NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Research Council (US) Committee on Population; Casterline JB, editor. Diffusion Processes and Fertility Transition: Selected Perspectives. Washington (DC): National Academies Press (US); 2001.
Diffusion Processes and Fertility Transition: Selected Perspectives.
Show detailsJOHN B.CASTERLINE
By the early decades of the twentieth century, it was apparent that a historic change had occurred in childbearing patterns in Western societies. Whereas in the past women had borne an average of five or more children by the end of their reproductive years, under the newly emerging fertility regime, the norm was roughly two children per woman (Livi-Bacci, 1999). The decline had occurred in the space of a generation or two, and was in no sense a silent revolution. Instead, this was a development noted by social scientists and laymen alike, and widely regarded as a fundamental departure from childbearing patterns in the past (Szreter, 1993). Among social scientists, the decline in fertility was generally viewed as but one element in a larger transformation in the family and its functions, with far-reaching implications for Western societies (Davis, 1945). The fertility decline in the West, largely completed prior to World War II, has been followed in the second half of the twentieth century by comparable fertility declines in Asia, Latin America, and, most recently, Africa (United Nations, 2000). By any criteria, the fertility decline of the late nineteenth and twentieth centuries must be ranked as among the more profound social changes of this era. Accordingly, it has begged for explanation, and social scientists have applied themselves to this task from the 1930s to the present.
Not surprisingly, the earliest efforts to explain the decline in fertility linked it with the other social and economic changes that were themselves a major preoccupation of social scientists, particularly the massive economic transformation labeled “industrialization,” and the concomitant shift in settlement patterns toward greater concentration of populations in towns and large urban centers. Industrialization and urbanization, it was argued, resulted in a substantial increase in the costs of rearing children and a decrease in the benefits children conferred on older generations (Thompson, 1929; Davis, 1945; Notestein, 1945, 1953). By mid-century, arguments along these lines held sway as the dominant explanation for fertility decline in the West. Some scholars also emphasized the importance of mortality decline as a precondition (Davis, 1963), and in the emerging field of social history there was an interest in the causal contribution of changing notions about family life (Aries, 1962, 1980; Caldwell, 1982) and declining adherence to long-dominant religious and ethical systems (akin to the secularization argument rearticulated by Lesthaeghe and collaborators in the 1980s [Lesthaeghe, 1983; Lesthaeghe and Surkyn, 1988]).
By the time rigorous quantitative research was initiated on the European fertility decline and the emerging declines in Asia and Latin America, the presumption was that variables such as modes of production, urbanization, and levels of schooling, themselves indicators of basic economic and social structural changes that had taken place in these societies, would largely account for the decline in fertility. Hence it came as some surprise when researchers associated with the European Fertility Project at Princeton discovered that the empirical associations between the standard battery of economic and social indicators and fertility decline in fact were rather modest in strength (van de Walle and Knodel, 1967; Knodel and van de Walle, 1979; Watkins, 1986). In hindsight, this may have been a mistaken conclusion, drawn from aggregate-level studies that were incapable of detecting the many linkages at the household level between social and economic change and demographic change (see, e.g., Kertzer and Hogan, 1989). In any case, confronted with these findings from the Princeton project, scholars turned to other explanations to augment, or even to supplant, the dominant theoretical framework in which the primary causal forces underlying fertility decline were mortality decline and the paradigmatic economic and social changes that occurred in Europe in the nineteenth and early twentieth centuries.
One set of alternative explanations that came to the fore has usually been collected under the label “diffusion.” As we shall note below, the arguments classified as “diffusion theories” vary somewhat in their emphasis, and particularly in what they regard as the unique causal contribution of diffusion theory. What unites them is an overarching model of social change in which attitudes and behaviors become more prevalent in a population through their spread from some individuals to others, through informal face-to-face social interaction or at a distance through the mass media (Rogers, 1962; Brown, 1981; Valente and Rogers, 1995). Diffusion theory usually stresses the innovative nature of the attitudes or behaviors that spread—the common phrase is “innovation diffusion”— but for most scholars this is not an essential feature of this theoretical perspective. For example, Retherford and Palmore (1983), in one of the more extended reviews of the contribution of diffusion theory to research on fertility, distinguish “discontinuous” from “continuous” innovations, the former being entirely novel introductions into a population and the latter involving modification of something already present in the population (and thus not an innovation in the strictest sense of the term). Instead, what sets diffusion explanations apart from the mainstream theories that were formulated in the early and middle decades of the twentieth century is the assertion that fertility decline is not simply an adaptive response to changes in demographic, economic, and social structures; rather, it reflects the spread of certain key attitudes (e.g., about the costs and benefits of children) and behaviors (e.g., birth control technologies). For most scholars who have argued the case for diffusion theory, the crucial point is that the spread of attitudes and behaviors is not bound tightly to societal structural changes, rather, it has an independent dynamic of its own, and hence can account for a unique portion of the variation in the timing and pace of change (Bongaarts and Watkins, 1996). Some scholars have gone so far as to propose that diffusion theory can substitute for theories that feature economic and social structural changes (Cleland and Wilson, 1987; Watkins, 1991). The more common stance is that the two sets of explanations are complementary, not competing, with diffusion theory adding further independent factors to an enlarged theory of fertility decline (Retherford, 1985; Montgomery and Caster line, 1996). These are Cleland's (this volume) “pure” and “blended” diffusion models, respectively.
Early efforts to apply diffusion theory to fertility change were not submitted as challenges to the dominant social scientific theories of demographic transition; rather, they were directed to the more practical and programmatic goal of accelerating the adoption of contraception (Bogue, 1967; Palmore, 1967; Rogers, 1973). Although not recognized at the time, in hindsight the first articulation of a diffusionist argument that ran counter to, or added a significantly new element to, the demographic transition theory developed by Davis, Notestein, and others, was Coale's address to the International Union for the Scientific Study of Population (IUSSP) General Conference in Liège (Coale, 1973). In this influential paper, Coale proposes that sustained marital fertility decline has three preconditions, one of which can be viewed as primarily a function of societal structural changes (birth control must be perceived as advantageous) and the other two of which can be viewed as somewhat independent of changes in social and economic structures (birth control must be “within the calculus of conscious choice” and couples must have at their disposal the means of avoiding births). In the decade following this paper, scholars increasingly came to perceive that the latter two preconditions could change through diffusion processes, that is, the spread through a population of attitudes and behaviors (Knodel and van de Walle, 1979; Retherford and Palmore, 1983). As Lesthaeghe (this volume) reminds us, a great virtue of Coale's framework is that it acknowledges the causal contribution of both the diffusion of innovative attitudes and behaviors and changes in economic and social structures (the latter affecting parental calculations about the costs and benefits of children). In Coale's framework neither set of explanatory factors is neglected, in contrast to much of the literature of the past two decades that has touted the contribution of one set at the expense of the other. Lesthaeghe cogently argues that much needless dispute would disappear if the field returned to Coale's full framework of three preconditions—each one necessary—for fertility decline.
While scholars struggled with the intellectual challenge of isolating the causes of the fertility declines that occurred in Europe in the past and that were under way in developing countries in the present, public policy concerns about rapid population growth in developing countries provided an entirely separate motivation for examining the explanatory power of diffusion theories. If the main obstacle to fertility decline in developing countries was not that couples did not perceive birth control to be in their interests (the argument of Davis's [1967]) famous dismissal of the potential returns from investments in family planning programs), but rather that birth control was regarded as either unacceptable or infeasible, then programs that attempted to inform couples about birth control, legitimize its practice, and make services and supplies more conveniently available and affordable would have the potential to accelerate the decline in fertility. Beginning in the 1960s, there was an enormous financial investment in “purposive diffusion” of birth control (Retherford and Palmore, 1983) through government and private family planning programs. Whether these programs were having a net impact on the timing and pace of fertility decline, and how substantial that impact was in relation to the financial cost of these programs, were research questions of utmost importance beginning in the 1970s and continuing into the 1990s (Bongaarts et al., 1990). Because these questions have been addressed in numerous other articles and volumes, they are not singled out for separate attention in this volume. However, it is worth noting that one very important issue, not considered at any length in this volume, is the extent to which social diffusion might condition the impact of formal programs, either amplifying or dampening their impact (Montgomery and Casterline, 1998).
Once the concept of innovation diffusion gained general currency in the field around 1980 (through, for example, widely read articles by Knodel [1977] and Knodel and van de Walle [1979]), the term “diffusion” appeared with increasing frequency in the literature on demographic change. Initially this reflected the frustration described above with the apparently poor performance of explanatory theories that ignore diffusion dynamics (Cleland and Wilson, 1987; Watkins, 1990; Bongaarts and Watkins, 1996; Kirk, 1996). Increasingly in the 1990s, the concept has been employed in fresh empirical studies that have yielded direct evidence supportive of one or more variants of diffusion theory (Montgomery and Casterline, 1993; Rosero-Bixby and Casterline, 1993; Entwisle et al., 1996; Kohler, 1997; Rutenberg and Watkins, 1997; Munshi and Myaux, 1998; Kohler, 2000). Diffusion explanations have been applied not only to fertility decline but also to mortality change (Montgomery, 2000), and to the experiences of both historical and contemporary populations.
The main objective of the January 1998 workshop organized by the National Research Council (NRC) Committee on Population, from which the papers contained in this volume are drawn, was to assess the potential contribution to our understanding of fertility decline of explanations that invoke the concept of diffusion. The twofold question motivating this workshop was, “How might diffusion dynamics affect the timing and pace of fertility change, and how might the magnitude of those effects be ascertained through rigorous empirical research?” This question reflected the committee's unease with the existing research literature. The committee noted several tendencies in the literature that undermined the contribution of innovation diffusion theory. The first was the tendency to construct arguments that relied heavily, or exclusively, on innovation diffusion theory or, worse, only portions of innovation diffusion theory (as described below). The committee felt the need for a more balanced treatment of diffusion in fertility theory and research, a treatment that allowed full play to the potentially powerful implications of the concept of innovation diffusion while not losing sight of the other determinants that together make up a comprehensive theory of fertility change. A second weakness in the literature was the common failure to employ the concept of diffusion in a manner that would permit rigorous assessment of causal contribution. That is, although it may be descriptively accurate to say that fertility decline “diffuses,” in the sense of reaching societies and subgroups at different times, the same descriptive generalization applies to virtually all social change. Surely if diffusion theory is to make a significant contribution to demography, it must be informative about the determination of the timing and pace of demographic change. It was with these concerns in mind and with the aim of contributing to a more balanced and insightful treatment of diffusion effects on reproductive change that the committee organized the Workshop on Social Processes Underlying Fertility Change in Developing Countries.
This introductory chapter begins with a synopsis of the treatment of diffusion in the fertility literature, highlighting the inadequacies in this literature alluded to above. From this emerges a theoretical stance, labeled here “the social effects model,” that combines the causal effects of innovation diffusion and other determinants (such as demographic and socioeconomic factors). Variants of this model are the primary focus of the papers in this volume, although other diffusionist perspectives are also represented here. Key features of the social effects model are reviewed, with reference to the papers in this volume. The Introduction concludes with a discussion of theoretical developments and empirical investigations that promise to advance our understanding of the contribution of innovation diffusion to reproductive change.
KEY ELEMENTS IN DIFFUSION THEORY
The influential pieces in the literature on fertility decline that invoke the term “diffusion” are, with few exceptions, subscribing to the same theoretical propositions. The crux of innovation diffusion theory is an argument that has two closely linked, yet distinguishable, key elements that correspond to the two terms in the phrase “innovation diffusion”: fertility decline is the consequence of the increased prevalence of attitudes and behaviors that were previously very rare or absent in the population (i.e., they are innovative), and their increased prevalence is the consequence of the spread of these attitudes and behaviors from some segments in the population to others (i.e., a diffusion process). Although it is natural to join these two elements together—and, indeed, absent the other, the explanatory contribution of each element is severely weakened—a frustrating aspect of much of the pertinent research literature is the extent to which one element has been stressed to the neglect of the other. The consequence has been a diversity in theoretical emphasis that, because it has often gone unnoticed, has led to some confusion. We identify three distinct emphases in the research literature on fertility change that have been heavily influenced by innovation diffusion theory and concepts.
In some portions of the literature, the emphasis is on the innovation component of “innovation diffusion.” This has yielded two distinct bodies of work:
- One body of work stresses the innovativeness of the exercise of deliberate fertility control. Modern fertility control as the diffusion of a behavioral innovation (use of certain types of birth control techniques and technology, heretofore unknown or rare) is the main theme of this literature.
- A second body of work that focuses on the innovation half of the “innovation diffusion” concept stresses the spread of novel ideas. These are the so-called ideational theories of fertility transition. The basic argument is that declines in fertility occur because of the growing strength of certain knowledge, attitudes, and values.
One might characterize these two literatures as concerned with the question “What diffuses?”
Another set of papers, by focusing on the phenomenon of “diffusion,” draws somewhat different conclusions about the explanatory contribution of innovation diffusion theory:
- This third body of work focuses on the social dynamics of the spread of innovative information and behaviors, such as birth control practices. The fundamental premise motivating this literature is that changes in the attitudes and behaviors of some individuals can influence the likelihood that other individuals will change their attitudes and/or behaviors. This describes a social dynamic that in the aggregate over time results in a diffusion process, that is, the spread of attitudes and behaviors through the population. In theory, this dynamic can alter both the timing and the pace of fertility decline, and hence is properly classified as a causal factor.
One might characterize this body of work as preoccupied with the question “How does diffusion occur?”
By no means are these various literatures in contradiction with each other, nor are they even exclusive of each other. There is, in fact, considerable overlap among the three. One can view the differences as primarily, although not entirely, a matter of different weightings of the two key elements of innovation diffusion theory. However, this tendency to slight one or the other key element has seriously limited the contributions of innovation diffusion theory to the explanation of fertility change. This is apparent if each literature is subjected to a critical review.
Behavioral Innovation
This body of work stresses the innovativeness of the exercise of deliberate fertility control. Fertility decline is the consequence of the spread of innovative birth control techniques and technologies. The spread can be spontaneous or directed.
This is the issue that Carlsson (1966) isolates in his seminal piece on fertility decline in nineteenth-century Sweden. At the time of Carlsson's article, it should be noted, there was already a relatively well-developed literature on the diffusion of new technologies, particularly agricultural technologies (Ryan and Gross, 1943; Beal and Bohlen, 1955; Griliches, 1957). Carlsson sets two arguments in opposition: fertility decline as an adjustment to changed social and economic circumstances that employed already known and accepted behaviors, against fertility decline as the consequence of the widespread adoption of birth control techniques that were for the most part unknown or unacceptable to previous generations. Carlsson concludes that the Swedish evidence on balance refutes the second argument.
The most articulate and sustained effort to defend the behavioral innovation argument came from scholars involved in the Princeton European Fertility Project. A cardinal principle in this project was that in pretransition populations, family limitation was an alien concept, and effective techniques to avoid pregnancy were largely unknown. This is implicit in Coale's (1973) three preconditions for marital fertility decline, one of which is that means of birth control must be known and available. This is the “able” condition in Lesthaeghe's contribution to this volume. That birth control was innovative behavior in historical European (and selected non-European) populations has been argued persuasively by Knodel and van de Walle, both separately (Knodel, 1977; van de Walle, 1992) and jointly (Knodel and van de Walle, 1979). Watkins (1986) has adhered to this interpretation of the historical European evidence. However, most of the evidence that birth control was a behavioral innovation is indirect.
During the past decade, further efforts have been made to uncover more direct evidence. Limited empirical materials from the European past can be brought to bear on this question; however, the gradual accumulation of evidence from literary and other sources raises doubts about the validity of the Princeton position in its simplest and purest versions (Szreter, 1993; Santow, 1995). More direct evidence can be gathered for contemporary non-European populations. In these populations, some researchers see clear evidence in pretransition societies of widespread awareness and acceptance of fertility regulation techniques, if not for limiting family size, then certainly for the purpose of spacing births (Bledsoe et al., 1994; Mason, 1997). By no means is there consensus on this issue, however. Cleland (this volume), updating an earlier review (Cleland and Wilson, 1987), concisely summarizes evidence for maintaining his position that pretransition populations are not familiar with birth control techniques, especially for the purpose of limiting the number of births, and therefore their widespread adoption during fertility transition should be regarded as truly innovative behavior.
Much dispute remains, therefore, on the relatively straightforward questions of whether the increasing prevalence of contraception as fertility transition progresses represents the spread through the population of innovative techniques and technologies and whether this spread is the critical catalyst to transition. For our purposes, an equally important question is: What might the answer to this question contribute to our understanding of the underlying determinants of fertility transition? If the answer is affirmative—that practice of birth control is the adoption of innovative behavior— how much has our understanding of the causes of fertility decline been advanced? The gain is far less than the intensity of the debate would imply, in our view, because this literature tends to neglect the related question of how the innovative birth control techniques and technologies reach individuals. The latter is a question about the diffusion process, and must be addressed if the argument is to provide any explanatory leverage. That certain behaviors were once alien and then become commonplace is itself a limited causal insight, more a description of social change than a theory for why and how the change came about.
The arguments of most of the scholars named here are not as crude as the previous paragraphs might suggest. The proposition that birth control is innovative behavior typically is coupled with a recognition that childbearing motivations change over the course of transition, and that the reasons why these motivations change is itself a question that must be addressed. Cleland (this volume) proposes that mortality decline is the key factor motivating fertility decline in contemporary developing countries; and Lesthaeghe (this volume) emphasizes that for fertility to decline, all three of Coale's preconditions must be present, including the perception that restricting fertility is advantageous (the “ready” condition).
In short, whether or not the increased prevalence of birth control during the course of fertility transition constitutes behavioral innovation is undeniably a significant question. Were it to be answered in the negative, this would be a serious blow to innovation diffusion theory's contribution to causal models of fertility transition. Yet by itself this question is insufficient, as becomes clear when we ask what an affirmative answer would contribute to theories of reproductive change. An emphasis on behavioral innovation must be complemented by theory and research on diffusion processes, that is, those processes through which behavioral innovations spread through a population. This is the emphasis of the second body of work reviewed below.
Ideational Change
A second body of work is closely related to the first because in both literatures the heart of the argument is the nature of the innovation rather than how the diffusion conies about. While the first body of work emphasizes behavioral innovation, the second emphasizes the spread of innovative ideas. Since Cleland and Wilson's (1987) influential piece, it has been common to speak of ideational theories of fertility transition. The basic argument is that fertility declines because of the presence of certain knowledge, attitudes, and values that either were not present previously or that grow significantly in strength. Although the innovative character of the ideas figures into the argument—and ultimately makes this argument difficult to distinguish from the behavioral innovation argument just discussed—this literature sets itself apart by its determination to contrast ideas from material conditions as possible causes of fertility change. Note that material conditions can include technological innovations, such as methods of contraception. Among the ideas identified in this literature are the notion of family limitation, knowledge/attitudes/values about modern contraception, and ideas about family behavior (the roles of women and children).
Falling under this label are major contributions to the literature on fertility transition that differ significantly among themselves in the substance of their arguments. Caldwell's theoretical pieces on fertility transition written in the late 1970s and collected in his 1982 book (Caldwell, 1982) attribute considerable causal power to the spread of Western ideas about family life, through schools and through the mass media. But the foundation of his theory is an argument that changes in modes of production—material conditions—alter the costs and benefits of children, and in this respect his theory is much larger than the ideational change argument (and, indeed, is on the whole compatible with the conventional demographic transition theory formulated several decades earlier). Similarly, Freedman (1979) suggests that new consumer aspirations, diffused through international networks of communication and transportation, constitute one of the powerful motivations to reduce family size. In doing so he adds content and specificity to a theory that at its core emphasizes the determining power of changes in social and economic conditions. Perhaps the most sophisticated contribution that might be classified in this body of work is Lesthaeghe's research on fertility change in Europe (Lesthaeghe, 1983; Lesthaeghe and Surkyn, 1988). Lesthaeghe argues that secularization and the emergence of an emphasis on individual autonomy and self-actualization together explain important features of fertility trends and differentials in Europe in the twentieth century. But like Caldwell, Lesthaeghe also attributes considerable causal power to changing economic structures.
In contrast, Cleland and Wilson (1987) explicitly reject the notion that fertility decline can be explained by changes in structural conditions. Instead, they argue, the relatively autonomous spread of information and values about fertility regulation has been the primary stimulant of contemporary fertility transitions. Cleland repeats, and updates, this crucial point in his contribution to this volume. After reviewing the accumulated empirical evidence up to the present, he maintains tine position articulated in his widely cited piece with Wilson, namely that “the case that the idea of marital fertility regulation was a true innovation both in Europe and elsewhere remains robust despite widespread skepticism” [our italics].
There are many disturbing aspects of the baldest statements of the ideational change argument. Several of these have already been identified by Mason (1992) and Burch (1996). For one thing, the favoring in ideational explanations of ideas about contraception is arbitrary and unnecessary. There is every reason to give equal or greater weight to ideas that influence the demand for children, namely, ideas about the costs and benefits of children, the roles of women and children, and so forth. The more fundamental problem with this literature, however, is its implicit behavioral model. It is common in this literature to perceive ideas and material conditions as alternative, even competing, causes of fertility change, a definition of the terms of the debate that demands that ideas can be separated from material conditions. Most social science theory does not accommodate such a relationship between ideas and material conditions: Palloni (this volume) makes this point by drawing on mainstream sociological theory, and Carter (this volume) recounts the rejection by anthropology early in the twentieth century of the notion of autonomous cultural diffusion, supplanted by functionalist theory that emphasized the capacity of societies to invent their own idiosyncratic solutions to common human problems. Most social scientists recognize that ideas are grounded in social and economic institutions (see review in Hechter, 1993). This insight is valid at any level of aggregation, from the individual to the society. From this perspective, the disjuncture between material conditions and ideas found in some of the pieces in this body of work is a fiction. This key point is underscored by empirical work carried out in developing countries during the past two decades, some of it inspired by the ideational argument, that plainly reveals that the ideas that bear on fertility and family planning decisions more often than not are ideas about material conditions—changing labor and commodity markets, new economic opportunities, and so forth (Casterline, 1999b). An opposition between material conditions and ideas simply does not fit such empirical evidence, and this should come as no surprise: cultural systems cannot be detached from social and economic systems to the degree that some of this literature presumes.
A rejection of the causal contribution of changes in material conditions does not necessarily follow from an emphasis on ideational change. As already argued, innovation diffusion theory can augment, rather than displace, theories that focus exclusively on the causal effects of changing societal structures. Most social science theory accords some limited autonomy to ideas. Exercising this limited autonomy, certain innovative ideas may have led fertility transition, indeed acted as the key catalyst to fertility transition. At issue, then, is whether it is plausible that ideas led the way, and what those ideas might be. On these points, too, there are some reasons for skepticism about the validity of the ideational change argument.
Ironically, social scientists have as often been impressed with the resistance of cultural systems to change as with their capacity to stimulate change. “Cultural lag” is a venerable, if simplistic, concept in sociological research on development and social change (Ogburn, 1922). Some demographers point to the rapidity of the decline of fertility in many societies as evidence of the causal contribution of changing ideas, the assumption being that ideas (values, norms) are generally capable of more rapid change than social and economic structures (Bongaarts and Watkins, 1996). This assumption is sensible, and yet there is a substantial amount of social science research on societal change that reveals that norms and values are slow to change because they are so deeply wedded to individual and group identities (Geertz, 1973). In the case of fertility, the more common case may be that fertility reduction is an effort to maintain existing norms and values in the face of changing material conditions (Casterline, 1999a). That is, family limitation is a new strategy for achieving long-standing goals (Montgomery and Chung, 1999). Santow and Bracher (1999) beautifully describe how this generalization applies to the decline of fertility among southern Europeans in Australia. And some analyses of the East Asian fertility transitions—the most complete to date of the non-European transitions—conclude that they were motivated above all by a desire to achieve socioeconomic goals that were grounded in established familial norms and values (Greenhalgh, 1988).
In short, on its own the ideational change argument is unsatisfactory on several counts. A divorce of ideas from structural conditions is artificial; in fact, it may often be ideas about material circumstances that are most influential in reproductive decision making. By specifying an opposition between ideas and material circumstances, tine ideational change argument impedes the development of a satisfactory theory of fertility change. It follows that the ideational change argument must be embedded in a larger theory that encompasses a more complete set of causal factors. A second point is that, as with the behavioral innovation argument reviewed above, the ideational change argument should be accompanied by theory and empirical research on how ideas become more prevalent—diffuse—in a population. That certain ideas come to the fore and stimulate changes in reproductive behavior is more a description of change than an explanation. Propositions about how ideational change comes about (its timing, its pace) can be derived from the diffusion portion of “innovation diffusion” theory, as discussed below. It is not surprising that many of the prominent pieces in the ideational change literature propose that the ideas establish and strengthen themselves in a society through a diffusion process (e.g., Cleland and Wilson, 1987). What is lacking in this body of work is a formal and rigorous development of diffusion processes as a causal force. Like the literature reviewed above that emphasizes behavioral innovation, the main deficiency of the ideational change literature is that the governing theory and the empirical research are incomplete.
Social Dynamics
A third body of work that draws on innovation diffusion theory is more concerned with the diffusion process, that is, with the question “How does diffusion occur?” The focus of this literature is the social dynamics of the spread of innovative information and behaviors. By social dynamics, we mean the interdependencies among individuals in their behavioral decisions, in this case with respect to reproductive behavior. The key premise underlying this body of work can be concisely stated as follows (see Palloni, this volume): “Changes in the knowledge and behaviors of some individuals affect the likelihood that other individuals will change their knowledge and/or behaviors.” Concretely, one might posit that if one woman in a community begins using a modern contraceptive, for example, this in itself changes the likelihood that other women in the community will adopt contraception, net of other characteristics of the women and the community. To borrow language from epidemiology, a social contagion process occurs.
Like the two literatures just reviewed, the social dynamics literature tends to concern itself with information and behaviors that are innovative. The explanatory contribution of this argument is not limited, however, to ideas and actions that are novel. Social dynamics can help account for increased prevalence of already existing knowledge (e.g., the advantages of small families) or behaviors (e.g., coitus interruptus), provided that other conditions have changed. Moreover, an emphasis on knowledge and values is characteristic of this literature, but without the tendency found in the ideational change literature to set ideas and material conditions against each other. One can view the social dynamics argument, therefore, as subsuming key elements of both the behavioral innovation and ideational change arguments.
What distinguishes the social dynamics literature is its attention to the diffusion process itself. This process can be viewed as an emergent outcome of the accumulated decisions of many individuals. The argument is that the very dynamics of this process can influence individuals to make decisions that differ from the decisions they would make in isolation from this process. For example, individuals who are predisposed to use a modern contraceptive might not do so because so few of their peers use contraception or because of the struggles with side effects that they have observed among their friends. Or, in contrast, individuals who have been reluctant to use a modern contraceptive might feel secure in beginning use once they observe that many of their friends use one and seem to derive benefit from use. Elsewhere, we have proposed that two mechanisms provide the behavioral foundation for social dynamics (Montgomery and Casterline, 1996; see also Palloni, this volume): social learning, the process through which individuals gain knowledge from others (through informal or formal social interaction, and including the mass media); and social influence, the process through which some individuals exert control over others, by virtue of their power or authority. Social learning and social influence are both types of what we shall term social effects, a key concept in this essay. Social learning and social influence are perhaps the most powerful and pervasive types of social effects that bear on fertility, but other types of social effects also can be identified (see next section).
Although there are early contributions to this literature by Bogue (1967), Palmore (1967), Freedman and Takeshita (1969), Rogers (1973), Crook (1977), and Rogers and Kincaid (1981), it is during the past decade that research on social dynamics and reproductive behavior has intensified. Several teams of researchers have submitted both conceptual and empirical pieces: Casterline, Montgomery, and collaborators (Rosero-Bixby and Casterline, 1993; Montgomery and Casterline, 1993; Lee and Casterline, 1996; Montgomery and Casterline, 1996); Watkins and collaborators (Watkins, 1990; Bongaarts and Watkins, 1996; Buhler et al., 2000; Kohler et al., 2001); and Entwisle and collaborators (Entwisle et al., 1996; Godley, 2001). Burch (1996) and Kohler (1997) are other strong contributors to this literature. This work has drawn inspiration from a rapidly expanding literature in sociology (especially the social network literature: Marsden and Friedkin, 1993; Valente, 1995; Hedström and Swedberg, 1998) and in economics (especially the literature on social learning and the literature on neighborhood effects: Case and Katz, 1991; Bikhchandandi et al., 1992; Ellison and Fudenberg, 1993; Shiller, 1995; McFadden and Train, 1996; see also Arrow, 1994). These literatures describe models similar in structure that assume essentially the same underlying behavioral mechanisms. Related literatures in epidemiology, geography, and communication are reviewed in Valente (1995).
For the purposes of the current overview of innovation diffusion theory and fertility transition, the crucial point is that the social dynamics argument adds distinct, and plausible, elements to causal models of fertility change, what we have termed social effects. Social effects can accelerate or retard the process of fertility change, a point that is developed further in this introduction and in the papers collected in this volume. Explaining fertility change is a matter of accounting for the timing and pace of change, and social dynamics can affect either one. More debatable is whether the social dynamics argument has anything to say about why fertility transition occurs at all. This can be restated as a question of the contribution of social effects to the determination of “equilibrium” levels of fertility. Durlauf and Walker (this volume) use economic theory to describe several types of social effects, with an emphasis on fertility change but with some commentary on theoretically plausible effects on equilibrium levels as well. Similarly, research in social psychology indicates that social influence can modify behaviors even under conditions of relative stability in the surrounding social, economic, and cultural structures (Cialdini and Trost, 1998). Whether this occurs to any meaningful extent with respect to fertility is an issue that requires more theoretical and empirical investigation. For now, assessing the causal impact of social effects on the timing and pace of fertility change is sufficient challenge and is the primary focus of the papers in this volume.
It is natural to imagine that social effects operate through informal social interaction—that is, through social networks—and hence it is hardly surprising that much of the recent empirical work on social effects on reproductive behavior has included the collection of extensive data on social networks. In adopting social network models, fertility researchers can draw on the theory, concepts, and tools of a rich subfield of sociology (Degenne and Forse, 1999) that have been applied to the problem of innovation diffusion from the 1950s (Coleman et al., 1957; Coleman et al., 1966) to the present (Valente, 1995). However, personal social networks are but one means through which social effects might operate, and it would be a mistake to limit social effects to this channel. In particular, it is clear that the mass media are another channel through which one set of individuals can affect another. Individuals become aware of what other persons are thinking and doing by reading newspapers and magazines, listening to radio, and watching television. Although exposure through social networks and through the mass media are clearly fundamentally different modes of contact with other persons, both can be channels for social effects as defined above, and both fit within the general social effects model to be described in the next section.
An important question is what relationship exists between social effects that operate through personal networks and social effects that operate through the mass media. To begin with, it is plausible that each set of effects conditions the nature and magnitude of the other. In this vein, in their review of mass media effects on reproductive behavior, Hornik and McAnany (this volume) suggest that exposure to other persons through social networks can amplify or dampen the effects of mass media exposure, both by affecting the receptivity of individuals to mass media messages, and, once individuals are inclined to adopt an innovative idea or behavior acquired through mass media exposure, by encouraging or discouraging them from acting on their desires. Alternatively, the two channels may substitute for each other as sources of information or influence (Valente and Saba, 1998). Yet another possibility is that mass media exposure leads to a modification of the patterns and/or content of interpersonal communication (Valente et al., 1996). Because of these various possible relationships between social effects via personal networks and via the mass media, social network analysis alone cannot provide a complete assessment of the contribution of innovation diffusion theory. It is with this in mind that this collection of papers includes the Hornik and McAnany review of the research literature on the mass media and fertility.
The literature that has concerned itself with social dynamics—how diffusion occurs—yields testable hypotheses about the timing and pace of fertility change (see next section of this essay; also Palloni, this volume). Despite this considerable strength, the social dynamics argument falls short of providing anything like a sufficient foundation for a theory of fertility change, for the simple reason that it provides little guidance as to why individuals might be prepared to change their reproductive behavior and which innovations will have appeal. Hence, although the social dynamics argument is attractive on formal grounds because it lends itself naturally to the articulation of causal propositions, the argument must be said to lack essential content. As we have stressed throughout this section, satisfactory explanatory theory must join innovation and diffusion, the former providing content and the latter describing process, and it must recognize the causal contribution of societal structural changes.
From this review of the three distinct thematic emphases in the literature on innovation diffusion and fertility change—behavioral innovation, ideational change, and social dynamics—two principal conclusions emerge. First, each emphasis on its own is incomplete and, in particular, is unable to support full-fledged theory about the causes of fertility transition, that is, why onset is early or late and why pace is slow or rapid. Second, the literature that emphasizes social dynamics—that is, diffusion processes, how diffusion occurs—has been the latest to develop and would seem to offer particular advantages when it comes to the formulation of explanatory models. For this reason, the present collection of papers is weighted toward this latter emphasis—how diffusion occurs— although the question of what diffuses is discussed at some length by several contributors (Cleland, Lesthaeghe).
THE SOCIAL EFFECTS MODEL
The formulation and empirical investigation of the social effects model is guided by the premise that changes in the knowledge and behaviors of some individuals affect the likelihood that other individuals will change their knowledge and/or behaviors. This describes a social contagion process or, following Erbring and Young (1979), an “endogenous feedback” process. The usual hypothesis is that social effects operate in addition to (or “on top of”) other determinants of changes in fertility behavior (Palloni, this volume).
The premise stated above rings true; it possesses “face validity.” But we should press ourselves and ask why such social dynamics might occur. In the previous section, social learning and social influence were identified as specific mechanisms through which social effects might operate, in this instance to affect the timing and pace of fertility change. It may be helpful to make this less abstract by describing those circumstances under which it is highly plausible that these sorts of mechanisms might be in play (Montgomery and Caster line, 1993):
- (1)
When individuals are uninformed about behavioral choices they might make, for example, information about available contraceptive technologies. Those individuals who learn about, or who adopt, certain contraceptive methods can serve as sources of information for others. Or advertisements in the mass media might bring new contraceptive technologies to the attention of individuals. In these circumstances, the social effects consist of information flow.
- (2)
When individuals are uncertain about the benefits and costs of certain fertility decisions they might make. Risk aversion can be an impediment to the adoption of innovative behaviors that would appear to offer net benefit to the individual. The experiences of some individuals offer concrete demonstration to others of the possible benefits and costs of making the same reproductive choices. In these circumstances, the social effects can be termed demonstration effects.
- (3)
When social norms prohibit certain reproductive behaviors, for example the use of induced abortion to limit family size. If group norms are determined, in part, by the behavior of group members, then individual decisions to adopt innovative behavior can modify the group norms that others later confront when contemplating adoption of innovative behaviors (Bicchieri et al., 1997). (This effect is potentially very powerful if the violation of group norms by merely a minority of the group is sufficient to undermine those norms and render innovative behaviors acceptable.) In these circumstances, the social effects consist of change in normative context.
An emphasis on the circumstances in which social effects are likely to be powerful is an important feature of our argument. The expectation is that effects of substantial magnitude are only likely to operate in rather special circumstances, namely where lack of information, risk aversion, or norms are obstacles to the adoption of behavior that is otherwise desirable from the standpoint of individual (or couple) cost-benefit calculus. There is a danger of either underestimating or exaggerating the magnitude of social effects. On the one hand, conventional models that either omit or do not feature social effects may miss entirely their tremendous potential to powerfully accelerate or decelerate changes in attitudes and behaviors, as is clearly evident in formal simulations (Rosero-Bixby and Casterline, 1993; Burch, 1996; Hedström, 1998; and other literature reviewed in Durlauf and Walker, this volume). On the other hand, the temptation to exaggerate the magnitude of social effects takes two forms: first, failure to recognize that these effects in their full strength are probably limited historically and to certain social contexts; and, second, failure to recognize that social effects are but one set of factors in a larger model of the determinants of fertility.
As should be clear from the critique above of the ideational change argument, by no means are social effects restricted to knowledge about contraception. Included within demonstration effects are effects on the perceived costs and benefits of children: through interaction with others, couples may modify their perceptions of the net value of an additional child. Similarly, although it is not clear from the definition above, included within information flow is the spread of new algorithms for weighing the many separate costs and benefits of children. As Mason (1992) points out, calculation of the net value of children is a complex task, and hence couples almost certainly rely on simplified calculation procedures that are part of the cultural toolkit. Revision and reinterpretation of these procedures is an ever-present possibility, following discussions with other persons or observations of their experiences (the latter being an example of demonstration effects), or as a result of mass media exposure that alters what Hornik and McAnany (this volume) term the “frame” that guides individuals when making decisions.
It is essential to be clear about the contribution of social effects to causal models of fertility change, a concern of several papers in this volume (particularly Palloni). A simple algebraic representation will assist in structuring this discussion:
Yi,t=Xi,t β+α∑Yj,t–1Wj+εi,t (1)
where:
Y is an indicator of fertility behavior
X are sets of conventional determinants of fertility
W are weights for the salience of the jth person for the ith person
i denotes the ith person or couple
j denotes the jth person or couple
t denotes time period
and
β, α, ε are parameters
This is a model for fertility dynamics, hence the explicit subscripting by time. Imagine Yi to be innovative fertility behavior, namely termination of childbearing after two live births. This behavior is affected by the conventional determinants X and, in addition, by exposure to the fertility behavior of others Yj. A more elaborate formulation might include on the right-hand side, in addition to the fertility behavior of others (Yj), their reproductive knowledge and attitudes, and perhaps also other behaviors and attitudes that might plausibly affect fertility decision making (e.g., other persons' views about how much schooling children should obtain). Parameter α, which in the simple expression of equation (1) is a single parameter that summarizes the cumulative effects of the behaviors of all persons j (as weighted by Wj), represents an overall social effect, that is, the combined effects of the behaviors of other persons j on the fertility behavior of person i. This social effect is assumed to operate with some lag. This equation is a concise representation of more elaborate models of the same form developed in Palloni (this volume) and, among recent contributions to the literature, Marsden and Friedkin (1993), Strang and Tuma (1993), Valente (1995), Friedkin (1998), and Van den Bulte and Lilien (in press).
Nothing in the formulation of equation (1) requires that persons j from which the social effect a originates be confined to the personal social network of person i. The only requirement is that person i be aware of the behaviors of persons j and not indifferent to those behaviors (i.e., persons j are salient social actors for person i). This formulation allows for social effects operating indirectly and at a distance, including through the mass media.
A fundamental feature of equation (1) is that it encompasses both social effects and the effects of the conventional determinants X. In this specification, the two types of effects are combined additively. This is consistent with the argument presented earlier in this Introduction and by other scholars (Mason, 1992; Burch, 1996; and, in this volume, Palloni, Carter, and Lesthaeghe) that there is no theoretical basis for setting structural and innovation diffusion explanations in opposition to each other. Equation (1) can be viewed as one algebraic articulation of Cleland's (this volume) “blended model.”
Further features of equation (1) deserve some attention because they point to issues that are examined by one or more papers in this volume. A first point is that the equation does not require that social effects result in more rapid fertility change. For the social effects to accelerate fertility decline, two restrictions must be placed on α. First, it must be positive in sign (i.e., the weighted social effect across all j must be positive). Second, and a corollary of the first, those individuals who have changed their behavior (the innovators) must be more salient to individual i, that is, they must have larger Wj, than other individuals X. The first restriction rules out negative feedback effects, such as rumors about detrimental health side effects of contraception. But negative feedback effects are by no means uncommon, especially with respect to innovative behaviors and technologies about which little is known, such as modern contraceptive technologies (Lesthaeghe, this volume; Casterline and Sinding, 2000). The second restriction means simply that innovative fertility behaviors must have more appeal than customary fertility behaviors. It is plausible that this is often the case on the eve of fertility transition, if the social and economic calculus has changed in such a manner as to make additional children less valuable to parents (e.g., because of substantial mortality decline). Or the calculus may not have changed, but technologies heretofore unavailable happen to satisfy long-standing desires. Such may be the case, for example, with respect to the adoption of modern medical technologies and methods of personal hygiene that lead to improvements in child survival. Hence a more complex but realistic specification than equation (1) would make the social effect α conditional on the outcome of individual cost-benefit calculus: an individual i will be especially alert for behaviors of other individuals j that are in his or her interest. Note that this line of reasoning gives clear primacy to cultural, social, and economic explanations for fertility change. These account for the fact that individuals are prepared to adopt innovative behaviors.
Without these two restrictions, the social effects model, rather than offering an explanation for the rapidity of many fertility declines, instead provides good reason to expect fertility to be resistant to change. Indeed, perhaps it was the relative absence of retarding social effects that explains the rapid declines in countries such as Thailand and Colombia (Knodel et al., 1984, at times imply as much for the case of Thailand), and the dominance of certain social effects that explains the slow pace of decline in other settings such as Pakistan (Sathar and Casterline, 1998). In this vein, Potter (1999) argues that various sorts of social effects have contributed to the maintenance of contraceptive practices in Brazil and Mexico that have proven to be highly disadvantageous to the health of women and undesirable on other grounds as well.
A related issue is how to account for the decisions of the earliest adopters of innovative behavior. As Pollak and Watkins (1993) point out, a pure social effects model cannot account for the behavior of the vanguard group (termed “trendsetters” by Pollak and Watkins), and this is a weakness of any theory that places heavy weight on social effects. This problem is addressed in equation (1) through the Xi,t β term—the effects of conventional determinants, what Durlauf and Walker (this volume) refer to as “exogenous forcing variables,” and presumably the stimulus for the earliest adopters. This alone does not resolve the problem, however, because if this model is to enjoy any advantage over models that exclude the social effects term (α∑Yj,t–1Wj) the vanguard group must exert more influence than others in the population. Given these two problems, it is not surprising that social effects models perform much more effectively as ex post explanations than as predictive theories; ex ante it is, in practice, difficult to know which persons will assume the role of trendsetters and why these persons, and not others, will exert disproportionate influence on the behavior of others (Kreager, 1993; Pollak and Watkins, 1993). This is but one aspect of a larger theoretical problem raised by Carter (this volume) and elaborated on by Carley (this volume): individuals are not passive recipients of relatively limited amounts of discrete information about the attitudes and practices of other persons; rather, they must sift regularly among large volumes of information coming from persons both nearby and distant (e.g., through the mass media), much of it contradictory. Typically the outcome of this process will be a reliance on some pieces of information and not others (i.e., selective social learning), and, perhaps of more profound importance, a transformation of the information received, so that it fits better with past experience and/ or with existing beliefs. How to explain the relative salience of the voluminous bits of social information to which individuals are exposed is among the most challenging problems confronting researchers who wish to employ the social effects model. For this reason, the joining of cognitive psychology and social network research, as in Carley (this volume), may be critical to the formulation of successful social effects models.
This discussion draws attention to one further issue about the social effects model, namely the role of perceptions. Although the social effects term on the right-hand side of equation (1) contains the behaviors of persons j, ordinarily it is not those behaviors themselves that matter but rather person i's perceptions of those behaviors. As Durlauf and Walker (this volume) note, it is the expectation of the choices of others in the populations that bears on the decision of any particular person (see also Valente et al., 1997; Montgomery and Chung, 1999). This again makes the case for an integration of models for cognition and social interaction, for which Carley's review (this volume) provides much helpful guidance.
IMPROVING RESEARCH ON SOCIAL EFFECTS
Although several of the seminal contributions to the literature on innovation diffusion and fertility change were published a decade or more ago, systematic rigorous research on this topic—both theoretical and empirical research—is still in its infancy, with much of this work very recent. This is especially the case when it comes to research guided by the social effects model. The work to date can be faulted for its simplicity, and yet this is a common feature of a research literature in the early stage of development. In addition to reviewing existing research, the authors of the chapters in this volume identify ways in which research on social effects might be improved. The needed improvements are both conceptual and practical. We consider four ways in which research on this topic might be advanced. The first three are conceptual, with each one having direct implications for the design of empirical research: consideration of a broader set of types of social effects, better specification of the structure of social relations, and more explicit attention to the dynamics of social systems. The fourth concerns data collection requirements.
Types of Social Effects
As noted above, the mechanism for social effects that has received the most attention to date is social learning, that is, that individuals obtain information from others (about the likelihood of children dying, about the costs and benefits of children, about contraceptive technology, etc.) that informs their reproductive decisions. It is plausible that social learning can exert a powerful effect on reproductive decision making, but other types of social effects also deserve consideration. Already mentioned above was social influence, that is, that some individuals have the power to constrain the decisions of others (due to authority, deference, cumulative obligations, etc.). It is natural to group these two types of social effects together, because in structure they closely resemble each other, as reflected in expressions such as equation (1).
In contrast, the type of social effect that sociologists term social comparison (Carley, this volume; Palloni, this volume) can take an altogether different form. Theories of social comparison have a very long heritage in sociology (Festinger, 1954; Merton, 1968). The fundamental notion is that an individual assesses his or her needs and well-being through comparison of his or her circumstances to those of others. The conclusions drawn from this comparison are a function of the relative status of the individuals, and this is what gives social comparison theory its power and complexity. Individuals are assumed to respond differently to a recognition that higher status individuals have certain attitudes or behaviors than to a recognition of the same attitudes and behaviors among lower status individuals. One particular form of social comparison is social competition, and another is social emulation (Hedström, 1998; Palloni, this volume). Although social emulation generates effects that are nicely represented by equation (1) and similar expressions, social competition effects would appear to require a different specification, in which the degree of dissimilarity between the individual and his or her peers enters explicitly into the modeling. To date there has been little effort to examine how social comparison and its subtypes (including emulation and competition) might affect reproductive decisions.
Another type of social effect that can be subsumed under social influences but is distinct enough to be noted separately is social coercion (Molm, 1997). In all societies, individuals make some decisions under orders from others. The orders may be issued in personal relationships or, at the other extreme, in codified rules that are enforced through institutionalized power. This applies to reproductive decisions, if not fertility outcomes then the direct determinants of fertility: marriages can be arranged, and contraception and induced abortion can be prohibited. In northern Ghana, for example, senior men are granted decision making authority over many aspects of young women's lives, and in effect operate as gatekeepers for the diffusion of innovative reproductive behaviors (Adongo et al., 1997).
Much of the literature on social effects on reproduction presumes relatively passive social exposure. This can simplify the modeling of social effects, particularly if one is prepared to assume that social exposure is exogenous to reproductive decision making. It is clear, however, that in many instances this assumption is untenable: individuals make choices about with whom they interact and to what they are exposed (Carley, this volume). Indeed, at the extreme individuals actively seek information that might assist them in making decisions, about reproduction and other types of outcomes (Pescondido, 1992; Boulay, 2000). Whether information that has actively been sought can be assigned a causal role is a matter of dispute, and raises basic philosophical questions about the nature of causality in the social sciences (Pearl, 2000): Can factors that are deliberately employed by individuals to achieve desired ends be regarded as “causal” in any sense? What is the causal standing of mediating variables? However one answers these basic questions, the possibility (indeed, virtual certainty) of active information seeking certainly complicates the assessment of social effects on reproduction and, more concretely, demands that equation (1) be augmented with an equation for the determination of social exposure (i.e., social effects α∑Yj,t–1Wj as an outcome).
An entirely different set of social effects can be grouped under the concept of social capital. Social capital refers to the access to resources, of all kinds, provided by social relations. It can be viewed as a property of individuals and higher aggregations, such as local communities. Since James Coleman developed this concept and coined the term in the late 1980s (Coleman, 1988), social capital has been the subject of a burgeoning body of research, initially in sociology (Putnam, 1995; Portes, 1998) and more recently in economics (Knack and Keefer, 1997). The concept has become extremely influential in the development literature, with an accumulating body of empirical studies indicating that individuals and communities possessing certain types of social capital fare better in terms of standard development outcomes (Woolcock, 1998; Narayan and Pritchett, 1999). For the purposes of modeling reproductive change, the value of the concept is threefold. First, it adds to the explanatory models for various determinants of fertility, such as schooling, income, and health status. In this respect social capital is only an indirect determinant of fertility, and thus does not enter directly into models of fertility such as equation (1). Second, social capital may bear directly on the perceived costs and benefits of children. Augmenting one's social capital can be a motivation for having, or not having, children. This argument is developed in Astone et al. (1999) and tested empirically with survey data from the United States in Schoen et al. (1997). (Although use of the term social capital is recent, this particular argument has a longer history in the fertility literature. For example, it figures prominently in Caldwell, 1982.) As Palloni (this volume) points out, social structure itself is transformed by changes in fertility, and individuals may recognize this and take this into account when making decisions about reproduction. Third, and more germane to social effects as defined here, social capital as a property of individuals and communities—to whom individuals are connected, the resources they can obtain through those relationships, their trust in those relationships—can affect the scope and magnitude of social effects on fertility. In terms of equation (1) and the social effects term α∑Yj,t–1Wj, the concept of social capital encompasses both the composition of the j other persons and the structure of the Wj, that is, the salience attached to the knowledge and actions of those other persons. In this respect, although the concept of social capital would not appear to bear on the basic structure of the social effects model, it may well lead to significant improvements in the application of this model in empirical research, informing decisions about the content of data collection instruments and the specification of equations at the analysis stage (Lin, 1999).
Structure of Social Relations
There has been a tendency in the fertility literature to view social effects as the outcome of informal social interaction in local personal social networks. These networks are often treated as unstructured, homogeneous, and static. As Carter (this volume) and Carley (this volume) stress, this can simplify social relations to such an extent as to lead to serious bias in the assessment of the nature and magnitude of social effects. To rectify this shortcoming, theory and empirical research must be improved in a number of respects.
First, there must be more precise measurement of patterns of informal and formal social interaction. This can be viewed as a question of how to determine the composition of persons j and the structure of weighting matrix Wj, Heterogeneous mixing will be the norm in virtually all settings, with important implications for the expected magnitude of social effects (Akerlof, 1997) and for the design of data collection exercises. The literature on social networks—conceptualization and measurement—is now well developed and provides more than adequate instruction (Strang, 1991; Wasserman and Faust, 1994; Degenne and Forse, 1999). As Carley proposes (this volume), the modeling of social network effects needs to be more attentive to complexity in network “nodes” (types of actors) and network “ties” (types of linkages and the exchanges that occur through them). An excellent example of the lack of development of the fertility literature in this respect is its neglect of the “structural equivalence” argument of Burt (1987, 1992). According to this argument, rather than being swayed by the overall prevalence of certain attitudes and behaviors in their social network, individuals are more affected by the attitudes and behaviors of those persons j with whom they share a “structurally equivalent” network position. This argument has clear implications for the construction of the matrix Wj. In the fertility literature, apparently only Valente (1995) has tested Burt's influential hypothesis in empirical analysis.
Second, patterns of social relations are not static but rather undergo continual evolution and transformation, and this must be taken into account in any effort to assess the nature and magnitude of social effects. Theory and methods for considering social network evolution are under active development (Carley, 1999, this volume). Almost certainly this implies longitudinal observation in empirical research.
Third, research on social effects on fertility must allow for both localized and long-distance effects. The mass media (Hornik and McAnany, this volume) are the most dominant means for social effects over a distance, but indirect social relationships can take other forms as well, as described in the burgeoning literature on “globalization” (Calhoun, 1992; Kearney, 1995). A particular type of model that deserves more attention in the fertility literature is the two-step model of Katz and Lazarsfeld (1955), in which innovative ideas and behaviors are transmitted relatively long-distance to elites (perhaps via the mass media), who in turn affect other individuals in their local communities. Watkins and Hodgson (1998) describe such a process with respect to the diffusion of fertility control in Kenya: Kenyan elites were exposed to innovative reproductive behaviors in other countries—inadvertently through their schooling and professional activity and deliberately through the recruitment efforts of international agencies—and then subsequently undertook local activities that eventually modified the attitudes and behaviors of other segments of the Kenyan population. There is a risk, however, of placing too much emphasis on social effects that transcend and penetrate small-scale groups; the continuing importance of local communities should not be overlooked (Cox, 1997; Watts, 1999).
Social Systems
Social effects can be represented in simplified form in equations such as equation (1). However, because these effects operate over time and consist of interdependencies among community members, an assessment of the full impact of social effects on social change can only be obtained through the construction and estimation of social systems that contain the implied feedbacks. As yet, research on fertility change has hardly begun to entertain such system models, although a few scholars have made initial efforts in this direction (Durlauf and Walker, this volume; Kohler, 2000). (See also Gregersen and Sailer, 1993; Hallinan, 1997.)
Feedbacks are a fundamental feature of these system models. It is also likely that the models will need to allow for social effects that are nonlinear in form: thresholds, ceilings, and marked variation in the magnitude of the effects as the prevalence of attitudes or behaviors evolves (Durlauf and Walker, this volume; Hornik and McAnany, this volume; Palloni, this volume). This is a common proposition in the sociological literature; see, for example, the influential pieces by Granovetter and Soong (1983, 1986). (See also literature reviewed in Valente, 1995.) An intriguing concept is “tipping point”—that once an attitude or behavior achieves a certain prevalence in a community, adoption by others in the community becomes much easier and occurs rapidly. This notion has recently caught the imagination of a popular audience through Gladwell (2000).
The data requirements for the construction and estimation of social system models are daunting. Surely for the foreseeable future the models will be highly simplified and will obtain parameters by assumption rather than empirical measurement. Nevertheless, interest in such models is growing in demography (see, e.g., Blanchet, 1998), and it is reasonable to expect substantial progress over the next few years. Among other dividends, this effort, if accompanied by continued disciplined empirical work, should improve our capacity to assess the nature and magnitude of social effects on reproductive behavior.
The Need for Empirical Data
In light of the several-decades heritage of interest in applying innovation diffusion theory to the study of fertility transition, it is somewhat puzzling that the literature contains so few rigorous empirical studies. The most influential conceptual pieces were published more than a decade ago, and several of them more than two decades ago (Coale, 1973; Knodel and van de Walle, 1979; Watkins, 1986; Cleland and Wilson, 1987). As of the early 1990s, nearly all the empirical research that attempted explicitly to test hypotheses derived from innovation diffusion theory had been carried out either by Donald Bogue and his students or by Everett Rogers and his students (see reviews in Retherford and Palmore, 1983 and Valente, 1995). It is by any measure a rather small body of empirical work, especially in the context of an explosion of empirical research on fertility and fertility transition during the 1970s and 1980s. How can this neglect of social diffusion processes be explained?
Certainly unavailability of the necessary empirical data accounts in part for the paucity of research. The major survey programs—the World Fertility Survey and the Demographic and Health Survey—have collected very little of the information required for estimation of any variant on the social effects model. This in turn can be interpreted as an implicit rejection, or at least lack of interest, in innovation diffusion theories (Cleland and Wilson, 1987). There is surely some truth to this, but in our view the obstacles are as much practical as ideological. The data requirements are demanding and, more importantly, entail somewhat different data collection designs than have been standard in the field. The key features of data collection that would permit the estimation of the social effects model are:
- Measurement of social exposure, including some of the following: informal social interaction with kin, friends, neighbors, and workmates; formal social interaction with program agents (health and family planning workers, school teachers); and mass media exposure.
- Measurement of individuals' perceptions of the attitudes and behaviors of other persons.
- Prospective data collection, so that social exposure and perceptions at earlier times can be related to later attitudinal and behavioral transitions.
Longitudinal designs, achieved through prospective data collection, are especially critical for obtaining valid estimates of causal effects, such as those specified in equation (1) (Palloni, this volume).
In principle, social effects should be considered at all levels and via all mechanisms: personal social networks, local social organizations, influential elites, the mass media, and program personnel (health workers, school teachers, and so forth). In practice, simplification is unavoidable: no one data collection exercise can afford to obtain data that permits estimation of social effects at all these levels. In any case, if the social effects model is specified in full generality, it admits too many possibilities and lacks sufficient structure for these effects to be precisely and confidently identified (Montgomery and Casterline, 1998). Researchers have no recourse but to engage in some simplification, primarily through deletion, in their investigation of social effects. This requires an informed and in-depth understanding of the structure of social relations in the society where the investigation is occurring, as proposed by Carter (this volume). A good rule of thumb is that researchers' protocols for sampling social networks (informal, formal, and long-distance) should to the extent possible mimic the sampling habits of the actors under investigation (Palloni, this volume).
All this seems a daunting undertaking. And yet data collection carried out in the 1970s (as reviewed in Retherford and Palmore, 1983, and Valente, 1995) and during the past decade (Kincaid et al., 1996; Valente et al., 1997; Entwisle and Godley, 1998; Boulay, 2000; Casterline et al., 2000; Kohler et al., 2001) demonstrates that it is feasible to design projects that conform to the principles just enunciated. Despite the recent progress, it remains the case that the more imposing barrier to research on innovation diffusion and reproductive behavior is not the underdevelopment of theory but rather the lack of data that will support rigorous empirical testing of theory already in place. A number of the papers in this volume nicely demonstrate that a rich collection of concepts and theories are awaiting empirical investigation (Cleland, Palloni, Carter, Carley).
THE PAPERS IN THIS VOLUME
Despite an interest in innovation diffusion theory among demographers that extends back at least to the 1960s, and the publication more than a decade ago of widely read pieces that argued vigorously that research on the determinants of fertility change should give far more attention to diffusion dynamics (Knodel and van de Walle, 1979; Watkins, 1986; Cleland and Wilson, 1987), research on the contribution of diffusion remains undeveloped. One reason for this, just noted, is the scarcity of data that will support the estimation of the models implied by innovation diffusion theory, including the basic model we have termed the “social effects model.” A more fundamental explanation for this state of the field, however, is that many of the efforts to date have employed incomplete or imbalanced conceptualizations of diffusion effects.
The aim of this collection of papers is to fill in some of the existing gaps and achieve a better balance than has characterized the literature to date. A deliberate effort has been made to represent the various social science disciplines that have given systematic attention to diffusion processes (either recently or in the past)—sociology (Palloni), anthropology (Carter), social and cognitive psychology (Carley), and communication sciences (Hornik and McAnany). (Economics is the major oversight; an economic analysis was presented at the 1998 workshop [Durlauf and Walker, this volume].)
As indicated above, the existing literature on fertility transition that was influenced by innovation diffusion theory tends to focus either on innovation—What are the innovative attitudes or behaviors that diffuse?—or on diffusion—By what process do attitudes and behaviors spread through the population? The latter has been given far less attention than the former, and hence this is the emphasis of the majority of the papers in this volume. Palloni reviews the evolution of theory and models of diffusion in sociology, and then presents and critiques a more complicated version of the social effects model of this introduction. Carley provides a concise yet comprehensive overview of research findings from social and cognitive psychology that speak to the general question of how individuals learn from, or are influenced by, other persons. All models implicitly, if not explicitly, make assumptions about the nature of interpersonal learning and influence. For social effects models of fertility change to become more powerful and precise, they must be informed by the behavioral research that Carley summarizes. Hornik and McAnany tackle the important problem of social effects through the mass media (with particular reference to effects on reproductive behavior). It is clear that in the contemporary world, this is an important channel for innovation diffusion, and that to ignore this channel is to run the risk of obtaining a biased impression of the impact of diffusion dynamics on reproductive decision making.
In the three other papers in this volume, far more attention is given to the issue of the content of diffusion processes—What are the innovative attitudes and behaviors that diffuse, and to what extent does this explain fertility change? Cleland reviews the fertility literature of the past four decades, for both historical Europe and the contemporary developing countries. From this he concludes that the diffusion of knowledge, acceptance, and technologies of birth control provides a parsimonious and compelling explanation for the onset of fertility decline in historical Europe and, while not as decisive for declines outside Europe in the second half of the twentieth century, it nevertheless stands as one of the primary underlying causal forces. Lesthaeghe revisits Coale's (1973) three preconditions of sustained marital fertility decline, which as noted earlier was an early theoretical statement that can be viewed as (implicitly) arguing for a central causal role for innovation diffusion. Lesthaeghe argues, illustrating his point through analysis of recent Demographic and Health Surveys data, that in positing a causal role for innovation diffusion one need not deny the central causal contributions of changes in the demand for children, itself a response to societal structural changes (demographic, social, economic). Finally, Carter observes that anthropology early in the twentieth century accorded substantial causal power to cultural diffusion, only to conclude that this was an inadequate explanation for much of the meaningful variation among societies. Carter's chapter serves as a caution against excessive enthusiasm for innovation diffusion theory. As noted earlier, the research literature on fertility transition contains examples of such excess enthusiasm. One of the conclusions that it is hoped the reader will take away from this collection of papers is that innovation diffusion is but one component in a more elaborate causal process that also involves factors such as mortality decline and economic transformation, and that the most revealing models will take due account of all these causal forces.
REFERENCES
- Adongo, P.B., J.F.Phillips, B.Kajihara, C.Fayorsey, C.Debpuur, and F.N.Binka 1997. Cultural factors constraining the introduction of family planning among the Kassena-Nankana of Northern Ghana. Social Science and Medicine 45(12) :1789– 1804. [PubMed: 9447629]
- Akerlof, G.A. 1997. Social distance and social decisions. Econometrica 65(5):1005–1027.
- Aries, P. 1962. Centuries of Childhood: A Social History of Family Life . New York: Knopf.
- 1980. Two successive motivations for the declining birth rate in the West. Population and Development Review 6(4):645–650.
- Arrow, K. 1994. Methodological individualism and social knowledge. American Economic Review 84(2):1–9.
- Astone, N.M., C.A.Nathanson, R.Shoen, and Y.J.Kim 1999. Family demography, social theory, and investment in social capital. Population and Development Review 25(1):1–32.
- Beal, G.M., and J.M.Bohlen 1955. How Farm People Accept New Ideas . Cooperative Extension Service Report 15. Ames, IA: Cooperative Extension Service.
- Bicchieri, C., R.Jeffrey, and B.Skyrms 1997. The Dynamics of Norms . Cambridge, Eng.: Cambridge University Press.
- Bikhchandandi, S., D.Hirshleifer, and I.Welch 1992. A theory of fads, fashion, custom and cultural change as information cascades. Journal of Political Economy 100(5):992–1026.
- Blanchet, D. 1998. Nonlinear demographic models and chaotic demo-dynamics. Population: An En glish Selection (Special Issue): 139–150. [PubMed: 12157941]
- Bledsoe, C., A.G.Hill, U.D'Alessandro, and P.Langerock 1994. Constructing natural fertility: The use of Western contraceptive technologies in rural Gambia. Population and Development Review 20(1):81–114.
- Bogue, D.J., editor. , ed. 1967. Sociological Contributions to Family Planning Research . Chicago: Community and Family Study Center, University of Chicago.
- Bongaarts, J., W.P.Mauldin, and J.F.Phillips 1990. The demographic impact of family planning programs. Studies in Family Planning 21(6):299–310. [PubMed: 2075620]
- Bongaarts, J., and S.C.Watkins 1996. Social interactions and contemporary fertility transitions. Population and Develop ment Review 22(4):639–682.
- Boulay, M. 2000. The Influence of Information-Seeking Strategies on Social Network Composition and Contraceptive Adoption Among Women in Rural Nepal. Unpublished paper presented at the annual meeting of the Population Association of America, March 23–25, 2000, Los Angeles.
- Brown, L.A. 1981. Innovation Diffusion: A New Perspective . New York: Methuen.
- Buhler, C., H.-P.Kohler, and S.C.Watkins 2000. Who Influences Contraceptive Use in S.Nyanza District, Kenya?: Evidence from a Social Network Study. Unpublished paper presented at the annual meeting of the Population Association of America, March 23–25, 2000, Los Angeles.
- Burch, T.K. 1996. Icons, straw men, and precision: Reflections on demographic theories of fertility decline. Sociological Quarterly 37(1):59–81.
- Burt, R.S. 1987. Social contagion and innovation: Cohesion versus structural equivalence. Ameri can Journal of Sociology 92(6):1287–1335.
- 1992. Structural Holes: The Social Structure of Competition . Cambridge, MA: Harvard University Press.
- Caldwell, J.C. 1982. Theory of Fertility Decline . London: Academic Press.
- Calhoun, C. 1992. The infrastructure of modernity: Indirect social relationships, information technology, and social integration. Pp. 205–236 in Social Change and Modernity , H. Haferkamp, editor; and N.Smelser, editor. , eds. Berkeley: University of California Press.
- Carley, K.M. 1999. On the evolution of social and organizational networks. Research in the Sociology of Organizations 16:3–30.
- Carlsson, G. 1966. The decline of fertility: Innovation or adjustment process. Population Studies 20(2): 149–174. [PubMed: 22084908]
- Case, A., and L.Katz 1991. The Company You Keep: The Effects of Family and Neighborhood on Disadvantaged Youths . National Bureau of Economic Research Working Paper No. W3705. Cambridge: NBER.
- Casterline, J.B. 1999. a Conclusions. Pp. 357–369 in Dynamics of Values in Fertility Change , R.Leete, editor. , ed. Oxford: Oxford University Press.
- 1999. b The Onset and Pace of Fertility Transition: National Patterns in the Second Half of the Twentieth Century . Policy Research Division Working Paper No. 128. New York: Population Council.
- Casterline, J., M.Montgomery, S.Green, P.Hewett, D.Agyeman, W.Adih, and P.Aglobitse 2000. Contraceptive Use in Southern Ghana: The Role of Social Networks. Unpublished paper presented at the annual meeting of the Population Association of America, March 23–25, 2000, Los Angeles.
- Casterline, J.B., and S.W.Sinding 2000. Unmet need for family planning in developing countries and implications for population policy. Population and Development Review 26(4):691–723.
- Cialdini, R.B., and M.R.Trost 1998. Social influence: Social norms, conformity, compliance. Pp. 151–192 in The Hand book of Social Psychology, Volume II , D.T.Gilbert, editor; , S.T.Fiske, editor; , and G.Lindzey, editor. , eds. New York: McGraw-Hill.
- Cleland, J., and C.R.Wilson 1987. Demand theories of the fertility transition: An iconoclastic view. Population Stud ies 41(1):5–30.
- Coale, A.J. 1973. The demographic transition reconsidered. Pp. 53–72 in International Population Conference, Liège, 1973, Volume I . Liège, Belgium: International Union for the Scientific Study of Population.
- Coleman, J. 1988. Social capital in the creation of human capital. American Journal of Sociology 94:S95-S120.
- Coleman, J., E.Katz, and H.Menzel 1966. Medical Innovation: A Diffusion Study . New York: Bobbs Merrill.
- Coleman, J., H.Menzel, and E.Katz 1957. The diffusion of an innovation among physicians. Sociometry 20:253–270.
- Cox, K.R., editor. , ed. 1997. Spaces of Globalization: Reasserting the Power of the Local . New York: Guilford Press.
- Crook, N. 1977. On social norms and fertility decline. Journal of Development Studies 14(4):198–210.
- Davis, K. 1945. The world demographic transition. Annals of the American Academy of Political and Social Science 237:1–11.
- 1963. The theory of change and response in modern demographic history. Population Index 29(4):345–366. [PubMed: 12335951]
- Degenne, A., and M.Forse 1999. Introducing Social Networks . London: Sage Publications.
- Ellison, G., and D.Fudenberg 1993. Rules of thumb for social learning. Journal of Political Economy 101(4):612–643.
- Entwisle, B., and J.Godley 1998. Village Networks and Patterns of Contraceptive Choice. Unpublished paper presented at National Academy of Sciences Workshop on Social Processes Underlying Fertility Change in Developing Countries, January 29–30, 1998, Washington, D.C.
- Entwisle, B., R.R.Rindfuss, D.K.Guilkey, A.Chamratrithirong, S.P.Curran, and Y. Sawangdee 1996. Community and contraceptive choice in rural Thailand: A case study of Nang Rong. Demography 33(1):1–11. [PubMed: 8690134]
- Erbring, L., and A.A.Young 1979. Individuals and social structure: Contextual effects as endogenous feedback. So ciological Methods and Research 7(4):396–430.
- Festinger, L. 1954. A theory of social comparison processes. Human Relations 7:114–140.
- Freedman, R. 1979. Theories of fertility decline: A reappraisal. Social Forces 58(1): 1–17.
- Freedman, R., and Y.Takeshita 1969. Family Planning in Taiwan: An Experiment in Social Change . Princeton: Princeton University Press.
- Friedkin, N.E. 1998. A Structural Theory of Social Influence . Cambridge, Eng.: Cambridge University Press.
- Geertz, C. 1973. An Interpretation of Culture . New York: Basic Books.
- Gladwell, M. 2000. The Tipping Point: How Little Things Can Make a Big Difference . Boston: Little, Brown.
- Godley, J. 2001. Kinship networks and contraceptive choice in Nang Rong, Thailand. Interna tional Family Planning Perspectives 27(1):4–10.
- Granovetter, M., and R.Soong 1983. Threshold models of diffusion and collective behavior. Journal of Mathematical Sociology 9:165–179.
- 1986. Threshold models of interpersonal effects in consumer demand. Journal of Eco nomic Behavior and Organization 7:83–99.
- Greenhalgh, S. 1988. Fertility as mobility: Sinic transitions. Population and Development Review 14(4):629– 674.
- Gregersen, H., and L.Sailer 1993. Chaos theory and its implications for social science research. Human Relations 46(7):777–802.
- Griliches, Z. 1957. Hybrid corn: An exploration in the economics of technical change. Econometrica 25:501–522.
- Hallinan, M.T. 1997. The sociological study of social change. American Sociological Review 62(1):1–11.
- Hechter, M. 1993. Values research in the social and behavioral sciences. Pp. 1–30 in The Origin of Values , M.Hechter, editor; , L.Nadel, editor; , and R.E.Michod, editor. , eds. New York: Aldine de Gruyter.
- Hedström, P. 1998 Rational imitation. Pp. 306–327 in Soci al Mechanisms: An Analytical Approach to Social Theory, P.Hedström and, editor; R.Swedberg, ed, editor. s. Camb ridge, Eng.: Ca: m bridge; University Press.
- Hedström, P., and R.Swedberg 1998. Social mechanisms: An introductory essay. Pp. 1–31 in Social Mechanisms: An Ana lytical Approach to Social Theory , P.Hedström, editor; and R.Swedberg, editor. , eds. Cambridge, Eng.: Cambridge University Press.
- Katz, E., and P.Lazarsfeld 1955. Personal Influence: The Part Played by People in the Flow of Mass Communications . New York: Free Press.
- Kearney, M. 1995. The local and the global: The anthropology of globalization and transnationalism. Annual Review of Anthropology 24:547–565.
- Kertzer, D.I., and D.P.Hogan 1989. Family, Political Economy, and Demographic Change . Madison: University of Wisconsin Press.
- Kincaid, D.L., S.Pathak, and S.N.Mitra 1996. Communication Networks and Contraceptive Behavior in Bangladesh. Unpublished paper presented at the annual meeting of the Population Association of America, May 9–11, 1996, New Orleans.
- Knack, S., and P.Keefer 1997. Does social capital have an economic payoff? A cross-country investigation. Quar terly Journal of Economics 112:1251–1288.
- Knodel, J. 1977. Family limitation and the fertility transition: Evidence from the age patterns of fertility in Europe and Asia. Population Studies 31(2):219–249. [PubMed: 22077839]
- Knodel, J., N.Havanon, and A.Pramualratana 1984. Fertility transition in Thailand: A qualitative analysis. Population and Development Review 10(2):297–328.
- Knodel, J., and E.van de Walle 1979. Lessons from the past: Policy implications of historical fertility studies. Population and Development Review 5(2):217–245.
- Kohler, H.-P. 1997. Learning in social networks and contraceptive choice. Demography 34(3):369–383. [PubMed: 9275246]
- 2000. Social interactions and fluctuations in birth rates. Population Studies 54(2):223–237.
- Kohler, H.-P., J.R.Bereman and S.C.Watkins 2001. The density of social networks and fertility decisions: Evidence from South Nyanza District, Kenya. Demography 38(1):43–58. [PubMed: 11227844]
- Kreager, P. 1993. Anthropological demography and the limits of diffusionism. Pp. 313–326 in Inter national Population Conference, Montreal 1993 , Volume IV. Liège, Belgium: Ordina.
- Lee, R.D., and J.B.Casterline 1996. Introduction. Pp. 1–15 in Fertility in the United States: New Patterns, New Theories , J.B.Casterline, editor; , R.D.Lee, editor; , and K.A.Foote, editor. , eds. Supplement to Population and Devel opment Review 22. New York: Population Council.
- Lesthaeghe, R. 1983. A century of demographic and cultural change in Western Europe: An exploration of underlying dimensions. Population and Development Review 9(3):411–435.
- Lesthaeghe, R., and J.Surkyn 1988. Cultural dynamics and economic theories of fertility change. Population and De velopment Review 14(1):1–45.
- Lin, N. 1999. Building a network theory of social capital. Connections 22(1):28–51.
- Livi-Bacci, M. 1999. The Population of Europe: A History . Malden, MA: Blackwell Publishers.
- Marsden, P.V., and N.E.Friedkin 1993. Network studies of social influence. Sociological Methods and Research 22(1):127– 151.
- Mason, K.O. 1992. Culture and the fertility transition: Thoughts on theories of fertility decline. Ge- nus 48(3–4): 1–14.
- McFadden, D.L., and K.E.Train 1996. Consumers' evaluation of new products: Learning from self and others. Journal of Political Economy 104(4):683–703.
- Merton, R.K. 1968. Social Theory and Social Structure . New York: Free Press.
- Molm, L.D. 1997. Coercive Power in Social Exchange . Cambridge, Eng.: Cambridge University Press.
- Montgomery, M.R. 2000. Perceiving mortality decline. Population amd Development Review 26(4):795–819.
- Montgomery, M.R., and J.B.Casterline 1993. The diffusion of fertility control in Taiwan: Evidence from pooled cross-section, time-series models. Population Studies 47(3):457–479. [PubMed: 11613198]
- 1996. Social learning, social influence, and new models of fertility. Population and Devel opment Review (Supplement) 22:151–175.
- 1998. Social Networks and the Diffusion of Fertility Control . Policy Research Division Working Paper No. 119. New York: Population Council.
- Montgomery, M.R., and W.S.Chung 1999. Social networks and the diffusion of fertility control in Korea. Pp. 179–209 in Dynamics of Values in Fertility Change , R.Leete, editor. , ed. Oxford: Oxford University Press.
- Munshi, K., and J.Myaux 1998. Social Effects in the Demographic Transition: Evidence from Matlab, Bangladesh. Unpublished paper, Department of Economics, Boston University.
- Narayan, D., and L.Pritchett 1999. Cents and sociability: Household income and social capital in rural Tanzania. Economic Development and Cultural Change 47(4):871–898.
- Notestein, F.W. 1945. Population—the long view. Pp. 36–57 in Food for the World , T.W.Shultz, editor. , ed. Chicago: University of Chicago Press.
- 1953. Economic problems of population change. In Proceedings of the Eighth International Conference of Agricultural Economists . London: Oxford University Press.
- Ogburn, W.F. 1922. Social Change . New York: B.W.Huebsch, Inc.
- Palmore, J.A. 1967. The Chicago snowball: A study of the flow of influence and diffusion of family planning information. In Sociological Contributions to Family Planning Research , D.J. Bogue, editor. , ed. Chicago: Community and Family Study Center, University of Chicago.
- Pearl, J. 2000. Causality: Models, Reasoning, Inference . Cambridge, Eng.: Cambridge University Press.
- Pescondido, B.A. 1992. Beyond rational choice: The social dynamics of how people seek help. American Journal of Sociology 97:1096–1138.
- Pollak, R.A., and S.C.Watkins 1993. Cultural and economic approaches to fertility: Proper marriage or mesalliance? Population and Development Review 19(3):467–496.
- Portes, A. 1998. Social capital: Its origins and applications in modern sociology. Annual Review of Sociology 22:1–24.
- Potter, J. 1999. The persistence of outmoded contraceptive regimes: The cases of Mexico and Brazil. Population and Development Review 25(4):703–740.
- Putnam, R.D. 1995. Bowling alone: America's declining social capital. Journal of Democracy 6:65–78.
- Retherford, R. 1985. A theory of marital fertility transition. Population Studies 39(2):249–268.
- Retherford, R., and J.Palmore 1983. Diffusion processes affecting fertility regulation. Pp. 295–339 in Determinants of Fertility in Developing Countries , Volume 2 , R.A.Bulatao, editor; and R.D.Lee, editor. , eds. New York: Academic Press.
- Rogers, E.M. 1962. Diffusion of Innovations . New York: Free Press.
- 1973. Communication Strategies for Family Planning . New York: Free Press.
- Rogers, E.M., and D.L.Kincaid 1981. Communication Networks: Toward A New Paradigm for Research . New York: Free Press.
- Rosero-Bixby, L., and J.B.Casterline 1993. Modelling diffusion effects in fertility transition. Population Studies 47(1):147–167.
- Rutenberg, N., and S.C.Watkins 1997. The buzz outside the clinics: Conversations and contraception in Nyanza Province, Kenya. Studies in Family Planning 28(4):290–307. [PubMed: 9431650]
- Ryan, R., and N.Gross 1943. The diffusion of hybrid seed corn in two Iowa communities. Rural Sociology 8(1): 15–24.
- Santow, G. 1995. Coitus interruptus and the control of natural fertility. Population Studies 49(1):5–18.
- Santow, G., and M.D.Bracher 1999. Traditional families and fertility decline: Lessons from Australia's Southern Europeans. Pp. 51–77 in Dynamics of Values in Fertility Change , R.Leete, editor. , ed. Oxford: Oxford University Press.
- Sathar, Z., and J.B.Casterline 1998. The onset of fertility transition in Pakistan. Population and Development Review 24(4):773–796.
- Schoen, R., Y.J.Kim, C.A.Nathanson, J.Fields, and N.M.Astone 1997. Why do Americans want children? Population and Development Review 23(2):333– 358.
- Shiller, R. 1995. Conversation, information and herd behavior. American Economic Review 85(2): 181–185.
- Strang, D. 1991. Adding social structure to diffusion models: An event history framework. Socio logical Methods and Research 19(3):324–353.
- Strang, D., and N.Tuma 1993. Spatial and temporal heterogeneity in diffusion. American Journal of Sociology 99(3):614–639.
- Szreter, S. 1993. The idea of demographic transition and the study of fertility change: A critical intellectual history. Population and Development Review 19(4):659–701.
- Thompson, W.S. 1929. Population. American Journal of Sociology 34(6):959–975. United Nations.
- 2000. World Population Prospects, the 1998 Revision: Volume III, Analytical Report . New York: United Nations.
- Valente, T.W. 1995. Network Models of the Diffusion of Innovations . Cresskill, NJ: Hampton Press.
- Valente, T.W., P.R.Poppe, and A.P.Merritt 1996. Mass-media-generated interpersonal communication as sources of information about family planning. Journal of Health Communication 1:247–265. [PubMed: 10947363]
- Valente, T.W., and E.M.Rogers 1995. The origins and development of the diffusion of innovations paradigm as an example of scientific growth. Science Communications 16(3):242–273. [PubMed: 12319357]
- Valente, T.W., and W.P.Saba 1998. Mass media and interpersonal influence in a reproductive health communication campaign in Bolivia. Communication Research 25(1):96–124.
- Valente, T.W., S.C.Watkins, M.N.Jato, A.van der Straten, and L.-P.M.Tsitsol 1997. Social network associations with contraceptive use among Cameroonian women in voluntary associations. Social Science and Medicine 45(5):677–687. [PubMed: 9226791]
- Van den Bulte, C, and G.L.Lilien in Medical innovation revisited: Social contagion versus marketing effort. American press Journal of Sociology .
- van de Walle, E. 1992. Fertility transition, conscious choice and numeracy. Demography 29(4):487–502. [PubMed: 1483538]
- van de Walle, E., and J.Knodel 1967. Demographic transition and fertility decline: The European case. Pp. 47–55 in Contributed Papers Sydney Conference , International Union for the Scientific Study of Population. Canberra, Australia: Australian National University Press.
- Wasserman, S., and K.Faust 1994. Social Network Analysis: Methods and Applications . Cambridge, Eng.: CambridgeUniversity Press.
- Watkins, S.C. 1986. Conclusions. Pp. 420–449 in The Decline of Fertility in Europe , A.J.Coale, editor; and S.C. Watkins, editor. , eds. Princeton: Princeton University Press.
- 1990. From local to national communities: The transformation of demographic regimes in Western Europe, 1870–1960. Population and Development Review 16(2): 241–272.
- 1991. More Lessons from the Past: Women's Informal Networks and Fertility Decline. Unpublished paper presented at the International Union for the Scientific Study of Population (IUSSP) seminar on The Onset of Fertility Decline in Sub-Saharan Africa, Harare, November 19–22,1991, Zimbabwe.
- Watkins, S.C., and D.Hodgson 1998. From Mercantilists to Neo-Malthusians: The International Population Movement and the Transformation of Population Ideology in Kenya. Unpublished paper presented at National Academy of Sciences Workshop on Social Processes Underlying Fertility Change in Developing Countries, January 29–30 , 1998, Washington, D.C.
- Watts, D.J. 1999. Small Worlds: The Dynamics of Networks Between Order and Randomness . Princeton: Princeton University Press.
- Woolcock, M. 1998. Social capital and economic development: Towards a theoretical synthesis and policy framework. Theory and Society 27:151–208.
Footnotes
John B.Casterline is senior research associate at The Population Council, New York.
- PubMedLinks to PubMed
- Diffusion Processes and Fertility Transition: Introduction - Diffusion Processes...Diffusion Processes and Fertility Transition: Introduction - Diffusion Processes and Fertility Transition
- Atlapetes aconitase-1 (ACO1) gene, intron 15.Atlapetes aconitase-1 (ACO1) gene, intron 15.PopSet: 2260376545PopSet
- PopSet Links for Nucleotide (Select 2260376552) (1)PopSet
- NADH dehydrogenase subunit 3 (mitochondrion) [Atlapetes personatus personatus]NADH dehydrogenase subunit 3 (mitochondrion) [Atlapetes personatus personatus]gi|2260392522|gb|USS19609.1|Protein
- Atlapetes personatus paraquensis voucher AMNH:DOT4895 NADH dehydrogenase subunit...Atlapetes personatus paraquensis voucher AMNH:DOT4895 NADH dehydrogenase subunit 3 (ND3) gene, complete cds; tRNA-Arg gene, complete sequence; and NADH dehydrogenase subunit 4L (ND4L) and NADH dehydrogenase subunit 4 (ND4) genes, complete cds; mitochondrialgi|2260392709|gb|OL792683.1|Nucleotide
Your browsing activity is empty.
Activity recording is turned off.
See more...