Fall 1984, Vol. 16, No. 3
Revolutionary War Pension Records and Patterns of American Mobility, 1780–1830
By Theodore J. Crackel
An American will build a house in which to pass his old age and sell it before the roof is on; he will plant a garden and rent it just as the trees are coming into bearing; he will clear a field and leave others to reap the harvest; he will take up a profession and leave it, settle in one place and soon go off elsewhere with his changing desires.
—Alexis de Tocqueville, 1835
Tocqueville's observations have been reiterated by each generation of American historians and social commentators. In fact, Oscar and Lillian Handlin were so impressed with this trait that they entitled their recent social history of Americans in the Revolutionary War A Restless People.1 The significance of this characteristic was first drawn out nearly a century ago by Frederick Jackson Turner, but long association with these notions has not been confirmed by detailed studies of mobility in early America. One can only surmise that the reason was lack of data. The federal census, the usual source, did not lend itself to mobility studies until 1850 when it began to record the state of birth of subjects. Until now no other source has proven to be an adequate substitute.
This essay is a status report on a continuing study of geographic mobility in post-revolutionary America. The data for the study is drawn largely from the Revolutionary War pension applications and associated papers and reports and supplemented by state and local records.2 A preliminary study of the mobility of some 13,500 Revolutionary War veterans from across the country revealed that, in the years after the war, more than half the pensioners (54 percent) had left the state in which they had lived at the beginning of the revolution.3 This was (and is) a remarkable level of mobility. For similar age groups in the years from 1850 to 1950, the rate fluctuated between 32 and 38 percent.4
This essay concerns a much more detailed study of some 1,400 New Jersey veterans, a roughly 50 percent sample. The basic source of data is, of course, the pension applications which have been generally described in an essay by Constance Schulz that appeared in Prologue last year.5 Only a word or two in addition needs to be said about them as a source for mobility studies. At a minimum, the records contain information on the point of entry into revolutionary war service and the location at which they resided at the time of application; that alone is more than we know about any other such extensive group for this period. For a few veterans this is about as much as one learns concerning mobility, but for most applicants the level of detail is much greater. In addition, other sources were consulted to provide more detail. For example, information from 332 tax rolls from 93 New Jersey townships in 13 counties were manually linked to data from the pension applications. Among those who did not leave the state, linkage was achieved with one or more of the lists in 80 percent of the cases; some 58 percent of these 1,400 pensioner files were successfully linked to the New Jersey ratable lists. Many of those who moved, it would seem, did so before they married or turned 21 when they were added to the lists. Where linkage was made we obtained both economic data, including the value of real and personal property owned, and locations in time and space that added to our knowledge of these men. For those who claimed to have moved, the census and some tax records in the states to which they moved were searched. Of those who moved and had not been linked to New Jersey tax records, over a third were successfully linked in one or more other states. To date over 70 percent of all applicant records have been successfully linked in one or more states. Extensive further linkage is anticipated when county property records are consulted.
The most comprehensive individual records dealing with mobility are those drawn from application files for pensions under the act of 1832. That year Revolutionary War pensions, which hitherto had been given only to regular Army veterans—the Continentals—or disabled veterans, were authorized for all who had served at least six months in any of the military forces during the war. For the most part this meant those who had served in the various state militias, though it also included naval personnel, state line troops, and certain contract civilians such as teamsters. The depositions taken to substantiate the required service are a remarkable record in themselves, as John Dann's book of eyewitness accounts of the Revolution drawn from them reveals.6 The depositions, however, are more than a collection of personal accounts of service—as fascinating as these can be. They are rich with data concerning Revolutionary War veterans and their families and a unique record of the life and time of this generation.
Among the extensive data these lists provide are the responses of the applicants to two questions posed to all who sought pensions under the act of 1832. The first question was "Where and in what year were you born?" It would be another 20 years—years too late for most of this generation—before the federal census (in 1850) began collecting such information. A second, three-part interrogatory was-from the perspective of one interested in geographic mobility—even more remarkable. "Where," they asked, "were you living when called into service; where have you lived since the Revolutionary War, and where do you now live?" The responses to the questions allow us to study the geographic mobility of this generation at a level of detail never before possible.
Over 15,000 men responded to these queries, often providing us with a year by year, state by state, county by county record of their moves in the fifty years following the termination of the war. A few examples will suffice to show the range of possibilities.
Michael Shurts of Lebanon township, Hunterdon County recorded that "at the time (was first called out I lived where I now live, and have lived there ever since, except about 9 years that I lived about 2 miles from there but in the same township of Lebanon."7 Those, like Shurts, who reported only moves within the county in which they had resided at the beginning of the war were counted as non-movers. His report is typical of this class.
The reports of those who moved were as varied as the moves themselves: Cornelius McCollum "lived in . . . Gloster [County, New Jersey] after the war until twenty-nine years since [about 1805] when he moved to the farm in Late township in Clermont county [Ohio] In which he has ever since resided and now resides." Isaac Johnson reported that "after Peace was declared he remained in New Brunswick [New Jersey] till the year 85 when he moved to the city and state of New York." He left New York City for a time but returned in 1824 and had lived there ever since. Samuel Skelly didn't apply for his pension until 1851. He had returned to his home in the town of Newark, In Essex County, New Jersey, at the end of the war, but moved to New York as soon as the British left the city. That would place his first move in late 1783 or early 1784. He later returned to Newark and while there was called out for militia duty in the Whiskey Rebellion. Upon returning, in 1794, he again moved to New York, but stayed only a few years. He reported in 1851 that he had lived over fifty years in New Jersey's Sussex County where he made his pension application. Asa Beach swore that "when he was first called into service he resided at Morristown, New Jersey, resided there till after the revolutionary war. In 1784, he thinks, he removed to Catskill, the state of New York. Lived there five years & then removed in 1790 to the town of Chaslton in the county of Saratoga, State of New York. Lived there till 1827 then removed to Milton in the same county [Saratoga] where he has continued to reside ever since & where he now resided." These files are typical of the class who moved. The level of detail in the 1832 applications is quite adequate to the needs of the migration researcher.
John Anson's file is not necessarily unique but it is rich in detail and may mark an upper limit of what researchers can reasonably expect to find. He reported that
since the Revolution and after being discharged he went home to the town of Maidenhead [in Hunterdon—now part of Mercer—county] and lived . . . about five years. [He then went on] a voyage to Sea, was gone one year, landed in Philadelphia, [and] went from there to Cambridge, Washington County, State of New York. [He] stayed about one year, went to Canada, lived about three years, came from Canada to Arlington in Bennington County, State of Vermont. [There he] stayed about 2 years . [He] went from there [back] to Cambridge N[ew] York, lived about 8 years, married, moved to Chenango County, New York in the town of Oxford, and lived there 18 years, [and] lost his wife. [He then] went to Camillus, Onondaga County, N[ew] York, [and] lived about five years.
(At this point Anson digresses a bit to tell of his service in the War of 1812 and of his capture at Fort Niagara in 1814.) After the war he lived for about four years in Erie County and then "went back to Camillus in Onondaga County [and then to] Cayuga County and [then on to the] town of Scott, Cortland County, lived about five years, from there moved to Perrysburgh [Cattaraugus County, New York] where he now lives and has for three years last past."
This is an extraordinary record, both in its extent and in its detail, but the researcher who follows the narrative closely might have caught an internal contradiction. Anson seems to arrive at the year 1820 before getting to the War of 1812. That is a small problem, however, because there is more than enough detail to cross-check his movements in the numerous state and federal censuses of New York for the period;8 and additional data is available in the county records. Few men moved as often as John Anson, but most moved more than once and many moved four, five, or six times.
A number of this study's tentative findings suggest that further analysis will yield significant new understanding of this period in American history. Among the important opportunities the study provides is the chance to examine some of the specific patterns and implications of post-Revolutionary American migration. Before examining these findings, the question of how representative the pensioners are of the adult male population at large must be considered. If these men are broadly representative of their generation, then their pension records and the linked data will allow researchers to study geographic mobility in the post-Revolutionary period in a degree of detail never before attained. Two questions suggest themselves. How well did pensioners represent veterans as a whole? and How well did veterans represent the general population they were drawn from?
The examination suggests that pensioners were quite representative of the whole veteran population. In fact, it seems that a very high proportion of the veterans who survived actually applied for pensions. This is true of both the Continentals who were initially pensioned in 1818 and the state militia men (and others) who were provided for in 1832. Between 80 and 90 percent of the eligible survivors must have submitted applications. In addition, there seems to be no particular bias in the composition of the element that did not apply—it does not, for example, represent the most wealthy 10 to 20 percent.9 The officers in the sample, the group that comes closest to a differentiated upper class, are found in all grades from ensign to general, and in total numbers proportional to those found originally in the revolutionary army. Moreover, when the property lists were required in 1820 to substantiate the claims of indigence, a significant number of applicants showed substantial holdings and a few appeared to be rather wealthy men. Roughly one third of the 18,000 on the list in 1820 were initially removed because of the amount of property they owned.10
Just as the pensioners were found to be representative of the veterans, the veterans also seem quite representative of the adult male population as a whole. To begin with, those who served the nation in the Revolutionary War must have numbered between 175,000 and 200,000, or 40 percent or more of all men who were of military age.11 Participation was biased toward the younger men; the research shows that over 50 percent of those aged eighteen to twenty-two must have served for at least six months.12 Several of these birth cohorts provided even higher percentages. Of those born in 1760, who were between the ages of fifteen and twenty-one during the war years, nearly three in every four seem to have served. When one discounts the numbers who remained loyal to the Crown—estimates run as high as one third of the whole population—it is clear that a very substantial portion of the available able-bodied young men in some age groups must have served in one manner or another.13
Samples drawn from the 1780s ratable lists of nine New Jersey towns tend to substantiate these findings and demonstrate how well the veterans mirrored the larger society. Of the sample, 39 percent could be identified as veterans. (This sample of the adult men of the towns includes some number—probably 20 to 25 percent—who were beyond the age of forty-five. A quick calculation shows that the veterans represented about half the men of military age.) The tax rates of these veterans identified them as very nearly a cross section of the whole population. Except for the wealthiest, (the upper 3 or 4 percent), veterans constituted between 37 and 43 percent of every rate group.14 Among the wealthiest ratables, the veterans numbered only three of thirteen, or 23 percent. The variance in this last group may simply reflect the likelihood that the civil leadership was drawn from it, diminishing the number available for military service. This supposition is bolstered by the relative distribution of veterans when measured in terms of land holdings. Here veterans represent a rather even proportion in each group from lowest to highest. Veterans constituted from 35 to 42 percent of each land holding group—from those with no land to those with very sizable holdings. This suggests that those who held large areas of improved land (the wealthiest farmers) were as likely to have served as the landless or the smaller land holders. It also suggests that wealthy men who did not perform military service had their holdings tied up in something other than improved land and were probably merchants or lawyers or members of other such groups that traditionally performed the civil leadership roles.
The veterans—Continentals and state militia taken together—do seem to have been quite representative of the whole adult male population, at least in economic terms. Because this study focuses on mobility, however, one more comparison is necessary. Were veterans somehow different in terms of mobility? Were they more mobile? Could they simply have been the most adventurous, or could service away from home (often in other states) have influenced them? No previous study has asked these questions of this generation. The pension files are of minimal help in this area because they provide no data base for the non-veteran population. The issue can be addressed by implication, however. The distribution of pensioned veterans and the male population of their age group as of 1830 can be determined. This calculation leads to the conclusion that in the trans-Appalachian regions populated from the original states, the pensioners account for very nearly the same proportion of their age group as they had half a century before on the other side of the divide.15 That is to say, non-pensioners (in fact, non-veterans) moved at approximately the same rate as the pensioned veterans.16 This fact rather strongly suggests that the movement found in the pensioner sample will be broadly representative of the base New Jersey population it was part of. The issue begs more definitive analysis.
The study of these New Jersey men has yielded a number of preliminary findings and promises many more. One finding of the study seems particularly interesting. It was noted, even in the process of data collection, that a large number of men reported moving in the years immediately following the war. A closer examination revealed the surprising extent of that early out-migration. Fully half of the men who would ultimately move made their first move before 1785—that is, within three years of the end of hostilities.17 This is more than 30 percent of all the veterans and (by implication) of all the men in the veteran age group. In the immediate postwar years the out-migration from the established regions of New Jersey approached and may have exceeded 10 percent a year through 1785. After that, the outward flow seems to have leveled off at a rate of roughly 3 percent a year for the next decade. The data suggest that nearly two-thirds of the veteran age population made one or more moves in the years after the war, and that the vast majority of those moves had occurred before 1795.18 Even the lower rate during the decade after 1785 is significantly higher than previously suspected. Jackson Turner Main, in his remarkable study of The Social Structure of Revolutionary America estimated that "perhaps 15 percent of the population left in the course of a decade" and that "probably 40 percent of the population moved during a few years' time."19 The substantial difference between the rate of mobility reported here and that noted by Main is actually understated, for he counted intracounty moves not included in this study.
Unlike the mobility which Main recorded, which, "insofar as it can be traced, ordinarily consisted of a removal from one town to an adjacent town, or perhaps a few towns away," our movers frequently covered substantial distances. Over half of the population ultimately made state to state moves.20 The pension files—retrospective as they are—let us take the full measure of this generation's mobility.
An explanation of this very remarkable postwar phenomenon certainly must lie in the war itself. One might suspect some pent-up demand, but there was a continuing flow of at least the magnitude reported by Main throughout the war years. It has been suggested that the war obliterated the legal restrictions on movement that the proclamation line of 1763 had established and that the Quebec Act of 1774 had tended to reinforce, but the proclamation line restrained few who wished to move in the years before hostilities began. The fact is that in most areas there was so much empty land east of the line that the line was hardly constraining. Where that was not the case, it was often violated. For that matter, the first moves of the New Jersey veterans seldom crossed the line, even after the war had obviated it. Their first moves were made into the empty regions east of the line that had long been available for settlement.21 The obvious answers have little explanatory power. Land hunger, pent-up demand, and the lifting of restrictions simply prove unsatisfactory in explaining the magnitude and extent of this vast out-migration.
The evidence suggests a very different hypothesis. This movement reflects not pent-up demand but pent-up tensions created by the revolutionary nature of the war and the radical nature of the rhetoric developed to support the cause: a pent-up force that both compelled and expelled the populace. Some historians who have found the Revolution a conservative affair have supported that finding by pointing up its failure to produce the aftermath of terror and reaction that has typified so many other revolutions.22 Robert Palmer, in describing the "ambivalence of the American Revolution," asserted that "Americans, when they constituted their new states, tended to reconstitute much of what they already had. They were as fortunate and satisfied a people as any the world has known."23
The findings of this study suggest that Americans may have been less satisfied than Palmer suspected, but that in lieu of violence, they vented their dissatisfaction by moving. The ability to uproot and move may have served to relieve post-revolutionary pressures. For many, the wilderness lay only a few days walk to the west. This ready access to underpopulated areas provided a safety-valve of seeming opportunity that rendered unnecessary the violent explosion that marked the aftermath of so many other revolutions.
The pension application files open another window into this generation. They offer us a unique source of data that can be used to address many questions. They give us a new opportunity to observe and draw understanding from many of the hitherto inarticulate. The pension records are one branch of our military history that promises to take us beyond battles.
Theodore J. Crackel is editor in chief of the Papers of George Washington at the University of Virginia in Charlottesville.
1. Oscar Handlin and Lillian Handlin, A Restless People: Americans in Rebellion, 1770–1787 (1982).
2. The Revolutionary War Pension Application Files are part of the Records of the Veterans Administration, Record Group 15, in the National Archives. The files are available on microfilm as Revolutionary War Pension and Bounty-Land Warrant Application Files, 1800–1900, National Archives Microfilm Publication M804 (1974).
3. Theodore J. Crackel. "Longitudinal migration in America, 1780–1840: A study of Revolutionary War Pension Records," Historical Methods. 14 (Summer 1981): 133–137.
4. 0nly in 1850 did the federal census begin to gather and record data on state of birth.
5. Constance B. Schulz, "Revolutionary War Pension Applications: A Neglected Source for Social and Family History," Prologue 15 (Summer 1983): 103–114.
6. John C. Dann, ed., The Revolution Remembered: Eyewitness Accounts of the War for Independence (1980).
7. This and similar quotes that follow are drawn from the pension application files.
8. In addition to the federal censuses of 1790, 1800, 1810, 1820, and 1830, there are state censuses for 1795, 1801, 1807, 1814, 1821, 1825, and 1835.
9. It is difficult to account for those who did not apply. Some, whose widows applied in later years, appear simply to have been too proud to apply for pensions which implied indigence. Others, like some who applied many years late, may simply have been so isolated that a convenient opportunity never occurred.
10. Some of these were restored in the next few years and all were again eligible for pensions as a result of the act of 1832.
11. These numbers include approximately 25,000 who had been killed or had died in the war.
12. The men in our sample generally had at least six months service, the minimum required for pensions by the act of 1832. The pensioners represent at least 50% of the men in these birth cohorts. Our number would be greater if all those who served less than six months were counted. Some who applied probably served less than six months but are counted in our number because they applied (even if rejected).
13. There has been some work to support the assertion that those who served in the Continental Army were something less than average colonial Americans. But if they were poorer, or more marginal, their militia brethren (drawn from the balance of society) must have constituted a group as unrepresentative in the opposite direction. When the two groups are merged we would expect to find a population much more representative of the whole than either group of veterans alone.
14. The ratables were not grouped on the lists but were arbitrarily categorized for the study as follows: those whose total tax rate equaled 1–49 pounds (58%, of the whole group); 50–99 (20%); 100–199 (15%); 200–299 (49%); and 300+ (3%).
15. In the few states that had previously been populated (Florida, Louisiana, and Mississippi, for example) the proportion of pensioned veterans is substantially lower.
16. That is not to say that they followed exactly the same patterns of movement. In an earlier report I noted a tendency of southern pensioners to move northward without any substantial countervailing move southward on the part of northern veterans. The population distribution just addressed suggests that non-veterans (particularly southerners) may not have followed this same trace. Veterans tended to be underrepresented in the southern tier of states and somewhat overrepresented in the most northerly. This finding suggests the need for a further detailed investigation, particularly of southern veteran groups.
17. This number is based on the rate at which men disappeared from the lax rolls or failed to appear on the rolls as expected at majority, indicating that they had already departed. This conclusion is buttressed by application statements and other evidence that establishes the date of the first move of 40% of the movers in these years. If even a small proportion of the number for whom the first move could not be dated fell into this period the 50% figure would easily be reached. My considered judgment is that the actual figure will be somewhat in excess of 50% when more definitive data are obtained on those whose date of first move is now unknown.
18. 0f those for whom we are quite certain as to whether or not they moved (about 95% of the whole), 63% can be counted as movers. Over 70% of the movers were on their way before 1795. I strongly suspect that most of the 5% whose move status remains unknown will ultimately prove to be movers and that most will have moved prior to 1795. Their files suggest movement. Those whose files did not suggest movement have almost all been resolved into the non-mover category. Typical of those that remain unresolved might be a widow's application filed in a state other than New Jersey. She obviously had moved, but there was no certain indication whether she had made the move before or after the death of her husband.
19. Jackson Turner Main, The Social Structure of Revolutionary America (1965), p. 193.20. Ibid., p. 183.
21. That is not to say that some large scale speculative schemes were not frustrated, but simply to point out that for those who wished to relocate there was land as readily available in those years as it would be in the years immediately after the war.
22. Benjamin Fletcher Wright, Consensus and Continuity, 1776–1787 (1958), pp. 1–4.
23. R. R. Palmer, The Age of Democratic Revolution, 2 vols. (1959), 11: 232.