Published in Agricultural History, XIX (January, 1945): 31-37. Transcribed into hypertext by Andrew Chrucky, August 12, 2004.
A POST MORTEM ON THE LABOR-SAFETY VALVE THEORY
FRED A. SHANNON
Department of History, University of Illinois
Since 1935 there has been a growing suspicion among historians that the venerable theory of free land as a safety valve for industrial labor is dead. Out of respect for the departed one even the newer textbooks on American history have begun to maintain silence on the subject. For generations the hypothesis had such a remarkable vitality that a dwindling remnant of the old guard still profess that they observe some stirrings of life in the assumed cadaver. Consequently, it seems that the time has arrived for the reluctant pathologist to don his gas mask and, regardless of the memphitis, analyze the contents of the internal organs. Are the stirrings in the body an evidence of continued animation, or merely of gaseous and helminthic activity? Before the corpse is given a respectable burial this fact must be ascertained beyond any possible doubt.
There can be no question as to the venerable age of the decedent. Thomas Skidmore foretold him as early as 1829 in The Rights of Man to Property! George Henry Evans and his fellow agrarians of the 1840s labored often and long in eulogy of the virtues of the safety valve they were trying to bring into existence. The Working Man's Advocate of July 6, 1844, demanded the realization of "the right of the people to the soil" and said:
That once effected, let an outlet be formed that will carry off our superabundant labor to the salubrious and fertile West. In those regions thousands, and tens of thousands, who are now languishing in hopeless poverty, will find a certain and a speedy independence. The labor market will be thus eased of the present distressing competition; and those who remain, as well as those who emigrate, will have the opportunity of realizing a comfortable living.1
Long before Frederick Jackson Turner tacitly admitted the validity of the theory,2 even the name "safety valve" had become a middle-class aphorism. The idea was so old and so generally held that it was commonly repeated without question. The Republican Party had so long made political capital of the Homestead Act and its feeble accomplishments that the benefit to the industrial laborer had become an axiom of American thought. Turner, himself, made only incidental use of the theory as a further illustration of his general philosophy concerning the West. Apparently he made no effort to examine the basis of the safety-valve assumption. Had he done so, no doubt the theory would have been declared dead forty or fifty years ago, and the present autopsy would have been made unnecessary. It was some of the followers of Turner who made a fetish of the assumption, but in recent years few if any have gone so far as to say that Eastern laborers in large numbers actually succeeded as homesteaders.
The approach has been shifted. An early variation of the theme was that the West as a whole, if not free land alone, provided the safety valve.3 This, as will be seen, was no more valid than the original theory. Another idea, sometimes expressed but apparently not yet reduced to a reasoned hypothesis, is that land, in its widest definition (that is, total natural resources), constituted a safety valve. This is merely one way of begging the question by proposing a new one. Besides, it is easy to demonstrate that as new natural resources were discovered the world population multiplied to take advantage of them and that the old problems were quickly transplanted to a new locality. It can readily be shown that the monopolization of these resources prevented their widest social utilization and that the pressure of labor difficulties was no less intense in new communities than in the old. Witness the Coeur d'Alene strike in Idaho in the same year as the Homestead strike in Pennsylvania. But the natural-resources-safety-valve theory will require a thorough statement and exposition by one of its adherents before an examination can be made. The manufacture of such a hypothesis will be a tough problem, in view of the fact that, ever since the development of the factory system in America, labor unrest has resulted in violently explosive strikes rather than a gentle pop-oft of steam through any supposed safety valve. The question will have to be answered: If any safety valve existed why did it not work? Since it did not work, how can it by any twist of the imagination be called a valve at all?
Another turn of the argument is a revival of the supposition of Carter Goodrich and Sol Davison (further expounded) that while no great number of industrial laborers became homesteaders, yet the safety valve existed, because it drained off the surplus of the Eastern farm population that otherwise would have gone to the cities for factory jobs. So, free land was a safety valve because it drew potential industrial labor to the West.4
Again, the question immediately arises: Why did this potential safety valve not work? Was it really a safety valve at all or was it merely a "whistle on a peanut roaster"? There can be no confusion of definitions involved. There is only one definition of the term: "An automatic escape or relief valve for a steam boiler, hydraulic system, etc." Under the catch-all "etc." one may just as well include "labor unrest." Obviously the safety valve is not for the benefit of the steam, water, or labor that escapes from the boiler, hydraulic system, or factory. It is to prevent the accumulation of pressure that might cause an explosion.
A safety valve is of use only when pressure reaches the danger point. This is where the trouble comes with the labor safety valve in all of its interpretations. It certainly was not working at the time of the Panic of 1837, or in the depression following the Panic of 1873, when over a million unemployed workmen paced the streets and knew that free lands were beyond their reach. It was rusted solid and immovable during the bloody railroad strikes of 1877 and the great labor upheaval of the 1880s. When the old-time Mississippi River steamboat captain "hung a nigger" on the arm of the safety valve when running a race, it can be positively asserted that his safety valve as such did not exist. This belief would doubtless be shared by the possible lone survivor picked maimed and scalded off a sycamore limb after the explosion.
No responsible person has ever tried to deny that at all times in America some few of the more fortunate laborers could and did take up land. But this seepage of steam which went on almost constantly did not prevent the pressure from rising when too much fuel was put under the boiler, and the seepage almost stopped entirely whenever the pressure got dangerously high. It was not till the 1830s, when the factory system in America began to bloom and the labor gangs were recruited for the building of canals and railroads, that any situation arose which would call for a safety valve. The shoemaker or carpenter of colonial days who turned to farming did not do so as a release from an ironclad wage system, as millions between 1830 and 1900 would have liked to do if they could. It was an era of slipshod economy and easy readjustment, where no great obstacle was put in the way of misfits. Even if one admits that a scarcity of free labor for hire was one of the minor reasons for the late development of a factory system, and that the choice of close and cheap land kept down the supply, yet a far greater reason was the scarcity of manufacturing capital. When the factory system began, it was easy to import shiploads of immigrant laborers. The same could have been done a generation or two earlier if there had been the demand.
But perhaps a more substantial argument is needed to answer so attractive a hypothesis as that of the potential safety valve. At first glance this new idea has some charm. Certainly the Western farms did not create their own population by spontaneous generation. If not Eastern industrial laborers, then undoubtedly Eastern farmers must have supplied the initial impulse, and each Eastern farmer who went west drained the Eastern potential labor market by one. But the question is: Did all the migration from East to West amount to enough to constitute a safety valve for Eastern labor? Did not the promise of free land, and such migration as actually occurred, simply lure million of Europeans to American shores, seeking farms or industrial jobs, the bulk of the newcomers remaining in the East to make possible a worse labor congestion than would have existed if everything west of the Mississippi River had been nonexistent The answer is so simple that it can be evolved from census data alone. The post mortem can now be held. If a sufficient domestic migration did take place with the desired results, then there was a safety valve, and there is no corpse of a theory to examine. If not, then the theory is dead and the body can be laid to rest.
The first question to be answered is: How large a surplus of farm population developed and where did it settle between 1860 (just before the Homestead Act) and 1900 (by which date the last gasp of steam is admitted to have escaped from the safety valve)? Here close estimates must substitute for an actual count, for before 1920 the census did not distinguish between actual farm and nonfarm residence. But the census officials did gather and publish figures on the numbers of persons employed for gain in the different occupations, and, wherever comparisons can be made, it is noticeable that the ratio of farm workers to all other persons receiving incomes has always been relatively close to the ratio between total farm and nonfarm population. On this basis of calculation (the only one available and accurate enough for all ordinary needs), in forty years the farm population only expanded from 19,000,000 to 28,000,000, while the nonfarm element grew from somewhat over 12,000,000 to 48,000,000, or almost fourfold. Villages, towns, and cities gained about 18,000,000 above the average rate of growth for the Nation as a whole, while the farm increase lagged by the same amount below the average. These figures are derived from a careful analytical study of occupations, based on census reports, which shows the number of income receivers engaged in agriculture creeping from 6,287,000 to 10,699,000, while those in nonfarm occupations soared from 4,244,000 to 18,374,000.5
Small as was the growth of agricultural population, it must be noted further that over 35 percent of the farms in 1900 were tenant-operated,6 and 43 percent of all farm-income receivers were wage laborers.7 This leaves only 22 percent as owner operators. But even though 25 percent is conceded, this would allow only 7,000,000 people in 1900 living on farms owned by their families, except for some sons who were also wage laborers or tenants. But the total national population increase was nearly 45,000,000 in forty years. Though the zealot may choose to ignore the fact that at least some of the farm workers owned their own land even in 1860 and may accept the figure for 1900 as growth alone, yet he has put but a small fraction of the increased population of the United States on such farms anywhere in the Nation, and hardly enough to consider in the West. This is not the way safety valves are constructed.
A further analysis of the data reveals that only 3,653,000 farms in 1900 were operated, even in part, by their owners. But at the same time at least 21,000,000 farm people were tenants and wage laborers and their families on the total of 5,737,000 farms in the Nation.8 These laborers were rarely any better off financially (often worse) than the toiling multitudes in the cities. These were not persons who had found release either from the farms or the cities of the East on land of their own in the West. The bulk of them were still east of the Mississippi River. These were not potential competitors of the city workers. They were actual competitors, for the hard living conditions of each group had a depressing effect on the economic status of the other. Neither element had the opportunity, the finances, the experience, or the heart to try their luck in the West.
These incontestable facts and figures play havoc with the assumption that "perhaps most" of the Eastern boys who left their "ancestral acres" migrated "to the West to acquire and develop a tract of virgin soil."9 There just was not that much of an increase in the number of farms between 1860 and 1900. Only 3,737,000 units were added to the 2,000,000 of the earlier year, and 2,000,000 of the total in 1900 were tenant-operated.10 How large a proportion of the Eastern boys who left their fathers' farms could have become by any possibility the owners of the fraction of the increase in farms that lay in the West?
Here the potential-safety-valve advocates spoil their own argument. One of them stresses the great fecundity of Eastern farmers, "a dozen children being hardly exceptional."11 At only the average rate of breeding for the whole Nation, the 19,000,000 farm population of 1860, with their descendants and immigrant additions, would have numbered about 46,000,000 by 1900. But barely 60 percent of that number were on farms anywhere in the country at the later date, and only 7,000,000 could have been on farms owned by themselves or their families. If farmers were as philoprogenitive as just quoted, then by 1900 the number of persons of farm ancestry must have been closer to 60,000,000 than 46,000,000, and the increase alone would amount to at least 40,000,000. But the growth of farm population was only 9,000,000, and, of these, little more than 2,000,000 could have been on farms owned by their families. If it could be assumed that all the augmentation in farm population had been by migrating native farmers, by 1900 there would have been 31,000,000 of farm background (as of 1860) residing in the villages, towns, and cities; 9,000,000 would have been on new farms or subdivisions of old ones; of these, nearly 7,000,000 would have been tenants or hired laborers and their families, depressing industrial labor by their threat of competition; and about 2,000,000 would have been on their own farms, whether "virgin soil" of the West or marginal tracts in the East. But it would be taking advantage of the opponent's slip of the pen to trace this phantasy further. The law of averages is enough in itself to annihilate the safety valvers' contention. By the use of this conservative tool alone it will be realized that at least twenty farmers moved to town for each industrial laborer who moved to the land, and ten sons of farmers went to the city for each one who became the owner of a new farm anywhere in the Nation.
As to the farms west of the Mississippi River, it is well known that many of them were settled by aliens (witness the West North Central States with their large numbers of Scandinavians). Here is a theme that might well be expanded. The latest exponent of the potential-labor-safety-valve theory declares that "potential labor was drained out of the country, and to secure it for his fast expanding industrial enterprise, the manufacturer must import labor from Europe."14 Anyone must admit that a fraction of the surplus farm labor of the East went on new farms. But how does this additional immigrant stream into the cities affect the safety valve? The immigrants may not really have increased the industrial population. It has often been contended that, instead, the resulting competition restricted the native birth rate in equal proportion to the numbers of the newcomers. Apparently this must remain in the realm of speculation. Be this as it may, the immigrants, with their background of cheap living, acted as a drag on wages, thus making the lot of the city laborer all the harder. This is not the way that even a potential safety valve should work.
But, returning to the West, there is a further fact to be considered. The total population west of the Mississippi River in 1860 was about 4,500,000. In 1900 it was just under 21,000,000.13 Surely the "fecund" Westerners must have multiplied their own stock to about 12,000,000 by the latter date. In the same forty years some 14,000,000 immigrants came to America.14 By 1900, with their descendants, they must have numbered half again as many, or 21,000,000, for it has not been contended that immigrant competition lowered the immigrant birth rate. On this point the census data are not altogether satisfying. Foreign-born persons and their American-born children (counting only half of the children of mixed American and alien parentage) numbered 23,673,000. No doubt the survivors of the foreign-born counted in the Census of 1860, together with their later children, would reduce the alien accretion since 1860 to the 21,000,000 estimate. If anyone can prove that this should be cut still a few more million, he will not greatly change the estimates that follow.
The Western States, in proportion to their total population, had proved amazingly attractive to the immigrants. Though over 19,087,000 of the 1900 count (including those with only one foreign-born parent) lived east of the Mississippi River, 7,112,000 were in the States (including Louisiana and Minnesota) to the west of the same line. In the eleven Mountain and Pacific States they were 47.6 percent of the total population, the figure reaching 61.2 in Utah, 57.3 in Montana, and 54.9 in California. Nevada also had a majority. Kansas and Missouri alone of the West North Central group had less than 40 percent of alien parentage, while the percentage in North Dakota was 77.5, in Minnesota 74.9, and in South Dakota 61.1. In round numbers Minnesota had 1,312,000, Iowa 958,000, California 815,000, Missouri 741,000, Nebraska 503,000, Texas 472,000, and Kansai 403,000. Aside from Texas the numbers, as well as the percentages, in the West South Centra States were low.15
In 1860 the trans-Mississippi West contained 653,000 persons of foreign birth,16 but the number of their American-born children was not given. Even if the survivors and the children numbered over a million, by 1900 those twenty-two States still had 6,000,000 of post-1860 immigrant stock. If the estimate for the increase of the pre-1860 element is too low, so, it can be countered, were the totals of the Census of 1900. Grandchildren were not counted, and mature immigrants of the 1860s could have had a lot of grandchildren by 1900. All the descendants of the pre-1860 immigrants were included in the estimate of 12,000,000 for the increase of the inhabitants of 1860, whereas all after the first descent are excluded from the post-1860 immigrant posterity. On the other hand let it be conceded that 12,000,000 by internal expansion and 6,000,000 by immigration, or 18,000,000 in all, is too much. This would leave only 3,000,000 of the West in 1900, or one-seventh of the total, accounted for by migration from the Eastern States. The calculator can afford to be generous. Subtract two million from the internal expansion and another million from the alien stock, and add these to the migrants from the Eastern States. Suppose, then, that 6,000,000 of the West's population of 1900 was of pre-1860 Eastern United States origin, and three times that many foreigners and their children had come into the East to replace them. It all simmers down to the fact that the West acted as a lure to prospective European immigrants, either to take up lands, to occupy vacated city jobs, or to supply the demands of a growing industry. In any case the effect was just exactly the opposite of a safety valve, actual or potential.
Now the question is in order as to how many of those Eastern boys who left their "ancestral acres" and migrated "to the West" actually were able "to acquire and develop a tract of virgin soil." As will soon be demonstrated, only 47.1 percent of the Western population of 1900 lived on farms. By the same ratio, a mere 2,826,000 of the exaggerated number of the Eastern stock (as listed above) were farm residents. There were barely more than 2,000,000 farms west of the Mississippi in 1900.17 If two-sevenths of the population was Eastern in origin, it may be assumed that the same proportion of the farming was done by them. This would give them less than 572,000 units to operate as owners, managers, tenants, or hired laborers. But in the West, as in the Nation as a whole, the ratio of tenants and hired laborers to all farmers was very high. A full 35 percent of all Western farms were occupied by tenants. The high ratio in the West South Central region affects the average for all somewhat, but there were several other States that approximated the worst conditions. The percentage in Nebraska was 35.5, in Kansas 33.9, in Iowa 33.6, in Missouri 30.6, and in South Dakota 21.9.18 But, also, slightly over 40 percent of all Western farm-income receivers were wage laborers.19 If these same ratios apply to total population on the farms, then well over 1,130,000 of the Eastern element in the West were wage laborers' families; more than 989,000 were on tenant holdings; and less than 707,000 occupied farms owned by themselves. This means that there was only one person on such a family possession for each twenty-five who left the farms of the Nation in the preceding forty years. But perhaps this number is a little too small. No doubt a good number of the hired laborers were also the sons of the owners. Also, though many of the wage workers in the West lived with their families in separate huts on the farms, another considerable number were single men (or detached from their families) who boarded with the owner. How much this situation affected the given figures is uncertain. But here is something more substantial. Only 65 percent of the farms, or less than 372,000 in all, were owner-operated. Here, then, is the number of those tracts of "virgin soil" taken up and kept -- one for each forty-eight persons who left their "ancestral acres" in the East, or possibly one family farm for each ten families. What a showing for the potential safety valve!
One point remains: Urban development in its relation to safety-valve theories. Between 1790 and 1860 the percentage of persons in cities of 8,000 or more inhabitants grew from 3.3 to 16.1; the number of such places from 6 to 141; and their population from 131,000 to 5,000,000. Over half of this growth took place after 1840. The city was already draining the country. But this was only the curtain raiser for the act to follow. In the next forty years the number of cities was multiplied to 547, their inhabitants to 25,000,000, and their percentage of the total population to 32.9. They had grown more than twice as fast as the Nation at large.20 The same rule applies to all municipalities of 2,500 and over, as their population expanded from 6,500,000 to 30,400,000.21 The cities may have bred pestilence, poverty, crime, and corruption, but there is no evidence that they bred population that rapidly. Immigration alone cannot explain the phenomenon, for, if the entire number of immigrants after 1860 is substracted from the nonfarm population of 1900, the remainder still represents twice the rate of growth of farm population.
It is conceded that the bulk of the immigrants settled in urban localities, and it has been demonstrated that the great bulk of the surplus of farm population did the same. For that matter, outside the Cotton Belt, the majority of the westward-moving population did not settle on farms. When the Eastern city laborer managed to pay his fare or "ride the rods" westward, he, like the migrating farmer, was likely to establish himself in a mining camp, town, or city, where, as in the Coeur d'Alene region of Idaho, he found that he had exchanged drudgery in an Eastern factory for equally ill-paid drudgery (considering living costs) in a Western factory or mine. The urbanized proportion of the population west of the Mississippi River, where 1,725,000 new farms had been created,22 very nearly kept pace with the national average. In 1900, when almost half (47.1 percent) of America's people were living in incorporated towns and cities, the ratio west of the Mississippi River was over three-eighths (38.1 percent). Minnesota exceeded, while Missouri, Iowa, and Nebraska nearly equaled the national ratio. The combined eleven Mountain and Pacific States rated even higher than Minnesota, with 50.6 percent of their population in incorporated places. It was only the Dakotas and the West South Central States that were so overwhelmingly rural as to keep the trans-Mississippi West below the national ratio.23 On the basis of the gainfully employed, always a better measure, the West showed a still higher proportion of nonfarm population. The census figures for 1870, 1890, and 1900 are used in the accompanying table to illustrate this point.24
In each decade, the Far-Western regions were well below the national ratio of agricultural to town and city labor, and to 1890 they were far below. In 1870, outside the West South Central States and Iowa, the figure averaged 44.3 percent for seventeen Western States compared with 47.4 percent for the United States. In the next twenty years, when free land was presumed to be the greatest lure of the West, the towns gained on the farms till the latter included only 46.5 percent of the Western total in spite of the still preponderantly rural character of the West South Central division. Then in 1890, according to the legend, the gate to free land flew shut with a bang, and the urban-labor safety valve rusted tight forever. Yet, the increase in agricultural population in the next ten years was nearly a fourth larger than the average for the preceding decades. Whereas the city had been draining labor from the farm before 1890, now that the theoretical safety valve was gone the Western farm was gaining on the Western city. Good land -- free, cheap, or at speculators' prices -- undoubtedly was more abundant before 1890 than afterward. Before that date, without cavil, this land had helped keep down rural discontent and unrest. A small percentage of surplus farmers and a few other discontented ones in periods of hard times, had been able to go west and take up new farms, but many times that number to had sought refuges, however tenuous, in the cities. Whether this cityward migration left the most intelligent and energetic or the duller and more indolent back on the farm is relatively immaterial so far as the release of pressure is concerned. Such evidence as has been uncovered shows no decided weight one way or the other.
This much is certain. The industrial labor troubles of the 1870s and 1880s, when this potential safety valve was supposed to be working, were among the most violent ever experienced in the Nation's history. Steam escaped by explosion and not through a safety valve of free land. On the other hand, down to 1890 the flow of excess farmers to the industrial centers was incessant and accelerated. When hard times settled down on the farms of the Middle West, as in the 1870s, Grangers could organize, antimonopoly parties arise, and greenbackers flourish; but the pressure was eased largely by the flow of excess population to the towns. No doubt the migrants would have done better to stay at home and create an explosion. Instead, they went to town to add to the explosive force there. Farm agitation died down when a few reforms were secured, and the continued cityward movement retarded its revival.
However, after 1890 this release for rural discontent began to fail. The cities were approaching a static condition and were losing their attraction for farmers. This condition continued until between 1930 and 1940 there was virtually no net shift of population between town and country.25 In the 1890s when the city safety valve for rural discontent was beginning to fail, the baffled farmer was at bay. Drought in the farther West and congestion in the cities left him no direction to go. He must stay on his freehold or tenant farm and fight. Populism in the 1890s was not to be as easily diverted or sidetracked by feeble concessions as had been Grangerism in the 1870s. In the forty years after 1890, the farmers, balked increasingly in their cityward yearnings, began to take far greater risks than ever before in their efforts to conquer the arid regions. Four times as much land was homesteaded as in the preceding decades.26 Great things were accomplished in the way of irrigation and dry farming; but also great distress was encountered, great dust bowls were created, and great national problems of farm relief were fostered.
Generalization alone does not establish a thesis, but already there is a substantial body of facts to support an argument for the city safety valve for rural discontent. Nevertheless old stereotypes of thought die hard. Quite often they expire only with their devotees. It has been proved time after time that since 1880, at least, the old idea of the agricultural ladder has worked in reverse. Instead of tenancy being a ladder up which workers could climb to farm ownership, in reality the freeholder more often climbed down the ladder to tenancy. Yet there are people in abundance who still nourish the illusion that their old friend remains alive. There is no reason for assuming that in the present instance the truth will be any more welcome than it has proved to be in the past. There never was a free-land or even a Western safety valve for industrial labor. There never was one even of the potential sort. So far did such a valve fail to exist that the exact opposite is seen. The rapid growth of industry and commerce in the cities provided a release from surplus farm population. The safety valve that actually existed worked in entirely the opposite direction from the one so often extolled. Perhaps the growth of urban economy also, on occasion, was rapid and smooth enough to absorb most of the growing population without explosive effect. Once the people concentrated in the cities, there was no safety valve whatever that could prevent violent eruptions in depression periods. Of this, the die-hards also will remain unconvinced. The persons who mournfully sing that "The old gray mare, she ain't what she used to be" seldom are ready to admit that she never did amount to much.
The post mortem on the theory of a free-land safety valve for industrial labor is at an end. For a century it was fed on nothing more sustaining than unsupported rationalization. Its ethereal body was able to survive on this slender nourishment as long as the supply lasted. But when the food was diluted to a "potential" consistency, it was no longer strong enough to maintain life. Death came from inanition. The body may now be sealed in its coffin and laid to rest. Let those who will consult the spirit rappers to bring forth its ghost.