Contemporaries, historians, and social scientists have long acknowledged the central role that internal migration played in the growth and development of American society as well as in the personal lives of American individuals, families, and communities. With the first seeds planted in Jamestown and Plymouth, Americans began spreading ever outward, continuing a pattern established in England and reinforced by the transatlantic crossing (Thornthwaite, 1934; Rutman, 1965; 1975; Hansen, 1940; Demos, 1970:9-11; Morgan, 1975:423-32; Breen and Foster, 1973; Kulikoff, 1986a:162-3; Kelly, 1979; Adams and Kasakoff, 1984; Prest, 1976; Main, 1965: 164; Bielinski, 197?; Mathews, 1909; Turner, 1920; Bailyn, 1986; Craven, 1926:57; Wrightson, 1982; Clark and Souden, 1988; Horn, 1988). In thousands of antebellum newspaper editorials, books, diaries, travellers’ accounts, and personal letters, Easterners bemoaned the great loss of people and capital to the West while Westerners beckoned everyone to join in the bonanza, respectively condemning and applauding the great migratory habit of Homo Americanus. Federal census reports beginning in 1850 officially confirmed the great East‑West movement of population. Frederick Jackson Turner in his famous “frontier thesis” speech of 1893 placed internal migration (i.e., the “frontier safety valve”) at the heart American historiography. Economists have linked much of early American economic growth to the extensive migration of labor and capital to regions of higher wages and greater economic opportunity.
Nevertheless, despite the centrality of migration to American development, the historical aspects of the phenomenon have received scant attention. Beginning in the 1930s with the rapid increase in rural‑urban migration, social scientists began to intensively analyze internal migration, but only two studies, that of C. Warren Thornthwaite (1934) and Everett S. Lee (1957), extended their analysis back into the 19th century and these offered little more than estimated net state migration totals. During the 1930s and 1940s, historians like James C. Malin (1984), Frank L. Owsley (1945; 1949), and Barnes Lathrop (1949) began to tap manuscript census records for information on domestic migration but the tedious work attracted few followers. Only with the coming of social mobility studies, following Merle Curti (1959) and Stephan Thernstrom (1964), and the rise in the 1960s of the “new social history” did internal migration become a major topic of interest, but even then most studies limited analysis to calculating community persistence rates, balking at the immense difficulty of tracking individuals across both space and time.
In the 1980s, a renewed interest in internal migration began to appear on the horizon, perhaps as part of a growing frustration with so much ignorance over such a crucial element in American life. Bernard Bailyn (1986) has even called for a new American historical synthesis centered on migration to and within British North America, and two general surveys of the literature have seconded the tremendous potential for research in this area (Parkerson, 1983a; Kulikoff, 1986a). Social and economic historians have begun to trace migrant heads of households across space and time using census indexes to finally link both social and geographical mobility (Stephenson, 1974; 1980; Schaefer, 1985; 1987; 1989; Steckel, 1988a; 1989). Studies, including several unpublished dissertations, have opened new sources of information like family genealogies, colonial militia and war pension records, state censuses, and land sales data (Davenport, 1982; 1984; Parkerson, 1982; 1983b; Bosworth, 1978; 1980; Oberly, 1982; 1986; Villaflor and Sokoloff, 1982; Steckel, 1983; Adams and Kasakoff, 1980; 1984b; Crackel, 1981). With so many new avenues tapped, the 1990s would seem ripe for a synthetic effort.
However, on almost every point besides simply acknowledging the importance of geographical mobility in American history, historians and social scientists have reached little consensus about the exact nature of that mobility, apart from its general westward and cityward trend. Even on such a crucial point as identifying a link between geographical and social mobility, some stress the positive “pull” of greater economic opportunity, others the negative “push” into wandering poverty, while still others a homeostatic holding pattern. Although migration and community persistence studies have generated much empirical evidence, the lack of a solid theoretical framework has thus far prevented any synthesis.
Unfortunately, for the most part migration theory has not really advanced beyond Ravenstein’s (1885; 1889) hundred‑year‑old overly mechanical push-pull “laws,” developed from census analysis in industrializing England. Most “theory” remains simply at the level of empirical regularities observed in rural‑urban migration in developing and developed nations. But a general theory of migration has yet to appear, perhaps reflecting the very complexity of migration impacted by many psychological, cultural, social, demographic, economic, and other factors. Overall, migration studies prove most unsatisfying for their failure to place migration within the total context of available choices, the individual and household decision-making matrix.
The most sophisticated approach to the study of internal migration in agrarian regions like early America builds on the extensive literature of “peasant” economic development and the “transition to capitalism” debate.[ref]This interpretation of economic development follows a modified formalist-ecological anthropological approach in contrast to more explicit substantivist or Marxist approaches, while recongizing that these approaches share much overlap (e.g., Cook, 1973; Earle and Johnson, 1987:9-10). For an interesting attempt to set up a rigorous test of the approaches, see Orlove (1986).[ref] Social scientists have used economic development models almost exclusively in one direction: “modernization” models based on a presumed historical understanding of Western economic development (particularly Great Britain and the northern United States) applied to third world development programs. The failure of many of these programs (such as the “green revolution”) and the imperial nature of this relationship have dampened enthusiasm for the widespread applicability of such development models in more recent years. However, economic anthropologists since the 1960s have been doing much interesting work on how these third world nations actually developed (or did not develop) which has profound implications for the understanding of the origins of modern Western society. American historians have ignored much of this literature based on peasant economies, simply assuming that such models would have little applicability to the United States which was uniquely born modern, capitalist, and developed.[ref]Aversion to such peasant migration theories on the assumption that peasants do not migrate reflects a view long dismissed in the social science literature. See, for example, Jackson (1969); Sauers (1974); Kosinski and Prothero (1975); Brandes (1975); Yang (1979).[/ref] Although work has begun to identify much evidence of traditional and pre‑capitalist behavior among early Americans, historians have yet to escape the ultimate bugaboo of American uniqueness which colors all of American historiography. Historians who dabble in the peasant literature do so for insights into the unique American experience, while disavowing any explicit comparability. This study presumes that the United States experience did not fundamentally differ from the experience of developing nations today and that theoretical and comparative work provide an excellent framework to help organize research and raise questions to be tested against the historical evidence.
This paper synthesizes a behavioral model of “peasant” economic development built on a “peasant” (“subsistence economy”) ideal typology, contrasted with a “capitalist farmer” (“market/political economy”) ideal typology. The “peasant” ideal type posits that peasant farmers were “interested first in obtaining an adequate output per head and second in maximizing their leisure,” “were not much concerned with profit, that their principal interests were subsistence and the long‑term security of the farm, that they did not try to maximize production of cash crops and marketed only their surplus, that they avoided risk and were suspicious of innovations” (Grigg, 1979:69; McCusker and Menard, 1985:298). Earle and Johnson (1987:12) trace the logic of the “subsistence economy” to formal economics and optimal foraging theory in animal populations: “The goal is not to maximize production but to minimize the effort expended in meeting household needs.” In general, a specific mix of strategies that minimizes procurement costs remains stable unless upset by changes in population density, technology, or the environment. The model presented here presumes that each peasant household participates in the market and thus the subsistence economy interacts with the political economy.[ref]”The political economy involves the exchanges of goods and services in an integrated society of interconnected families” (Earle and Johnson, 1987:13). See also Wilkinson (1973); Chayanov (1986); Zipf (1949); Spooner (1972); Simon (1977:146-50); Boserup (1965). The “leisure preference” hypothesis, although a useful heuristic concept, might actually reflect an aversion to routinization, seasonal labor bottlenecks within the existing agricultural system (e.g., Carr and Menard, 1989), and the necessity or preference for “alternative” domestic (including social) maintenance activities, which neo-classical economic theorists (who ignore the role of values in the utility function) might consider inefficient, but all of which characterize modern Western society close enough to dismiss perpetuation of the “myth of the lazy peasant” (Hymer and Resnick, 1969).[/ref]
In the “market/political economy,” “capitalist farmers” “were latent entrepeneurs, willing to take risks and accept innovation, who found their drive for profits frustrated by high factor prices, primitive technologies, poor transportation networks, and weak markets” (McCusker and Menard, 1985:298). Whereas peasant behavior centers on minimizing labor necessary to maintain a “competency” without undue risk, the profit-oriented capitalist farmer, although limited like the “peasant” by a “bounded rationality,” more willingly risks capital and works harder with no a priori limits, constantly calculating and acting on differential economic opportunities in an attempt to maximize profits.[ref]On “bounded rationality,” see Simon (1957).[/ref] Although both “peasants” and capitalist farmers” use household labor, the key for the peasant household lies in household survival; this contrasts with an individualistic capitalist farm household which operates far closer to an n-person cooperative game than an organic unit and more likely uses non-household labor (Becker, 1981; Sundstrom and David, 1988).[ref]Problematically, slave labor can serve, at least in the antebellum Southern context, as capital and labor for the capitalist farmer (Fogel and Engerman, 1974), as an extension of the domestic patriarchal household (Fox-Genovese, 1983), or as something in between (Wright, 1978).[/ref]
McCusker and Menard (1985:304-5) identify the different implications of the two models for the process of internal migration: “In the subsistence model, ‘push’ factors dominated migration decisions: migrants moved away from overcrowded settlements with their poor prospects for economic independence rather than toward better market possibilities; and movement was financed out of family resources. In the market model, the ‘pull’ of better prospects predominated: migrants moved toward chances for commercial agriculture rather than away from depressed conditions; and it was merchants and land developers who provided much of the needed capital.”
The formalist-ecological “peasant” model of economic development employed here falls under the rubric of “population pressure” theory. Although traceable directly to the work of Thomas Malthus and his “positive” and “preventative” checks, most of this body of theory actually follows the anti‑Malthusian argument, beginning in 1818 with George Ensor, which stresses that most societies avoided a Malthusian catastrophe through implementing many different demographic and economic strategies thus enabling agricultural output to rise as rapidly as population (Ensor, 1818:205-12,371; Grigg, 1980a:61-4; Simon, 1977:3).[ref]Malthus himself noted “that if it were not for population increase, no motive…would be sufficiently strong, to overcome the acknowledged indolence of man, and make him proceed in the cultivation of the soil” (Simon, 1977:149, citing Malthus, 1803:491).[/ref] Ensor refuted Malthus’s dismissal of the potential of emigration advancing “a mobility and support‑capacity thesis which has been essentially corroborated,” theorizing “that people would migrate from settled to unoccupied areas until support capacities were reached at approximately equal levels of living; the choicest unused land would be sought by the nearest peoples who lived at the lowest standards. Ensor’s ‘theory’ was narrowly‑based on the principle of searching for food, but it did bear the elements of a minimax principle‑‑viz., least effort and most need” (Adams, 1969:36,435). In more recent years, two key lines of anti-Malthusian thought have developed:
(1) “demographic response” following Kingsley Davis’s “multiphasic response” with a primary focus on fertility control in modern industrializing nations (Davis, 1963)
(2) “economic response” following lead of Ester Boserup with an emphasis on agricultural intensification in response to population pressure on the land in primitive agrarian nations (Boserup, 1965)
While often recognizing the intricate relationship between economic and demographic behavior, most studies focus on one or the other to reduce the inherent complexities in “total” analysis. However, a few studies have taken a more complex Ensorian approach (e.g., Nicholls, 1969; Simon, 1977:183-203; Murray, 1977:22-6; Friedlander, 1983:250-1). Following Grigg (1980a; 1980b), responses to population pressure fall into six general categories:
(1) “agricultural extensification” to increase agricultural output by putting additional land into production within present agricultural system
(2) “agricultural intensification” to increase agricultural production through frequency of cropping, reduced fallow, adoption of new technology, improved methods, switch to higher yielding crops, and regional specialization (Boserup, 1965; Hirschman, 1959; Dumond, 1965; Geertz, 1963:36; Von Thunen, 1966; Slicher van Bath, 1963; Gourou, 1966; Colin Clark, 1967, 1969; Clark and Haswell, 1974; Wilkinson, 1973; Harris, 1977; Simon, 1977:158-82; Earle and Johnson, 1987; Turner, Hanham, and Portararo, 1977).[ref]Grigg (1979) provides the best critique of this school of thought. In a less secure land tenure and labor system, agricultural reform can possibly generate population pressure, displacing farm laborers and tenants from the land (e.g., Friedlander,1983:266‑7). Simon (1977:158-82) contrasts the predominant Malthusian “invention‑pull” (although not explicitly part of Malthus) with the anti‑Malthusian “population push” hypotheses; both accept an exogenous role for inventions, but Boserup’s “population push” model posits that the adoption of “new” knowledge and agricultural reform depends on prior population pressure on existing resources, viewing population pressure as a necessary but not sufficient cause for economic growth. Simon actually finds both these hypotheses complementary by restricting the “invention‑pull” hypothesis to labor‑saving inventions and the “population push” hypothesis to inventions which produce more output with absolutely more labor (although, contrary to Boserup, with less labor per unit of output).[/ref]
(3) “non‑agricultural diversification” to increase non-farm income through domestic industry, day laboring, and temporary and seasonal migration as surplus labor makes rural industry profitable (Thirsk, 1961; Jones, 1968; Mendels, 1972)
(4) “fertility control” to limit the number of births whether through regulation of births within marriage (e.g., contraception, abortion, sterilization, abstinence, infanticide, etc.) or “nuptuality control” (e.g., higher age of marriage and increased celibacy). Studies emphasizing this response note like Ensor that most preindustrial societies maintained a variety of methods for limiting population growth within a “homeostatic demographic regime,” whereas Malthus admitted only the “moral” solution of delayed marriage and abstinence (Malthus, 1798; Davis, 1963; Hajnal, 1965; Wrigley, 1966, 1969:108-43; Friedlander, 1969; 1983; Wilkinson, 1973:37‑9; D.S. Smith, 1977; Harris and Ross, 1987). Concomitant with this hypothesis, populations tend to grow most rapidly in situation with abundant resources, such as early America (Malthus, 1798; Birdsell, 1957; Dumond, 1965:303).
(5) “migration” to remove from the present place of residence on a permanent or semi-permanent basis, shares many similarities with “agricultural extensification” (rural-rural or frontier migration) and “non-agricultural diversification” (rural-urban migration), with migration simply implying a relatively greater break with the place of origin. Studies stressing this approach, often draw upon the concept of a population “safety valve,” both in a third-world peasant context (Friedlander, 1969; 1983; Brandes, 1975:14) and an early American frontier context (Turner, 1920; Wilkinson, 1973; Rutman, 1975, 1986; D.S. Smith, 1980).
(6) “sociopolitical adaptation” (or “agricultural involution”) to change institutions of access, especially “pressure reducing innovative arrangements in the manner of allocating cropping ground within the society” in response to population pressure, leading frequently to lower standards of living (Murray, 1977:32; Geertz, 1963).[ref]Murray (1977) interestingly describes a response to population pressure in Haiti very similar to the 17th century English norm of sending children to work on neighboring farms because of insufficiency of land for inheritance, providing perhaps a possible way to explain that enigmatic English practice.[/ref]
Of these different responses to population pressure, the exact solution set remains temporally, spatially, and culturally specific.
Empirical evidence reveals, ceteris paribus, a ubiquitous preference for simple extensive development, reflected in perpetuation of and regression to slash-and-burn agricultural systems under conditions of relative land abundance (Boserup, 1970:110; Davis, 1963:353; Turner, 1920:21-22; Dumond, 1965:315-8; Adams and Kasakoff, 1984a:215). Similarly, many studies treat out-migration as a “safety valve” at the heart of any adaptation strategy, with alternative responses inversely correlated with migration opportunities (Davis, 1963:355; Friedlander, 1969:359-60; 1983:265,268; Moore, 1945:119; Hawley, 1950:Chap. 9; Adams and Kasakoff, 1984a:218-9; Zelinsky, 1971).
Following Malthus, social scientists often equate “population pressure” with a true subsistence crisis, but many of the recent works disavow such a simplistic notion, preferring a culturally-defined balance between “levels of living and aspirations.” This definition emphasizes not absolute poverty, hunger, or some collective standard of living but more a sense of deprivation, declining relative to other families in the same reference group especially in conditions of rising prosperity (Davis, 1963:352,355-6,362; Glass, 1965; Friedlander, 1969; 1983; Dumond, 1965:317-8; Simon, 1977:165).[ref]Recent work on fertility changes in “modern” societies projects a similar argument, stressing adjustment of nuptuality and fertility behavior to balance earnings potential with a desired standard of living (Banks, 1954; Easterlin, 1973; 1978:126-7).[/ref]
Although some historians and social scientists might question the applicability of peasant developmental models to America, there are really no a priori reasons for assuming so.[ref]I am at present working on another paper which will address this question directly.[/ref] The question is not whether American farmers were peasants (whether of the substantivist, Chayanovian, formalist, or Marxist variety), capitalist farmers, or something in between. Such definitions often rest denotatively on unmeasurable variables and connotatively on the “you know one when you see one” syndrome, which obviates definitions that could possibly include those individuals, like American farmers, whom one “knows” were not peasants. Rather, the value of any typology lies in its ability to form testable hypotheses (i.e., avoiding post hoc circular explanations) conforming to rules of simplicity and generality (i.e., Occam’s Razor) which successfully predict or retrodict observable regular, repeatable (i.e., non-idiosyncratic) individual and aggregate human behavior (which in an historical framework proves more difficult since the historian can only infer behavior from material and literary remains). The power of any typology rests in its ability to explain both stasis and change, with process models stressing temporal continuity preferable to transformation models under the assumption of a basic continuity in human behavior over time as well as commonality across space, allowing for variations in evolutionary historical experience.[ref]Wyatt-Brown (1975) provides a precedent for explicit use of an ideal typology applied to the antebellum South. Hahn (1983) superficially introduces a “peasant” and “capitalist farmer” ideal typology but quickly abandons an etic behavioral for an eclectic emic/etic mental approach, and frequently confuses the ideal with real types. The approach presented here combines Marvin Harris’s (1968;1979) “cultural materialism” paradigm for the “science of history” with Robert Berkhofer’s (1969) “behavioral approach to historical analysis.”[/ref]
Indeed, the peasant framework, although rarely explicitly described as such in American historiography, informs much of the work on the nature of the family farm and agricultural practices in early America. The “population pressure” concept of “standard of living and aspirations” and the “peasant” emphasis on meeting household needs compares very favorably with the traditional notion of “competency” which several scholars have found to drive early American behavior (e.g., Turner, 1920; Christopher Clark, 1979; Bushman, 1981:240; Hahn, 1982:42-3; Vickers,1990). Eclectic similarities to the peasant literature reflected in the work of “substantivist-Marxist” historians like Henretta (1978), Christopher Clark (1979; 1990), Merrill (1977), Kulikoff (1989), and Hahn (1982; 1983); “formalist-ecological” scholarship like Wright (1978), Otto (1983; 1985; 1989a; 1989b), and Wilkinson (1973); the “formalist-Marxist” work of Weiman (1983; 1985; 1987; 1989); the “Celtic schoo”l of McDonald and McWhiney (1980); the more balanced historical approach of Allman (1979), Rutman (1985; 1986), Schlotterbeck (1980; 1982), Bushman (1981), and Vickers (1990); and the seminal (but rather neglected) work of cultural geographer Milton B. Newton (1967; 1974) demand a closer examination of the benefits that a more rigorous anthropological peasant framework can bring to an understanding of early American development. Due to space limitations the rest of this paper will limit development of this framework to the “demographic response” to population pressure within early America, saving for a second paper an analysis of the “economic response” and changes wrought by commercialization, industrialization, and urbanization.
Although not fully synthesized, “population pressure” models have a long history in the United States, beginning with the work of Malthus himself who pointed to America as proof that only a rich abundance of land could stave off the disaster of Malthusian population growth (Malthus, 1798; Potter, 1965:662-3; Wilkinson, 1973: 148). George Tucker (1843:101‑7), using available aggregate census data, perhaps first recognized the negative impact of “increasing density of numbers” on fertility in a society with a relatively high standard of living far removed from a true subsistence crisis. Tucker identified a highly uniform inverse relationship between population density and natural increase across space and time, East and West, North and South, the same empirical pattern that social scientists are still trying to explain.
Most recent “anti-Malthusian” interpretations of the historical American experience highlight “fertility control” and “migration” as the key responses to population pressure, with the most sophisticated work following the lead of Yasuba (1962) who identifies population pressure on limited land resources and delayed marriage as the key causal factor in the tremendous fertility decline in the 19th century United States.[ref]Other major studies noting the link between land abundance and fertility include Thompson and Whelpton (1933) and Taeuber and Taeuber (1958).[/ref] These works, including Forster and Tucker (1972); Easterlin (1976a; 1976b; 1978); Leet (1975; 1976; 1977; 1978); McInnis (1977), Lindert (1978), Temkin-Greener and Swedlund (1978:40), Laidig, Schutjer, and Stokes (1981), Schapiro (1982; 1986), Newell (1986), and Atack and Bateman (1989), for the most part follow Easterlin’s “bequest model” emphasizing the causal interrelationships between farm acreage values, returns on capital, inheritance practices, and the desire of a father to pass on land to his children, which lead to efforts to limit the number of children or move west as land prices rise under more dense settlement. Combining the leads of Frederic LePlay on inheritance patterns and Kingsley Davis’s “multiphasic response,” Easterlin extends the “competency” model of early American behavior to not just the immediate household but future branch households since the “concern about loss of status for his children most motivates the farmer who goes to great efforts to avoid risking a short fall” (Easterlin, 1976b:65). “The main pre‑conditions for our model are chiefly an agricultural organization based on family farm or peasant units, the institution of multi‑geniture or its equivalent, and mortality conditions sufficiently favorable to assure a reasonably good prospect of the survival of infants to adulthood” (Easterlin, 1976b:73, emphasis added).
However, the “bequest model” and other “population pressure” variants, despite achieving some sort of consensus among historians, remain largely untested and continue to rest on highly questionable assumptions. Much effort has gone into attempts to refine and isolate the “population pressure” variable (Potter, 1965:677-8; Forster and Tucker, 1972; Easterlin, 1976b; Leet, 1978; Schapiro, 1986; Connell et al., 1976:7-9; Vinovskis, 1978b; D.S. Smith, 1987:77), but still the concept remains quite nebulous. Steckel (1980:332; 1985:132-3) notes quite correctly that aggregate correlations of population density factors with fertility might mask more significant causal relationships (suggested in the steady decline of frontier fertility over time), while tests of the model at the individual household level have proven far less successful than at the aggregate county or state level (McInnis, 1977; Easterlin et al., 1978). Both of these points indicate the need to look more closely at the cultural and community context in which households make demographic decisions and the relative nature of the standard of living and aspirations as the key to population pressure rather than absolute land availability, well illustrated in the finding that New Englanders and Southerners retained to a high degree their deviant fertility patterns when they settled in Ohio under conditions of similar land availability (Leet, 1978).
The present understanding of agrarian inheritance patterns remains highly undeveloped (Bogue, 1976). Some studies note a shift from primogeniture to more egalitarian multigeniture with respect to land inheritance under increasing population pressure in the 19th and 20th centuries (Newell, 1986; Otto, 1989a), while Rutman and Rutman (1984b) and Greven (1970) note the opposite in the 17th and 18th centuries, and Easterlin (1976b:64) and Schapiro (1986) presume multigeniture fully in place by the 17th century with little change thereafter. Davis (1963:353) entirely downplays the role of the “inheritance system” as more symptomatic than causal simply reflecting the prevailing demographic solution, a conclusion supported by Gallman’s (1982) comparison of age of marriage patterns in the colonial Chesapeake and New England. Also, although Easterlin (1976b:72-3) finds that such a model explains “closely linked patterns of economic and demographic change that have reoccurred in state after state,” combining North and South in a single explanation as Tucker (1843), Yasuba (1961), and Forster and Tucker (1972) had done before him, demographic historians have all but ignored the applicability of the model to the early South.[ref]For one case study supporting the “land availability” hypothesis in the antebellum South, see Schlotterbeck (1980:79-120).[/ref]
The major problem with the “bequest model” for the study of early America lies in questionably presuming people delay marriage because of the number of children they want to have rather than real cultural constraints linking marriage to the establishment of an independent household. Easterlin (1976b:63) draws inspiration from Greven’s (1970) work on colonial Andover, and thus a long tradition linking age of marriage to inheritance patterns and land availability in pre-industrial societies (e.g., Malthus, 1798; Davis and Blake, 1956:215-8; Ohlin, 1961; Hajnal, 1965:133-4; Wrigley, 1978; D.S. Smith, 1973; Gross, 1976:78), while many studies have identified adjustments in nuptuality (both the rise in the age of marriage and celibacy) as the key to the decline in the child-woman ratio in the American colonial and antebellum era, especially in the South where marital fertility control proved negligible (Yasuba, 1962:135; Steckel, 1980:350; Sanderson, 1979; David and Sanderson, 1986:309-13; Haines, 1986:380; Newell, 1986; Wahl, 1986:420-1; Sundstrom and David, 1988:191).[ref]On the other hand, in analyzing the fertility ratios of New York counties in 1845, Vinovskis (1976a:74,82) finds evidence to support the impact of changing marriage patterns on fertility but, based on his study of Massachusetts, he rejects the notion that marriage patterns could have accounted for such a change, favoring instead attention to birth control methods like coitus interruptus. The national studies examined here raise grave doubts about Vinovskis’s attempt to define American demographic history as Massachusetts writ large.[/ref] But the “bequest model” negates the primacy of nuptuality by shifting from the readily understandable context of a couple attempting to set up a household in order to marry to the dubious context of a couple deciding whether to marry based on twenty-year futuristic projections of the standard of living of their potential offspring. Indeed, Dan Scott Smith (1987:76) in a recent overview of the literature concludes: “Declining fertility as a consequence of later and less universal marriage requires no special theory of the uniqueness of the American experience; America was simply becoming Europeanized in its fundamental economic environment.” For this reason Yasuba’s more general “population pressure” model emphasizing nuptuality adjustment, land availability, and independent household establishment continues to prove a more useful starting point than Easterlin’s “bequest model.”[ref]D.S. Smith (1987:77), summarizing recent research on American demographic history, concludes that Yasuba’s “assessment o f the main patterns still stands.” For similar criticisms of the “bequest model,” see Bogue (1976). Interestingly, Easterlin’s (1978:126-7) more general model of fertility abandons inheritance and the “bequest model” for a “modernization” approach to fertility. However, the argument presented here should not imply dismissal of the role of long-term projections on demographic decision making, only that Easterlin and his followers have not clearly delineated the entire decision-making matrix or explicitly defined the relationship between nuptuality and fertility control. However, evidence suggests that long-term income projections may provide the key to migration decision-making and, since “marital migration” makes up a key element of migration, clearly “futuristic projections” may play a critical role in nuptuality patterns. On migration and income, see Kearl, Pope, and Wimmer (1980).[/ref]
More general critiques of the “population pressure” model for fertility generally fall into two camps: neo-classical economics and “modernization” theory. Sundstrom and David (1988), reflecting the neo-classical economic “capitalist farmer” approach, emphasize an “old‑age security motive” for raising children with demands for family labor regulated by intrafamilial bargaining in an n-person cooperative game. They trace the 19th century fertility decline to the commercial and industrial revolution which generated increasing labor market opportunities outside the agricultural sector, thus giving children a distinct bargaining advantage and reducing the adult demand for children. Similarly to the criticisms of the “bequest model,” however, the Sundstrom-David model patently confuses nuptuality and fertility responses, questionably presuming that people delay marriage based on futuristic projections of the number of children they want to have take care of them in old age. Other economic historians, reflecting unease with equating fertility behavior and the availability of land in an era of general land abundance, reverse the “population pressure” causality hypothesizing that bigger families tended to migrate to the frontier (Wright, 1974:307; Steckel, 1980a:344,348-9).[ref]Steckel (1980a; 1980b), in the most thorough analysis of antebellum Southern white fertility to date, does not test the “land availability” hypothesis, preferring total wealth to land availability as a superior predictor of the ability of a household to support a family. Steckel, however, acknowledges the ambivalence of the relationship between total wealth and fertility due to the negative impact of increasingly higher costs of children (indeed, the basis on which many modern demographers explain lower fertility among the wealthy); his conclusions reflect this ambivalence revealing no systematic impact of wealth, apart from a substantial lower fertility among households with zero personal and real wealth (which he finds highly correlated with urban residence).[/ref] Several early American studies cite evidence supporting such an hypothesis (A.G. Smith, 1958:34; Taylor, 1979:32; Oakes, 1982:81; Hammel, Johansson, and Ginsberg, 1983; Otto, 1985:196-7); however, Steckel (1989:211-17), in the most systematic test of this hypothesis, finds no systematic link between number (or sex ratio) of children and migration.
Like Davis and Friedlander, “modernization” theorists stress the role of the “standard of living and aspirations” to demographic and economic behavior; but, like Sundstrom and David, they link the causal role to commercialization, urbanization, and industrialization, highlighting the accompanying “isms” and subsequent “broad changes in American society” (Potter, 1965:677-8; Vinovskis, 1976a:76-9; Modell, 1971; Wells, 1975). Interestingly, Vinovskis, the preeminent critic of the “population pressure” school, seems most troubled over the failure of the “population pressure” model to explain the parallel decline in both urban and rural fertility, yet Davis (1963:355-6) finds that the “population pressure” model does just that as both rural and urban individuals respond to the changing circumstances fostered by population growth and redistribution. More descriptive than explanatory, modernization theory, drawn like most Western economic development models from a limited understanding of the demographic and economic experience of northwest Europe and the northern United States, offers little help to the study of rural peoples. Most importantly, the theory provides no inherent direction (let alone causation) for changes in demographic behavior, except from presentist assumptions that smaller families are more desirable and modern people more migratory. Education and literacy only make information available and do not dictate utilization (cf. Newell, 1986:300).[ref]However, Von Thunen and central place theories, to be discussed in a second paper, could provide a way to merge the “population pressure” and “modernization” approaches, a potential reflected possibly in Modell’s (1971) study which identifies a transformation to smaller families in antebellum Indiana when the local rural population exceeds a critical level of nonagricultural employment and D.S. Smith’s (1987:80) suggestion that land availability might proxy distance to a significant urban center, hypotheses which have not been rigorously tested.[/ref]
Davis (1963:354-5) likewise rejects explanations of change or persistence based on value systems divorced from economic and demographic forces; such explanations dichotomize modern “declines in marital fertility (a neo‑Malthusian transition)” and traditional “adjustments in nuptuality (a Malthusian transition),” “biological and social control” and “independent choice,” “social sanctions” and “family sanctions” (Haines, 1986:380; Wrigley, 1969:192; 1978: 148-52; Easterlin, 1978:120-3). Peasant “failure to feature contraception and abortion was not due to ‘traditional attitudes’…but to the availability of an alternative which fitted the interests and structure of peasant families in the evolving economy” (Davis, 1963:355). The exceedingly poor state of nuptuality theory reflects an obsessive presentist concern with modern marital fertility control (shared equally by the “bequest model,” “modernization” theory, and the Sundstrom-David hypothesis) which leads demographers to ignore the basic “homogeneity” between traditional nuptuality and modern fertility control as well as the many alternative fertility adjustments made within marriage in most “pre-modern” societies (Wrigley, 1985; Coale and Watkins, 1986; Sundstrom and David, 1988:191; Smith, 1987:74).[ref]Davis (1963:356‑7) notes that among Polish peasants the poor both married later and stopped their reproduction earlier (perhaps by abstinence and abortion), while births per year (from marriage to birth of last child) remained constant across various landholding groups. In a highly suggestive article, Osterud and Fulton (1976) forward a rationalization hypothesis in which traditional nuptuality control (due to population pressure on land) gives way to modern marital fertility control (due to market expansion increasingly putting a price tag on family labor, thus demanding conscious planning of both the size and use of family labor); however, they acknowledge that in 19th century New England, nuptuality and marital fertility control occurred simultaneously, indicating no sharp break between traditional and modern worlds.[/ref] Furthermore, this presentism divorces fertility responses from alternative economic, demographic, and social responses to population pressure and exaggerates social discontinuity, dismissing the high level of individual household decision making in “traditional” societies and social constraints in “modern” society, the very problems which the framework presented here seeks to correct.
In contrast to the work on fertility, social scientists have devoted little attention to testing the implications of the “population pressure” model for the study of early American migration. Yet the link of population pressure to internal migration in the American context, like that for fertility decline, also has long roots and many branches. Most of these studies emphasize the ability of frontier land abundance to act as a “safety valve” or “homeostatic governor,” thus avoiding a true subsistence crisis or decline in the standard of living (Ravenstein, 1889:278-94; Turner, 1920; Wilkinson, 1973:164-5; Rutman, 1975; Kelly, 1975:95-6; Bielinski, 197?:18; D.S. Smith, 1980:19; 1982).[ref]Thornthwaite (1934:1) considers migration the American norm with resistance due to physical isolation, inertia, prejudice, ignorance, and other physical and cultural factors.[/ref] Historians of the South, following Craven (1926), traditionally turn to a similar but more negative theme linking migration directly to soil exhaustion and concomitant land engrossment by large planters, forcing small farmers to search for new lands farther West, although the most thorough analyses move closer to a “population pressure” interpretation stressing “relative” rather than “absolute” exhaustion and scarcity since much undeveloped land remained in older areas (Gray, 1933, 1:196-8,2:641; A.G. Smith, 1958:32-3; Walsh, 1977:402-5; Baird, 1990:61-8; Wooten, 1953:62; Schlotterbeck, 1980; Crawford, 1988a:216,232).
Many of these studies hypothesize an “optimum density” model of infilling (net in-migration) until population reaches an “optimum density,” followed by overflowing (net out-migration) to maintain the proper density, a process suggested for the historical American experience as early as 1934 (Thornthwaite, 1934:1-2; Gallaway and Vedder, 1971: 628-32; Rutman, 1975; 1986:173-4; Easterlin, 1976b: 72; Schlotterbeck, 1980:96,110; Jones, 1981:24,70-1; McCusker and Menard, 1985:254; Schapiro, 1986:114-5n).[ref]Inspired by ecological distribution theory, Hudson (1969:371) posits a very similar location model for rural agricultural settlement with the “density‑dependent conditions” centering, like Papenfuse (1972) and Rutman (1975), on the “lower limit on the size of farm that can be operated economically.” Like “population pressure” models described by Davis and Friedlander, Rutman’s (1975) “homeostatic governor” links migration as the key response to the level of economic opportunity, impacted by socio-cultural attitudes toward migration, yet reflects more Simon’s “invention-pull” than any “population-push” model since Rutman treats the level of economic opportunity, including the shift to non-agricultural opportunities, as an exogenous variable independent of population density.[/ref] Others criticize this infilling-overflowing model for distorting the wave-like (rather than cell-like) frontier migration pattern, ignoring the relative cultural and spatiotemporal nature of “population pressure,” and minimizing the impact of tremendous gross migration in an overly exclusive emphasis on net migration (Baird, 1990:78-92). Adams and Kasakoff (1984a:213), seconded by D.S. Smith (1982b:453-4), argue for a “constant process of moving in and out of towns [driven by differential land prices] rather than a ‘golden age’ of filling up and a subsequent period of ‘spilling over’ once towns were filled.”[ref]Historians simply follow the lead of economists and anthropologists who have applied “optimum density” models derived from neo-classical microeconomics and optimal foraging theory to explain many of the patterns of human behavior found around the world, while other scholars criticize this heuristic device for often masking complicating cultural factors which differentiate humans from animals (some discussed below in describing the “stickiness” of the migration response). Birdsell (1957:54) finds that, for prehistoric man, “emigration appears to begin to occur well before a population reaches its presumed maximum under the existing economic order‑‑when, in fact, the population has reached little more than half its presumed comfortable maximum. If this is constant, it would suggest a lengthy anticipation of the possibilities of population pressure and a tendency to deal with the problem before it becomes acute” (cited in Dumond, 1965:310). Parkerson (1982) presents data, however, suggesting that much of the high turnover found in American frontier communities may actually reflect high record-linkage failures rather than out-migration.[/ref]
Recently demographic historians have begun moving toward a synthesis along the lines of Zelinsky’s (1971) “mobility transition” hypothesis to link early American fertility and migration behavior in a “population pressure” framework, although the diverse American experience has made periodization rather difficult, suggesting perhaps a superior strategy to focus on Easterlin’s wave-like process (Easterlin, 1976b; Nugent, 1981; Jones, 1981; Schapiro, 1986; Swierenga, 1983). Dan Scott Smith (1980; 1982b:453-4; 1987) presents most of the elements for such a synthesis but has thus far left the connections only implicit. These works identify a similar process, North and South, East and West, with transition from traditional demographic regime regulated by land population pressure, rural-rural and frontier out-migration, and nuptuality control to a “modern” regime regulated by non-agricultural economic opportunity, rural-urban and urban-urban out-migration, and marital fertility control.
These attempts to combine both fertility and migration behavior in one model have highlighted, contrary to Dumond (1965) and Friedlander (1969), that out-migration, whether rural-rural or rural-urban, did not relieve the pressure for fertility regulation (Easterlin, 1976b:46).[ref]Dumond (1965:315‑8), following Davis and Blake (1956), prioritizes extensification over intensification over fertility limitation as preferred responses to population pressure, with the fertility regulation acting only as a last resort to avoid Malthusian positive checks when subsistence can not be improved. In a later work, Friedlander (1983:265,268) plays down the exclusivity of migration responses noting that nuptuality, marital fertility, and migration act as substitute responses in agricultural districts and that nuptuality adjustments have less impact in regions where higher marriage ages and alternative nonagricultural employment opportunities exist. Jones (1981) presents evidence from 18th century Massachusetts which supports Friedlander, finding that marriage ages after peaking under conditions of high population pressure began dropping with the increase in non-agricultural opportunities.[/ref] This conclusion, however, far from disproving Friedlander or championing American uniqueness, highlights the importance of understanding the entire response matrix of peoples in historical and cultural context, as well as differentiating between aggregate and individual responses. While some individuals choose to migrate, some always choose to stay; and under conditions of rapid population growth, those that choose to stay face greater pressures to adapt in other ways, and thus migration does not negate alternative responses to population pressure as Friedlander seems to suggest.[ref]Barlett’s (1980) outline of a total decision-making framework incorporating ecological, cognitive, and aggregate statistical approaches provides a good starting for resolving the inherent differences between aggregate level and the much more complex individual household level analysis of economic and demographic behavior.[/ref] Migration may indeed be the key to the total response strategy, but even “free” migration is a “sticky” process with “imperfections” “of many types: family attachments, ties to a locality, imperfect information, and, particularly, financial barriers to the acquisition of land and equipment and the cost of migration to the western lands” (Wright, 1979:665; D.S. Smith, 1980:19-20; Morrison, 1977:63-4; Easterlin, 1976b:67; Walsh, 1987).[ref]Easterlin (1976b:69-70), however, seeks to argue a fertility control preference, reasoning that “A farmer felt obligated to give a child a proper start in life, whether he was staying at home or making his fortune elsewhere. Thus, the fact that one’s children might eventually migrate to urban areas or the frontier in no way relieved one of the costs of providing for them. In consequence, fertility declined despite ample opportunities for out‑migration.” Schapiro (1986:42n), however, in the most sophisticated application of the “bequest model” to both fertility and migration, falls much closer to a “sticky” migration interpretation, linking the reduced demand for children in older areas despite availability of land in West, to the considerable costs of migration and settlement and ties to locality. Kearl, Pope, and Wimmer (1980) show 19th century rural migrants faced considerably difficult decisions balancing the “pull” of higher incomes against the significant loss in wealth accompanying the move due to lost earnings, transportation costs, etc.; this suggests migrants based their decisions more on long-term projections of competency rather than short-term profits, a point seconded by Danhof (1969:101-29).[/ref]
Indeed, this very “stickiness” in the process accounts for differing interpretations of “Malthusian crises” and “homeostatic governors” in New England towns (Grant, 1961; Lockridge, 1971; Greven, 1970; Cook, 1970; Gross, 1976; Rutman, 1975; D.S. Smith, 1980:20) and the colonial Chesapeake (Earle, 1975; Kulikoff, 1986b; Papenfuse, 1972; Marks, 1979b; Rutman, 1986; Baird, 1990). Many of the “crisis” studies conclude that eventually a migration response kicks in (e.g., Lockridge, 1971:481-2; Greven, 1970), while Rutman (1986:173) concludes “The truth encompasses a bit of both of these views.” Yet the dichotomization continues to reign. A shift from explaining local and regional variations to the framework presented in this paper, with special attention to fleshing out the “stickiness” of the migration response and the alternative responses of non-migrants, would help resolve these dichotomies while pushing the debate forward to examine more rigorously the nature of early American society.[ref]Under criticism from economists for equating “overcrowding” with a “crisis of subsistence affecting a majority of the society,” Lockridge (1971:488) reinterprets his overcrowding “in terms of ‘population pressure’ which in lieu of sufficient viable alternatives was reducing the standard of living” of many individuals, a definition quite in conformity with the framework presented here. Yet lacking a theoretical direction, Lockridge never really considers any alternatives, let alone tries to analyze why individual households did or did not respond to those alternatives. Indeed, the entire debate over Malthusian crisis versus homeostatic governor suffers severely from an overreliance on an inductive, empirical approach which, as Lockridge himself recognizes, allows the data to dictate the model. Lockridge’s (1971:491) call “to go back to the evidence and build a structure of hard data and rigorous analysis” will prove totally self‑defeating without a solid social science framework to organize that data and analysis.[/ref]
This paper has outlined the “demographic response” to population pressure within the agrarian context of early America (excluding rural‑urban migration for later anaylsis along with the “economic” shift to non‑agricultural occupations). A second paper will analyze the “economic response” including the initial regression from the English alternate husbandry to a slash‑and‑burn agricultural system, the later impact of agricultural reform movements, and the abandonment of the land for mills, small towns, and larger cities. This second paper will also more explicitly compare and contrast the response differences between the “peasant” and “capitalist farmer” ideal types to economic and demographic changes, developing hypotheses which should allow testing of the two typololgies against the historical evidence. Furthermore, the analysis will identify the critical interdependent link between population pressure, market forces, and economic development through incorporation of Von Thunen and central place theories. Finally, this next paper will conclude with a summary of the role of internal migration within the total demographic‑economic regime for both the “subsistence economy” and “market/political economy” models, setting up a research agenda for analyzing migration at both the individual household, community, and aggregate regional levels in the antebellum South in order to test the two models.
[cite]