The Travels of a T-Shirt in the Global Economy

As technological innovation increased productivity, higher productivity in turn lowered prices. The lower prices spurred demand for textiles, which then left England starving for raw cotton. Once the British masses had a taste of “gaiety of dress,” there was no turning back. The cheap cotton clothing available to the masses was the historical equivalent of today’s $5.99 cotton T-shirt. Then, as now, consumer demand was behind the push and pull of world trade flows.

Of course, British demand for cotton does not fully explain American success in meeting that demand. Indeed, at the takeoff of the Industrial Revolution, the United States did not seem like a promising source of cotton at all. As Figure 2.1 shows, in 1791, the U.S. share of world cotton production was almost too small to be counted. The American South produced barely 2 million pounds of cotton in 1791, a minuscule amount compared to the output of producers elsewhere. It is doubtful that producers in Asia (primarily India), with production of nearly 400 million pounds, perceived much of a competitive threat from the American South.

The boom in U.S. cotton production that happened next was astounding. In 10 years, American production increased by 25 times. And by the outbreak of the Civil War, the South was producing more than a billion pounds per year, approximately two-thirds of the total world production. Cotton production was overwhelmingly export oriented. From 1815 to 1860, cotton constituted approximately half of the value of all U.S. exports, and more than 70 percent of all American cotton produced was exported, primarily to England.² In a relatively short period of time, American cotton farmers had trounced their foreign competition.

The victory did not come cheaply. First, the single-minded concentration of capital, labor, and entrepreneurial energies into cotton production left the American South far behind the North in broader industrial development, a gap that has narrowed decisively only during the past 25 years. Second, early American cotton production took place mostly, though not entirely, on slave plantations, and there is little doubt that this system of human captivity contributed significantly to the “productivity” of the American cotton grower. And while plantation slavery was undoubtedly the most horrible of the many labor systems in U.S. economic history, as we will see, slavery is not the only instance in which a horrific—or at least objectionable—labor system played a role in the production and trade of cotton clothing such as T-shirts. On this issue, today’s trade skeptics have a point.

Slavery was the first significant American “public policy” that served to protect cotton growers from the perils of operating in a competitive market. For a number of reasons, relying on a competitive labor market—rather than on captive slaves—was a risk that growers were loath to assume, and it was also a risk that would have likely precluded the explosive growth in American cotton production.

Growing cotton in the antebellum South was mind-numbing, backbreaking physical labor. Beginning in mid-spring, the ground would be prepared for planting with hoes, and later, mule-drawn plows. Following planting, the battle of the weeds began. The tender cotton plant was not able to hold its own against the rapacious weeds, and so required the constant help of workers who guarded the young plants against their encroachment. Indeed, numerous journals and diaries reveal that keeping cotton “out of the grass” was perhaps the planters’ biggest worry and the most physically demanding work.³ Weeding and thinning continued, although at a slower pace, almost until the four-month harvest season began in late summer. On a large plantation, one worker could prepare, plant, weed, and harvest about 18 acres of cotton.

Critically, the timing and intensity of each of these tasks was dictated by the weather, so the growers were unable to predict their labor requirements beyond the weather forecast. During a very rainy spring, each field had to be weeded up to six times, which doubled the labor requirement during that season. The harvesting of cotton was perhaps the most unpredictable task. (Even today, Nelson and Ruth Reinsch cannot plan for Thanksgiving travel.) Cotton cannot be picked either in the rain or while still wet, and it typically takes three to four days to dry. A few days of rain, then, might leave pickers idle for a week. But once the cotton was open and dry, it needed to be picked as soon as possible, so that the tender fluffs did not blow away or fall to the ground. Cotton that had been rained on became spotted and weaker, so often planters tried hurriedly to get the cotton picked as rain clouds approached.

These exacting and unpredictable labor requirements were impossible to meet while relying on the market. As Gavin Wright has argued, farm labor markets in the American South barely functioned, if in fact they existed at all.⁴ Farms were geographically dispersed, which made communication and transportation difficult. The very low population density, combined with uneven labor requirements throughout the year, as well as poor information flows, meant that a farmer who relied on the “market” to meet his labor needs might not be able to harvest his crop at any price. The problem of farm labor, then, was not limited to a shortage of workers or high wages. Rather, the problem was the absence of a well-functioning market where farm workers and growers could transact with any degree of effectiveness. Relying on the market to supply the right number of workers at the right time was a business gamble that cotton farmers preferred to avoid.

Even with a functioning labor market, however, it is doubtful that workers would have been attracted to opportunities as wage hands in cotton production. As a very early student of the cotton economy noted, “the difficulty or impossibility of inducing the whites to become wage earners while they were in contact with cheap land is undoubtedly the chief reason why the cotton industry in the country was developed by slave instead of by free labor.”⁵ Of course, the same could be said of blacks. In the absence of slavery, blacks as well as whites would prefer a farm of their own to work as wage hands. And, in the early years of the American South, land was available to all comers.

In summary, free labor—black or white—was unlikely to be attracted to wage work on Southern cotton farms, because of both the poor functioning of labor markets and the superior alternative available to these workers—the family farm. Slavery, then, allowed cotton farmers both a way to avoid the risks associated with transacting in the labor market and a way around the family labor constraint. Slavery also enabled the growers to cultivate greater acreage. The greater acreage in turn allowed cotton production to increase. The average farm size in the cotton South was nearly twice that of the free states of the North, and there was a strong positive relationship between farm size and relative cotton production, at least for farms below 600 acres.⁶ Put simply, large farms were slave plantations, not family farms, and it was the slave plantations that produced most of the world’s cotton by 1860.

My first care has been to select a proper place for my “Quarter” well protected by the shade ... and to erect comfortable houses for my negroes.... A large house is provided as a nursery for the children where all are taken at daylight, and placed under the care of a careful and experienced woman, whose sole occupation is to ... see that they are properly fed and attended to. ... I have a large and comfortable hospital provided for my negroes when they are sick ... [and] I must not omit to mention that I have a good fiddler, and keep him well-supplied with catgut, and I make it his duty to play for the negroes every Saturday night until twelve o’clock.⁸

Lest we be tempted to sign up, the writer later notes that his solicitous human resource policies reduced, but did not eliminate, the need for whipping. Whatever its motivation, paternalism clearly strengthened the control of the planter over his slaves and served as a governance mechanism. And when combined with constant monitoring, and the positive and negative incentives that ruled the workday, the planter’s domination was complete.

To summarize, slavery was the first in a set of evolving public policies that served to insulate farmers from the perils of the market. American success in producing large volumes of cotton for world markets required a reliable supply of farm labor, but this labor was likely both unwilling and unavailable through a market mechanism in the pre-Civil War South. But slave ownership alone did not assure productivity. To induce slaves to perform the repetitive and exhausting tasks associated with cotton production, planters used a complex blend of governance mechanisms, including positive and negative incentives, paternalism, and monitoring. Many elements of the command-and-control factory system, of course, survive today in many industries. And complicated blends of incentive and monitoring mechanisms survive as well.

The lessons of the early American cotton industry are relevant for modern debates. America’s early dominance of the cotton industry illustrates that commercial success can be achieved through moral failure, an observation especially relevant for T-shirts, which critics allege are produced under sweatshop conditions not far removed from slavery. But the early story of American cotton also reveals a critical lesson for the market-phobic: It was not the perils of the labor market but the suppression of the market that doomed the lives of the slaves. More generally, the tactic of suppressing and avoiding markets rather than competing in them continues today to be a viable business strategy, particularly in agriculture but also in other industries. This ability to suppress and avoid competition, as we will see, is often the result of a power imbalance between rich and poor, an imbalance that persists in world cotton agriculture today.

With the labor problem “solved” by slavery, unlimited land to the West, and unlimited demand from the East, the pieces were still not quite in place for American cotton’s victory. In their westward expansion, cotton growers encountered perhaps the greatest production bottleneck in American economic history. Once they had pushed farther than 30 miles from the Atlantic coast, the cotton growers found that the lustrous and strong Sea Island cotton demanded by British mills would not bloom. Only Upland cotton, with a shorter fiber and stickier seed, would grow further west. However, while Sea Island cotton could be separated from the seeds with a simple roller gin modeled on an ancient device from India (the Churkka gin), this device was unable to separate the sticky seeds in Upland cotton from the lint.

The severity of this supply bottleneck is difficult to overestimate. A young and healthy slave could pick up to 300 pounds of cotton each day. Even children could typically pick 100 pounds per day. With the seeds, however, the cotton had no market. Since the roller gins would not remove the seed from Upland cotton, slaves were required to pick the seeds out by hand. So sticky and stubborn were the seeds, however, that a slave could clean no more than 1 pound per day. England’s mills would die of cotton starvation at this pace.

So if it hadn’t been Eli Whitney, it likely would have been someone else, and soon. In the fall of 1792, the necessary ingredients for entrepreneurial success converged: a production bottleneck, an idea, a source of capital, and a way to make a profit. For developing countries today, the important part of the story is not Eli—poor countries have plenty of smart and inventive people—it is the convergence of all the ingredients necessary for forward leaps.

Dear Parent:

I received your letter of the 16th of August with peculiar satisfaction and delight. It gave me no small pleasure to hear of your health and was very happy to be informed that your health and that of the family has been so good since I saw you.…I expected to have been able to come [home to] Westboro’ sooner than I fear will be in my power. I presume, sir, you are desirous to hear how I have spent my time since I have left College. This I conceive you have a right to know and that it is my duty to inform you and should have done it before this time.…

On the way to Savannah, Whitney had met the widow and family of Major General Greene of Revolutionary War fame. Mrs. Greene took a liking to the polite young man and invited him to spend a few days on the family’s plantation before continuing his journey. When a group of Revolutionary War officers who had served under General Greene came to the plantation to pay their respects to his widow, the conversation soon turned to the pressing need for a mechanism to separate Upland cotton from its seeds so as to meet the British demand. The seeds, the planters were sure, were the only obstacle to their fortunes. “Gentlemen,” Mrs. Greene remarked, “apply to my young friend, Mr. Whitney,—he can make anything. ”

I went from N. York with the family of the late Major General Greene to Georgia. I went immediately with the family to their plantation ... with an expectation of spending four or five days.... During this time I heard much said of the difficulty of ginning Cotton, that is, separating it from its seeds. There were a number of very respectable gentlemen at Mrs. Greene’s who all agreed that if a machine could be invented which would clean the cotton with all expedition, it would be a great thing for both the Country and the inventor.

I involuntarily happened to be thinking on the subject and struck out a plan of a machine in my mind, which I communicated to Miller (who ... resides in the family, a man of respectability and property). He was pleased with the Plan and said that if I would pursue it and try an experiment to see if it would answer, he would bear the whole expense, I should lose nothing but my time, and if I succeeded we would share the profits.…

The machine worked, of course. Whitney’s simple and elegant model was quickly duplicated throughout the South. The good news was that during the next eight years, cotton production rose 25-fold, and by 1820, more than 90-fold. The bad news was that more than any other single factor, Eli Whitney’s cotton gin solidified the slave plantation in the cotton South. For the growers, it was good while it lasted. For the men and women who had been bought and sold and bred and whipped and captured and fiddled to, it was good when it ended.

At the beginning, as Figure 2.1 shows, other countries continued to produce cotton in relatively stable quantities while American production soared. It was not a matter, then, of American producers squashing the competition with low-cost and efficient production. Instead, for the older cotton producers, it was business as usual. But business as usual was not good enough.

British demand for cotton had exploded with the new textile machinery and the burgeoning consumer class. It was not a matter of steady growth in demand, not a curve that the old cotton producers could ride profitably on into retirement. The British Industrial Revolution was a lightning bolt in cotton’s story, like the cotton gin or the boll weevil or emancipation, which changed everything ahead. By 1860, Britain was consuming over a billion pounds of cotton per year, which was considerably more than the entire production of the world, excluding the United States.¹⁰

An explosion in demand required an explosion in supply. The question, then, becomes why the supply exploded in the United States rather than in the countries that had been the world’s major producers since the beginning of the cotton trade. The question of American success becomes more intriguing when we note the remarkable lengths to which the British went—quite unsuccessfully—to reduce their risky dependence on American cotton.

Put simply, modern markets did not yet work in India or China, in cotton or in anything else. As economic historian David Landes advises, a useful way to understand why something in economic history did or did not happen at a certain place and time is to ask, Who would have benefited?¹¹ If cotton growers in India or China could have benefited by increasing their productivity, improving quality, and selling cotton to British mills, they would have done so. It appears, though, that they would not have benefited; the risks were too great, the rewards likely minimal. Capitalism of the type that rewards an idea, an improvement, or an initiative, had not yet taken hold in Asia. The foundations were lacking.

First, there were no property rights, or as Francois Bernier, a Frenchman who lived in India during the seventeenth century, wrote, no mien et tien (no mine and yours).¹² There were no incentives to improve age-old methods, to learn, to grow more, to do better. The agricultural workers were at the mercy of rulers who were often absent, and who changed and moved frequently. And even if wealth had been created, Bernier wrote, it had to be hidden lest it be extorted or seized.¹³

In China, too, cotton growers would not have benefited. Under the tyranny of the emperor, there was little reason to take a business risk in the modern sense of the term. As a Christian missionary remarked in the late 1700s, “Any man of genius is paralyzed immediately by the thought that his efforts will win him punishment rather than rewards.”¹⁴ As Landes notes, too directly for most tastes, China’s “cultural triumphalism and petty downward tyranny made [the country] a reluctant improver and a bad learner.”¹⁵ Culturally, the Qing dynasty, which ruled China from the 1600s until the early 1900s, displayed an aversion to all things Western, and to change in general. A Jesuit passing through commented that the Chinese were “more fond of the most defective piece of antiquity than of the most perfect of the modern...”¹⁶ In other words, all of the Eli Whitneys in China had no reason to try.

On the surface, of course, the American cotton victory over India and China appeared to be due to slavery. An 1853 observer confidently noted that American cotton growers’ “superiority” was due to the “cheap, and reliable labor they derive from that patriarchal system of domestic servitude.”¹⁷ While certainly it was slavery that allowed the cotton factories on the plantations to produce such enormous volumes of cotton, India and China, too, had millions of people who were made to work for nothing by tyrannical rulers, millions of people who could not say no. Why these people were never organized to produce large volumes of cotton for export is another matter entirely.

Thus, while slavery allowed farmers to evade the risks of the labor market, it does not explain why other countries failed to seize the opportunities presented by the Industrial Revolution. The institutions necessary to support factory-style cotton production—property rights, incentive structures, what is today called “governance”—also had an important role to play. Governance still has an important role to play, which will remain the challenge for many poor cotton-producing countries. As we will see, all of the Eli Whitneys in Mali, Burkina Faso, and Benin often still have little reason to try.

This dire prediction about the demise of civilization rested on the importance of cotton to the industrial centers of the Northern states and Europe. The giant textile mills that lined the rivers of the new industrial centers depended on the South to supply cotton. This bit of fluff, the boll as big as a fist yet lighter than a breath, reigned supremely, if not benevolently, over the world’s new economic order. Southern cotton had a God-given monopoly. Because it could not be grown either in the Northern states or in England, Hammond reasoned, the industrial world would bow to cotton, and the South had nothing to fear:

But while the Civil War eliminated slavery, the cotton economy of the South survived because public policy evolved to continue to protect the growers from the perils of the labor markets. Labor requirements in cotton production remained highly seasonal, and the challenge was still to have sufficient labor available at critical but unpredictable times in the cotton cycle. However, transacting in a labor market was fraught with risk, as the market still offered no guarantees about either the price or availability of labor at these critical times. Without the tight control of slavery, landowners needed an alternative system to bind labor to their land upon demand. The labor system that emerged—tenant farming, or “sharecropping”—fit the bill.

In exchange for their labor, the landowner provided the sharecroppers with housing and food (known as the furnish) as well as the right to hunt and to fish. By providing housing and food, rather than cash, the landowner bound the worker to the property and assured himself of labor at critical times. The worker was contractually bound as well, since he was indebted to the landlord through the harvesting of the crop.

A wide variety of public policies were instituted to bind the sharecroppers to the land and insulate the cotton growers from the risks of transacting in the labor market.²⁰ Gradually, the legal definition of sharecropper shifted in favor of the landowners, especially through the passage of crop lien laws.²¹ These laws changed the status of the sharecropper in the courts to a laborer who was paid wages in crops rather than a tenant with ownership of a share of the crop. The difference was critical. As a laborer, the sharecropper could not offer his crop for lien because it technically belonged to the landowner. The crop lien laws, then, shut the sharecropper out of the capital markets while widening access to capital for the landowners. Other laws, such as vagrancy laws and “alienation of labor” laws (which protected the landowner from having his labor hired away) also served to bind the sharecropper to the land. At the same time, planters opposed public schooling for blacks and poor whites, so illiteracy and lack of education kept the balance of power in the sharecropping arrangement heavily in favor of the planter, and limited the alternatives of the workers.

Moreover, the contractual arrangement between sharecropper and landowner left the sharecropper little hope of climbing out of subsistence. The sharecropper’s dream—to own land—was thwarted by a cycle of perpetual debt whereby the sharecropper’s share of each harvest was barely enough to settle the year’s debts, and by exclusion from external capital markets. A remark reportedly made by Louis XIV of France is apt: “Credit supports agriculture as a cord supports the hanged.”²²

It took all them six bales to pay Mr. Curtis. In the place of prosperin’, I was on a standstill.... I had not a dollar left out of the cotton.... Mr. Curtis had Mr. Buck Thompson furnish me groceries ... kept a book on me.... [Mr. Curtis] paid Mr. Thompson and I paid him—the deal worked that way—out of my crop. So he made somethin off my grocery bill besides gettin half my crop when the time came.²³

Cobb’s biography repeats this theme year after year. Some years, there was a little cotton left after paying the landlord; in other years, there was not enough to settle the debts and Cobb had to start the next year in the hole. Thanks to creative accounting, it was typical to come out even. In Macon County, Alabama, researchers uncovered a remarkable coincidence: 62 percent of black sharecroppers had come out even for the year in 1932.²⁴

Ironically, the success that the planters had in devising public policies to keep the workforce docile and uneducated soon began to backfire. When the boll weevil began to ravage the southern cotton crop in the early 1900s, government extension programs were mobilized to spread advice to farmers on how to combat the weevil and save their crops. The news and advice reached the large farms and the educated farmers, but often passed by the poor and illiterate sharecroppers, black and white, who had to fend for themselves.²⁵ In 1921, approximately 30 percent of the cotton crop—predominantly that produced by small sharecroppers—was lost to the weevil.²⁶ Many were pushed off the land. Ned Cobb remembered the time well:

That was boll weevil time... these white folks told the colored people if you don’t pick them cotton squares off the ground and destroy them boll weevils we’ll quit furnishin’ you. Told em that—puttin the blame on the colored man for the boll weevil. Couldn’t nobody pay his debts when the weevil et up his crop. ²⁷

The requirements for successful large-scale cotton farming had changed little from the pre-Civil War South. The landowners still required large numbers of workers to be available on demand to plant, weed, and harvest the crop at the whim of the weather. Relying on a labor market in the modern sense of the term was still fraught with risk and expense. How would the planter be assured that the market would provide for labor requirements when the weeds bloomed or the cotton opened? And what if the market wage went up or help was hired away by competitors?

Creative solutions abounded.³⁰ Planters imported monkeys from Brazil and tried to teach them to pick cotton, but the animals in the end were uncooperative. And geese, it turned out, will weed a cotton field when fenced in, and the farmers discovered that only two geese could weed an acre of cotton. They also discovered, however, that geese could not be trained not to trample cotton plants, and that insecticide is also goosicide. For a time, farmers also used flamethrowers to weed cotton fields, but taking fire down the rows of their livelihood proved too difficult for most. In the end, neither monkeys nor geese nor fire could accomplish the tasks as well as a captive labor force.

The Taft cotton ranch, near Corpus Christi, occupied 39 percent of the land of San Patricio County.³¹ The ranch was organized as a corporation, but in reality it was a community in which people’s lives—not just their work—were hierarchically managed for the purpose of cotton production. The ranch had company housing, schools, and churches segregated along ethnic lines for whites, Mexicans, and blacks. Like the “furnish” provided to old South sharecroppers, workers were paid partly in scrip, which could be redeemed only at company stores. Finally, like the plantation owner who kept his “fiddler well-supplied with catgut,” the Taft Ranch provided holidays, music, and festivities as well, again designed for the three different ethnic groups. This entire system, of course, served to ensure that workers were around when the cotton needed to be planted, weeded, and harvested. The new cotton factories did not so much influence public policy, they were public policy over vast stretches of Texas.

These large and tightly controlled production systems were hailed as models of the farms of the future, models of productivity, efficiency, and profitability. Once again, successful large-scale cotton production depended on a factory system in which large numbers of workers were available on demand to complete the repetitive chores associated with weeding, planting, and picking. Once again, success depended on avoiding—not competing in—the labor market.

Keep the Fiddler Well-Supplied with Catgut

Eli Meets a Venture Capitalist

Where Was the Competition?

All God’s Dangers Ain’t a White Man

Cotton Factories Arrive in Texas