The Fertility Transition in New England: The Case of Hampton, N.H., 1655-1840
By Lawrence J. Kilbourne
Reprinted here (with permission) from chapter eleven of the book "Generations and Change: Genealogical Perspectives in Social History", edited by Robert M. Taylor, Jr. and Ralph J. Crandall, Published by Mercer University Press, 1986. All rights reserved.
One familiar characteristic of modernization is a shift from high to relatively low levels of fertility. Throughout most of Western Europe this transition occurred in the late nineteenth century. In the United States it apparently happened much earlier. Yasukici Yasuba's pioneer work in American demography demonstrated that the refined white birthrate in older areas of settlement was already quite low by 1800, while Coale and Zelnick confirmed that a downward trend in U.S. fertility had begun well before 1850. But precisely when the decline began remains in doubt.
R. V. Wells, using techniques of family-reconstitution analysis, discovered a group of mid-Atlantic Quakers whose fertility fell sharply in the late eighteenth century and continued to fall thereafter. Although Wells makes no claims that his Quakers were typical -- rather, he suggests that they merely foreshadowed tendencies subsequently generalized throughout society -- the possibility nevertheless arises that a shift toward reduced fertility may have occurred in many late colonial communities.
Efforts to date the transition inevitably raise questions of method and motive. Most population historians credit changes in early-modern natality to variations in marriage age. Wells ascribes the diminished size of his Quaker families to intramarital fertility control. Discussions of motive usually associate reduced fertility with the emergence of an urban-industrial society. Such a society, so theory holds, lowers the economic value of children while creating consumption patterns that compete with them as sources of satisfaction. Since the consumption patterns and life-style of the urban middle class are affected first, this group is considered most likely to initiate efforts at birth control.
Despite its plausibility, the urban-industrial explanation of fertility decline has not lacked critics. Yasuba stressed the importance in America of population density and availability of farmland. More recently, Maris Vinovskis has attempted to move beyond urban-rural dichotomies, emphasizing instead the pervasive influence of an increasingly commercialized, market-oriented society. Following Vinovskis's lead, Nancy Osterhud and John Fulton see the spread of family-limitation practices at Sturbridge, Massachusetts, during the early nineteenth century as the combined result of land pressure, which provided motivation, and a "commercial revolution," which fundamentally altered familial attitudes, inducing habits of rational planning and allocation.
This essay deals with Hampton, New Hampshire, between the mid-seventeenth and mid-nineteenth centuries, and is an effort to shed further light on the origin and causes of fertility decline in early America. Although much work on Hampton's economic and social history is still undone, the rough outlines are known. Founded in 1638 on a sandy strip of the Atlantic coast, Hampton remained a predominantly agricultural community until well into the nineteenth century. The growth of Portsmouth, N.H., however, exerted a strong influence on all the towns in its hinterland, and by the mid-eighteenth century farming at Hampton had long been market-oriented. While the early industrialization of southern New Hampshire affected the town indirectly, Hampton itself never became an industrial center. Nonetheless, between 1815 and 1850 the town did undergo a modest commercial revolution. Due to its excellent beach, it became a popular summer resort -- a development reflected in the increasing number of nonresidents paying property taxes, and in the establishment of several hotels such as the Eagle House, founded in 1830, and the Boar's Head Hotel, first opened in 1826. Further evidence of commercial expansion during these years is the construction of two new roads (one joining town and beach) and the completion in 1840 of a railroad linking the community to Boston.
The genealogical section of Joseph Dow's History of Hampton, supplemented by local and family histories, furnished the data for this essay. Though the genealogy contained information on more than 1,200 families, only 281 completed unions that met Henri and Gautier's Type 1 criteria were selected for analysis. These families were then divided into seven marriage cohorts. Material on pre-1700 families was relatively scarce; therefore, those unions that I was able to reconstruct were combined into a single cohort covering the second half of the seventeenth century. The other cohorts range over periods of twenty to thirty years.
Since genealogical data is often regarded as suspect, a preliminary defense of my source is required. A random check of material found in Dow against vital records preserved by the town clerk at Hampton revealed an insignificant margin of error (between two and three percent). In addition, I examined surviving tax records both to determine Dow's comprehensiveness and as a test for economic bias. The percentage of households on tax lists for 1647, 1709, 1800, and 1820 omitted by Dow varied from a low of 1.4 (1647) to a high of 9.3 (1820) with an overall omission rate of 6.3 percent. Virtually all of these households belonged to the poorer half of the population. The genealogy is consequently biased somewhat in favor of more prosperous members of the community, but not sufficiently to distort seriously my statistics. In the absence of reliable colonial censuses, the proportion of indigent non taxpayers left out of the genealogy cannot be estimated accurately. I must assume that at Hampton, as elsewhere in New England, extreme poverty was rare and that those indigents who may have been omitted were not numerous enough to jeopardize my conclusions. In any case, my checks indicate that Dow's level of comprehensiveness is remarkably high, comparable to the most complete vital records.
With these methodological problems put aside, I can now turn to the population trends that emerge from the data. The most notable feature of fertility in completed families is the decrease over time in the mean number of children born. Between the first and last cohort (Table 1), family size dropped by 3.4 births. Abrupt decreases occurred in the mid-eighteenth and early nineteenth centuries, dividing Hampton's demographic history into three distinct phases. An extremely high rate of reproduction characterized the first three cohorts—roughly eight children per union. A slight downward fluctuation occurred just after 1700, but rapid recovery ensued. Not until 1760 did a permanent decline commence. While still high, natality in cohorts married in the late eighteenth century was markedly lower than during preceding periods. In the final phase, covering 1800 to 1840, this falling trend descended to five children per marriage.
The mean number of births per union, therefore, was already in decline by the mid-eighteenth century. A steady increase in female marriage offers the simplest explanation for such a trend and, as Table 2 indicates, marriage age for women did go up. Between 1655 and 1840, mean age at first union rose by slightly more than three years. But assuming an average birth interval of 24 months, an increase of 3.02 years would result in a loss of only 1.5 births, less than half of the actual decrease. The increase in marriage of 1.58 years between the mid-seventeenth and mid-eighteenth century could have reduced family size by at most only one child, rather than the actual reduction of more than two. A more sophisticated test for the effects of changes in marriage age is to multiply the difference in decimal years by the 20-24 age-specific fertility rate of the cohort in question. Doing this, it is evident that changes in female marriage age accounted for sixty-three percent of the decline in births for women married during the first half of the nineteenth century; however, only a third of the decrease in family size between 1760 and 1799 stems from this source. A full explanation must consequently take into consideration changes in intramarital fertility.
Table 3 shows age-specific fertility -- the best indicator of intramarital fertility -- for each of our seven cohorts. A careful comparison reveals that 1760-1779 marks a boundary between a period of fluctuating but high fertility, and a period of falling fertility. A clearer picture emerges by combining the data into larger groupings (Table 4). Once again, Hampton's demographic history divides into three distinct phases. Marriages begun during the town's first hundred years were extremely fecund. In contrast to this initial phase, fertility in unions begun after 1800 was quite low. The late eighteenth century formed a transition. While rates, particularly in the later age groups, were lower than in the preceding era, they remained significantly higher than they would be during the first half of the nineteenth century.
These statistics link decreases in family size firmly to changes in marital fertility. They do not establish, prior to 1800, that deliberate practices designed to limit births caused these changes. Populations in which family-limitation practices are widespread usually have a convex fertility curve. This typical shape results from an abrupt decline in age-specific rates between the wife's late twenties and early thirties as couples, having produced a desired number of children, make efforts to avoid further pregnancies. When graphed, the rates for 1655-1759 form a concave curve, showing no sign of fertility control (Figure 1). The graph for the combined post-1800 cohorts has the convexity associated with contraception. But the graph for unions begun during the latter half of the eighteenth century is neither convex nor concave;`rather, it descends smoothly from the high plateau of age groups twenty to twenty-nine.
When applied to the 1760-1799 cohorts, other standard tests to determine fertility control also prove negative or inconclusive. Fertility rates for women wedded before and after age twenty-five differed little; telltale jumps in the interval between penultimate and last births did not occur; and mothers at the birth of their last child were only one year younger at the close of the eighteenth century than they had been at the close of the seventeenth.
Of course, biological factors can also reduce a population's fertility. Unfortunately, current research renders this hypothesis an unlikely explanation for the eighteenth-century fertility decline in Hampton. Biologically induced lower fecundity results from a deterioration in nutrition, or from genetic damage caused by excessive intermarriage. Although data on early American nutrition remains rare, New England diets in the late colonial period seem to have improved. Starting around 1740, inventories left among the probate records of Middlesex County, Massachusetts, reveal a steady rise in both the quantity and quality of food. As to the possibility of genetic debility, a study of endogamy patterns at Hampton, terminating with males born before 1720, concluded that husbands, in general, picked their spouses from a fairly broad cross-section of the geographic, economic, and social spectrums.
Deliberate family limitation accounts best for the early decline in the town's fertility. The ambiguity in the 1760-1799 statistics probably reflects a lack of consensus regarding ideal family size. Table 5 gives a percentile distribution by cohort of births per union. The modal number of children born to women in the first three cohorts fell between eight and nine. After 1800 the mode pay is in the range of two to five as extremely large families became unusual. Typically, couples married in the late eighteenth century had from six to seven offspring. The next most common frequencies, however, were four to five births and eight to nine. The aggregate effect of statistical measures that superimpose such diverse fertility behavior would be to obscure traces of a significant minority practicing birth control within a community whose fertility is still largely unregulated. Highs and lows would cancel each other out, and age-specific rates for the population as a whole would, when graphed, slope smoothly rather than turning convex.
The data, therefore, allows me to draw three conclusions with reasonable certainty: (1) Like Well's Quakers, the number of births per union at Hampton started to decline in the latter half of the eighteenth century; (2) while a rising marriage age played a role, changes in marital fertility were more important; (3) reduced fertility resulted from voluntary choice rather than biological debility.
Is it possible to determine which groups initiated fertility control? The small size of the samples on which family-reconstitution studies are based hinders attempts to measure the impact of economic variables. Nevertheless, by using what information is available on the economic status of a few of the families in Dow's genealogy, some light can he shed on this question. Taking the state tax for 1800, I divided all couples in which the wife was born between 1750 and 1774 into three groups and computed age-specific fertility rates for each (Table 6). The town's richest citizens had the highest reproductive rates; the poorest had the lowest. Low fertility among Hampton's less prosperous taxpayers is unsurprising and easily explained. Due to nutritional deficiencies the poor in early-modern communities usually had the lowest fecundity. Of more interest are the age-specific rates for couples in the middle range ($1.00 to $3.00). In these unions fertility fell precipitously after the wife's twenty-fifth year, precisely the expected pattern among couples practicing birth control.
A simple analysis of the demographic mechanisms by which fertility is lowered leaves aside larger issues of social causation. Obviously hypotheses stressing urbanization or industrialization do not fit conditions at Hampton. Though one of the smallest of Rockingham County's townships, Hampton commenced a long-term decline in its fertility a full half-century before New England experienced even the first stirrings of industrialization.
As it would at Sturbridge Village a few years later, land pressure provided Hamptonians with the strongest motivation for smaller families. On the whole, Hampton's demographic history conforms to what James Henretta has called the "morphology" of a New England town: there is a growth phase during which the earliest settlers, graced with an abundance of land, multiply prodigiously, followed by a period of contraction as the favorable land-people ratio reverses. Probably Hampton first felt the pangs of population pressure quite early, perhaps just after 1700, but solved the problem through out-migration, sending its surplus inhabitants to settle the newly opened territories of inland New Hampshire. Local and family histories mention numerous offspring who in the middle decades of the eighteenth century pioneered homesteads in Maine and along the Merrimack and Connecticut valleys.
Yet unlike Sturbridge, the origins of the specific mentalité that persuaded Hampton's citizens to choose intramarital fertility control (as opposed to more traditional methods of reducing population pressure) cannot be traced to a concurrent commercial revolution. Chronic depression characterized the three decades extending from the mid-1760s to the mid-1790s. The struggle for independence and its aftermath disrupted trade, while monetary retrenchment in the 1780s created further hardship. Even wartime inflation, which stimulated so many sectors of the colonial economy, helped New Hampshire farmers very little since the cost of labor outstripped rising food prices.
At Hampton, the evolution of a rationalistic, risk-calculating mentalité, which Osterhud and Fulton see as a psychological precondition for the spread of family-limitation practices, preceded the town's demographic crisis. By the early eighteenth century, Portsmouth had already woven Hampton, along with the other towns of its Piscataqua hinterland, into an aggressive, market-oriented, entrepreneurial economy. The economic adversities of the late eighteenth century acted as a catalyst, triggering a demographic. response for which the requisite mentalité already existed.
Most likely, land scarcity also motivated the second sharp decrease in fertility that occurred after 1800. Here, however, other factors must be considered. Diversification of the local economy as Hampton began its slow transformation into a summer resort, combined with new consumption patterns and labor options created by the industrialization of nearby communities, may have lessened both the economic value and emotional satisfaction conferred by a numerous progeny. The many distinguished private schools founded in New Hampshire toward the turn of the nineteenth century — Hampton Academy among them — suggest that the "quality" of children had become important. Moderately prosperous couples, unable to afford both a higher standard of living and a large family, would be the first to feel the influence of these forces and to respond by attempting to avoid unwanted pregnancies.
Whatever the proximate causes behind the fertility transition, psychological changes are of greatest importance. In a wide-ranging theoretical article, R. V. Wells has argued that family planning rests on an assumption that life chances are subject to rational manipulation. The spread of birth control reflects a society's enhanced sense of mastery over its fate. Doubtlessly the increasingly commercialized nature of the late colonial economy tended to induce the type of mentalité that Wells describes, but political upheaval also played a role. The collapse of colonial oligarchies and the spread of egalitarian ideas during the Revolution powerfully reinforced feelings of individual identity and personal autonomy. The data from Hampton suggest that in the closing decades of the eighteenth century many Americans seized control of their reproductive as well as their political destinies.
[*]This paper was written while I was a research assistant to Professor David H. Fischer of Brandeis, whose encouragement and advice I would like to acknowledge gratefully.
Yasukici Yasuba, Birth Rates of the White Population in the United States. 1800-1860 (Baltimore: Johns Hopkins University Press, 1962) 50-72; Ansley Coale and Melvin Zelnick, New Estimates of Fertility and Population in the United States (Princeton: Princeton University Press, 1963).
R. V. Wells, "Family Size and Fertility Control in Eighteenth-Century America: A Study of Quaker Families," Population Studies 25 (1971): 73-82.
One recent study of a Massachusetts township found evidence of family limitation in the middle decades of the eighteenth century. A family-reconstitution analysis of Philadelphia's social elite, however, found no evidence of intramarital fertility control prior to 1825. See Edward Byers, "Fertility Transition in a New England Commercial Center: Nantucket, Massachusetts, 1680-1840," Journal of Interdisciplinary History 13 (Summer 1982): 17-40; Louise Kantrow, "Philadelphia Gentry: Fertility and Family Limitation among an American Aristocracy," Population Studies 34 (1980): 21-33.
Maris Vinovskis, "A Multi-Variate Regression Analysis of Fertility Differentials among Massachusetts Townships and Regions in 1860," in Historical Studies in Changing Fertility, ed. Charles Tilly (Princeton: Princeton University Press, 1978) 225-57.
Nancy Osterhud and John Fulton, "Family Limitation and Age at Marriage: Fertility Decline in Sturbridge, Massachusetts, 1730-1840," Population Studies 30 (1976): 481-94.
Our chief source for the social, economic, and political history of Hampton is Joseph Dow's History of the Town of Hampton, New Hampshire, vol. 2 (Salem MA, 1893).
Ibid.; Type 1 families are those in which both the marriage date and end-union date as well as the wife's birth date are known. Completed unions are defined as those unions that lasted into the wife's forty-fifth year. For a fuller discussion of the techniques of family reconstitution, see Louis Henri's Manuelle de Demographie Historique (Geneve: Droz, 1967), or E. A. Wrigley, An Introduction to Historical Demography (London, 1966) 91-159. Type 1 families are discussed in the opening chapter of Louis Henri and E. Gautier's model study, La Population de Crulai, Paroisse Normande (Paris: Presse Universitaire de France, 1958).
Sally McMahon, "'Provisions Laid up for the Family': Toward a History of Diet in New England, 1650-1850" (Brandeis Uni versity).
James M. Gallman, "Small Words and Large Worlds: Barriers and Bonds in Colonial New Hampshire" (Brandeis University).
James Henretta, "The Morphology of New England Society in the Colonial Period," in The Family in History, ed. Theodore Rabb and Robert I. Rotberg (New York: Harper and Row, 1971) 171-78. See also Kenneth Lockridge, "Land, Population and the Evolution of New England Society, 1630-1790," Past and Present 39 (1968): 62-80.
"For the economic plight of New Hampshire during these years, see James H. Flannagan, Trying Times: Economic Depression in New Hampshire (dissertation, Georgetown University, 1972); Richard Francis Upton, Revolutionary New Hampshire: An Account of the Social and Political Forces Underlying the Transition from Royal Province to American Commonwealth (Hanover NH: Dartmouth College Publications, 1936).
R. V. Wells, "Family History and Demographic Transition," Journal of Social History 9 (Fall 1975): 1-19.
"For an account of the evolution of politics and political ideas in late eighteenth-century New Hampshire, see Jere R. Daniel', Experiment in Republicanism: New Hampshire Politics and the American Revolution, 1742-1794 (Cambridge: Harvard University Press, 1970).