For over thirty years now, social historians have examined nineteenth-century household patterns and living arrangements in an effort to understand how families supported and reproduced themselves over the life course These activities were important, for they sustained national populations in the past. The living arrangements of the elderly provide a fundamental insight into Victorian family values and behaviour. As families mature, life events in parents’ and children’s lives converge, such as death, childbirth, widowhood and marriage, and the needs of parents and children began to change. Confronting these changing needs could engender conflict, co-operation or both among family members. Examining the residential patterns of mature families illuminates the social mores and practices adopted when families must respond creatively to these various life course changes. For this reason, studying the household arrangements of the elderly and their children is a useful way to identify and understand family systems as a whole.

This dissertation examines the living arrangements of the elderly and their children in Victorian Canada and the United States. A comparative work, this study contrasts the processes of inter-generational co-residence north and south of the border. It is based on a fully integrated set of census data from 1871 Canada and the United States in 1880, and offers the first truly consistent international comparison of nineteenth-century household structure. The dissertation demonstrates that the Canadian elderly lived with unmarried children to a significantly greater extent than did their aged American counterparts. Conversely, the American elderly resided with unrelated persons in greater numbers than did the Canadian aged. These disparities were directly related to differences in the timing of Canadian and American sons’ and daughters’ departure from home. Differences in the life course transitions of youths were in turn a product of distinct demographic structures and economic opportunities and similar gender norms north and south of the border. Qualitative research on intergenerational relationships outside the household, as well as lateral ties with siblings, friends and neighbours, demonstrate the important role played by proximity in sustaining close, supportive relationships. This evidence confirms the importance of studying co-residence patterns within the household. It also suggests that collective emotional differences between Canadian and American families resulted from structural differences in their living arrangements.

North American historians have depicted the elderly and their living arrangements in various ways. Drawing together the Canadian and American literature on the history of the elderly as well as the history of family behaviour in general inspires questions about how inter-generational family patterns compared north and south of the border. Many American scholars, particularly those writing before 1980, emphasized the nineteenth-century elderly’s dependence on their children. By focusing on the impact of modernization on the elderly, David Hackett Fischer and W. Andrew Achenbaum set the agenda for historical research on the elderly, centering their own and subsequent debates on the assumption that the elderly's status in the family and in American society had declined over the past two hundred years. Fischer and Achenbaum, joined by John Demos, Daniel Scott Smith and Tamara Hareven disagreed about the causes and periodization of this change in status; their arguments are summarized in Carole Haber and Brian Gratton’s synthesis, Old Age and the Search for Security: An American Social History. However, American historians of the 1970s all founded their research upon the assumption that the status of the elderly had indeed declined by the end of the nineteenth century. Subsequent research on social welfare provision for the aged was modeled on the same question of status, producing top-down studies which focused on welfare providers’ attempts to control the elderly and delimit their identity. Like the cultural studies of the 1970s, early work on social welfare provision for the aged attempted to identify a single status for an age group. This entire exercise was inappropriate since the elderly as a group have always featured too many internal differences according to gender, race, class and ethnicity to fit into a single category. Furthermore, these historians' emphases on status and social control worked to deny the elderly's own agency.

The field of old age history in the United States finally began to transcend its preoccupation with declension when historians turned from cultural evidence to routinely-generated data such as the manuscript census. This work dovetailed with more general work on the household structure of Victorian Americans. During the late 1970s and 1980s, life course historians such as Lutz Berkner, Glen Elder, Howard Chudacoff and Tamara Hareven qualified Peter Laslett's emphasis on the predominance of nuclear family household structure in Western societies by pointing out that many individuals experienced a variety of living arrangements throughout their life course. In doing so, these historians pointed out that many aged women and men had experienced far fewer discontinuities in work and family life than was originally supposed. For example, Howard Chudacoff and Tamara Haraven asserted that the elderly actually rarely experienced an empty nest syndrome. These historians also began to highlight differences among the elderly. In "Life Course, Norms and the Family System of Older Americans in 1900," Daniel Scott Smith used the 1900 Public Use Microdata Set to examine the impact of various factors such as sex, marital status and work status on the elderly's living arrangements. Smith asserted that the co-residence of the elderly with children was not primarily forced by economic necessity; indeed, the black elderly, as the most economically deprived group, were less likely to live with children than any other group. Some of these scholars still made assumptions about the elderly’s dependence. E.A. Hammel asks "in what kinds of groups have the elderly been nurtured" and speaks of the elderly as a "general social burden" Howard Chudacoff and Tamara Haraven as well as Dan Scott Smith convey the same view of the elderly, and portray the family as "the basic support system,", the "welfare institution," for older people before the turn of the century. However, these historians implicitly countered the status declension thesis by emphasizing the elderly's overall familial integration. Furthermore, by using their demographic data to analyze internal differences among the elderly, these historians made it impossible for old age historians to ignore the effects of gender, race, ethnicity and class.

A growing tide of believers maintain that old age did not herald inactivity and absolute dependence. Andrejs Plakans and Charles Weatherell describe aged peasants in the eastern European serf estate, Pinkenhof, and the "dozens of light but important chores farmstead life entailed." They argue that, in performing these tasks, aged farmers "maintained a link with the world of work that was rarely, if ever, severed." Research by Steven Ruggles marked another development in the historiography of old age in the United States and made further inroads in the declension thesis. He did so partly by demonstrating the irrelevance of declension in the light of the demographic and economic realities of the elderly’s living patterns, and partly by highlighting how Victorian family values increasingly esteemed multi-generational family connections. In his 1987 study, Prolonged Connections, Ruggles used microsimulation models to postulate high rates of extended family living in nineteenth-century U.S., arguing that demographic transitions at the time had increased the biological opportunities for the formation of extended families. Since he found that incidences of extended families were more frequent among the bourgeoisie than among the industrial working class, Ruggles asserted that the extended family was a luxury rather than a functional strategy. In his 1994 article, "The Transformation of American Family Structure," Ruggles turned to the fuller and longer-range data available in the Integrated Public Use Microdata Set (IPUMS). Ruggles concluded once again that scholars must account for demographic conditions and the resultant pool of available kin alongside economic and cultural influences when studying family and household composition. Ruggles contended that delayed marriages and decreasing mortality increased Victorian Americans' chances of living in three-generation households.

On the basis of their research into the Victorian elderly’s living arrangements and income, scholars such as Steven Ruggles, Brian Gratton, Carole Haber, Frances Rotondo and Cheryl Elman have concluded that the elderly during the late nineteenth century were neither burdens upon their children nor dependents of the state. Rather, they contend that the industrial revolution and the steadily increasing wages it brought actually improved the elderly’s ability to provide for their own retirement. Using data from U.S. budgetary surveys and censuses, Gratton and Rotondo argued that the late nineteenth-century elderly could avoid indigence by tapping into their children’s earnings. Cheryl Elman found that the presence of nonnuclear elders, such as grandparents, in American homes in 1910 increased the odds that teens would be in school, and concluded that these elderly individuals were not a burden to the household. Instead, she argued, "Older persons could be and no doubt were productive by helping with child care and family activities such as housework, gardening or farm chores…freeing up time of other [household] members." Carole Haber and Brian Gratton describe large numbers of elderly persons in the late nineteenth century who sold their rural farms or urban businesses and moved to villages in order to live autonomously in their old age. In more general economic histories of nineteenth-century rural America, Jeremy Atack, Peter Passell and Allan Kulikoff emphasize aged farmers’ role in ensuring their children’s future by bequesting farm property or some other kind of wealth.

The history of old age is a fairly new field in Canada. Until the 1990s, Canadian scholars writing on old age demonstrated a top-down perspective or a preoccupation with the present rather than the past. They described the elderly's fates at the hands of old age home managers or chronicled the rising proportion of the elderly in the Canadian population. By the late 1980s, however, a few Canadian historians began to take an interest in the perspectives and strategies of the elderly themselves. Most did so by studying how the elderly or their families used institutions such as poorhouses and old age homes to their own advantage. James Snell has revised the history of pension policies in twentieth-century Canada by exploring how the elderly reacted to Canada's pension apparatus, forming interest groups to petition for increases in pension payments. Beyond these publications, however, few Canadian historians have focused directly on the history of elderly people.

Instead, the Canadian elderly have received largely indirect attention in general studies of the family, household structure and family economics in nineteenth-century Canada. These works, like their recent American counterparts, have also tended to picture the Victorian elderly as empowered and important decision-makers. Canada’s social historians have often adopted a family strategies perspective which emphasizes co-operation between generations: "Par nécessité, les générations étaient solidaires les unes des autres. Les enfants avaient bésoin de l’aide de leurs parents pour leur éstablissement, et les parents de celle de leurs enfants pour assurer leurs besoins matériels dans leur vieillesse." Historians such as Gérard Bouchard, J.I. Little, David Gagan, Gordon Darroch, Lee Soltow and Bruce Elliott have documented the tremendous energy parents devoted to establishing all their children on landholdings nearby. They picture the elderly as paternal gate-keepers, responsible for the fates of their children. Historians north and south of the 49^th parallel now try to cope with the limitations of the family strategies perspective by paying greater attention to intra-familial conflict and inequity, often based on gendered roles.

Given Canadian and American historians’ shared views of nineteenth-century mature families, did Victorian Canada and the United States constitute a common arena of family behaviour? American historians seldom consider this question, preferring to highlight contrasts with England and Europe. As Béatrice Craig tartly observes, "…aux yeux des Américains, le Canada dans son ensemble pourrait tout aussi bien être situé sur une autre planète." Allan Kulikoff relates late nineteenth-century America’s changing family systems to unique American events: the Revolution and the Civil War. "We have emphasized the American Revolution, arguing that the origins of yeoman and rural capitalist classes can be traced to the egalitarian impulses that conflict incited. The history of rural America in the nineteenth century would have been far different without it. The Civil War completed the bourgeois revolution begun by the Revolution." His analysis implies the irrelevance of the Canadian experience.

Canadian scholars, on the hand, frequently include comparisons to the United States in their research, especially comparisons to the American Northeast and Old Northwest. Two scholars, David Gagan and Alan Brookes, have pictured the mid- to late-nineteenth century United States as a land of plentiful work opportunities in comparison to Ontario and Nova Scotia. Based on his findings for Peel County, Gagan suggested that young men in many rural Ontario counties experienced limited access to land by the 1870s. In contrast, stated Gagan in a brief reference to the United States, "In America, the farmers’ frontier lasted until 1890." Alan Brookes made a similar argument when examining youths’ behaviour in late nineteenth-century Canning, Nova Scotia. In this district, the limited availability of farmland coupled with general economic decline sparked an exodus to the United States. Both Gagan and Brookes maintain that restricted economic opportunities prolonged children’s state of dependence in their parents’ households, bringing about "a change in both household and family structure, and in individual experience." Though they emphasized relatively restricted Canadian opportunities during the late nineteenth century, neither Gagan nor Brookes made systematic comparisons of Victorian Canadian and American household structure. Gordon Darroch and Lee Soltow have recently disputed Gagan’s theory that a land crisis occurred in rural Ontario, asserting that Ontarians’ rapid rise in land ownership with age shows they "continued to pursue a landed family patrimony as if the option remained viable." Darroch and Soltow emphasized Canadian-American similarities by demonstrating virtually identical rates of property ownership in Ontario and the northern United States in 1870 and 1871. However, like Gagan and Brookes, Darroch and Soltow left comparisons of Canadian and American household patterns to speculation.

Such speculation has formed the basis of work by two historians of French Canada, Béatrice Craig and Gérard Bouchard. Theorizing more broadly, these historians argue that the northern United States and Canada can be seen as a regional unit, in which the predominance of family farming led to common forms of family reproduction. Bouchard contends that an "open system" of family reproduction, marked by parents’ steady efforts to establish all their sons on farms of their own in return for security in old age, characterized Canada and the northern American states alike. Craig noted that eighteenth- and nineteenth-century Canadians and Americans adopted common family strategies to cope with the similar challenges of establishing their children. She counsels social historians of Quebec, English Canada and the United States to engage in more comparative research of these regions in order to enrich their understanding of North American families’ mores, motivations and behaviour. Bouchard and Craig’s hypotheses dovetail with research questions raised by Ruggles’ work and that of English historian, Richard Wall. According to Ruggles, the great majority of American elderly men and women who could have lived with their children and grandchildren actually did. However, English historian Richard Wall maintains that the majority of the nineteenth-century English elderly lived apart from their kin. Perhaps the seemingly distinct American pattern discovered by Ruggles was actually part of a broader North American experience, as suggested by Bouchard and Craig. This is a tantalizing prospect. The comparative study of household structure in nineteenth-century Canada and the United States affords the opportunity to understand the extent of similarity or difference in Canadian and American family behaviour. Did the Victorian elderly and their families in Canada and the United States alike practice strategies which differed from those practiced in the "Old World" because they lived on a continent which offered a different set of challenges and opportunities? This dissertation responds to this challenge.

This study is founded on the belief that household structure is heavily influenced by economic opportunity, by demographic structures including mortality, fertility, and age at marriage, and by social practices such as indivisible inheritance. The interpretations given in this dissertation have also drawn upon feminist analysis. Victorian Canadians and Americans shared similar gender norms. The power dynamics forged by these gendered expectations informed intergenerational relations and life course patterns. Previous historical research has convinced me that North American families did not react blindly or passively to the challenges of new farm development or of industrialization but rather maximized family ties to cope with nineteenth-century changes. At the same time, I am sensitive to criticisms of the "family strategies" approach. These criticisms are twofold. First, feminist scholars argue that family history can assume too much uniformity of purpose among families. Past studies of family strategies have ignored internal family conflicts or the fact that the needs of individual family members such as daughters could be subordinated to the interests of other family members. Greater sensitivity to both unity and disunity of family interests has encouraged both Canadian and American historians to picture a family system model in which intergenerational relations involved a degree of co-operation and exchange alongside conflict and control. By directing my research at the individual level, employing a life course rather than family cycle perspective, I analyze the behaviour of family members separately and attempt to understand their unique needs. Steven Ruggles has also criticized the family strategies approach to understanding human behaviour for its suggestion that "human motivation can be reduced to rational calculation of maximum utility." Ruggles acknowledges that material circumstances influenced human behaviour in important ways. However, he asserts that "there is no aspect of human behavior that is in its entirety the product of rational calculation for material gain," and lists a host of irrational motives for family patterns, "Jealousy, altruism, status anxiety, love, pride and social obligations..." While broad, demographic surveys of Canadian and American family behaviour and life course patterns risk portraying these North Americans as mechanistic automatons, as explained below, the dissertation draws upon documentary evidence from diaries and letters to reflect both the measured and the more emotional motivations which informed residential choices.

Finally, at the heart of this dissertation is its comparative approach. By choosing two North American countries, I approach my topic with the "most-similar nations" comparative strategy explained by Ann Shola Orloff as one which "minimizes variation along potentially explanatory dimensions" in order to "specify the causal role of [a] particular factor or set of factors." In the late nineteenth-century, Canada and the United States featured a complex array of similarities and differences. Both countries were "New World" nations with high rates of immigration and high proportions of rural dwellers; in each nation, waged labourers concentrated their efforts in primary resources or manufacturing. However, both marked and subtle differences distinguished the two North American nations during the nineteenth century. Far more Americans than Canadians lived in urban areas, laboured as professionals and hailed from places other than the British Isles. By 1871, Canada had experienced a much smaller rate of population growth than her southern neighbour; she was, from the perspective of many of her inhabitants, a newer nation. Whereas American socio-political ideology had been shaped in the cauldron of revolution and racial conflict, Canadian ideology was variously rooted in loyalism, Catholicism, ethnic and language conflict, and the exultation of "Anglo-Saxon northerness". While both nations shared a Victorian ethos of gendered social roles and family privacy, middle-class anglophone Canadians more consciously explained their behaviour in terms of their British heritage. Neither nation had yet developed comprehensive, federal-level social welfare structures; both had largely left social welfare to local authorities, private charity and churches. The United States' higher proportion of urban centres offered a somewhat more developed private charity apparatus than that found in English-Canada. Conversely, the Catholic church in Quebec had a developed a denser array of orphanages, schools, old age homes, poorhouses and hospitals than found in any other province or state. This grouping of national similarities and differences make these countries eminently suitable for "most-similar nations" comparative analysis.

Recent intellectual studies and historiographical articles have advocated a continentalist approach to Canadian and American history. In his survey of works on contemporary Canadian literature, social values and political culture, Allan Smith found that comparing Canadian and American societies has enabled investigators "to refine [their] understanding of several important aspects of Canadian life." John Mack Faragher has pointed out the settlement of North America took place "on a vast continental stage"; to understand these processes requires "nothing less than the comparative history of North America societies." Matthew Mulcahy and Russell Menard explained how comparative approaches such as those espoused by Faragher could enable historians to synthesize new, multicultural works in American history without reverting to a narrow national history which emphasized American exceptionalism. This trans-national, demographic study will confront American—and Canadian—exceptionalism by comparing and contrasting the household patterns of aged parents and children in Victorian Canada and the United States. Crucial to this aim is the unique primary source which forms the basis of this dissertation.

The key to this research project lies in its data, data which facilitates its comparative aims. Comparing family patterns across boundaries has traditionally helped scholars to understand the relationships among economies, cultures and household strategies. In fact, comparing and contrasting work from different countries has often led to major theoretical breakthroughs in the historiography of the family. For example, essays collected in one of the earliest volumes on this topic, Household and Family in Past Time, prompted Peter Laslett to conclude that a "Western Family type" of relatively late marriages and small, nuclear families dominated preindustrial Europe, as well as North America, and contrasted with family types in southern and mediterranean Europe. More recently, in the concluding essay to Aging in the Past: Demography, Society and Old Age, David Kertzer examined research on the elderly and their families in Britain, Europe and the United States. After comparing and contrasting these studies, Kertzer theorized that both the neolocal and stem-family models of household formation were overly reductive and proposed instead a new model for mature households in the past, nuclear reincorporation. Gérard Bouchard hypothesized that common family reproduction practices characterized nineteenth-century Canada and the northern United States through a survey of other histories. Thus, comparing and contrasting secondary works on family history has led to significant theoretical advances. However, in the absence of directly comparable household measures, insights drawn on the basis of secondary source surveys are inconclusive.

Transnational demographic history is a natural next step in comparing family patterns north and south of the 49^th parallel. Although independent studies could be used systematically to explore the family landscape of North America, a more effective approach to North American demographic history is one which uses routinely-generated sources from both Canada and the United States. However, the incompatible format and coding of the Canadian and American census data sets poses many obstacles to comparative research. To facilitate the research requirements of this dissertation, I have adopted a new methodology to make truly comparative demographic history possible. Following the initiative of the Minnesota Historical Census Projects, which integrated eleven American census PUMS from 1850 to 1990, I joined the 1880 U.S. PUMS with a similar Canadian data set to create an integrated series of census data from the United States and Canada. While the U.S. data originated with the Minnesota Historical Census Projects, the Canadian census data came from a PUMS of the 1871 Canadian census created by Gordon Darroch and Michael Ornstein at York University in 1979. The 1871 Canadian PUMS includes Ontario, Quebec, New Brunswick and Nova Scotia.

For the first time, census data is available to address the possibility that nineteenth-century North Americans shared a collective family dynamic of intergenerational relations. The fully integrated nature of this data set allows me to use nationality alongside other characteristics such as marital status, gender and occupational class when analyzing the comparative effects of socio-economic and demographic characteristics on the elderly’s chances of living with their children. These data also offer information at the region, state and province level, allowing me to explore not only national, but also regional and state/provincial similarities and differences. Doing so allows me to transcend, when necessary, a view of Victorian North America as a continent with distinct nations to a view of North America as a collection of regions.

The reliability of these comparative measures are what make this study of Victorian living arrangements truly unique. Previous Canadian, American and English demographic family histories have utilized different measurement schemes, analyzing family formation sometimes by the individual and sometimes by the household. These studies have also employed different classifications of individual characteristics such as occupation. These differences have made historians' findings difficult to compare. In contrast, this findings of this comparative study are based on completely comparable data.

Though census data help us to understand family patterns and dynamics, they shed little light on the motivations and feelings which underlay such behaviour. As a result, in addition to these quantitative data, I have consulted diaries and letters written between 1850 and 1890 and held at the Ontario Provincial Archives and the Ohio State Archives. The men and women whose lives provide insights for this dissertation are described in the following "Cast of Characters". It is hoped that their stories will lend a personal dimension to the analyses discussed throughout this dissertation, contextualizing household and life course patterns, illustrating the rational and irrational dimensions of family behaviour, and illuminating the significance of observed Canadian-American differences.