A Glass Half Empty: Latina Reproduction and Public Discourse

Latina reproduction and fertility have become ground zero in a political war--not just of words but of public policies and laws. This article builds on a theoretical framework that includes issues of stratified reproduction which characterize some women as reproductive threats to society. From an examination of the discourse found in 10 national magazines over a 35-year period beginning in 1965 emerge three interrelated themes concerning Latina reproductive threat: 1) high fertility and population growth; 2) reconquest; and 3) overuse of medical and other social services. The final section examines data on reproduction and fertility collected from Latinas and Anglo women in Orange County California to explore the "truth claims" associated with Latina reproduction and fertility. The findings suggest that Latinas do not begin sexual activities at a relatively early age nor do they have relatively more sexual partners than Anglo women. Most Latinas have used birth control pills at some point in their lives. Latinas generally have fewer than two children per woman. Mexican-origin women born or raised in the Untied States had fewer children than adult immigrants and their differences from Anglo women were insignificant. Ordinary least squares (OLS) regression finds that age marital status education and language acculturation are more important than ethnicity for understanding fertility. (authors)

(women of Latin American origin in the United States) reproduction. Two key questions come out of this review and frame subsequent sections. First, how have Latina reproduction and fertility been constructed? To examine this question, Latina fertility and reproduction are analyzed as key intertwined concepts in a national public discourse on immigration, in a manner suggested by Fraser and Gordon's (1994) research on the keyword "dependency" in the welfare state. The genealogy of Latina "fertility and reproduction" as "threats to U.S. society" are traced in the visual and textual discourse found in 10 national magazines traced over a 35-year period, beginning in 1965 and continuing up to the end of 1999 (Chavez 2001). As Ginsburg and Rapp (1995:6) observe, "Representations provide the arena in which cultural understandings and hierarchies are produced, contested, and revealed." By tracing representations and characterizations of Latina fertility and reproductive capacities, we can generate important questions that become the focus of examination in the next part of this article.
Discourses that construct people with "dangerous," "pathological," and "abnormal" reproductive behaviors and beliefs are not simply of academic interest. There are real political and economic consequences to such constructions. In California, for example, the perceived threat of Latina fertility, especially among immigrants, was central to the "Save Our State" movement that led to Proposition 187, which sought to curb undocumented immigration by denying undocumented immigrants social services, particularly prenatal care and education for their children. Bette Hammond, one of the organizers of Proposition 187, characterized Latina immigrants in her hometown in a way that emphasized the threat of reproduction: "They come here, they have their babies, and after that they become citizens and all those children use social services" (Kadetsky 1994:418). Pete Wilson, governor of California from 1991 denying undocumented immigrant women prenatal care a top priority of his administration (Lesher and McDonnell 1996). The 1996 welfare reform law also targeted medical services for immigrant women (Fix and Passel 1999). As we shall observe below, the popular discourse of Latina reproduction is decidedly alarmist in that it becomes part of a discourse of threat and danger to U.S. society and even national security, which is underscored in a post-911 world. Thus discourses, because they not only filter reality but help construct what is taken for "real," have important material implications that warrant examination.
The second key question examined here is: Is the construction of Latina reproduction and fertility accurate, or is the story more complicated? The doxa, or taken-for-granted beliefs, of Latina reproduction derived from the previous analysis are examined from the vantage point of data on Latina reproductive behavior collected in Orange County, California. In essence, I use my research on media representations of immigration-related issues (Chavez 2001) to raise questions that are examined through the use of empirical data collected in another research project I was also involved in (Chavez et al. 1995;Chavez et al. 1997;Chavez et al. 2001). Although the two research projects were independent, the use of one research project to generate research questions for analysis with data from another exemplifies the serendipitous possibilities and even benefits of combining research in this way. The data on Latina fertility and reproduction were collected for a study of cancer and Latinas and were not collected with the thought of refuting a public discourse on Latina fertility. The Orange County data can also be compared to national data to pinpoint differences and similarities between the local and the national. Questioning the factual bases of the discourse surrounding the politics of Latina reproduction may, hopefully, contribute to the formulation of a new way of thinking about reproduction, immigration, and social change.
Before proceeding, it is important to examine the theoretical developments in anthropology on the politics of reproduction. An emerging anthropology of reproduction in general, and Latina reproduction in particular, raise critical issues and questions that guide the analysis undertaken here.

Anthropology, Reproduction, and Latinas
Faye Ginsburg and Rayna Rapp (1991) argue that "to reproduce" has many connotations. At the very least, it is important to distinguish biological reproduction from social reproduction. Both aspects of reproduction, as well as their intersection, are often sites of political confrontation. In societies with competing and often unequal social groups split along various lines of race, ethnicity, sexuality, and immigration status, the biological and/or social reproduction of one or all of those groups can be the target of public debate and state policies aimed at controlling reproduction (Horn 1994;Lock and Kaufert 1998). As Ginsburg and Rapp (1991:314) note "Throughout history, state power has depended directly and indirectly on defining normative families and controlling populations." In the process, regimes of representation can emerge in which particular groups are said to be pathological, even "dangerous" to the larger society. (See Gutiérrez [1999] for a thorough discussion of the racial politics of Latina reproduction.) Ginsburg and Rapp (1995) utilize Shellee Colen's (1990) concept of "stratified reproduction" to describe how for some groups women's reproduction is characterized positively, while that of other women is "disempowered." As they note, "The concept of stratified reproduction helps us see the arrangements by which some reproductive futures are valued while others are despised" (Ginsburg and Rapp 1995:3).
One particularly insidious example of stratified reproduction is that of the "black welfare mother" image used so effectively in political discourse (Fraser and Gordon 1994). As Dorothy Roberts (1997b:3) argues, society has blamed poor black mothers for "perpetuating social problems by transmitting defective genes, irreparable crack damage, and a deviant lifestyle to their children." African American women pose a "reproductive threat" that, as we shall see below, is different from that posed by Mexican immigrant women and their daughters, but both groups have faced the stigma of society's surveillance of their reproductive capacities. As Roberts (1997b:7) argues, "welfare reform measures that cut off assistance for children born to welfare mothers all proclaim the same message: The key to solving America's social problems is to curtail Black women's birth rates." (For research on sterilization of Latinas, see Lopez 1998, Vélez-Ibáñez 1980, and Vélez-Ibáñez 1999) Ginsburg and Rapp (1991:331) also suggest the importance of examining discourse, which is part of the analysis presented here. As they note, "The powerful tools of discourse analysis can be used to analyze 'reproduction' as an aspect of other contests for hegemonic control, such as state eugenic policies, conflicts over Western neocolonial influences in which women's status as childbearers represents national interests, or fundamentalist attacks on abortion rights as a part of a campaign to evangelize the American state." The analysis of a national discourse on Latina reproduction presented below will attempt to draw out these "other contests for hegemonic control." Ginsburg and Rapp's observations, especially concerning discourse, have implications for Latinas and the politics surrounding their reproduction. As Foucault (1977) argues, discourse produces objects of knowledge and meaning (see also Hall 1977). "Latinas" exist and "reproduction" exists, but "Latina reproduction" as an object of a discourse produces a limited range of meanings, often focusing on their "over" reproduction and their fertility and sexuality, which are "out of control" in relation to the supposed social norm. Latina biological reproduction combines with its social reproduction to produce fears about the population growth of Latinos in American society, which in turn positions them as a possible threat to the "nation," that is, the "people" as conceived in demographic and racial terms. Reproduction here is an ideological concept that defines normative fertility levels (that of Anglos or non-Latino whites) and their opposite: the nonnormative, stigmatized, "high fertility" level of Latinas and the sexual behavior that produced it.
Or put another way, Anglo women's fertility is not only normative, but they also possess "subject status," which Link (1991:40) defines as "an autonomous, responsible, quasijuridical person of sound mind, as in a legal subject." In contrast, Latinas do not possess subject status; their behavior is irrational, illogical, chaotic, and, therefore, threatening. The simple dualism inherent in the rendering of a social group as not possessed of subject status works well when constructing enemy images who threaten the life and well-being of those with subject status, be they individuals or nations.
Carole Vance (1991) adds another important dimension to the politics of reproduction. Because of the increased interest in reproduction and sexuality in the 1980s and early 1990s, Carole Vance titled her article "Anthropology Rediscovers Sexuality." Vance also emphasized the politics surrounding the concept of sexuality: "For researchers in sexuality, the task is not only to study changes in the expression of sexual behavior and attitudes, but to examine the relationship of these changes to more deeply-based shifts in how gender and sexuality were organized and interrelated within larger social relations" (ibid.:876). For Latinas, this means their lives as wives and mothers are subject to redefinition by the larger society that views them in comparison to more "modern" U.S. women (Glenn 1994). In particular, Euro-American women's roles are more broadly defined to include education and work outside the home, and their sexuality and reproduction are positively viewed against the "Other" women of the Third World, including Mexican immigrant women and U.S.-born women of Mexican descent, with their "high" reproductive levels. (See Wilson [2000] for an in-depth discussion of Latina reproduction in relation to control of immigration from Mexico.) Complicating what we know about Latina fertility and reproduction is that it has been the subject of social science interest, and construction, since at least the early 1970s (Gutiérrez 1999). For much of this time, the emphasis has been on high fertility levels, especially among Mexican-origin women, with less emphasis on the rapid drop in fertility rates among Mexican and Mexican American women between the 1960s and 1990s (see below). As Amaro (1988:6) observed, "The social science literature has often portrayed Mexican-American women as sacrificed to childbearing…. An assumption behind these evaluations of Mexican-American women is that traditional cultural values and religious traditions promote attitudes favorable to continuous childbearing, opposition to contraception, and opposition to abortion." The few studies that have examined actual behavior among Latinas have found some important differences from stereotypical characterizations. For example, Marchi and Guendelman (1994) found that Latina girls in their study had lower rates of sexual activity than non-Latina girls, which they attributed to Latino cultural norms. They noted that with "increasing acculturation to U.S. norms and values, Latina girls engage in sexual activities at an earlier age and are more likely to have births out of wedlock" (ibid.:210). Amaro (1988) also found most of the Mexican American women in her study favored contraceptive use, most had used one or more contraceptive methods, and they sometimes desired smaller families than they actually had. Similarly, Stroup-Benham and Trevino (1991) found that in 1979, 61 percent of Hispanic women nationwide has used oral contraceptives, almost as high a rate as non-Hispanics (68%).
Finally, research in Mexico also suggests problems with common characterizations of Mexican women's fertility and reproduction. Carole Browner (1986) found that women in a rural Mexican village generally wanted fewer children than the number promoted by government policies. Hirsch (1998:540-541) found that fertility rates have declined dramatically in Mexico-from 7 to 8 children per woman before 1970 to 4.4 children per woman in 1980 to 3.8 children in 1986 to 3.4 in 1990-according to Mexico's Consejo Nacional de Población (National Population Council [CONAPO]). In 2002, Mexico's fertility rate had dropped to 2.9 children born to a woman during her lifetime, compared to 2.1 for U.S. women, according to the Population Reference Bureau (2003). CONAPO (2003) puts the fertility rate lower-at 2.4 children per Mexican woman in 2000. Clearly, Mexico has experienced dramatic decline in fertility rates over the last few decades. Declines in fertility are undoubtedly greater for younger Mexican women than these averages for fertility indicate. Hirsch attributes the drop in Mexican women's fertility to changing beliefs about marriage relationships, delaying having children, spacing births out more than in the past, and increased contraceptive use.
At the very least, these studies suggest that Latina fertility and reproduction are more complex than generally characterized in the social science literature. The section that follows examines recurring themes in how Latina reproduction and fertility are characterized in the discourse found in national magazines.

Latina Fertility and Reproduction in National Magazines
To explore popular characterizations of Latina fertility and reproduction, I draw from a study in which I examined selected issues of 10 national magazines published between 1965 and the end of 1999 (Chavez et al. 2001). 3 Magazines were selected if their cover mentioned, in image or text, immigration or alluded to immigration in some direct way. A total of 76 magazines were selected and their covers and pertinent articles copied for analysis. In the space available here, I would like to suggest the contours of the recurring themes and issues related to Latina fertility and reproduction found in these magazines.

High Fertility and Population Growth
The first theme focuses on population growth. The contribution of Latino immigrants and their children to the nation's population growth was viewed as particularly problematic, given the pressure of environmental and population groups, such as Zero Population Growth, Inc. As Leonard F. Chapman, Jr., commissioner of the Immigration and Naturalization Service (INS), commented in a 1974 interview in U. S. News & World Report (1974:30): "We're very close in this country to a zero population growth through births. As we get closer to that zero growth, immigration will become an even larger percentage of the population increase." Social science constructions of the Latina, particularly Mexican American, fertility "problem," often intersected with the characterizations found in the magazines. For example, Alvirez and Bean (1976:271), citing INS Commissioner Chapman's (1974) estimates of the growing Mexican American population, noted that "The most noticeable feature of the Mexican American family is its size relative to other groups in America. The fertility of Mexican Americans is substantially higher than other groups." At the time, the average size of Mexican American families (4.4) was about one person larger than for all Americans (3.5 persons per family) (Alvirez and Bean 1976:280-281). Alvirez and Bean also observed that Mexican women's fertility rates were subject to change from urbanization and social mobility, which is substantiated by later empirical findings. As Bean, Swicegood, and Berg (2000) note, the mean number of children ever born to Mexican-origin women has decreased dramatically between 1970 and 1998. All Mexican-origin women in the U.S. between 18 and 44 years of age had 1.81 children, well below zero population. Non-Hispanic white women between the same ages, however, only had 1.27 children according to these data, so there still exists a "gap" of 42 percent (ibid.).
The theme of Mexican women's "high" fertility surfaced on U.S. News & World Report's July 4, 1977, cover, which carried the headline: "TIME BOMB IN MEXICO: Why There'll be No End To the Invasion of 'Illegals.'" The accompanying article clarified that the "time bomb" was Mexico's population and its expected growth rate. The article stressed that the fertility of Mexicans and their inability to produce jobs for their population would lead to greater pressure for immigration to the U.S. in the future. Importantly, U.S. News & World Report drew the reader's attention to the external threat posed by the reproductive capacity of Mexican women, a threat that was also internal since Mexican immigrant women's and their U.S.-born children's high fertility levels were implicated in the rapidly growing U.S. Latino population.
The growth of the U.S. Latino population was often paired with the decline in immigrants from Europe and the declining proportion of whites in the U.S. population. For example, Newsweek's January 17, 1983, issue reported that between 1970 and 1980, Latinos grew by 61 percent, largely because of immigration and higher fertility rates, and because since the mid-1960s there were 46.4 percent fewer immigrants from Europe. The politics of fertility and reproduction are not limited to immigrant Latinas, but includes U.S.-born Latinas, whose high fertility is characterized as partly responsible for demographic changes occurring in the nation's racial composition. An example is provided by John Tanton, an ophthalmologist from Michigan, who was once president of Zero Population Growth and the founder of the Federation for American Immigration Reform. He was also an ardent promoter of population control, restricting immigration, and making English the official language of the United States. He wrote a now infamous memorandum in 1988 about Latina fertility and "the Latin onslaught": "Will Latin American immigrants bring with them the tradition of the mordida (bribe), the lack of involvement in public affairs, etc.? Will the present majority peaceably hand over its political power to a group that is simply more fertile?…On the demographic point: Perhaps this is the first instance in which those with their pants up are going to get caught by those with their pants down!" (Conniff 1993:24).
The National Review's June 22, 1992, issue featured a cover illustration of the Statue of Liberty standing with a very serious expression on her face and her arm straight out with palm up in a halting gesture. The text informed us that she is redirecting the flow of immigrants to another country: "Tired? Poor? Huddled? Tempest-Tossed? Try Australia. Rethinking Immigration." In the feature article, "Time to Rethink Immigration?," Peter Brimelow (1992:45) found that Hispanics are particularly troublesome. "Symptomatic of the American Anti-Idea is the emergence of a strange anti-nation inside the U.S.-the so-called Hispanics." Brimelow used Latinos as a bullypulpit from which to launch a diatribe about bilingualism, multiculturalism, multilingual ballots, citizenship for children of illegal immigrants, the abandonment of English as a prerequisite for citizenship, the erosion of citizenship as the sole qualification for voting, welfare and education for illegal immigrants and their children, and congressional and state legislative apportionment based on populations which include illegal immigrants (ibid.). Latino social and biological reproduction were the basis for Latinos being characterized as a "problem" in The National Review.
The alleged high fertility of Latinos was part of an apocalyptic vision of the future in the February 1994 issue of The Atlantic Monthly. In the article titled "The Coming Anarchy," Robert D. Kaplan (1994:75) foresees a "new cartography" in which political borders as fixed and abrupt lines are replaced by "buffer entities." The Latino buffer entity replaces the precise U.S.-Mexico border. This new map is "an ever-mutating representation of chaos" that changes in response to migrations of people, explosions of birth rates, and disease. 177 VOL. 63, NO. 2, SUMMER 2004 The threat of Mexican fertility to American society continues into the 21st century. Samuel P. Huntington (2004:32) raises the alarm in Foreign Policy: In this new era, the single most immediate and most serious challenge to America's traditional identity comes from the immense and continuing immigration from Latin America, especially from Mexico, and the fertility rates of those immigrants compared to black and white American natives.

Reconquest
A second theme characterizes Latino social and biological reproduction as a taking over, or "reconquest," of the United States. Key to this theme is the example provided by the French-speaking Canadians (the Quebec model) who sought separation from English-speaking Canada. The reconquest theme surfaced in the U.S. News & World Report's December 13, 1976, issue, which featured the headline "Crisis Across The Borders: Meaning to U.S." The cover's image was a map of North America with two arrows, both beginning in the U.S., one pointing to Mexico and one pointing to Canada. The problem in Canada was Quebec, where many Frenchspeaking residents were pushing for greater sovereignty and even separation from the English-speaking provinces. The crisis in Mexico was the potential for increased migration to the United States. The "Quebec problem" would come to serve as a metaphor, or civic lesson, for the "Mexican problem" in later years. For example, U.S. News & World Report's March 9, 1981, issue featured an illustrated map of the North American continent, including Mexico. The United States was the focal point of the map and the stars and stripes of the U.S. flag covered it. To the north was Canada, with the image of a Mountie holding the Canadian flag and a French-Canadian holding the Quebec flag in one hand and raising his other hand in a defiant, closed-fisted gesture toward the Mountie. To the south was Mexico. The text said the image is about "OUR TROUBLED NEIGHBORS-Dangers for U.S." The cover's image seemed to suggest that Mexican immigration and the growing Mexican-origin population will pose a problem for the United States much as the Quebecois movement did for Canada.
Reproduction, immigration, and the Quebec threat, or "reconquest," came together in U.S. News & World Report's August 19, 1985, cover. Its headline announced: "The Disappearing Border: Will the Mexican Migration Create a New Nation?" The cover's image rendered the two nations, the U.S. and Mexico, through the strategic use of colors. Central to the image were U and S in large block letters, their color white. The US sat in a field of green and rested atop smaller letters forming the word MEXICO. These letters were in red and sat on a field of yellow. The red of MEXICO bled into the white of the US letters, made possible by the disappearance of the lines (borders) between the letters. Without the borders, a one-way flow moved up (north) in the image. Little people were drawn in stereotypical fashion to suggest Mexicans migrating north. The accompanying article, titled "The Disappearing Border" (Lang and Thornton 1985:30), establishes a "reconquest" theme: Now sounds the march of new conquistadors in the American Southwest. The heirs of Cortés and Coronado are rising again in the land their forebears took from the Indians and lost to the Americans. By might of numbers and strength of culture, Hispanics are changing the politics, economy and language in the U.S. states that border Mexico.
Their movement is, despite its quiet and largely peaceful nature, both an invasion and a revolt. At the vanguard are those born here, whose roots are generations deep, who long endured Anglo dominance and rule and who are ascending within the U.S. system to take power they consider their birthright. Behind them comes an unstoppable mass-their kin from below the border who also claim ancestral homelands in the Southwest, which was the northern half of Mexico until the U.S. took it away in the mid-1800s. Like conquistadors of centuries past, they come in quest of fabled cities of gold. America's riches are pulling people all along the continent's Hispanic horn on a great migration to the place they call El Norte.
The often repeated alarm of a Mexican take over was raised again by Samuel P. Huntington in 2000. Even though it comes after the timeframe of my sample of magazines, it illustrates how persistent this theme is in public discourse: The invasion of over 1 million Mexican civilians is a comparable threat [as 1 million Mexican soldiers] to American societal security, and Americans should react against it with comparable vigor. Mexican immigration looms as a unique and disturbing challenge to our cultural integrity, our national identity, and potentially to our future as a country (Huntington 2000:22).

Overuse of Medical and Other Social Services
A third theme focuses on undocumented immigrants' use of welfare and other social services, displacement of U.S. citizens from jobs, and crime. U.S. News & World Report's April 25, 1977, issue focused on these topics, beginning with the cover headline: "Border Crisis: Illegal Aliens Out of Control?" The invasion metaphor raised the specter of a nation under siege, with its national security at stake: "On one point there seems little argument: The U.S. has lost control of its borders" (Kelly 1977:33). But the specific "out of control" behavior emphasized in the magazine is the use of welfare and related social services, which threaten the economic security of the nation.
Themes often become intertwined, especially those of Latina biological and social reproduction, immigration, and the overuse of social services. Both, U.S. News & World Report (March 7, 1983) and Newsweek (June 25, 1984) published covers that serve as examples. U.S. News & World Report's cover announced "Invasion From Mexico: It Just Keeps Growing," emblazoned over a photograph of men and women being carried across a canal. At the head of the line was a woman sitting on a man's shoulders. Newsweek used a similar cover, a photograph of a man carrying a woman across a shallow body of water. The woman wore a headscarf and a long shawl. The man carried her handbag, which suggests she is traveling somewhere, moving with a purpose and for an extended amount of time. She held a walking cane. The title read: "Closing the Door? The Angry Debate Over Illegal Immigration. Crossing the Rio Grande." Featuring women so prominently on the covers of these two national magazines while warning of an "invasion" sent a clear message about fertility and reproduction. Rather than an invading army, or even the stereotypical male migrant worker, the images suggested a more insidious invasion, one that included the capacity of the invaders to reproduce themselves. The women being carried into U.S. territory carry with them the seeds of future generations. The images signaled not simply a concern over undocumented workers, but a concern with immigrants who stay and reproduce families and, by extension, communities in the United States. These images, and their accompanying articles, alluded to issues of population growth, use of prenatal care, children's health services, education, and other social services.
While not exhaustive, this review of national magazines illustrates the issues in a politics of Latina reproduction in popular discourse. The underlying premise of these themes is that U.S.-born Latinas and Latin American immigrants have extreme, even dangerous, levels of fertility in comparison to an imagined native population. Is this premise accurate? To find out, I examined data collected on Latinas and Anglo women residents of Orange County, California, to which I now turn.

Latina Fertility and Reproductive Behavior
In the early 1990s, I was part of a study designed to examine beliefs about breast and cervical cancer and the use of cancer screening examinations (Chavez et al. 2001;Hubbell et al. 1996aHubbell et al. , 1996b. The data also included information related to fertility and reproduction. While we did find some differences between Latinos and Anglos, they were modest in comparison with the rhetoric surrounding Latina fertility. Moreover, it became important to look at reproduction and fertility not as something fixed and immutable, but as reflecting differences among Latinas, who are often glossed as homogenous and impervious to change. Orange County is the third most populous county in California, with 2,846,289 inhabitants counted in the 2000 U.S. Census (U.S. Census Bureau 2004). It covers an area of 789 square miles, is largely urban, and contains 34 cities and numerous unincorporated communities. Latinos accounted for 30.8 percent of the county's population in 2000. Most Latinos are of Mexican heritage, but Latino immigrants from other nations in Latin America, particularly Central America, also live in the county. Latinos are found in greater concentrations in the northern half of the county, which includes Santa Ana, where about 4 out of 5 residents are Latino. The southern half of the county has been an area of rapid growth in new middle, upper-middle, and exclusive residential communities. Latino immigrants often work in south county communities but find less expensive housing in the many working class communities in the northern part of the county.

Data Collection
Trained bilingual women interviewers from the Field Research Corporation in San Francisco conducted our telephone survey from September 1992 to March 1993. Eligible participants were English-or Spanish-speaking women, 18 years of age or older, who were not institutionalized, and who identified themselves as white (Anglo, Caucasian, non-Hispanic white) or Latino (Hispanic, or more specific ethnic identifiers such as Mexican or Mexican American). We sought a larger subsample of Latino respondents to examine variation within the population. The telephone survey used a cross-sectional sample of random-digit telephone listings to identify eligible subjects. Both listed and unlisted numbers appeared in the listings, avoiding potential bias due to exclusion of households with unlisted numbers (Survey Sampling 1990). Telephone survey findings may not be generalizable to families without telephones. In Orange County, however, approximately 94 percent of Latinos and 99 percent of Anglos have telephones (SDC 1995). Another potential limitation of the study might be that it would not find hard-to-reach members of the population, the homeless, and those engaged in street-corner employment and migrant agricultural labor. This may be more of a bias, however, for male than female Latinas, who are less likely to be homeless or seek day work by standing on street corners (Chavez 1998).
Our survey randomly selected both households and respondents within households-the woman 18 years or older who had the most recent birthday. The cooperation rate was 78.5 percent. 4 Latina respondents could choose to answer the questions in Spanish or English. We pilot tested the questionnaire, tested its content validity, and translated it from English to Spanish to English. The final questionnaire included inquiries about demographic characteristics and questions related to fertility and reproduction. It also included a previously validated five-point "acculturation" scale (Marin et al. 1987) that measured acculturation primarily on the use of Spanish or English (e.g., to read with, speak with, think with, used as child, and speak with friends). We included this language-acculturation measure because it offered a greater range of variation than the dichotomous foreign-born/U.S.-born variable. Moreover many people who lack English proficiency face significant obstacles accessing medical services in the United States, including information on reproductive services (Solis et al. 1990). (See Chavez et al. 1997, Hubbell et al. 1996a, and Hubbell et al. 1997 for a detailed discussion of methods and a summary of general findings.) Since this study was not focused on reproduction and fertility per se, it is limited in the data it provides. For example, there is information on use of birth control pills, but not other methods of contraception. Despite this limitation, the data that are available provide interesting information on Latina reproductive behavior. Table 1 presents the nationality of survey respondents. About a third (33.6%) of the survey's 803 Latina respondents were born in the United States. Most U.S.-born Latinas were of Mexican descent, but many others traced their heritage to different nations. Most of the Latina immigrants (428, or 80%) surveyed were born in Mexico, but the survey also included Latina immigrants from other countries, including 24 from El Salvador. Table 2 provides a summary of respondents' sociodemographic characteristics. The major difference is between Latina immigrants and both U.S.-born Latinas and Anglo women. Latina immigrants, with a mean age of 33, were younger than both U.S.-born Latinas (mean age 37) and Anglo women (mean age 44). Latina immigrants had received fewer years of education (mean 9 years) than U.S.-born (mean 13 years) and Anglo women (mean 15 years). Latina immigrants had been in Orange County, on average, threefourths of the time they had been in the United States. Immigrants were also more likely than both U.S.-born Latinas and Anglo women to be married and have at least one child in the household under 18 years of age. Since Latina immigrants were young, married, and in the early stages of their reproductive cycles, they were more likely than the other women to be homemakers (amas de casa). An overwhelming majority of Latina immigrants earned less than $20,000 a year. U.S.-born Latinas generally lived in households earning above $20,000 a year, with a fourth above $50,000 a year. Almost half of the Anglo women lived in households earning above $50,000 a year.

Interviewee Characteristics
Scores on the language-acculturation index indicate that immigrant Latinas were much less likely to use English than U.S.-born Latinas. Out of the five points possible, one point for each question (for example, speaking English at home is one point), Latina immigrants had a 1.6 mean score (standard deviation .83). U.S.-born Latinas, on the other hand, had a mean score of 4.2 (standard deviation .85), which means that most of them used English for much of their communication needs. This score is used as a measure not only of language use, but also as an indicator of acculturation (Marin et al. 1987).

Fertility and Reproduction
Before examining the number of children women have, there are four other factors related to reproduction that deserve attention: age when sexual relations are initiated, the number of sexual partners, age at first child, and use of birth control pills. Although discourses and ideologies shape the truth for political ends, their truth claims are still subject to examination. The discourse on Latina reproduction would suggest that Latinas and Anglo women would differ significantly in these fertility-related variables. Table 3 presents information on the age when the women sampled in Orange County initiated sexual intercourse. Latina immigrants were somewhat less likely than U.S.-born women, both Latinas and Anglos, to begin engaging in sexual intercourse under age 18. On average, Anglo women (18.1 years) began sexual relations about a year younger than all Latinas surveyed (mean = 19.0), a significant difference (tvalue = -3.71; p = < 0.001). The difference is insignificant when U.S.-born Latinas (mean age = 17.9) are compared to Anglo women (t-value = .63; p = .530), but is significant when Latina immigrants (mean age = 19.5) are compared to Anglo women (t-value = -5.07; p = < 0.001). Latina immigrants were about a year-and-a-half older than Anglos when they initiated sexual intercourse.
Latinas and Anglo women also varied on the number of reported sexual partners (Table 4). Latina immigrants were more likely to report having had two or less sexual partners. Harvey et al. (1997) also found that Mexican immigrants were unlikely to have had two or more sexual partners. Anglo women were more likely than Latinas to report having had five or more sexual partners. Latinas generally (mean 2.5) and Anglo women (mean 6.3) differed significantly in the mean number of reported sexual partners (t-value = 8.78; p = < 0.001). Once again, U.S.-born Latinas (mean 4.3) did not differ significantly from Anglo women (t-value = 1.61; p = 0.11), but Latina immigrants (mean 1.8 sexual partners) did (t-value = 10.36; p = < 0.001).
There were also significant differences in the age when women had their first children (Table 5). Few Anglo women had their first child under age 18, compared to Mexican immigrants and U.S.-born Chicanas (women of Mexican descent). Anglos were more likely than Latinas to have their first child after age 25. The mean age at which Anglo women had their first child (24.3) was significantly older than Latinas generally (21.6) (t-value = 6.78; p = < 0.001), U.S.-born Latinas (21.9) (t-value = 4.11; p = < 0.001), and Latina immigrants (21.4) (t-value = 6.90; p = < 0.001). In contrast to a stereotype    to the discourse on Latina fertility. Although two-thirds of Latinas generally had used birth control pills, they were still significantly less likely to have done so than Anglo women, most of whom had used the pill (X 2 = < 0.0001).
We turn now to the number of children ever born to Latinas and Anglo women in the survey. This variable refers to the number of children a woman had at the moment of the interview, not the total number of children she will have in her lifetime. Table 6 presents the number of children by various age categories to indicate the influence of age and to take into account the different age structures and fertility patterns of Latinas and Anglo women. Examining only women up to 44 years of age-the convention in most studies of fertility-may capture a majority of women during their peak years of fertility, but it leaves out more older women among Anglos than Latinas. About 59 percent of the Anglo women surveyed were between 18 and 44 years of age, compared to 83 percent of the Latinas.
Latinas between 18 and 30 years old had only 1.2 children, and Anglo women in this age group had .7 children-a significant difference. Given that both groups are still early in their reproductive years, this number will increase, but by how much is difficult to predict. For both Latinas and Anglos, the trend is toward fewer children. With each age category, the number of children increases, but women in their 30s and early 40s have fewer children than women 45 and older. The key age category is 18-44 years of age. Both Latinas and Anglo women in these prime childbearing years are below the 2.0 children per woman required for population replacement and well below the 2.1 children needed for population growth.
Age is only one factor influencing fertility. How does a Latina's immigration history and generation in the United States influence the number of children she has? Table 7 presents the number of children born to women in our Orange County study between the ages of 18 and 44 in relation to immigration and generation patterns. Since Mexican-origin women have been the subject of heightened surveillance and the target of much of the discourse on Latina fertility, the table provides information on them separately. The table also provides the national data on children ever born to Mexican-origin women provided by Bean et al. (2000).
There are a couple of distinct comparative advantages to using these data, which were obtained by pooling individual records of women of childbearing age from the June 1986 and June 1988 Current Population Surveys (CPS). First, the information on the number of children ever born is broken down into generation in the United States, from immigrants to third and later generations. Second, the CPS data are only a few years earlier than the data collected in our study in Orange County.
What is striking about Table 7 is the low number of children among almost all the women. Anglo women in Orange County, with 1.22 children per woman, have fewer than the 1.27 children for Anglo women nationally. Latinas generally  also had fewer than 2.0 children per woman. With 1.93 per woman, Mexican-origin women showed a dramatic decrease from the 4.4 children per woman found in the early 1970s. Mexican immigrants who migrated to the U.S. as adults (16 years old or older) had the highest number of children per woman (2.31), but their rate is lower than that found among their counterparts nationally. But the number of children born to Mexican immigrants who migrated as children (under 16) fell to 1.55 per woman, only 22 percent higher than Anglo women (although statistically significant; t-value = -2.14; p = 0.033). Immigrants from countries other than Mexico had a mean of 1.81 children per woman. When all Latin American immigrants, including Mexicans, are examined together, they had 2.08 children per women, a rate which demographically replaces the parents but contributes minimally to growth. The fertility story is even more dramatic when we examine U.S.-born Latinas. Second generation Mexican-origin women (Chicanas) actually had fewer children per woman (1.17) than Anglo women (statistically insignificant; t-value = .30; p = 0.768). This is lower than for second-generation Mexican Americans nationally. Third-generation Mexicanorigin women had 1.42 children per woman, which is more than second-generation Mexican-origin women but still only 16 percent more than Anglo women in Orange County (also insignificant; t-value = -1.08; p = 0.280). Third-generation Mexican-origin women in Orange County had fewer children on average than their counterparts nationally. When all U.S.-born Latinas sampled in Orange County are examined together, they had a mean of 1.28 children per woman, only 5 percent more than Anglo women in Orange County and almost equal to Anglo women nationally.
Finally, I wondered if the significant difference in the mean number of children born to Latinas and Anglo women would hold up in a multivariate analysis, which accounts for the influence of other variables. Table 8 presents the findings from an ordinary least squares (OLS) regression using the number of children ever born as the dependent variable. The independent variables include Latina/Anglo (values = 0, 1), married (0 = not married, 1 = married), education (total years), age (total years), and the language-acculturation variable (five point scale). Income was not included because it correlated closely with language-acculturation. Latinas who scored low on their integration into English-speaking U.S. culture and society also generally had lower incomes. A variable indicating U.S. or foreign birth was also not included since it too correlates highly (.9) with the language-acculturation variable (see Table 2). In addition, separate regression analyses were run for different age categories to account for Latina and Anglo women's differing age structures.  Table 8 summarizes the results of the regression analyses. In Model 1, younger women's (18-30) fertility is examined. Age, education, and marital status predicted the number of children women have in this age category. As a woman ages she is more likely to have children. Married women have more children than unmarried women. The more years of education a woman has, the fewer children she has. Among these relatively young women, language-acculturation did not predict how many children they had. Finally, although Latinas did have more children than Anglos, the difference was not significant when these other variables were taken into consideration.
Model 2 examines fertility among women aged 31-44. Similar to the younger women, age, education, and marital status were significant predictors of how many children the women have had. Also significant here, however, was the language-acculturation variable. The more integrated Latinas are into English language usage and, thus, English-speaking society, the fewer children they had.
Model 3 examines women 18 to 44 years of age, the primary reproductive years. All the same variables are predictors of fertility. Ethnicity, at .06, was not a significant predictor of fertility for these women. However, some might argue that at this level of significance, ethnicity must at least be considered an important predictor of fertility for these women.
Model 4 includes women of all ages in the analysis. Age, education, marital status, and language-acculturation were significant predictors of fertility. Ethnicity, however, was once again not significant for understanding fertility. Roberts (1997b:8) observes that welfare reform and policies to regulate fertility are propelled by powerful stereotypes. "Myths are more than made-up stories. They are also firmly held beliefs that represent and attempt to explain what we perceive to be the truth. They can become more credible than reality, holding fast even in the face of airtight statistics and rational argument to the contrary." The taken-for-granted assumption in the discourse on Latina fertility is that they are a population with "their pants down" and, thus, their reproductive behavior poses serious threats to the nation. As the discourse in popular magazines underscores, Latina reproduction and fertility threaten the nation's demographic future, in terms of size and ethnic-racial composition, provide the basis for a potential takeover or reconquest of U.S territory, and hasten a destabilization of the nation's medical and other social services. The data on Latina reproductive behavior examined here cannot possibly refute the deeply held beliefs upon which such cataclysmic stories are based. However, the evidence presented here does not support the pejorative view of Latina reproduction-related behavior. Latinas do not begin sexual activities at a relatively early age nor do they have relatively more sexual partners than Anglo women. They may have their first child a couple of years younger than Anglo women, but on average they are over 20 years old when they do so. 5 Importantly, most Latinas have used birth control pills at some point in their lives. These findings are not evidence for reproductive behavior that is out of control.

Latina Fertility Reconsidered
Moreover, Latinas are not static when it comes to fertility. They, like other women in the United States, Mexico, and the world in general, have experienced rather dramatic declines in fertility. In terms of the number of children ever born to a woman, Latinas in Orange County compare favorably with Latinas nationally. All Latinas have fewer than 2.0 children per woman. Mexicans who immigrated to the U.S. as adults and second-and third-generation Mexican Americans (U.S.born) had fewer children, on average, than their counterparts nationally. All U.S.-born Latinas had almost the same number of children as Anglo women nationally, a low 1.28 children per woman. These findings suggest that reproductive behavior can vary among Latinas as their life experiences change. Future research could determine in greater ethnographic detail how context and life experiences influence reproductive behavior among Latinas.
Multivariate analysis suggests that several factors influence the number of children women have, factors that are as, if not more, important than being Latina or Anglo. Age, education, and marital status consistently predict whether women have more or less children. In addition, increasing facility with English, perhaps because it increases sources of knowledge about reproduction control, leads to fewer children among Latinas. Ethnicity was an important, but not a statistically significant, variable for understanding fertility differentials in the 18-44 age group. This is important given the theoretical discussion presented earlier on stratified reproduction, which emphasizes that Latinas in general are poorer and have less access to health services than Anglos.
How do these empirical findings and their interpretation "speak to" the politics of Latina reproduction? The discourse surrounding Latina reproduction is actually about more than reproduction. It is also about reinforcing a characterization of Anglos as the legitimate Americans who are being supplanted demographically by less-legitimate Latinos. A recurring theme in this discourse is the image of Latinas and their "comparatively high" fertility as a threat to the Anglo population. This is a powerful image, one that provides fuel for political actions such as California's Proposition 187. As such an image becomes part of "common sense," it makes it difficult to interpret events from different perspectives. For example, the politics of Latina fertility have obscured a rather dramatic story of reproduction over the last 30 years. Latinas and Anglos both have fewer children today than they did three decades ago. This trend toward fewer children is not peculiar to the United States; it is found in most of the world, with the industrialized nations having the lowest birth rates.
In the discourse on Latina fertility, comparisons assume that the extreme decline in birth rates among Anglo women is a positive value against which equally dramatic declines among Latinas inevitably come up short. Latina fertility seems destined to be viewed as a "glass half empty." An unasked question is: at what point do extremely low birth rates become problematic? The implications of falling birth rates was the subject of the Los Angeles Times recent headline, that read: "Nation's Birthrate Drops to Its Lowest Level Since 1909" (Zitner 2003). The implications have to do with family structure, how communities spend money, how the nation finances retirement, and pressure for immigration.
What are the implications of fertility rates well below zero population growth and the increasingly high value placed on having ever fewer children, especially in industrialized societies? At the present time, a pattern of extreme fertility decline in industrialized nations increases pressure for immigration to satisfy labor demands and to slow down national population declines. Latina fertility levels may be more reasonable from a societal point of view than the continually lower fertility rates among Americans in general and white women in particular.
From an anthropological perspective, comparisons are relative and reflect the taken-for-granted values of the person or society doing the comparison. Shifting assumptions that valorize white women's fertility levels no matter how low they drop would alter the way Latina fertility is represented. Rather than Latinas being characterized as having "comparatively high" birth rates, Anglo women may be characterized as having "comparatively low" birth rates. Would it be just as possible to make the following observation: the abnormally low fertility rates of Anglo women are leading to demographic changes and increased pressure for immigration? In Japan, with slightly higher fertility rates than Anglo women in the United States (1.38 per woman), there are rewards for families who produce more than two children (French 2000;Newsweek 2000:23). On the other hand, given the fertility rate of some European countries, the U.S. rate is "comparatively high"; for example, 19 percent higher than Spain's 1.07 mean children per woman (Wools 2000:47). Should we therefore describe U.S. fertility as pathologically high in comparison to that of Spanish women?
This shift would refocus the discourse on Latina reproduction. Rather than singling out Latinas and their "fertility problem" as the cause of negative demographic changes (proportionally fewer Anglos), more attention might be paid to understanding the social, economic, and cultural influences on decreasing fertility among all women. It might also spur a societywide discussion about the relationship between fertility and immigration-and the value of children for the reproduction of a nation's population.

HUMAN ORGANIZATION
Notes 1 With passage of major reforms in the nation's civil rights and immigration laws, 1965 was a watershed year in U.S. history. It also marks the beginning of the most recent period of large-scale immigration. 2 In the late 19th and early 20th centuries, the period of the last large wave of immigration, the politics of reproduction were also central to anti-immigrant discourses, most notably in the eugenics movement (Gould 1981;Marks 2002;Roberts 1997a). 3 The magazines used in the research were American Heritage, Time, Newsweek, U.S. News & World Report, The New Republic, The Nation, The National Review, The Atlantic Monthly, Business Week, and The Progressive. 4 The cooperation rate is defined as the number of completed interviews divided by the sum of the completed interviews and refusals by eligible women [1,225/(1,225+336)]. 5 This is not to minimize the issue of teenage pregnancies. Latinas have relatively more teen births than Anglo and non-Hispanic African Americans, but here, too, there has been a decline. Pregnancy rates for black and white teenagers between 15 and 19 years of age fell 23 and 26 percent respectively from 1990 to 1997. Latina teen pregnancy rates only began falling in 1994, but they fell 11 percent from that time to 1997 (Ventura et al. 2001). These data do not indicate marital status, the father's involvement, or extended family relations for the mother and child, important factors when considering life opportunities.