Some Aspects of the Tariff Question

The cotton manufacture grew up,—to recapitulate summarily,—during the period of interrupted foreign trade which preceded the war of 1812 and continued through the war until 1815. It was systematically and successfully developed during the time of the early protective movement which set in with the tariff of 1816; it maintained itself unshaken notwithstanding the gradual reduction of duties carried out in 1833-40 under the provisions of the compromise tariff act of 1833. A marked advance took place in the decade 1840-50, perhaps stimulated by the higher duties of the tariff act of 1842, but at all events not checked by the lower duties of the act of 1846. From 1846 to 1857 cotton goods were subjected to a simple
ad valorem duty of 95 per cent, and from 1857 to 1861 to one of but 24 per cent. The industry progressed rapidly and grew to large dimensions during this period of moderate duties. Not only did it grow at home, but it reached out to foreign markets. A considerable export trade developed,—conclusive proof, if not of complete independence from protection in every branch, at least of a stage of development to which the young industries argument could no longer apply.
*52

Nevertheless the further growth of the industry since the civil war suggests some questions which are related to the arguments for protection to young industries, and some other questions which bear on the more general problems of the international division of labor. To these attention will be given in this present chapter.

The rates of duty on cotton goods since 1860 tell a somewhat curious story. In the tariff act of 1861, enacted before the war, specific duties were substituted for the
ad valorem duties of 1846 and 1857; with the declared intention, and in the main probably with the effect, of simply changing the method of levy, not the height of the tariff.
*53 But the change to the specific system soon led to unexpected consequences. During the war, the price of raw cotton went up to extraordinary figures. The average price for 1864 was over fifty cents a pound in gold; and for more than ten years after the war it continued to be at a high level. Not until the close of the decade 1870-80 did it fall to something like the normal figures (ten to twelve cents a pound) that had prevailed before 1860. The prices of cotton goods went up correspondingly, the rise being, of course, most marked in the heavier and cheaper goods for which the raw material was the largest item in the expenses of production. Naturally the specific duties were raised correspondingly. As the prices of cotton and cotton goods had gone up five-fold, so the duties on the goods went up in a similar ratio. On the cheapest grade of unbleached cloth, for example, the rate in 1861 had been one cent per yard; it became five cents per yard in 1864.

The price of cotton began to decline as soon as the war closed; within a year or two it declined greatly. The duties on cotton goods as raised in 1864 became proportionately heavier. Even the rates fixed in 1861 had been prohibitory on the cheap goods; those of 1864 became very heavy, often prohibitory, on goods of medium and finer grades. A reduction was to have been expected; but it was long postponed, and when finally made, still left a high range of rates. Such was the case, as is well-known to all students of our tariff history, with all the protective duties of the war period: it was their prolonged retention, largely through inertia, that caused the protective system to become so extreme. In the case of cottons, the duties, raised to an especially high pitch in 1864, were not overhauled systematically until the general revision of 1883. Even then they were reduced to figures that left them prohibitory for all the cheaper grades of goods. The duty on the lowest class was left at two and one-half cents a yard, amply sufficient to shut out any possibility of importation; and those on most other grades remained correspondingly high. In the protective tariff acts that came after 1883,—those of 1890, 1897, and 1909,—the same process of cautious reduction of the duties on the cheaper grades was continued. By 1897 the duty on the lowest class had gone down to one cent a yard, precisely the figure of 1861. This was still a “safe” rate. So were the corresponding rates on the lower grades generally,—on yarns of the coarser counts, and on the cheap and medium grades of woven fabrics, whether in the gray, or bleached, or printed and dyed.

Meanwhile, as the protective system was extended and stiffened, another movement appeared. The specific duties were differentiated more and more; and side by side with the reduction of the rates on the lower classes of goods, there went a steady increase in those on the dearer goods. In each successive act the same general scheme (that of 1861) was maintained: the specific duties being adjusted first according to the number of threads per square inch of cloth and then according as the cloth was bleached, dyed, printed. In the acts of 1897 and 1909 still another method of differentiation was added,—the number of square yards to the pound,
i.e., the weight per square yard; the fabrics within each class being subjected to higher duties as they were lighter in weight. It is not important for the present discussion to follow the changes in detail: it will suffice to indicate the general trend by noting the maximum duties on the finest fabrics. The maximum was in 1883 6 cents a yard; in 1890 6¾ cents; in 1897, 8 cents; and in 1909, 12½ cents.
*54 Part and parcel of the same tendency was the increase in the dragnet rate,—the general
ad valorem rate on manufactures of cotton not specifically enumerated. The dragnet clause levied, in 1861, a duty of 30 per cent; in 1883, one of 35 per cent; in 1890, 40 per cent; in 1897 and 1909, 45 per cent. The cotton schedule, comparatively simple in 1861, became extremely complex,—so much so that the significance of the rates and gradations of duty was difficult to follow, and the rates became susceptible to the sort of manipulation indicated by the term “joker.” The unusually intricate provisions in the cotton schedule of 1909 gave opportunity for veiled and disguised increases of duty which contributed much to the feeling of suspicion and revolt aroused by this last step in the ultra-protectionist series.
*55

The tariff of 1913, it need hardly be said, made a great breach in this huge and complicated structure. It substituted for the mass of intricate and heavy specific duties a simple system of moderate
ad valorem duties. These were graded, ranging from a minimum of 5 per cent to a maximum of 30 per cent. The lowest rate imposed (5 per cent) was on the coarsest yarns; the highest (30 per cent) was on the finest woven fabrics.
*56 The change in the figures of the statute-book was very great. But, as will appear presently, the effect on the cotton manufacture was in most cases negligible. Only on the finer goods was the reduction of real consequence. At the date of writing these pages (1914) it is still uncertain what will be the effects of the changes on the finer goods.

What effects can be traced to the high duties maintained throughout the half-century that followed the civil war? Those on fabrics of cheaper grade,—the staple goods of the industry,—were quite prohibitory. Those on finer goods, though not in all cases prohibitory, were put up notch by notch in the successive protectionist acts, with the design of promoting the manufacture of such goods within the country. The free trader might be led to predict that the extreme rates on the ordinary goods, and the exclusion of foreign competition as regards them, would lead to something like stagnation in this part of the domestic industry. On the other hand, the high effective rates on the dearer goods might be expected by the protectionist not only to put an end to their importation and cause domestic goods to be substituted for them, but also to bring about some results of the young industries type,—improvements in the field newly opened for the Americans, and attainment of independence or at least indications of some approach to independence.

First a general survey may be made of the growth of the industry at large. The following figures indicate how steady and great was the increase in domestic production, how comparatively small were the imports.

It will be seen that the domestic industry grew rapidly and without check. The best single indication of the extent and growth of such an industry as the cotton manufacture is in the number of spindles; and on this the statistics have been sufficiently accurate. The spindles in 1910 were more than five times as many as in 1860,—twenty-seven millions as compared with about five millions. The same rate of growth is indicated by the consumption of raw cotton; this also increased five-fold. The value of the product (a figure to be used with much more caution) also increased nearly five-fold. The number of persons employed increased distinctly less, about three-fold,—an indication of a growing effectiveness of labor, such as any manufacturing industry may be expected to show. The stationary number of separate establishments is also in accord with the general trend of modern industry; production is on a larger scale, the individual establishment becomes greater, the total number of establishments does not keep pace with the growing volume of production.

The imports, on the other hand, show no considerable change, except in the very last decade. As in the case of silks, they remain not far from constant absolutely, and thus become a steadily diminishing proportion of the total supply. In 1860 they were, in value, still not very far from one-third of the domestic output; in 1910, little more than one-tenth.
*58 It will be shown presently that these general figures need much explanation. The continuing imports are in large part specialties; those which really compete with the domestic products are even less considerable than the figures would indicate. It is clear, however, that a very great increase in the cotton industry has taken place within the country. Here also the protective system would seem to have succeeded in attaining at least one object,—a great preponderance of domestic supply, a lessening dependence on imports.

Proceeding now to a more detailed consideration of the several branches of the industry, let attention be given first to the manufacturing of the cheaper grades. This was the earliest to be established, and the only one that flourished before the civil war. As has just been noted, it seems to have already reached in that period the stage of independence. The foreign (British) competitors were not feared, except possibly in times of exceptional depression in the foreign markets. Exports on a considerable scale had begun. Even the comparatively moderate duties of 1861 had been virtually prohibitory on the cheaper goods; they were prohibitory beyond doubt through the half-century after the war duties of 1864. The domestic manufacture in this branch has, therefore, gone its own way, quite untroubled by foreign competition.

This part of the cotton manufacture remained, after the war as before, quantitatively by far the most important. In 1905, the census report on the industry stated that “almost three-fourths of all the woven goods reported fall under the classification of coarse or medium counts,—print cloths, sheetings, and shirtings, drills, ticks, denims and stripes, duck and bagging.”
*59 Over one-half of the yarn spun in American mills was in 1905 and in 1910 of the low counts (1 to 20) used for distinctly coarse goods. Five-sixths of the remainder was of counts still low (20 to 40),—what might be called low-medium counts.
*60 In other words, only one-twelfth of the quantitative output (pounds) could be reckoned as spun for the fine or better medium goods. The great growth which has taken place in the industry has therefore been predominantly in that branch already firmly established before the system of high protection was applied.

With this growth in the manufacture of the ordinary (cheaper) goods, a marked change has taken place in geographical distribution. Until 1880, New England and the middle states were almost the sole seats of the industry. After that date a rapid growth took place in the south (chiefly in South Carolina, North Carolina, and Georgia) until by 1910 this region became comparable in importance with the northern states. The goods made in the south have been almost exclusively of the ordinary grades; and this circumstance has much affected the character of the industry elsewhere. The northern mills, especially those of New England, felt the competition of the south on the cheaper grades and turned more than before to the finer. In the older seats of the industry, therefore, the diversification has been greater than the general figures indicate. The finer goods are made almost exclusively in the north, and chiefly in New England; and hence they form in the last mentioned region a much more important constituent than they do in the country at large. Yet even here much the greater part of the manufacture is still given to the cheaper goods.
*61 What causes have influenced the great growth in the south, and the tenacious hold even of the cheaper grades in the north, will be considered as we proceed.

More significant, however, than the volume of growth was the technical development of the cotton manufacture. Both the changes which took place in the American mills and those which failed to take place are instructive. It is chiefly in the manufacture of the cheaper goods that machinery and methods were remodelled; as regards dearer goods there has been least tendency to divergence from the practices of European rivals, especially of Great Britain. It will be convenient to describe briefly the technical changes that most affected the American industry, proceeding then to a consideration of their bearing on the tariff problems.

During the half-century the two fundamental processes in the mills—spinning and weaving—underwent changes almost revolutionary as regards the cheaper goods which constitute the bulk of the American output.

In spinning, the great change has been the extraordinary growth of ring spinning
*62 and the decline of mule spinning. The following figures show what sort of transition has taken place:

It will be seen that the number of ring spindles has increased without halt, both absolutely and relatively. The number of mule spindles, on the other hand, has hardly increased at all. Though there was some gain in the twenty years from 1870 to 1890, a loss followed from 1890 to 1910, so that in the last-named year the total of mule spindles exceeded but little that of 1870. At the outset (1870) the two kinds were in use half and half; at the close (1910) the ring spindles had increased nearly ten-fold, and constituted five-sixths of the total.

The mule spindle is in essentials that invented by the English pioneers in the industrial revolution of the eighteenth century. As perfected by Crompton, it involved the placing of a large number of spindles on a single stand or carriage which moves to and fro, spinning on its movement one way only, and getting ready for the next spinning on the return movement. Ring spinning is more recently invented, still more recently of wide use. The essential of the device is a small steel ring, through which passes the roving (the smoothed and partitioned sliver of flimsy cotton) and in passing is given the twist which pulls the fibres together into yarn. Of American invention (1828), it came into extensive use in the United States even before the war. After the war, and particularly in the decade from 1880 to 1890, it was immensely improved by a series of subsidiary changes, and took the commanding place in the industry indicated by the figures just given.
*64

The industrial differences between the two methods of spinning can be stated without entering on the complicated mechanical details. The ring, in brief, is better adapted for coarser yarns, for economy of space, for large-scale operations, for that combination of spinning and weaving in the same establishment which has always been the rule in American mills, and, last but not least, for the utilization of labor little skilled. The ring spins continuously, not intermittently as does the mule; and for this reason, as well as for others, the ring produces more per spindle. The ring puts more strain on the yarn, and hence is available primarily for the coarser yarns; yet the march of improvement has made it available for yarns less coarse than in the earlier stages of its use. It yields yarn comparatively harsh, and not acceptable where a softer quality is needed (
e.g., for most hosiery) or where much sizing is to be put into the fabric (as is commonly done in England). The ring winds the yarn on wooden bobbins of appreciable size and weight; and the yarn thus wound and mounted is more expensive to transport than that which comes from the mule. Hence arises an obstacle to specialization between spinning and weaving; ring spinning strengthens the general American practice of combining the two in one establishment.

Perhaps most important of all is the difference in the kind of labor force required. Mule spinning is a trade, and mainly a man’s trade. The spinner is a skilled workman, or at least comes close to that grade. In Great Britain the trade is often hereditary. It has been stated to the present writer, by conversant persons, that only a boy who has grown up in a mill can become a good mule spinner. The statement doubtless is exaggerated, but doubtless rests on a basis of fact. Mule spinners have strong unions; they cannot be readily replaced when they strike. They are often accused by the manufacturers of being a turbulent and unruly set, of clannishly opposing the entrance of recruits into the trade, of having a trade-union monopoly; all of which are indications that, though the degree of skill may be exaggerated, the men must have some of the qualities of the skilled handicraftsman. The ring, on the contrary, is more automatic, needs less continuous and alert watching, can be operated with little need either of strength or skill. Ring spinning has been very greatly improved in the United States during the half-century; and the improvements have taken the direction of making the machinery more self-acting, less in need of skilled attention, less liable to breakdown and repair. Ring spinners are always women and children, who can be easily trained and easily replaced.

The difference in the needed quality of labor goes far to account for the unequal distribution of mule spindles and ring spindles in the various seats of the American industry. Mule spinning in the United States is confined almost entirely to the north. Even there it is overshadowed by its rival; while in the south there is virtually no mule spinning at all. The figures from the Census of 1910 again tell the story.
*65

The progress of invention in ring spinning machinery has been characteristic. A series of Yankee machinists and manufacturers experimented with the various refinements of the device, vied with each other in offering the cotton manufacturers different variants, added improvement to improvement, until by a process of selection and survival the well-nigh perfect machine was developed. The number of revolutions per minute had been 5,500 in 1860, and became 9,000 by 1890. The operation of the spindle was declared by the foremost expert to be “so near absolute perfection that it would seem as though no changes were required.”
*66 Yet after that date the speed of revolution was raised to 10,000 per minute, even to more in some cases; some such figure being apparently the maximum for the device as it now (1910) stands.
*67 No more labor, no more power, no more space were required for the improved spindle; the doubling of speed meant a doubling of output.

Even more important than the changes in spinning were those in weaving. As has already been noted, weaving was carried on by Americans with special aptitude and success from the very beginnings of the modern textile industries. The power loom was put into use—nay, virtually invented—in the cotton mills of the United States contemporaneously with its introduction in England.
*68 By the first third of the nineteenth century the weaving processes in American mills were found by a skilled observer to be at least equal to those in England, perhaps superior.
*69 And in the closing decade of that century a new invention, that of the automatic loom, was perfected in the United States and adopted almost universally for the cheap and medium goods.

The ordinary power loom in a sense is automatic; the weaver is no more than an attendant who simply sees that the machine runs as it should. The degree of attention, however, varies greatly according to the nature of the material turned out. On some goods the power loom weaver can operate but one loom, as did the hand loom weaver before him; and he must have some of the qualities of the skilled artisan. Such is the case with finer woolen and silk fabrics; and, as has already been noted for silks and will be pointed out presently for woolens also,
*70 these are the branches of the textile industries which are not easily domiciled in the United States. But on plain cotton goods of the cheaper grades the power loom had long been developed to the point where the mechanism largely took care of itself, and where a weaver could attend to six or eight looms, sometimes even more. One operation, however, had not been subjected to the machine, and thereby a limitation had remained on its uninterrupted working,—that of replacing the supplies of weft as they were exhausted. The yarn is wound on bobbins; as one bobbin is emptied by the loom, another must be put in its place in the shuttle, and the thread from this other must be attached to the shuttle which moves to and fro in the loom. The chief business of the weaver on the ordinary power loom is to replace bobbins as they are emptied, and to attach the thread of the fresh bobbin to the shuttle. On the average the loom has to be stopped once in eight minutes to accomplish these two closely-related steps. The automatic loom achieved the crowning triumph of carrying out both without the use of the human hand.
*71 A magazine is attached to the loom, containing a supply of filled bobbins, which are automatically transferred to the loom shuttle. The shuttle itself is automatically threaded by the motion of the loom; and this takes place whether the bobbin is completely emptied or whether its thread is by accident broken before emptying. In either case the shuttle automatically catches up a thread from a fresh bobbin, and the loom continues to work without interruption. The unhygienic process of attaching the fresh thread to the bobbin by the weaver’s sucking it in is done away with.
*72 If a warp (not weft) thread breaks, the loom stops automatically, and the weaver ties the broken ends. The weaver now has become more than ever a mere attendant, keeping an eye on the looms and seeing what is wrong when they are brought to a stop by the automatic devices. The commonest cause of stoppage is the breaking of a thread, and the commonest task of the weaver is to tie a broken thread.

Weaving on the modern power loom, whether of the automatic type or the ordinary type, calls for no strength or special skill. It is not, to be sure, reducible to simple routine as completely as ring spinning. Some alertness is required; and the weaver gets the pay of the average factory worker. But women can be used as well as men, and they seem to be equally efficient. More important, so far as concerns the automatic loom, is the possibility of a more highly developed division of labor. A separate staff (of young persons, boys or girls) can be given the simple task of keeping the magazines charged with bobbins; the weaver can be relieved of this, and called on solely to keep his (or her) eye on the looms. The oiling of the looms and dusting of the floors can be turned over to another set of unskilled persons.
*73

Yet some skilled labor remains indispensable, and on the automatic loom perhaps even more so than on the ordinary loom. The loom fixer, a highly expert mechanic, must be in attendance, to correct any defect in the working of the complicated mechanism or order the transfer of a loom to the repair shop if something serious has happened. The skilled artisan is by no means dispensed with in the modern development of machinery. His sphere of action is merely shifted, and his skill is turned where most needed. This is one of the reasons why machinery which is dubbed “automatic” can never be transported to regions where there is abundance of cheap and unskilled labor, but labor of that kind only. It calls for much more than mere tending and feeding. It must be supervised and kept in order; there must be intelligent and experienced foremen and superintendents, and a staff of skilled mechanics, such as these very loom fixers. However perfected the machine,—nay, the more it is perfected,—the human hand and the human brain are still indispensable.

The Northrop automatic loom,—so named from one among the inventors by whom it was worked out,—illustrates several matters noteworthy in the history of modern inventions. In the first place, it was deliberately planned, and brought to the point of success after prolonged and expensive experimenting. A number of inventors were kept at work on it for years. Some sixty patents were taken out or applied for in the course of the experiments; and the instance is one among many to show that the patent system, however ill adjusted it may have been in some of its details, serves to stimulate invention and still more to promote investment in inventions calling for long and expensive trial. When finally ready to be put on the market, a demonstration of its efficiency had to be given; and the firm which developed it had to shoulder the additional experiment and investment of equipping a large cotton mill in which the loom was first used in manufacturing on a considerable scale. It required this kind of proof, highly effective, but necessitating a still further commitment of funds, to bring the automatic loom into wide use. Convincing the demonstration was. It became clear that, whereas a weaver could attend to eight ordinary looms, he could look after twenty, twenty-four, even thirty automatic looms. Though the capital outlay was larger (the automatic looms are much more expensive), the saving in current labor was so great that the cost of weaving was cut in two. The use of the loom spread with great rapidity, and soon this process dominated the manufacture of the ordinary grades of plain cotton goods.
*74

An instructive aspect of this development is that it has by no means stood alone. The Draper Company had competitors and imitators. A host of inventors and mechanics were vying with them. So it was with the ring spindle just described; there also the leaders did not stand alone, but were spurred on by many keen rivals. So it is, indeed, with every forward movement, whether in literature, in the fine arts, in science, in the mechanic arts. The genius who reaches the crowning achievement is not isolated; he is borne forward by the sweep of a large movement. And every such movement has a character of its own,—the impress of the influences, little understood as regards their relative strength or their channels of operation, of environment, historic growth, the inborn and inherited qualities of a people. So it has been with the various inventions and changes which have marked the industrial growth of the United States throughout its history, and not least during the last half-century.

In the case of the automatic loom—to return to this—rivals and improvers soon appeared. So far as concerned the original field of the Northrop loom, its primacy seems to have been little shaken; there was rivalry, possibly an improvement in one detail or another, but no marked advance. In a neighboring field, however, a striking advance was stimulated by its success. The Northrop loom and its direct rivals were suitable only for plain cloth, or goods with the simplest stripe or figure. They were not available for the ginghams and checks in which weft yarns of more than one color are used. Such fabrics were made on the so-called drop-box looms,—a variant of the ordinary power loom. Here again, systematic experimenting, continued over ten years and more, resulted in a further elaboration of the machine’s competence, a further extension of the range of automatic action. Another well-known firm
*75 put on the market in 1905, a decade after the introduction of the Northrop loom, a gingham loom in which bobbins containing separate colors were held in a magazine and automatically selected for the insertion of the colored threads in the chosen pattern of cloth. Advantages of the same kind as from the Northrop loom, and apparently no less in degree, were secured by this mechanism, when compared with the previous looms for parti-colored fabrics. There was no interruption for putting in fresh bobbins; and the number of looms which one weaver could attend was increased from six to sixteen,—here also more than double. Limitations still remained; the finer and more variegated goods cannot be subjected to this sort of treatment; it was available only for goods of standardized pattern, turned out on a considerable scale. But for the production of quantities of uniform goods on a large scale another striking improvement was achieved.