Between the Market and the State:The Problem of Prison Labor in the New Deal
This essay explores the efforts of federal officials in the early days of the New Deal to regulate, reform, and standardize American systems of prison labor. Eager to curtail the convict-lease system and other inhumane prison practices, yet fearful of idle convicts, officials in the Department of Justice collaborated with Franklin Roosevelt's new National Recovery Administration to create an industry "code" governing convict labor. Initially celebrated as a major step toward a humane and rational system of prison labor, the NRA code immediately ran afoul of powerful corporate groups, skeptical labor unions, and its own internal inconsistencies. Though the code was short-lived, it nevertheless bears study as the first effort to create a national framework for a system of rehabilitative—not profit-driven—prison labor.
convict labor, National Recovery Administration, Prison Labor Compact, New Deal
In 1931, penologist Louis Robinson asked a disarmingly direct question with the title of his book: should prisoners work? Robinson's rhetorical device did not suggest the author's uncertainty regarding his argument. An experienced criminologist at Swarthmore College, Robinson knew what he thought, and was out to persuade his readers. Like any good scholar, he adroitly utilized sources and statistics, and advanced his case methodically. But Robinson also knew that his subject matter was controversial, vexing, and even paradoxical. It did not lend itself to easy answers and ultimate solutions—no periods or explanation marks would do. On the contrary, the propriety of prison labor had inflamed and eluded the best thinking of lawmakers, businesspeople, judges, moralists, reformers, and ordinary workers since the origins of the republic.1
Robinson's question, then, was straightforward but sneaky. Surely prisoners, who had broken society's laws and violated the social contract, deserved punishment. Most ordinary citizens viewed prison labor in this light, Robinson averred, "as old as the story of Adam, on whom God laid the penalty of work for disobedience." Indeed, work had been considered essential to the penitentiary ever since the opening of the first modern prisons in the 1820s. At Pennsylvania's Eastern State Penitentiary and New York's Auburn Prison, America's first penologists anticipated, in the words of historian David Rothman, that prison labor "would become not an oppressive task for punishment, but a welcome diversion, a delight rather than a burden." Work would rehabilitate the law breaker and prepare him to rejoin productive society. These pioneering penitentiaries created small-scale factories within prison walls, hoping the discipline of labor would make inmates choose to be productive citizen-workers on their return to freedom.2 [End Page 77]
Behind this benevolent scrim, however, the question of convict labor involved a thicket of political, economic, and moral conundrums. While penologists and criminologists claimed a legitimate interest in reforming criminals through work, the inescapable fact that prisoners' labor was unpaid, often coerced, and frequently served the financial interest of private manufacturers effectively transformed the convict into a kind of slave—an analogy made even by prison officials such as New York's Thomas Osborne. The fact that so many convict workers following the Civil War were African Americans, and that crime itself became so insidiously tied to white fears of blackness, underscored the exploitive and illiberal essence of prison labor in a society premised on contract freedom.3
The Thirteenth Amendment, ratified in the wake of the Civil War, only amplified this connection. While the amendment abolished slavery and "involuntary servitude," it provided one crucial exception: forced labor was legitimate if it served as "punishment for a crime." As legal scholar Jacobus tenBroek noted decades ago, the Thirteenth Amendment "has never lived up to it historic promise" to protect exploited laborers and expand individual freedom; in that regard, tenBroek concluded, the amendment was "insignificant." Shielded by the newly amended Constitution, states aggressively expanded ever-more exploitative systems of convict labor. Historians like Alex Lichtenstein, David Oshinsky, and Karin Shapiro have ably documented and analyzed the insidious growth of the convict-lease system in the New South, where a brutal form of quasi-slavery placed mostly black prisoners at the mercy of private employers who paid the state a usage fee far below prevailing wage rates for free workers. Convicts were a cheap and pliant labor pool for textile manufacturers, mine owners, and other private businessmen who discovered they had a vested economic interest in maintaining a steady stream of law-breaking bodies to risk in hazardous occupations. Given the vulnerability and social invisibility of convicts, it is hardly surprising that their working conditions were often dreadful, and sometimes lethal.4
But Brian Greenberg and Rebecca McLennan have importantly clarified that the problem of convict leasing was a national, not a regionally southern, phenomenon, and posed a major social and ideological challenge to the ideal of free labor. In fact, as Steve Fraser and Joshua Freeman have observed, the Thirteenth Amendment's protection for prison labor existed "precisely because [labor] had long since become the dominant form of punishment throughout the free states." During the "long depression" of the late nineteenth century (from the 1870s through the 1890s), northern prisoners produced over $35 billion worth of goods (in today's dollars). Little wonder, then, that workers and labor unions throughout the country understood convict labor as one of the greatest evils of the Gilded Age. For laborers, the possibility of a reserve army of unpaid convicts was the ultimate cudgel that employers wielded [End Page 78] against proudly free workers willing to strike to defend their dignity and assert their equal citizenship. Along with the eight-hour day, the abolition of prison labor was often the major political issue that galvanized independent labor parties in the nineteenth century.5
While historians have illustrated how reformers ultimately chipped away at the prison labor system during the Progressive Era, they have paid far less attention to the 1930s, which are usually framed as either a denouement from the bad old days of the 1870s–1910s, or an uneventful precursor to the revivified "prison industrial complex" that began stirring in the 1970s and sprawls incubus-like across the United States today.6 Such an omission is unwarranted. In fact, the early years of the New Deal marked a period of tumult, experimentation, high idealism, and unforeseen outcomes—a crucial pivot point for convict labor in the United States. From one direction, congressional lawmakers were, for the first time, aggressively attacking the worst abuses of convict labor. From the other side, states, which had become dependent on prison industries (notably in the South), scrambled to keep inmates at work. Even states that had adopted progressive prison reforms feared that the sudden collapse of all prison industries portended a crisis of discipline and a betrayal of the fundamental goal of rehabilitation. And in Washington, DC, a generation of lawyers and penologists reared on progressive principles faced an unprecedented challenge and a unique opportunity: to realize the rehabilitative potential of work and eliminate inhumane exploitation, all the while protecting the livelihoods of free workers from the threat of convict labor. In the New Deal, lawyers and penologists like James Bennett, Linton Collins, Sanford Bates, and Donald Richberg found a new set of tools to rework the problem of prison labor.7
In particular, they turned to the National Recovery Administration (NRA), centerpiece of the so-called first New Deal. The overall strategy of the NRA was to end the depression by eliminating overproduction, controlling prices, and raising wages. It thought to accomplish this by allowing sectors of industry to openly cooperate without fear of antitrust laws. Across the country, businesses developed "codes"—essentially, voluntary compacts, under the jurisdiction of federal officials—to govern their corner of the economy in a more "planned" and rational manner.8 Why, prison officials wondered, could the NRA not benefit the prison economy in much the same way as it promised to help the free economy?
The Prison Labor Code (PLC) that emerged from the NRA in 1934 seemed to embody many long-sought goals of progressive penology. It contained a series of measures designed to safeguard prisoners from abusive situations—prohibiting "dangerous" [End Page 79] working conditions, for instance, and mandating a forty-hour-maximum work week. It clarified that its conditions applied uniformly to all forms of prison labor and all prison products (although, critically, it exempted agricultural prison labor, which meant that inmates at places like Louisiana's infamous Parchman Farm remained without federal protection). Most significantly, the code expressly stipulated that prison goods could never be sold for "lower than the fair current price," thereby safe-guarding free industry and free labor from the undercutting effects of convict-made products. The centrality of this provision can be discerned in the final name of the code: "The Compact of Fair Competition in Prison Industries."9
The possibilities of the NRA and the PLC were exhilarating to prison activists. In one fell swoop, the code promised to eliminate abusive labor systems, while preserving the ideal that prisoners should work. It claimed to mollify both industry and organized labor by controlling prices. And, perhaps most importantly, it seemed to signify a movement toward the national standardization of prison labor without sacrificing the states' own role or authority in governing its prison systems. As a major investigation into the PLA would later note, reformers at the time felt that the compact "was going to achieve at one blow what fifty years of public agitation and striving had failed to secure."10 On April 19, 1934, Franklin D. Roosevelt signed the code, and, in the Prison Labor Authority, American prison labor found its first seminational governing structure. That the code entailed such fatal compromises, engendered such passionate resistance, and collapsed so quickly and completely illustrates the powerful undertow of Louis Robinson's question.
As the New Deal took shape in 1933, a series of deeper historical forces were converging that recast both the political economy and cultural meaning of convict labor. Since the 1860s, a network of reformers and unionists had consistently targeted the exploitations of prison labor. While southern states leased convicts directly to mines, railroads, and lumber mills, northern prisons employed a "contract" system, whereby private businessmen were essentially granted the authority to operate their factories within the prison; for instance, at the outset of the Civil War, a Bronx-based iron molder locked out his unionized workers before moving his factory to Sing Sing to exploit the contract system. In the 1880s, when the Knights of Labor revivified the labor movement, Knights assemblies often accrued considerable political influence at the state level, and in some states the Knights were able to modify or outlaw the contract system. Organized labor delighted in these few victories, but it is not at all clear that the new "piece price" system—in which wardens gained control over the prison environment but prisoners continued to produce "piece" goods for private contractors, who then haggled over the price—was an improvement. Indeed, a warden in an Ohio prison claimed that the "piece price system does more injury to the laboring [End Page 80] classes outside [than the contract system], because it enables the contractors to sell their products at lower prices.… The prison suffers and the workingman does not see that he is hoodwinked." It was not until 1894 that New York—the national model in progressive prison reform—finally enacted a system that prohibited the two contract models and envisioned a "progressive" model in which labor was considered in solely rehabilitative terms and not as a revenue stream.11
Still, the energies of reformers and labor activists were changing the norms of acceptable conditions. In 1915, President Woodrow Wilson's Commission on Industrial Relations publicly deplored the "evil results" of convict labor and urged its banishment from "any civilized community." Many states took heed. Some outlawed the use of convict labor or prohibited the sale of convict-produced goods. Others, such as Minnesota and Michigan, pioneered a "state's use" system that allowed prisoners to work but ensured that prison goods were produced only for state agencies, not for sale on the open market. Yet others, like New York, even devised modes of paying wages to working convicts in a reasonably fair and transparent manner. By the time the stock market crashed in 1929, public sentiment had already firmly turned against the private exploitation of prison labor; in the early 1930s, when all work was scarce, few supported the idea of creating jobs for convicts. At the same time, a new wave of socially consciousness "prison movies," such as The Big House (1930) and Mervyn LeRoy's 1932 film I Am a Fugitive from a Chain Gang, placed the audiences' sympathy squarely with the abused and ill-treated prisoners. In short, by the time that New Dealers converged on Washington in the early 1930s, a moral consensus had congealed among the public: exploiting prisoners for profit was wrong.12
On the other hand, several tributaries of public anxiety regarding lawlessness and crime were also converging into a river of social angst. Throughout the early twentieth century, brutalized and overcrowded prisoners responded with desperate riots. In 1929, thirteen hundred inmates at New York's Clinton Prison attempted to breach the walls and set fire to the prison. In Kansas, the prison coal mine witnessed four major insurrections between 1927 and 1937, including one that resulted in a four-day armed stand-off and another that involved over four hundred convicts holding twenty hostages. Similar incidents occurred in Philadelphia, Colorado, and other locations across the country. Many Americans worried that a "new breed" of criminal was filling the country's jails—one beyond rehabilitation and therefore beyond pity. Crime rates, which ballooned dramatically during the 1920s, fueled this public anxiety. Meanwhile, as Jeffrey Adler has argued, traditional sources of justice seemed incapable of enforcing the law and ensuring safety: police departments often proved corrupt, while juries acquitted sympathetic lawbreakers with disturbing frequency. Alarmed by the perceived breakdown of law and order, the federal government [End Page 81] ordered the first major federal study into law enforcement: the so-called Wickersham Commission, which submitted its reports to President Herbert Hoover in 1930. The Wickersham reports supported a significant strengthening of the national police force (the Federal Bureau of Investigation) and more punitive sentencing guidelines. If progressive reformers argued against convict labor on a humanitarian basis, many other Americans in the early 1930s were prepared to reject convict labor because it did not seem punitive enough.13
Finally, federal law itself was rewriting the rules of convict labor in the early 1930s. A decade earlier, spurred by manufacturers in the garment industry, the government had initiated a major study of prison industries. The final report, which severely criticized the use of convicts in the existing prison economy, proved divisive and controversial. (Sanford Bates, director of the federal Bureau of Prisons during the New Deal, refused to sign it, arguing that the report ignored the rehabilitative and disciplinary value of prison labor.) Nevertheless, the report persuaded Congress to pass the Hawes-Cooper Act in 1929. The law, set to take effect in 1934, did not directly assault convict labor but intended to pressure the system by targeting the interstate trade in prison-made products. This strategy was not new; in fact, Congress had considered and nearly passed similar legislation as early as 1888, though political momentum for that law abated following the collapse of the Knights of Labor. Hawes-Cooper stipulated that any goods produced by convicts in one state would, on shipment across state lines, immediately become subject to the laws of the receiving state. Therefore, if the receiving state had a law that prohibited prison-made goods from being sold on the open market, the imported goods would be banned from sale.14
The ultimate goal of Hawes-Cooper was to incentivize states to finally end the practice of convict leasing and to instead embrace what was known as the "state-use" system. The idea of state-use was straightforward. Rather than leasing convicts out to private contractors, or using convicts to produce piecework that could be sold to private merchants, states would develop carefully selected industries directly on prison grounds. There, prisoners produced goods needed for the maintenance of the prison or required by a state agency: items like work shirts, license plates, or simple tools. These prison-made goods were limited only to products not otherwise produced by a "free" enterprise within the state, and could be purchased only by the state, not sold on the open market. The animating concept behind the state-use system was the decommodification of the prison economy—sundering it from the normal rules of market competition to protect both prisoners and private industry. This approach had been accepted, as one memorandum to the solicitor general observed in December of 1933, as "a fair, humane, and reasonable compromise by both labor [End Page 82] and capital." General trends in prison management reflected this fact. In the 1880s, only 25 percent of states used the state-use system, but by the early 1930s, this number was up to 65 percent, while the convict-lease system withered. For the experts of the New Deal—committed to economic rationalization, efficiency, coordination, and a tripartite relationship between businesses, labor, and the state—the state-use approach to prison labor seemed ideal.15
The move away from convict labor and toward a state-use system seems to confirm the broader liberal narrative of American progress; it comforts us with the notion that humanitarian impulses won out in the end. The reality is less sanguine. For one thing, prisoners' working conditions at places like San Quentin's jute mill, Folsom's rock quarry, or Texas's enormous prison cotton plantations remained brutal and dangerous, and reflected harsh hierarchies of racism and violent masculinity.16 Quite apart from the fact that prisons in the 1930s could hardly be described as cradles of human kindness, the "progress narrative" also obscures the basic centrality of coerced labor to even the most reformist vision of prison discipline. In fact, the ascension of the state-use system (and, with it, the Prison Labor Compact) did not mark the end of the idea of prison labor but its entrenchment at the heart of progressive theories of rehabilitation. Louis Robinson's own attitude toward prison labor was emblematic. All informed thinkers of his time, Robinson thought, agreed that prisoners should not be exploited or abused, nor should their labor threaten the livelihood of free workers. But without productive work, Robinson concluded, the reform of criminals was impossible. Prison work, as documents of the time often framed it, "is primarily therapeutic" and rehabilitative. In the context of the penitentiary, reformers believed that prison labor's fundamental character was psychological and social, not economic.17
From this perspective, the main crisis in the prison system was not labor exploitation but "unemployment." When the Department of Labor surveyed prison labor in 1932, it estimated that, despite a growth in the inmate population, the proportion of prisoners who worked had shrunk from 75 percent when the department first investigated in 1885 to a mere 52 percent. James Bennett, a leading advocate of prison labor within the federal Bureau of Prisons, averred that the real number was closer to 25 percent in 1934. A 1933 memorandum prepared for the US attorney general argued that prisoners' unemployment was a "serious problem," leading to idleness, depression, and discontent. Even the reformist American Prison Association agreed, arguing in 1931 that "without work, there is bound to be a deterioration and breakdown of prison discipline." As James Bennett emphatically put it in a letter to [End Page 83] the Commerce Department, the entire New Deal approach was "built around one fundamental axiom, viz: That prisoners must work."18
New Dealers therefore pursued two goals: to compel prisoners to work, and to have this labor performed within a state-use system. But the reality was that New Dealers in Washington, even with the implementation of Hawes-Cooper looming in late 1933, could not compel the adoption of a state-use approach. Punishment had always been primarily the business of states and municipalities, and prison conditions varied widely across regions and even within states. Howard Gill, a superintendent of the State Prison Colony in Massachusetts, assessed the law in 1931 (three years before it was implemented) and predicted that Hawes-Cooper would merely "shift the control [of convict leasing] from a small number of large contractors operating in a national market to a large number of small contractors operating in a state market." In other words, Hawes-Cooper simply balkanized states into those that used a state-use model and those that permitted private convict leasing; it did not actually dismantle convict leasing. Indeed, Hawes-Cooper reinforced a federalist model, in which each state pursued its own prison labor policy, at the very moment that New Dealers were hoping to rationalize prison labor according to national standards.19
Politics were at play on the opposite side of this argument, too. In state legislatures, politicians asked why law-abiding taxpayers should foot the bill to house dangerous felons in prisons if private contractors were willing to pay for the privilege of accessing an otherwise idle labor pool. The deep American tradition of federalism had splintered police agencies into hundreds of distinct authorities, each with the ability to arrest and imprison citizens, depending on the crime. And, in each of these jails and prisons, the rules and conditions governing prison labor differed. The resulting problems were predictable. While most state prison administrators wanted to diversify their prison industries and to achieve a greater national unity of prison labor policies, this had not happened before the New Deal because of, in the words of one report, "an inability to get the states to agree on a nation-wide program and the impossibility of operating efficiently small production units." The promise of the New Deal was the possibility for building national mediating institutions that could create uniform guidelines and transparent standards for the operation of the prison economy. Nevertheless, federal administrators faced a difficult question: how far could—or should—the federal government tread into the sacrosanctity of local control?20
Not only was federal authority limited in pushing states into a state-use system, but it could not even assure its more basic and central goal: keeping prisoners at work. Most states, these officials recognized, were flatly unprepared for the [End Page 84] implementation of Hawes-Cooper. Many federal prison administrators worried that the law would not only affect the interstate trade of prison-made goods but also undermine all prison industries—and with it, they felt, any chance of prison discipline or rehabilitation. Some states, particularly in the underdeveloped South, depended on the interstate sale of prison-made goods to finance their precarious, Depression-ravaged budgets and had no viable substitute on the horizon. These problems emerged immediately after the law's passage, when Alabama—one of the most notorious utilizers of the convict-lease system—filed suit in federal court against nineteen other states that had cited Hawes-Cooper in banning the sale of work shirts by a contractor who had used Alabama prison labor. As Alabama's attorney argued, the state had invested over $300,000 to build prison mills and relied on its share of sales from these work shirts to maintain its inmates at labor. More significantly, Alabama invoked the notion of federal authority over interstate commerce in an ironic twist that ultimately eroded federal authority over prison labor. It argued that other states could not challenge Alabama's prison-made exports because they could not regulate interstate commerce: that was a federal responsibility.21
There was a bemusing, if not cynical, irony in hearing Alabama, a state with a long history of antipathy to federal intervention in local affairs, argue against states' rights on the matter of prison-made goods. Meanwhile, penologists of the New Deal looked warily on the outcome of the Alabama case and the ultimate impact of Hawes-Cooper. As Alabama v. Arizona worked its way to the Supreme Court in late 1933 (just months after Roosevelt took office), another ironic twist emerged: lawyers working for the US attorney general urged the Justice Department to support Alabama, fearing that, if it lost, prison labor would be tarred as fundamentally inimical to society. (The Supreme Court ultimately denied Alabama's suit in 1934.) That liberal criminal justice reformers in Washington would go so far as to defend Alabama's convict-lease arrangement—despite their historical, philosophical, and moral opposition to convict leasing—illustrates the depth of their concern about the public's rejection of any sort of prison-labor programs.22
A sense of panic—even crisis—overwhelmed both state and federal penologists when they imagined a world without prison labor in the early 1930s. Above all, the NRA's Prison Labor Code was the product of this fear and the looming sense of crisis, as prison administrators imagined chaos and disorder exploding across the country in the absence of discipline-inducing labor. The Prison Labor Code was an eleventh-hour emergency measure designed to allow states the ability to dance around Hawes-Cooper, while also guaranteeing the rehabilitative potential of prison labor and advancing a measure of humanitarian reform. States that signed the code agreed to abide by its rules on prices, hours, and wages and to be governed by a board consisting of nine members (six elected by the participating states, and three appointed [End Page 85] by the President). As administrators pushed ahead in March of 1934, thirty-one states were poised to accept the NRA's leadership.23
New Dealers hoped that, with this balancing of interests, they could enlist the cooperation of business and labor leaders, utilizing the power of the state as a mediator and rule setter while maintaining a commitment to liberal constitutionalism. The position of organized labor was particularly interesting. Labor, after all, had a deep and abiding hatred of convict labor, and it now represented one of the most important voting blocs in Roosevelt's New Deal coalition. But PLA reformers spotted a small oasis of common ground. They noted that labor leaders despised the use of convicts to undercut free workers and undermine unions—but they were not insensitive to the life-affirming value of prison work per se. For instance, John P. Frey of the iron molder's union acknowledged, "No one would be absurd enough to claim that no convict should do any work whatever." PLA advocates were especially quick to point to American Federation of Labor president Samuel Gompers, who, in 1916, wrote to the Oklahoma State Federation of Labor that, "since human beings are compelled by nature to have some kind of self-employment in order to preserve mental sanity and physical health, the State or any other agency that would deprive individuals under its power of opportunities for employment would justly be regarded as inhuman and barbarous." The abiding heart of the criticism from both Frey and Gompers was with the contract and convict-lease systems, not with the general notion of work behind bars. Therefore, to the extent that they propelled the shift to a state-use system, PLA administrators garnered a slim measure of support from organized labor. Roosevelt's selection of United Garment Workers president Thomas Rickerts to serve on the PLA further signaled a promising direction: for the first time, a unionist actually had an ability to influence the shape of the prison economy.24
But if the architects of the PLA expected huzzahs, they were quickly and bitterly disappointed. From the beginning, it was clear that the ultimate success of the code hinged on its ability to separate the free economy from the prison economy and to prevent prison-made goods from undercutting free-market goods. Prevailing thinking held that this would best be accomplished in a state-use system. But, as with Hawes-Cooper, the code had no power to actively promote a state-use system. Nor did it have the authority to ban convict leasing. In fact, by allowing states that continued to rely on contracting to participate in the compact, the code effectively shielded those states and, arguably, perpetuated a despised system that had little momentum outside the world of the code. Moreover, the code allowed signatories to continue the practice of interstate trade in prison-made goods. To critics, the code was effectively an end-run around Hawes-Cooper.
Given these two serious flaws, charges of the code's failure were certain and emerged almost immediately. In Louisiana, tomato canners complained that the [End Page 86] Angola prison farm artificially depressed the market; where free-market canned tomatoes sold at $0.60 per pound, prison tomatoes sold at only $0.40. Worse, the quality of prison-made cans was so "sorry" that the state Charity Hospital would not even buy the prison-packed tomatoes; the produce was dumped on the open market, where it undercut the fruits of free labor. In Virginia, the operator of a rock company that ground gravel for state roads saw his orders drop by almost 80 percent after the state required that all contracts go to the lowest bidder; his company was undercut by the state penitentiary. "We do not believe you fully understand the situation," the president of the Belmont Trap Rock Company wrote to the PLA: competing with prison labor meant the end of his business and unemployment for his twenty-two workers. Likewise, in Illinois, strong evidence suggested that the state prison had put the Columbia Quarry Company out of business by selling lime dust at $0.60 per pound, when the minimum cost to even produce that amount was $0.80. Throughout the state, a local field agent found, small quarries "complained to me bitterly" about competition from prison labor.25
Problems emerged not only within states, but between them, as well. A Chicago saddle maker realized that sales of his whips were being undercut by the Bardall Company from West Virginia, which utilized convict labor. When the Chicagoan (named Schmidt) traveled to West Virginia to induce his competitor to raise prices to reasonable levels, Bardall did not repent his strategy but instead suggested collusion in producing illegal convict-produced goods—a prospect that would mean the end of Schmidt's shop and unemployment for his workers. In this case, the PLA was able to sanction the Bardall Company, but it also discovered frustrating loopholes that surely struck some businesses as Kafkaesque. The Trojan Wire Specialty Company, for instance, had operated for years in Troy, New York, but suddenly discovered that it could not match the prices of the Gatch Brush and Wire Goods Company, which operated out of the Baltimore city jail. This seemed to be a clear violation of the terms of the PLC, to which both New York and Maryland were signatories. But, as Linton Collins of the PLA responded, the code only policed labor at state institutions. The seventeen workers at the Trojan Wire Company who faced unemployment apparently had no recourse when confronted by convict labor at a municipal house of correction.26
More was at stake than just individual businesses, important though they might have been to the individuals involved. The efforts of reformers to develop a more nationalized approach to prison labor also touched on fundamental philosophical approaches to political economy. And, if the NRA seemed hopeful to progressives and liberals, it struck others as a dangerous, even tyrannical, example of the [End Page 87] centralization of power and enlargement of the executive branch. Henry Hanson, a representative of the marking devices industry, acidly made this point to PLA officials. His industry, Hanson observed, had pioneered the production of license plates, "only to see the sentimentalities of pseudo-criminologists combine with the distorted economics of self-serving individuals" who shifted license-plate production to prisons, thus forcing states to duplicate an industry that already existed. This inefficiency hurt both industry and taxpayers, Hanson claimed, and also led to substandard quality. Over 200,000 Illinois motorists were forced to return their prison-produced 1935 license plates after the plates rusted out in a matter of months. The disaster of Illinois's license plate production was more than an isolated instance of prematurely rusty metal. To Hanson, it also portrayed "one more example of the inevitable economic waste inherent in the attempt of government to exercise the proper function of private industry." The struggle over the fate of the PLA was therefore also a conflict over the proper function of government and the relationship between the state and the economy.27
No dispute illustrated this political conflict in starker terms than the virulent opposition the PLA attracted from the cotton garment industry. Federal officials responsible for the PLA felt that the hostility of garment makers to prison labor was vastly overblown and hypocritical; in 1931, only 604 prisoners were counted at work on 15,000 spindles, as compared to the nearly 29 million spindles employing about 330,000 free workers.28 Still, the textile industry had a potent political voice, and it carried deep historical reverberations. The emergence of the textile industry in the 1810s marked the dawn of America's industrial age; for nearly one hundred years, the spread of mills from bucolic New England to the resplendent New South propelled the nation's economic growth and spurred a wave of factory building that quickly outstripped most of Europe. Textile factories had served as the cradle of the American working class, revolutionizing both the experience of labor and the social identity of laborers.
But by the early 1930s, America's textile industry faced a pivotal moment. The cotton garment trade had been mired in its own depression for nearly a decade before the stock market crash of 1929. Tariffs prevented American manufacturers from selling surplus goods overseas. Shifts in fashion and taste increasingly favored newer products like rayon and nylon. Decentralization led to cutthroat competition and ruinous overproduction, which in turn forced wages ever downward. Technological and market changes were not the only challenges facing textile companies. In the spring of 1934, just as the PLA was being signed into law, over 300,000 textile workers across the country staged a massive general strike—the largest labor conflict ever to rock the industry.29
Faced with these challenges, the garment industry responded furiously to the PLA. The opposition from the Cotton Garment Code Authority (or CGCA, the [End Page 88] trade group organized under the NRA to regulate textile production) and private manufacturers flamed up just weeks after Roosevelt authorized the PLA. Ideological and material interests quickly coalesced on both sides of the matter. In May, the CGCA filed a formal complaint with Hugh Johnson, Roosevelt's choice to head the NRA, protesting the fact that prison-made goods would be entitled to their own NRA "blue eagle" badge. The "blue eagle" logo signified compliance with the NRA and was meant to unify business, labor, and the public around the early New Deal; for many, it symbolized the entirety of the first New Deal program. The blue eagle's "We Do Our Part" motto inspired a patriotic sense of togetherness and collective commitment in the face of the Depression, and signaled to the public that whatever item bore a blue eagle had pledged itself to the wage and price guarantees of the New Deal. Granting prison-made goods a blue eagle, in fact, represented a total reversal of nineteenth-century reformers' efforts to have a "prison made" label stamped like a scarlet letter on all prison-made products. The opposition of the CGCA, then, was both practical and ideological: practical because, without the badge, prison-made goods would be effectively barred from a code-governed marketplace; and ideological because removing the blue eagle would symbolically strip prison labor from the sociopolitical legitimacy that a code would otherwise confer.30
To Sam Lewisohn, chairman of the PLA, the CGCA's opposition smacked of self-interested special pleading. Implicitly threatening to resign if the PLA code was rescinded, Lewisohn wrote to Hugh Johnson that the PLA "was to be a realistic attempt on the part of the Federal Government to remedy such evils as have grown up in connection with the employment of prisoners." The efforts of "public spirited persons" to humanize prison labor, Lewisohn insisted, should not be sacrificed to manufacturing interests. Like earlier reformers, Lewisohn saw himself and fellow travelers in the PLA as being "above" special interests, working only for the broader social good, and locked in conflict with greedy capitalists. As a memo to the attorney general noted that just before the code took effect, the only major critics of the PLA seemed to be "employers and manufacturing interests and generally in such a way as to indicate that they were more interested in profit than in the question on employment [i.e., rehabilitative labor].… If many of the critics of this system should turn their attention to the outworn, unfair, and vicious systems in vogue in their own States they would find much more to criticize and remedy." Writing in Survey, James Bennett was even more critical. Bennett noted that, when labor costs were low, garment manufacturers were slow to invest in new technologies. As soon as wage and hour standards were introduced with the NRA, however, "the manufacturers called in the efficiency experts," and within a year operating speeds had nearly doubled, thus displacing prisoners' former market niche. "The greedy machines have swallowed the jobs formerly allotted to prisoners," he wrote.31 [End Page 89]
Assessing this situation, Sanford Bates noted, "We have infinitely more opposition from the manufacturers, who are interested in profits, than from the unions, who are interested in humanity." But the controversy over the blue eagle threatened a revolt from labor, too. When the PLA met in April of 1934, both Thomas Rickerts of the United Garment Workers and Sidney Hillman of the Amalgamated Clothing Workers agreed that prisons could not possibly meet the standards of the cotton garment code, and therefore prison products could never be given a blue eagle that reflected this code. When the PLA started affixing blue eagles under the Prison Labor Compact code in 1935, the United Garment Workers local in St. Louis erupted in protest. Petitions to remove the blue eagle from prison-made goods flooded in, accompanied by cartoons featuring unemployed workers staring wistfully into a prison; the caption read, "They Have Our Jobs!" Seeking organized labor's help in resolving this controversy, Linton Collins reached out to Joseph Briegel of the Chicago Federation of Labor. Briegel responded that, in learning of the conflict, he was more committed than ever to end prison competition with free labor. However, recognizing that taxpayers seem to want something for supporting the prisons, he puckishly suggested a solution: "That every article manufactured in prison that would in any manner compete with Free Labor… be piled up in a safe spot in the prison yards and burned as a funeral pyre. (The date set for these fires to be on Labor Day, as a reminder to the world that Prison Competition with Free Labor is a sacrilege to humanity)."32
The PLA made its own mistakes in 1934, fueling even more confusion and controversy around its program. During the late summer, President Roosevelt issued an executive order limiting the workweek in the textile industry to thirty-six hours, in an effort to prevent layoffs and curb overproduction. Prison mills, however, somehow did not receive the same order, and for several weeks prisoners were operating with a forty-hour week, "to the disadvantage of free industries," as the Bureau of Prison's James Bennett admitted. In retaliation, the CGCA amended its code to allow disciplinary action against any cotton garment manufacturer that shipped prison-made goods into a state that prohibited open sale of prison-made garments. On the surface, this decree simply seemed to mimic the thinking of the Hawes-Cooper Act; to the PLA, however, such an action constituted a "usurpation of the police powers of the States by the CGCA."33
By the autumn of 1934, the CGCA and PLA were in a state of virtual warfare. Erwin Feldman, attorney for the Needle Trades Association, claimed that tensions were at a "boiling point." Urging President Roosevelt to issue an executive order "prohibiting the further marketing of these [prison-made] products," Feldman [End Page 90] claimed that prison officials—signatories to the PLA compact—had openly violated its terms, inferring that they could not be paying the scales agreed to in the compact. Moreover, he argued, the garment trade had been misled when it expanded its operations in the expectation that prison production would greatly decrease after the passage of Hawes-Cooper, resulting in overproduction. Furthermore, there was pressure to reduce the workweek while maintaining wages. "Industry demands immediate action and can wait no longer," Feldman declared. But PLA officials, writing to the reformist judge Joseph Ulman late in 1934, argued that private firms were themselves responsible for overcompetition and low prices. They also suggested that firms were simply exploiting the "bugaboo" of prison labor to avoid the president's thirty-six-hour workweek order. The PLA had offered to work with the CGCA to set a "base price" reflective of market conditions, but the CGCA had a fundamentally unshakeable reason for refusing to cooperate: it simply did not view the PLA as socially legitimate.34
As James Bennett admitted to the American Prison Association in September of 1934, the "baffling problem" of prison labor was "at the crossroads." Just weeks after Bennett's hand-wringing speech, President Roosevelt intervened in the dispute by issuing an executive order creating a blue-panel committee to hold hearings and write a report on the crisis. Chaired by Baltimore judge Joseph Ulman, the so-called Ulman Committee was a living link between the progressivism of the early twentieth century and the coalescing liberalism of the New Deal. Ulman, a supporter of the state-use system who once described American prisons as "the last word in social stupidity," had been associated with prison reform for nearly twenty years by the time he was appointed chairman of the group. Frank Tannenbaum, the second panel member, was a noted historian of the Mexican Revolution with a background in radical politics and the author of a 1926 book that excoriated convict labor. The third and final member of the committee, W. Jett Lauck, was a veteran economist with long ties to progressive political figures in Washington and alliances with the labor movement (Lauck would soon become a key player in the nascent United Automobile Workers union). All three men were deeply invested in "disinterested" public service, but all were also widely experienced and well-known in reformist circles. They all knew of the depredations that had historically menaced convict labor, and, given the political mood in the mid-1930s, they were not going to respond with undue warmth toward industrial interests. In sum, the Ulman Committee's members fit the mold of their era as prototypic technocratic liberals who exemplified the spirit of the New Deal.35
Yet, despite the committee's political and philosophical homogeneity, its report on the conflict between the industry and prison labor was thorough and nuanced. As the Ulman Committee's members acknowledged, "We were dealing with a complex [End Page 91] problem," not a simple matter of right and wrong. Before producing the final report, over the course of five weeks the committee heard the testimonies of hundreds of witnesses, all with a stake in the fate of convict labor. From representatives of the cotton garment industry the committee heard "absolute disdain" for the idea of coordinating or regulating prison labor. "Right or wrong," the committee concluded, the garment industry was "prepared to fight on this issue to the bitter end" and had even gained support from unions and from progressive groups like the National Consumers League. The labor movement, meanwhile, absolutely rejected the idea that prison goods could bear a blue eagle, arguing that they were lacking any such "legal or moral right." Representatives of the American Federation of Labor (AFL) expressed no confidence at all in the administrators of the PLA, claiming that both society and morality were on free labor's side and that prison products were by nature "outlaw goods." Although the AFL recognized the state-use system as fair and humane, the report stated, it "declare[d] war to the death" on all other forms of prison labor.36
At the same time, the committee agreed, prisons faced "a grave crisis" because of the industry's and unions' opposition to their products. Without work, prisons stood on the cusp of "riot and bloodshed," and the whole prison system threatened to degenerate into a social wasteland filled with broken people and ruined lives. Without work, nearly all prison officials seemed to agree, prison served no rehabilitative purpose. But difficult questions regarding the specific nature of the PLA troubled committee members. There seemed little compelling reason, for instance, why prisons focused exclusively on garment production or allowed prison-made garments to be sold on the open market. Despite the insistence by code defenders that the PLA forced prisons to meet the same standards of hours, wages, and prices as free industry, these claims obscured "the obvious fact that compensation paid to a State by a prison contractor is not precisely the same thing as wages paid a worker for his family's support." Pointedly, the committee—staffed as it was by men with experience in the labor movement—raised the obvious but critical fact that prison officials were asking to manufacture products bearing blue eagles, which "can by no stretch of the imagination be said to have been produced by labor invested with the right to collective bargaining." In other words, prisons were suddenly asking to participate in a "free" economy with unfree workers, at the very moment that the labor movement stood poised to secure legal protections, such as collective bargaining rights, that had eluded it for a century. Hovering between the market and the state, prison labor remained a circle that seemingly could not be squared.37
By the mid-point of 1935, it was clear to everyone that the Prison Labor Authority was a doomed experiment. "Regretfully and with extreme reluctance," the Ulman Committee had concluded in November of 1934, the PLC had not and [End Page 92] could not solve the problem of prison labor, despite its idealism and good intentions. Ironically, in fact, the NRA and PLA had "inadvertently reversed" the trend away from commercialized prison labor by propping up the existing system. "The true function of the prison," the committee declared, "is neither to make profit for private contractors nor to make profit for the State." If the prison did not rehabilitate its citizens, then it had failed in its fundamental social purpose and must be either reformed or abolished. The PLA's lack of support from Ulman and his colleagues meant a new approach was inevitable. A few months after their report's release, the Supreme Court ruled unanimously in its famous Schetcher decision in May of 1935 that the entire National Industrial Recovery Act was unconstitutional. The NRA and all of its codes—including the PLC—were on their way out.
The Prison Labor Code might well be judged a failure: a quickly aborted sub-agency that rightly disappeared into historical obscurity. However, the PLC stimulated significant political and cultural ripples that contributed to prison labor's evolving place in American political economy. For one thing, the shortcomings of the PLC prompted legislators into action, as they scurried to safeguard the intention of the Hawes-Cooper law. In 1935, responding to the furor surrounding the PLC, Congress passed the Ashurst-Sumners Act, which strengthened Hawes-Cooper's restrictions by turning any interstate shipment of convict-made goods into a federal offense, while the Walsh-Healy Act of 1936 outlawed convict labor on federal contract work. Even then, some businesses and unions remained unsatisfied because Ashurst-Sumners was effective only if the receiving state had a law on the books prohibiting the importation of prison-made goods. The amended Sumners-Ashurst Act of 1940 fixed this loophole, making it a federal crime to knowingly transport convict-made goods in interstate commerce for private use, regardless of laws in the individual states.38
If the PLC lit a fuse that finally ignited an explosion of congressional legislation, it also led to a vast expansion of public knowledge about the actual conditions of labor in prisons across the country. Perhaps the key conclusion of the Ulman Committee was the discovery that almost all of the existing data on American prisons was obsolete. The Department of Labor had sporadically documented prison labor since the mid-1880s, but such information was unhelpful in understanding the situation in the 1930s. Indeed, if society had broadly agreed that the state-use system and its program of rehabilitative labor should be guiding norms, then much of the existing information, based as it was on scattered studies of a predominately convict-lease system, offered limited guidance for either professional criminologists or the general public.
Little wonder that, as the PLC disappeared, many of its officials and supporters (including Louis Robinson and Joseph Ulman) filtered into the Prison Industries Reorganization Committee (PIRC) and began the most systematic national survey of prison labor ever attempted. From California, PIRC investigators criticized the woeful overcrowding at San Quentin prison (the nation's largest) and warned that, without [End Page 93] a strongly diversified state-use system, the culture of the prison "will be reduced to the most bitter, unregenerate and anti-social of the prisoners." From Kansas, they warned that the state prison seemed like a tinderbox, having witnessed "several spectacular prison breaks, mine riots, and other disturbances" in recent years; they urged an end to overcrowding, the closure of a massive and dangerous prison coal mine, and stronger reliance on prison farms. And, from Georgia, investigators noted the inmate population's racial disparity (of the state's 4,437 inmates in 1936, 3,349 were black), the use of humiliating chain gangs on public works, and the sometimes-horrific holding pens (which, investigators urged, should be destroyed "in the interests of decency and humanity"). Under the scrutiny of federal investigators, stirrings of reform were evident. The very fact that a Georgia governor opened the state's prison system to federal investigators and subsequently pledged to modernize it illustrates the 1930s-era move away from the idea of prison labor as a valuable commodity and toward a model focused on rehabilitation. While fair-minded and fact-based, the reports were by no means value neutral. Indeed, the thirteen detailed studies conducted by the PIRC created a pool of knowledge-as-advocacy. The entire purpose of accumulating facts and expert recommendations was to eliminate inhumane conditions, generate greater unity and efficiency, and push states toward robust state-use prison economies.39
Given this legacy, the Prison Labor Code echoes oddly but urgently in the context of the early twenty-first century. Just as many political values of the New Deal era came under assault with the emergence of the New Right in the 1970s, so, too, did many ideas about crime, punishment, and the economy of prison labor. In part, this shift reflected the tumultuous 1960s, an era of not only rising crime rates but also deep social unrest and numerous urban uprisings. As historians such as Michael Flamm have demonstrated, conservative politicians adroitly associated lawlessness with liberalism (and, by extension, with black civil rights), promising a return to "law and order" with a harsher and more punitive approach to (black) crime. Meanwhile, as historian Elizabeth Hinton has recently argued, much of the "war on crime" was ironically rooted in liberal policies of the Great Society. Driven by forces on both the Left and the Right, the American prison population exploded at mind-boggling rates. In 1970, fewer than 500,000 people were imprisoned; by 2014, that number had swollen to 2.5 million (even as crime rates declined), making America the largest penal state in the history of the world.40
As attitudes toward crime shifted, so, too, did New Deal–era norms regarding the appropriateness of prison labor. Beginning in 1979, the Justice System Improvement Act—a piece of legislation drafted by a conservative think-tank—once again allowed for-profit corporations access to prison labor. This legal retreat from the premises of New Deal prison reform perfectly coincided with the expansion of [End Page 94] "get tough" anticrime laws and an enormously expanding prison population. Indeed, during the 1990s, thirty states legalized contract prison labor to private firms. In California, Proposition 139 even provided tax credits for firms who built plants on prisons. A handful of enormous, prestigious companies, such as AT&T and Microsoft, hired inmates in the early 1990s before backing away after receiving critical press coverage. Journalist Christian Parenti, exploring the issue in 1996, interviewed one garment producer who had moved his operation from a maquiladora in Mexico into California's San Quentin prison because of the lower operating costs. As Parenti warned at the time, prisons "are rapidly becoming maquiladoras in our midst." Bill Gates may not have wanted Microsoft customers to associate his fortune with convict labor, but many other types of companies—like textile producers—are much less vulnerable to bad press and much more sensitive to tight margins. As of 2012, over 600,000—and possibly close to a million—prisoners were fully employed, often earning between $0.13 and $0.32 an hour while working for private contractors.41
In the unraveling of ideas and efforts associated with progressive penology, including even the flawed efforts of the PLC, Louis Robinson's 1931 question has reemerged with haunting vigor. But his query—"should prisoners work?"—is not simply about whether society should keep prisoners occupied. As the history of the PLC indicates, it really means, "Should cheap convict labor be exploited by cost-cutting governments and/or profit-seeking corporate entities, or should it be a tool for rehabilitating and empowering lawbreakers?" The question, ultimately, is about the very nature of our society.
MATTHEW PEHL is an associate professor of history at Augustana University in Sioux Falls, South Dakota. His book, The Making of Working-Class Religion, was published in 2016. He is currently working on a labor history of police officers during the "urban crisis" of the late 1960s and 1970s. He welcomes comments at firstname.lastname@example.org.
7. One notable exception is Ethan Blue's fine social history, Doing Time in the Depression.
9. Documents relating to the Prison Labor Authority (PLA), including the compact itself, are contained in Records of the National Recovery Administration, National Archives of the United States, College Park, MD, RG 9, entry 334, box 1 (hereafter NRA). Where possible, folder titles will be indicated.
10. "NRA: Report of Committee on Competition of Products of Cotton Garment Industry with Products of Prison Labor…" November 26, 1934, NRA, box 1.
12. U. S. Commission on Industrial Relations, Final Report, 234; Weyand, "Wage Systems in Prisons." On the impact of prison movies, see Doherty, Pre-code Hollywood, 157–70.
15. "Committee to Promote the State Use Prison System of the National Committee on Prisons and Prison Labor," February 28, 1935, in NRA, box 6, folder, "Prison Labor—Publications"; "Memorandum for the Solicitor General: in re: State of Alabama v. State of Arizona et al.," December 22, 1933, NRA, box 1; for numbers on the state-use system, see Bureau of Labor Statistics, "Prison Labor in the United States," 2.
17. Quote on therapeutic labor in "Comments of the Prison Labor Authority on the Report of the Ulman Committee," February 8, 1935, NRA, box 1.
18. James V. Bennett, "Prison Labor at the Crossroads," September 18, 1934, in NRA, box 1; "Official Bulletin" of the Prison Labor Authority, November 1, 1934, NRA, box 1.
20. "Memorandum for the Solicitor General: in re: State of Alabama v. State of Arizona et al.," December 22, 1933, NRA.
21. Alabama v. Arizona 291 U.S. 296 (1934).
22. "Memorandum for the Solicitor General: in re: State of Alabama v. State of Arizona et al.," December 22, 1933, NRA.
24. Gompers is quoted in an unattributed but thorough "Memorandum for the Solicitor General: in re: State of Alabama v. State of Arizona et al.," December 22, 1933, NRA, box 1; John P. Frey, "The Competition of Convict Made Goods," in NRA, box 6, "Interpretations-Publications."
25. William J. Kreb to Code Authority Canning Agency, TL, May 3, 1935, NRA; Belmont Trap Rock Company to Linton Collins, TL, September 26, 1934, NRA; E. K. Krause to V. P. Ahearn, TL, October 10, 1934, NRA.
26. Wayne Dismore to Linton Collins, TL, November 7, 1934, NRA; Linton Collins to Clark Cipperly, TL, April 15, 1935, NRA.
27. Henry Hanson to Linton Collins, TL, April 8, 1935, NRA.
28. "Memorandum for the Attorney General: Prison Labor Policies," September 22, 1933, NRA.
30. On the movement for the "prison made" label, see McKelvey, "Prison Labor Problem," 258.
31. "Official Bulletin" of the Prison Labor Authority, November 1, 1934; Sam Lewisohn to Hugh Johnson, TL, May 25, 1934, NRA, box 1; "Memorandum for the Attorney General: Prison Labor Policies," September 22, 1933; Bennett, "American Prisons."
32. Sanford Bates, "Should Prisoners Work When Other People Cannot" (n.d.), NRA, box 6; "Minutes of the Meeting of the Prison Labor Authority, April 13, 1934," NRA, box 4; Joseph Briegel to Linton Collins, June 6, 1935, NRA, box 4, folder, "Prison Labor General."
33. James Bennett to Herbert Rose, TL, July 9, 1934; "Official Bulletin" of the Prison Labor Authority, November 1, 1934.
34. The Cotton Garment Record, December 28, 1934, NRA; Howard Gill and James Bennett to Joseph Ulman, TL, November 22, 1934, NRA.
35. Bennett, "Prison Labor at the Crossroads"; for Ulman quote, see "U.S. Penal System Scored as 'Stupid,'" New York Times, May 24, 1936. On Tannenbaum, see Merrill, "Even Conservative Unions Have Revolutionary Effects."
36. "NRA: Report of Committee on Competition of Products of Cotton Garment Industry with Products of Prison Labor…" November 26, 1934.
37. "NRA: Report of Committee on Competition of Products of Cotton Garment Industry with Products of Prison Labor…" November 26, 1934.