Their argument for stability and persistence rests on some 16,000 life histories collected from genealogists and family historians. Never exploited before on the national scale, this evidence is in several ways superior to data derived from the census because it enables the authors to trace individual migrants, identify reasons for their moves, and examine the variations in migration across occupation, marital status, age, gender, family position, and stages of the life cycle. Chapter 2 discusses the sample and carefully assesses its biases in relation to known population characteristics from census records. Much over-represented are married men and women who lived long lives, as compared to the under-representation of single men and women and children who died young. The ability to study the mobility of those who remained single is thus impaired. On the other hand, the geographical coverage of the sample is close to being representative.

One of the shortcomings here and in other parts of the analysis is the absence of measures of spread around the means of distributions. To take the mean age of death as an example, the reporting of the means for the periods 1750-1819 (69.9 years), 1820-49 (68.6 years), 1850-89 (72.6 years), and 1890-1930 (74.4 years) reflects the bias toward long-lived adults as well as the real change toward increased longevity, but the omission of corresponding standard deviations deprives us of knowing the range and shape of the age at death distributions. The authors, it must be said, do not claim representativeness for their data and are open about its limitations. Serious students of migration will want to read this chapter with care.

Equipped with this unique if somewhat skewed collection of hard-wrought data, the authors challenge much conventional wisdom, while clarifying and revising theory and empirical research in migration history. As for conventional wisdom, the notion that British people today move more often than their forbearers is said to be mistaken. This conclusion buttresses studies in recent decades showing that pre-industrial populations of Britain and Western Europe were more mobile than previously thought [1]. In addition to the number of moves individuals made, several other aspects of migration are studied across time, including the distance of moves, the destinations and paths followed, and the characteristics of the migrants themselves. To trace persistence and change, historical periods are defined and compared, the major ones being 1750-1839, 1840-79, 1880-1919, and 1920-94.

As for the distance migrants moved, the predominant pattern from 1750 consisted of short distance movements, typically within the same region. Although they grant that the proportion of long-distance moves increased over the period and especially in the twentieth century, the increase, they argue, was a small feature in the overall stability of migration patterns. Given the revolutionary improvements in transportation during nineteenth and twentieth centuries, they continue, the degree of stability seems remarkable—a point to which we shall return. Who typically moved? In the eighteenth and nineteenth centuries, the family histories show that they were usually young adults moving in family groups, and less frequently on their own.

With these continuities in place, several theories of migration are laid to rest or revised. The hypotheses argued by Zelinsky, a major voice in the modernization school of migration theory, held among other things that modernization would go hand in hand with increased and accelerating rates of mobility and with an expanding geographical extent of migration (Zelinsky, 1971, 1979, 1983 et 1993). Pooley and Turnbull find no support at all for these ideas in their data.

More influential than Zelinsky’s theory are the so-called “laws” of migration advanced by E. G. Ravenstein in the 1880s (Ravenstein). They stand up a bit better to the Pooley and Turnbull critique. One such law, that the bulk of migration since 1800 involved short distance moves within the same region, is fully borne out. And Ravenstein’s Newtonian insight that every migratory current flowing in is paired with a counter current flowing out is upheld as well. In communities, therefore, the arrival of newcomers was regularly accompanied by other inhabitants departing—another aspect of what can be called “steady state mobility”. A third of Ravenstein’s propositions—that longer distance moves tended to be from smaller to larger communities, up the urban hierarchy so to speak—proves to hold only for London and not elsewhere in Britain. Moreover, moves from large to small settlements—down the hierarchy—were nearly as common as moves from small to large settlements, according to the family history data. In rejecting the idea of stage migration, Pooley, interestingly, sets aside the argument he made with Lawton in a previous publication (Lawton, Pooley, 1992, 127-130).

Set aside also are Ravenstein’s hypotheses about gender and rural to urban migratory flows. Contrary to these hypothesis, men did not tend to move over longer distances than women. Except in cases of emigration to foreign lands, the life histories show that women and men were more like fellow travelers than the notion of separate spheres would suggest, resembling each other in the distances they moved and in age, marital status, and family position. The conventional picture that Ravenstein helped promote of rural inhabitants moving to cities while city folk stayed put also fails to find support. Instead, a pattern of rural and urban equality proves the rule: rural inhabitants and towns people were more or less equally prone to pick up stakes and move on.

After laying out the general patterns of mobility over geographic space and historical time, Pooley and Turnbull turn to explore the reasons and causes that can account for the patterns. This lucid and painstaking effort is marked throughout by an admirable insistence on complexity over simplification. In weighing the evidence and suggesting explanations, they suggest how the interaction of multiple factors was at work. Their treatment of employment is a good example. As we might expect, the search for work was indeed the single most important reason behind the movements of family groups and individuals. But work alone, they tell us, was the sole motivation in only 1/3 of the 16,000 cases studied. Rather, in the majority of cases, the decision to move was influenced not merely by employment but also by considerations of marriage and housing, as well as one’s family position, age, occupation, and other factors. The decision to pick up stakes and move, we learn, was something of an overdetermined event.

The regional character of migration to secure jobs, to cite another instance of multiple determinants, resulted from the interplay of economic circumstances, psychological outlooks, and occupation. In prosperous times, moving beyond one’s region was limited by the high cost of long-distance resettlement. And when hard times set in, the desire to stay put and tough it out in a community where one was known intensified. The inclination to stay put also varied by occupation. Skilled workers typically moved further a field than their less skilled compatriots, while farmers and agricultural laborers were less mobile than any other group.

Not every aspect of migration was the story of persisting patterns. The effects of declining family size, shifting attitudes to familial obligation, and different stages of the life course reveal intriguing changes in mobility. After the 1880s, the decline in mean family size meant that average sized families were more likely to move over longer distances than before. With the swelling of domestic service in the Victorian period, the mobility of young women in service was increasingly determined by the mobility of their employees. Migration after marriage—the key event in the life course—changed considerably from the eighteenth to the twentieth century. In the century of Dr. Johnson and Jane Austin, only 4 of every 10 newly married couples moved into their own homes, while in 1890 to 1930 period nearly 7 out of every 10 couples did so upon marriage. At work here, the authors remind us, were numerous other changes, including the increased supply of inexpensive housing, higher real wages, and a greater voice for women in household decisions.

The propensity to relocate after retirement—or at the end of the life cycle—increased dramatically during the xx^th century. Among those born between 1750 to 1839, about 1/3 of women and men changed residence after leaving the work force, often to be near their children. In the cohort born a century later (1890-1930), nearly ¾ moved to another residence upon retirement. Pooley and Turnbull point out that the meaning of these figures is difficult to pin down. In sorting out the complexities, they note that mere survival into retirement age was not as unlikely in the eighteenth century as might be expected. In both the eighteenth and twentieth-century cohorts the mean age of death was above 69.2 years—a higher figure than that for the underlying population because their data is biased in favor of adults. But that is not all. The likelihood of relocating upon retirement rose in the twentieth century for other reasons, chief among them being the increased availability of suitable housing, greater affluence, expanded welfare benefits, smaller families, and increased longevity. In view of the biased data and so many factors at work, they offer the qualified conclusion that some part of the difference in their figures is attributable to real increase in mobility upon retirement.

Here and elsewhere, the upward trend of longevity is a complication—one that often lurks in the background as a confounding factor. It casts its shadow, for example, over the major argument that mobility—as measured by the average number of moves—remained relatively stable from 1750 on. To make their case the authors point to the likely under enumeration of short distance migration in the xviii^th and early xix^th century data, thereby closing the gap between the average number of moves for 1750-1819 cohort (3.3) and that for the 1890-1930 cohort (7.0). If in this way they “bump up” the earlier figure, they whittle down the later one by subtracting, in effect, the facilitating influence of modern transportation systems. Instead of clarifying things, this reasoning muddies the water and siphons off the meaning of the numerical results. Adjusting measures of migration to neutralize the enabling factor of modern transportation seems an odd and a historical maneuver.

Arguably, a better tack is to bring longevity out of the shadows and to the fore. Other things being equal, people who live longer are apt to move more often. To take this into account and improve comparisons of mobility over time, one could begin by standardizing the historical cohorts by age. Then one could compare, say, the number of moves per average life span. Others more statistically savvy than I can doubtless suggest better methods or refinements to disentangle longevity from the propensity to move and thus control the confounding of the two. Once disentangled, the results would likely offer better evidence for the claim of stability than the argument now advanced.

As this reservation makes clear, Pooley and Turnbull have produced a book that challenges received wisdom and stimulates its readers. A work of originality in its conception and in its use of sources unexploited before, their reconstruction of individual movements in space and time is painstaking, their identification of patterns is thorough, and the reasons behind them are laid before us in admirable breadth. After reading the book the idea that mobility has been a stable feature of British life since the mid-eighteenth century becomes clear and generally compelling. Questions remain, of course, and debates over specific claims will continue fruitfully. Many questions will concern the sample and its stated and unstated limits. All of this will take us considerably further in the effort at understanding migration and mobility in past time. In this and other ways the book is a large success.