Henry George, Protection or Free Trade:

An Examination of the Tariff Question, with especial Regard to
the Interests of Free Trade [1886]

Edition used:
Protection or Free Trade, An Examination of the Tariff Question, with especial Regard to the Interests of Free Trade (New York: Doubleday, Page, and Co., 1905).

Chapter IX

PROTECTIVE tariffs differ from revenue tariffs in their object, which is not so much that of obtaining revenue as that of protecting home producers from the competition of imported commodities.

The two objects, revenue and protection, are not merely distinct, but antagonistic. The same duty may raise some revenue and give some protection, but, past a certain point at least, in proportion as one object is secured the other is sacrificed, since revenue depends on the bringing in of commodities; protection on keeping them out. So the same tariff may embrace both protective and revenue duties, but while the protective duties lessen its power of collecting revenue, the revenue duties by adding to the cost of home production lessen its power of encouraging home producers. The duties of a purely revenue tariff should fall only on commodities not produced in the country; or, if levied on commodities partly produced at home should be balanced by equivalent internal taxes to prevent incidental protection. In a purely protective tariff, on the other hand, commodities not produced in the country should be free and duties should be levied on commodities that are or may be produced in the country. And, just in proportion as it accomplishes its object, the less revenue will it yield. The tariff of Great Britain is an example of a purely revenue tariff, incidental protection being prevented by excise duties. There is no example of a purely protective tariff, the purpose of obtaining revenue seeming always to be the original stock upon which protective features are grafted. The tariff of the United States, like all actual protective tariffs, is partly revenue and partly protective, its original purpose of yielding revenue having been subordinated to that of giving protection, until it may now be best described as a protective tariff yielding incidental revenue.

As we have already considered the revenue functions of tariffs, let us now consider their protective functions.

Protection, as the word has come to be used to denote a scheme of national policy, signifies the levying of duties on the importation of commodities (as a means) in order (as an end) to encourage domestic industry.

Now, when the means proposed in any such scheme is the only means by which the proposed end can be reached, it is only needful to inquire as to the desirability of the end; but when the proposed means is only one of various means we must satisfy ourselves that it is the best. If it is not, the scheme is condemned irrespective of the goodness of its end. Thus the advisability of protection does not, as is generally assumed, follow the admission of the advisability of encouraging domestic industry. That granted, the advisability of protection is still an open question, since it is clear that there are other ways of encouraging home industry than by import duties.

Instead of levying import duties, we might, for instance, destroy a certain proportion of imported commodities, or require the ships bringing them to sail so many times round the world before landing at our ports. In either of these ways precisely the same protective effect could be secured as by import duties, and in cases where duties secure full protection by preventing importation, such methods would involve no more waste. Or, instead of indirectly encouraging domestic producers by levying duties on foreign goods, we might directly encourage them by paying them bounties.

As a means of encouraging domestic industry the bounty has over the protective system all the advantages that the system of paying public officers fixed salaries has over the system prevailing in some countries, and in some instances in our own, of letting them make what they can. As by paying fixed salaries we can get officials at such places and to perform such functions as we wish, while under the make-what-youcan system they can only be got at places and in capacities that will enable them to pay themselves, so do bounties permit the encouragement of any industry, while protection permits only the encouragement of the comparatively few industries with which imported commodities compete. As salaries enable us to know what we are paying, to proportion the rewards of different offices to their respective dignity, responsibility and arduousness, while make-what-you-can may give to one official much more than is necessary, and to others not enough, so do bounties enable us to see and to fix the encouragement to each industry, while the protective system leaves the public in the dark and makes the encouragement to each industry almost a matter of chance. And as salaries impose on the people much lighter and more fairly apportioned burdens than does the make-what-you-can system, so is the difference between bounties and protection.

To illustrate the working of the two systems, let it be assumed desirable to encourage aerial navigation at public expense. Under the bounty system we should offer premiums for the building and successful operation of air ships. Under the protective system we should impose deterrent taxes on all existing methods of transportation. In the one case we should have nothing to pay till we got what we wanted, and would then pay a definite sum which would fall on individuals and localities in general taxes. But in the other case we should have to suffer all the inconveniences of obstructed transportation before we got air ships, and whether we got them or not; and while these obstructions would, in some cases, more seriously affect individuals, businesses and localities than in others, we should never be able to tell how much they distorted industry and cost the people, or how much they stimulated the invention and building of air ships. In the one case, moreover, after aerial navigation had proved successful, and the stipulated bounties had been paid, the air-ship men would hardly have the audacity to ask for more bounties, and would not be likely to get them if they did. In the other case, the public would have grown accustomed to the taxes on surface transportation, while the air-ship proprietors, if they had not convinced themselves that these taxes were necessary to the continued prosperity of aerial navigation, could readily pretend so, and would have, in opposing their repeal, the advantage of that inertia which tends to the continuance of anything that is.

The superiority of the bounty system over the protective system for the encouragement of any single industry is very great; but it becomes greater as the number of industries to be encouraged is increased. When we encourage an industry by a bounty we do not discourage any other industry, except as the necessary increase in general taxation may have a discouraging effect. But when to encourage one industry we raise the price of its products by a protective duty, we at the same time produce a directly injurious effect upon other industries that use those products. So complicated has production become, so intimate are the relations between industries, and in so many forms do the products of one industry enter into the materials or processes of others, that what will be the effect of a single protective duty it is hard for an expert to say. But when it comes to encouraging not one nor a dozen, but a thousand different industries, it is impossible for human intelligence to trace the multifarious effects of raising the prices of so many products. The people cannot tell what such a system costs them, nor in most cases can even those who are supposed to be its beneficiaries really tell how their gains under it compare with their losses from it.

The "drawback" system is an attempt to prevent, so far as exports are concerned, the discouragement to which the protection of one industry subjects others. Drawbacks are bounties paid on exports of domestic goods to an amount which it is calculated will compensate for the addition a duty on material has made to their cost. But drawbacks not only leave home prices undiminished, but while fruitful of fraud, can only in small part prevent the discouragement of exports, since it is only on goods into which dutiable commodities have entered in large proportion and obvious ways that drawbacks are allowed, or that it is worth the while of the exporter to attempt to collect them. In 1884, for instance, the United States paid out a larger sum in drawbacks on copper than was received in duties on copper, yet it is certain that very many exports into which copper entered, and which were therefore enhanced in cost by the duty, got no drawback whatever. And so of drawbacks on refined sugar, for which we are paying a sum greatly in excess of the duties collected on the raw sugar, though many of our exports, such as those of condensed milk, syrups and preserved fruits, are much curtailed by these duties.

The substitution of bounties for protection in encouraging industry would do away with the necessity for such inefficient, fraud-provoking, and back-action devices. Under the bounty system prices would not be raised, except as affected by general taxation. Each encouraged producer would know in dollars and cents how much encouragement he got, and the people at large would know how much they paid. In short, all and even more than protection can do to encourage home industries can be done more cheaply and more certainly by bounties.

It is sometimes asserted, as one of the advantages of tariff duties, that they fall on the producers of imported goods, and are thus paid by foreigners. This assertion contains a scintilla of truth. An import duty on a commodity of which the production is a closely controlled foreign monopoly may in some cases fall in part or in whole upon the foreign producer. For instance, let us say that a foreign house or combination has a monopoly in the production of a certain article. Within the limits of cost on the one hand and the highest rate at which any can be sold on the other, the price of such article can be fixed by the producers, who will naturally fix it at the point they conclude will give the largest aggregate profits. If we impose an import duty on such an article they may prefer to reduce their profit on what they sell to this country rather than have the sale diminished by the addition of the duty to the price. In such case the duty will fall upon them.

Or, again, let us suppose a Canadian farmer so situated that the only market in which he can conveniently sell his wheat is on the American side. Wheat being a commodity of which our home production not merely supplies home demands, but leaves a surplus for export, the duty on wheat does not add to price, and the Canadian farmer so exceptionally situated that he must send wheat to this side although there is no general demand for Canadian wheat, cannot get back in enhanced price the duty he must pay.

The two classes represented by these instances suggest all the cases in which import duties fall on foreign producers.8 Such cases, too unimportant to be considered in any estimates of national revenue, are only the rare exceptions to the general rule that the ability to tax ends with the territorial limits of the taxing power. And it is well for mankind that this is so. If it were possible for the government of one country, by any system of taxation, to compel the people of other countries to pay its expenses, the world would soon be taxed into barbarism.

But the possibility of exceptional cases in which import duties may in part or in whole fall on foreign producers, instead of domestic consumers, has in it, even for those who would gladly tax "foreigners," no shadow of a recommendation for protection. For it will be noticed that the cases in which an import duty falls on foreign producers, are cases in which it can afford no encouragement to home producers. An import duty can only fall on foreign producers when its payment does not add to price; while the only possible way in which an import duty can encourage home producers is by adding to price.

It is sometimes said that protection does not increase prices. It is sufficient answer to ask, how then can it encourage? To say that a protective duty encourages the home producer without raising prices, is to say that it encourages him without doing anything for him. Wherever beneath this assertion, as regardless of fact as it is of theory, there is any glimmering of reason, it is either in the notion that protective duties do not permanently add to prices, because they bring about such a competition between home producers as finally carries prices down to the previous level; or else in a confused idea that it would be an advantage to home producers to be secured the whole home market, even if at no higher prices.

But as to the first, the only way in which a protective duty can increase home competition in the production of any commodity is by so increasing prices as to attract producers to the industry by the superior profits to be obtained. This competition, when free to operate, ultimately reduces profits to the general level.9 But this is not to say that it reduces prices to what they would be without the duty. The profits of Louisiana sugar growing are now, doubtless, no larger than in other occupations involving equal risks, but the duty on sugar does make the price of sugar very much higher in the United States than it is in England, where there is no duty upon it. And even where there is no reason in natural or social conditions why a commodity should not be produced as cheaply as in any foreign country, the effect of the net-work of duties, of which the particular duty is but a part, is to increase the cost of production, and thus, though profits may fall, to keep prices above the point of free importation. Did the price of a protected article fall to the point at which the foreign product could not be imported were there no duty, the duty would cease to protect, since the foreign product would not be imported if it were abolished, and the producers for whose protection it was imposed would cease to care for its retention. In what instance has this been the case? Are any of our protected industries less clamorous for protection now than they were forty years ago?

As to the second notion, it is to be observed that the only way in which a protective duty can give the home market to home producers is by increasing the price at which foreign products can be sold in it. Not merely does this increase in the price of foreign products compel an increase in the price of domestic products into which they enter, but the shutting out of foreign products must increase the price of similar domestic products. For it is only where prices are fixed by the will of the producer that increase or decrease in supply does not result in increase or decrease of price. Thus, while the newspaper business is not a monopoly, the publication of each individual paper is, and its price is fixed by the publisher. A publisher may, and in most cases will, prefer increased circulation to increased prices. And if competition were to be lessened, or even cut off, as, for instance, by imposing a stamp duty on, or prohibiting the publication of all the newspapers of New York save one, it would not necessarily follow that the price of that paper would be increased. But the prices of the great mass of commodities, and especially the great mass of commodities which are exported and imported, are regulated by competition. They are not fixed by the will of producers, but by the relative intensity of supply and demand, which are brought to an equation in price by what Adam Smith called "the higgling of the market," and hence any lessening of supply caused by the shutting out of importations will at once increase prices.

In short, the protective system is simply a system of encouraging certain industries by enabling those carrying them on to obtain higher prices for the goods they produce. It is a clumsy and extravagant mode of giving encouragement that could be given much better and at much less cost by bounties or subsidies. If it be wise to "encourage" American industries, and this we have yet to examine, the best way of doing so would be to abolish our tariff entirely and to pay bounties from funds obtained by direct taxation. In this way the cost could be distributed with some approach to fairness, and the citizen who is worth a million times more than another could have the satisfaction of contributing a million times as much to the encouragement of American industry.

I do not forget that, from the bounties given in the colonial days for the killing of noxious animals to the subsidies granted to the Pacific railroads, experience has shown that the bounty system inevitably leads to fraud and begets corruption, while but poorly accomplishing the ends sought by it. But these evils are in separable from any method of "encouragement," and attach to the protective more than to the bounty system, because its operations are not so clear. If protection has been preferred to bounties it is not that it is a better means of encouragement, but for the same reason that indirect has been preferred to direct taxation—because the people do not so readily realize what is being done. Where a grant of a hundred thousand dollars directly from the treasury would raise an outcry, the imposition of a duty which will enable the appropriation of millions in higher prices excites no comment. Where bounties have been given by our States for the establishment of new industries they have been comparatively small sums, given in a single payment or in a subsidy for a definite term of years. Although the people have in some cases been willing thus to pay bounties to a small extent and for a short time, in no case have they consented to regard them as a settled thing, and to keep on paying them year after year. But protective duties once imposed, the protected industry has always been as clamorous for the continuance of protection as it was in the beginning for the grant of it. And the people not being so conscious of the payment have permitted it to go on.

It is often said by protectionists that free trade is right in theory but wrong in practice. Whatever may be meant by such phrases they involve a contradiction in terms, since a theory that will not agree with facts must be false. But without inquiring into the validity of the protective theory it is clear that no such tariff as it proposes ever has been or ever can be made.

The theory of free trade may be carried into practice to the point of ideal perfection. For to secure free trade we have only to abolish restrictions. But to carry the theory of protection into practice some articles must be taxed and others left untaxed, and, as to the articles taxed, different rates of duty must be imposed. And as the protection given to any industry may be neutralized by protection that enhances the price of its materials, careful discrimination is required, for there are very few articles that can be deemed finished products in relation to all their uses. The finished products of some industries are the materials or tools of other industries. Thus, while the protection of any industry is useless unless sufficient to produce the desired effect, too much protection is likely, even from a protective standpoint, to do harm.

It is not merely that the ideal perfection with which the free trade theory may be reduced to practice is impossible in the case of protection, but that even a rough approximation to the protective theory is impossible. There never has been a protective tariff that satisfied protectionists, and there never can be. Our present tariff, for instance, is admitted by protectionists to be full of the grossest blunders.10 It was only adopted because, after a long wrangle, it was found impossible to agree upon a better one, and it is only maintained and defended because any attempt to amend it would begin a scramble out of which no one can tell what sort of a tariff would come. This has been the case with every former tariff, and must be the case with every future tariff.

To make a protective tariff that would even roughly accord with the protective theory would require in the first place a minute knowledge of all trade and industry, and of the manner in which an effect produced on one industry would act and react on others. This no king, congress or parliament ever can have. But further than this, absolute disinterestedness is required, for the fixing of protective duties is simply the distribution of pecuniary favors among a crowd of greedy applicants. And even were it possible to obtain for the making of a protective tariff a body of men themselves disinterested and incapable of yielding to bribery, to threats, to friendship or to flattery, they would have to be more than human not to be dazed by the clamor and misled by the representations of selfish interests.

The making of a tariff, instead of being, as the protective theory requires, a careful consideration of the circumstances and needs of each industry, is in practice simply a great "grab" in which the retained advocates of selfish interests bully and beg, bribe and log-roll, in the endeavor to get the largest possible protection for themselves without regard for other interests or for the general good. The result is, and always must be, the enactment of a tariff which resembles the theoretical protectionist's idea of what a protective tariff should be about as a bucketful of paint thrown against a wall resembles the fresco of a Raphael.

But this is not all. After a tariff has been enacted, come the interpretations and decisions of treasury officials and courts to unmake and re-make it,11 and duties are raised or lowered by a printer's placing of a comma or by arbitrary constructions, frequently open to grave suspicion, and which no one can foresee, so that, as Horace Greeley naively says (Political Economy, p. 183):

"The longer a tariff continues the more weak spots are found, the more holes are picked in it, until at last, through the influence of successive evasions, constructions, decisions, its very father could not discern its original features in the transformed bantling that has quietly taken its place."

Under the bounty system, bad as it is, we can come much nearer to doing what we want to, and to knowing what we have done.