utilities which are the subject of exchange, should be regulated by cost of production. This, it is asserted, is the only just law. But such a conclusion is of doubtful validity. Even granting that cost of production forms a just basis for prices, no one claims that actual business prices of each particular commodity vary with the cost. The application to railway rates is plain. The cost of service theory might logically demand that the sum total of charges should vary with the cost, but not that the price of each individual transaction should be fixed by its cost of service. Even were this practicable, which we have seen is not the case, it would not be theoretically defensible.

The principle of value of service has a firmer foundation. Railway charges cannot, indeed, be fixed like prices in general, simply by demand and supply. This is the mistake of the railway officials who attempt to justify all charges.1 Railway transportation is more than a simple business; it is a semi-public occupation, a public trust. Hence the necessity of restricting the inequalities of every-day commercial practice. But to oppose the abuses of a principle is quite another thing from demurring to the principle itself. The value-of-service doctrine, correctly understood, simply applies the methods already followed in certain public relations. It fixes charges according to the ability to pay the same principle that is recognized in taxation. Charging what the traffic will bear is a rough way cf stating that the charges are proportioned to the capacity or ability of the articles that compose the traffic. It will not be questioned that the endeavors to develop traffic can be realized only by making lower charges for the cheaper (and distant) goods. But the element of justice is introduced as soon as we show that such a method graduates charges according to ability. Of course it does not follow that all rates actually charged are just rates. The inference simply is that the principle of value, as a principle, is not open to the objections often

1 So de la Gournerie, Études économiques sur l'exploitation des chemins de fer (1880), pp. 118, 119; Grierson, Railway Rates (1886), p. 68; and most of the American writers.

urged. The ability of an article to pay, its capacity to contribute to the payment of the expenses, is an undeniably valid basis for rates. As it is well expressed by the Cullom Committee:

The capacity of each commodity to contribute to the payment of the fixed charges is measured by the extent to which the cost of its transportation fixes its market value and determines the question of its movement. In the case of commodities like coal, stone, ore, beef, corn, lumber, etc., the freight charge constitutes the principal item of cost to the consumer; however small may be his contribution to the general burden, it is relatively greater than that made by the consumer of high-priced articles, such as clothing or dry goods, e.g., the selling-price of which is not appreciably affected by the freight charges, even though unreasonably high.1

And what is true of the cheap goods is true of the distant goods. For the purposes of transportation they stand on precisely the same footing and are subject to the same conditions. The principle again applies equally well to passenger traffic. Even in the United States there are virtually different classes, and the higher fares for the better service may be upheld on the principle that the passengers in the higher class cars possess more ability to pay large fares than those in ordinary or emigrant cars. The value of service principle is based on supposed ability to pay.

But now the difficult question arises. We have shown that the low-priced wares possess less ability to pay than the highpriced wares. Does it follow that the more valuable wares, by reason of their greater ability to pay, should be charged higher rates than the average, or than would otherwise be the case, in order to compensate for the lower rates of the cheaper goods? Does the principle of value imply this compensating action, and is this principle of compensation valid and just? This is the puzzling question. To give a precise answer is not so simple as it might appear. We may, indeed, assert with decision that difference in value implies a difference in ability to pay, but it is rather arbitrary and hazardous to assert exactly what relation

1 Report, p. 185.

there is between value and ability. Shall an article of double the value pay twice the rate; and if not, why not? The difficulty, in fact, is exactly the same that is met with in the problems of taxation. One of the fundamental principles of equity in taxation is that contributors should pay taxes in proportion to their ability. A rich man ought to pay more than a poor man; the difficulty arises when we must determine exactly how much more he ought to pay. Is the difference of ability proportional to their property, or to their income, or to their expenses? Or, again, should the difference of ability be measured not by a proportional, but by a progressive, scale of taxation, — should there be a progressive property, or a progressive income, or a progressive expense tax, rather than a proportional tax? None of these questions can be declared definitely settled by the science of finance. The answers are necessarily vague because of the relativity in the test of ability.

Exactly the same considerations are applicable to railway tariffs. Difference in value implies difference in ability. The cheap articles possess less ability than the dear goods, and should thus pay lower rates. But to determine how much higher rates the others should pay is not a self-evident proposition. The question is a relative one, and the rates may vary within wide limits. It is precisely because the question is a relative one that the many abuses of railway management have arisen. This relativity, the possibility of making undue differences within the limits of the just principle, becomes therefore a strong argument in favor of some form of public regulation.1 The unhampered railway management may pursue the correct policy of charging what they think the service is worth, but their opinions may vary within wide limits. There is, in other words, such a possible elasticity or flexibility in the methods of fixing the details that the actual charges may be far from adequately satisfying all demands. This fact above all others has earned for the doctrine of charging what the traffic will bear the deserved

1 Cohn, Untersuchungen über die englische Eisenbahnpolitik, Bd. iii (1883), S. 84, concludes that the railways must therefore be owned and managed by the state. But such a conclusion is not at all necessary.

epithet of "hap-hazard" estimate.1 The doctrine of free competition and uncontrolled liberty does not follow from the principle of value of service as the foundation of railway tariffs.

But at all events one point has been gained. The principle of value, within these wide limits, is a principle which not only does determine railway tariffs, but which, although liable to abuse, is a correct principle. It is just because it is founded on the principle of ability. It is neither new nor "outrageous." It is not only a just principle, but, as we have seen, the only practicable principle. The cost of service doctrine can no longer put forth the exclusive claim of justice as the basis of railway tariffs.

One exaggeration, however, must be avoided. The principle of value, we said, implies the doctrine of compensation. But this does not imply that the higher charges on the dear goods or local traffic are higher than they would be if there were no lower charges on the cheap goods or through traffic. Were the rates on the cheap or long-distance traffic to be raised, it would not be transported at all; and since its contribution to the fixed expenses would fall away, the whole expense would necessarily be borne by the dear and short-distance traffic. The rates on the latter would have to be increased to make good for the loss of the former; the dear and local freight would pay even more than it pays now. Those who object to the principle of value because it unduly raises the charges on high-class and local business thus utterly fail to perceive that in many cases it produces just the contrary effect. The principle of value often lowers the rates on the dear goods, and renders possible the transportation of the cheap goods. It is the long-distance traffic which has enabled the American railways to reduce their charges, through as well as local, far below the level of European tariffs. Pro rata charges, or even rates based solely on cost of service, would give us tariffs much higher than those in actual use; they would level up, not level down.

Classification and differential rates are thus legitimate and necessary expedients: legitimate, because based on value; neces

1 Sir B. Samuelson, Report on Railway Goods Tariffs, etc. (1886), p. 20.

sary, because without them railway transportation would become vastly less efficient. The same analysis would show the illegiti, macy of personal discriminations, even in wholesale transactions. Classification and local discrimination reduce rates for the traffic which is less able to pay; personal discrimination reduces rates for the traffic which is better able to pay. Reduced rates to large shippers increase the advantages of the strong; rates fixed according to value tend to diminish the disadvantages of the weak.] Preferential rates are wrong because not based on the principle of value; differential rates are right because following the doctrine of value. Preferential rates invert the considerations of ability; differential rates maintain the principle of ability.

But we must not be misunderstood. While the principle of charging what the traffic will bear is essentially just and legiti mate from the railway standpoint, from the public standpoint it must be regarded as a subordinate principle. Value of service puts into the hands of the railways practically a power of taxation. It is indeed not entirely an arbitrary power, since the charges are partially regulated by water competition. But in its essence it is a power of taxation—a taxation often cunningly masked in the methods of classification and discrimination. From the public standpoint we maintain the great principle of equal treatment for all persons and all business. This is the general rule; the principle of value must be viewed as a legitimate qualification of the rule of equal treatment. But it must be shown in every particular case that the service is of varying value. From the public standpoint in other words the burden of proof must rest on the railways. Charging what the traffic will bear is just, but its application is so elastic that the justice must be demonstrated in each instance. To leave the application of the principle to the discretion of the railway results in the chaotic, almost barbaric, condition of actual charges during railway wars. The only rational method to reconcile public and private interests is to lay down the rule of equal treatment for all persons and places, and to admit the principle of value as a necessary infraction of the rule. But the necessity of the in