Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 208
Diffusion Processes and Fertility Transition: Selected Perspectives 7 Mass Media and Fertility Change ROBERT HORNIK AND EMILE MCANANY Does the spread of access to mass media in a society affect the rate of decline in fertility? This essay organizes the thicket of issues bundled in that simple question and then reviews the relevant evidence, which will provide some tentative answers. We close with some comments about what we believe to be the most promising ways to think about this issue. There is evidence of a very substantial association between access to mass media and the level of fertility in a country. That is true at the most aggregated level, when national levels of per capita television access are used to predict fertility rates. There is a strong association of the natural log of fertility with televisions per 1,000 population for 144 countries for which data are available (Figure 7–1). Excluding four oil-rich outlier countries, a regression equation fitting a linear and quadratic term shows that televisions per capita accounted for 74 percent of the variance in fertility in 1997. This is a substantially better prediction of fertility than one gets from measures of gross national product per capita (indicated by an index of purchasing power parity) or indices of female education.1 Robert Hornik is Wilbur Schramm professor of communication and health policy in the Annenberg School for Communication at the University of Pennsylvania. Emile McAnany is the Walter E.Schmidt, S.J.professor of communication and chair of the communication department at Santa Clara University. The authors extend a special note of gratitude to R. Kirkland Ahern, Jennifer Horner, and Jo Stryker for research assistance, as well as Joseph Cappella for insightful suggestions in the preparation of this paper.
OCR for page 209
Diffusion Processes and Fertility Transition: Selected Perspectives FIGURE 7–1 Fertility by television (N=144), normal log of fertility rate, televisions per 1000 population. The association of media access and fertility is also found at the subnational level; for example, Potter and his colleagues (1997) show the significant association between television ownership and fertility at the municipality level in Brazil. It is true also at the individual level; for example, Westoff and Bankole (1997) show the association between various types of media access and contraceptive behaviors and childbearing intentions in seven African countries. In that study, married women who are exposed to television are two to six times more likely to be current users of contraception than the nonexposed comparison group. However, although these associations are consistent with a claim of media effects on fertility, they are far from definitive. They are not definitive, of course, because causal inferences from cross-sectional bivariate correlations are always tentative. Even though these associations generally are reduced but still significant when likely alternative explanations for them (such as education, urban residence, or occupation) are statistically controlled, inferences remain unsure. But they are also not definitive because, even if we were willing to make causal inferences from them, they do not tell us how the observed associations came about. There is quite a broad set of explanations that might suggest mechanisms through which access to mass media in a society might affect fertil-
OCR for page 210
Diffusion Processes and Fertility Transition: Selected Perspectives ity, even if we were willing to make a causal inference about that relation. We develop five major alternative (but not mutually exclusive) hypotheses. The first is about effects of the medium itself, the way it is used, separately from its content. The next three focus on the effects of ideas, each highlighting a different possible effect of the ordinary content of radio or television. The last considers effects of deliberate programs aimed at fertility control. It will turn out that tests of this last area of effect, evaluations of family planning communication programs, are the most common on the ground, but we suspect that the middle three hypotheses may deserve the most attention. FIVE HYPOTHESES ABOUT MEDIA EFFECTS ON FERTILITY Economic and Time Competition Effects First, most hypotheses about media effects focus on their content, the effects of exposure to ideas in the media. However, the central fact of exposure to media is the act of exposure—the time spent with the medium. Given the fixed boundaries of the 24-hour day, time spent with mass media may mean time lost from other activities. This effect may be particularly important when there are sharp increases in television ownership in individual homes. Individual ownership and the resulting heavy use of television may preclude or make some other activities more difficult, such as socializing with others at a central meeting place. If enough people are engaged with television in their homes, then the availability of such places for socializing in the community may be lost. This is an argument made strongly by Putnam (1995) in his oft-quoted paper “Bowling Alone,” which attributes the decimation of Americans’ social involvement to the spread of television. How might this relate to fertility? We can only speculate. In some societies courting behavior precedes marriage. If the amount of socializing young people do is reduced by watching television at home and age at marriage is responsive to the amount of socializing potential mates do with marriage waiting until a certain amount of socializing has occurred, then television might produce delays in age at marriage with resultant reductions in fertility. Or, if nonmarital sex and resulting pregnancy are a function of the amount of time people spend socializing, and television means people spend more time at home and have less opportunity for nonmarital sex, there may be less nonmarital fertility. Television may also compete with marital sex. If sex is, in part, a response to the lack of other things to do, and television watching takes up time that would otherwise be spent in sexual activity, fertility reductions might result. Or, contrarily, if television keeps spouses at home, and
OCR for page 211
Diffusion Processes and Fertility Transition: Selected Perspectives they spend more time in each other’s company, the opportunities for sex may be increased. Anecdotal reports of increased births about nine months after power blackouts in the United States are consistent with this suggestion. Availability of power and access to television may suppress fertility. Clearly the credibility of these noncontent-based mechanisms depends on the statement of plausible paths of effect and evidence to corroborate them. A second noncontent-related path of effect relates to the cost of purchasing a television. In poorer countries such a purchase is a significant expense relative to annual income. In Brazil, for example, there is evidence that television sets are often the first major consumer purchase, even in households with quite low incomes. If fertility reflects a deliberate allocation of scarce resources to children versus other forms of expense, then the decision to purchase and maintain a television may directly compete with childbearing, and reduce fertility. The more usual version of this competing use of resources hypothesis, which focuses on the materialist content of television and its influence on individual interests in consumption, is discussed below. These time and cost models are less commonly proposed than ideational models, perhaps for good reason. We turn to them now. Mass Media Effects on Elites Public policy in a country may favor or constrain fertility control. There is good evidence that deliberate family planning efforts have accelerated the decline in fertility (Freedman, 1997). There is also a fair consensus that other forms of public investment, in areas such as health care, female education, and social security, have important effects on fertility decisions. The question to be raised here is what leads elites to allocate resources to such programs? Is it possible that the ideas of elites on these issues are in some part the reflection of the world they see pictured on their television screens? Over a number of years beginning in the 1960s, American television programming dominated other countries’ entertainment (Varis, 1984.) In recent years, however, a number of larger countries have managed to produce more of their own prime time fare, which generally is more popular. Even so, such national programming reflects cosmopolitan points of view from the national capital or other large urban centers, and may still reflect the production norms of foreign program sources. Elites may have more direct contact with foreign cultures, regardless of television use, than do nonelites and especially rural audiences, but television and other media may reinforce and add to these direct influences on elites within a country. Thus values such as female equality, small families,
OCR for page 212
Diffusion Processes and Fertility Transition: Selected Perspectives open discussion of sexuality and contraception, and the acceptance of medical and other technology as a solution to problems all may be typical of imported or cosmopolitan national content and affect how elites view the world and what policy decisions they make. Mass Media Effects on Population Values Indirectly Relevant to Fertility Decisions Like the elites described above, general populations, if they watch television, watch content that typically comes from outside national boundaries, or they watch programs produced by writers and directors who are influenced by an international standard or a national cosmopolitan view. Although there are exceptions, the values of audiences who are not yet committed to reduced fertility may differ from those implicit in television content. That content, full of advertising of bright goods, may glorify consumption: when viewers must choose between consumption and childbearing, they may move toward consumption. Also, that content may tell stories which legitimate values associated with reduced fertility: a soap opera tells the story of a woman who delays marriage so she can achieve in a career. The value supporting achievement over early marriage influences viewers, which turns into lessened fertility. Locally produced dramas show that well-educated women are respected or that their education has a high economic return—parents accept the value of girls’ education; increases in girls’ education result in reduced fertility. (See very similar arguments in Faria and Potter, 1999; Westoff and Bankole, 1997; Kottak, 1990.) Fadul and colleagues (1996) show in a thematic analysis of popular telenovelas how, over a twenty-five-year period in Brazil, women’s roles have changed from housewives, teachers, and nurses to other professional roles outside the home and even to being portrayed as CEOs of major corporations. The authors also point to the consistently smaller family size on these soap operas than the current norms in Brazil, and how soap operas provided an early introduction to controversial ideas such as divorce, abortion, and premarital sex to a large Brazilian audience. News broadcasts may highlight reductions in child mortality, so that couples may take those reductions confidently into account when they decide how many children to bear. Coale (cited in Mason, 1997) argued for the importance of acceptance of the idea that it is possible to limit family size as a precondition to the fertility transition. Small families on television may carry that message to viewers. Television programs may say what it means to be modern, and that to be modern is good. These value diffusion effects may occur regardless of whether there is any explicit mention of the value of reduced fertility.
OCR for page 213
Diffusion Processes and Fertility Transition: Selected Perspectives Unintended Direct Mass Media Effects on Fertility Decisions Media content may not only carry messages about values, which are indirectly related to fertility; they may also carry content that bears directly on the subject. These messages may be included in news programming, in dramatic or other content, for their news or dramatic value, without any conscious intention to influence behavior. Soap operas may have scenes with couples deciding to limit family size or delaying having children or scenes with unmarried couples using birth control to avoid pregnancy. News programs may provide information about available methods of fertility control, or about services available in local clinics, or about new governmental policies in support of reduced fertility. Mass Media Effects on Fertility Through Deliberate Information, Education, and Communication Programs Family planning programs in many countries have tried to use mass media-based programs to influence contraceptive decisions and behavior—sometimes trying to increase awareness of the need for fertility control, sometimes providing information to increase utilization of available services, sometimes marketing specific products. Those programs may use a variety of approaches: didactic programs, advertising approaches, entertainment education approaches. They may use a variety of educational strategies: providing straightforward information; modeling desirable behaviors (e.g., showing “typical” people going into clinics for services, using condoms, or discussing family planning with partners); or emphasizing social norms. These programs are not always different in purpose than ordinary population education efforts, with one exception: in each case, program developers use mass media as the primary channel, assuming that media will improve cost-effectiveness and fidelity. They may believe that mass media are potentially less effective per person reached than alternative strategies (through outreach workers or through clinic-based staff), knowing that a field agent has the great advantage of being able to respond to the needs of individuals. But they expect that the heightened control over the messages delivered, and particularly the “reach and frequency” achieved through wide exposure to mass media, should counterbalance that advantage. MAJOR ISSUES TO CONSIDER WHEN THINKING ABOUT MEDIA EFFECTS These are all paths of influence for media effects. Regardless of the path, however, it would be a mistake to think about media influence as a
OCR for page 214
Diffusion Processes and Fertility Transition: Selected Perspectives simple direct learning phenomenon: that people learn about an idea or a behavior from mass media and they immediately put it into practice. Little mass communication theory relies on such an assumption. Evaluation of the evidence about media effects has to take into account the process through which the effects might occur. Before turning to the presentation of evidence about each hypothesis, we review some central issues in mass communication effects theory, relevant to developing expectations for what the evidence might show. The Issue of the Social Process Around Media Effects Media may be influential because of their direct effects on what people know and do, but they are also likely to be effective because they affect the social process around individual decision making. Media content may serve as a stimulus for discussion in which multiple individuals exposed to the same message develop a shared response; exposure to media content may create permission for open discussion of taboo topics in social networks; people considering adopting a new behavior inconsistent with the way things are done locally may gain from media exposure a sense of belonging to a broader social network; one individual exposed to media content may pass on what he or she learns to others inside his or her social network. Fadul et al. (1996) documented how controversial topics were introduced into Brazilian society over time through soap operas. Kothari (1997) explored the controversy surrounding Islamic practices regarding women that were introduced in popular soap operas in Pakistan by important women writers. A soap opera in Mexico (San Jose Mercury, 3 January 1998, 4A) stirred controversy by directly attacking in its story line the macho double standard about marital fidelity. In all cases the public nature and popularity of national television drama opens up opportunities to discuss sensitive social issues, including those regarding sexuality, divorce, abortion, and vasectomy. This social view of the effects process has an important implication for the search for evidence. It says that the effects of mass media diffuse beyond those who are directly exposed to the content. Then comparisons between those individuals who are or are not exposed to media may tend to underestimate effects. Thus if person A is exposed to televised messages and person B isn’t, but person A shares the content of the messages with person B, then a measure of exposure will show a difference between them but such exposure will not be associated with effect, even if there are effects. This idea of a diffused effect, likely to be lost if one focuses on individual differences, also can operate at a higher level of aggregation. If a certain proportion (about 30 percent) of a village is exposed to televi-
OCR for page 215
Diffusion Processes and Fertility Transition: Selected Perspectives sion, then the entire village may be substantially exposed because the messages are shared throughout the social network. A comparison of villages with 30 percent versus 60 percent access may not show the true effects of exposure, because the effects of exposure may be diffused in both places. We are often appropriately concerned about the ecological fallacy when we risk making erroneous inferences about individuals from aggregated data. But also we need to be careful on the other side. We can miss an effect if the evidence is based on individual comparisons of exposed and not exposed individuals, but the appropriate unit of effect is not the individual but the social network with whom the individuals affiliate. Media May Be Effective Because They Change the Frame People Use to Think About an Issue Media may be influential because people learn specific messages, but they are also influential because they influence the frame through which people process information (compare Iyengar and Kinder, 1987). This phenomenon is sometimes called agenda setting (compare McCombs et al., 1997). Here is an example regarding a woman’s decision to obtain a modern method of contraception. Assume that two beliefs influence her decision: her (lack of) belief that her living children will survive to adulthood, and her belief that repeated pregnancies will make her less attractive to her husband. Media content may provide information about one or another of these beliefs, but it also may provide messages that influence which of the beliefs should matter more in her calculus of the contraceptive decision. If television tells her she ought to weigh her potential attractiveness to her husband more highly, then she might adopt contraception, without changing the level of either belief. These possible framing effects of exposure to media content may complement effects on the levels of beliefs or attitudes. Research that focuses only on level changes but ignores complementary framing effects may underestimate effects. The Issue of Active Audiences Versus Universal Effects (Strong Audiences Versus Strong Media) Since the beginnings of mass media research, a consistent controversy has been the extent to which effects are universal, widely shared among audience members, or contrarily, highly conditional on characteristics of individual audience members and their social networks. One view focuses on the common effects, stressing the widely shared experience of media viewing, and while recognizing that individuals will vary in response, they are mostly interested in the main effects (compare Gerbner et
OCR for page 216
Diffusion Processes and Fertility Transition: Selected Perspectives al., 1994; Schiller, 1976). These theorists are interested in how society at large is affected by media, and assume that those effects will be large. Others insist that audiences make use of media largely to solve their own problems and are highly active in the process, reading the content, making sense of it, from within their particular belief system and social context (compare Katz and Liebes, 1990). These scholars tend to expect more limited effects of media as media influence is constrained by the characteristics of audiences. This theoretical dispute has important implications when we begin the search for evidence for effects of mass media on fertility. If the universal effects hypothesis is substantially true, then main effects are the appropriate place to focus our analyses, whether those were effects on knowledge or attitudes or behavior. We would look for noticeable average effects across the entire population. If, in contrast, effects are highly conditional on the beliefs and contexts of individuals, then main effects may be small, and only interaction terms will show effects. In that case a focus on the average correlation in a population, or the average effects of a media-based purposive program, may be misleading; if they are small they may be obscuring larger effects for particular subgroups for which the message was most relevant. An example: a media program that provides information about clinic services may be quite effective for the subset of people who are currently seeking contraception and are unaware of available services. An evaluation that asks for the average change in the population may fail to include enough of those people to show worthwhile effects. Early AIDS education evaluations foundered when they looked for increased condom use among all sexually active adults. But when they focused on people who reported “casual” sexual encounters, they found sharp increases in condom use (compare Dubois-Arber et al., 1997). The Issue of Effects on Awareness Versus Effects on Behavior An additional controversy bears on the issue of where to look for effects. The established diffusion literature stresses a claim that media effects are largely on awareness, on getting people to think about a new behavior. In contrast, that literature is skeptical about whether media have much influence on decisions to adopt behaviors (Rogers, 1995), expecting interpersonal sources to be more influential at decision-making stages. The diffusionist view is analogous to a more recently popular version of the same argument, called stages of change theory. That theory expects people at different stages of a change process (labeled precontemplation, contemplation, ready for action, action, and maintenance— which are similar to the awareness, behavior, and other categories used
OCR for page 217
Diffusion Processes and Fertility Transition: Selected Perspectives by diffusion theorists) to be affected by different types of messages (Prochaska and DiClemente, 1983) rather than different channels. A contrary position is more inclined to credit media with direct influence on behavior change, pointing to the examples of advertising effects and some other evidence (compare Hornik, 1989). That hypothesis expects direct learning effects, but also credits media with effects because they initiate person-to-person discussion and diffusion, which produce change, and because they crystallize latent demand—telling people how to act to realize what it is they have chosen to do. The Issues of the Nature of the Relationships Between Media Exposure and Fertility Outcomes This issue of what media can affect also relates to two final issues—the time lag for effects and the shape of the relationship. Some effects will occur rapidly with a limited exposure to a message. For example, women ready to adopt a modern contraceptive may respond quickly to information about eased access to a clinic. On the other hand, some behaviors are more deeply ingrained in the belief system and social context of individuals. Individuals might be affected by media content but require a broad change in the public communication environment, with mass media and other channels of communication providing new messages supportive of changed behavior, but then one would not expect instant response. Rather, the time lag of effects can be in years or even decades. For example, there has been an extraordinary decline in cigarette smoking in the United States, so the current level is half or less what it was at the height of smoking prevalence. But that change occurred at a slow pace, one or two percent per year, often not detectable in the short term with typical samples we can study. Programs that operate for a year or two, and evaluations that are only able to detect substantial and rapid change, may not be informative for change depending on communication effects on world views. Gradual change over longer periods also suggests that consistent reinforcement of the change messages is necessary. Even though advertising campaigns, for example, can create important short-term changes in buyers’ behavior, the longer term strategies of companies such as Coca-Cola or Nike is to reinforce short-term gains with follow-up ads. If the mass media are generally promodern, as many have argued since Lerner (1958) first introduced the idea of modernization and media’s role in the process, then their reinforcing value may be considerable in promoting not only consumerist desires but also modern and even Western views about fertility ideals and practices. A related concern is about the shape of the curve associating media exposure to effects, both for individuals and for larger aggregates. At the
OCR for page 218
Diffusion Processes and Fertility Transition: Selected Perspectives individual level, will the effects of media exposure be the same when an individual moves from no exposure to one hour of daily exposure as they are when he or she moves from three hours to four hours of exposure? Or, will the effects on a village be the same when the tenth percentile of households gets televisions as when the fiftieth percentile gets television? Some effects might be larger at the high levels of exposure (e.g., the effects on decline in social interaction); others may show little additional effect after a low level of exposure is achieved (effect on knowledge about the existence of contraceptive methods). This issue explores three overlapping concerns: the absolute quantity of exposure, the distribution over time of the exposure, and the time lag before the exposure produces effects. Any assessment of the effects of mass media needs to ask whether the evidence used as the basis for inference is consistent with hypothesized expectations for how each of these operate. Thus far we have laid out the various paths through which mass media might influence fertility, and pointed to a series of conceptual controversies in mass media research that bear on the search for evidence. These discussions followed a brief presentation of the global evidence for an association between media spread and fertility. We now turn to a more extended treatment of the evidence. It contains two major sections: evidence for overall effects of media spread on fertility, and evidence for effects of specific programs that have used mass media to influence fertility. The first of these is relevant to all of the first four hypotheses. Each would predict an overall association between media access and fertility. The second section tests the final path of direct effects of purposive programs. EVIDENCE ABOUT OVERALL EFFECTS Studies Showing Associations at the Aggregate Level We have noted the very substantial correlation (.86) between television access and fertility rates across 140 countries for which data were available (excluding 4 oil-rich outliers). The relation is still strong (.76) when the analysis is restricted to lower income countries (the countries classified as below the median on the purchasing power parity index used by the World Bank). There are two typical threats to a causal inference from this type of association. One is a concern with causal order. Is it possible that the association reflects not the effects of television on fertility but the effects of fertility on television ownership? After all, we did previously argue that television and children compete for scarce family resources. Isn’t it
OCR for page 229
Diffusion Processes and Fertility Transition: Selected Perspectives cerned with defining how the reception process takes place and how audiences make sense of the ideas that come to them in the media (Katz and Liebes, 1990; Moores, 1993). There has been little attempt until now to link this body of evidence with evidence of change in receivers’ attitudes or behaviors, partly because the focus was exclusively on reception processes and partly because changes were viewed as longer term and sometimes indirectly related to the ideas contained in the programs. The link between reception and outcome is critical and one that Freedman (1997) has recently highlighted for understanding changes in fertility preferences due to family planning programs. He suggests that qualitative studies can add to the evidence about how ideational processes take place. He points out that the little evidence that we do have suggests the importance of discussion of demographic issues in social networks. He concludes: Interaction in social networks is likely to be an important basis for diffusion and legitimization of ideas about smaller families and birth control. But where do these ideas come from? Establishing their origins and pathways in the growing national and international networks, in the direct and indirect influence of local programs, and in the changing life situations of the population is a primary challenge, (p. 9) The argument of models three and four above is partly based on the assumption that large numbers of people are exposed daily to information about issues relating to norms and values of family relations, women’s roles, and sexuality coming from metropolitan or international mass media sources that generate reactions and discussions among audiences over long periods of time. Serial dramas like the Latin American telenovelas or equivalent genres in Asia, Africa, and the Arab countries often raise demographically relevant issues (Fadul et al., 1996; McAnany and La Pastina, 1994). Moreover, there is consistent evidence that audiences strongly identify with these stories and discuss issues of norms and values in family, neighborhood, and social groups (Brown, 1994, for English-speaking countries; McAnany and La Pastina, 1994, for Latin America; Mankekar, 1993, for India; Diase, 1996, for Egypt; and Kothari, 1997, for Pakistan). In an ongoing study of Brazilian television and demography (Potter et al., 1997; Faria and Potter, 1999; Rios-Neto et al., 1998; McAnany, 1997), a series of qualitative studies of audience reception and incorporation have been carried out to delineate how demographically relevant ideas enter social networks and circulate beyond the family. The preliminary evidence (Hamburger, 1998; La Pastina, 1997) suggests several pathways and processes: (1) new ideas about nontraditional roles for women and taboo ideas, such as divorce, abortion, and vasectomy, reach the public sphere over time through television narratives that reach all levels of a
OCR for page 230
Diffusion Processes and Fertility Transition: Selected Perspectives national audience; (2) audiences engage with these narratives, discuss the issues raised in them, and compare and apply them to their own lives; (3) other media such as newspapers, magazines, and radio, not only reinforce and popularize the stories but discuss issues and debates raised by government, religious, and social groups; (4) mothers and fathers sometimes intervene in television viewing by their children to discuss issues about sexuality, authority, and social roles; most parents express the conviction that their children are influenced by the models, norms, and values presented in the stories. All of the evidence suggests a pathway for entry of new ideas concerning modern value issues confronting national audiences. Some behavioral outcomes are immediate, such as the buying of recorded music appearing in the series or adopting new fashions and styles of characters, but most are longer term and less direct, such as attitudes about women’s work roles, the value of virginity before marriage, or the legitimization of divorce for women in loveless marriages. Historical records of the first treatment of such sensitive issues as divorce or abortion in a public medium such as national television (Fadul et al. 1996) provide some basis for tracking the debate surrounding such issues. The long-term nature of such changes, however, makes it difficult but not impossible to establish the link between changes in fertility practice and the spread of television (thus the integration of various sources of evidence within a single project like the one in Brazil is a critical step). There is one more conclusion to the audience reception literature: there are a wide variety of reactions to messages received. The macro studies reviewed earlier suggest that media exposure is somehow related to a series of similar practices resulting in lowered fertility. One is left with the impression that it is a single, simple process of influence. What ethnographic audience studies reveal is that audiences represent a wide variety of positions on value and normative issues. The commonality of audiences is that they receive on a daily basis the stories and situations that they will discuss and debate with family and neighbors. Moreover, this process is repeated again and again over months and years, with varying degrees of intensity, depending on the issue that a particular story may raise. The process outlined here is not, for the most part, an ideational campaign from government or other social agency. It is often random and even chaotic. The media, especially radio and increasingly television for developing countries, carry many messages in their programs, but in many countries the serial narratives making up the prime listening and viewing hours deal consistently with family matters that legitimate or at least open for debate problems facing their societies regarding women and their roles, families, premarital sex, and other modernizing pressures. Scriptwriters with an intent of raising debate (compare Kothari,
OCR for page 231
Diffusion Processes and Fertility Transition: Selected Perspectives 1997, on women’s issues in Pakistani soaps) often consciously introduce these issues, but there may not be any immediately observable change. Looked at in the long term, however, and linked with studies of media exposure and changes in fertility beliefs and practices, the process of ideational change takes on a clearer path. DISCUSSION Thus far we have painted a complex picture of evidence around the hypothesis that the spread of mass media is associated with reductions in fertility. The correlational evidence, both cross-sectional and longitudinal, but not associated with discrete programs, is consistent with such an effect. The evidence from discrete programs is much less convincing about broad effects, although it provides some evidence for (short-lived) increases in demand for clinic services. The evidence that falls between the two types, reporting on the association of exposure to family planning messages but not tied to a specific program, is more optimistic. However, we recognize there are problems with these data. The correlational studies do not permit us to reject the rival hypothesis that selection, the effects of unmeasured exogenous variables, accounts for the observed association. The discrete program evaluations have suffered from mischance in data collection, or too short a time frame for the intervention programs and for the estimation of effects, or too complex an environment to attribute effects cleanly. In some sense, the very attempt to make them open to study as discrete interventions may have worked against producing or measuring their success. The problem of openness to study relates directly to the issue of a proposed model of effect. How is it that mass media are to have their effects? The review of specific evidence just completed corresponds to two particular models of media effects. The first model often underpins the evaluations of discrete programs. They assume that a focused input can produce specific effects in the short term. This may be a reasonable model when the effects involve shifting people from intending to engage in a behavior to actually engaging in the behavior. A good example from the evidence presented here would be a program designed to stimulate those people who are already intending to use modern methods to actually visit the clinic to obtain a contraceptive. The evidence from several programs of increases in clinic demand, without corresponding evidence of a population change in contraceptive use, is consistent with this model. It would suggest that the success of short-term programs might rely on a substantial population of people who are ready to act and just need a final push. Although in these family planning examples this group was apparently small, limiting the size of population effects, that does not have to be the
OCR for page 232
Diffusion Processes and Fertility Transition: Selected Perspectives case. In another area, encouraging immunization of young children, there is evidence that the ready-to-act population is much larger and media campaigns can stimulate rapid changes in immunization rates. In the Philippines, the proportion of urban children fully immunized by their first birthday increased from 32 percent to 52 percent as a result of a one-year media campaign. It apparently worked mostly by teaching parents who were having their children vaccinated, but on a delayed schedule, that they needed to bring their children in before their first birthdays (Zimicki et al., 1994). However, there is also a second model of media effect, particularly relevant when behavior is not so close to realization. People may be aware of the new behavior but do not yet intend to do it, and then the model of effect may be different. Media may still have profound effects but the process may be slower and have a different character. Then the process of influence may take substantial time, and require reinforcement through repetition and transmission through multiple channels. This complex process of effects is described in studies of how people respond to prime time soap operas (compare McAnany and La Pastina, 1994; Brown, 1994). A program exposes television owners to messages directly; it generates conversation among those who watch it and with others not exposed; newspapers may regularly report about ongoing story lines, especially those stories that cause controversy. Soap operas, and other programs, address similar issues. This constant lapping against the shore, rather than a single large wave, may produce slow, long-term changes in values underlying fertility intentions. There is some recent work analyzing health communication programs in the more developed world that has bearing here. There have been a series of discrete intervention programs directed toward smoking and other heart disease-related behaviors. These have included the Stanford Heart Disease Prevention Program, the Minnesota Heart Health Program, and the Community Intervention Trial for Smoking Cessation (COMMIT) trial (Farquhar et al., 1990; COMMIT, 1995; Luepker et al., 1994). Each has compared treatment and control groups to measure the effects of media (and other IEC efforts) on behavioral outcomes. Each has had to conclude that the evidence is inconsistent with an effect, or at best that the evidence is ambiguous. These discrete programs may contrast with a series of uncontrolled efforts at public education that are associated with massive changes in behavior. To point to just two: the Swiss AIDS program was associated with an increase from 8 percent to 56 percent of young people always using condoms for casual sex (Dubois-Arber et al., 1997); the National High Blood Pressure Education Program (NHBPEP) was associated with a 57-percent drop in stroke rate in a 12-year period (Roccella and Lenfant, 1992). (See Hornik, 1997, for a fuller discussion of these cases and others.)
OCR for page 233
Diffusion Processes and Fertility Transition: Selected Perspectives One explanation for the difference between the small success of the discrete programs and major successes of the uncontrolled programs may be that the evaluations were of radically different quality. But a more important distinction may come from an understanding about how communication programs have big effects. First, there is relatively little evidence for big differences in exposure to messages over time between the treatment and control areas in the discrete studies, while for the messy programs (like Swiss AIDS and NHBPEP), the levels of exposure achieved are much higher. This may be less of a problem in many developing country contexts in which the media environment is less competitive. Indeed, in many of the evaluations summarized here, authors are able to point to quite substantial levels of basic exposure to messages. People may also change behavior when many aspects of their environments communicate new messages in a repeated and reinforcing way over time. They may change when there is substantial diffusion of those messages through social networks in the ways previously described. That is the way that programs that are considered strong family planning programs operate. Elsewhere one of us has speculated about the way that someone might initiate new practices related to high blood pressure: How might the NHBPEP have worked? One can imagine how the process of change occurs: a person sees some public service announcements and a local TV health reporter’s feature telling her about the symptomless disease of hypertension; she checks her blood pressure in a newly accessible shopping mall machine, those results suggest a problem; she tells her spouse who has also seen the ads and encourages her to have it checked; she goes to a physician who confirms the presence of hypertension, encourages her to change her diet and then return for monitoring. Meanwhile the physician has become more sensitive to the issue because of a recent Journal of American Medical Association article, and some recommendations from a specialist society, and a conversation with drug detailer, as well as informal conversations with colleagues and exposure to television discussion of the issue. The patient talks with friends at work or family members about her experience; they also increase their concern and go to have their own pressure checked. She returns for another checkup and her pressure is still elevated although she has reduced her use of cooking salt; the physician decides to treat her with medication. The patient is ready to comply because all the sources around her, personal, professional and mediated, are telling her that she should (Hornik, 1997:49–50). We would suspect that the story, while different in the details for a developing country, would not be radically different in essence for adoption of family planning behavior.
OCR for page 234
Diffusion Processes and Fertility Transition: Selected Perspectives This may be why many of the discrete evaluations are unable to find population-level effects that stick. They are evidence about media effects on fertility within the first model. If the question at issue here is how media affect fertility in the broadest sense, although these projects get the lion’s share of research, they may represent only quite a small share of the influence. They examine a small proportion of all the media content developed to affect fertility because for most programs media effects cannot be isolated from other program effects. Of greater moment, they are an even smaller proportion of all the media materials that may affect fertility, if we speculate that most fertility effects are due to hypotheses two, three, and four stated at the start—the ideational effects of ordinary media content. In that case, if what we are interested in is the long-term large effects of multifaceted programs, and of natural exposure to media content, the discrete evaluations are much less interesting than are the correlational studies. The correlational studies, both cross-sectional and longitudinal, address the model of effect that is most likely to have produced big fertility effects. And the correlational studies, for all their methodological tentativeness, support the claim that the spread of mass media affects fertility. The evidence base is certainly open to challenge, but here are conclusions and speculations based on what is available: There is good evidence that short-term campaigns can affect clinic demand, which may relate to shortening the time lag between intention and behavior for those who are ready to act. The evaluations of those campaigns have been less successful in establishing that they produce detectable population-level changes in behavior. Access to mass media is substantially related to fertility (or fertility-related behavior), at three levels of aggregation—individual, municipal, and national—even when other likely causes of the relation are controlled. There is some evidence that mass media access predicts subsequent changes in fertility-related behavior. This second type of evidence is the strongest support for the argument that the spread of media affects fertility on a scale that is important. However, even if that claim was accepted, the mechanism of effect is undetermined. We have suggested at least four paths for such influence: medium noncontent effects, influence on elites, influence on basic values of general audiences, and influence on fertility-specific knowledge of general audiences. Nonetheless, we would speculate that if the spread of mass media has effects on fertility, it reflects a complex social process rather than a medium effect or a discrete learning process: multiple channels, providing reinforcing messages, over time, producing inter-
OCR for page 235
Diffusion Processes and Fertility Transition: Selected Perspectives personal discussion and a slow change in values, and working at a level of social aggregation higher than the individual. Most of our discussion has focused on two types of effects: those related to discrete mass media interventions, and those related to general media access and exposure to ordinary media content. But there is a third type of possible effect: the effects of continuing messages delivered through the IEC efforts of general profamily planning programs. The explicit evidence relevant to these programs is that self-reported attention to media family planning messages, not in the context of a discrete intervention, is related to fertility-related behavior in the African sites studied by Westoff and Bankole. Most programs incorporate such efforts, but they are of unknown heft and quality, and of unknown effect. These programs cannot separate IEC or mass media messages from the rest of their activities, but they operate over the long term and with constant reinforcement of messages. In some ways they may behave more like the general media content, long-term effects model than they do the discrete, short-term effects model. More attention to these operational programs may produce evidence of long-term effects not seen with the short-term evaluations of discrete programs, just as the messy NHBPEP showed effects that the better evaluated but less comprehensive Stanford and COMMIT programs did not. NOTES 1. Analysis based on 1997 data from World Development Indicators (World Bank, 1999); the predicted variance including the omitted countries is still a very substantial 66.1 percent. 2. The self-reported exposure to family planning messages on radio is also controlled in these analyses. Because that measure is likely to be highly correlated with radio exposure, it probably artifactually reduces the observed associations with overall media use. REFERENCES Bankole, A., G.Rodriguez, and C.F.Westoff 1996 Mass media messages and reproductive behavior in Nigeria. Journal of Biosocial Science 28(2):227–239. Bertrand, J.T., R.Santiso, S.H.Linder, and M.A.Pineda 1987 Evaluation of a communications program to increase the adoption of vasectomy in Guatemala. Studies in Family Planning 18(6):361–370. Brown, M.E. 1994 Soap Opera and Women’s Talk: The Pleasure of Resistance. Thousand Oaks, CA: Sage Publications.
OCR for page 236
Diffusion Processes and Fertility Transition: Selected Perspectives Community Intervention Trial for Smoking Cessation (COMMIT) 1995 I.Cohort results from a four year intervention. American Journal of Public Health 85:183–192. Cook, T.D., and D.Campbell 1979 Quasi-Experimentation. Boston, MA: Houghton-Mifflin. Diase, M. 1996 Egyptian Television Serials, Audiences and “The Family House”: A Public Health Enter-Educate Serial. Unpublished Ph.D. dissertation, University of Texas at Austin. Dubois-Arber, F., A.Jeannin, E.Konings, and F.Paccaud 1997 Increased condom use without other major changes in sexual behavior among the general population in Switzerland. American Journal of Public Health 87(4):558– 559. Fadul, A., E.McAnany, and O.Torres Morales 1996 Telenovela and Demography in Brazil: A Working Paper. Unpublished paper presented at the International Association for Mass Communication Research, August 18–22, Sydney. Faria, V.E., and J.E.Potter 1999 Television, telenovelas and fertility change in northeast Brazil. Pp. 252–272 in Dynamics of Values in Fertility Change, Richard Leete, ed. Oxford, Eng.: Clarendon Press. Farquhar, J.W., S.P.Fortmann, J.A.Flora, C.B.Taylor, W.L.Haskell, P.T.Williams, N. Maccoby, and P.D.Wood 1990 Effects of communitywide education on cardiovascular disease risk factors. The Stanford five-city project. Journal of the American Medical Association 264(3) :359– 365. Ferreros, C. 1990 Social Marketing of Condoms For AIDS Prevention in Developing Countries: The Zaire Experience. Washington, DC: Population Services International. Freedman, R. 1997 Do family planning programs affect fertility preferences? A literature review. Studies in Family Planning 28(1):1–13. Gerbner, G., L.Gross, M.Morgan, and N.Signorielli 1994 Growing up with television: The cultivation perspective. In Media Effects: Advances in Theory and Research, J.Bryant and D.Zillman, eds. Hillsdale, NJ: Lawrence Erlbaum Associates. Hamburger, E. 1998 Producing Culture in Contemporary Brazil: Theories and Practices Involved in Production and Reception of Telenovelas. Working paper. Population Research Center, University of Texas at Austin. Harvey, P.D. 1991 The Correlation Between Consumer Prices and Per Capita Sales of Condoms in Seventeen Social Marketing Programs. Washington, DC: Population Services International. Hornik, R.C. 1989 Channel effectiveness in development communication programs. In Public Communication Campaigns, 2d ed., R.Rice and C.Atkins, eds. Beverly Hills: Sage. 1997 Public health education and communication as policy instruments for bringing about changes in behavior . Pp. 45–60 in Social Marketing: Theoretical and Practical Perspectives, M.Goldberg, M.Fishbein, and S.Middlestadt, eds. Hillsdale, NJ: Lawrence Erlbaum Associates.
OCR for page 237
Diffusion Processes and Fertility Transition: Selected Perspectives Hornik, R.C., ed. 2001 Public Health Communication: Evidence for Behavior Change. Mahnah, NJ: Lawrence Erlbaum Associates. Hornik, R.C., J.McDowell, J.Romero, and R.Pareja 1987 Communication and Health Literacy: Evaluation of the Peru Program 1984–1985. Philadelphia: Annenberg School, University of Pennsylvania and Washington, DC: Academy for Educational Development. Inkeles, A., and D.H.Smith 1974 Becoming Modern. Cambridge, MA: Harvard University Press, Iyengar, S., and D.R.Kinder 1987 News That Matters: Television and American Opinion. Chicago: University of Chicago Press. Kane, T.T., M.Gueye, I.Speizer, S.Pacque-Margolis, and D.Baron 1998 The impact of a family planning multimedia campaign in Bamako, Mali. Studies in Family Planning 29(3):309–323. Katz, E., and T.Liebes 1990 Interacting with Dallas: Cross cultural readings of American TV. Canadian Journal Of Communication 15:45–65. Kincaid, D.L., A.P.Merritt, L.Nickerson, S.Buffington de Castro, M.P.de Castro, and B.M. deCastro 1996 Impact of a mass media vasectomy promotion campaign in Brazil. International Family Planning Perspectives 22:169–175. Kothari, S. 1997 From Genre to Zanaana: Women and Urdu Drama Serials in Pakistan. Unpublished Ph.D. dissertation, University of Texas at Austin. Kottak, C.K. 1990 Prime Time Society: Anthropological Analysis of Television and Culture. Belmont, CA: Wadsworth. La Pastina, A. 1997 Watching “O rei do gado” in Timbabuba: A Short Account of an Ethnographic Research. Working paper. Population Research Center, University of Texas at Austin. Lerner, D. 1958 The Passing of Traditional Society. New York: Free Press. Luepker, R.V., D.M.Murray, D.R.Jacobs Jr., M.B.Mittelmark, N.Bracht, R.Carlaw, R. Crow, P.Elmer, J Finnegan, and A.R.Folsom 1994 Community education for cardiovascular disease prevention: Risk factor changes in the Minnesota Heart Health Program. American Journal of Public Health 84(9): 1383–1393. Mankekar, P. 1993 National texts and gendered lives: An ethnography of television viewers in a north Indian city. American Ethnologist 20(3):543–563. Mason, K.O. 1997 Explaining fertility transitions. Demography 34(4):443–454. McAnany, E. 1997 Television and Demographic Transition in Brazil: A Model for Studying Long Term Change. Unpublished paper prepared for Long Term Consequences Through Mass Media, University of Saarlandes, February, Saarbrucken, Germany. McAnany, E., and A.La Pastina 1994 Telenovela audiences: A review and methodological critique of Latin American research. Communication Research 21(6):828–849.
OCR for page 238
Diffusion Processes and Fertility Transition: Selected Perspectives McCombs, M, D.L.Shaw, and D.Weaver, eds. 1997 Communication And Democracy: Exploring The Intellectual Frontiers In Agenda-Setting Theory. Mahwah, NJ: Lawrence Erlbaum Associates. Moores, S. 1993 Interpreting Audiences: The Ethnography of Media Consumption. Thousand Oaks, CA: Sage Press. Piotrow, P.T., D.L.Kincaid, M.J.Hindin, C.L.Lettenmaier, I.Kuseka, T.Silberman, A. Zinanga, F.Chikara, D.J.Adamchak, M.T.Mbizvo, W.Lynn, O.M.Kumah, and Y.-M.Kim. 1992 Changing men’s attitudes and behavior: The Zimbabwe male motivation project. Studies in Family Planning 23(6 p 1):365–375. Piotrow, P.T., J.G.Rimon 2d, K.Winnard, D.L.Kincaid, D.Huntington, and J.Convisser 1990 Mass media family planning promotion in three Nigerian cities. Studies in Family Planning 21(5):265–274. Potter, J.E., R.M.Assuncao, S.M.Cavenaghi, and A.J.Caetano 1997 The Spread of Television and Fertility Decline in Brazil: A Spatial Temporal Analysis. Unpublished paper presented at the XXIIIrd IUSSP General Conference, October 11–17, Beijing, China. Prochaska, J.O., and C.C.DiClemente 1983 Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting & Clinical Psychology 51(3):390–395. Putnam, R.D. 1995 Bowling alone: America’s declining social capital. Journal of Democracy 6(1):65–79. Rios-Neto, E., P.Miranda-Rebeiro, and J.Potter 1998 I Saw It on TV: Television and Demographic Change in Brazil. Paper presented at the Workshop on the Social Dynamics of Fertility Change in Developing Countries, National Research Council, January 28–30, Washington, DC. Roccella, E.J., and C.Lenfant 1992 Family physicians and the battle against hypertension. (Editorial) American Family Physician 46(5):1390. Rogers, E.M. 1995 Diffusion of Innovations, 4th ed. New York: Free Press. Rogers, E.M., D.Solomon, and R.Adhikarya 1978 Further Directions for USAID’s Communication Policies in Population. Stanford: Institute for Communication Research. Rogers, E.M., P.W.Vaughan, R.M.A.Swalehe, N.Rao, P.Svenkerud, and S.Sood 1999 Effects of an entertainment-education radio soap opera on family planning behavior in Tanzania. Studies In Family Planning 30(3):193–211. Ross, J.A., and W.P.Mauldin 1996 Family planning programs: Efforts and results, 1972–1994. Studies in Family Planning 27(3):137–147. Schiller, H.I. 1976 Communication And Cultural Domination. White Plains, NY: Pantheon Books. Storey, D., M.Boulay, Y.Karki, K.Keckert, and D.M.Karmacharya 1999 Impact of the integrated radio communication project in Nepal, 1994–1997. Journal of Health Communication 4(4):271–294. Valente, T.W., Y.M.Kim, C.Lettenmaier, W.Glass, and Y.Dibba 1994 Radio promotion of family planning in the Gambia. International Family Planning Perspectives 20:96–100. Varis, T. 1984 The international flow of television programs. Journal of Communication 34(1):143– 152.
OCR for page 239
Diffusion Processes and Fertility Transition: Selected Perspectives Westoff, C.F., and A.Bankole 1997 Mass Media and Reproductive Behavior in Africa. Demographic and Health Surveys Analytical Reports No. 2, April. Calverton, MD: Macro International. Westoff, C.F., and G.Rodriguez 1995 The mass media and family planning in Kenya. International Family Planning Perspectives 21:1, 26–31, 36. World Bank 1999 World Development Indicators. Washington, DC: World Bank. Yoder, P.S., R.C.Hornik, and B.Chirwa 1996 Evaluating the program effects of a radio drama about AIDS in Zambia. Studies in Family Planning 27(4):188–203. Zimicki, S., R.C.Hornik, C.C.Verzosa, J.R.Hernandez, E.de Guzman, M.Dayrit, A.Fausto, M.B.Lee, and M.Abad 1994 Improving vaccination coverage in urban areas through a health communication campaign: The 1990 Philippine experience. Bulletin of the World Health Organization 72(3):409–22.
Representative terms from entire chapter: