close correlation of the distance charge to what each commodity will bear. It assumes that any rate, however low, which will yield a surplus over expenses directly incidental to the increment of traffic and which thus contributes something toward indivisible joint costs, serves not only the carrier by increasing his gross revenue, but at the same time lightens the burden of fixed expenses upon the balance of the traffic. This principle of joint cost, so clearly set forth by Professor Taussig,[250] is fundamental and comprehensive. It pervades every detail of rate making. But it rests upon two basic assumptions which, while generally valid, are not universally so. In the first place each increment of traffic must be new business, not tonnage wrested from another carrier and offset by a loss of other business to that competitor. And secondly, each increment of traffic must be economically suitable to the particular carriage in contemplation.

The first of these assumptions fails wherever two carriers mutually invade each other's fields or traffic. Each is accepting business at a virtual loss, all costs including fixed charges on capital being taken into account, in order to secure the increment of business. Each gain is offset by a corresponding loss. It is the familiar case of the rate war. A less familiar aspect of the matter is presented when traffic is disadvantageously carried by two competing roads, each diverting business from its natural course over the other's line. The sum total of traffic is not increased. Each carries only as much as before but transports its quota at an abnormal cost to itself. This may, perhaps, swell gross revenues; but by no process of legerdemain can the two losses in operating cost produce a gain of net revenue to both. And each increase of unnatural tonnage, where offset by a loss of natural business, instead of serving to lighten the fixed charges, becomes a dead weight upon all the remaining traffic. The commonest exemplification of this is found in the circuitous transportation of goods, instances of which will be given later.

The second case in which the principle of joint cost fails to justify charges fixed according to what the traffic will bear may arise in the invasion of two remote markets by one another; or, as it might be more aptly phrased, in the overlapping of two distant markets. A railroad is simultaneously transporting goods of like quality in opposite directions. Chicago is selling standard hardware in New York, while New York is doing the same thing in Chicago. Prices are the same in both markets. Of course if the two grades of hardware are of unequal quality, or if they are like goods produced at different cost, an entirely distinct phase of territorial competition is created. But we are assuming that these are standard goods and that there are no such differences either in quality or efficiency of production. What is the result? Is each increment of business to the railroad a gain to it and to the community? The goods being produced at equal cost in both places, the transportation charge must be deducted from profits. For it is obvious that the selling price cannot be much enhanced. The level of what the traffic will bear is determined not, as usual, by the value of the goods but by other considerations. The traffic will bear relatively little, no matter how high its grade. The result is that the carrier, in order to secure the tonnage, must accept it at a very low rate, despite the length of the haul.

This is the familiar case of the special or commodity rate granted to build up business in a distant market. Special rates confessedly form three-fourths of the tonnage of American railways, as has already been said. The assumption is usually made that such traffic is a gain to the railways, justified on the principle of joint cost as already explained. But does it really hold good in our hypothetical case? There is a gain of traffic in both directions, to be sure. But must it not be accepted at so low a rate that it falls perilously near the actual operating cost? It is possible that even here it may add something to the carriers' revenue, and thereby lighten the joint costs in other directions. But how about the community and the shipping producers? Are any more goods sold? Perhaps the widened market may stimulate competition, unless that is already keen enough among local producers in each district by itself. The net result would seem to be merely that the railroads' gain is the shippers' loss. There is no addition to, but merely an exchange of, place values. Both producers are doing business at an abnormal distance under mutually disadvantageous circumstances. It may be said, perhaps, that the situation will soon correct itself. If the freight rates reduce profits, each group of producers will tend to draw back from the distant field. This undoubtedly happens in many cases. But the influence of the railway is antagonistic to such withdrawal. It is the railway's business to widen, not to restrict, the area of markets. "The more they scatter the better it is for the railroads." "Keep everyone in business everywhere." And if necessary to give a fillip to languishing competition, do so by a concession in rates. Is there not danger that with a host of eager freight solicitors in the field, and equally ambitious traffic managers in command, a good thing may be overdone, to the disadvantage of the railway, the shippers and the consuming public?

An objection to this chain of reasoning arises at this point. Why need the public or other shippers be concerned about the railways' policy in this regard? Is not each railway the best judge for itself of the profitableness of long-distance traffic? Will it not roughly assign limits to its own activities in extending business, refusing to make rates lower than the actual incidental cost of operation? And are not all low long-distance rates, in so far as they contribute something toward joint cost, an aid to the short haul traffic? The answer will in a measure depend upon our choice between two main lines of policy; the one seeking to lower average rates, even at the expense of increasing divergence between the intermediate and the long distance points, the other policy seeking, not so much lower rates as less discriminatory rates between near and distant points. In the constant pressure for reduced rates in order to widen markets it is not unnatural that the intermediate points, less competitive probably, should be made to contribute an undue share to the fixed sum of joint costs. The common complaint today is not of high rates but of relative inequalities as between places. It is a truism to assert that it matters less to a shipping point what rate it pays than that its rate, however high, should be the same for all competing places. This immediately forces us to consider the consumer. What is the effect upon the general level of prices of the American policy of making an extended market the touchstone of success, irrespective of the danger of wastes arising from overlapping markets? That the result may be a general tax upon production is a conclusion with which we shall have later to do. Such a tax, if it exists, would go far to offset the profit which unduly low freight rates in general have produced. In short, the problem is to consider the possible net cost to the American people of our highly involved and most efficient transportation system. Our markets are so wide, and our distances so vast, that the problem is a peculiarly American one.

Having stated the theory of these economic wastes, we may now proceed to consider them as they arise in practice. Concrete illustration of the effect of disregard of distance naturally falls into two distinct groups. Of these the first concerns the circuitous carriage of goods; the second, their transportation for excessive distances. Both alike involve economic wastes, in some degree perhaps inevitable, but none the less deserving of evaluation. And both practices, even if defensible at times, are exposed to constant danger of excess. It will be convenient also to differentiate sharply the all-rail carriage from the combined rail and water transportation. For as between railroads and waterways, the difference in cost of service is so uncertain and fluctuating that comparisons on the basis of mere distance have little value.

Recent instances of wasteful and circuitous all-rail transportation are abundant. A few typical ones will suffice to show how common the evil is. President Ramsay of the Wabash has testified as to the roundabout competition with the Pennsylvania Railroad between Philadelphia and Pittsburg by which sometimes as much as fifty-seven per cent. of traffic between those two points may be diverted from the direct route. "They haul freight 700 miles around sometimes to meet a point in competition 200 miles away."[251] Chicago and New Orleans are 912 miles apart, and about equally distant—2,500 miles—from San Francisco. The traffic manager of the Illinois Central states that that company "engages in San Francisco business directly via New Orleans from the Chicago territory, and there is a large amount of that business, and we engage in it right along."[252] Wool from Idaho and Wyoming may move west 800 miles, to San Francisco; and thence via New Orleans over the Southern Pacific route to Boston.[253] This case, therefore, represents a superfluous lateral haul of nearly a thousand miles between two points 2,500 miles apart. The Canadian Pacific used to take business for San Francisco, all rail, from points as far south as Tennessee and Arkansas, diverting it from the direct way via Kansas City.[254]

Goods moving in the opposite direction from San Francisco have been hauled to Omaha by way of Winnipeg, journeying around three sides of a rectangle by so doing, in order to save five or six cents per hundred pounds.[255] Between New York and New Orleans nearly one hundred all-rail lines may compete for business. The direct route being 1,340 miles, goods may be carried 2,051 miles via Buffalo, New Haven (Indiana), St. Louis and Texarkana.[256] A generation ago conditions were even worse, the various distances by competitive routes between St. Louis and Atlanta ranging from 526 to 1,855 miles.[257] New York business for the West was often carried by boat to the mouth of the Connecticut river, and thence by rail over the Central Vermont to a connection with the Grand Trunk for Chicago. To be moved at the outset due north 200 miles from New York on a journey to a point—Montgomery, Alabama—south of southwest seems wasteful; yet the New York Central is in the field for that business.[258] The map herewith, prepared in connection with the Alabama Midland case, shows the number of lines participating in freight carriage between New York and the little town of Troy, Alabama. It is nearly as uneconomical as in the old days when freight was carried from Cincinnati to Atlanta via the Chesapeake and Ohio, thence down by rail to Augusta and back to destination.[259] It was common for freight from Pittsburg to go by boat down to Cincinnati, only to return by rail via Pittsburg to New York at a lower rate than on a direct shipment.[260] Even right in the heart of eastern trunk line territory, such things occur in recent times. The Cincinnati, Hamilton and Dayton prior to its consolidation with the Pere Marquette divided its eastbound tonnage from the rich territory about Cincinnati among the trunk lines naturally tributary. But no sooner was it consolidated with the Michigan road than its eastbound freight was diverted to the north—first hauled to Toledo, Detroit and even up to Port Huron, thence moving east and around Lake Erie to Buffalo.[261] In the Chicago field similar practices occur. Formerly the Northwestern road was charged with making shipments from Chicago to Sioux City via St. Paul. This required a carriage of 670 miles between points only 536 miles apart; and the complaint arose that the roundabout rate was cheaper than the rate by the direct routes. I am privately informed that the Wisconsin Central at present makes rates between these same points in conjunction with the Great Northern, the excess distance over the direct route being 283 miles. Complaints before the Elkins Committee[262] are not widely different in character. Thus it appears that traffic is hauled from Chicago to Des Moines by way of Fort Dodge at lower rates than it is carried direct by the Rock Island road, despite the fact that Fort Dodge is eighty miles north and a little west of Des Moines. The Illinois Central, having no line to Des Moines, pro-rates with the Minneapolis and St. Louis, the two forming two sides of a triangular haul. An interesting suggestion of the volume of this indirect routing is afforded by the statistics of merchandise shipped between American points which passes through Canada in bond.[263] The evidence of economic waste is conclusive.

A common form of wastefulness in transportation arises when freight from a point intermediate between two termini is hauled to either one by way of the other. Such cases are scattered throughout our railroad history. One of the delegates to the Illinois Constitutional Convention of 1870, cites, as an instance of local discrimination, the fact that lumber from Chicago to Springfield, Illinois, could be shipped more cheaply by way of St. Louis than by the direct route.[264] And now a generation later, it appears that grain from Cannon Falls, forty-nine miles south of St. Paul on the direct line to Chicago, destined for Louisville, Kentucky, can be hauled up to St. Paul on local rates and thence on a through billing to destination, back over the same rails, considerably cheaper than by sending it as it should properly go.[265] The Hepburn Committee reveals shipments from Rochester, New York, to St. Louis, Minneapolis or California, all rail, on a combination of local rates to New York and thence to destination.[266] Presumably the freight was hauled three hundred miles due east and then retraced the same distance; as New York freight for southern California is today hauled to San Francisco by the Southern Pacific and then perhaps three hundred miles back over the same rails. Even if the rate must be based on a combination of low through rates and higher local rates, it seems a waste of energy to continue the five or six hundred miles extra haul. Yet the practice is common in the entire western territory. From New York to Salt Lake City by way of San Francisco is another instance in point.[267] Of course a short haul to a terminal to enable through trains to be made up presents an entirely different problem of cost from the abnormal instances above mentioned.[268]

Carriage by water is so much cheaper and as compared with land transportation is subject to such different rate-governing principles, that it deserves separate consideration. Mere distance, as has already been said, being really only one element in the determination of cost, a circuitous water route may in reality be more economical than direct carriage overland. Yet beyond a certain point, regard being paid to the relative cost per mile of the two modes of transport, water-borne traffic may entail economic wastes not incomparable to those arising in land transportation. In international trade, entirely confined to vessel carriage, a few examples will suffice for illustration. Machinery for a stamp mill, it was found, could be shipped from Chicago to San Francisco by way of Shanghai, China, for fifteen cents per hundredweight less than by way of the economically proper route. Were the goods ever really sent by so indirect a route?[269] It would appear so when wheat may profitably be carried from San Francisco to Watertown, Massachusetts, after having been taken to Liverpool, stored there, reshipped to Boston, thereafter, even paying the charges of a local haul of nearly ten miles;[270] or when shipments from Liverpool to New York may be made via Montreal to Chicago, and thence back to destination.[271] I am credibly informed that shipments of the American Tobacco Company from Louisville, Kentucky, to Japan used commonly to go via Boston. Denver testimony is to the effect that machinery, made in Colorado, shipped to Sydney, Australia, can be transported via Chicago for one-half the rate for the direct shipment; and that on similar goods even Kansas City could ship by the carload considerably cheaper by the same roundabout route. Conversely straw matting from Yokohama to Denver direct must pay $2.87 per hundred pounds; while if shipped to the Missouri river, five hundred miles east of Denver, and then back, the rate is only $2.05.[272]

As a domestic problem, water carriage confined to our own territory has greater significance in the present inquiry. Purely coastwise traffic conditions are peculiar and in the United States, as a rule, concern either the South Atlantic seaports or transcontinental business. As to the first-named class, the volume and importance of the traffic is immense. Its character may be indicated by a quotation from a railroad man.

Allowing a constructive mileage of one-third for the last named water haul,[274] many of these even up fairly well with the all-rail carriage; although a route from New York to Kansas City by way of Savannah, Georgia, would appear to be an extreme case, owing to the relatively long haul by rail.[275] The increasing importance of Galveston and the necessity of a back haul to compensate for export business make it possible for that city to engage in business between New York and Kansas City, although the roundabout route is two and one-half times as long as the direct one.[276] As compared with these examples, it is no wonder that the competition for New York-Nashville or New England-Chattanooga business by way of Savannah, Mobile, or Brunswick, Georgia, is so bitter. The roundabout traffic thus reaches around by the southern ports and nearly up again to the Ohio river.[277]

The second great class of broken rail and water shipments consists of transcontinental business. Goods from New York to San Francisco commonly go by way of New Orleans or Galveston,[278] as well as by Canadian ports and routes.[279] In the opposite direction, goods are carried about 1000 miles by water to Seattle or Vancouver before commencing the journey east. But more important, as illustrating this point, is the traffic from the Central West which reaches the Pacific coast by way of Atlantic seaports. As far west as the Missouri, the actual competition of the trunk lines on California business has since 1894[280] brought about the condition of the "blanket" or "postage stamp" rate. The same competitive conditions which open up Denver or Kansas City to New York shippers by way of New Orleans or Galveston, enable the Southern Pacific Railroad or Cape Horn routes to solicit California shipments in western territory to be hauled back to New York, and thence by water all or part of the way to destination. How important this potential competition is—that is to say, what proportion of the traffic is interchanged by this route—cannot readily be determined.

Transportation over undue distances—the carriage of coals to Newcastle in exchange for cotton piece goods hauled to Lancashire—as a product of keen commercial competition may involve both a waste of energy and an enhancement of prices in a manner seldom appreciated. The transportation of goods great distances at low rates, while economically justifiable in opening up new channels of business, becomes wasteful the moment such carriage, instead of creating new business, merely brings about an exchange between widely separated markets, or an invasion of fields naturally tributary to other centres. The wider the market, the greater is the chance of the most efficient production at the lowest cost. The analogy at this point to the problem of protective tariff legislation is obvious. For a country to dispose of its surplus products abroad by cutting prices may not involve economic loss; but for two countries to be simultaneously engaged in "dumping" their products into each other's markets is quite a different matter. In transportation such cases arise whenever a community, producing a surplus of a given commodity, supplies itself, nevertheless, with that same commodity from a distant market. It may not be a just grievance that Iowa, a great cattle raising state, should be forced to procure her dressed meats in Chicago or Omaha;[281] for in this case some degree of manufacture has ensued in these highly specialized centres. But the practice is less defensible where the identical product is redistributed after long carriage to and from a distant point. Arkansas is a great fruit raising region; yet so cheap is transportation that dried fruits, perhaps of its own growing, are distributed by wholesale grocers in Chicago throughout its territory. The privilege of selling rice in the rice-growing states from Chicago is, however, denied by the Southern Railway Association.[282] An illuminating example of similar character occurs in the Southern cotton manufacture, as described by a Chicago jobber:

The overweening desire of the large centres to enter every market is well exemplified by recent testimony of the Chicago jobbers.[284]

All this record implies progressiveness, energy, and ambition on the part of both business men and traffic officers. Nothing is more remarkable in American commerce than its freedom from restraints. Elasticity and quick adaptation to the exigencies of business are peculiarities of American railroad operation. This is due to the progressiveness of our railway managers in seeking constantly to develop new territory and build up business. The strongest contrast between Europe and the United States lies in this fact. European railroads take business as they find it. Our railroads make it. Far be it from me to minimize the service rendered in American progress. And yet there are reasonable limits to all good things. We ought to reckon the price which must be paid for this freedom of trade.

One further aspect of economic waste may be mentioned, especially as bearing upon Federal regulation so far as it affects carload ratings and commercial rivalry between remote middlemen in the large cities and provincial jobbing interests. The actual cost of handling small shipments being about one-half that of carriage by carloads, the cheapest way in which to supply, let us say, the Pacific slope or Texas territory, is to encourage the local jobber who ships by carload over the long haul. For, obviously, distribution by less-than-carload lots from New York, or even Chicago direct, direct to the cross-road store, is bound to be a wasteful process by comparison.[285] But in addition there are also, of course, the social factors to be considered, which are of even greater weight.

The causes of economic waste in transportation are various. Not less than six may be distinguished. These are: (1) congestion of the direct route; (2) rate cutting by the weak circuitous line; (3) pro-rating practices in division of joint through rates; (4) desire for back-loading of empty cars; (5) strategic considerations concerning interchange of traffic with connections; and (6) attempts to secure or hold shippers in contested markets. These merit consideration separately in some detail.

Congestion of traffic upon the direct line is a rare condition in our American experience. Few of our railways are over-crowded with business. Their equipment may be overtaxed, but their rails are seldom worked to the utmost. Yet the phenomenal development of trunk line business since 1897 sometimes makes delivery so slow and uncertain that shippers prefer to patronize railways less advantageously located, even at the same rates. The congestion on the main stem of the Pennsylvania railway between Pittsburg and Philadelphia is a case in point.

Special rates or rebates often divert traffic. The weak lines, in that particular business, are persistently in the field and can secure tonnage only by means of concessions from what may be called the standard or normal rate. The differential rate is an outgrowth of this condition. The present controversy over the right of the initial line in transcontinental business to route the freight at will involves such practices. The carriers insist that they can stop the evil only by the exercise of choice in their connections. An interesting recent example is found in the Elkins Committee testimony. It appears that lumber from points in Mississippi destined for Cleveland instead of going by the proper Ohio river gateways was diverted to East St. Louis. The operation was concealed by billing it to obscure points,—Jewett, Ill., near East St. Louis, and Rochester, Ohio,—and there issuing a new bill of lading to destination:

Wherever a large volume of traffic is moving by an unnatural route, the first explanation which arises therefore is that rebates or rate-cutting are taking place.[287]

A third cause of diversion of traffic is akin to the second; and concerns the practices in pro-rating. Much circuitous transportation is due to the existence of independent transverse lines of railway which may participate in the traffic only on condition that it move by an indirect route. This situation is best described by reference to the following diagram. Let us suppose traffic to be moving by two routes passing through points B and C, and converging on A, which last-named point might be Chicago, St. Louis, New York or any other railroad centre. Cutting these two converging lines of railway, we will suppose a tranverse line passing through B and C. Obviously the proper function of this railway is as a feeder for the through lines, each being entitled to traffic up to the half-way point, D. But over and above serving as a mere branch, this road, desirous of extending its business, has a powerful incentive to extend operations. The longer the tranverse haul, the greater becomes its pro-rating division of the through rate with the main line. Traffic from C is of no profit to the tranverse road so long as it is hauled directly to A. But if hauled from C to the same destination by way of B, the profit may be enhanced in two ways. In the first place the pro-rating distance is greater; and secondly, such traffic from C not being naturally tributary to the main line B A but merely a surplus freight to be added to that already in hand, the main line A B is open to temptation to shrink its usual proportion of the through rate in order to secure the extra business. This same motive may on proper solicitation induce the other main line C A to accept traffic from B and its vicinity. The result is a greatly enhanced profit to the cross line and circuitous carriage of the goods in both directions around two sides of a triangle. Only recently in a case in Texas the Interstate Commerce Commission found that two roads thus converging on a common point were each losing to the other traffic which rightfully was tributary to its own line. In a recent case, ninety-nine per cent. of the business from Chatham to New York was moving over a route 249 miles long, when it might have gone directly only 144 miles, by pro-rating with another road.[288] Our illustrative examples are not fanciful in any degree.[289]

This roundabout carriage becomes of course increasingly wasteful in proportion to the width of angle between the main lines converging on the common point. And several cases indicate that in extreme instances the two main lines may converge on a common point from exactly opposite directions, while the transverse or secondary road or series of roads forms a wide and roundabout detour. The well known Pittsburg-Youngstown case, cited in the original Louisville & Nashville decision in 1887, serves as illustration. The Pennsylvania was competing from Pittsburg directly eastbound to New York with certain feeders of the New York Central lines which took out traffic bound for the same destination but leaving Pittsburg westbound.[290] Other instances of the same phenomenon occur at Chattanooga, where freight for New York may leave either northward or southward, at Kansas City and in fact at almost any important inland centre.

Another extreme form may arise even in the competition between two parallel trunk lines cut transversely by two independent cross roads. One of these latter may induce traffic to desert the direct route, to cut across to the other trunk line, to move over that some distance and then to be hauled back again to a point on the first main line where it may find a "cut" rate to destination. Grain sometimes used literally to meander to the seaboard in the days of active competition between the trunk lines. Wheat from Iowa and northern Illinois finally reached Portland, Maine, by way of Cincinnati in this manner, with a superfluous carriage of from 250 to 350 miles:

Another witness describes the route as follows:

This ruinous diversion of freight seems to have been dependent upon the existence of active competition at Detroit and ceased when the Grand Trunk came to an agreement with the American lines. But there can be no doubt that wherever these cross lines exist there is a strong tendency toward diversion. In the recent hearings of the Senate Committee on Interstate Commerce on railway rate regulation, a railroad witness again describes the operation: