Never again, swore Progressive reformers in the early twentieth century. Devotees of efficiency, they were appalled at what they saw as the fiscal and human waste created by a Civil War pension system that had consumed more than forty percent of the federal budget in the late nineteenth century [1]. Never again would they rely on a “blundering, plundering, endless, happy-go-lucky pension policy” that reduced veterans to “parasites when they are able to be self-supporting” [2]. Never again would they do “everything possible to make the cripple a failure” [3]. As these Progressives saw it, Union army casualties had simply been “cobbled up as well as the surgery of 1863 could do it, given a pension, and turned loose” [4].

Progressive reformers had a better idea, one that was “honest, efficient, patriotic, and businesslike” [2]: occupational rehabilitation. This approach, thought to have originated among physicians and philanthropists around the turn of the twentieth century, became national policy in the wake of the First World War. The premise of these early vocational laws was that a person with a disability might achieve acceptance into the larger community by “overcoming” the impairment with restoration and then obtaining employment [5]. The reformist plan for veterans of the First World War would thus help them avoid the fate of Civil War soldiers. Progressives proceeded with their own program of expert rehabilitative assistance to return people with disabilities to work. Disability compensation would still exist, but it would be tied to occupational rehabilitation.

This article ventures beyond prior studies of these eras and their assumptions by investigating the actual employment effect of the federal government’s benefits for Union veterans. We draw on physical examinations in Union army pension records and on items in census returns. We also make use of additional information derived from the Union army samples designed by the Center for Population Economics at the University of Chicago (“CPE”).

Our two central questions stem from the criticisms Progressives leveled at federal pensions and other forms of assistance for Union veterans. Did they discourage gainful employment by rewarding veterans for not working? Or did they serve as a safety net for ex-soldiers with disabilities? Our findings evidence the complex and contradictory impact of the public–private partnerships for Union army veterans and veterans of the First World War on occupational rehabilitation and employment. These efforts, based on different notions of disability rights, occurred well before modern notions of disability rights under the Americans with Disabilities Act of 1990 (“ADA”) and its central accommodation principle [5, 6].

Medical to Rehabilitation Models of Disability

Developments in veterans’ benefits were influenced by the “medical model” of disability—the presumption that disabilities could be rectified by approved medical intervention [5]. By the turn of the twentieth century, the medical model’s dominant emphasis was on transforming bodies [7]. It focused primarily on the individual whose disability precluded participation in the economy and society. The government was to provide resources to enable such individuals, if worthy, to “overcome” their impairments [5].

The Progressive policy, embodied in the Smith-Sears Act of 1918, built on the “wonderful strides [that] have been made in the science and art of restoring maimed men to physical soundness” [4]. Employed under supervision in “curative workshops,” wounded veterans would learn the skills and self-confidence “to become again a man among men” [4, 8].

The Smith-Fess Act of 1920 extended this approach to civilians. Both of these acts, and some subsequent laws, promoted the reintegration of persons with disabilities into civil employment. They offered rehabilitative services to people “who, by reason of a physical defect or infirmity, whether congenital or acquired by accident, injury, or disease, [are], or may be expected to be, totally or partially incapacitated for remunerative occupation.” The Randolph-Sheppard Act of 1936, for instance, created a federal program to employ qualified blind people as vendors on federal property [5].

A second type of law conferred monetary and other benefits on groups of persons with disabilities. The Social Security Act of 1935 established a federal and state system of health services for “crippled” children [5]. In 1954, the act was amended to provide monthly benefits for eligible workers who acquired disabilities. It was amended again in 1972 to provide benefits to limited categories of poor persons with disabilities [5].

The certitude of Progressive advocates in their reforms implies that rehabilitation and modern ADA workplace accommodation principles were unknown before the twentieth century. But one should not mistake their conviction for knowledge. The Civil War era, in fact, witnessed unprecedented public–private efforts to facilitate employment for veterans with disabilities. Congress required preferential treatment of veterans with disabilities in federal hiring and encouraged preference in private employment [9]. Schools in the new national system of soldiers’ homes trained clerks and teachers, and workshops taught “new trades suited to [residents’] particular disability” [10].

Philanthropists funded employment agencies whose benefits included temporary allowances, educational placement, and tools for “sick men who could gain a trifle by working at home” [11]. Commercial colleges, including the Illinois Soldiers’ College, actively recruited veterans for retraining [12]. And public and private sectors collaborated on prosthetics. The federal government purchased prosthetics from manufacturers until 1870, afterward offering veterans a subsidy of $50 for an artificial arm and $75 for a leg [13, 14].

To be sure, these endeavors fell short of an all-out campaign for veterans’ return to work, and they were not mindful of the barriers to work the disabled encountered. Multiple agencies administered the policies, and doctors and politicians viewed productivity as a matter of individual character [12]. Moreover, veterans’ aging shifted their needs from employment to income support. In 1904, President Theodore Roosevelt issued an executive order providing that old age itself was a “disability” covered by pension law, even if no medically disabling cause was claimed. In 1907, Congress authorized pensions based solely on a veteran’s age and length of military service [15].

Nor did the employment efforts obviate the basis of reformers’ critiques, which rested in particular on the requirement that pension recipients be unable to “perform manual labor” [16]. These criticisms and the caveats noted here have a common feature: they depend far more on exegesis of policies than on the real-life experiences of the men affected.

Recent scholars have begun exploring the actual work lives of Civil War veterans with disabilities. Some draw on contemporary accounts of postwar unemployment, describing civilian ambivalence about veterans’ employability, fruitless job searches, and ex-soldiers begging on street corners [17,18,19,20]. Other studies note disabled veterans’ transition to clerical occupations and suggest that most soldiers returned to meaningful employment [12, 21]. This article expands the focus on ordinary veterans by investigating recently available data on Union army recruits’ life course and the circumstances of civilian peers.

Data and Operationalization

Data for addressing these questions come from two large samples, one of Union army recruits and a comparison sample that links individuals from the 1870 federal census to the 1880 enumeration. The Union army sample combines several datasets designed by the University of Chicago’s CPE, mentioned above. The samples are part of the Early Indicators of Later Work Levels, Disease, and Death (“EI”) project. Researchers began with 331 randomly selected infantry companies, eventually expanding the project to comprise 427 companies of white recruits and 169 companies of the U.S. Colored Troops. Extant military and pension records were coded for all soldiers in these units, with ongoing linkage to U.S. censuses [22]. This article also relies on an EI-compiled sample of survivors of the Andersonville prison. This latter sample consists of men who lived to 1900, as does some of the census linkage in other EI samples, so the article’s analyses reflect a modest survivor bias. The EI samples include different sampling rates, with oversampling of urban recruits and African Americans; all analyses in this article are weighted to match Union veterans’ geographic distribution in the published summary of the 1890 federal census [23].

Our second resource draws on data created for the Integrated Public Use Microdata Series (“IPUMS”). Having collected samples from sixteen U.S. censuses, compilers at the University of Minnesota produced seven pairs of linked subsamples, one of which includes individuals who appear in both the 1870 and 1880 enumerations [24,25,26]. For comparability between veterans and the general population we selected all men aged 35 to 60; approximately 95 percent of the EI sample veterans were between these ages in 1880. The IPUMS sample includes veterans as well as civilians (veterans were approximately one-fifth of men ages 35 to 60 in 1880) [27]. Analyses in this article employ case weights assigned by IPUMS researchers to compensate for biases in linking individuals between censuses.

The 1880 census is this article’s main focus because veterans were then in the typically most active work years, and because it included a unique indicator. In 1880 (and 1890, whose census returns were mostly destroyed by fire), enumerators were to ask if each individual was “maimed, crippled, bedridden, or otherwise disabled” [28]. The terms were not further defined and the results were never published, but requiring the choice invited amputees and their families to evaluate their condition.

This variable hints at the meaning of impairment to men who did not write about their lives, but it hinges on whether enumerators and respondents took the question seriously. Other evidence suggests that they did. Exceptions existed (and will be discussed below), but veterans who wrote about their condition tended to answer the census question accordingly.

Eli Watkins, who responded to a physician’s inquiry by declaring that he was “allmost entirely disable,” had been reported as crippled in the 1880 census [29]. Ira Broshears, whose amputee penmanship essay reported that he was “a cripple with a broken constitution hastening probably to an early grave,” was likewise listed as disabled [30]. On the other hand, Joshua Lawrence Chamberlain, wounded six times and given the Medal of Honor for bravery at Gettysburg’s Little Round Top, “was committed to appearing and acting as able-bodied as possible” and kept his severe condition from the census-taker [31]. John Robinson wrote that he was able to “chop my own firewood harness and drive my own team and do all the general labour on my farm,” despite a missing arm [32]. There is no entry in Robinson’s disability column.

The occupational indicator, the response to the census question on “profession, occupation, or trade,” is less straightforward than it may seem. This query reflected the traditional objective of distinguishing the labor force from those who were considered non-productive. Census designers cooperated in merging occupation and identity: enumerators were instructed to “tell intelligibly what [a person] is” (emphasis in original) [33].

Yet our more familiar concept of work irregularity was emerging at the same time, spurred by the Panic of 1873. That year’s financial collapse reverberated throughout the economy, causing business failures, wage cuts, and layoffs [34]. Authorities initially scoffed at reports that one-third of Americans with occupations had no work, but they came to acknowledge a distinction between identity and employment [35, 36]. Federal census administrators introduced a question about months unemployed in 1880 (but inconsistent collection precluded publication), and states began to collect data on unemployment [36]. Yet most Americans would have accepted the older concepts of “occupation” and “employment,” and this article treats them accordingly. When individuals responded with “no occupation,” “none,” “retired,” “disabled,” “at home,” or gave no answer, they acknowledged separation from the labor force.

The dependent variable derived from these responses reflects three possibilities for labor-force participation. Men who reported an occupation in 1870 and 1880 are the reference group of consistent participants. Individuals who claimed no employment in 1880 had joined the ranks of those who, in the estimation of a state official, “take no part in the work of life” [36]. Yet there was another possibility—men who had no occupation in 1870 but reported employment in 1880. This alternative belied the traditional logic: having lapsed into indolence, individuals should not have regained the will to work, particularly at a period of economic depression. But some veterans and civilians did return to employment, and our analysis includes their behavior as a contradiction of prevailing assumptions.

Independent variables include four dichotomies. Racism constricted employment choices for African Americans in the late nineteenth century.Footnote 1 “The only place in the world of labor that the colored man can win,” wrote a social activist, “is the place that no one else wants. He may sweep down the subway steps, run the elevators in cheap apartment houses, act as porter in stores, where the work is heavy and the pay small” [37].

To assess the effect of older age on unemployment, we divide men in the samples at age forty. A study of contemporary workers finds a “marriage-unemployment gap” with higher rates of exit from the labor force among single men, so we examine the impact of having never been married [38]. We also include a dummy variable for self-reported physical disability in 1880. We use multinomial logistic regression for multivariate modeling.

Results

Table 1 shows univariate measures for the principal variables. When men reported employment, their leading occupations differed little by veteran status and disability. More than 40 percent were farmers, and approximately one-fourth as many were laborers (including farm workers). Other occupations were much smaller in number: carpenters, the next largest group, made up less than 5 percent of workers, and clerical work, a prominent opportunity for veterans with disabilities, accounted for 2 percent of ex-soldiers and smaller proportions of the IPUMS population [21]. Amputees have also drawn close scrutiny as exemplars of Civil War disability [21, 39,40,41,42,43]. They were predictably concentrated among “crippled” veterans, but five out of six of those citing disabilities had some other impairment.

Table 1 Univariate measures, EI+ samples of Union veterans and IPUMS++ sample of men ages 35–60 linked from 1870 to 1880 censuses
Full size table

Looking beyond specific occupations reveals contrasting prospects and behavior. Only 8 percent of men in the general population reported no occupation in 1880, but 13 percent of those self-identified as “crippled” were then without work. Nonetheless, their unemployment was far below the nearly 50 percent found among people with disabilities in an analysis of data from 2006, and undoubtedly much lower than in the current pandemic crisis [44, 45].

Being an ex-soldier added another set of distinctions. Veterans’ unemployment rate was lower than that of their peers. Given the scale of their disabilities—24 men per thousand in the EI samples claimed to be crippled versus 18 per thousand in the general IPUMS population—and the availability of government benefits allegedly breeding idleness, veterans seem more rather than less likely to have left the labor force. Their early job struggles were indeed genuine, but veterans benefited from a hiring preference that evidently extended well beyond civil service. For every report of ex-soldiers begging in public, newspaper advertisements hinted that joblessness would not last. Employers called for “disabled soldiers or other men of energy,” offered “great inducements to disabled soldiers,” and occasionally warned that “none but disabled soldiers and sailors need apply” [46,47,48].

Multivariate analysis is the next step in sorting out influences on veterans’ behavior and the consequences of federal pensions. Table 2 summarizes multinomial logistic regressions for the two samples, showing the average effect of each characteristic on the estimated probability of sample members’ appearance in an occupational category. The control variables function in ways both predictable and unanticipated. If all men had been single, their probability of being without an occupation in 1880 would have been 5 percent to nearly 10 percent higher than if all had married (probability changes of .048 for veterans and .094 for men in general). Yet single veterans and civilians also disproportionately returned to work after being jobless in 1870. This paradox is a reminder that instability is a significant dimension of labor-force participation; transition into and out of the workforce went with singlehood. Older men, on the other hand, showed a lower probability of resuming work. Influences of marriage and aging on unemployment were processes rather than single occurrences.

Table 2 Estimated effects on labor-force status, EI+ samples of Union veterans and IPUMS++ sample of men ages 35–60
Full size table

Race operated in a similarly complex way. Being African American had no statistically significant effect on returning to work in either group examined in Table 2. The impact of race on unemployment in 1880, however, seems to have diverged sharply between veterans and the general population. Net effects such as these are useful measures, but this apparent contrast illustrates the need for caution in comparing across samples. The opposite effects of race in the upper panel of Table 2 stem primarily from lower unemployment among white veterans. Probabilities underlying the net effects are actually similar for black men: their estimated probability of 1880 unemployment is .034 for veterans and .039 for African Americans in general (not shown). For whites, on the other hand, the unemployment estimates are .025 for veterans and .082 for all white men, producing a net positive effect of race for veterans and a negative for men in general. This observation leaves little evidence of employment instability among African American men, affirming a scholar’s suggestion that late-nineteenth-century racism “obliged [black men] to take jobs that were menial, yet as a rule, more stable” [36].

Table 2 shows a similarity and two key differences in employment experiences of veterans and their peers. The marriage and aging gaps affected both groups, as did the stability of African American employment. On the other hand, we have seen that veterans were in demand as workers, a distinction that provides perspective on employment stability and changes. And being “crippled” made another difference. Reported disability falls short of statistically significant effects on unemployment in the IPUMS sample, but it is the best predictor of 1880 joblessness among veterans. It would be surprising if the extent and variety of the Civil War’s imprint on veterans’ bodies had not disproportionately kept them from working.

But critics at the time saw it differently, blaming pensions for veterans’ supposed idleness. Condemnation would grow more strident after Congress ended the war-connected disability requirement in 1890, but attacks on pensions were already underway in the 1880s [15, 49]. Commentators accused policy makers of destroying veterans’ will to work by putting political pandering ahead of disability compensation. “The old soldiers have been corrupted,” a journalist declared. “They win pensions and live in idleness…. They become simple mendicants and worthless village bar-room idlers” [50]. Critics especially blamed lump-sum retroactive pensions, authorized in 1879, for “the demoralization of the ex-soldiers.” New payments “emptied [soldiers’ homes] for weeks,” the residents returning “little by little, suffering the miserable consequences of a long debauch” [51].

Table 3 addresses such allegations by examining the workforce implications of pension income. Detractors implied that veterans would readily abandon their usual occupation for the promise of a government subsidy. It is a simple matter to include pension amounts in a regression, but interpretation presents challenges. Federal military pensions were a combination of allowances for specific injuries and payments for medically determined functional disabilities. Most awards applied to the latter, based on examining physicians’ judgment of the extent of “disability for the performance of manual labor” [16]. This evaluation was meant to substitute medical expertise for self-diagnosis, but veterans’ own assessment of their disability is equally useful. Looking back to Table 1, a correspondence appears between pensions and self-reported disability, but they were far from identical. Fewer than half of veterans who claimed to be crippled had been approved for a pension before 1880. Some had failed to apply because they did not know about the aid, but most had disabilities originating after the war and were thus ineligible for a pension. To maintain the distinction between the two perspectives on disability, we report separate regressions incorporating pension income and census-reported disability.Footnote 2

Table 3 Estimated probabilities of labor-force status, EI+ samples of Union veterans
Full size table

The central question for this analysis is whether economic inducement or physical limitations influenced veterans’ employment behavior. Expectations offer guidance: pension recipients should have validated contemporary criticism by using their allowance to leave the workforce by 1880, and they should have avoided any return to work. Pensions should also have outweighed self-reported disability in inducing veterans to relinquish employment.

Table 3 reports estimated probabilities of labor-force experience for the two key variables. Focusing on probabilities themselves rather than net effects allows assessment of representative pension amounts. The models underlying the table control for the variables included in Table 2. Table 3 confirms the expectation about joblessness in 1880: the upper panel shows that the higher his monthly payment, the higher the probability that a pensioner was without work in 1880. It appears that a guaranteed income lured veterans away from the labor force.

But other findings defy suppositions. Work-averse pensioners should not have returned to employment after an interval of joblessness, yet the lower panel of Table 3 indicates that they did. As some pensioners were leaving the workforce after having been employed in 1870, others were reclaiming an occupation, with a probability that increased as the pension went up.

The last column of Table 3, substituting self-reported disability for pension income, further undermines the presumption of a work disincentive. The 1880 unemployment probability for “crippled” veterans overshadows those for pensions, except for veterans with especially large awards. For the 95 percent of EI sample members who received pensions of less than $18 per month, joblessness was more closely linked to this broad indicator of disability than to their government payment.

The linkage reverses for veterans with larger pensions. The estimated probability of unemployment in 1880 rises steadily with monthly payments, reaching an estimate of 73 percent for the $72 maximum (no EI sample members with pensions above $30 were re-employed by 1880, so larger pensions are excluded from the lower panel). These might be the men suspected of trading work for benefits, but they were hardly the ones targeted by contemporary commentators. A pension of $18 required a lost hand or foot or its equivalent; $24 awards were intended for arm or leg amputees or labor “incapacity”; those losing a hand and foot (or their use) qualified for $36 pensions; the $72 maximum was reserved for total blindness, loss of both hands or feet, or dependence on an attendant [16]. An editor pointed out that “a man who lost one arm or both arms, or one foot or both feet in the service may live 30 years, but would any one laugh at such a man if he claimed that he was disabled” [52]. On the other hand, another article declared that “rheumatism, rupture, and all of the long list of ‘ills that flesh is heir to’… have been made the basis of unholy, unscrupulous and shameless raids upon the generosity of a grateful people” [53]. Most veterans who sought pensions for illness or injury would have been eligible for an $8 rate; Table 3 suggests that this income was a poorer predictor of unemployment than was a self-assessment unrelated to pension-seeking.

Pension fraud was real enough [15, 49, 54], but nothing in our analyses points to widespread corruption and idleness among veterans. Our findings instead underscore the role of instability in the nineteenth-century work experience. The consistent presence of single men in both unemployment categories in Table 2 shows their tendency to move into and out of the labor force. Table 3 suggests that though their reasons surely differed, veterans with disabilities also disproportionately entered and exited employment. Whether measured by pension income or the broader self-report of being crippled, veterans with disabilities were frequently unemployed in 1880, but they were also more apt to have returned to work than were those without disability.

Implications

This article offers three unique perspectives on Union army veterans and the nineteenth-century workplace. First, the system of pension benefits illustrates the erratic advent and pursuit of the “medical model” of disability after the Civil War. The conviction that disability is an individual abnormality correctable through skilled intervention underwent fits and starts in the late nineteenth century. Policies for veterans were no exception. The federal artificial-limb program reflected the belief that prosthetics could repair amputees’ disability. “Those who cannot wear [a limb],” declared a senator, “are in a far worse condition than those who can. They are less capacitated for business or for labor than those who can wear artificial limbs” [55]. Policy makers also allowed for “overcoming” other disabilities. Congress included “recovery from disability” as cause for terminating a pension, and imposed biennial examinations to detect recuperation.

But experience undermined these expectations. Assistance with artificial limbs continued for the rest of the century, with disappointing results. Administrators reported in the 1890s that “few of those who were furnished with an artificial arm called for a renewal of it,” proving that “its usefulness is regarded as nil” [56].

Prosthetic legs were only moderately more successful. Fifty-two leg amputees in the EI samples related their experience with prosthetics when they were examined for increased pensions. Only one-third wore the limb regularly; their modal job type was that of clerk. These veterans’ prosthetic use appears to have affirmed the dictum offered by Oliver Wendell Holmes. “At an age when appearances are realities,… it becomes important to provide the cripple with a limb which shall be presentable in polite society, where misfortunes of a certain obtrusiveness may be pitied, but are never tolerated under the chandeliers.” The modal occupation type among non-wearers was skilled laborer. These amputees exemplified the “plain working-man” Holmes also described, for whom “an old-fashioned wooden leg” was “the best thing for his purpose” [57]. Appearance seems to have been at least as important as function in the prosthetic program.

The belief in overcoming disability was equally unrealistic. In more than 20 years of pension administration, officials dropped fewer than two percent of pensioners for “recovery” [58]. Biennial examinations were discontinued in the 1870s, but suspicions about recovery remained. As a member of Congress put it, “Who knows whether [a pensioner] has recovered or not? No one but himself. Many a man carries a pleasant face and an agreeable smile when he is enduring constant suffering and distress” [59].

Other initiatives at the time likewise demonstrate the tenuousness of a medical theory of “correcting” veterans’ disabilities. However limited in scope the approaches were, the adoption of veteran preference in federal hiring and the offering of separate workshops in soldiers’ homes were endorsements of adapting work opportunity to disability rather than the reverse. Models of disability also collided with politics: veterans’ advocates, administrators, and politicians wrestled with the meaning of disability (including a quandary over the distinction between “disability to perform manual labor” and “incapacity to perform any manual labor”) while pondering how best to turn monetary benefits into votes [15, 16, 49, 54].

As long as Civil War veterans remained a political force, accusations of corruption and demoralization gained little traction [49]. As suggested by the quotes at the beginning of this article, however, a new generation in a new century witnessing a world war discovered common ground with pension opponents. Agreeing with earlier detractors that the Civil War pension system was a massive waste of money and human capital, Progressive reformers vowed to devise an efficient program of support and rehabilitation for a new generation of veterans.

A final perspective calls attention to the agency of the veterans themselves. Answering census questions may not seem to be an act of self-determination, but some queries called for just that. Identifying an occupation allowed a degree of judgment, albeit constrained by the instructions quoted above. Most census items focused on facts: age, sex, birthplace, relation to household head, occupation, and so on. Nineteenth-century census officials could have attempted to factualize a disability question in the manner of modern enumerations, which ask about “long-lasting conditions” and refer to specific limitations on activities such as walking, climbing stairs, or working [28]. The post-Civil War enumeration dealt instead in adjectives: “crippled,” “maimed,” and “disabled.” The census did include items for those who were deaf, blind, or “insane,” but few veterans in the EI samples answered this question.

Given this scope for discretion, individuals (or informants) exercised judgment in ways that might puzzle a modern analyst. Michael Redmond lost his right leg at the Battle of Antietam. Physicians examining him for an increased pension concluded that Redmond’s stump was too short for a prosthesis; he reported that “he wore an artificial leg on or off for about a year, and has not worn it for 18 years” [60]. Redmond was listed “at home” in the occupation entry for the 1880 census, but not reported as disabled. On the other hand, Herman Koch, a former private in the 73rd New York Infantry, was rejected for a pension in 1881 when examining physicians pronounced him “healthy” [61]. Koch had nonetheless judged himself disabled in the previous year’s census.

Personal as these decisions were, they also exhibited tendencies. Likely because of the plentiful job opportunities alluded to above, veterans with and without disabilities claimed employment more often than did men in general. Sometimes veterans with disabilities gave up an occupation, but for reasons more complex than eagerness to live at government expense. Comparison of pensions and census-listed disability suggests that disability outweighed the value of benefits in most veterans’ occupation reports. Pensioners’ inclination to resume working after joblessness implies that disability undermined occupational stability, not the will to work. Judging from our findings, nineteenth-century Americans with disabilities wanted what their twenty-first-century counterparts want—work at a meaningful occupation [44].

Conclusion

Progressive reformers of the twentieth century, promoting a perception that has endured, focused on the supposed motivation-killing effect of pensions and on fraud and largesse when advocating for a rehabilitative approach for World War I veterans. Champions of empiricism, they seized on accounts of pension abuse and scarcity of reported rehabilitation. No strangers to moralizing either, Progressives appropriated the longstanding supposition that people would rather live on government handouts than work [62]. They proceeded confidently with their new program of expert rehabilitative assistance to return people with disabilities to work.

For those who were disabled, the pension scheme for Union veterans set new boundaries in their encounters with the state. The medical model on which the approach was based seldom questioned the environment in which people with disabilities were forced to function, nor did it consider their rights. The Progressive occupational rehabilitation policy then extended the medical model to embrace occupational rehabilitation to restore men to “physical soundness,” that they might learn or re-learn the skills of work, but similarly skirted questions of environment or rights. This is a dramatic change in perspective—from a medical status to be isolated or pitied, to rehabilitative potential to overcome or cure [63]. Both the medical and rehabilitation models, however, promoted the reintegration of persons with disabilities into civil employment, albeit from different perspectives.

Both models long predated the rise of the disability rights movement, which culminated in passage of the ADA [5, 64]. Only recently has this modern civil rights or “social” model of disability questioned the physical and social environment in which people with disabilities are forced to function. The social model of disability rights accepts that individual differences are to be accommodated as part of the human experience and as a matter of civil rights. At the heart of the ADA is the view that social institutions must remove attitudinal and structural barriers confronting people with disabilities [65,66,67].

Barrier removal remains a particular challenge in the occupational readjustment of twenty-first-century service members. Dissonance between military culture and that of the civilian workplace deters some veterans, especially those with psychological traumas, from requesting adaptations [68]. Whatever their circumstances, the ADA’s most prominent command for the equal employment of people with disabilities is the accommodation principle.