College Women and Fertility Values 9781400876051

Citation preview

COLLEGE WOMEN AND FERTILITY VALUES

COLLEGE WOMEN AND FERTILITY VALUES

BY

CHARLES F. WESTOFF AND

RAYMOND H. POTVIN

PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY 1967

Copyright © 1967, by Princeton University Press ALL RIGHTS RESERVED

L. C. Card 66-26589

Printed in the United States of America By The Maple Press Company, York, Pennsylvania

FOREWORD Two surveys of the fertility of American couples conducted dur­ ing the 1950's revealed an initially mystifying difference in the apparent influence of education on the fertility of Catholic and non-Catholic women.1 Non-Catholic women who had completed four years of secondary schooling had about the same fertility as those with less than four years, and those who had attended college had fertility that was somewhat lower. But when the fertility of Catholic women in these educational groupings was examined, an unmistakable tendency toward higher fertility among women with more education was evident. The researchers identified the source of this difference in pat­ tern when they prepared separate cross tabulations of educa­ tional attainment and fertility for Catholic women whose school­ ing had been wholly, partly, or not at all in church-operated institutions. It then became clear that amount of education had essentially the same relation to fertility among Catholic women who had attended only secular schools as among non-Catholic women, and that the anomalous positive association involved women whose education was at least partly under Catholic auspices. The positive association was especially strong among women whose education had been wholly in Catholic institutions. These findings left unanswered the question of whether educa­ tion in American Catholic institutions tends in some way to 1 One survey was the first of a series of studies of American fertility begun by the Survey Research Center at the University of Michigan, and the other was the first phase of a continuous project conducted at the Office of Population Research, Princeton University. The findings are reported in: R. Freedman, P. K. Whelpton, and A. A. Campbell, Family Planning, Sterility and Population Growth (New York: McGrawHill, 1959); P. K. Whelpton, A. A. Campbell, and J. E. Patterson, Fertility and Family Planning in the United States (Princeton: Princeton University Press, 1966); C. F. Westoff, R. G. Potter, Jr., P. C. Sagi, and E. G. Mishler, Family Growth in Metropolitan America (Princeton: Princeton University Press, 1961); C. F. Westoff, R. G. Potter, Jr., and P. C. Sagi, The Third Child (Princeton: Princeton University Press, 1963).

FOREWORD

inculcate or strengthen attitudes favoring high fertility, or whether some process of selection causes women with high fer­ tility predilections to seek their higher levels of education in Catholic schools. The principal purpose of the research reported in this book was to provide an answer. The authors have not examined the effects of education on Catholic women alone. In the interest (initially) of providing relevant comparisons or statistical controls, they have analyzed the effects of various forms of college education and various differences in background on attitudes toward fertility of women of each of the principal American religious persuasions. Thus the findings of this book are a contribution to the sociology of American religion and education as well as to demography. The authors of this book collaborated fully at every stage of their research, although the first draft of each chapter was an individual responsibility. Chapters 1-5 and 8 were drafted by Westoff; Chapters 6, 9, and 10 by Potvin; and Chapters 7, 11, and 12 were produced jointly. ANSLEY J. COALE, Director

Office of Population Research Princeton University

PREFACE Since this study involved the participation of students in fortyfive colleges and universities, we have numerous debts of grati­ tude to discharge. In some instances, however, the administrator responsible for permitting us to distribute questionnaires pre­ ferred institutional anonymity. In view of this we feel we should maintain the anonymity of all institutions in our sample and simply express our appreciation to all of the presidents, deans, and professors for their splendid and frequently enthusiastic cooperation. Our introduction to the Catholic institutions in the sample was facilitated greatly by the invaluable help of Dr. Roy J. Deferrari, Director of the Program for Affiliation, Catholic Uni­ versity of America, who contacted the presidents of Catholic women's colleges to secure permission for us to administer our questionnaires. The fact that we received complete cooperation from the Catholic schools is undoubtedly due in no small mea­ sure to his intervention. During the months that we were collecting data, the study was greatly benefited by the managerial talents of Dr. Boris Karashkevych, then a research associate of the Office of Popula­ tion Research, who is currently Assistant Professor of Sociology at Hollins College, Virginia. Dr. Karashkevych was responsible for most of the myriad details and logistics of the questionnaire administration, coding, and punching, and the early stages of data processing. We owe our single greatest debt of appreciation to Dr. Karashkevych. Numerous other persons participated in the data-processing phase. Sally Ann Freedman of Peabody, Massachusetts, worked diligently as a research assistant at the Office of Population Re­ search during the summer of 1964. W. Rudolph Struse, III, also a research assistant at the Office during the summer of 1965, prepared and processed many of the statistical analyses on the IBM 7094 at Princeton. Similar contributions were made by Millicent Taplow of the Office of Survey Research and Sta-

PREFACE

tistical Studies at Princeton University and by two Princeton undergraduates, Johnson M. Hart and Orin Merrill. On a more senior level, we are grateful to Carl Helm of the Department of Psychology and Roald Buhler of the Computer Center for their help in adapting various computer programs to our needs. (This work made use of computer facilities supported in part by National Science Foimdation Grant NSF-GP 579.) On the clerical side we are indebted to Lorna Harvey and Hazel Chafey of the OfHce of Population Research for their manuscript typing. We would also like to acknowledge the co­ operation of Dr. Robert Larson, Chairman of the Department of Sociology, Seattle University, for making available clerical help during the summer of 1965. Last but hardly least, we would like to acknowledge the sup­ port of the Ford Foundation and the Council on Human Rela­ tions of Princeton University for a supplementary grant that facilitated the expansion of our sample. The study was conducted at the Office of Population Research, which is supported by sustaining grants from the Rockefeller and Ford Foundations. CHARLES F. WESTOFF RAYMOND H. POTVIN

CONTENTS Foreword

ν

Preface

vii

List of Tables

a PART I · INTRODUCTION

1. Background, Scope, and Method

3

PART II · THE EFFECTS OF HIGHER EDUCATION

2. Types of Colleges and Family-Size Preferences

21

3. Higher Education and Family-Size Preferences

29

4. Higher Education and Family-Planning Intentions

50

5. Higher Education and Fertility Values within Selected Controls

60

6. Higher Education and Some Beliefs about Marriage, Family, and Career

87

7. Conclusions and Discussion of the Effects of Higher Education 120 PART III · THE INFLUENCE OF SOCIAL FACTORS AND BELIEFS

8. Social and Personal Characteristics and Fertility Values

129

9. Beliefs, Family-Size Preferences, and Family-Planning Intentions 163 10. Intergroup Distances and Belief Structures

191

11. Multivariate Analyses of Correlates of Family-Size Preferences 204 12. Conclusions and Discussion of Fertility Values Index

Factors Affecting 220 235

ix

TABLES 1. Number of Institutions by Type in the Sample 2. Distribution of Non-Catholic and Catholic Schools by the Proportion of Questionnaires Returned by Firstand Fourth-Year Students 3. Number of Children Desired by Women, by Year and College or University 4. Distribution of Schools by Average Number of Children Desired by First- and Fourth-Year Women 5. The Average Number of Children Desired by Women in Small and Large Nonsectarian Coed Institutions and Catholic Women's Colleges 6. Number of First- and Fourth-Year Women in the Sample by Religion, Type of College, and Type of Elementary and High School for Catholics 7. Number of Children Desired by Women with No Religious Preference, Protestant Women, Jewish Women, and Mormon Women by Year and Type of Institution 8. Number of Children Desired by Catholic Women by Year and Type of Institution, Subdivided by Type of High School and Elementary School 9. Number of Children Desired by Catholic Women in Catholic Colleges, by Type of Elementary and High School and Whether the College Attended is Coeduca­ tional or a Women's College 10. CoeflBcients of Correlation for Catholic Freshmen Be­ tween Selected Variables and Choice of a Nonsectarian or Catholic College 11. Number of Children Desired by Women, by FertilityPlanning Intentions and Religion 12. Intentions to Plan Spacing and Number of Children of Women with No Religious Preference, Protestant Women, Jewish Women, and Mormon Women, by Year and Type of Institution 13. Intentions of Catholic Women to Plan Spacing and

11

12

22 26

26

32

36

40

46

47 52

53

TABLES

14.

15.

16.

17.

18.

19.

20.

21.

22. 23.

24.

25.

Number of Children, by Year and Type of Institution, Subdivided by Type of High School and Elementary School Number of Children Desired, by Religion, Type of Institution, and Year, by Women Whose Parents Were High School or College Graduates Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by Education of Parents Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Occupational Class of Father Proportion Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by Occupational Class of Father Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Nationality Back­ ground of Mother Number of Children desired by Women, by Religion, Type of Institution, and Year, by Nationality Back­ ground of Father Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by Nationality Background of Mother Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by Nationality Background of Father Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Religion of Parents Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by Religion of Parents Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Woman's Percep­ tion of the Importance of Religion to Her Parents Proportion of Women Intending to Plan Both Number

54

62

63

65

66

67

68

69

70 73

74

75

TABLES

26.

27.

28.

29.

30.

31.

32.

33.

and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by the Woman's Perception of the Importance of Religion to Her Parents Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Religion of the Woman's Best Girl Friend and the Religion of Most of Her Girl Friends Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by the Religion of the Woman's Best Girl Friend and the Religion of Most of Her Girl Friends Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Woman's Per­ ception of the Importance of Religion to Her Best Girl Friend and Closest College Friend Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by the Woman's Perception of the Importance of Religion to Her Best Girl Friend and Closest College Friend Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Number of Religious Practices the Woman Observes and the Im­ portance She Attaches to Religion Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institu­ tion, and Year, by the Number of Religious Practices the Woman Observes and the Importance She Attaches to Religion Percent of Women Agreeing or Disagreeing That a Married Couple is Obligated to Want a Large Family, by Religion, Year, and Type of Institution Percent of Women Agreeing with Selected Statements on Marriage and Children, by Religion, Year, and Type of Institution

76

78

79

80

81

84

85

90

93

TABLES

34. Percent of Women Agreeing with Selected Statements on the Mortality of Family-Size and Birth Limitation, by Religion, Year, and Type of Institution 35. Percent of Women Agreeing with Selected Statements on Religion, Family Planning, and Birth Control by Religion, Year, and Type of Institution 36. Percent of Women Agreeing with Selected Statements about the Catholic Image, by Religion, Year, and Type of Institution 37. Percent of Women Intending to Work or Have a Career, by Religion, Year, and Type of Institution 38. Percent of Women Agreeing with Selected Statements on Family and Career, by Religion, Year, and Type of Institution 39. Percent of Catholic Women Educated in Nonsectarian High Schools and Enrolled in Catholic Colleges Agree­ ing with Statements on Family "Normative" Beliefs, by Year and Type of College 40. Number of Children Desired and the Proportion of Women Intending to Plan Spacing and Number of Children, by Religion 41. Number of Children Desired by Women, by Her Re­ ligion and the Religion of Her Parents 42. Number of Children Desired by Women, by Religion and Religion of Most Girl Friends and Best Girl Friend 43. Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and Religion of Parents 44. Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and Religion of Most Girl Friends and Best Girl Friend 45. Distributions of Family Sizes Women Consider "Large" and "Small," by Religion

99

103

104 111

113

116

131 131

133

133

134 136

TABLES

46. Proportion of Women Desiring Large and Small Families, by Religion 47. Number of Children Desired by Women, by Religion and Their Perception of the Size of a "Large" and a "Small" Family 48. Proportion of Women Intending to Plan Both Number and Spacing of Children, by their Perception of the Size of a "Large" and "Small" Family 49. Number of Siblings of Women by Religion 50. Number of Children Desired by Women, by Religion and Number of Siblings 51. Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and Number of Siblings 52. Number of Children Desired by Catholic Women, by Number of Siblings and Scale of Religious Practices 53. Number of Children Desired by Women, by Religion and Occupation of Father 54. Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and by Occupa­ tion of Father 55. Number of Children Desired by Women, by Religion and by Education of Parents 56. Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and by Education of Parents 57. Number of Children Desired by Women, by Religion and by Main Nationality Background of Parents 58. Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and by Main Nationality Background of Parents 59. Number of Children Desired by Catholic Women, by Main Nationality Background of Parents and Scale of Religious Practices 60. Proportion of Catholic Women Intending to Plan

137

138

139 140 141

142 143 144

145 146

147 148

149

150

TABLES

61. 62.

63. 64.

65. 66.

67.

68.

69. 70.

71. 72.

73.

Number and Spacing of Children, by Main Nationality Background of Parents and Scale of Religious Practices Number of Children Desired by Women, by Religion and Selected Measures of Religious Behavior Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and Selected Measures of Religious Behavior Number of Children Desired by Women, by Religion and Scale of Religious Practices Observed Intentions of Women to Plan Both Spacing and Number of Children, by Religion and the Scale of Re­ ligious Practices Observed Number of Children Desired by Women, by Religion and Importance of Religion to Parents Number of Children Desired by Women, by Religion and by the Importance Attached to Religion by Their Best Girl Friend and Their Closest College Friend Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and Importance of Religion to Parents Proportion of Women Intending to Plan Both Spacing and Number of Children, by Religion and by the Im­ portance Attached to Religion by Their Best Girl Friend and Their Closest College Friend Number of Children Desired by Women, by Religion and Religiousness of Significant Others Intentions of Women to Plan Both Spacing and Num­ ber of Children, by Religion and Religiousness of Significant Others Number of Children Desired by Women, by the Im­ portance They Attach to Religion Proportion of Women Intending to Plan Both Spacing and Number of Children, by the Importance They Attach to Religion Mean Number of Children Desired by Women and

151 153

154 155

156 157

157

158

158 159

159 160

160

TABLES

74.

75.

76.

77.

78.

79.

Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that a Married Couple is Obligated to Want a Large Family Mean Number of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that Having Children is the Most Important Function of Marriage Mean Number of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Dis­ agreement with the Statement that if People Do Not Intend to Have Children They Ought Not Get Married Mean Number of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that the Ideal Family Has at Least Five or Six Children Mean Number of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Dis­ agreement with the Statement that it is Morally Wrong to Have More Children than One Can Afford Mean Number of Children Desired by Women and Percent Planning Number and Spacing, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that a Family Should Have as Many Children as Possible and God Will Provide for Them Mean Number of Children Desired by Women and Percent Planning Number and Spacing, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that Voluntary Restriction of Family Size (by Morally Acceptable Means) in Order to Improve One's Standard of Living Is Morally Wrong

164

166

167

168

170

171

172

TABLES

80. Mean Number of Children Desired by Women and Percent Planning Number and Spacing, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that It Is a Perfectly Legitimate Function of Religion to Determine the Proper Means of Birth Limitation 81. Mean Number of Children Desired by Women and Percent Planning Number and Spacing, by Religion and Type of Institution, in Agreement or Disagreement with the Statement that the Repeated Use of Methods of Birth Control is Physically Injurious to Women 82. Mean Number of Children Desired by Catholic Women and Percent Planning Number and Spacing, by Type of Institution, in Agreement or Disagreement with the Statement that the Catholic Church Regards Large Families as Morally More Desirable than Small Families 83. Mean Number of Children Desired by Catholic Women and Percent Planning Number and Spacing, by Type of Institution, in Agreement or Disagreement with the Statement that Catholics Are Morally Bound to Support Laws that Outlaw the Sale of Birth Control Devices 84. Mean Number of Children Desired by Catholic Women and Percent Planning Number and Spacing, by Type of Institution, in Agreement or Disagreement with the Statement that the Catholic Interpretation of Family Planning Is Presently Undergoing Change 85. Niunber of Children Desired by Women and Percent Planning Spacing and Number, by Religion and Type of Institution, on the Basis of Whether They Do or Do Not Intend a Career 86. Mean Number of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or

174

176

177

178

179

181

TABLES

Disagreement with the Statement that It Is Perfectly All Right for a Woman to Have a Career and a Family at the Same Time 87. Mean Number of Children Desired by Women and Percent Intending to Plan Spacing and Number Ac­ cording to Intent to Pursue a Career and Agreement or Disagreement that a Career and Family Are All Right 88. Mean Number of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Dis­ agreement with the Statement that a Woman Should Derive Her Main Satisfactions in Life from a Family Rather than from a Career or Intellectual or Com­ munity Activities 89. Mean Niunber of Children Desired by Women and Percent Planning Number and Spacing of Children, by Religion and Type of Institution, in Agreement or Dis­ agreement with the Statement that a Mother with Young Children Should Not Work If She Does Not Have To 90. Group Means on Items, by Religion and Type of Institution 91. Distance Measures Relating the Ten Groups 92. Religious Groups: Rotated Factor Matrices 93. Multivariate Analysis with Number of Children De­ sired: (a) Zero-Order Correlations; (b) nth Order Partial Correlations 94. Values of Multiple Correlations with Number of Children Desired for: (1) All Original Variables; (2) Excluding Images of Small and Large Families 95. Multivariate Analysis of Beliefs and Other Selected Factors with Number of Children Desired: (a) ZeroOrder Correlations; (b) Kth Order Partial Correlations 96. Multivariate Analysis for Total Sample with Number of Children Desired: (a) Zero-Order Correlations; (b)

183

186

187

188 192 196 199

209

210

214

TABLES

8th Order Partial Correlations; (c) 6th Order Partial Correlations; (d) 5th Order Partial Correlations; (e) 3rd Order Partial Correlations 216 97. Rotated Factor Structure for the Total Sample 218

PART I INTRODUCTION

CHAPTER 1

Background, Scope, and Method One of the more interesting aspects of the unanticipated rise in the fertility of American women during the 1950's is that it evidently reflected a genuine shift upward in the number of children young couples considered desirable and that it coin­ cided with types of social changes heretofore regarded as produc­ ing lower fertility—rising levels of living and increasing propor­ tions of young persons being exposed to higher education. While social theorists have reflected occasionally on the complexity of the relations between income and fertility and the possibly changing significance of income for fertility at advanced stages of economic development, less ambiguity has characterized the assumed connections between higher education and fertility. In demographic terms, higher education is connected with lower fertility through the mechanisms of deferred marriage and greater use of the means of fertility control. In more socialpsychological terms, the assumption is that persons (especially women) receiving higher education develop interests and values that compete with the attraction of home, family, and children. In addition, higher education in nonsectarian institutions is pre­ sumed to diminish religious values and, to the extent that religion and fertility are associated, a further significance for lower fertil­ ity is implied. Concerns such as these form the main focus of the research reported in the first part of this monograph. More specifically, our interest in the influence of higher education on family-size values stems from a current study of social and psychological factors affecting fertility in the United States. This study features a longitudinal design, with a sample of mothers interviewed originally in 1957 some six months after the birth of their second child. They are currently being interviewed for a third and final time. A second series of interviews was conducted in 1960, in which the factors associated with the advent of a third child were studied. One of the prominent findings of this study was

I. INTRODUCTION

that religion is the most important social determinant of familysize preferences as well as of actual fertility. Catholic couples expressed desires for the largest families and seem well on the way toward achieving this goal. Jewish families want the fewest children and appear to have exercised the greatest degree of control in family planning. Protestant couples fall in between these two extremes, being closer to Jewish couples in the number of children desired but more nearly similar to Catholics in their control of fertility.1 The influence of religion on fertility seems to operate primarily through its effect on the number of children desired and only secondarily through the mechanisms of fertility control. The comparatively ineffective fertility planning of Catholic couples should be viewed more as a consequence of a strong positive orientation toward large families, with a resulting casual attitude toward carefully controlled child-spacing in the earlier years of marriage, rather than as a simple result of their concentration among users of the less effective rhythm method of contraception.2 These particular findings grew out of an examination of the connections between amount of education and success in con­ trolling fertility. Among Protestants and Jews there is some ten­ dency for successful fertility planning to be associated with in­ creasing education, while among Catholics, paradoxically, the opposite pattern prevails—the more educated Catholic women seem to be least successful in controlling fertility. This paradox was resolved by considering the kind of education to which a Catholic woman had been exposed, and the explanation seemed to lie in Catholic education. In other words, if the com*C. F. Westoff, R. G. Potter, Jr., and P. G. Sagi, The Third Child (Princeton: Princeton University Press, 1963), p. 89. 1This does not imply that there would be no change in popular definitions of the ideal size of family if other methods such as the oral pill were officially approved. It also does not imply that the rhythm method is just as intrinsically effective as any other method, but it is clear that its effectiveness increases sharply when it is used following the achievement of desired family size, compared with its effectiveness in the earlier birth intervals in which spacing is the main concern.

4

BACKGROUND, SCOPE, AND METHOD

parison is confined to Catholic women educated in nonsectarian schools, the relation of amount of education to fertility planning is in the same direction as in the Protestant and Jewish samples. On the other hand, among Catholic women educated in Catholic schools and especially in Catholic colleges, the association of education with fertility-planning success is sharply negative.8 A similar set of relationships was uncovered in connection with the number of children desired. Among Catholics with no educa­ tion in Catholic schools, there is no correlation at all between educational attainment and family-size preferences, while among women educated in Catholic schools and colleges a strong, posi­ tive correlation emerges.4 As noted above, the influence presumably exerted by Catholic education appeared to occur primarily at the college level. At the high school level, there seemed to be some association with fertility behavior and attitudes only among women, and even here the association appeared weaker than at the college level. At the time we advanced the following speculations about this question: Why does fertility respond to Catholic school education pri­ marily at the college level? ... A number of possible expla­ nations suggest themselves. Students in Catholic colleges re­ ceive instruction in the Catholic philosophy of marriage and the family. It may well be that Catholics who attend non-sectarian institutions are exposed to values which are more anti­ thetical to those espoused in Catholic institutions at the higher level of learning than at lower levels of education. Also the four more years of education in the Catholic system can be viewed as simply that many additional years of exposure to the value system in the same sense that graduates of Catholic high schools have had more exposure than Catholics who do not reach that level of education. And then there is the 'G. F. Westoff, R. G. Potter, Jr., P. C. Sagi, and E. G. Mishler, Family Growth in Metropolitan America (Princeton: Princeton Univer­ sity Press, 1961), p. 218. ' Westoff, Potter, and Sagi, The Third Child, p. 117.

I. INTRODUCTION

possibly greater selectivity of the more religious individuals who continue their education in the Catholic school system. And finally, as we observed in our first analyses those attending Catholic schools are weighted with persons of Irish background which probably implies a stricter interpretation of Catholic doctrine than other ethnic backgrounds. It is possi­ ble that this type of Catholic may be found disproportionately at the higher educational attainment levels. This possibility was examined and confirmed.® In part, our current research can be viewed as an attempt to determine which of these speculations is in fact the explana­ tion. We have included a sample of nonsectarian colleges, how­ ever, not only to examine the selectivity hypothesis, but also to assess more directly the implications of higher education for the family-size values of Protestants and Jews as well. The question of selectivity requires further elaboration since it is one of the central concerns of the current study if not, in some ways, the primary reason for conducting the study. Although it applies theoretically to both Protestants (there are Protestant denominational colleges as well) and Catholics, our discussion will be limited to Catholics since our earlier work related to Catholic schools. The main question involved in the selectivity problem is to determine how much of the relationship between education in Catholic schools and fertility reflects the influence of the formal educational system and the continuous reinforcement of Catholic values resulting from association with persons likely to have similar attitudes and values, and how much reflects the initial selection of religiously affiliated colleges by young persons from more religious backgrounds. In order to unravel this "input" and "output" effect, we had included in the second round of interviews with our panel of married couples some questions aimed at estimating the religious­ ness of the home environment in which the respondent spent her childhood. With this crude index and the extent of education in Catholic schools and colleges, we concluded that although both factors affected fertility independently, the influence of " Westoff, Potter, and Sagi, The Third Child, p. 99.

6

BACKGROUND, SCOPE, AND METHOD

the educational experience seemed to be the stronger.· However, we were not really satisfied with the quality of the evidence. Married adult women with two and three children were being asked to reconstruct the religiousness of their parents' home several decades in the past. Not only is unreliability introduced by the simple memory factor but one's present adult attitudes can distort memory of childhood experiences in unpredictable ways. Furthermore, we are concerned with the nature of the causal inference about the effect of a particular type of college education on subsequent behavior and attitudes when many other experiences associated with a college education could have intervened between graduation and marriage and childbearing. It would be much more desirable to examine its "effect" at the time, that is, at the beginning and end of the college experi­ ence. This is exactly what the present study attempts to do. There is the possibility of a delayed or "sleeper" effect, of course. Higher education may affect fertility behavior in later years without its effects being apparent in the earlier years. There is no obvious reason for making this assumption, but, nevertheless, our later analyses of the effects of higher education on the possibly intervening variables of attitudes toward mar­ riage, family, and career are concerned with this question. In its most narrow focus, therefore, the study is an attempt to determine whether the observed high fertility of women grad­ uates from Catholic colleges and universities is related to the educational experience or whether it reflects the selective admis­ sion into these institutions of women already oriented favorably toward large families. In the most general sense, the study de­ scribed in the first part of this volume represents an inquiry into the effects on family-size values of different types of higher education of women of all religions. CHARACTERISTICS OF WOMEN ASSOCIATED WITH FERTILITY VALUES

The second part of this volume is devoted to an examination of the social characteristics and beliefs of women that may have 'Westoff, Potter, and Sagi, The Third Child, pp. 100-102.

7

I. INTRODUCTION

some bearing on their family-size preferences and attitudes to­ ward family planning. This analysis is independent of type of college or whether the individual is in her first or fourth year and is addressed mainly to the question of whether the fertility values of women of different religions are related to such variables as their socioeconomic origins, religiousness, size of family in which they were raised, nationality, and similar background characteristics. The general theory underlying some of these analyses derives from propositions and assumptions about the process by which fertility values are learned during early adolescence (see Chapter 7). In addition, some analyses are guided simply by traditional interests in the demographic study of differential fertility. THE SAMPLE OF INSTITUTIONS

The population we desired to sample was defined generally as freshmen and senior women in four-year institutions of higher education throughout the continental United States. In view of the considerable number of studies involving non-random samples of colleges and the resultant difficulties of generalization and interpretation, we felt that it was especially desirable to draw a probability sample of institutions. Partly in order to increase the homogeneity of the population and partly for theoretical reasons, various types of institutions were excluded, such as junior colleges, theological and religious schools, professional and art schools, colleges with enrollments under 300, and a number of other categories.7 The purified group that survived these exclusions constitutes a total of 683 schools that can best be described as four-year universities, liberal arts colleges, and teachers' colleges with full-time undergraduate female enrollments of at least 300 students. 'These other categories include institutions for men only, technical and semiprofessional schools, and Negro colleges. Actually, a Negro Catholic college was included in the Catholic sample, and we subse­ quently included a Negro nonsectarian school for comparative purposes, but since these are very small samples, we have excluded them from the present report.

BACKGROUND, SCOPE, AND METHOD

Since our theoretical interests are heavily concentrated on the connection of Catholic higher education with fertility values, we were especially concerned with securing an adequate repre­ sentation of such institutions in our sample. This consideration led to the decision to stratify disproportionately the sample of institutions by type of control. Of these 683 schools the propor­ tion of institutions of higher education under Roman Catholic control that admit women is 20 percent (the proportion of women in Catholic schools among all women enrolled in colleges and universities is about 13 percent). Since we wished to com­ pare the "effects" on fertility values of Catholic and nonsectarian institutions, we decided to draw an equal number of schools from each type. The design is complicated further by the pres­ ence of schools under the control of Protestant and other nonCatholic denominations that account for 26 percent of all nonCatholic institutions (and 13 percent of all women enrolled in non-Catholic institutions). We accommodated these schools in the sample in proportion to their enrollments. The remaining forms of stratification imposed on the sample were intended simply to insure the correct representation of certain characteristics. The major attribute is whether the insti­ tution admits only women or is coeducational. Although we had no strong conviction that this would be particularly relevant to fertility values, there is a very sharp difference in the propor­ tion of women enrolled in coeducational institutions among Catholic schools compared with non-Catholic institutions. Among Catholic schools, 72 percent of the women are enrolled in women's institutions and 28 percent in coeducational schools. Among non-Catholic schools, on the other hand, fully 92 percent of all women are in coeducational institutions. In itself this contrast is quite suggestive of possible differences in the two types of educational experience. The third stratification applied was the region of the country in which the institution is located. This was imposed only to guarantee the appropriate representation of schools in all areas of the country.

I. INTRODUCTION

These three criteria generated 24 strata—three types of control (nonsectarian, Catholic, other denominational), women's and coeducational institutions, and four regions of the country (northeast, north central, south, and west), from which we were to draw at random our sample of institutions. The last consideration prior to taking a random selection from each of these strata is the fact that these schools vary widely in the size of (female) enrollments. If we were to draw the sample of institutions proportionately from the 24 strata, each institution would be weighted equally with no distinction made between institutions with 300 women and those with ten and twenty times that number. Since we wish ultimately to generalize about the effects of college experience on the fertility values of American women, it is important that we do not run the risk of loading the sample, for example, with very small colleges which might reflect only a very small proportion of the educa­ tional experience of American women. To ensure drawing a sample of institutions proportionate to the prevailing distribution of schools of different sizes, we further divided the 24 strata into three size classes: schools with fewer than 1,000 women (but at least 300), schools with between 1,000 and 3,000 women, and those with more than 3,000 women. Actually this size classi­ fication could be imposed only on the nonsectarian coeducational institutions. On other strata, such as the Catholic coeducational institutions and nonsectarian women's colleges, we used only one division at the line of 1,000 women students. Other cutting points had to be used in the smaller groups.8 The final step was to determine the number of colleges that could be included in the sample. This decision was influenced by budgetary limitations and the number of students desired for adequate bases for statistical analysis. The net result of this consideration was a decision to aim for a total probability sample * In Protestant coeducational schools, a cutting point of 700 was used, and Catholic women's colleges were divided into three classes: 1,000 and over, 600-999, and 300-599. The small size of Protestant women's colleges precluded representing them in the sample.

10

BACKGROUND, SCOPE, AND METHOD TABLE 1 Number of Institutions by Type in the Sample Nonsectarian, coeducational Nonsectarian, women only Protestant, coeducational Catholic, coeducational Catholic, women only Total

15 3 3 8 16 45

of 48 institutions divided about equally between Catholic and non-Catholic schools. The final sample is described in Table 1. This distribution reflects not only the randomness of the selec­ tion procedures, but also the fact that several colleges declined to cooperate. On the whole, however, the level of cooperation was excellent, with only 3 of the 48 schools originally included declining cooperation. All 24 Catholic schools cooperated although, as noted below, there is some variation in response rates among the different schools. THE SAMPLE OF STUDENTS

Ideally, we were interested in including all freshmen and senior women in the institutions in our sample, an ideal that was compromised in two ways. Some universities, especially some of the state universities, have such large enrollments that the administration of questionnaires to all eligible students would have been a prohibitive task, especially since we were completely dependent upon the goodwill and cooperation of local faculty or administration. We were able to maintain this ideal in all 24 Catholic institutions and in 14 of the 21 non-Catholic schools. In the 7 remaining institutions with very large enrollments we employed various sampling ratios. The second compromise was unavoidable—some students are not sufficiently motivated to complete questionnaires.9 This type * The questionnaire is reproduced in the Appendix. The instrument was developed on the basis of three extensive pretests in the Washington, D.C., and New York City areas.

I. INTRODUCTION

of nonresponse is, of course, a source of bias. For many reasons, including the type of student body, the method of administering the questionnaires, and the energy and interest invested by the persons responsible, the response rates varied considerably, espe­ cially among the nonsectarian schools. The method of adminis­ tering the questionnaires, which were quite short (they required about 20 minutes to complete), varied mainly with the size of the institution. In some institutions they were administered in classrooms or an auditorium either by us or by an interested faculty member or person in the administration of the institution itself. In others, primarily the large schools, they were mailed to the students to be returned to us. Where possible we tried to secure the names of the first-year women in order to make possible a follow-up study three years later. We did not exert any pressure, however, since our main concern was to maximize the response rate (and the reliability of response) in the current study. In all instances, confidentiality was stressed. The Catholic schools produced higher response rates (with some conspicuous exceptions), mainly because they are smaller schools and also probably because they are more centrally orga­ nized (Table 2). Both the size and relation of the student to ad­ ministrative authority at the typical state university would make TABLE 2 Distribution of Non-Catholic and Catholic Schools by the Proportion of Questionnaires Returned by First- and Fourth-Year Students Response rate

90 percent and higher 80-89 70-79 60-69 50-59 40-49 30-39 20-29 Number of schools

Non-Catholic Schools

Catholic Schools

1

4

1

4

4 3 4 4 3

2 4 2 4 7

5 9 4

2

2

1 21

— 21

4 8 5 1 1 2 3 — 24

1 1 1 3 24

BACKGROUND, SCOPE, AND METHOD

it difficult indeed to imagine assembling all freshmen and senior women in the auditorium at noontime to fill out our questionnaires. The response rates for each institution are based upon enroll­ ment figures for freshmen and senior women obtained from each college. In many cases these were only approximate. Some of the returns were defined as ineligible, however, if they were from those who were nonwhite, foreign born, born prior to 1939, or seniors who had not been freshmen at the same institution. The color, nativity, and age restrictions were imposed simply to promote further homogeneity. Transfer students were elimi­ nated because the logic of our analysis requires the assurance that the fourth-year students had been exposed to all four years at the same institution. The response rates reflect simply the total number of returned questionnaires divided by the enrollment. As such, the nonrespondents include persons absent from school, groups simply missed at the time the questionnaires were administered, and refusals to cooperate. A more refined measure would be desirable if we had some estimate of the proportion of eligibles in the "population," but such refinement is unobtainable. The overall response rate for all institutions and classes com­ bined was 72 percent. After eliminating ineligible returns, we had about 15,000 questionnaires returned. SOME METHODOLOGICAL CONSIDERATIONS

The main analytical question prompting this research is whether higher education influences the number of children that women would like to have. This question is an expansion of our initial interest in whether education in a Catholic college or university increases the number of children desired or whether the fact10 of higher fertility and family-size desires among Catho­ lic women educated in Catholic colleges could be attributed to initial selection of women who for various reasons might be more disposed toward large-family preferences than Catholic women choosing non-Catholic schools. Our present inquiry en10 Westoff,

Potter, and Sagi, The Third Child, pp. 94—102.

13

I. INTRODUCTION

compasses not only this comparison but also the question of whether the family-size orientations of Protestant and Jewish as well as Mormon women are influenced by higher education. For Protestant women we have included in our sample three colleges under Baptist, Lutheran, and Nazarene church control respectively, as well as private and public nonsectarian schools. Evaluation of the influence of a particular experience upon values calls ideally for a "before and after" design in which the values of the same individuals are assessed at two points in time—before and after the experience. The classical experi­ mental design also requires an adequate control group on which the same measurements are taken. The present research com­ promises this model for various practical reasons, but nevertheless permits reasonable inferences to be drawn from the analysis. Instead of following the same individuals from the beginning to the end of their college experience,11 we have compared the attitudes of first- and fourth-year students currently enrolled. The logic of such a design (well known in demography as the "synthetic cohort" technique and used frequently for inferring time trends from cross-sectional age data) is that the difference in attitudes or preferences observed between first- and fourthyear students can be attributed to the intervening experience of the years at college. In other words, the freshmen of today are considered to be the seniors of today three years later.12 The methodological weaknesses of this design13 are numerous, 11 Actually we have made a considerable effort to secure the names of the first-year students in order to permit following up the same individuals three years later if it seems desirable. 12 For various practical reasons we were not able to contact the fresh­ men students early in the academic year (1963-64) so that we do not have an unbiased picture of their fertility values prior to any ex­ posure to college. Our assumption is that the values are not likely to change radically in the few months prior to our measurement. Ideally, of course, their attitudes should have been assessed before any exposure to college. a For a review of many of the shortcomings of such a design, see A. H. Barton, Studying the Effects of College Education (New Haven: Hazen Foundation, 1959).

BACKGROUND, SCOPE, AND METHOD

although some of the serious shortcomings can be modified in the analysis. One problem, which could be shared by the longi­ tudinal design as well, is that seniors and freshmen are different not only in the amount of exposure to higher education but also in age. This means that any difference observed due to increasing age and maturity could be attributed erroneously to education. Essentially, this is a problem of providing an adequate control group, for example, random samples of 18- and 22-yearold unmarried women not in college. This was not done for practical reasons, and the only insurance we have against this danger is that our hypotheses call for different patterns of "change" in the two classes in different kinds of institutions. In other words, our protection is that if we observe a "decrease" in the number of children desired by Catholic women in nonCatholic institutions, for example, and an "increase" among Catholic women in Catholic institutions, we can rule out the age difference since it is the same generally in all institutions. A more difficult problem, and one that is more intrinsic in the synthetic cohort approach, is the susceptibility of the design to changes in society at large. This is a particularly thorny prob­ lem in the present context and one that cannot be dismissed lightly or resolved completely by any special handling of the data. During the last few years an increasing amount of publicity has been devoted to population questions and to the position of the Catholic church on the whole issue of family limitation. Very few days now elapse without some mention of this problem in the press and one must assume that discussions of the possi­ bility of some reinterpretation of the traditional position of the Catholic church have penetrated college campuses throughout the country. We have tried to collect some data in our question­ naire that would provide an idea of just how much of this activity was known to students, but there is no fully adequate way of estimating how much, if at all, Catholic women's atti­ tudes have been influenced by this discussion. It is our impres­ sion, based on some admittedly unsystematic group interviews with Catholic college women on several campuses, that although

I. INTRODUCTION

they are aware of the currency of the problem in theological circles, they are doubtful that any radical change will be forth­ coming. It certainly does not seem to be the case that any basic change in definitions of desired family size has occurred. And one could always reason that any changes occurring might affect freshmen and seniors equally, thereby maintaining the difference but still affecting Catholic and non-Catholic comparisons. Simi­ larly, the values of non-Catholic women may be affected by public discussion of the detrimental consequences of rapid popu­ lation growth and themes of responsible parenthood, but this is admittedly a rather tenuous hypothesis. A third difficulty involved in the assumption that the seniors represent the freshmen three years later is the fact that in many schools there is a considerable attrition over the period. Students will drop out of school for academic reasons, marriage, or nu­ merous other personal reasons, which means that the senior class may be quite a biased residue of the freshmen class. Whether the types of bias involved are relevant to the present research interests can only be conjectured. It seems reasonable to assume that dropouts for purposes of marriage carry implications for actual fertility, but it is not altogether obvious that initial family-size preferences are involved. Our main concern with the characteristics of the dropouts, of course, is their possible effect on the freshman class and the reliability of the comparison with the senior class. Thus, if it could be determined that individuals with orientations toward larger families tended to drop out and our comparison of average family-size preferences of the two classes revealed no difference, we would be masking the fact that there may be a real increase in the value of the variable being measured. There is no completely satisfactory solution to this problem,14 but the sharp edges of its dangers can be blunted considerably "For an early statement of methodological aspects of this problem, see Stephen M. Corey, "Attitude Differences between College Classes: A Summary and Criticism," Journal of Educational Psychology XXVII, No. 5 (May 1936), 321-330.

BACKGROUND, SCOPE, AND METHOD

in two ways. In many of the higher quality schools, the attrition is small over the four-year period so that comparison among these schools15 will be relatively pure. Secondly, in schools with higher drop-out rates we can compare first- and fourth-year women coming from segments of the population in which drop­ outs will be lowest—for example, among daughters whose mothers attended college.18 Although such approaches are ad­ mittedly inconclusive they should help considerably in minimiz­ ing the problem (see Chapter 5 ) . The student who transfers from one institution to another during her college career poses still another problem, part of which can be easily resolved and part of which can only be described. Our sample of students in all schools has been purified to contain only women who have never attended another insti­ tution of higher education. The freshmen samples, however, con­ tain varying proportions of young women who will eventually graduate from different schools. Thus, the freshmen group in­ cludes this impurity, similar in some respects to its inclusion of eventual dropouts, but these students are probably more simi­ lar in relevant characteristics to the freshmen who will remain four years than are eventual dropouts. SUMMARY

The research reported in the first part of this volume is di­ rected toward determining the effects of higher education on fertility. We have focused more specifically on the effects of higher education of women on their attitudes toward family size and their intentions to plan the spacing and number of children. The study is an outgrowth of earlier research on Ameri­ can fertility, which indicated a relationship between education " This "solution" is complicated by the fact that our research interests included analysis of the possible influence on family-size values of differ­ ent types of schools which, in turn, have different rates of attrition. " There remains the logical possibility that such groups will differ substantially in the "effects" of the college experience on family-size values and that we will not be able to extend these observations to other groups.

I. INTRODUCTION

in Catholic colleges and the number of children desired as well as actual fertility. The current study is designed to determine whether the mechanism of this association is selectivity or the actual educational experience. Although stimulated initially by questions about Catholic fertility, the study was broadened to include women of all major religions—Protestants, Jews, Mormons, and women with no religious preference—as well as nonsectarian and Protestant institutions of higher education and schools under Catholic auspices. The population to be sam­ pled was defined as freshmen and senior women in four-year colleges and universities throughout the United States. A proba­ bility sample of 48 institutions was drawn, stratified for type of control (half Catholic), size, coed and women's colleges, and region of the country. Data were collected by questionnaires from women in 45 of these schools. In some of the larger institu­ tions, samples of students were drawn but in most schools we aimed (though response rates varied) at including all first- and fourth-year women. The overall response rate for all institutions and classes combined was 72 percent, which provided us with a total sample of approximately 15,000 women. The strategy of the analysis is to compare the fertility values of first- and fourth-year women classified by religion and type of institution. This synthetic cohort technique assumes that com­ parison of the two classes will reflect the changes that would take place if we actually followed a single group from the begin­ ning to the end of the four-year experience. The several meth­ odological problems of this assumption are discussed at length. The second part of this report presents the results of an inquiry into the social characteristics and relevant attitudes of college women in relation to their fertility values. These analyses are con­ ducted independently of the woman's class in college or type of college attended but are addressed to the general question of the characteristics of women of different religions who want large or small families and who differ in their intentions to plan fertility.

PART II THE EFFECTS OF HIGHER EDUCATION

CHAPTER 2

Types of Colleges and Family-Size Preferences The design of the sample of colleges and universities is based on the assumption that institutions of certain types (e.g., Catho­ lic women's colleges) randomly selected will yield distributions of family-size preferences and differences between classes varying only within sampling error. Such homogeneity within strata would then encourage classifying individuals within groups and ignoring the distinction by specific institution. There is an ob­ vious economy of data processing to be gained if various consoli­ dations can be justified. The detailed family-size preferences for first- and fourth-year students in 44 institutions1 are presented in Table 3 and sum­ marized in Table 4. The distribution of schools by the average number of children desired (Table 4) reveals a striking contrast between Catholic and non-Catholic schools and considerable homogeneity within these two categories. With the exception of one nonsectarian university, there is simply no overlap at all in average family-size preferences among women in Catholic and non-Catholic institutions. All Catholic schools fall above the aver­ age of 4.4 children and are concentrated in the 5.0 to 5.9 interval. And, with the exception of the one school already noted, all nonCatholic schools fall below the level of 4.0 children, being clus­ tered mostly in the 3.5 to 3.9 interval. Comparison of coeducational and women's schools suggests no difference in nonsectarian schools (the number of women's colleges is very small in this group, of course) and only a small difference among Catholic schools where women in coedu­ cational schools appear to show preferences for slightly smaller families on the average than students in Catholic women's colleges. The relationship between size of the school, measured in terms 1One university with largely Mormon students is omitted from the tabulation because we planned to analyze this group separately regardless of the distribution of their family-size preferences.

21

Jl. THE EFFECTS OF HIGHER

EDUCATION

TABLE 3 Number of Children Desired by Women, by Year and College or University SCHOOL

CLASS

No.

Percent Number Mean Standard Number of Children Desired or 3 4 5 6 or Tottd Demof less more otion JVomen

2

NONSECTARIAN COED SCHOOLS 50

51

52

54

57

58

59-60

61-62

63

66

67

68

1

17

37

29

5

12

100

192

3.7

1.6

4

16

40

32

3

9

100

106

3.6

1.5

1

11

30

33

13

13

100

224

3.9

1.6

4

9

29

46

7

9

100

140

3.7

1.2

1

20

31

38

6

5

100

361

3.5

1.2

4

21

34

34

6

5

100

194

3.4

1.1

1

21

26

33

9

11

100

121

3.7

1.7

4

19

32

32

7

10

100

197

3.5

1.3

1

17

37

31

7

8

100

521

3.5

1.4

4

19

37

30

7

7

100

136

3.5

1.3

1

19

25

35

7

14

100

149

3.7

1.5

4

22

37

28

3

10

100

60

3.5

1.5

1

21

24

40

8

7

100

201

3.6

1.5

4

27

24

30

8

11

100

119

3.5

1.4

1

16

32

33

9

10

100

809

3.7

1.4

4

20

38^ 35

3

4

100

326

3.3

1.1

1

20

31

33

9

7

100

172

3.6

1.5

4

17

35

35

6

7

100

144

3.5

1.2

1

18

29

31

9

13

100

115

3.8

1.5

4

32

34

28

4

2

100

50

3.1

1.0

1

12

19

28

13

28

100

196

4.6

2.2

4

9

24

36

15

16

100

123

4.2

1.5

1

25

27

30

3

15

100

143

3.7

1.6

4

18

25

45

6

6

100

49

3.6

1.6

22

TYPES OF COLLEGES AND FAMILY-SIZE

PREFERENCEi

TABLE 3 (continued) SCHOOL

CLASS

No.

Number of Children Desired Percent Number Mean Standard Devi2 or 3 4 5 6 or Total of ation less more Women

71

1 4

14 16

34 36

36 35

6 6

10 7

100 100

250 116

3.7 3.5

1.4 1.1

74-75

1 4

1 2

14 13

33 35

34 36

18 14

100 100

545 437

3.7 3.6

1.4 1.3

80

1 4

13 13

38 43

31 31

7 10

11 3

100 100

137 112

3.7 3.5

1.4 1.0

81

1 4

18 26

34 35

34 27

8 6

6 6

100 100

346 172

3.5 3.3

1.2 1.3

82

1 4

16 13

33 23

32 49

11 9

8 6

100 100

301 130

3.7 3.7

1.4 1.1

90

1 4

15 13

26 61

42 17

7 —

10 9

100 100

168 23

3.7 3.2

1.4 1.2

91

1 4

12 11

26 35

43 44

11 6

8 4

100 100

310 113

3.9 3.6

1.4 1.0

92

1 4

20 14

26 43

38 32

5 9

11 2

100 100

79 44

3.9 3.4

1.9 0.9

10

1 4

6 3

7 7

20 26

17 22

50 42

100 100

152 82

5.8 5.5

2.6 2.2

11

1 4

3 2

12 5

23 55

15 18

47 20

100 100

68 44

5.8 4.8

2.5 1.7

12

1 4

7 6

6 16

27 34

10 20

49 24

100 100

96 50

5.3 4.7

2.0 1.8

13

1 4

2 2

5 8

25 21

21 18

47 51

100 100

163 61

5.7 5.9

2.1 2.3

NONSECTARIAN VI^OMEN'S COLLEOI:s

PROTESTANT COED SCHOOLS

CATHOLIC COED S(3HOOLS

23

II. THE EFFECTS OF HIGHER EDUCATION TABLE 3 (continued) SCHOOL

CLASS Number of Children Desired Percent Number Mean

No.

2 or less

3

4

5

6 or more

Total

of Women

Standard Deviation

15

1 4

3 4

7 33 7 3 32 29

50 32

100 100

30 28

5.8 5.3

2.5 2.0

16

1 4

4 16 32 16 5 14 36 15

32 30

100 100

399 171

4.9 4.8

1.9 2.0

18

1 4

3 7

53 45

100 100

221 74

5.9 5.3

2.2 2.1

20

1 4

4 10 6 6

44 34

100 100

186 53

5.7 4.9

2.4 1.4

30

1 4

1 11 4 7

31

1 4

2 1

32

1 4

— 10

33

1 4

4 10 4 7

34

1 4

4 2

35

1 4

36

1 4

4 1

37

1 4

5 3

38

1 4

5 22 17 6 26 16 23 19 30 24

CATHOLIC WOMEN'S COLLEGES

22 19 24 16

47 49

100 100

148 55

5.6 5.6

2.1 2.0

5 24 19 7 22 21

50 49

100 100

161 91

6.3 5.6

2.6 1.9

6 26 5 16

26 21

42 48

100 100

31 19

5.9 5.5

2.4 2.3

29 14 24 19

43 46

100 100

119 54

5.5 5.9

2.3 2.6

21 25

36 41

100 100

168 180

5.1 5.6

1.9 2.1

34 16 24 25

39 44

100 100

103 45

5.4 5.5

2.1 1.6

6 2

28 15 26 32

47 39

100 100

116 85

5.5 5.5

2.1 1.8

6 6

27 15 27 19

47 45

100 100

150 91

5.5 5.5

2.2 2.0

2 11 28 18 1 14 30 20

41 35

100 100

123 100

5.4 5.1

2.2 1.8

9 29 4 28

1 10 — 7

TYPES OF COLLEGES AND FAMILY-SIZE PREFERENCES TABLE 3 (continued) SCHOOL No.

CLASS Number of Children Desired Percent Number Mean Standard 2 or 3 4 5 6 or Total of Deviless more Women ation

39

1 4

11 3

7 7

23 32

17 14

42 44

100 100

110 135

5,,2 5,.4

2.2 2.0

40

1 4

2 1

7 4

27 28

17 21

47 46

100 100

126 75

5,.7 5,.5

2.1 1.8

41

1 4

2 5

9 5

24 14

14 14

51 62

100 100

63 42

6..1 6 .0

2.6 2.1

42

1 4

4 —

8 5

26 27

18 16

44 52

100 100

180 94

5..4 5 .3

2.1 1.4

43

1 4

5 4

14 5

25 43

14 10

42 38

100 100

85 42

5,.3 5 .1

2.2 2.1

44

1 4

3 —

4 13

21 37

17 25

55 25

100 100

29 11

6 .0 4 .9

2.3 1.4

1 4

4 11

8 6

20 28

21 5

47 50

100 100

75 18

5 .6 5 .0

2.1 2.0

45

of total female enrollment, and family-size preference was also examined to determine whether the larger schools might produce (or attract) women with less familistically oriented values (Table 5). Only a slight suspicion of such a relationship emerges for women in nonsectarian coeducational institutions while among Catholic women's colleges2 no such pattern appears at all. Examination of freshman-senior differentials also warrant our combining the individual institutions into general classes. The only possible reservation is that the differences in family-size preference between freshmen and seniors enrolled in Catholic coeducational schools seem slightly larger than in Catholic 'These are the only two groups large enough to tabulate for this analysis.

II. THE EFFECTS OF HIGHER EDUCATION

TABLE 4 Distribution of Schools by Average Number of Children Desired by First- and Fourth-Year Women Mean Number of Children Desired

3.0-3.4 3.5-3.9 4.0-4.4 4.5-4.9 5.0-5.4 5.5-5.9 6.0-6.4 Number of schools

Non sectarian Coed

Non sectarian All Women

Protes­ tant Coed

Catholic Coed

Catholic All Women

1

4

1

4

1

4

1

4

1

4

—

3 10 1

3

1 2

— 3

2 1

—

—

—

—

1 1 6

4 2 2

8

8

— 6 7 3 16

1 5 9 1 16

13 —

1

14

14

3

3

3

3

—

women's colleges (Table 3) and a single Catholic college group­ ing might obscure some variations. As subsequent analysis will reveal, however, the distinction is relevant only for a very small proportion of our study population and we decided to deal with this issue separately as a possible modification of our basic con­ clusion about the influence of Catholic higher education. In view of the patterns revealed in Table 3, the decision was made to classify the colleges into three groups: nonsectarian schools, Protestant schools, and Catholic schools. Protestant schools have not been merged with nonsectarian institutions de­ spite the similarity of their family-size preference distributions TABLE 5 The Average Number of Children Desired by Women in Small and Large Nonsectarian Coed Institutions and Catholic Women's Colleges Size and Type of School

Nonsectarian coed schools with at least 2,000 women Nonsectarian coed schools with less than 2,000 women Catholic women's schools with at least 600 women Catholic women's schools with less than 600 women

1

4

3. 6 3. 8 5 .5 5 .6

3. 5 3..7 5 .5 5 .5

TYPES OF COLLEGES AND FAMILY-SIZE PREFERENCES

because a different hypothesis exists about the effect on fer­ tility values of education in church-sponsored institutions by comparison with education in nonsectarian schools. And finally, one other school not included in these tabulations is retained for separate analysis because of a high concentration of Mormon students who exhibit a distinctive fertility-value system. THE COLLEGE EXPERIENCE AND CHANGE IN FAMILY-SIZE VALUES

In the following three chapters the main question of our study—whether higher education affects fertility values—will be probed within various refined controls that are absent in Table 3. For example, the religion of the student is not controlled in this tabulation, which is especially pertinent for the nonsec­ tarian schools. The overall picture certainly does not seem to indicate any great differences between first- and fourth-year students. Virtu­ ally all of the non-Catholic institutions reveal slight "declines" in average family-size preferences from first- to fourth-year stu­ dents with these differences being concentrated in the 5 to 10 percent range. Although at a significantly higher level, the pat­ tern among Catholic schools is predominantly the same, that is, a slight decline in average family-size preferences from firstto fourth-year students especially in coeducational schools, but there are more Catholic than non-Catholic schools that deviate from this pattern. Refined tests of particular hypotheses about the effects of higher education on fertility values are deferred to the following chapters. THE COLLEGE EXPERIENCE AND CONSENSUS ON FAMILY-SIZE VALUES

Although little difference appears to emerge between firstand fourth-year students in the size of family they prefer on the average, women in their fourth year evidently think alike more often on the subject than do first-year women. In other words, there is much more homogeneity of preference of family

II. THE EFFECTS OF HIGHER EDUCATION

size among seniors than among freshmen. In 37 of the 44 com­ parisons in Table 3 the standard deviation of the distribution on number of children desired is lower, and frequently substan­ tially lower, among seniors. Such a finding is not difficult to understand.3 These women have lived together and have been exposed to similar stimuli for four years and it would be strange indeed if their thinking on all kinds of subjects did not become more alike.4 In the case of family-size values, there is the additional process of matu­ ration, which could be responsible for reducing the extremes of the distribution. There is, of course, also the possibility that the process of attrition over four years (both dropouts and transfers) selects students on the extremes and thereby reduces the variance, an effect that only a longitudinal study could satis­ factorily isolate. * P. E. Jacob (Changing Values in College, New York: Harper & Bros., 1957) reports similar findings in his review of college studies. A contrary result is reported in Harold Webster, "Changes in Attitudes During College," Journal of Educational Psychology, XLIX (June 1958), 109-117. On the basis of a longitudinal study of 274 Vassar students, in which a variety of attitudinal scales were administered, Webster concluded: "Results are interpreted as supporting those person­ ality theories which emphasize increasing complexity, differentiation, ability and independence during late adolescence; they do not support the view that college students become more alike in their general atti­ tudes while attending college" (p. 116). 4A supplementary analysis is planned to determine whether such convergence occurs on different measures of behavior and of values and attitudes, and whether certain groups are more or less vulnerable to the process.

CHAPTER 3

Higher Education and Family-Size Preferences Our central hypotheses relate to the differences expected in fam­ ily-size preferences among different religious groups and between first- and fourth-year students of each major religious preference in each different type of educational institution. They are based on assumptions about the differential impact of the secularization process as it affects students in nonsectarian and denominational colleges. Numerous other analyses, containing many additional variables, are included elsewhere in this report. Some of these are elaborations of analyses implied by the following hypotheses; others reflect specific substantive interests in the social correlates of the number of children desired. SPECIFIC HYPOTHESES

1. Women with no religious preference, in nonsectarian institutions Of all groups considered, these women are expected to register preferences for the smallest families and to experience no change after four years of college. 2. Protestant and Jewish women in nonsectarian institutions We expect the family-size preferences of Jewish women to be slightly lower than Protestant women and expect both groups to experience a decline (that is, fourth-year students tend to prefer smaller families than first-year students) over the four years of college. 3. Protestant women in Protestant institutions We have included only three such colleges in our sample and thus do not represent all Protestant denominations. With this restriction, our expectation is that these women will prefer larger families than Protestant women in nonsectarian institutions and that this higher level will be maintained during their college 29

II. THE EFFECTS OF HIGHER EDUCATION

years; that is, no change in average number of children desired is expected. 4. Protestant women in Catholic institutions Although this group is numerically insignificant, it represents an interesting theoretical combination. On the assumption of peer-group influence and the factors disposing the girl to select such a college at the outset, we expect their family-size prefer­ ences to be the same as those of Protestant women attending Protestant colleges, but expect the desired number of children to increase after four years of the experience. 5. Mormon women in nonsectarian institutions Special interest attaches to the fertility of members of the Church of Jesus Christ of Latter-Day Saints because they consti­ tute a high-fertility non-Catholic population. As a matter of fact, one recent, very pertinent study1 in 1956 of the fertility of college graduates who graduated 10 and 25 years before indicated that Mormon graduates might have even higher mari­ tal fertility than graduates of Catholic colleges, although differ­ ences in age at marriage obscured the comparison. In the present study, we have included a sample of 388 Mormon women from one nonsectarian university. The hypothesis to be tested is that the average size of family desired by Mormon women will be greater than that desired by all other non-Catholic women but not as high as the family-size preferences of Catholic women, especially women educated in Catholic schools. Given the non­ sectarian nature of the university from which the sample was drawn, our expectation is that the average family-size preferences of fourth-year Mormon women will be lower than that of women in their first year. CATHOLIC women, irrespective of both the type of college they currently attend and whether their previous education was in '"College Study Report—1956," Population Bulletin, XII (1956). This report was prepared by the staff of the Population Reference Bureau with P. K. Whelpton and Patience Lauriat.

30

ON FAMILY-SIZE PREFERENCES

Catholic schools, are expected to express preferences for families larger than any of the groups above. 6. Catholic women in nonsectarian institutions A. WOMEN WHO ATTENDED NONSECTARIAN HIGH SCHOOLS:

We expect such women to prefer larger families than their non-Catholic fellow students but to experience a similar decline over the four years. B. WOMEN WHO ATTENDED CATHOLIC HIGH SCHOOLS:

These women are expected to prefer larger families than Catholic students who attended nonsectarian high schools and to experience an even greater decline during the college years. 7. Catholic women in Catholic institutions A. WOMEN WHO ATTENDED NONSECTARIAN HIGH SCHOOLS:

These women are expected to prefer larger families than their Catholic classmates from high school who elected to attend nonCatholic colleges and to prefer fewer children than their class­ mates in college who attended Catholic high schools. The num­ ber of children they prefer is expected to increase over the four-year period. B. WOMEN WHO ATTENDED CATHOLIC HIGH SCHOOLS:

This group is expected to express preferences for the largest families of all groups considered and to experience little change in this preference during the college years. Each of these four subgroups of Catholic women has been divided further into two categories on the basis of whether the elementary school attended was Catholic. These eight subsamples range in size from a high of 3,904 women whose education has been exclusively Catholic at all levels to a low of 45 women in nonsectarian colleges who had attended a Catholic high school and a nonsectarian elementary school. Since these eight groups 31

11. THE EFFECTS OF HIGHER EDUCATION

are subdivided again in order to compare first- and fourth-year students, some strain on sample size and therefore the stability of statistical estimates results, but the sample is large enough so that in only 3 of the 16 groups does the number of cases fall below 100. Our interest in Catholic elementary education is merely an extension of the same logic connecting family-size values and Catholic education at higher levels. Within each of the four groups above (6a and b, and 7a and b) the expectation is that women with a Catholic elementary education would exhibit the higher fertility orientation. These various combinations of religion, type of institution, and type of elementary and secondary schools attended by Catholics produce 14 subsamples, each divided into first- and fourth-year TABLE 6 Number of First- and Fourth-year Women1 in the Sample by Religion, Type of Colle and Type of Elementary School and High School8 for Catholics Type of College and Year

Religion of Student

NONSECTARIAN

1 226 None Protestant 3,682 Jewish 529 Catholic: Nonsectarian high school and: Nonsectarian elementary school 423 167 Catholic elementary school Catholic high school and: Nonsectarian elementary school 29 Catholic elementary school 229 229 Mormon All women 5,514

4 197 1,836 335

Totc

CATHOLIC

PROTESTANT

14

14

*

81 *

*

21 *

269 101

450 322

270 151

16 122 159 3,035

146 2,524

104 1,380

*

*

613

201

*

*

*

*

*

*

*

*

*

*

*

*

*

*

3,523

1,926

613

201

42 6,43 8«

1,41 74 25 4,25 3i 14,81

* Number not shown if frequency is less than 10. all transfer students, foreign-born persons, nonwhites, persons born before IS and persons of religions other than those listed above. These exclusions amount to 2, women, most of whom are in the category of transfer students. ' The classification of elementary school and high school refers to the type of school fi which the person graduated. 1 Excludes

ON FAMILY-SIZE PREFERENCES

students for a grand total of 28 groups. The number of women in each group is shown in Table 6. They range in size from 16 (Catholic women in their fourth year at nonsectarian institutions who attended a Catholic high school and a nonsectarian elemen­ tary school) to 3,682 (Protestant first-year students in nonsec­ tarian colleges.) Our hypotheses involve three kinds of comparisons—across religions, within the same religion but across types of colleges, and across time as inferred by the comparison of first- and fourth-year students. A special inquiry into the difference between Catholic women's colleges and Catholic coeducational institutions concludes the analysis. RESULTS 1. Women with no religious preference, in nonsectarian institutions

The first hypothesis—that women with no religious preference would express desires for the smallest families—is sustained (see Table 7). Of all 14 groups of freshmen women, none expressed desires for families as small as the 3.2 average recorded by women who report no religious preference, and seniors as well indicate preferences for the smallest families (an average of 3.1 children). Of all groups in the analysis, this group shows the highest proportion expressing preference for no children—7.3 percent among freshmen and 4.4 percent of seniors.2 On the suspicion that no religious preference and an interest in a career might be associated and that an interest in a career might ex­ plain the higher proportions not wanting any children, we inves­ tigated the responses to the question, "Do you intend to have a career outside the home during most of your married life?" in relation to the number of children desired, and discovered that this indeed is the explanation. Among women with no re­ ligious preference about half answered "yes" to this question 1 The tabulation is confined to women who intend to marry. These proportions seem high in view of the fact that voluntary childlessness seems to have all but disappeared on the American scene.

33

II. THE EFFECTS OF HIGHER EDUCATION

and almost all of the women desiring no children are in this category. Although women with no religious preference constitute only about 3 percent of the total sample, they are interesting for theoretical reasons in that they may be thought to represent the absence of a religious effect as such on the attitude toward family size. Many of these women, of course, may be rejecting organized religion actively, others may simply be indifferent. Although an expression of no religious preference does not imply automatically that the girls are not religious, there is a definite statistical association in our data. These women have by far both the lowest scores on a scale measuring participation in religious activities and, perhaps more critically, the highest pro­ portion of those replying that religion held no importance at all for them. Turning to the second part of the first hypothesis—that there would be no decline after four years of college—we do find a slight drop in the expected direction (from 3.2 to 3.1) but the magnitude of the difference is not statistically significant and the most reasonable conclusion is that there is no difference. The reasoning behind the hypotheses is that a nonsectarian edu­ cation exerts a secularizing influence on any religious elements supporting a high fertility value and, in the case of persons with no religious preference, such an effect is assumed to be minimal. 2. Protestant and Jewish women in nonsectarian institutions A. PROTESTANT WOMEN:

Before combining women of all Protestant denominations, the family-size preferences of the larger denominations were exam­ ined separately. These are the Baptist, Lutheran, Methodist (in­ cluding Evangelical and United Brethren), Presbyterian (includ­ ing Reformed Church of America), Episcopalian (including An­ glican), Congregational (including Evangelical Reformed 34

ON FAMILY-SIZE PREFERENCES

United Church of Christ), Christian (including Disciples of Christ), and all others combined. The variation in the average number of children desired across all eight groupings was 3.4 to 3.6—a very homogeneous set of groups indeed.3 As a matter of fact, the averages of all eight groups are closer to the Protes­ tant average than any is to the Catholic average or, for that matter, to that of the Jewish group or to the average of women with no religious preference. This pattern of results clearly justi­ fies for our purposes the merging of all denominations under the label "Protestant." Our hypothesis about the effect of education in a nonsectarian institution on the family-size values of Protestant women is that it will bring about a reduction in the number of children desired. A comparison of the appropriate means reveals a slightly lower value (3.4) among seniors than among freshmen (3.5) and, although even this small difference with such large samples is unlikely to have occurred by chance,4 the most reasonable con­ c l u s i o n i s t h a t t h e r e i s v e r y l i t t l e if a n y c h a n g e ( T a b l e 7 ) . B. JEWISH WOMEN:

The number of children desired by Jewish women is hypothe­ sized to be lower than the number preferred by Protestant women and, as with Protestants, is expected to decline after four years in nonsectarian colleges and universities. Both of these expectations are confirmed, with the differences between Protestants and Jews (by class) and the difference between firstand fourth-year Jewish students being statistically significant (Table 7). The latter difference, however, is just barely signifi­ cant statistically (at the 0.05 level) and again the most reason3 The

same homogeneity among Protestant denominations was reported in a study of family-size preferences and actual fertility among married women in the largest metropolitan areas of the United States. See Westoff, Potter, Sagi, and Mishler, Family Growth in Metropolitan America, p. 182, and Westoff, Potter, and Sagi, The Third Child, p. 81. * The difference between the means of these two samples could occur by chance in theory less than 0.001 times.

II. THE EFFECTS OF HIGHER EDUCATION

TABLE 7 Number of Children Desired by Women with No Religious Preference, Protestant Womei Jewish Women, and Mormon Women by Year and Type of Institution Religion of Woman

None

Protestant

Jewish

R-otestant

Type of Institution and Year

3

4

5 6 or Total more

Nonsectarian 1 4

11 5

24 32 18 26 32 27

8 7

8 3

100 100

205 182

3.2 3.1

1.8 1.3

Nonsectarian 1 4

1 1

17 34 34 19 36 33

7 5

7 5

100 100

3,609 1,751

3.5 3.4

1.3 1.1

Nonsectarian 1 4

1 2

18 40 30 18 43 31

7 5

4 2

100 100

514 319

3.4 3.2

1.2 0.9

Protestant 1 4

1 1

14 27 40 13 43 34

9 5

9 4

100 100

600 195

3.8 3.4

1.5 1.0

1

14 35 32 24 33 33

7 10 5 5

100 100

77 21

3.7 3.3

1.5 1.0

4 20 37 11 27 4 15 36 13 31

100 100

229 154

4.6 4.8

1.7 1.8

Protestant Catholic 1 4 Mormon

Percent Number Mean Standai Devi­ ation

Number of Children Desired

Less 2 than 2

Nonsectarian 1 4

—

1 1

able conclusion is that if higher education does depress the fam­ ily-size preferences of Jewish women, its effect is very slight indeed. 3. Protestant women in Protestant institutions

Our sample of colleges includes three Protestant institutions— one in the southeast, one in the southwest, and one in the north central part of the country. Although such a sample is too small to represent adequately the full range of such colleges, we do have enough students altogether in the three (800) to provide

ON FAMILY-SIZE PREFERENCES

some reliable generalizations about a small segment of this population. The hypothesis asserts that Protestant women attending Prot­ estant schools will prefer larger families than Protestant women attending nonsectarian schools (due to the selectivity of more religious girls to religious institutions) and that the preference will remain unchanged after four years. The first part of this hypothesis is sustained—freshmen women in the Protestant col­ leges express a preference for an average of 3.8 children com­ pared with the 3.5 average desired by Protestant women in their first year in nonsectarian institutions (Table 7).5 The second part, however, is not sustained, as the seniors in Protestant insti­ tutions express preferences for significantly® fewer children (3.4) than the freshmen (3.8). In fact, the downward change is even greater than that observed over the four years in nonsectarian institutions (3.2 for freshmen and 3.1 for seniors). 4. Protestant women in Catholic colleges Although less than 2 percent of all Protestant women in our sample attend Catholic colleges and universities, they are an interesting group theoretically. The decision to attend a Catholic institution is undoubtedly influenced by propinquity as well as religious considerations. Nevertheless, the suspicion arises that such women probably come disproportionately from homes in which one of the parents is Catholic. This expectation is con­ firmed by our data, which indicates that 11 percent of these girls have Catholic mothers and 15 percent have Catholic fathers—proportions that are much higher than the correspond­ ing estimates of 1 and 2 percent for Protestant girls in nonsec­ tarian institutions and 1 percent for both parents of girls in Protestant colleges. The presumed influence of a Catholic parent, as well as asso­ ciation with Catholic women, led us, therefore, to hypothesize that Protestant women attending Catholic colleges would prefer •Difference is significant at the 0.001 level. " Two-tailed test significant at the 0.01 level.

37

II. THE EFFECTS OF HIGHER EDUCATION

larger families than Protestant women going to nonsectarian institutions—at about the same level as Protestant women at­ tending Protestant schools. Both expectations are supported (Table 7) but the first comparison—between first-year Protes­ tant women in Catholic colleges and those in nonsectarian schools—is not statistically significant, although the difference in means (3.7 compared with 3.5) is in the expected direction. The second part of our general hypothesis calls for an increase in the number of children desired after four years, on the reason­ ing that exposure to Catholic students and their large-family values would have some effect. This expectation proves wrong, however, and, although the difference is not statistically signifi­ cant, the reverse actually occurs—seniors prefer 3.3 children compared with 3.7 among freshmen. In fact, the whole pattern is indistinguishable from Protestant women attending Protestant schools. 5. Mormon women in nonsectarian institutions The expectation that Mormon women educated in a nonsec­ tarian institution would desire more children on the average than other non-Catholic women is confirmed; the average number desired is 4.7 children (Table 7). We also hypothesized that Mormon college women would prefer families smaller than those desired by Catholic women, especially those educated in Catholic schools. In fact, the average of 4.7 does fall in between the number desired by Catholic women whose education has been exclusively secular (4.2) and women educated solely in Catholicschools (5.6). The hypothesis that four years of nonsectarian educational experience would reduce the size of family preferred is not sus­ tained. The difference between first- and fourth-year women is in the opposite direction from that expected (4.6 for freshmen and 4.8 for seniors) and statistical tests fail to support the con­ clusion that the two samples differ at all. REGARDLESS of whether the college they attend is Catholic or

whether their earlier education was received in Catholic schools,

ON FAMILY-SIZE PREFERENCES

women who identify themselves as Catholic desire considerably larger families than all except Mormon women (whose familysize values are exceeded only by Catholics educated mainly in Catholic schools). The family-size preferences of all other nonCatholic women range between 3.1 and 3.8 children, while the number of children desired by Catholic women in college (all kinds of colleges) ranges between 4.1 and 5.7. Catholic women in their first year express a preference for 5.3 children and those in their senior year say they want 5.1 children on the average. This "decline" of the average results from a downward shift in the proportion of women saying they want very large fam­ ilies—preferences for 7 or more children decline from 22 percent among freshmen to 14 percent among seniors. Preferences for 2 or fewer remain negligible in both classes—5 and 4 percent in the first and fourth year respectively—a figure that differs markedly from the proportions among non-Catholics, which range from 13 to 35 percent. The following section is devoted to a report of the "effects" of Catholic education upon attitudes toward family size. All relevant statistics are summarized in Table 8. 6. Catholic women in nonsectarian colleges A. WOMEN WHO ATTENDED NONSECTARIAN HIGH SCHOOLS:

This group was hypothesized to prefer smaller families than any other group of Catholics. With the further subdivision on the basis of Catholic or non-Catholic elementary school, we are able to isolate Catholic women whose education has been exclu­ sively secular, the group that is expected to prefer the smallest families of all eight categories of Catholic women. The distribu­ tion of the number of children desired (Table 8) clearly supports this hypothesis, with an average of about 4.0 children preferred. This is significantly lower than the number preferred by all other Catholic women. The second subgroup is comprised of women whose elementary education was in a Catholic school and whose subsequent educa­ tion has been in nonsectarian institutions. These women, as ex-

11. THE EFFECTS OF HIGHER EDUCATION

TABLE 8 Number of Children Desired by Catholic Women by Year and Type of Institution, Subdivided by Type of High School and Elementary School Elementary High School School

NS

NS

College Year

Number of Children Desired

Less than 2 NS

1 4

—

2

2

3

4

5

Per­

6 or cent more Total

Num­ Mean Stana ber ard Devi atior,

11 25 31 14 12 23 40 10

17 14

100 100

397 253

4.1 4.0

1.7 1.3

C

NS

NS

1 4

1 1

9 18 34 12 6 19 45 12

26 16

100 100

153 98

4.4 4.3

1.7 1.6

NS

C

NS

1 4

—

4 11 19 6 50 31

—

58 6

100 100

26 16

6.4 3.3

2.9 1.2

1 4

—

3 12 35 18 6 16 39 14

33 25

100 100

217 118

5.0 4.5

1.8 1.4

7 14 33 16 5 12 40 16

29 27

100 100

433 263

4.8 4.8

2.1 1.8

3 12 31 15 2 9 32 16

38 41

100 100

307 126

5.1 5.3

2.0 1.9

C

NS

C

NS

C

C

NS

NS

C

C

NS

C

C

C

C

6

—

1

8

1 4

—

1 4

—

1 4

—

2 10 26 20 3 7 28 15

42 47

100 100

137 96

5.5 5.5

2.1 2.1

1 4

—

2 2

48 43

100 100

2370 1296

5.7 5.4

2.2 1.9

1

—

7 25 18 7 28 20

NS = Nonsectarian. C = Catholic.

pected, want slightly more children on the average (0.3 more)7 than Catholic women whose education has been exclusively secular and they, in turn, prefer smaller families than any of the other Catholic groups. The slightly higher family-size preferences of the women who attended Catholic elementary schools cannot, of course, be inferred to be the simple result of this educational experience; it is quite probable that girls from more religious * Significant at the 0.05 level.

ON FAMILY-SIZE PREFERENCES

families were sent to parochial schools and that their preference for larger families may reflect this initial selectivity more than the educational experience. This entire question of selectivity as opposed to the institutional effect is central to our major concern in this study and we shall return to it more systemati­ cally later. The second part of the hypothesis asserts that the size of family desired by Catholic women educated in nonsectarian high schools would decline after four years in a nonsectarian college. The average number of children is slightly lower among seniors in both subgroups but the differences are not large enough to sustain the hypothesis. Evidently women whose secondary school education was secular are not particularly vulnerable to any possible further secularizing influences on family-size atti­ tudes in nonsectarian colleges. In this sense, the pattern of change in family-size values is identical with that of most nonCatholic women in nonsectarian colleges. B. WOMEN WHO ATTENDED CATHOLIC HIGH SCHOOLS:

As hypothesized, Catholic women in their first year of nonsec­ tarian colleges who attended Catholic high schools desire more children (5.1) than their college classmates who went to nonCatholic high schools (4.2). As we shall see subsequently, this is about at the same level (4.9) as Catholic women in their first year at Catholic colleges who attended nonsectarian high schools but lower than the number desired by women who at­ tended both Catholic high schools and Catholic colleges (5.7). Subdivisions of this group on the basis of the type of elemen­ tary school education produces two categories of unequal size. There are very few Catholic women in nonsectarian colleges (3.3 percent in our sample) who attended the combination of a Catholic high school and a non-Catholic elementary school; only the group whose previous education was exclusively Catho­ lic is numerically important. As expected, there is a statistically significant drop in the number of children desired in both groups as one moves from

II. THE EFFECTS OF HIGHER EDUCATION

the first to the fourth year. There appears to be an especially sharp drop within the small subgroup of the women who at­ tended nonsectarian elementary schools (from 6-4 to 3.3 children desired), but, as indicated, the frequencies are very small, and although the difference is beyond any reasonable chance fluctua­ tion, it should probably not be taken too seriously. In both groups the decline is due to a drop in the proportion of women pre­ ferring 5 or more children. 7. Catholic women in Catholic colleges A. WOMEN WHO ATTENDED NONSECTARIAN HIGH SCHOOLS:

The first part of our hypothesis for this group is clearly sup­ ported—graduates from nonsectarian high schools electing to attend a Catholic college prefer larger families (4.9) than their Catholic classmates who went to a nonsectarian institution (4.2). The second part—that women who moved on to Catholic col­ leges from nonsectarian high schools would prefer fewer children than their college classmates from Catholic high schools—is also clearly sustained. The women comprising this latter group prefer a greater number of children than those in any of the larger Catholic student populations—an average of 5.7 children. These differences appear among first-year students and thus raise questions about the "influence" of previous education. Sub­ dividing our group of Catholic college women who attended nonsectarian high schools on the basis of whether their elemen­ tary school was Catholic again confirms our expectation about the direction of the difference, though the magnitude is small— those who attended a Catholic elementary school prefer 5.1 chil­ dren compared with 4.8 preferred by Catholic first-year students who had gone to a nonsectarian elementary school. Although this evidence consistently suggests selection—initial selection of more religious girls into Catholic grade schools and further, perhaps greater, selection into Catholic high schools, and finally into Catholic colleges—the question whether familysize values are sustained or even increased in Catholic colleges

ON FAMILY-SIZE PREFERENCES

remains pertinent. The last part of our hypothesis calk for preferences for larger families among fourth-year students than among first-year women in Catholic colleges on the assumption that the values encountered in the Catholic college (at both formal and informal levels) promote a norm of high fertility. A comparison of the two classes does not reveal any change at all—women begin with a preference for 4.9 children as fresh­ men and seniors report the identical number. There is, however, a slight change—from 5.1 to 5.3—among women who attended Catholic elementary schools, but it is not statistically significant. It appears that the "effect" of a Catholic college education on the family-size values of women educated in nonsectarian high schools is sustainment of the initial level. But when these groups are cross-classified by type of Catholic college, as in Table 9, an interesting though not completely consistent pattern emerges. Among women with nonsectarian elementary and high school education, seniors in women's colleges want 0.3 more children while seniors in coeducational schools want 0.6 fewer children. This reversal, however, is not maintained among women from Catholic elementary schools and nonsectarian high schools. The data suggest the possibility that the Catholic co­ educational colleges reduce and the women's colleges increase the family-size preference of students educated exclusively in nonsectarian institutions, perhaps because the former, enrolling more nonresident students, serve less as a counteracting Catholic reference group and the level of interaction and reinforcement is probably greater in the smaller, residential women's colleges. However, the data are tenuous because of the small number of cases in some categories and hence are only suggestive. B. WOMEN WHO ATTENDED CATHOLIC HIOH SCHOOLS:

The overwhelming proportion (94 percent) of women in Catholic colleges who attended Catholic high schools also at­ tended parochial schools. As hypothesized, this group of fresh­ man women whose education has been exclusively in Catholic schools express preferences for the largest families (5.7 children)

II. THE EFFECTS OF HIGHER EDUCATION

of all groups under consideration. Almost two thirds say they would like to have five or more children and fully 10 percent say they want 9 or more children.8 Probably such responses should not be taken too seriously and might be regarded as the im­ mature sentiments of late adolescents just out of high school. However, such an interpretation is weakened somewhat by the fact that the proportion wanting large families remains quite high among seniors who are in their early twenties. Although the proportion in the category of women preferring 9 or more children drops to 6 percent, the proportion of women preferring 6, 7, or 8 children remains about the same. One might still maintain that unmarried women just graduating from college have not yet faced the realistic demands of raising a family and that they will quickly modify their attitudes after they have their first few children. And perhaps so. Nevertheless, there is sufficient evidence that married Catholic women who have at­ tended Catholic schools and who already have children still want, and will have, a much larger number of children than other Catholic women. One study of a probability sample of married women with two children living in the largest metro­ politan areas of the country indicates that Catholic women who attended Catholic colleges as well as Catholic high schools and parochial schools still wanted an average total of 5.1 children in their families. And this question was posed six months after they had had their second child.9 Still, one might argue that the reference is only to what women say they would like to do rather than to their actual behavior. But three years later, when these same women were interviewed again, 54 percent 8 The actual mean (and the value used in all calculations) is 10.7 children. A small fraction of women (16 of the 2,370 freshmen women in this Catholic educational group) who replied to the question on number of children desired with non-numerical statements, such as "It's up to God," or "As many as God sends," are also included in the "9 or more" category. * At the same educational level, Catholic women who had no Catholic school education preferred a total of 3.4 children. See Westoff, Potter, Sagi, and Mishler, Family Growth in Metropolitan America, p. 219.

44

ON FAMILY-SIZE PREFERENCES

had experienced two additional pregnancies during the interval, with only 4 percent reporting no pregnancy.10 In another study of the fertility of college graduates, the conclusion was drawn that by "the time the class of 1946 reaches its 25th anniversary, it is quite probable that in the average number of children per married graduate and per parent, the alumni of Catholic colleges will exceed those of Mormon colleges."11 And the alumni of Mormon colleges in that study exhibited by far the highest fertility of any group. In view of this kind of evidence, it seems reasonable to assume that the high-fertility values of these Catho­ lic college women in our sample are not to be regarded too lightly. Returning to our hypotheses, the smaller of the two groups— women who attended a nonsectarian elementary school and then went on to a Catholic high school and continued in a Catholic college—want only slighdy fewer children (an average of 5.5) than do women educated exclusively in Catholic schools, which is consistent with our overall generalization about the implica­ tions for fertility values of the kind of elementary school attended. The final part of the hypothesis asserted that women educated in Catholic high schools would experience little change in their family-size preferences after four years in a Catholic college. Actually there is a drop—from 5.7 to 5.4—which is statistically significant,12 so that formally the hypothesis is refuted. Among this group, attendance at a coeducational or women's college appears relevant only for those students with a nonsectarian elementary school education, with seniors in women's colleges desiring more children on the average than freshmen, and seniors 10 Less than 20 percent of other Catholic women who attended college reported 2 pregnancies and over one third reported none. See Westoff, Potter, and Sagi, The Third Child, p. 98. 11 "College Study Report-1956," Population Bulletin, XII (1956). This report was prepared by the staff of the Population Reference Bureau with P. K. Whelpton and Patience Lauriat. 12A two-tailed test puts the probability of the difference between the means at 0.05 percent.

II. THE EFFECTS OF HIGHER EDUCATION

TABLE 9 Number of Children Desired by Catholic Women in Catholic Colleges, by Type of Elementary and High School and Whether the College Attended is Coeducational or a Women's College Elementary School

High School

Catholic Coed Colleges

Freshmen MEAN N NS C NS C

NS NS C C

4.9 5.4 6.1 5.8

( 189) ( 130) ( 55) (1,068)

Catholic Women's Colleges

Seniors MEAN N 4.3 5.9 5.0 5.4

( 82) ( 46) ( 33) (472)

Freshmen MEAN N

Seniors MEAN N

4.7 4.9 5.1 5.7

5.0 5.0 5.7 5.4

( 244) ( 177) ( 84) (1,302)

NS = Nonsectarian. C = Catholic.

in coeducational institutions wanting fewer than the freshmen (see Table 9). Again, however, the number of cases involved is small, and, among the great majority, women from Catholic elementary schools and high schools, the pattern in the two types of college is the same. Generally the following conclusion emerges. There appears to be a tendency in the Catholic colleges for the number of children desired to decline slightly except for students with a nonsectarian elementary or high school education enrolled in a women's college, who appear to experience some increase. Whether this is the result of the college experience itself is prob­ lematical; clearly the magnitudes involved in the freshman-se­ nior differentials are such that for the great majority of students the Catholic college appears simply to sustain a high-fertility norm if, indeed, it plays any active role at all. In any case, the differentials due to selectivity are consistently greater than those that can be inferred as the effect of the college experience, with the one exception of Catholic women from Catholic high schools enrolled in nonsectarian institutions.

The Nature

of SELEcnvnY

Not only do the data suggest that Catholic colleges are more often selected by high-fertility-oriented women but when the

(18: ( 8( (6: (82-

ON FAMILY-SIZE PREFERENCES

major factors related to family-size preferences (associations to be discussed in Chapters 8 and 11) are in turn related to the choice of a college, significant relationships appear. The data in Table 10, based on all Catholic freshmen in nonsectarian colleges and a 25 percent random sample of freshmen in Catholic colleges (to equalize the number of cases in each category and thus to maximize the variance of the dependent variable: the choice of a Catholic or nonsectarian college) show the correla­ tions between choice of a nonsectarian or Catholic college with type of secondary and elementary school attended, the impor­ tance of religion to parents and girl friends, various measures of religiousness, number of siblings, ethnic background, and the extent of Catholic education of the parents. The main result is that the initial correlation of 0.26 between choice of a Catholic college and family-size preferences drops to 0.07 when these other factors are held constant. This finding further corroborates the important role in the association TABLE 10 CoefiScients of Correlation for Catholic Freshmen1 between Selected Variables and Choice of a Nonsectarian or Catholic College* Number of children desired Type of secondary school Type of elementary school Importance of religion to self Importemce of religion to mother Importance of religion to father Importance of religion to girl friend Frequency attends mass Frequency receives communion Number of siblings Irish or French mother vs. all other nationalities Irish or French father vs. all other nationalities Extent of Catholic education of parents

.26 .46 .38 .14 .14 .17 .21 .37 .45 .05 .05 .04 .21

1 Calculations are based on a total of 1,671 Catholic freshmen equally divided between nonsectarian and Catholic colleges. 2 Nonsecteirian college is coded zero and Catholic college one, so that a positive correlation indicates an association of the particular variable with the probability of attending a Catholic college (with this probability for all Catholic freshmen being artificially set at 50 percent for purposes of this analysis).

11. THE EFFECTS OF HIGHER EDUCATION

between type of college and family-size preferences played by the selection of students on characteristics associated with both factors. Given this selectivity, there is always the possibility that the influence on family-size preferences of the college experience itself is masked or even augmented by the interplay of these factors of selectivity. The existence of a "suppressant" variable that would conceal predictive power or of another factor that is a more basic predictor can be determined through multiple-re­ gression analysis,13 so we calculated for the Catholic sample a series of Ilth order correlations between college class and num­ ber of children desired with these same selectivity factors partialed out. The results simply confirm our conclusions. For women from nonsectarian high schools in nonsectarian colleges the zero order correlation of —0.03 between college class and family-size preference dropped to 0.00; for women from nonsec­ tarian high schools in Catholic colleges the correlation changed from 0.00 to 0.03 and for women from Catholic high schools in Catholic colleges the correlation dropped from —0.05 to —0.02. The fact that no substantial differences develop rein­ forces our conclusion that the college experience had little or no influence on these groups. Qn the other hand, for women from Catholic high schools now enrolled in nonsectarian colleges, although the correlation between college class and number of children desired drops from —0.22 to —0.12, the influence of the college experience is not completely eliminated and hence cannot be explained completely by these factors of selectivity. SUMMARY

This chapter contains a detailed description and tests of the hypotheses connecting the number of children desired with higher education and religion. A total of 14 subgroups are identi­ fied, classified by religion, type of college, and, among Catholics, type of secondary and elementary schools. Comparisons of all of these groups support the various hypotheses involving the " Related analyses are also presented in Chapter 5 in more detail.

48

ON FAMILY-SIZE PREFERENCES

relation of religion, religiousness as inferred from type of school, and family-size preferences. The smallest families (an average of 3.2 children) are desired by women expressing no religious preference, followed by Jewish and Protestant women in nonsectarian institutions and Protestants in Protestant and Catholic schools. Mormon women, preferring an average of 4.7 children, lie between Catholic women in nonsectarian schools and those in Catholic schools. The largest families (5.6 children) are de­ sired by Catholic women whose education has been in Catholic schools. The hypotheses asserting various kinds of change in the number of children desired to be associated with education in different types of institutions (modified by prior education for Catholics) are mostly unsupported by the data. With few exceptions, the average number of children desired by seniors is slightly lower than are desired by freshmen, a difference that might easily arise from the fact that seniors are several years older and closer to the realities of marriage and childbearing than freshmen. Among Catholics, there is some evidence that higher education is associated with a reduction in the number of children desired, among women who graduated from Catholic high schools and are now enrolled in nonsectarian colleges and among women in Catholic coeducational colleges whose previous education was exclusively in nonsectarian schools, but the differ­ ences among Catholics in general appear to reflect much more the influence of other factors than that of the educational experience.14 14 Since this chapter was written, a replication of our study con­ ducted in Belgium (see Maurice Szykman, La dimension souhaitee de la famille, mimeographed report, 1965) concludes also that higher-fertilityoriented women are selected into Catholic institutions. No substantial differences in family-size preferences between freshmen and seniors were found except for a 0.6 decline among Catholic women in Catholic institutions. The author suggests however that this difference is probably the result of influences independent of the university.

CHAPTER 4

Higher Education and Family-Planning Intentions The ultimate number of children that a woman has depends not only on the size of family she prefers but also on her readiness and ability to practice some form of family limitation. Family planning consists in planning both the spacing and the number of children in the family. Other research1 on married couples has indicated that control over spacing has been extremely weak, partly because couples have only vague spacing preferences and partly because of such involuntary factors as pregnancy wastage. Although virtually all couples eventually practice some form of family limitation during their marriage,2 there is considerable variation in the time at which they begin. According to the 1960 national survey of fertility,3 only one out of every three married couples used some method of family planning between their marriage and first child. With the recent development of radically new methods of contraception, such as the oral pill and more especially the intrauterine plastic devices, along with all of the attendant publicity, the probabilities seem high that control of fertility will begin much earlier in marriage. Such a change would undoubtedly mean a longer interval between births in the early period of marriage and although completed family size would not be affected necessarily, the chances are that it probably would be reduced. It will be especially interest­ ing to see what changes will occur after the new intrauterine contraceptive devices become popular and the psychology of fertility is radically changed from the ordinary situation in which the decision-making involves repeated steps to prevent pregnancy 1

Westoff, Potter, and Sagi, The Third Child. 'Only 4 percent of fecund white couples in the U.S. as of 1960 had never used or did not expect to use some form of fertility regulation. See P. K. Whelpton, A. A. Campbell, and J. E. Patterson, Fertility and Family Planning in the United States (Princeton: Princeton Uni­ versity Press 1966), p. 184. ' Ibid., p. 223.

ON FAMILY-PLANNING INTENTIONS

to a situation in which pregnancy must be the result of a de­ liberate decision and action to eliminate the obstacle to conception. In view of these changes, as well as the more general sig­ nificance for fertility of attitudes toward family planning, we included several questions in the present study designed to permit evaluation of the "effects" of higher education on such attitudes, as well as to examine their interrelations with religion and other social variables. In this section we shall concentrate on the effects of the experience of higher education on intentions to plan fer­ tility by comparing the combined distributions of responses by freshmen and seniors to two questions, both providing "yes" and "no" response options: "Will you try to plan the spacing of your children?" and "Will you try to plan the number of children in your family?" The interpretation that Catholic respondents attached to the term "plan" in both of these questions can only be conjectured. The two questions appear early in the questionnaire (Questions 16 and 17) after a series of rather innocuous items on educa­ tional background and immediately following several questions on intentions to marry and to pursue a career, and on the num­ ber of children desired and expected. The battery of items on attitudes toward birth control does not appear until some twenty-five questions later. Nevertheless, it is certainly possible that some Catholic women interpreted these questions to imply the use of contraceptive methods not permitted by their church even though the two questions themselves do not contain any such referent. It is quite literally true that a "Yes" response to both questions would be quite consistent with the assumption that fertility would be regulated by the approved method of periodic continence and in fact not at all inconsistent with the formal teaching of most Catholic theologians. (Further discus­ sion of this position is contained in Chapter 6.) The responses to these two questions have been combined into a single distribution of four categories: (1) women who intend to plan both spacing and number; (2) those who intend

II. THE EFFECTS OF HIGHER EDUCATION

to plan spacing but not number; (3) women who say they will plan the number but not the spacing, and; (4) those who say they do not intend to plan either spacing or number. Our expectations about the distribution of responses of firstand fourth-year students and the relationships among religions, and across types of primary and secondary school education TABLE 11 Number of Children Desired by Women, by Fertility-Pliinning Intentions and Religion Fertility-Planning Intentions

Number of Children Desired Percent 2 or 3 4 5 6 or Total

fewer

Mean

Number

more

NO RELIGIOUS PREFERENCE

Intend to plan:

Number and spacing Number only Spacing only Neither

34 32 22 22 28 28

7 5 6 16

100 100

3.1 3.9

28 22 14 22 14

100

4.1

355 18 3 14

100 100 100 100

3.4 3.7 4.3 4.4

5,640 170 256 184

100 100 100 100

3.3 3.9 3.7 3.9

776 21 24 17

100

4.4

*

PROTESTANTS

Intend to plan:

Number and spacing Number only Spacing only Neither

19 18 6 12

35 32 24 18

34 6 6 35 6 9 41 12 17 29 17 24 JEWS

Intend to plan:

Number and spacing Number only Spacing only Neither

20 41 30 5 4 14 33 24 14 15 — 46 37 17 — 19 41 30 6 4 MORMONS

Intend to plan:

Number and spacing Number only Spacing only Neither

6 21 39 10 24

100 100

6.0 6.3

319 6 33 22

100 100 100 100

4.5 4.6 5.6 6.0

2,616 257 1,447 1,986

•

3 3 21 15 58 5 18 36 12 29 CATHOLICS

Intend to plan:

Number and spacing Number only Spacing only Neither

8 17 36 16 23 7 17 33 17 26 1 6 26 20 47 2 5 22 16 55

* Not enough cases to compute statistics.

52

ON FAMILY-PLANNING INTENTIONS

of Catholic women, follow exactly the same logical structure of the hypotheses about the number of children desired. There is the empirical question, however, of the interdependence of the two variables. The association of desired family size with fertility-planning intentions (Table 11) is in the expected direc­ tion (smaller families preferred by women intending to plan fertility), but the fact that the vast majority of non-Catholics in­ tend to plan but nevertheless vary considerably in their family-size preferences precludes inferring one variable from the other. Even among Catholic women, who differ much more in attitude toTABLE 12 Intentions to Plan Spacing and Number of Children of Women with No Religious Preference, Protestiint Women, Jewish Women, and Mormon Women, by Year and Type of Institution Ugton of man

>ne

Dtestant

vish

otestant

rtestant

ormon

Type oj Institution and Year

Plan Spacing and Number Percent

Plan Number Not Spacing Percent

Plan Spacing Not Number Percent

Plan Neither Percent

Percent Total

Nonsectarian 1 4

89 92

5 4

1 1

5 2

100 100

209 158

Nonsectarian 1 4

88 95

3 1

5 2

4 1

100 100

3,619 1,758

Nonsectarian 1 4

91 95

3 2

4 1

2 2

100 100

518 323

Protestant 1 4

87 94

4 3

5 1

4 2

100 100

601 196

Catholic 1 4

82 85

3 5

9 5

6 5

100 100

77 20

Nonsectarian 1 4

79 91

3 —

11 5

7 4

100 100

226 157

Number

II. THE EFFECTS OF HIGHER EDUCATION

ward fertility planning, the distributions of the number of chil­ dren desired within planning categories are highly variable. The relationships between higher education and fertility-plan­ ning intentions are (Tables 12 and 13) consistent by and large with the generalizations about the number of children desired. The great majority—from 79 to 95 percent—of non-Catholic women (Table 12) say they intend to plan both the spacing TABLE 13 Intentions of Catholic Women to Plan Spacing and Number of Children, by Year and T of Institution, Subdivided by Type of High School and Elementary School Elementary School

High School

College

Year

Plan Spacing and Number Percent

NS

NS

NS

1 4

62 66

C

NS

NS

1 4

NS

C

NS

C

C

NS

Plan Number not Spacing Percent

Plan Spacing not Number Percent

Plan Neither Percent

Percent Total

8 4

11 10

19 19

100 100

4 2

59 63

3 6

14 8

24 23

100 100

1

1 4

35 56

8 13

15 6

42 25

100 100

NS

1 4

38 62

4 2

19 19

39 18

100 100

2 1

NS

C

1 4

45 47

6 5

21 25

28 24

100 100

4 2

C

NS

C

1 4

40 40

4 4

25 27

32 29

100 100

3 1

NS

C

C

1 4

39 52

6 1

22 21

33 26

100 100

1

C

C

C

1 4

31 42

4 3

25 28

40 28

100 100

23 13

NS = Nonsectarian. C = Catholic.

Nui

ON FAMILY-PLANNING INTENTIONS

and the number of children in their family. The proportion intending not to plan either spacing or number never exceeds 7 percent. The proportions of Catholic women (Table 13) intending to plan their families completely are much lower regardless of the type of schooling they have had or the type of college they currently attend. The proportions range from a high of 66 per­ cent (seniors in nonsectarian colleges whose past education was exclusively secular) to 31 percent (freshmen in Catholic colleges whose past education was exclusively Catholic). As expected, the proportion of those who say they will plan neither is consider­ ably higher among Catholics than among non-Catholics, ranging roughly from 20 to 40 percent. Some of the motivations involved are revealing. The following excerpts were written in short essays by the women projecting their future life. Case 13165, Catholic: I do not believe in birth control . . . I think a large family with cooperation among members would be better than the strained relationship I believe birth control would foster between mother and father. Case 54032, No Religion: Children deserve educated parents. They deserve attention. They deserve love. For these rea­ sons I will limit the size of my family, an adult cannot raise children properly if he cannot devote time to them. Case 54069, Catholic: I would like a large family (mostly boys) and I think it unnatural to 'plan' the size of a family. Case 79076, Mormon: I don't plan on having children until my husband and I are well established and have some­ thing to offer our children. Case 79167, Catholic: I am a pre-med major so I hope to be a doctor by then . . . I shall plan my family because with a larger number of children, it's very difficult to give each child enough attention and proper guidance. Case 91018, Protestant: In my opinion people who have

II. THE EFFECTS OF HIGHER EDUCATION

large families and do not plan are either selfish or ig­ norant. The population explosion is fast becoming a seri­ ous problem and people should look ahead and think of future generations. An interesting difference between Catholics and non-Catholics emerges when we compare the ratio of responses in the two intermediate categories—those who reply that they intend to plan spacing but not number, and those who intend to control the number but not the spacing. Among non-Catholics (except Mormons), the two responses are about equally divided on the average, but among Catholics, women who intend to plan spac­ ing but not number are considerably more numerous than women subscribing to the obverse pattern. The difference is most striking among women with Catholic education. The ex­ planation lies presumably in an interpretation of their religion as sanctioning the spacing of births to protect the mothers' health, for example, but as being opposed or less favorable to the conscious control of the number of children since this would appear to be more akin to an objectionable negative ideology of contraception. This explanation ought not be pushed too far, however, since an average of about half of all Catholic women favor controlling both the spacing and number of their children. THE INFLUENCE OF HIGHER EDUCATION

What can we infer from these tabulations about the influence on plans for fertility control of four years' experience in institu­ tions of higher education? Does secular education through both its formal and its informal processes increase the proportion of women who intend to control fertility? Are Catholics more affected than others? Does Catholic education depress this proportion or sustain it at a low level? How much of the differ­ ences among Catholic women can be attributed to the "effects" of different types of education compared with the importance

ON FAMILY-PLANNING INTENTIONS

of selective recruitment of women from backgrounds of varying religiousness? In virtually every comparison, the proportion replying that they intended to plan both spacing and number is higher among fourth-year than among first-year women. For the most part the differences are slight, ranging from 3 to 12 percent—a differ­ ence that might easily be due simply to the increase in age and maturity and the greater imminence of marriage and childbearing for women close to college graduation.4 There is one instance where the magnitude of the differentials supports the possibility that the college might have some secu­ larizing effect, although the inference is by no means unambigu­ ous. Catholic women who have attended Catholic high schools begin college with only about a third feeling that they will plan their families. Those who go on to nonsectarian colleges show about the same proportion in this planning category as those who attend a Catholic college, which discourages the explanation of obvious selectivity. The difference between freshmen and se­ niors in the nonsectarian schools is well over 20 percent. Compar­ ing the largest two groups—those who attended Catholic ele­ mentary and high schools—the proportion of women enrolled in nonsectarian colleges who intend to plan number and spacing "increased" from 38 to 62 percent and among those attending Catholic colleges the proportion changed from 31 to 42 percent. The differences in the proportion reporting that they intend to plan neither spacing nor number are consistent with the differ­ ences in the category of planners—from 39 to 18 percent for women in nonsectarian colleges and from 40 to 28 for girls in Catholic colleges. On the other hand, selectivity does seem to be the main factor if we compare the graduates of nonsectarian high schools who are now in nonsectarian or Catholic colleges. The propor­ tion of prospective planners in the first group is about 60 percent 4Another

possible explanation might be a selective attrition of women who are not oriented toward planning in general.

57

II. THE EFFECTS OF HIGHER EDUCATION

in the first year and rises to about 65 percent in the fourth year while in Catholic colleges the corresponding proportions are 43 and 45 percent. It is quite clear that the differences reflect initial values more than any effects of the college experience. Cross-classification of women in Catholic coeducational or women's colleges affects this pattern only slightly. However, some mention should be made of the finding that among students with an exclusively nonsectarian education virtually no difference (actually 2 percent, which is not statistically significant) exists between freshmen and seniors in the women's colleges but a 12 percent "increase" is evidenced in coeducational institutions. Since the freshmen differential in both types of institutions is only 5 percent, selectivity may not be the only explanation. This finding parallels that in the chapter on family-size preference. CONCLUSIONS

The overall generalizations about the effects of higher educa­ tion on attitudes toward family planning that emerge are as follows: 1. Protestant, Jewish, and Mormon women, and students with no religious preference all experience an increase in the propor­ tions intending to plan their families, but the differences for the most part are slight and could easily be due to factors other than the college experience. 2. Although Catholic women also become more inclined to plan, there is a considerable amount of variation by types of schools attended. There does appear to be some secularizing effect of a nonsec­ tarian education on the graduates of Catholic high schools, al­ though one could argue that women who elect not to continue in Catholic schools may have vulnerabilities of a delayed nature that are not so apparent in the first year. Whatever the mecha­ nism, Catholic high school graduates who attend nonsectarian universities show a sharp increase in family-planning intentions so as to become indistinguishable from their Catholic classmates

ON FAMILY-PLANNING INTENTIONS

who attended nonsectarian high schools. Catholic high school graduates who attend a Catholic college also become more ori­ ented toward family planning but to a much lesser extent, and the proportion intending to plan both spacing and number of children by graduation is considerably lower than that among their high school classmates who went to nonsectarian colleges. 3. The entire difference between Catholic graduates from nonsectarian high schools who attend Catholic colleges and those who attend nonsectarian colleges is selectivity, evidenced by sharp initial differences in the first year.

CHAPTER 5

Higher Education and Fertility Values Within Selected Controls The rather elaborate analyses described in the following pages were undertaken to spell out in greater detail summary analyses presented in Chapter 3 and to increase our confidence in the reliability of our generalizations about the "effect" of higher education on fertility values by imposing a considerable number of statistical controls on the comparisons of first- and fourth-year students in the form of socioeconomic and religious subdivisions. The discussion and tables included relate to both the number of children desired and family-planning intentions. Although the latter dimension (in order to save space) is indexed only by the proportions of women replying that they intend to plan both spacing and number of children, the generalizations would not differ if the entire distribution were presented. COMPARISONS WITH SOCIOECONOMIC CONTROLS

Before concluding finally that for most women the college exerts little or no influence on the average number of children desired, it is desirable to compare the fertility values of firstand fourth-year students within certain controls. Several consid­ erations argue for imposing socioeconomic controls. There is some association between social class background and the colleges that young women attend and we want to be absolutely certain that our inferences about the "effects" of different types of insti­ tutions are not confused with differential class compositions. Perhaps more directly pertinent is the problem of attrition and its possible implications for the comparison of first- and fourthyear students.1 We know that significant fractions of women entering college never graduate from that college, either because they leave school altogether or because they transfer to another institution. To the extent that such attrition is selective on fac­ tors associated with fertility values, any inference about change in 1

See Chapter 1 for a fuller discussion of this methodological problem.

60

II. THE EFFECTS OF HIGHER EDUCATION

this variable over four years is correspondingly adulterated. One partial solution to this problem is to compare the responses of women who are least likely to leave school before graduation, such as the daughters of women who themselves were college graduates. The present focus of interest is not on the relation of socioeco­ nomic factors to fertility attitudes—a subject to which we shall return later—but rather the "effects" of the college experience on the fertility values of women drawn from more homogeneous (and therefore more comparable) backgrounds. Not all of the categories of religion and types of college have a large enough representation in the sample to permit the neces­ sary further subdivisions, so we are excluding Protestants in Catholic colleges, and ignoring the subdivision of Catholic women according to the type of elementary school attended and the type of Catholic college they are now attending. CONTROL ON EDUCATION OF PARENTS

The variations of family-size preferences and family-planning intentions by class, institution, and religion follow essentially the same pattern within two selected homogeneous educational categories of parents (high school graduates and college grad­ uates) as that observed for the total sample (Tables 14 and 15). In the non-Catholic samples, there are only slight differ­ ences between the classes and most of these are in the direction of preferences for smaller families and slight increases in propor­ tions of seniors intending to plan both the number and spacing of children. And, among Catholics, the general pattern observed for all Catholic women is also repeated. As noted earlier, the major deviation from the selectivity hypothesis occurs among Catholic women who attended a Catholic high school and moved on to a nonsectarian college, where there does appear to be a persistent difference between first- and fourth-year students. Although it is clear from the higher average family-size prefer­ ences of the all-Catholic group that some selectivity is involved when Catholic high school students elect nonsectarian rather

ON FERTILITY VALUES WITHIN HOMOGENEOUS GROUPS

TABLE 14 Number of Children Desired, by Religion, Type of Institution, and Year, by Women Whose Parents Were High School or College Graduates Education of Mother

Education of Fathe

Type of Institution

Year

None

Nonsectarian

1 4

3.2 3.1

3.1 3.2

3.0 2.9

(3.4) 3.0

2.9 2.9

Protestant

Nonsectarian

1 4

3.5 3.4

3.5 3.3

3.6 3.5

3.5 3.4

3.6 3.4

Jewish

Nonsectarian

1 4

3.4 3.2

3.4 3.2

3.3 3.4

3.5 3.2

3.4 3.0

Protestant

Protestant

1 4

3.8 3.4

4.0 3.6

3.8 3.3

3.7 3.3

3.6 3.6

Mormon

Nonsectarian

1 4

4.6 4.8

4.7 4.9

4.2 5.3

4.6 5.2

4.1 4.8

Religion of Woman

All Women

HIGH HIGH COLLEG COLLEGE SCHOOL GRADUATE SCHOOL GRASUA GRADUATE GRADUATE

HIGH SCHOOL

COLLEGE

Catholic

NS

NS

1 4

4.2 4.1

4.2 4.2

4.3 4.0

4.3 4.0

4.3 4.0

Catholic

C

NS

1 4

5.2 4.3

5.2 4.3

4.8 (4.5)

5.2 4.4

4.9 (4.1)

Catholic

NS

C

1 4

4.9 4.9

4.9 4.8

5.2 5.4

5.1 4.7

4.9 5.6

Catholic

C

C

1 4

5.7 5.4

5.6 5.5

5.9 5.7

5.5 5.2

5.9 5.4

( ) Mean based upon fewer than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

than Catholic colleges, there is nevertheless a considerable "de­ cline" in the average size of family preferred by those who attend nonsectarian colleges, while the size of the large families desired by the women with exclusively Catholic education tends to drop off only slightly. The same pattern holds for family-planning

11. THE EFFECTS OF HIGHER EDUCATION

intentions, in which there is very little evidence of selectivity. There is a sharp "increase" among Catholic high school students graduating from nonsectarian colleges while the "increase" among such graduates in Catholic colleges is much smaller. It is interesting to note that the selection of women moving TABLE 15 Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by Education of Parents rligion of 'oman

Type of Institution

Year

one

Nonsectarian

1 4

89 92

87 92

92 97

81 93

97 88

rotestant

Nonsectarian

1 4

88 95

88 95

89 95

87 96

89 96

:wish

Nonsectarian

1 4

91 95

93 95

90 93

91 99

93 94

rotestant

Protestant

1 4

87 94

86 88

84 93

91 94

89 93

Iormon

Nonsectarian

1 4

79 91

78 96

77 82

80 97

86 85

All Women

Education of Mother

Education of Father

HIGH HIGH COLLEGE COLLEGE SCHOOL GRADUATE SCHOOL GRADUATE GRADUATE GRADUATE

HIGH SCHOOL

COLLEGE

Iatholic

NS

NS

1 4

61 65

63 71

60 63

63 62

70 78

!atholic

C

NS

1 4

38 61

35 62

29 (55)

27 63

46 (64)

!atholic

NS

C

1 4

43 45

44 45

39 41

40 38

45 45

!atholic

C

C

1

31 43

34 43

25 39

32 42

30 46

4

( ) Proportion based upon fewer than 20 but at least 10 cases. NS = Nonsectarian C = Catholic.

ON FERTILITY VALUES WITHIN HOMOGENEOUS GROUPS

from a Catholic high school to a nonsectarian college apparently has the same effect on the level of family-size preferences as the obverse combination, that is, the selection of women moving from non-Catholic high schools to Catholic institutions of higher education. Both groups begin college wanting about 5 children on the average, but those in nonsectarian schools graduate with an average level about 20 percent lower while those in Catholic colleges remain the same. In fact, among daughters of parents who were college graduates and who went to a Catholic college from a non-Catholic high school, there is some indication of an increase over the four years. CONTROL ON OCCUPATIONAL CLASS OF FATHER

Comparison of the fertility values of women of different re­ ligions in different types of educational institutions is presented in Tables 16 and 17, tabulated by the major occupational class of the girl's father. Not all occupational categories are repre­ sented in the sample in sufficient number to permit detailed subdivision and some combinations (clerical with sales and skilled with semiskilled) were formed to increase statistical stability. In brief, the data reveal that social class background as re­ flected by the occupational class of the student's father seems to have no bearing on the difference between first- and fourthyear students either in the size of families they prefer or in their intentions to plan family size. CONTROL ON NATIONAUTY BACKGROUND

The nationality background of married couples, especially that of Catholics, has been catalogued as a social factor differentiating fertility and fertility planning mainly through its association with the number of children desired. Persons of Irish descent represent the high-fertility pattern with Italian background being associated with low fertility. Since there is a strong relationship between nationality and affinity with the Catholic educational system, an analysis of the influence of higher education on fam­ ily-size values should ensure comparability across types of institu-

II. THE EFFECTS OF HIGHER EDUCATION

TABLE 16 Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Occupational Class of Father Religion of Woman

Type of Institution

Year

All Women

Prof.

Mgr. Clerical Skilled Frop. and and Of. Semi­ Sales skilled

None

Nonsectarian

1 4

3.2 3.1

3 .2 3 .4

2.7 2.9

3.2 3.1

3.2 2.7

Protestant

Nonsectarian

1 4

3.5 3.4

3 .5 3..4

3.5 3.4

3.6 3.3

3.6 3.3

Jewish

Nonsectarian

1 4

3.4 3.2

3 .5 3 .2

3.3 3.2

3.4 3.2

3.1 3.5

Pirotestant

Protestant

1 4

3.8 3.4

3..7 3,.4

3.9 3.5

3.6 3.5

3.6 3.4

Mormon

Nonsectarian

1 4

4.6 4.8

4..6 5..1

4.5 4.7

4.6 4.5

4.7 4.7

HIGH COLLEGE SCHOOL

Catholic

NS

NS

1 4

4.2 4.1

4. 5 4..1

3.9 3.9

4.6 4.3

4.2 4.0

Catholic

C

NS

1 4

5.2 4.3

5. 1 4.2

4.8 4.5

5.8 4.7

5.3 4.0

Catholic

NS

C

1 4

4.9 4.9

5. O 5..1

5.1 4.9

4.6 4.6

5.1 5.1

Catholic

C

C

1 4

5.7 5.4

5.,9 5,.6

5.7 5.4

5.6 5.6

5.6 5.3

NS = Nonsectarian. C = Catholic.

tions. Although our evidence about the connections of nationality and fertility pertain mainly to Catholics, all religious groups are presented in Tables 18 to 21. In general, the evidence from calculations by the nationality background of both mother and father suggests no basic modifi-

ON FERTILITY VALUES WITHIN HOMOGENEOUS GROUPS

TABLE 17 Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by Occupational Class of Father Type of Institution

Year

All Women

Prof.

None

Nonsectarian

1 4

89 92

89 97

93 90

86 96

90 90

Protestant

Nonsectarian

1 4

88 95

89 97

89 96

87 97

86 93

Jewish

Nonsectarian

1 4

91 95

90 97

95 94

83 97

93 96

Protestant

Protestant

1 4

87 94

87 96

84 97

86 91

89 93

Mormon

Nonsectarian

1 4

79 91

76 79

75 95

88 (100)

82 (100)

Religion of Woman

Mgr. Clerical Skilled Prop. and and Off. Sales Semi­ skilled

HIGH COLLEGE SCHOOL

Catholic

NS

NS

1 4

61 65

59 67

64 69

65 60

60 64

Catholic

C

NS

1 4

38 61

44 64

42 61

(28) 75

37 58

Catholic

NS

C

1 4

43 45

41 45

42 41

45 57

47 46

Catholic

C

C

1 4

31 43

32 42

28 45

35 44

33 42

( ) Proportion based upon fewer than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

Nonsectarian

Protestent

Nonsectarian

Jewish

Protestant

Mormon

C

C

NS

C

Catholic

Catholic

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

1 4 1 4 1 4 1 4 1 4

1 4 1 4 1 4 1 4

All Women

Year

4.2 4.8 5.2 (4.3) 5.0 4.7 5.6 5.2

4.5 4.2 5.3 4.4 5.2 5.1 6.0 5.8

*

*

*

*

*

3.7 3.5 4.6 4.8

*

*

(2.8) (3.2) 3.5 3.6

Irish

*

2.8 3.3 3.6 3.3

British

4.3 3.8 4.8 (4.4) 5.1 5.1 5.7 5.3

2.7 2.9 3.5 3.4 3.2 3.1 3.7 3.3 (4.7) (5.1)

German

Polish, Russian, other Slavic, and Hungarian. * Less than 10 cases. Mean not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

1 Includes

NS

C

Catholic

NS

NS

Catholic

COLLEGE

Nonsectaritin

Protestant

HIGH SCHOOL

Nonsectarian

Type of Institution

None

Religim of Woman

4.5 (5.1)

*

*

•

*

(4.6) (4.3)

3.9 3.5 4.0 4.8

*

*

(2.8) (2.9) 3.7 3.4

Scandi­ navian

5.9 (5.2) 6.3 5.4

*

*

3.7 4.0

*

*

4.1 4.0 5.5 4.0 4.4 4.4 5.3 5.1

*

*

*

*

*

*

*

*

*

3.3 3.4

*

*

•

* *

*

French

•

Italian

4.1 4.0 4.6 3.9 4.5 4.8 5.1 5.1

*

*

*

*

3.8 2.9 3.9 3.8 3.4 3.3

Eastern European1

TABLE 18 Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Nationality Background of Mother

1 4 1 4 1 4 1 4 1 4

Nonsectarian

Nonsectarian

Nonsectarian

Protestant

Nonsectarian

HIOH COLLEGE SCHOOL NS NS

None

Protestant

Jewish

Protestant

Mormon

Catholic

C

C

NS

C

Catholic

Catholic

4.5 4.1 5.3 (4.6) 5.0 5.2 5.7 5.6

2.8 3.2 3.6 3.4 (3.6) (3.1) 3.7 3.4 4.9 5.0

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

British

All Women

4.4 4.3 5.3 4.5 5.3 5.3 6.2 5.7

*

*

*

*

*

*

2.8 (2.6) 3.5 3.3

Irish

4.4 4.1 5.0 (4.7) 5.3 4.9 5.6 5.4

3.4 3.0 3.5 3.3 3.3 3.1 3.5 3.4 4.6 4.5

German

Polish, Russian, other Slavic and Hungarian. * Less than 10 cases. Mean not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian. Π = flatlinltr.

1 Includes

NS

C

Catholic

1 4 1 4 1 4 1 4

Year

Religion of Woman

Type of Institution

*

5.1 (4.7)

*

*

*

•

*

*

4.8 (4.5) 5.8 5.3

*

*

4.1 (4.4)

4.0 3.7 4.9 (3.9) 4.6 4.6 5.1 5.1

*

3.8 4.2 5.3 4.0 4.5 4.4 5.4 5.0

*

*

•

*

*

*

*

3.5 3.1 3.3 3.7 3.4 3.3

Eastern European1

*

*

*

*

3.6 3.5

*

*

4.0 3.5 4.2 4.8

*

*

3.2 (3.3)

*

•

*

French

*

Italian

•

(2.8) (2.9) 3.6 3.7

Scandi­ navian

TABLE 19 Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Nationality Background of Father

1 4 1 4 1 4 1 4 1 4

Nonsectarian

Nonsectarian

Nonsectarian

Protestant

Nonsectarian

None

Protestant

Jewish

Protestant

Mormon

NS

NS

C

C

NS

C

NS

C

Catholic

Catholic

Catholic

Catholic

4

1

4

4 1

4 1

61 65 38 61 43 45 31 43

89 92 88 95 91 95 87 94 79 91

All Women

71 65 41 (50) 42 47 38 44

93 95 88 96 100 (100) 92 91 75 91

British

57 59 33 60 36 39 25 36

*

*

*

*

*

86 (94) 88 94 •

Irish

60 63 40 (64) 42 37 31 44

86 97 89 94 90 93 90 95 (81) (100)

German

Polish, Rumanian, Russian, other Slavic, and Hungarian. * Less than 10 cases. Mean not computed if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

1 Includes

COLLEGE

HIGH SCHOOL

1

Year

Religion of Woman

Type of Institution

28 (53)

*

*

*

*

(73) (69)

84 96 87 97

•

*

(90) (90) 82 94

Scandi­ navian

61 64 42 61 54 60 39 43

*

*

*

*

*

*

*

38 (58) 35 51

*

*

59 59 36 (67) 45 39 21 49

*

61 71

*

*

*

*

91 91 91 96

84 88

Eastern European1

•

*

*

•

•

92 100

• *

*

French

*

Italian

I ABLE 20 Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by Nationality Background of Mother

Nonsectarian

Protestant

Nonsectarian

Jewish

Protestant

Mormon

C

NS

C

Catholic

Catholic

Catholic

C

C

NS

NS

1 4 1 4 1 4 1 4

1 4 1 4 1 4 1 4 1 4

Year

61 65 38 61 43 45 31 43

89 92 88 95 91 95 87 94 79 91

All Women

55 62 46 53 45 38 29 35

93 90 89 96 (100) (91) 91 97 75 91

British

56 52 38 51 43 46 28 36

*

*

92 (100)

*

*

95 (100) 85 98

Irish

66 71 42 63 33 37 29 49

90 95 88 95 84 98 94 84 79 100

German

Polish, Rumanian, Russian, other Slavic, and Hungarian. * Less than 10 cases. Mean not computed if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian.

1 Includes

NS

Catholic

COLLEGE

Nonsectarian

Protestant

HIGH SCHOOL

Nonsectarian

Type of Institution

None

Religion of Woman

*

41 (58)

*

*

*

*

*

59 64 37 61 52 60 39 46

(41) (59) 34 47

*

*

58 (47)

*

*

71 71 37 (100) 41 35 34 49

*

*

*

* *

*

*

* •

91 86 89 94 91 95

Eastern European1

*

*

80 99 81 88

•

* •

*

86 87

*

90 (92)

•

*

French

•

Italian

*

(92) (100) 91 95

Scandi­ navian

TABLE 21 Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by Nationality Background of Father

II. THE EFFECTS OF HIGHER EDUCATION

cation of the generalizations already developed about the associa­ tion of higher education and family-size values. Although there are some apparently significant and expected variations by na­ tionality in the level of family-size preferences, particularly among Catholics, the pattern of variations is independent of the student's year in college, thus implying that the differences between the classes are unaffected. The main difference apparent between first- and fourth-year students is among those expressing no religious preference, where there are several reversals from the overall picture of slight de­ cline. Whether these differences reflect more than simply chance variation is difficult to assess. CONTROLS ON RELIGIOUSNESS

Perhaps the most important factor that should be examined in connection with the "changes" in family-size values during the college years is the religiousness of the student and her par­ ents. We have ample evidence, especially for Catholics, that religiousness is associated both with fertility values and type of school attended.2 There are three references to be considered: the girl's parents, her friends, and herself. Parents

Does it make any difference in the vulnerability of a girl's family-size preferences to higher education if her mother or father prefers a religion different from her own?3 The religious­ ness of her parents, although admittedly difficult to assess re­ liably, especially since our estimates are filtered through the views of the daughter, is especially critical for our hypotheses 2 For the influence of religiousness on the choice of college see A. Greeley, Religion and Career (New York: Sheed & Ward, 1963), p. 178, and A. Greeley and P. Rossi, The Education of Catholic Americans (Chicago: Aldine, 1966). 8 The religion of the mother and that of the father is compared separately on the basis of the student's religious preference. In almost all instances involving Protestants, Jews, and Catholics, if one parent holds a religion different from that of the student, the other parent will have the same religion as the daughter.

ON FERTILITY VALUES WITHIN HOMOGENEOUS GROUPS

about the effects of higher education on family-size values. As frequently reiterated in discussions of our statistical analyses, particularly in regard to Catholic women, we are concerned continually with the comparative significance of the selection of certain types of women into particular institutions versus the effects of the four years' experience in the institution. One of the obvious dimensions of selection into a nonsectarian or a Catholic college is the religiousness of the student's home en­ vironment, one of the best indices of which is the religiousness of her parents. Holding constant some estimate of this variable, as crude as it may be, over and above the type of high school attended, will contribute greatly to an evaluation of the com­ parative influence of selectivity and the educational experience. Keeping constant the religion of the parents and then com­ paring freshmen and senior women's family-size values by re­ ligion and type of institution (Tables 22-25) does not alter the pattern at all. One by-product of this analysis, however, is the apparent "increase" in family-size preferences among Catholic graduates of nonsectarian high schools who attend Catholic col­ leges and whose fathers either are Protestant or have no religious preference. This is probably a product of association with Catho­ lic women more than an effect of the formal educational experi­ ence itself. One dimension of relevance for fertility values is the religion and level of religiousness of the friends with whom the girl is in close contact during her college years. Again our focus is still limited to refining the test of the "effects" of higher education on family-size values, which we assume may operate differently if supported or impeded by reference groups of vary­ ing compatibility. There is, of course, no way of inferring that these groups or, more importantly, their religious characteristics remain constant over the four years and probably a considerable amount of change in particular individuals and some change in religious characteristics does occur, but despite such complex­ ity there is some gain in the use of such controls. The religion of either the student's best girl friend or most

Nonsectarian

Protestant

Nonsectarian

Jewish

Protestant

Mormon1

C

NS

C

Catholic

Catholic

Catholic

C

C

NS

NS

1 4 1 4 1 4 1 4

4 1 4

1

4

1

1 4 1 4

Year

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

All Women

*

5.1 5.0 5.4 5.2

•

*

*

*

*

*

4.4 4.4

3.8 3.4 (3.6) 4.8

*

*

3.1 3.1 3.5 3.4

*

*

*

*

*

*

*

4.2 4.0 5.2 4.4 4.9 4.9 5.7 5.4

4.4 (5.0) 5.9 5.3

*

*

4.2 4.5

*

•

*

*

*

*

*

*

*

*

*

*

*

*

2.8 3.1 3.6 3.3

(3.6) 3.1 3.5 3.4

•

*

•

3.4 3.2

*

•

3.5 3.1

4.7 5.2 5.6 5.3

*

*

4.5 4.2

3.8 3.4 3.9 4.9

*

•

3.3 3.0 3.5 3.4

*

*

*

*

*

*

*

*

*

*

*

*

3.4 3.2

*

*

3.3 3.0

Religion of Father PROTES­ JEWISH TANT

NONE

PROTES­ JEWISH TANT CATHO­ LIC

Religion of Mother

*

*

*

*

*

*

*

*

*

2.7 2.9 3.6 (3.8)

NONE

* Less than 10 cases. Meeui not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. 1 The means for women with Mormon parents are the same as for the "All Women" category. NS = Nonsectarian. C = Catholic.

NS

Catholic

COLLEGE

Nonsectarian

Protestant

HIGH SCHOOL

Nonsectarian

Type of Institution

None

Religion of Woman

XACJUCI ZZ

Number of Children Desired by Women, by Religion, Type of Institution, and Year, by Religion of Parents

4.1 4.0 5.2 4.3 5.0 4.9 5.7 5.5

*

*

*

*

*

*

(2.7) 3.2 3.5 3.8

CATHO­ LIC

Nonsectarian

Protestant

Nonsectarian

Jewish

Protestant

Mormon

NS

C

C

C

NS

C

Catholic

Catholic

Catholic

4

1

4 1 4

1

61 65 38 61 43 45 31 43

89 92 88 95 91 95 87 94 79 91

All Women

*

*

*

*

*

*

*

*

*

*

*

43 (36) 37 46

*

*

68 67

88 95 (80) (90)

*

* *

*

89 93 88 95

*

*

•

*

*

*

*

*

*

*

*

*

91 95

*

*

91 91

60 64 37 59 42 45 31 42

*

*

*

*

*

*

(84) 93 72 92

41 (62) 37 42

*

*

58 55

*

*

*

*

*

*

91 90 88 98

56 54 30 58

*

*

65 78

88 95 86 80

*

*

89 93 88 95

*

*

*

*

*

*

*

64 35 61 41 45 32 42

61

*

*

*

»

*

*

*

*

85 94 87 91

CATHO­ LIC

*

*

*

91 95

*

*

91 91

Religion of Father NONE

PROTES­ JEWISH TANT

CATHO­ LIC

Religion of Mother PROTES­ JEWISH TANT

*

94 91 89 (100)

NONE

* Less than 10 cases. Proportion not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian.

NS

NS

Catholic

1 4

Nonsectarian

Protestant

COLLEGE

1 4 1 4 1 4 1 4 1 4

Nonsectarian

None

HIGH SCHOOL

Year

Religion of Woman

Type of Institution

TABLE 23 Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by Religion of Parents

Nonsectariein

Protestant

Nonsectarian

Jewish

Protestant

Mormon

NS

C

C

C

NS

C

Catholic

Catholic

Catholic

1 4 1 4 1 4 1 4

4 1 4 1 4 1 4 1 4

1

Year

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

4.4 4.1 5.4 4.4 5.2 5.1 5.9 5.6

3.6 3.1 3.5 3.4 3.6 3.3 3.9 3.4 5.0 5.3

MUCH

All Women VERY

4.1 4.0 5.2 4.3 4.7 4.9 5.4 5.3

3.5 2.6 3.6 3.4 3.5 3.4 3.6 3.4 4.0 4.6

3.9 4.1 4.8 4.1 4.5 4.5 5.4 5.1 *

*

*

*

5.3 5.2

•

*

* *

*

*

*

*

*

(3.9) (4.3)

* *

(4.1)

* *

(3.6) (3.3)

*

*

2.6 (3.1)

*

2.6 3.2 3.4 3.5 3.3 2.9

3.1 3.1 3.5 3.4 3.2 3.2 3.5 (3.0) 3.8 4.6

NONE

*

VERY LITTLE

SOME

Importance to Mother MUCH

* Less than 10 cases. Mean not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

NS

NS

Catholic

COLLEGE

Nonsectarian

Protestant

HIGH SCHOOL

Nonsectarian

Type of Institution

None

Religion of Woman

4.2 4.3 5.4 4.4 5.3 5.1 6.0 5.8

3.3 3.7 3.5 3.4 3.5 3.4 3.9 3.4 5.1 5.5

VERY MUCH

4.3 3.9 5.3 4.3 4.8 5.0 5.5 5.4

3.2 2.8 3.5 3.3 3.3 3.4 3.6 3.5 4.1 5.0

MUCH

4.2 4.0 4.6 4.1 4.9 4.9 5.5 5.1

3.4 3.0 3.5 3.4 3.3 3.2 3.9 3.3 4.2 4.1

SOME

4.7 4.9 5.4 4.8

*

*

3.9 4.1

2.9 3.2 3.5 3.4 3.3 3.0 3.6 (3.1) 3.8 4.3

VERY LITTLE

Importance to Father

Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Woman's Perception of the Importance of Religion to Her Parents

4.0 (3.8) 5.8 5.5

*

•

4.7 (4.4)

•

(4.2)

•

*

2.8 3.0 3.6 3.3 3.7 (2.9)

NONE

Nonsectarian

Protestant

Nonsectarian

Jewish

Protestant

Mormon

NS

C

C

C

NS

C

Catholic

Catholic

Catholic

57 68 36 58 36 38 28 40

94 97 89 95 84 93 89 97 74 92

VERY MUCH

31 63 49 51 37 43

61 60

86 91 89 95 92 99 86 94 88 89

70 68 54 (56) 46 55 37 52

87 88 88 96 94 92 83 (75) 81 92

SOME

*

*

*

41 57

*

*

•

* * *

*

*

*

*

•

(90)

(75) (71)

*

*

*

92 (93) 85 (90) 88 (100)

NONE

*

93 97 86 96 88 97

VERY LITTLE

Importance to Mother

MUCH

* Less than 10 cases. Proportion not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian.

NS

NS

Catholic

61 65 38 61 43 45 31 43

Nonsectarian

Protestant

1 4 1 4 1 4 1 4

89 92 88 95 91 95 87 94 79 91

1 4 1 4 1 4 1 4 1 4

Nonsectarian

None

HIGH COLLEGE SCHOOL

All Women

Year

Religion of Woman

Type of Institution

51 64 33 62 35 41 26 38

100 95 87 95 91 97 89 95 76 89

VERY MUCH

60 63 31 55 44 44 34 43

89 88 89 95 88 94 87 95 68 82

MUCH

70 67 51 59 48 49 35 46

89 90 89 96 93 96 85 89 86 100

SOME

* *

52 (50) 39 44

*

47 53 37 57

52 (67)

*

•

*

*

89 94 91 93 86 (100)

KONE

*

64 70

89 96 86 98 92 96 84 (100) 82 96

VERY LITTLE

Importance to Father

TABLE 25 Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by the Woman's Perception of the Importance of Religion to Her Parents

II. THE EFFECTS OF HIGHER EDUCATION

of her girl friends has no apparent implications for our generali­ zations about the effects of higher education on either family-size preferences (Table 26) or family-planning intentions (Table 27). Although there is some tendency for the differences between classes to deviate more from the general pattern for women whose friends are of different religion, this could just as easily arise from the smaller samples involved as from any sociological sources. The same analysis (tabular presentation omitted for reasons of brevity) was undertaken in connection with the re­ ligion of the girl's most recent dates and again the pattern re­ mains unchanged. The same conclusion must also be drawn from the analysis (Tables 28 and 29) of the four-year differences by the impor­ tance of religion to the student's best friend and her closest college friend (who is frequently the same person, no doubt). The small variations by class that do occur seem to be merely chance phenomena. Religiousness of Self

The last reference, and one also subject to change, is the religiousness of the student herself—the frequency with which she attends religious services and participates in other organized as well as informal activities, and—a more subjective dimen­ sion—the importance in life she attaches to religion. Holding constant such indices of religiousness will presumably also con­ tribute some clarification of the role of the college in affecting values about family planning and family size, although the possi­ bility of change in religious values over the four years is itself theoretically one of the most important components of change in attitudes toward family size. The measure of the number of religious practices observed is a Guttman scale constructed on the basis of different content and different cutting points for Catholics and non-Catholics and is therefore not comparable in any direct sense. For Protestants, Jews, and women expressing no religious preference, the fre­ quency was determined with which she attends religious services,

Nonsectarian

Protestant

Nonsectarian

Jewish

Protestant

Mormon1

C

NS

C

Catholic

Catholic

Catholic

C

C

NS

NS

5.6 (5.3)

*

*

*

*

3.5 (4.6)

*

*

*

*

3.1 3.1 3.6 3.7 (3.1) (3.2)

4.2 3.9 4.6 (4.4) 5.0 4.9 5.8 4.9

3.2 3.1 3.5 3.4 3.6 3.4 3.8 3.4 4.1 (5.3)

*

*

*

*

•

*

4.3 4.2 5.3 4.1 4.9 5.0 5.7 5.5

*

*

4.7 (3.3)

*

*

* *

*

3.1 2.8 3.6 3.5 3.7 3.1

CATHOLIC

*

3.0 3.0 3.5 3.1 3.3 3.2

PROTES- JEWISH TANT

Religion of Best Girl Friend

NONE

5.6 (7.0)

*

*

•

*

(4.5) (3.6)

*

*

*

*

3.3 2.8 3.9 (3.8) (3.3) (3.4)

* * *

*

*

*

*

*

*

*

*

*

*

3.0 3.1 3.3 3.2 3.3 3.2

*

4.2 3.8 4.8 4.5 4.9 4.8

3.1 3.0 3.5 3.4 3.5 3.3 3.8 3.4 4.0 5.3

PROTES- JEWISH TANT

4.2 4.1 5.2 4.3 5.0 5.0 5.7 5.4

*

*

4C

•

*

*

*

*

3.1 (3.3)

CATHOLIC

Religion of Most Girl Friends

NONE

* Less than 10 cases. Mean not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. 1 The means for women with Mormon girl friends are the same as for the "All Women" category. NS = Nonsectarian.

NS

Catholic

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

Nonsectarian

Protestant

1 4 1 4 1 4 1 4

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

1 4 1 4 1 4 1 4 1 4

Nonsectarian

None

HIGH COLLEGE SCHOOL

All Women

Year

Type of Institution

Religion of Woman

TABLE 26 Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Religion of the Woman's Best Girl Friend and the Religion of Most of Her Girl Friends

Nonsectarian

Protestant

Nonsectarian

Jewish

Protestant

Mormon1

C

NS

C

Catholic

Catholic

Catholic

C

C

NS

NS

4 1 4

1

4

1

4

1

4

1

4

1

4

1

1 4 1 4

Year

61 65 38 61 43 45 31 43

89 92 88 95 91 95 87 94 79 91

All Women

*

*

•

•

#

•

76 (60)

*

*

•

*

95 95 86 97 (71) (100)

62 75 59 (60) 41 50 30 54

88 93 89 96 86 95 87 94 72 (92)

*

*

*

*

*

*

48 (69)

*

*

#

*

86 86 79 95 92 95

PROTES­ JEWISH TANT

58 54 32 58 41 44 31 42

*

*

*

*

87 92 87 95 91 90

CATHO­ LIC

Religion of Best Girl Friend NONE

24 (17)

*

*

*

•

(82) (64)

*

*

•

*

90 92 92 (94) (92) (80)

*

*

*

*

*

•

*

•

*

*

*

*

*

83 95 79 100 92 96

*

64 70 46 68 43 (60)

92 93 89 96 90 93 87 94 75 (85)

PROTES­ JEWISH TANT

55 58 31 56 41 44 31 43

*

*

•

*

(89)

(82)

90 (93) 82 94

CATHO­ LIC

Religion of Most Girl Friends NONE

* Less than 10 cases. Proportion not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. 1 The proportions for women with Mormon girl friends are the same as for the "All Women" category. NS = Nonsectarian. C = Catholic.

NS

Catholic

COLLEGE

Nonsectarian

Protestant

HIGH SCHOOL

Nonsectarian

Type of Institution

None

Religion of Woman

Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by the Religion of the Woman's Best Girl Friend and the Religion of Most of Her Girl Friends

C

C

C

NS

C

Catholic

Catholic

Catholic

4

1 4 1 4 1 4 1 4

4

1

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

All Women

4.2 4.2 5.6 4.5 5.1 5.2 5.9 5.6

3.2 (2.7) 3.6 3.5 3.6 3.3 3.7 3.3 5.0 5.2

VERV MUCH

4.4 4.2 5.1 4.4 4.9 4.9 5.5 5.3

3.5 3.2 3.5 3.4 3.3 3.0 3.9 3.5 4.1 4.5

MUCH

3.9 3.8 4.7 4.1 5.0 4.7 5.5 5.5

3.0 3.4 3.5 3.3 3.3 3.3 3.6 3.6 4.1 4.5

SOME

4.5 (4.2) 5.3 4.7

*

*

*

*

*

•

*

*

*

•

*

*

4.2 4.0

*

*

*

3.0 3.0 3.4 3.0 3.4 2.9

NONE

(3.7)

*

*

3.0 2.8 3.5 3.3 3.4 3.4

VER V LITTLE

Importance to Best Friend

* Less than 10 cases. Mean not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian.

NS

NS

NS

Catholic

Nonsectarian

Mormon

COLLEGE

4 1 4

Protestant

Protestant

HIGH SCHOOL

1

Nonsectarian

Jewish

1

Nonsectarian

Protestant

1 4

Year

Nonsectarian

Type of Institution

None

Religion of Woman

4.3 4.1 5.6 4.6 5.1 5.2 5.9 5.6

3.4 (3.3) 3.6 3.5 3.6 3.5 3.8 3.3 5.1 5.4

VERV MUCH

4.2 4.1 5.1 4.2 4.8 4.7 5.6 5.3

3.2 3.3 3.5 3.4 3.5 2.8 3.7 3.6 4.1 4.6

MUCH

4.0 4.1 4.8 4.1 5.0 4.6 5.7 5.3

3.1 3.2 3.6 3.3 3.3 3.3 3.7 3.3 4.0 4.2

SOME

5.0 (4.6)

•

4.2 3.8 5.5 (4.7) •

• • • •

*

*

4.5 (3.8)

*

*

•

• •

(4.1) (3.8)

3.2 2.9 3.6 3.4 3.5 3.0

3.0 2.8 3.5 3.3 3.4 3.3 *

NONE

VERV LITTLE

Importance to College Friend

TABLE 28 Number of CJhildren Desired by Women, by Religion, Type of Institution, and Year, by the Woman's Perception of the Importance of Religion to Her Best Girl Friend and Closest College Friend

Nonsectarian

Nonsectarian

Protestant

Nonsectarian

Protestant

Jewish

Protestant

Mormon

NS

C

C

C

NS

G

Catholic

Catholic

Catholic

61 65 38 61 43 45 31 43

89 92 88 95 91 95 87 94 79 91

53 60 33 50 41 43 26 40

68 (100) 89 93 88 91 90 94 71 89

MUCH

All Women VERY

56 59 39 64 40 44 35 45

92 89 88 96 92 94 87 94 90 95

MUCH

68 71 37 63 44 51 36 43

91 95 88 96 91 96 83 97 85 91

SOME

* *

(77) *

45 (53) 44 58

*

*

*

*

*

•

* *

*

*

*

74 79

*

*

94 93 82 100 87 100

NONE

*

85 98 91 94

86

90

VERY LITTLE

Importance to Best Friend

* Less than 10 cases. Proportion not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

NS

NS

Catholic

1 4 1 4 1 4 1 4

1 4 1 4 1 4 1 4 1 4

Nonsectarian

None

HIGH COLLEGE SCHOOL

Year

Religion of Woman

Type of Institution

53 48 25 64 39 42 27 39

86 (94) 89 93 86 96 88 95 72 91

VERY MUCH

54 63 41 58 42 49 35 46

71 92 87 97 87 96 88 94 86 91

MUCH

69 69 46 65 44 52 33 42

95 96 89 96 92 97 83 90 84 92

SOME

*

*

*

36 (63)

*

*

*

*

64 (82)

*

*

*

*

90 94 87 92 88 97

*

*

*

75 85

(100) (100)

*

*

95 86 89 96 96 94

VERY NONE LITTLE

Importance to College Friend

Proportion of Women Intending to Plan Both Number and Spacing of Children, by Religion, Type of Institution, and Year, by the Woman's Perception of the Importance of Religion to her Best Girl Friend and Closest College Friend

ON FERTILITY VALUES WITHIN HOMOGENEOUS GROUPS

prays at home, sees a minister or rabbi about personal problems, and attends a religion-sponsored group meeting or activity. For Catholics, there are five items referring to the frequency of at­ tending mass, receiving Holy Communion, attending other re­ ligious services, seeing a priest or nun about problems, and at­ tending church- or parish-sponsored meetings and activities. Although this scale of religious practices and the self-rating on the importance of religion show some positive association, especially among Catholics, both with the number of children desired and family-planning intentions (Tables 30 and 31), the same degree of relationship appears to exist for both classes of students, with the result that the differences between classes remain unchanged. SUMMARY AND CONCLUSIONS

We have now examined the variations in average family-size preferences and proportions intending to plan fertility of firstand fourth-year college women of different religions and in different types of educational institutions within a series of con­ trols. In addition to the religion of the student and the type of college (as well as the type of high school for Catholics), we have compared the two fertility variables by the two classes within the following: Eiducation of parents Occupational class of father Nationality background of parents Religion of parents Religiousness of parents Religion of friends Religiousness of friends Religious practices of the woman Religious self-image of the woman The objective of imposing this extensive set of socioeconomic and religious controls on the analysis has been both to reduce the vulnerability of our inferences to the unknown biases of

II. THE EFFECTS OF HIGHER EDUCATION

selective attrition and to increase our general confidence in the assumption that these comparisons are not being distorted by any possible differences in the college classes in social factors relevant to fertility. The results of this analysis do not in any systematic way qualify the impressions gained from our initial comparisons. The main conclusion of our study of the effects of higher education on family-size values now seems quite clear and as reliable as the design of this study will permit. With one possible significant exception, for most women the experience of higher education does not appear to have any substantial effect on the number of children that women would like to have. This conclusion does not mean that college women will ultimately have the same number of children as women of lesser educa­ tion (which, in fact, we know not to be true) or even that they will desire the same number of children, but, rather, that any differences are much more the result of some of the selective factors determining who does and who does not go on to college. Despite all of the plausible theories to the contrary, there is simply little evidence from our data to support the proposition that the college experience itself has any major effect on this attitude. The same generalization applies to attitudes toward family planning as revealed by analysis of women's intentions to plan the number and spacing of their children. Although seniors typi­ cally report intending to plan fertility in slightly higher propor­ tions than freshmen, this "change" could just as plausibly result from the difference in age and the imminence of marriage. The one possible exception to this general conclusion involves Catholic women who have attended a Catholic high school. Those graduates who attend a nonsectarian college do seem to "experience" significant declines in the number of children desired and a significant increase in the proportion intending to plan both the spacing and number of births. Comparative evidence suggests that this is not wholly selectivity (though some initial differences are apparent) and the most reasonable infer-

NS

NS

NS

Protestant

NS

None

Protestant

Jewish

Protestant

Mormon

C

NS

C

Catholic

Catholic

Catholic

C

C

NS

NS

1 4 1 4 1 4 1 4

1 4 1 4 1 4 1 4 1 4

Year

4.2 4.1 5.2 4.3 4.9 4.9 5.7 5.4

3.2 3.1 3.5 3.4 3.4 3.2 3.8 3.4 4.6 4.8

All Women

3.8 3.7 (4.1) (3.4) 4.5 4.3 4.9 4.3

3.4 3.6

C2-6)

3.0 2.9 3.4 3.3 3.2 3.1 3.4

4.1 3.7 4.5 4.1 4.1 4.4 5.1 5.0 4.3 4.4 5.3 4.8 5.0 4.9 5.5 5.2

3.1 (3.6) 3.5 3.5 3.4 3.4 3.4 3.2 3.1 3.2 3.2 3.8 (3.4) 3.6 3.6 4.7 4.4 4.6 *

*

5.3 5.3 6.0 5.8

*

5.6 5.4 5.1 5.6

*

*

(4.1) (5.5)

3.5 (3.1) 3.6 3.6 3.5 3.5 3.8 3.4 5.2 5.8

*

4.3 4.8

3.2 3.0 3.6 3.5 3.5 3.3 3.9 3.4 5.1 5.0

5.3 5.9 6.6 6.1

*

*

*

*

—

—

—

—

—

—

—

—

—

Number of Religious Practices Observed 0 1 2 3 4 5

The scale for non-Catholics does not include this point. * Less than 10 cases. Mean not presented if either class has less than 10 cases.

NS

Catholic

HIGH COLLEGE SCHOOL

Tyf e of Institution

Religion of Woman

4.5 4.4 5.5 4.7 5.3 5.3 6.0 5.7

(3.7) (3.5) 3.6 3.6 3.6 3.2 3.9 3.4 5.2 5.4

4.1 4.0 4.8 4.1 4.7 4.5 5.3 5.0

3.5 3.5 3.4 3.4 3.7 3.6 3.9 3.9

*

*

3.6 3.6 4.7 4.1 4.5 (4.5) 5.1 5.2

3.3 3.3 3.5 3.3 3.3 3.2 3.4 (3.1) 3.5 3.4

*

*

*

*

*

(4.7) (4.3)

*

*

*

•

*

*

*

4.3 4.2

* *

*

*

*

3.1 3.0 3.2 (2.5) 3.4 3.1

*

*

*

2.9 3.0 3.5 3.3 3.4 3.2

Importance Attached to Religio VERY MUCH SOME VERY NONE MUCH LITTLE

TABLE 30 Number of Children Desired by Women, by Religion, Type of Institution, and Year, by the Number of Religious Practices the Woman Observes and the Importance She Attaches to Religion

NS

NS

NS

Protestant

NS

None

Protestant

Jewish

Protestant

Mormon

C

Catholic

1 4 1 4 1 4 1 4

1 4 1 4 1 4 1 4 1 4

Year

61 65 38 61 43 45 31 43

89 92 88 95 91 95 87 94 79 91

All Women

53 63 46 62

*

*

74 71

91 94 89 96 95 98 93 (100) 83 100

64 73 55 68 58 52 35 47

90 96 92 96 94 (94) 98 91

*

*

52 54 34 59 46 47 36 45

78 75 88 95 90 95 81 96 68 87

32 36 27 37

*

*

37 48 37 41

*

(55) (36)

97 100 86 94 85 92 88 98 73 97

*

37 (60)

95 96 89 96 92 97 90 88 74 84

28 29 21 37

*

*

*

•

—

—

—

—

—

—

—

—

—

Number of Religious Practices Observed 0 1 2 3 4 5

scale for non-Catholics does not include this point. * Less than 10 cases. Proportion not presented if either class has less than 10 cases. ( ) Less than 20 but at least 10 cases. NS = Nonsectarian. C = Catholic.

— The

C

NS

Catholic

C

NS

C

Catholic

NS

NS

Catholic

HIGH COLLEGE SCHOOL

Type of Institution

Religion of Woman

54 53 31 53 38 42 28 38

(67) (100) 87 94 83 86 89 94 72 88

VERY MUCH

60 67 45 68 45 46 37 50

90 95 88 95 85 95 90 96

*

*

MUCH

82 79 48 (64) 51 (65) 44 49

97 97 89 97 94 99 83 (94) 84 (100)

SOME

68 (73)

• •

* *

*

* *

* *

* *

* *

*

* *

* *

« *

*

90 92 89 (100) 92 96

NONE

*

91 89 85 97 90 93

VERY LITTLE

Importance Attached to Religion

by the Number of Religious Practices the Woman Observes and the Importance She Attaches to Religion

ON FERTILITY VALUES WITHIN HOMOGENEOUS GROUPS

ence we can draw is that the experience of a non-Catholic educa­ tion and association with non-Catholic women has played some role in producing this effect. This does not mean that classes of women graduating from Catholic high schools theoretically could be distributed randomly to Catholic and nonsectarian col­ leges and presumably give rise to such changes in fertility values (number of children desired and intentions to plan fertility), since it is reasonable to assume that students switching to a non-Catholic educational institution will be more receptive even­ tually to secularizing influences. Regardless of how much is concealed selectivity and how much of the "change" is due to "institutional effect," the overriding conclusion for the great majority of women and institutions in our sample is that the experience of higher education appears to have little effect on fertility values.4 * Since this book went to press, a recent addition to the literature (A. Greeley and P. Rossi, The Education of Catholic Americans, Chicago: Aldine, 1966) concludes that "Catholic colleges do indeed have a powerful impact both on religious behavior and social attitudes, but only among those who have already gone to Catholic primary and secondary schools" (p. 181). Our own data do not support this con­ clusion in the area of fertility values as well as in some other areas not presented in this volume. Though we question some of the Greeley-Rossi conclusions because of what we feel to be a research design inadequate to assess selectivity, part of the explanation may be that our data are confined to women while the authors maintain that the influence of Catholic education is more pronounced upon men.

CHAPTER 6

Higher Education and Some Beliefs about Marriage, Family, and Career Influenced by Jacob's findings that higher education has rela­ tively little effect on a student's basic values,1 we have concluded tentatively that it is unlikely that higher education as such will have any substantial delayed effect upon family-size preferences in subsequent years. In an attempt to investigate this question further, we have analyzed student responses to a series of state­ ments on cognitive, cathective, and normative beliefs2 about mar­ riage and the family. Our focus is not on the relationship be­ tween preferred family size and beliefs, a discussion deferred to a subsequent chapter, but on the differentials between the religious and college groups of the study population and espe­ cially between freshmen and seniors within each of these groups. To the extent that no change in belief structure can be inferred to be the result of higher education, the possibility of a delayed effect on family-size preference becomes even less likely. If the reverse should be found, the possibility of such a delayed effect becomes a distinct possibility and a crucial subject for further research. Recent surveys have shown numerous relationships in the study of values and beliefs as explanatory factors of fertility 1P. E. Jacob, Changing Values in College, (New York: Harper & Bros., 1957), p. 53. The author recognizes the existence of a few socalled "value-producing institutions" but maintains they are exceptions. See also A. Greeley and P. Rossi, The Education of Catholic Americans, for a discussion of Catholic education. 8 Because of the varying nature of the statements we are abbreviating these items with the word "belief." In a strict sense beliefs are statements about the true nature of things while values are preferences for some state of being and norms are prescribed ways of acting. It can happen, however, that religious values and norms become beliefs (Cf. C. Glock and R. Stark, Religion and Society in Tension, [Chicago: Rand McNally, 1965], pp. 173ff.), and on this basis we are using the word "belief in a generic sense meaning cognitive, cathectic, or normative orienta­ tions. In fact, some authors consider an ethical norm as an "implement­ ing belief" (Ibid., p. 25).

II. THE EFFECTS OF HIGHER EDUCATION

and family-size preference.3 Religion is a most important deter­ minant of fertility and most religions have developed systems of beliefs about marriage and the family.4 For example, the Old Testament praises fecundity and in general the sacred writers accept sexuality, though they qualify it by warning against sexual desire as such even in marriage: "In guilt was I born and in sin my mother conceived me."5 Nonetheless, the existence of personal values in marriage was clearly recognized and formed the basis for the acceptance of contraception in exceptional cases by some of the Jewish authorities cited in the Talmud.6 Early Christianity, however, rejected it.7 Among Prot­ estant theologians, Luther and Calvin, while stressing the per­ sonal element in marriage, taught that procreation is "the only really positive purpose of sex."8 But the Lambeth Conference of 1930 and subsequent convocations have developed the notion of responsible parenthood and have sanctioned the limitation of fertility. The Mormon Church of Latter-Day Saints stresses both parenthood and the "eternal principle" of sex.9 Also, a review of Catholic textbooks on the subject reveals extensive concern with the function of sex in marriage, reproduction, and 8 For example, Westoff, Potter, Sagi, and Mishler, in Family Growth in Metropolitan America, explain the higher fertility of Irish Catholics by their more faithful adherence to the Catholic value system. See pp. 202-203. * See R. M. Fagley, "Doctrines and Attitudes of Major Religions in Regard to Fertility," a paper presented at the United Nations World Population Conference, Belgrade, Yugoslavia, August 3O-September 10, 1965 and his book, The Population Explosion and Christian Responsi­ bility (New York: Oxford University Press, 1960), pp. 95-209. 5 Psalms 51:5. 6 Yebamoth 34b; 35a; Niddah 45a. 'See Clement of Alexandria, Paedagogus 2.10.91.2, Die griechischen christlichen Schriftsteller der ersten drei Jahrhunderte (Leipzig, 1897), 12:212. For an excellent summary of early Christian thought on this subject see J. T. Noonan, Contraception (Cambridge: Harvard University Press, 1965). "W. Cole, Sex in Christianity and Psychoanalysis (New York: Oxford University Press, 1955), p. 131. ' J. Widtsoe, Rational Theology (Salt Lake City: Deseret News 1915), p. 147.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

family structure.10 These facts suggested the study of belief sys­ tems in four distinct though related areas relevant to family-size preference: (a) the relationship of marriage to procreation; (b) the relationship between personal responsibility and family size; (c) the competence of religion in matters of family plan­ ning; (d) the compatibility of marriage and career for women. In each area of concern, several questions were asked and the students answered along a five-point continuum ranging from "strongly agree" to "strongly disagree." Except for Table 32, positive responses have been combined into one category representing the percent agreeing with the statement. This deci­ sion was prompted by the desire to economize tabular presenta­ tion and in general does not affect the conclusions. Given Jacob's findings and the fact that there is little evidence from our data that the college experience itself has much effect on family-size preference, we expected little or no difference between freshmen and senior responses to statements about mar­ riage and the family with the one exception of Catholic women who attended Catholic high schools and are now enrolled in a nonsectarian college, the group in which there is some indica­ tion that higher education may affect fertility values. We did expect, however, considerable differences among the religious groups since the various religions differ in their definitions of parenthood and family beliefs. MARRIAGE AND CHILDREN

The great majority of women do not agree that a married couple are obligated to want a large family (see Table 32). Even among Catholics and Mormons, who themselves desire relatively large families on the average, only 4 or 5 percent recognize such an obligation. If response to this question can be accepted at face value, it seems that the women desiring many children are not in the main so oriented because of any consciously formulated duty. There is, nonetheless, a highly sig10See J. C. Ford and G. Kelley, Marriage Questions (Westminster: The Newman Press, 1964).

II. THE EFFECTS OF HIGHER

EDUCATION

TARI.E 32 Percent of Women Agreeing or Disagreeing That a Married Couple is Obligated to V\ a Large Family, by Religion, Year, and Type of Institution Religion of Woman

Type of Institution

Year Agree Agree UndeStrongly Some- cided or what Don't Know

Percent N Disagree agree Total Some- Strongly what

None

Nonsectarian

1 4

0 0

1 0

0 0

3 4

96 96

100 100

Protestant

Nonsectarian

1 4

1 0

1 0

2 1

12 10

84 89

100 100

Jewish

Nonsectarian

1 4

1 0

1 1

1 0

8 6

89 93

100 100

Protestant

Protestant

1 4

1 1

0 0

1 0

10 7

88 92

100 100

Protestant

Catholic

1 4

2 0

0 0

4 0

27 38

67 62

100 100

Mormon

Nonsectarian

1 4

0 2

1 2

2 2

12 11

85 83

100 100

HIGH

3 1

1

COLLEGE

SCHOOL

Catholic

!NS

NS

1 4

2 1

2 2

4 3

27 22

65 72

100 100

[C

NS

1 4

1 0

4 3

4 3

25 22

66 72

100 100

NS

C

1 4

1 2

3 4

5 3

29 24

62 67

100 100

C

C

1 4

1 1

3 1

4 3

29 22

63 73

100 100

NS = Nonsectarian. C = Catholic.

90

. 2 1

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

nificant differential between Catholics and other groups in the extent to which they disagree strongly with the statement.11 Ninety-six percent of the women with no religious preference, 90 percent of the Jews, 86 percent of the Protestants, 84 percent of the Mormons, but only 66 percent of the Catholics reject with intensity the obligation to have a large family. The differ­ ential between students with no religious preference and the other groups is also significant.12 It is interesting, moreover, to note that Protestants in Catholic colleges resemble their class­ mates more than other Protestants and that little difference exists between Catholics in nonsectarian and those in Catholic colleges. To what extent are these intensity differentials a function of some ambivalence stemming from other values or beliefs? For example, Catholics may reject an obligation to have a large family but feel obliged to accept a large family if they engage in sexual intercourse. In fact, Catholic theologians traditionally have taught that procreation is the primary end of marriage, while Jewish and Protestant scholars do not distinguish between the "three purposes or functions interwoven in human sexuality: the procreation of children, the fulfillment and completion of husband and wife, and the establishment of a stable environment for a mature family life."13 These various orientations are reflected in the students' agree­ ment or disagreement with such statements as: having children is the most important function of marriage; people not intending to have children ought not get married; the ideal family has at least five or six children. With one exception, a significantly greater proportion of Catholics and Mormons than Protestants, Jews, and women with no religious preference subscribe to these uThe

probability that this differential is due to chance is less than

0.001. 11P is at least 0.02. " Fagley, The Population Explosion and Christian Responsibility, p. 207. Recent declarations of Vatican Council II do not refer to this distinction between primary and secondary ends, which may foreshadow an eventual change in emphasis.

II. THE EFFECTS OF HIGHER EDUCATION

statements by as much as 70 percent for some comparisons14 (see Table 33). Also, fewer Mormons than Catholics believe that having chil­ dren is the most important function of marriage: 66 percent as compared to 77 percent;15 and fewer Mormons than Catholics agree that the ideal family has at least five or six children. In fact, O'Dea has suggested that the orientation of Mormons toward large families is not exclusively a function of the primacy of a procreative end in marriage, but is related also to a particu­ lar type of family and social structure linked to an agricultural and gemeinschaftlichte way of life.16 It seems probable that stu­ dents enrolled in a Mormon university would show even greater commitment to large families, since our data on Catholics reveal the choice of college to be selective on religiousness, which for both of these groups carries implications for fertility values. Since the Mormon church, however, may exert considerable influence on the value systems of nonsectarian colleges within the state of Utah (from which our sample institution was drawn) the differential may not be as great as for Catholics. Furthermore, Mormons overwhelmingly reject the statement that if people do not intend to have children they ought not to get married. Certainly there exists for them a religious injunc­ tion: "Let children be born into the earth! Let fatherhood and motherhood be the most honored of all the professions on earth."17 But the Church of Latter-Day Saints, while rejecting sexual promiscuity, does not simply tolerate sex in marriage for the begetting of children since "Sex is an eternal principle" for the Mormon and "the equivalent of sex has always existed and will continue forever."18 "Depending on the differentials and the number of cases, the proba­ bilities that these differences are due to chance range from 0.02 to less than 0.001. aP is less than 0.001. "See T. F. O'Dea, The Mormons (Chicago: University of Chicago Press, 1957). " Widtsoe, Rational Theology, p. 147. 18 Ibid., p. 146.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER TABLE 33 Percent of Women Agreeing with Selected Statements on Marriage and Children, by Religion, Year, and Type of Institution Type of Institution

xgion of men

Year

Children Most Im­ portant Function of Marriage1

No Five or Six Children: Children No Ideal3 Marriaget

PERCENT AGRBE

PERCENT AGREE

PERCENT AGREE

Number (Minimum Base)

ne

Nonsectarisin

1 4

17 14

2 5

8 5

( 209) ( 186)

itestant

Nonsectarian

1 4

35 28

5 5

14 9

(3,632) (1,757)

/ish

Nonsectarian

1 4

29 23

1 2

6 6

( 519) ( 323)

itestant

Protestant

1 4

34 25

5 6

16 4

( 606) ( 195)

itestant

Catholic

1 4

50 62

12 9

15 10

( (

irmon

Nonsectarian

1 4

68 65

8 4

38 35

( 228) ( 154)

78) 21)

HIGH SCHOOL

COIXEGE

NS

NS

1 4

64 56

22 24

32 24

( 566) ( 355)

C

NS

1 4

86 72

55 45

43 35

( 253) ( 136)

NS

C

1 4

79 78

39 55

43 42

( 751) ( 393)

C

C

1 4

85 74

59 61

54 45

(2,565) (1,425)

tholic

Having children is the most important function of marriage. If people don't intend to have children they ought not get married. The ideal family has at least five or six children. IS = Nonsectarian. ί = Catholic.

11. THE EFFECTS OF HIGHER EDUCATION

Except for the statement on the primary function of marriage, Protestants, Jews, and women with no religious preference do not differ much in their response to the statements, a small minority, from 1 to 15 percent, agreeing with them. On the primacy of procreation, however, more Protestants (33 percent) and more Jews (27 percent) agree than do women with no religious preference (16 percent). The differences are signifi­ cant,19 and this particular statement seems to discriminate es­ pecially well among all the religious categories. Type of institution in which the students are currently enrolled does not seem to be a factor in the response of Protestants, with one important exception: Protestants in Catholic colleges resemble their classmates more than other Protestants in their answer to whether having children is the most important func­ tion of marriage, with 52 percent agreeing as compared to ap­ proximately 31 percent in nonsectarian and Protestant institutions.20 Among Catholics, type of high school attended is related to the response on all statements. For example, in nonsectarian colleges 64 percent of the freshmen from nonsectarian but 86 percent of those from Catholic secondary schools agree that hav­ ing children is the most important function of marriage. Among students in Catholic colleges, the differential by type of high school, though statistically significant, is only 6 percent.21 On the other statements the differences range from 11 to 33 percent, in marked contrast to response to whether a married couple are obligated to desire a large family, which seems not at all related to previous education. The type of college in which Catholics are currently enrolled seems relevant also. For example, of those agreeing that at least five or six children constitute an ideal family, 11 percent more Catholic freshmen in Catholic colleges than in nonsectarian col­ leges subscribe to the statement, whether or not they attended a nonsectarian high school. On the other issues, differentials of 15 and 17 percent obtain between Catholic freshmen in CathP is less than 0.001.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

olic and nonsectarian colleges who attended nonsectarian high schools.22 These results suggest selectivity factors other than earlier formal education. Comparisons between Freshmen and Seniors

If such comparisons can be relied upon to indicate the influ­ ence of the college experience, the data reveal little or no impact among women with no religious preference and Jews, the differ­ ence between the two classes in either group never exceeding 6 percent. Among Protestants in nonsectarian colleges, however, the 7 percent differential in response to whether having children is the most important function of marriage, while not much larger, is statistically significant, 28 percent of the seniors compared to 35 percent of the freshmen agreeing.28 The difference is more pronounced for Protestants in Protestant colleges, but in Catholic institutions, though the difference is not significant, it is reversed, with 12 percent more of the seniors accepting the statement. The same phenomenon does not exist in response to ideal family size. Consistently fewer seniors than freshmen, no matter where they are presently enrolled, agree that at least five or six children constitute the ideal. While the differences are minor on this issue, there is no reversal for Protestants in Catholic colleges, which suggests again that the influence of a Catholic institution appears to be more pronounced when a belief is overtly taught. Freshman-senior differentials among the Mormons are minor and not significant. Again, it is interesting to speculate whether the same results would have been obtained in a Mormon college. The apparent influence of the college upon Catholic students, while not large, is more pronounced for some of the subgroups. As noted previously, however, a comparison of freshmen differen­ tials reveals considerable selectivity, which is generally a more important factor, especially for students who attended a nonsec­ tarian high school. One exception is the response pattern on the obligation of having a large family, where selectivity seems P is less than 0.001.

II. THE EFFECTS OF HIGHER EDUCATION

negligible and the 7 percent differential between freshmen and seniors from nonsectarian high schools enrolled in a nonsectarian college and the 10 percent differential between freshmen and seniors from Catholic high schools now attending Catholic insti­ tutions are significant.24 Analysis of responses to whether having children is the most important function of marriage and whether the ideal family has at least five or six children shows no differ­ ence between the two classes among nonsectarian high school graduates in Catholic colleges, but within the three other sub­ groups significant differences of 8 to 14 percent exist, with fewer seniors agreeing with the statements.25 Do these results imply for this first subgroup that the Catholic college functions as a check to a so-called "secularizing process" that is only now being experienced by Catholic college students from Catholic high schools and is continuing for Catholic graduates of nonsec­ tarian high schools in nonsectarian colleges? If so, the Catholic college cannot be the sole factor, given the response pattern of college students from Catholic high schools. Other elements related to selectivity appear crucial. The responses to the statement that people should not marry if they do not intend to have children reveals another pattern. Again, selectivity is a factor, but, nonetheless, 10 percent fewer seniors than freshmen in nonsectarian institutions who attended Catholic high schools agree that they should not, while 16 per­ cent more seniors than freshmen enrolled in Catholic colleges from nonsectarian high schools agree, not only a significant difference but a distinct reversal.28 This may be due to new "technical" information being received, namely, that the intent to exclude children by limiting the right to intercourse invali­ dates the marriage. The fact that as many as 40 to 45 percent of the seniors in Catholic colleges disagree does not invalidate this reasoning but suggests some confusion. While the statement is clear in ordinary English it is not precise enough for Catholic "The probabilities are less than 0.05 and 0.001. M P ranges from 0.02 to less than 0.001. xP is less than 0.001.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

theology and canon law, which distinguish between simple intent to exclude children from marriage and a limiting condition to the right of intercourse added to the consent in marriage. Ac­ cording to Catholic doctrine, the latter invalidates marriage while the former does not.27 The general conclusion that emerges is that, except for this statement that people who do not intend to have children should not marry, and except for students in Catholic colleges from nonsectarian high schools, whose values on the average seem stable over the four years, the freshman-senior differentials among Catholics suggest a slight though statistically significant change in belief that might reduce family size. This is especially pronounced for students who attended Catholic high schools and are now enrolled in a nonsectarian university. Selectivity, however, seems to be a factor much more important than the college experience. PROCREATION AND BIRTH LIMITATION

Even though the great majority of students in every religious group reject any obligation to have a large family, they differ considerably on ideal family size and the relative importance of procreation in marriage. More than any other group, Catho­ lics are oriented to an ideal family of at least five or six children and are more committed to the primacy of procreation. The Mormons resemble them closely but seem to accept a more positive function for sexuality as such in marriage. These findings suggest further that the preference of Catholics for larger fam­ ilies may be, not necessarily the result of an overt orientation to size, but the consequence of a belief system surrounding the use of sex. In any case very few students within the other groups accept the primacy of procreation, and most, from 64 to 76 percent, agree that it is morally wrong to have more children than one can afford. But again Catholics differ significantly from "See H. Davis, Moral and Pastoral Theology (London: Sheed & Ward, 1953), IV, 190-196.

II. THE EFFECTS OF HIGHER EDUCATION

the others, with from 41 to 54 percent agreeing28 (see Table 34). These varying orientations are reflected in the responses to a belief that God will provide for as many children as are born and to the perceived immorality of birth limitation as a means of improving one's standard of living. As expected, very few women with no religious preference (from 2 to 5 percent), Prot­ estants (from 2 to 13 percent), Jews (from 1 to 7 percent), and Mormons (from 2 to 8 percent) subscribe to either state­ ment. Among Catholics, on the other hand, approximately 30 percent of the students agree that voluntary restriction of family size to improve one's standard of living is morally wrong, and 15 to 36 percent say they believe that a couple should have as many children as possible and God will provide.29 Again, an interesting difference in the belief structure of the two highfertility-oriented groups, Mormon and Catholic, is suggested. As noted earlier, sex in marriage is a positive value in Mormon theology and the Church of Latter-Day Saints seems to take the position of ignoring the issue of birth control although it stresses "the need to provide earthly bodies for waiting spirits."80 In general, more women reject a moral obligation to limit their families because of need than agree that God will provide for the children. This raises the issue of where these women center responsibility of providing for children, or whether re­ sponsibility is an issue at all for some of them. At least partial evidence that the problem is not defined completely in moral terms can be surmised from the fact that in most groups more women accept limitation of family size to raise their standard of living than are willing to admit that it is morally wrong to have children they cannot afford. The implication may be that the decision about family size is to some extent one of personal preference and not simply an ethical consideration. Such speculation is supported in comments made by a few 28,28 P is less than 0.001. "See O'Dea, The Mormons, pp. 140-141.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

TABLE 34 Percent of Women Agreeing with Selected Statements on the Morality of Family-Size and Birth Limitation, by Religion, Year, and Type of Institution ieligion of Vomen

Type of Institution

ione

Nonsectarian

'rotestant

Nonsectarian

ewish

Nonsectarian

iTotestant

Protestant

iTOtestant

Catholic

vlormon

Nonsectarian

Catholic

HIGH SCHOOL

COLLEGE

NS

NS

C

NS

NS

C

C

C

Year

God Will Provide for Children1

Family Limitation for Im­ provement Morally Wrong'

PERCENT

PERCENT

PERCENT

AGREE

AGREE

AGREE

Number (.Minimum Bate)

1 4 1 4 1 4 1 4 1 4 1 4

66 73 67 67 65 70 66 66 64 76 64 71

2 O 5 2 1 3 4 1 8 O 2 8

5 3 11 5 7 5 13 9 18 14 12 12

( 209) ( 186) (3,618) (1,748) ( 519) ( 321) ( 604) ( 193) ( 77) ( 21) ( 229) ( 155)

1

54 52 45 52 52 47 41 46

20 15 35 22 28 19 36 21

26 20 37 21 33 30 32 21

( 560) ( 356) ( 252) ( 135) ( 750) ( 392) (2,554) (1,425)

4 1 4 1 4 1

4 1

More Children than Can Afford Morally Wrong1

It is morally wrong to have more children than one can afford. family should have as many children as possible and God will provide for them. * Voluntary restriction of family size (by morally acceptable means) in order to improve jne's standard of living is morally wrong. NS = Nonsectari an. G = Catholic. 1A

II. THE EFFECTS OF HIGHER EDUCATION

of the women on their questionnaires. For example, a Catholic woman (Case 10126) wrote: "Our standard of living will prob­ ably not be too high because I'd rather have the enjoyment [original underlining] of children." And a Protestant (Case 68082 stated: ". . . maybe we could have 8 or 9 children which would suit me fine—I'd love to be 'barefoot and pregnant' for months." While such comments are not typical, they are suggestive of an important dimension in the formation of family-size prefer­ ences. There is also the possibility that some women are simply resigned to accept their fate. This attitude was clearly expressed by one Catholic (Case 13292): "The size of our family will depend on the gift of ourselves we are asked to make—my only hope is that we be open to the demands made upon us." Nonetheless, a majority of the women recognize some obliga­ tion to limit their families because of need. A Protestant (Case 79020) worded it this way: "The family's size will be limited by the financial state we are in, for it is better to bring few children into the world with advantages and happiness waiting for them than to have many children who will find a life of misery awaiting them." As to a perceived obligation to limit family size because of need, unlike the belief in the primacy of procreation in marriage, no significant differences exist between Protestant freshmen at­ tending various types of institutions, but among Catholics there is a differential. On the perceived morality of having more chil­ dren than one can afford, significant differences of 9 and 11 percent exist between freshmen who attended Catholic high schools and those who attended nonsectarian high schools31 but not between freshmen in different types of colleges who attended the same type of high school. While a minority of Catholics believe that God will provide for any number of children desired, 8 to 15 percent fewer students educated in nonsectarian high schools agree, and for these women an added selectivity factor seems to exist in the choice of a college, with 8 percent more freshmen in Catholic institutions than in nonsectarian institutions 11P

is less than 0.02.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER agreeing. These differences, though small, are significant,32 and much the same phenomenon is observed in the responses to the statement that voluntary restriction of family size, even by morally acceptable means, is immoral if the purpose is to im­ prove one's standard of living. Comparisons between Freshmen and Seniors On these issues either no differential exists, as among the Mormons, or more seniors than freshmen take a position imply­ ing a greater commitment to family planning, but the differences among non-Catholics is never very large. Among the Catholic subgroups there is a consistent pattern of fewer seniors than freshmen agreeing that a family should have as "many children as possible and God will provide for them," the largest differentials (of 13 and 15 percent) existing between seniors and freshmen who attended Catholic high schools and who are now enrolled in either a Catholic or nonsectarian institution.33 In these instances selectivity does not seem to be a factor except for students who attended a nonsectarian high school. The very same pattern emerges in the response to the statement that voluntary restriction of family size to im­ prove one's standard of living is morally wrong. Again, students from Catholic high schools seem affected most by the college experience, whether now attending a Catholic college or a non­ sectarian institution. Of course, this may be a function of age, with the seniors being closer to marriage and more realistic. This pattern is not found in the Protestant and Jewish groups because in general they do not subscribe to these beliefs even as freshmen. RELIGION AND BELIEFS ABOUT FAMILY PLANNING

In The Third Child it was reported that members of various religious groupings (from all educational levels) differ consider" P is at least 0.002. * P is less than 0.01.

II. THE EFFECTS OF HIGHER EDUCATION

ably in their perception of their religion's position on birth con­ trol and family planning.84 Only 20 percent of the Protestants, nearly 30 percent of the Jews, but 87 percent of the Catholics see their religion as taking any stand on such issues, but very few, 6 percent or less in each category, define it as encouraging the limitation of family size. In the present study, however, from 20 to 33 percent of the college women agree that "all major religions in the United States are in favor of family plan­ ning," including their own religion. Relatively slight differences exist among the subgroups; 33 percent of the Mormons, 30 percent of the Catholics, 26 percent of the Protestants, 20 per­ cent of the Jews, and 25 percent of those with no religious prefer­ ence, accept the statement35 (see Table 35). Certainly the stu­ dents' interpretation of their own religion's position influences the answers, but the traditional image of the Catholic church as the only major religion opposed to birth control may be an­ other factor. The Catholic response, therefore, is all the more surprising, especially the relatively high 40 percent of Catholic seniors in Catholic colleges who disagree with that image. These results suggest many interpretations: they may be a function of recent change in official policy with which some Catholics at least are more familiar; they may mean that these students are more defensive or that they interpret family planning mainly in the context of the rhythm method. The hypothesis on perceived change seems confirmed by the fact that 34 percent of the Protestants in Catholic colleges agree that the major religions encourage planned families, while only 26 percent of the other Protestants agree.8® But such an inter­ pretation is open to question. Even though more non-Catholics than Catholics view the Roman Catholic Church as teaching that large families are morally preferable to small families, more also tend to agree that its position is presently undergoing change ** Westoff, Potter, and Sagi, The Third Child, p. 84. "The differences between Mormon, Catholic, and the other religious groups are significant; P range from 0.05 to less than 0.001. " P is less than 0.10 and the differential may be due to chance.

102

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER TABLE 35

Percent of Women Agreeing with Selected Statements on Religion, Family Planning, and Birth Control by Religion, Year, and Type of Institution Type of Institution

'eligion of Vomen

Tone

Nonsectarian

rotestant

Nonsectarian

;wish

Nonsectarian

rotestant

Protestant

rotestant

Catholic

lormon

Nonsectarian

!atholic

HIGH SCHOOL

COLLEGE

NS

NS

C

NS

NS

C

C

C

Year

All Religions Favor Planning1

Birth Control Physically Injurious*

Religion to Determine Means of Control*

PERCENT AGREE

PERCENT AGREE

PERCENT AGREE

1 4 1 4 1 4 1 4 1 4 1 4

22 27 25 29 18 23 24 25 34 33 33 33

2 2 9 4 6 2 10 3 16 20 12 12

9 6 20 16 12 8 23 17 19 14 22 17

( 208) ( 185) (3,619) (1,755) ( 519) ( 322) ( 603) ( 194) ( 78) ( 20) ( 228) ( 155)

1 4 1 4 1 4 1 4

25 37 30 38 23 34 26 40

17 10 31 18 27 34 42 39

37 30 55 53 48 58 58 62

( 565) ( 354) ( 253) ( 134) ( 752) ( 392) (2,554) (1,423)

Number {Minimum Base)

1 All

major religions in the United States are in favor of family planning. repeated use of methods of birth control is physically injurious to the woman. * It is a perfectly legitimate function of religion to determine the proper means of family mitation. NS = Nonsectarian. C = Catholic. !The

(see Table 36). It may be, therefore, that Catholics are inclined to see whatever change is taking place as a positive one in favor of planning but the others perceive it as an eventual move toward a neutral position—a definition of the situation consistent with the interpretation of their own religion's stand, if the find-

II. THE EFFECTS OF HIGHER EDUCATION TABLE 36 Percent of Women Agreeing with Selected Statements about the Catholic Image, by Religion, Year, and Type of Institution Religion of Women

Type of Institution

None

Nonsectarian

Protestant

Nonsectarian

Jewish

Nonsectarian

Protestiint

Protestant

Protestant

Catholic

Mormon

Nonsectarian

Catholic

HIGH SCHOOL

COLLEGE

NS

NS

C

NS

NS

C

C

C

Number (Minimum Base)

Large Church's Families Position Morally Undergoing Desirable1 Changei

Outlaw Sale of Birth Control Devices'

PERCENT AGREE

PERCENT AGREE

PERCENT AGREE

1 4 1 4 1 4 1 4 1 4 1 4

45 39 45 44 46 51 61 57 38 29 36 38

63 81 47 66 46 63 43 72 44 52 53 57

45 37 43 45 43 51 45 46 41 40 33 27

( 207) ( 185) (3,610) (1,749) ( 516) ( 320) ( 604) ( 196) ( 75) ( 20) ( 226) ( 153)

1 4 1 4 1 4 1 4

43 38 29 30 34 31 28 23

61 73 48 61 44 54 40 53

60 47 67 54 64 66 69 60

( 564) ( 355) ( 250) ( 135) ( 748) ( 394) (2,548) (1,425)

Year

1

The Catholic Church regards large families as more morally desirable than small familie Catholic interpretation of family planning is presently undergoing change. * Catholics are morally bound to support laws which outlaw the sale of birth control device NS = Nonsectarian. C = Catholic. 8 The

ings of the Princeton study can be extended to this college stu­ dent population.87 Furthermore, the image of the Catholic church varies accord­ ing to whether or not the student reports affiliation with some " Westoff, Potter, and Sagi, The Third Child, p. 84.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

religion. While approximately 47 percent of Jews and Protestants agree that the Catholic church regards large families as more morally desirable than small families, fewer of the nonaffiliated (42 percent) concur, but among Protestants themselves there is a considerable and statistically significant differential: 60 per­ cent of those in Protestant colleges, 45 percent of those in nonsectarian institutions, but 36 percent of those in Catholic institu­ tions agree with the statement.38 It may be that in Protestant schools religion courses stress Catholic-Protestant differences as they were traditionally perceived. There is also variation in the image of whether or not "Catho­ lic interpretation of family planning is presently undergoing change," with agreement ranging from 46 to 81 percent. Protes­ tants in Catholic colleges, whose response pattern resembles that of their Catholic classmates, as well as Mormons, see the Catholic church as less committed to change while Protestants in Protes­ tant colleges and women with no religious preference tend to see its position as undergoing change. Among Catholics, however, the more the members of a sub­ group define their church as favoring large families, the more they see its doctrine changing. Is it that because of their belief about their religion's position on family size they feel a greater need for change and thus define current developments? Or is it that those who are more sensitive to the covert implications of overt teaching are also more sensitive to the consequences of the present situation? In any case, the differentials among the subgroups are revealing. Thirty-seven percent of the Catho­ lics in nonsectarian institutions but only 27 percent of the women in Catholic colleges agree that their church regards large families as more desirable morally while 61 percent of the former but only 45 percent of the latter see its policy as changing.39 Are these two phenomena defensive reactions by Catholic-educated students or simply increasing perceptions of the subtleties in Catholic theology? These differences are not mainly the function "* P ranges from less than 0.10 to less than 0.001. w In both instances P is less than 0.001.

105

II. THE EFFECTS OF HIGHER EDUCATION

of the school in which these students are presently enrolled, since selectivity is a major factor to be considered. In fact, among Catholic women in nonsectarian institutions, 41 percent from nonsectarian high schools, compared with 30 percent of those from Catholic high schools, define their church's policy as favor­ ing large families. Among Catholic college women the propor­ tions are 33 and 25 percent.40 Their perception of change follows a similar pattern. In nonsectarian schools the proportions are 65 and 52 percent and in Catholic colleges 47 and 43 per­ cent.41 Women in nonsectarian institutions who attended nonsec­ tarian high schools tend to perceive the ethics of the Catholic church as favoring large families but also view this norm as undergoing change, an image rejected most often by those who attended Catholic high schools and are now enrolled in a Catho­ lic college. This may be an indication that women more fully socialized into the Catholic tradition are more preoccupied with relating past and present positions than in seeing differences. This commitment to tradition is confirmed by Catholic responses to whether Catholics are morally bound to support laws outlawing the sale of birth control devices. Recently, many theologians have denied this obligation,42 yet the more a student has been exposed to Catholic education the more she tends to agree (see Table 36). For example, 66 percent of those educated in Catholic high schools and now in Catholic colleges accept the moral obligation, compared with 55 percent of those edu­ cated in nonsectarian institutions.43 The latter resemble more the Protestant, Jewish, and "no religion" categories, wherein 41 to 46 percent consider Catholics bound to support laws against birth control devices. It is difficult to regard this simply as a result of exposure to current literature, since the press has 40

In both instances P is less than 0.001. the first instance P is less than 0.001 and in the second less than 0.02. 45 See, for example, the discussion by John Dunsford in G. Shuster and D. Barrett, The Population Problem (Notre Dame: University of Notre Dame Press, 1964), p. 91. aP is less than 0.001. 41In

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

consistently publicized the views of many a theologian on this issue. A certain amount of selective conditioning is probably the answer. Response to the statement that "repeated use of methods of birth control is physically injurious to the woman" is revealing (Table 35). The issue is not the truth of the statement but the difference in student perception. Very few non-Catholics, from 1 to 10 percent, agree with the statement. Protestants in Catholic institutions are an exception; they resemble Catholics in nonsectarian colleges, where 17 percent agree. But among Catholics in Catholic colleges the percentage rises to 38 and suggests some degree of "rationalization."44 It is interesting to note, however, that an important segment of the students, including Catholics, not only are uninformed about the position of the major religions in regard to family planning but also deny religion's competence in determining the morality of the methods of birth control. While approxi­ mately 20 to 40 percent believe religions favor planning, as many as 34 to 44 percent admit not knowing their position, and ap­ proximately 40 percent of the Jews and those with no religious preference, 35 percent of the Protestants, and 32 percent of the Catholics disagree. Obviously, since the objective reality can­ not be "yes and no," either the respondents have interpreted the statement in widely different ways or the answers are indica­ tive of a lack of communication on the part of the religions and probably indifference on the part of the students. The last alternative is certainly plausible, since, in the response pattern to the statement, "It is a perfectly legitimate function of religion to determine the proper means of family limitation," the majority of non-Catholics disagree, with only a very small proportion, 8 percent of those reporting no religion, 11 percent of the Jews, and 20 percent of the Mormons and Protestants accepting the statement. Among Catholics, however, a signifi­ cantly greater proportion, which increases with Catholic educa­ tion, agree, approximately 35 percent of the women educated 44P is less than 0.001.

II. THE EFFECTS OF HIGHER EDUCATION

exclusively in nonsectarian schools but as many as 60 percent of those educated in Catholic schools.45 The differential among Catholics does not appear to be associated exclusively with the type of institution in which they are enrolled, because students who attended a nonsectarian high school, whether now in a Catholic college or not, reveal a greater tendency (by as much as 13 percent if attending a nonsectarian college) to restrict the role of religion in this area of life.46 On this issue both religious affiliation and, for the Catholics, schooling in denomina­ tional institutions are strongly related to the acceptance of a legitimate role for religion in defining the morality of birth con­ trol methods. Nonetheless, the significant numbers among Catho­ lics who deny this role or who are undecided is especially sur­ prising, given their church's strong stand on the matter. Could it be that for some students current debate on the morality of certain forms of contraception is being interpreted as their church's relinquishing its competence to decide the issue? Comparisons between Freshmen and Seniors

To the extent that it is valid to assume that freshmen-senior differentials represent a cohort's college experience, non-Catholic women in non-Catholic institutions reveal little change of per­ ception and evaluation of religion's concern with family plan­ ning.47 The major exception is the statement that the Catholic church is changing its views on family planning, with approxi­ mately 20 percent more seniors than freshmen agreeing.48 Among Catholic students, however, the response to other statements seems affected by the college experience. " P ranges from 0.10 to 0.001. 4eP is less than 0.001. While the differential, for those attending a Catholic college, is small (4 percent), it is significant, P being less than 0.01. "While the differential does not exceed 5 percent for all these groups except for the Mormons, where 7 percent more seniors agree with the statement, it is nonetheless significant for certain subgroups. For example, the chance probability is less than 0.002 for the 4 percent difference among Protestants in nonsectarian schools. aP is less than 0.001.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

From 10 to 15 percent more Catholic seniors than freshmen in both nonsectarian and Catholic institutions agree that the major religions favor family planning.49 As to the legitimacy of religion's determining the proper means of family limitation, the type of school becomes an important correlate of whatever change takes place among Catholic students. In nonsectarian colleges a significant 7 percent fewer seniors than freshmen and in Catholic colleges a significant 5 percent more seniors agree and accept this function for religion.50 Selectivity is certainly an element to be considered, since 43 percent of the freshmen in nonsectarian but 56 percent of those in Catholic institutions agree with the statement in this first year of higher education. The college influence, nonetheless, should not be minimized, since in the first group the percentage is depressed, while in the second it is increased. The type of school in which they are enrolled also seems to affect their perception of the effects of birth control on the woman's health. Thirty-eight percent of both freshmen and seniors in Catholic colleges agree that the repeated use of birth control methods is physically injurious to women, but in nonsectarian colleges 9 percent fewer seniors, 12 as compared to 21 percent, agree with such a conclusion.61 The Catholic student's image of her church as favoring large families is roughly the same for freshmen and seniors in the subgroups, though there is a tendency for fewer seniors to accept the statement. Among the women in Catholic colleges, 29 per­ cent of the freshmen and 25 percent of the seniors define their church as regarding large families as more desirable morally.52 More seniors than freshmen (by approximately 14 percent) agree that Catholic interpretation of family planning is chang­ ing.53 If this were a real cohort it might be argued that seniors are now being exposed to current publicity about the Catholic church and its attempts to come to grips with the problem of " P is less than 0.001 in both instances. " P is less than 0.05 and 0.001. n P is less than 0.001. aP is less than 0.002. mP is less than 0.001.

II. THE EFFECTS OF HIGHER EDUCATION

family planning, but, since this is a synthetic cohort, the fresh­ men are also exposed to the same publicity. This may be a function of age, the fact that seniors may be more sensitive to such issues, and the fact that the college experience may simply make the students more aware of what is happening in this area. MOTHERHOOD AND CAREER

Some authors on marriage and the family imply a basic in­ compatibility between the role of mother and the pursuit of a career. Others point to the tensions that inevitably arise be­ tween these two roles, given traditional cultural patterns, and it is often stated that career-oriented women have lower fertility rates.54 To what extent do the women in our study population intend working after college and how many intend to continue after marriage? Is there a particular belief structure underlying their intentions and does the college experience affect it in any way? The great majority of the students, from 90 to 99 percent in every subgroup, intend to work after college, but significant differences appear among the groups in their intent to pursue a career after marriage. Almost half of the women with no religious preference, from 46 to 48 percent, intend to pursue a career (Table 37), compared with 20 to 30 percent of the Jews, Mormons, and Protestants and only 8 to 9 percent of the Catholics educated in Catholic high schools and colleges.55 "For example, see A. Clemens, Marriage and the Family (Englewood Cliffs, N.J.: Prentice-Hall, 1957), p. 89; R. O. Blood, Jr., "Long Range Causes and Consequences of the Employment of Married Women," Journal of Marriage and the Family, XXVII (February 1965), 43-47; F. I. Nye and L. W. HoflFman, The Employed Mother in America (Chicago: Rand McNally, 1963); G. C. Myers, "Labor Force Participa­ tion of Suburban Mothers," Journal of Marriage and the Family, XXVI (August 1964), 306-311; Freedman, Whelpton, and Campbell, Family Planning, Sterility and Population Growth, pp. 318-319; Whelpton, Campbell, and Patterson, Fertility and Family Planning in the United States, pp. 107-112. mP is less than 0.001.

ON BELIEFS ABOUT MARRIAGE, FAMILY, AND CAREER

TABLE 37 Percent of Women Intending to Work or Have a Career, by Religion, Year, and Type of Institution Religion of Women

Type of Institution

Year

Intend Work After College

Intend Career After Marriage

PERCENT

NUMBER

PERCENT

NUMBER

STone

Nonsectarian

1 4

98 98

( 188) ( 179)

48 46

( 210) ( 187)

3TOtestant

Nonsectarian

1 4

98 99

(3,293) (1,715)

26 29

(3,633) (1,758)

Fewish

Nonsectarian

1 4

98 97

( 466) ( 315)

28 30

( 519) ( 324)

Protestant

Protestant

1 4

99 99

( 546) ( 193)

20 24

( 605) ( 196)

( (

69) 20)

30 14

( (

78) 21)

Protestant

Catholic

1 4

100 95

Vformon

Nonsectarian

1 4

90 92

( 228) ( 156)

21 22

( 228) ( 155)

1 4

99 98

( 528) ( 350)

24 23

( 566) ( 355)

Satholic

HIGH SCHOOL

COLLEGE

NS

NS

C

NS

1 4

98 97

( 228) ( 133)

8 17

( 254) ( 135)

NS

C

1 4

98 99

( 692) ( 390)

14

11

( 754) ( 392)

1 4

98 98

(2,326) (1,380)

8 9

(2,569) (1,431)

C

C

NS = Nonsectarian. C = Catholic.

Ill

II. THE EFFECTS OF HIGHER EDUCATION

The data certainly suggest that traditional family values are associated with religious orientation, and that the longer the exposure to formal education in religion-affiliated schools the greater is the commitment. Significantly fewer Protestant women in Protestant colleges than in nonsectarian colleges plan a career after marriage.56 The differential for Catholics educated in non­ sectarian schools compared with those in Catholic schools is even greater, approximately 15 percent.67 These results, however, should not be interpreted to mean that it is only or mainly the college experience itself that creates the differential, since in fact the freshmen in these groups differ more from each other than do freshmen and seniors within groups. The data indicate considerable selectivity, with the more traditionally ori­ ented choosing to enroll in religion-affiliated schools. In general, among the various groups, agreement with the statement that it is perfectly all right for a woman to have a family and a career at the same time follows the same rank order as intent to pursue a career, though more women, approxi­ mately 60 to 80 percent of the Jews, Protestants, and women with no religious affiliation, and 40 to 55 percent of the Catholics and Mormons, accept the legitimacy of playing two roles than actually intend to have a career. Again, Protestants in nonsec­ tarian schools differ significantly from Protestants in Protestant colleges; 70 percent of the former but only 60 percent of the latter feel that a career and a family are all right.58 And the Catholic groups differ considerably from each other, with 40 percent of those educated in Catholic schools as compared to 66 percent of those educated exclusively in nonsectarian institu­ tions accepting the statement59 (see Table 38). Much the same pattern emerges from the responses to the statement that a woman should derive her main satisfactions in life from a family rather than a career of intellectual or community activities. Significantly fewer women with no reli­ gious preference (14 to 21 percent) agree, while 32 to 38 percent of the Jews, 40 to 53 percent of the Protestants, and 43 to P is less than 0.001.

ON BEUEFS ABOUT MARRIAGE, FAMILY, AND CAREER

TABLE 38 Percent of Women Agreeing with Selected Statements on Family and Career, by Religion, Year, and Type of Institution Type of Institution

Religion of Women

Year

Career and Satisfaction Mother of Family from Young All Righfi Family Children not Careert Should Not Work'

Number (.Minimum Base)

PERCENT AGREE

PERCENT AGREE

PERCENT AGREE

1 4 1 4

78 81 70 69

21 14 49 40

77 74 91 89

( 209) ( 186) (3,620) (1,752)

None

Nonsectarian

Protestant

Nonsectarian

Jewish

Nonsectarian

1 4

76 83

38 32

86 83

( 518) ( 321)

Protestant

Protestant

Protestant

Catholic

Mormon

Nonsectarian

1 4 1 4 1 4

58 61 61 52 57 55

53 45 50 48 52 51

95 94 92 95 95 95

( ( ( ( ( (

66 67 50 63 47 49 38 43

52 43 64 48 57 57 65 59

94 89 94 86 95 95 97 95

( 567) ( 356) ( 253) ( 136) ( 752) ( 395) (2,559) (1,426)

Catholic

HIGH SCHOOL

COLLEGE

NS

NS

1 4

C

NS

1

NS

C

C

C

1 4 1

4

4 1

604) 194) 77) 21) 227) 155)

It is perfectly all right for a woman to have a family and a career at the same time. woman should derive her main satisfactions in life from a family rather than from a career, intellectual or community activities. 3 A mother with young children should not work if she does not have to. NS = Nonsectarian. C = Catholic. 1A

II. THE EFFECTS OF HIGHER EDUCATION

65 percent of the Catholics and Mormons endorse the state­ ment.*0 Here also Catholic and Protestant education seems rele­ vant, but the data suggest a strong selectivity factor. Though there is considerable variation in career-orientation of the groups there is greater consensus that a mother with young children should not work if it is not necessary. Approxi­ mately 75 percent of the women with no religious preference, 85 percent of the Jews, and from 90 to 95 percent of the other groups agree with the statement. In this connection the women with no reported religious affiliation differ significantly from the others.81 And a significant 10 percent differential separates Jews and Catholics,®2 but no difference exists between Catholics and Protestants or within each of these groups. Comparisons between Freshmen and Seniors

No significant differences are found between freshmen and seniors in the proportion intending a career after marriage except for those Catholic women who attended a Catholic high school and are now enrolled in a Catholic college (9 percent more of the seniors manifesting a desire for a career after marriage) .68 The 16 percent fewer seniors among Protestants in Catholic colleges is not reliable because of the small number of cases. On the other hand, the greater proportion of Jewish seniors who agree that a career and family are all right, 83 percent as compared to 76 percent of the freshmen, is significant. So is the 5 to 13 percent differential among Catholic seniors and freshmen who attended a Catholic high school, whether now enrolled in a Catholic college or a nonsectarian institution.64 Furthermore, generally fewer seniors than freshmen believe that a woman should derive be: main satisfaction in life from a family. The differentials ai 0.9) than are the latter (Z) = 1.2). Along with the Mormons, Protestants

INTERGROUP DISTANCES AND BELIEF STRUCTURES

in Catholic colleges appear as a pivotal group with the least variance in D scores and about equidistant from the Catholics educated in Catholic schools and from women with no religious preference (D's = 2.9 and 2.8). Catholics from nonsectarian schools resemble most the Protes­ tants in Catholic schools (D = 1.0) and Mormons (D = 1.3). They are slightly closer to fellow Catholics in Catholic colleges who attended nonsectarian high schools (D = 1.5) than to fellow students who attended Catholic high schools (D = 1.6). Fur­ thermore, a greater distance separates them from women with no religious affiliation than from Catholics educated in Catholic high schools and colleges (D's = 3.3 and 2.2). On the other hand, Catholics educated in Catholic high schools and now en­ rolled in nonsectarian colleges resemble most the Catholics who attended nonsectarian high schools who are now in Catholic colleges (D = 0.5). In fact, these two groups are closer in belief structure than any two groups. They also resemble more the Catholics from Catholic high schools in Catholic colleges (D = 0.8) than they do the Catholics from nonsectarian high schools in nonsectarian colleges (D = 1.6). It is interesting to note the close clustering of the Catholic groups with the one exception of Catholics who have not experienced a Catholic secondary or college education. In effect, then, the D scores reveal that women with no re­ ligious preference, Jews, and Protestants in nonsectarian or Prot­ estant colleges resemble each other in their response pattern. Mormons, Protestants in Catholic colleges, and Catholics edu­ cated in nonsectarian high schools and colleges appear to consti­ tute a second cluster, and other Catholics with some Catholic secondary or higher education form a third grouping. FACTOR ANALYSIS

To gain further insight into the belief space of the various religious groups and its relation to family-size preference and planning intentions, we have employed multiple-factor analysis to compare not only the factor structure of each group's beliefs

III. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

but also the structure of the factors containing the number of children desired and intent to plan the number and spacing. In view of the clustering of groups discussed previously, we could have conducted the analysis for each cluster, but then we might have lost some of the variations between groups. On the other hand, the separate analysis of the ten groups did not seem feasible because of the great similarity between some of them and the small number of students in others. Since religious differences have been consistently relevant so far in our analysis we compromised and proceeded with five groups: women with no religious preference, Jews, Protestants, Mormons, and Catho­ lics. For this last group we included, besides the beliefs discussed previously, three others involving perception of the Catholic church's position on family size, and for all groups we added two more variables to the belief items—the number of children desired and intentions to plan the number and spacing of children. Table 92 presents two main factors after rotation to simple structure for non-Catholics and (because of the added items) three for the Catholics. The proportion of the variance of the number of children desired accounted for by the factors varies for the religious groupings with the communalities (h2), or the proportion of common-factor variance explained, ranging from 16 to 37 percent. In other words, the factors extracted appear least relevant for Jews and Protestants. The proportion of the variance of intentions to plan fertility also varies ranging from 11 to 45 percent, but in this instance the factors appear especially relevant for the Catholic group. In fact, planning intentions are more affected for these women than desired family size, while the reverse is true for the other groups. As suggested pre­ viously, this may be a function of the virtual universality of fertility-planning intentions among Jews, Protestants, and women reporting no religion. For Catholics, family planning is a more simple derivative of these beliefs about marriage, family, and religion than is the number of children desired. Table 92 further reveals that the factors are not equally impor-

~ ~

-

-

.16

II

Jews

. 31

.18

.09

.06

.30

.17

. 37

.23

.23

.06

.12

.22

.14

.06 - .16

.04

.22 - .24

. 33

- . 33

.30

.09

. 38

.01

.62

.12

.24

.26

.07

.12

.24

.29

.50

. 53

.42

.54

.48

.57

.21

.27

.37

.26

.64

.11

.08

. 35 .17

.09

.34

.34

.29 .51

.56

.22

.50

.52

.04 - .35

.33

.09

.28

.30 -.02

.18 -.04

.14 - .08 -.48

.12

.17

.48

II

Mormons

.12 -.46 - .10

.16

.27

. 45

. 52

.21

. 30

. 37

.40

.16

.28

.53

.05

.26

. 37

. 35

.42

.00 - . 37

.22 - .01

. 15

.04

.40 - .03

Protestants II h2

.11 -.34

.17

hI

.47

.02 - .39

.17 - .05

.23

. 13 - .33 - .03

.41 - .03

I

. 39

- .05 -.48

- .32

.28

No Religion II hi

.53 - .01

I

. 13

- -

all right 4. Mother should not work 5. Satisfaction from family, not career 6. Children most important function 7. No children then no marriage 8. Five or six children ideal 9. More children than can afford wrong 10. Obliged to want large family 11. God will provide for children

3. Career and family

ber and spacing

2. Intent to plan num-

desired

1. Number of children

Variable

TABLE 92 Religious Groups: Rotated Factor Matrices

.41

I

.25 - .01

Catholics II III

.35

.20

.17

.04

.00

.32

.51

.51

.35

.31

.42

.13

.50

.19

.21

.16

.13 - .42 - .05

.38

.20

.36

.25

.27

.41

.26

.44 .34

.19

.27

.30

.34

.18

. 10

.15

.45

.23

hi

. 10

.22

.05

.23

.24

.07

.24 - .10 - .37 -.07

.22 -.65 -.15 - .07

.37

hI

8

12. Birth control physically injurious 13. Religion to determine means 14. Birth control for advancement wrong 15. All religions favor planning 16. Catholic church favors large family 17. Catholics must outlaw sale of devices 18. Catholic church's position changing

VQ1'iable

-

-

-

-

-

-

-

.01

.06

.09

.08

.07

.01

.07

.25

.08

.28

.16

No Religioll II hi

.24

I

-

-

-

.03

.29

.09

.19

I

-

.08

.25

.20

.35

II

Jews

.07

.34

.10

-

-

-

.06

.22

.23

.24

.40

.21

- - - -

-

.02

.20

.07

.15

Mormtms II

.25

I

.01 - . 17

.16

.06

.17

Prote.rta/lls II hi

.33

I

- - -

-

.01

.15

.05

.16

hi

TABLE 92 (continued)

.13

I

.25

-

.24

.36

. 12

.14

.21

. 19

,I

.01

.%5

.09

. 21

.5' .26

.08 - . tt

.15

- . 17 - .23

. 12

.n .42 - . IT

.40

CatholitJ, II III

- .10 - .06

.03 -.32

. 20

.05 - .05

.08

hi

INTERGROUP DISTANCES AND BELIEF STRUCTURES

tant for family-size preferences within each group. Factor I in­ cludes from 62 to 74 percent of the explained variance of num­ ber of children desired for Mormons and Catholics but almost all of the explained variance for Jews, Protestants, and women with no religious preference. Nevertheless, this factor appears generally similar for all groups. It is defined principally by a large-family ideal and preference (or less intent to plan) with a religious dimension of obligation or trust in divine providence and rejection of an ethical norm that it is wrong to have more children than one can afford. The major difference between groups is that for non-Catholics the belief that a married couple is obligated to want a large family loads primarily on this factor while for Catholics it loads on Factor III along with perception of their church's position on large families. This is, however, a result of the fact that extra items were included and a third factor extracted for the Catholic group. Another interesting difference is that belief that having children is the most impor­ tant function of marriage and that a woman should derive her main satisfactions from her family and not a career both load about equally on Factors I and II for women with no religious preference but primarily on Factor II for the other groups. This second factor is of statistical importance for family-size preferences only within the Mormon and Catholic groups. This is somewhat strange, since in all groups it is defined generally by items stressing family- or child-orientation as opposed to career-orientation and other regression analyses reveal that the intention to pursue a career is related to the number of children desired for all groups but Jews. We have no explanation for this except that beliefs about a career for women is not the same phenomenon as intention to pursue a career. It further suggests, however, that the traditional link in value systems be­ tween family-orientation, as opposed to career-orientation, and larger families may not hold true for certain groups among today's population of college women. Among Catholics and Jews the belief that the repeated use of birth control methods is physically injurious and among Cath-

III. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

olics, Jews, and Protestants the belief that it is a legitimate function of religion to determine the proper means of family limitation load principally upon this factor also. The structure supports our earlier tentative conclusion that perception of birth control methods as injurious is at least in part a function of religious and child-oriented commitments for some groups. The Mormons, though committed on both counts, are a conspicuous exception. For Catholics, Factor III, defined primarily by the belief that the Catholic Church favors large families and the belief in an obligation to want a large family, is not statistically important for the number of children desired or planning intentions, a finding consistent with previous analyses. In other words, what­ ever religious dimension exists in the larger families preferred by Catholics cannot be equated, at least overtly, with an image of the Catholic church as favoring or obliging couples to have large families. This supports a previous finding reported in The Third Child: ". . . there is clearly no evidence to support the proposition that higher fertility among Catholics can be attrib­ uted to conformity to any direct encouragement of higher fertility."® SUMMARY

While multiple-factor analysis reveals that Jews, Protestants, and Mormons resemble each other in the covariation of items in belief space, a measure of the distances between items in belief space suggests that the Jews and Protestants are closer to the women with no religious preference and Mormons closest to Catholics, especially the Catholic students from nonsectarian high schools now enrolled in nonsectarian colleges. But the more Catholics are educated in Catholic institutions the more distant they become. For all groups the factors extracted from a battery of beliefs about marriage and the family, number of children desired, and family-planning intentions explain 16 to 37 percent of the ' Westoff, Potter, and Sagi, The Third. Child, p. 86.

202

1NTERGROUP DISTANCES AND BELIEF STRUCTURES

variance in family-size preference and 11 to 45 percent of the variance in family-planning intentions. The factors are especially relevant for Mormons and Catholics, the former mainly with reference to the number of children desired and the latter with reference to planning. It may be that family-size preference is less influenced for Catholics because of their concern with the effectiveness of rhythm. The first factor, defined mainly by a large-family ideal, an ethical obligation to have a large family, and trust in divine providence (or the reverse) is generally the same for all religious groups. The second factor, however, defined principally by child-orientation as opposed to career-orientation, is of statistical importance for family-size preference within the Mormon and Catholic groups only. Previous analysis had revealed that the intention to pursue a career is related negatively to the number of children desired for all groups but Jews. The explanation lies probably in the fact that approval of careers for women and per­ sonal intent to follow a career are not the same. But the finding does suggest that the traditional link is changing or has changed in the value systems of some of our college women. A third factor, extracted for Catholics and defined mainly by perception of the Catholic church's position on large families, revealed that their preference for larger families is not an overt function of that image. The influence of religion is a more subtle affair.

CHAPTER 11

Multivariate Analyses of Correlates of Family-Size Preferences In Chapters 8 and 9 we examined the relationships of a number of beliefs and social background variables individually with both the number of children desired and family-planning intentions. Quite frequently, especially in such areas as religiousness, where numerous measures are included, the question of redundancy arises. For example, we see apparently strong associations with fertility values among Catholics of both the frequency of receiv­ ing Communion and the importance attached to religion. Are these relationships merely different reflections of the same under­ lying association between a general factor of religiousness and fertility values, or is there some independence involved? Perhaps there are many Catholic women regarding religion as an impor­ tant part of their lives who do not manifest this value in con­ ventional adherence to formal norms. And it is at least imagin­ able that the association with t fertility values may not be tapping that part of the variance held in common by the two religious variables. It is this kind of question to which the present analyses are directed. In addition, we are interested both in determining how much of the variance in fertility values can be explained by a number of social and "belief' factors and in comparing the structure of factors affecting these values across religions. Our procedure will be to look first at a large group of variables for each religious grouping separately and, through the technique of multiple-regression analysis, to isolate those variables that remain the best predictors after considering all interrelationships. Then we shall examine the structure of factors related to fertility attitudes for the sample as a whole, treating religion as an inde­ pendent variable.1 This will permit some evaluation in compara1This will be done also through multiple-regression techniques with religion formulated as a dichotomous variable.

204

MULTIVARIATE ANALYSES

ble statistical terms of the power religion has for the determina­ tion of fertility values. THE DEPENDENT VARIABLE

In these multivariate analyses we shall concentrate on the number of children desired as the dependent variable and, unlike previous practice, we shall treat intentions to plan spacing and number of children as an independent variable. Although the logic of this procedure is certainly defensible, the decision not to retain this measure as a dependent variable was made on the empirical grounds that: (a) the overwhelming proportion of non-Catholics intend to plan both spacing and number of children so that little variance exists to be "explained," and (b) among Catholics, the variables associated with family-plan­ ning intentions are, by and large, the same as those related to family-size preferences. THE INDEPENDENT VARIABLES

The independent variables to be included in the analysis are those discussed in Chapter 8. They include the conventional background variables of education of parents, father's occupa­ tional status, number of siblings, behavioral and attitudinal mea­ sures of religiousness and the woman's career intentions (the last analyzed in Chapter 9). In addition, for Catholics, there are measures of the amount of education received in Catholic schools by the woman2 and her parents and whether her parents are of Irish descent. Two other variables—the images of "large" and "small" fam1Education of the woman in Catholic schools is represented in this analysis only in terms of the elementary and secondary level. Whether the college attended is Catholic is not included because the analysis is based upon the total sample of Catholic women, including firstas well as fourth-year students, which, in the former instance, com­ promises the presumption of "effect." More important, our earlier analy­ sis indicates that most of the association of Catholic higher education with fertility values is reducible largely to selectivity.

205

III. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

ilies—which have some tautological overtones for the number of children desired, are also included. However, additional analy­ ses excluding these variables have been conducted, so that their effects can be isolated. RESULTS FOR PROTESTANT AND CATHOLIC WOMEN

Turning attention first to the two larger groups of Protestants and Catholics, we can see, by comparing the values of statistics listed in Table 93, that the number of children desired is consid­ erably more amenable to "explanation" in the case of Catholic women than in that of Protestant women. Although more vari­ ables are included in the analysis for Catholics and, thus, the results are not strictly comparable with those for Protestants, the amount of variance controlled for Catholics—30 percent—is twice that for Protestants—15 percent (see Table 94). Excluding the two variables measuring images of small and large families makes the contrast even greater, with the amount of variance of family-size preferences explained dropping to 25 percent for Catholics and to only 7 percent for Protestants. Among Protestants, aside from these two variables, only inten­ tions to have a career3 after marriage appears to exert any inde­ pendent influence on the size of family preferred; none of the remaining variables retains independent predictive capability other than intentions to plan fertility. The number of children desired by Catholic college women is more predictable and forms a more heterogeneous number of variables. Intentions to plan fertility is the variable of greatest statistical importance, maintaining a partial correlation of 0.24 with all other 23 variables held constant. As with Protestants, images of a large or small family are also significant predictors of family-size preferences, reducible only partly to other corre­ lates of family-size preferences. There is, however, an interesting " It is important to emphasize that this question relates to whether the woman intends to pursue a career after marriage, not simply whether she intends to work for a few years, which is an almost universal intention or assumption among these college women.

206

MULTIVARIATE ANALYSES

difference in the comparative importance of the two variables for Protestants and Catholics. For Protestants, perception of a large family discriminates among personal family-size preferences better than the image of a small family. Among Catholic women, on the other hand, the perception of the small family is a slightly better predictor of the number of children desired than is the image of a large family. In each case we are describing a pattern of deviation from the group norm. As observed in Chapter 8, the number of siblings exerts a significant effect on the number of children desired by Catholic women. Although the magnitude of the partial correlation (0.14) is hardly impressive, it is important to note that the original zero-order correlation (0.25) with family-size prefer­ ences is not diminished entirely by holding constant such the­ oretically plausible connections of religiousness, Catholic educa­ tion, and Irish nationality background. The career plans of Catholic women, as well as those of Prot­ estant women, maintain an association with number of children desired, which is to some extent independent of all other vari­ ables in the analysis. Several other variables show some statistically significant but hardly impressive relationships. Whether the woman attended a Catholic high school maintains a partial correlation of 0.09 with all other variables held constant. Of course, this still begs the question of selectivity because our only index of the religious home environment is the woman's perceptions of her parent's religiousness, and even this is measured some years after high school. Nevertheless, it is suggestive that the type of elementary school attended is of no ultimate predictive significance once the type of high school is known. The partial correlations of the different measures of religious­ ness drop appreciably from the initial correlations with familysize desires, but their importance may be obscured by the fact that so many are included in the analysis. The two variables maintaining some residual significance for family-size preferences are the frequency of receiving communion and the respondent's

III. INFLUENCE OF SOCIAL FACTORS AND BEUEFS

perception of the importance of religion to herself. Whether her parents are Irish retains only a small trace of independent association. RESULTS FOR JEWS AND MORMONS

The Jews and Mormons are the two remaining numerically small groups expressing a religious preference. The relative im­ portance of each of the 17 independent variables is presented in Table 93. The number of children desired by Jewish women is about as "predictable" on the whole (Table 94) as for Protestant women, with only 13 percent of the variance responding to all 17 variables considered simultaneously. (For Protestants the proportion of the variance of number of children desired con­ trolled is 15 percent.) The family-size preferences of Mormon college women is much more responsive to the independent variables included here than that of any other religious group. With all 17 variables included, fully 42 percent of the variance is explained. And, as in the comparison of Protestants and Catholics, the contrast is even greater if we eliminate the somewhat artificial independent vari­ ables of the images of a large and small family. With this reduc­ tion to 15 independent variables, the variance of family-size preferences controlled drops to 6 percent for Jewish women and to 34 percent for Mormon women (Table 94). In general, the reasons for the number of children desired being more "predictable" for Mormons and Catholics than for Protestant and Jewish women are that there is much more of a direct implication of religiousness for fertility values among the former two and that there are numerous measures of reli­ giousness included in the analysis.4 Aside from the images of large and small families, the only 4 Related to this difference is the fact that there is greater homogeneity of response among Protestants and Jews to the question on fertilityplanning intentions and higher correlations between these and the de­ pendent variable among Catholics and Mormons. This contributes in a minor way to increasing the difference.

b

Jewish

a

b

Mormon

a a

b

Νφ Religim

.01 .07 .01 .94 .02 .03 .00 -.01 -.05 .01 .01 -.04 .04 .07 .02 .00 .01 .01 .03 .02 .03 -.29 -.21 -.20 -.16 -.12 -.12 -.21 -.10 -.05 -.17 -.14 -.35 -.24 -.12 -.09 -.29 -.20 -.14 -.11 .29 .29 .18 .21 .28 .25 .24 .15 .15 .38 .37 .23 .25 .21 .20 .15 .32 .21 .21 .28 .44 .07 .04 .00 -.11 .22 .08 .22 .06 .08 .24 -.03 .00 .08 .09 .92 .03 .14 .00 .08 .26 .01 .04 .05 .00 .05 .00 .14 -.02 .08 .00 -.01 .23 -.04 .04 .01 .16 .01 .05 -.03 .02 .26 -.02 .04 -.01 .14 -.01 .03 .06 -.03 .31 .17 .07 .03 .18 .04 .06 .04 .03 -.04 .35 .07 .06 .03 .05 .01 .15 .03 .05 .05 .29 .12 .05 .15 .08 .00 .25 .14 .10 .10 .00 .00 .00 .03 -.03 .03 .00 .06 -.02 -.03 .37 .01 .08 .00 .16 -.02 .05 .00 .01 .10 * • • * * * • • .01 .25 * * .11 .23 -.08 .06 .02 .13 .04 -.61 * • * » • * * * .31 .08 * * * • * • * * .09 .23 * * * * * * * * .19 .01 * * * * » * • * .13 -.02 * * * * * • » * .01 .13 * * * * • * * * .14 .04 • * • • * * * * .14 .06

For Catholics, frequency attends services other than Mass. * Not applicable.

1

Education of mother Education of father Career plans Intentions to plan fertility Perception of large family Perception of small family Importance of religion to self Importance of religion to mother Importance of religion to father Importance of religion to best girl friend to Importance of religion to best college friend § Frequency sees clergyman Frequency attends religious meetings Number of siblings Occupational status of father Frequency attends services1 Frequency attends Mass Frequency prays Frequency receives Communion Attends Catholic high school Attends Catholic elementary school Mother attended Catholic schools Father attended Catholic schools Irish mother Irish father

b

Catholic

a

a

b

Protestant

i/u)Lr. yj Multivariate Analysis with Number of Children Desired: (a) Zero-Order Correlations; (b) nth Order Partial Correlations

III. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

TABLE 94 Values of Multiple Correlations with Number of Children Desired for: (1) All Original Variables; (2) Excluding Images of Small and Large Families

Protestants R R* Catholics R R' Jews R R* Mormons R Rt None R R*

1

2

.39 .15

.26 .07

.55 .30

.50 .25

.36 .13

.25 .06

.64 .42

.59 .34

.47 .22

.28 ,08

variables showing some residual independent association with number of children desired among Jews are number of siblings and two measures of religiousness—perception of the importance the woman attaches to religion and the frequency of praying in the home. It is interesting to note that this self-perception measure shows no initial correlation at all with number of chil­ dren desired but it does produce a negative partial correlation of some magnitude. Although this may result from sheer chance, it may also imply that higher fertility values respond positively to traditional religious indices such as praying in the home and at the same time negatively to abstract perceptions of the philo­ sophical importance of religion. This is speculative, however, and the main picture is that very little of predictive value emerges at all for Jewish fertility values. In this negative context, it is interesting that only among Jews (for unknown reasons) does the intention to pursue a career fail to correlate with family-size preferences.

MULTIVARIATE ANALYSES

The structure of factors affecting the family-size preferences of Mormon women is completely different. The column of zeroorder correlation coefficients in Table 93 reveals a large number of significant values. As a matter of fact, with the exception of the socioeconomic variables of education and occupational status of the woman's father, which show no association at all with number of children desired (the same is true in the other religious groups as well), all of the remaining 14 variables dis­ close statistically significant correlations with the dependent vari­ able. But, of course, there is considerable redundancy, as revealed by the partial correlations. The nine measures in the area of religiousness are represented by two significant partial correla­ tions—religiousness of self and the frequency of consulting a clergyman. Images of large and small families, plans to pursue a career during marriage, and intentions to plan the number and spacing of children comprise the other variables remaining with significant correlations after all other variables are held constant. The number of siblings, initially disclosing a correlation of 0.29 with number of children desired, reveals a partial correla­ tion of only 0.08 with all 16 other variables controlled. Further experimentation indicates that this is a result of the intercorrelations involving the images of large and small families. When these two variables are excluded from the matrix, the partial correlation between the number of siblings and the number of children desired declines5 only to 0.17. RESULTS FOR WOMEN WITH No RELIGIOUS PREFERENCE

We can now approach explicitly the question of whether the structure of factors affecting fertility values among women with no formal religious preference resembles that of women who identify themselves as Protestants, Catholics, Jews, or Mormons. In other words, we can now assess whether the same variables affecting family-size preferences of women associated with major religions operate when not filtered through a superstructure of ' The multiple correlation drops from 0.64 to 0.59 with the two variables measuring the images of a large and small family excluded.

211

III. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

religious belief systems (which, as we have seen, vary greatly in any direct implications they may have for fertility values). The answer to our question is quite clear from the comparisons in Table 93 of the "no religious preference" group and the other groups. The same few variables associated with family-size preferences for Protestants and (to a lesser extent) Jews appear for women with no religious preference. Aside from the expected contribution of the items measuring perceptions of large and small families, the variables maintaining some tangible associa­ tion with the number of children desired, with all other variables held constant, are intentions to pursue a career after marriage and intentions to plan the number and spacing of children. Number of siblings also belongs in the group since its initial correlation (0.J0) with family-size preference drops to zero in the partial correlation, entirely because of the interrelation with the images of a large and small family. With these items elimi­ nated, the partial correlation is 0.09. Thus, the same cluster of associations appears as that observed among most of the other groups of women. As expected, there is no predictive significance for fertility of religious attitudes or behavior among women with no religious preference. The question of whether the number of children desired by women with no religious preference is, on the whole, more or less predictable than for women expressing a religious identifica­ tion is answered in Table 94. With all 17 variables included, 22 percent of the variance of family-size preferences is explained—lower than for Catholics and Mormons but greater than for Protestants or Jews. With the two questions on images of large and small families eliminated, however, the number of children desired among women with no religious preference is just as predictable (or, more accurately, just as unpredictable) as for Protestant and Jewish women. In general, women with no religious preference are virtually indistinguishable from Protestant and Jewish women in terms of the social factors related to fertility (at least those included in this study). Theoretically speaking, it is more enlightening

MULTIVARIATE ANALYSES

to reverse this proposition—that the structure of factors affecting family-size preferences among Protestants or Jews (or, even more meaningfully, among women who are neither Catholic nor Mor­ mon) is the same as that among women with no religious con­ nection. Only among Catholic and Mormon women do we en­ counter the statistically important operation of factors in addi­ tion to career plans, fertility-planning intentions, and number of siblings. This other factor, of course, lies in the area of reli­ gious attitudes and behavior, a reflection of the substantive value system of the religion itself. BELIEFS AS INDEPENDENT PREDICTORS

The data reported in Chapter 9 revealed that beliefs about marriage and the family are associated with the number of chil­ dren desired, especially among Catholics, Mormons, and women with no religious affiliation. Since some of these beliefs are also significantly related to career plans, intentions to plan fertility, or the various measures of "religiousness," the present analysis raises the question of whether beliefs, independently of these items, predict family-size preference within these three groups. In other words, given our present frame of reference, are the beliefs redundant? Do they add to the predictability of the num­ ber of children desired or are they reducible, for example, to being more or less "religious"? The question has theoretical significance because the dimen­ sions of religiousness are many and our measures tap but three: self-perception of religiosity, religious practice, and organiza­ tional commitment. Some of the belief items tap a fourth, an inner-worldly ethic. Other studies have shown that these dimen­ sions are not always highly correlated and influence behavior differently.6 To what extent, then, are they relatively indepen­ dent of each other in predicting fertility values? We shall at­ tempt an answer by using the same technique of multiple-regres'Glock and Stark, Religion and Society in Tension, pp. 18-37; and G. Lenski, The Religious Factor (Garden City, N.Y.: Doubleday & Co., 1961), passim.

III. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

sion analysis, retaining the variables that have significant partial correlations and including nine belief items that are associated (without being tautological) with the number of children de­ sired. Since previous analysis revealed little or no association between beliefs and family-size preference among Jews and Prot­ estants, we have omitted these two groups from the analysis. Table 95 shows that the zero-order correlations between most of the belief items and the number of children desired are sigTABLE 95 Multivariate Analysis of Belieft and Other Selected Factors with Number of Childrer Desired: (a) Zero-Order Correlations;1 (b) nth Order Partial Correlations No Religion a b Career plans Number of siblings Intentions to plan fertility Importance of religion to self Frequency sees clergyman Frequency attends religious meetings Frequency attends services2 Frequency attends Mass Frequency prays Frequency receives Comniunion Catholic secondary education Irbh mother Irish father Career and family all right Satisfaction from family and not career Children most important function No children then no marriage More children than can afford wrong Obliged to want large family God will provide for children Birth control physically injurious Birth control for advancement wrong

-.20 .10 -.14 .08 .03 .05 .10

-.17 .09 -.08 .06 -.02 .01 .03

Mormon a b -.29 .32 -.25 .44 .31 .35 .37

*

*

*

.04

.01

.23

*

*

•

*

*

*

*

*

-.01 .24 .21 .07 -.18 .13 .19 .07 .10

*

.11 .14 .12 .00 -.14 .11 .11 -.02 .02

Catholu a

-.14 .18 -.13 .21 .13 .03 .05 *

-.10

-.21 .25 -.30 .22 .18 .15 .16 .25

-

-

•

.31 * * .23 • * .14 * * .14 -.29 -.08 -.20 .16 .01 .12 .31 .15 .17 .22 .04 .20 -.15 -.05 -.15 .13 .01 .09 .30 .09 .22 .14 -.01 .11 .19 .02 .12 *

1 The differences between the zero-order correlations presented in this table and tho Table 93 are due to differences in the maximum and minimum number of cases give! different listing of items. 8 For Catholics, frequency attends services other than Mass. * Not applicable.

MULTIVARIATE ANALYSES

nificant for Mormons and Catholics but that they have little independent predictability, the partial correlations being quite small for both groups. In fact, the multiple correlations of 0.62 and 0.50 are not much different from those presented in the second column of Table 94. One exception to this generalization about independent predictability is found in the Mormon group for whom agreement or disagreement with the statement that having children is the most important function of marriage re­ tains an 18th order correlation of 0.15. For women with no religious affiliation, however, while the zero-order correlations are not very large, they are generally not diminished as much as for Mormons or Catholics, and in two instances they are actually increased. With the present listing of items the multiple correlation is 0.43 compared to the 0.28 of the second column in Table 94. In other words, for this group beliefs have some independent predictive power. The data suggest that, for Mor­ mons and Catholics, cognitive and normative beliefs about mar­ riage and the family are closely interwoven with other factors that retain significant partial correlations with the number of children desired, and that family-size preference, while associated with specific beliefs, is as much a function of a generalized ambience in which Mormons and Catholics find and orient themselves. THE TOTAL SAMPLE: ALL RELIGIONS COMBINED

Thus far we have explored the correlates of family-size prefer­ ences within each major religious category separately. We will now treat religion as an independent variable rather than as a control. We shall be able to estimate the proportion of the variance of number of children desired that can be predicted when we include religious preference as one of the predictors along with such other variables as career intentions, number of siblings, and religious behavior. Partly to economize the pro­ cedure but mainly to eliminate variables consistently showing no relationship with the dependent variable, we have reduced the number of independent variables to eight. In order to solve

111. INFLUENCE OF SOCIAL FACTORS AND BELIEFS TABLE 96 Multivariate Analysis for Total Sample with Number of Children Desired: (a) Zero-Order Correlations; (b) 8th Order Partial Correlations; (c) 6th Order Partial Correlations; (d) 5th Order Partial Correlations; (e) 3rd Order Partial Correlations Variable

Career plans Intentions to plan fertility Perception of large family Perception of small family Importance of religion to self Frequency attends services— Mass Frequency prays—receives Communion Number of siblings Catholic—non-Catholic Multiple correlation

a

b

C

d

e

-.26 -.41 .36 .37 .29

-.13 -.20 .20 .20 .03

-.14 -.22

-.15 -.21 .21 .21

-.16 -.23

.03

*

•

.37

.01

.03

*

*

.18 .26 .50

.07 .12 .19 .62

.08 .19 .24 .57

*

*

.13 .22 .61

.20 .29 .56

* *

* *

* Excluded from the analysis.

the problem created by the fact that Catholics and non-Catholics were asked different questions about religious practices, the vari­ able in Table 96 entitled "frequency attends services—Mass" means the frequency of attending Mass for Catholics and un­ specified religious services for non-Catholics. The variable labelled "frequency prays—receives Communion" also separates non-Catholics and Catholics on different dimensions in that order. In addition, religious preference has been incorporated as a dichotomous variable—the woman is either Catholic or not. At first, religious preference was treated as a series of five dummy variables, that is, each person was classified as belonging to one of the five groups or not, and four of these variables (all five cannot be used because of the indeterminacy resulting from the redundancy),' but no more information is gained than from ' See D. B. Suits, "Use of Dummy Variables in Regression Equations," Journal of the American Statistical Association, LII, No. 280 (December 1957), pp. 548-551.

MULTIVARIATE ANALYSES

the use of the Catholic-non-Catholic split, which has the highest correlation with family-size preferences (0.46) .8 The nine independent variables listed in Table 96 account for 38 percent of the variance (the multiple correlation is 0.62) of the number of children desired among the more than 14,000 college women in the sample. Excluding the images of large and small families reduces this to 33 percent but does not alter the values of the partial correlations except for number of sib­ lings. This particular variable influences family-size preferences, partly through its connection with the normative range of family sizes so that its partial correlation with the number of children desired is lower when these image questions are included. With these questions excluded, the number of siblings shows a partial correlation of 0.19 with the number of children desired, com­ pared with 0.12. Intentions to plan a career also retains predic­ tive capacity independent of the other variables. Along with family-size definitions, the best predictors are whether the woman is or is not Catholic and whether she intends to plan the number and spacing of her children. The three variables measuring reli­ giousness, despite some high initial correlations, are absorbed by the Catholic-non-Catholic variable and by fertility-planning intentions. Thus, eliminating the three religiousness variables makes little difference in the multiple correlation with number of children desired. With six variables, the value of the multiple correlation is 0.61 compared with 0.62; with four variables, eliminating the images of large and small families, the value is 0.56 compared with 0.57 (Table 96). The role of religiousness in the explanation of fertility values is perhaps better revealed by the structures of two factors shown in Table 97.9 Factor I contains virtually all of the com* In addition, if two of the four variables included in the multiple regression are Protestant or not and Catholic or not, these nth order partial correlations with family-size preferences are decreased to virtually zero because of the high intercorrelation between the two dummy variables, —0.80. 'Factors were extracted by the principal-axis method and rotated to simple structure by the quartimax solution.

111. INFLUENCE OF SOCIAL FACTORS AND BELIEFS TABLE 97 Rotated Factor Structure for the Total Sample Variablts

I

II

A'

Number of children desired Intentions to plan fertility Perception of large family Perception of small family Religiousness of self Frequency attends services—Mass Prays—receives Communion Number of siblings Career plans Catholic—non-Catholic

.71 -.57 .44 .47 .40 .57 .15 .32 -.32 .72

.03 .01 .05 -.01 .67 .61 .73 .02 -.09 .01

.51 .33 .20 .22 .61 .70 .57 .10 .11 .51

mon-factor variance of number of children desired, intentions to plan fertility, perceptions of large and small families, number of siblings, and the Catholic—non-Catholic variable. It also con­ tains the limited portion of the variance of the variables measur­ ing religiousness that is associated with the different fertility variables and with religious preference, that is, whether the indi­ vidual is or is not Catholic. Factor II, on the other hand, con­ tains only the part of the associated variance of the religiousness items (the major part) not connected either with being a Cath­ olic10 or with fertility values. SUMMARY

Multiple-regression analysis reveals that the number of chil­ dren desired is more "explainable" for Mormons and Catholics than for Protestants, Jews, and women with no religion. Further­ more, for the first two groups, along with images of large or small families, the intentions to plan fertility, the number of siblings, career plans, and some measures of religiousness retain a significant amount of independent predictability. For Catholics, attendance at a Catholic high school also maintains some associaω Since Catholics constitute 45 percent of the sample and Protestants 43 percent, the variable comes close to being a Catholic-Protestant split.

MULTIVARIATE ANALYSES

tion with the dependent variable. Among Jews, besides the images of large and small families, only the number of siblings and two measures of religiousness, religious self-perception and frequency of prayer, and among Protestants and women with no religious preference, only intentions to have a career and to plan fertility show some independent association. Though the zero-order correlations between the belief items and the number of children desired are significant for Mormons and Catholics, they retain litde independent predictability. The data suggest that both family-size preference and beliefs are a function of the overall religious and social system in which women find and orient themselves. This is especially noteworthy since, for women with no religious preference, the zero-order correlations, while not very large, are generally not diminished much and in two instances are increased. Analysis of the total sample with inclusion of a Catholic-nonGatholic variable as an index of religious affiliation shows that the number of siblings and intentions to have a career and to plan fertility retain their predictive capacity. Religious prefer­ ence, however, simplified into this Catholic-non-Catholic di­ chotomy, remains an important predictor of the number of chil­ dren desired.

CHAPTER 12

Conclusions and Discussion of Factors Affecting Fertility Values Part III of this report redirected the analysis from the effects of higher education on fertility values to the significance of certain characteristics of individuals and religious groups for the understanding and prediction of these values. The analysis has concentrated on the social origins and certain family and religious characteristics and on the belief systems of women of different religions on subjects of marriage, family, career, and religion. All of this work has been directed toward building an explanation of the differences in the number of children desired by women in college and, to a lesser extent, the variations among them in their intentions to plan the number and spacing of their children. Analysis of social and personal characteristics has revealed some basic similarities as well as differences in the factors affect­ ing the fertility values of women of different religious prefer­ ences. The main differences reflect the fact that the content of Catholicism, and, in different ways, also the religion of mem­ bers of the Church of Latter-Day Saints, carries more direct substantive connotations for fertility values than for Protestant or Jewish women, who, in their fertility attitudes, differ hardly at all from women with no religious preference. The family ideology of Mormons and Catholics is basically traditional. They tend to equate marriage and motherhood in their beliefs with a religious dimension informing the association. For Catholics there is moreover, a strong tendency to reject birth control as an integral component of such a belief cluster. For most of these students an orientation toward relatively large families is not seen as a duty or obligation, nor is it connected to any image of their church as teaching such an obligation, but it is nonetheless connected with the extent and the intensity of identification with the religious system. Their family-size pref­ erences appear as a commitment to a communal—traditional way

FACTORS AFFECTING FERTILITY VALUES

of life in which religion is the integrating factor. Though other research has shown that the process of secularization and increas­ ing rationality (in the Weberian sense) may have affected their economic orientations, their family values have not been affected, at least not to the same extent. We do have, however, some evidence of the secularization process at work among Catholic students who have been exposed to nonsectarian educational institutions. Our Mormon sample (also enrolled in a nonsectarian school) resembles this group the most. On the other hand, the process of secularization ap­ pears almost complete among Protestants and Jews. Their belief system resembles most closely that of women with no religious preference. They do not equate marriage and motherhood and their ideology is generally not involved with religiousness. One implication of this difference is that the number of chil­ dren desired is much more predictable for Catholics and Mor­ mons than for these other groups. Although certain characteris­ tics, such as career intentions and number of siblings, correlate with family-size preferences in all groups, additional factors oper­ ate among the first two, mainly in the area of religiousness. The dimension of religiousness, measured in terms of reported behavior and attitudes, can be regarded simply as an extension of the substantive significance of the religion. Thus, the indi­ vidual can be thought of as being "more or less Catholic," a variation that, unlike "more or less Jewish," has predictive sig­ nificance for fertility attitudes. We might infer from this that the factor that seems to be important is not the "otherworldliness" aspect of religiousness but rather the degree of involvement in or commitment to particular "inner-worldly" systems of beliefs that have some bearing on the formation of fertility values. This "bearing" assumes both direct and indirect forms. An obviously direct form is the Catholic position on birth control, which until recently did not emphasize the virtues of family planning even with sanctioned means. Although the effects are more subtle than those of explicit encouragement of large fam­ ilies, there seems little doubt that such a policy has certainly

111. INFLUENCE OF SOCIAL FACTORS AND BELIEFS

operated over the years to produce such an effect, whether or not intended. Perhaps of greater significance for Catholics and Mormons alike (as well as some Protestants, no doubt) are the more in­ direct forms of religious influence that define the ideal role of the woman. Marriage is considered a career in itself, one only to be entered if the goal is to have a family. This "family career," moreover, is apparently perceived to be so all-embracing in its demands on the woman and in its intrinsic importance that a career role in the occupational world is regarded as incom­ patible. There are definite sacred overtones to the status of motherhood and, in Catholicism as well as the Church of Latter-Day Saints to a much greater extent than in other religions, there is more integration of the religious and secular role. In a certain sense, the woman has a divine mission to perform. Nonetheless, this religio-familial orientation does not operate in a vacuum. It is certainly supported by the social systems of these religions, and perhaps this is one of the most subtle of influences upon family-size preference. In fact, it has been reported that "the more liberal [church] bodies resemble occa­ sional audiences or focussed crowds, while the more conservative groups tend to resemble moral communities in the Durkheimian sense of the word."1 It is reasonable to assume that the church body that constitutes, at the local level, at least a quasi-primary group will tend to generate primary-group orientations among its members. There is considerable evidence that for Mormons and Catholics religious affiliation does in fact serve as a primary source of informal social relations2 that can certainly affect fam­ ily-size preference.8 1Glock

and Stark, Religion and Society in Tension, p. 163. p. 164, and O'Dea, The Mormons. We are aware of Joseph FichteriS conclusion that the Catholic parish is not a primary group but our discussion is focused on the relative importance of religion as a source of primary relations. (See Social Relations in the Urban Parish (Chicago: Chicago University Press, 1954). * It might be argued that this reasoning might apply to the Jews also, but it is still a debatable question whether their communal orienta­ tion is primarily religiously or ethnically based. i Ibid.,

FACTORS AFFECTING FERTILITY VALUES

These groups, however, are not completely closed and are subject to influence by the secular society. Catholics educated in Catholic high schools but now enrolled in nonsectarian insti­ tutions are an example of such a process. Exposure to contradic­ tory norms often forces a reconsideration of beliefs and values that underlie conflicting rules of behavior. It is true that the nature of commitment to a religious system may preclude any such reexamination, but this is precisely an area that deserves further study. In point of fact, religious commitment is not a unidimensional phenomenon, and not all commitments are ulti­ mate or inexorably maintained. With which type and in which areas of commitment does a rational consideration of values and beliefs underlying contradictory norms become possible? The answer to such a question is fundamental to understanding the possible effect of competing reference groups upon the fertility values of religiously affiliated individuals. In the first section of this report we concluded that, except for Catholics from Catholic high schools now in nonsectarian colleges, higher education exerted relatively little influence on fertility values. In all probability the values about family and children are developed Ibng before the college age, a speculation that is reinforced by the persistence of a correlation between the number of children desired and the number of siblings. This entire process of socialization is certainly amenable to research. Beginning with girls when they are 9 or 10 years old, a study would follow them for 10 years with repeated inquiries into their images of ideal family size and determine the extent to which and conditions under which such early ideals change. Such research, although expensive, would seem extremely worth­ while, not only in intrinsic theoretical terms, but also for practi­ cal reasons; for, if social scientists are to be called upon to advise on ways of changing or maintaining such values, it should be quite useful to understand their origins.

APPENDIX

N3

O

to

Sample Number __

College Number.

4 S 6 7 8 9

Sophomore, 2nd term

Junior, 1st term

Junior, 2nd term

Senior, 1st term

Senior, 2nd term

Other fDlease eXDlain)

2» Have you previously attended any other college?

2 3

Freshman, 2nd term

Sophomore 1st term

1

Freshman, 1st term

1· Please indicate your class here:

4

3

2

1

1

2

3

4

5

6

7

8

0 12

3

4

8· How many years did you attend a Catholic high school altogether?

0

7· How many years did you attend Catholic elementary schools altogether?

(IF NEVER ATTENDED CATHOLIC SCHOOLS SKIP TO Q. 9)

Private non-denominational school

Other denominational school

Catholic school

Public school

6. From what kind of elementary school did you graduate?

This questionnaire is the main source of data being collected throughout the United States for a study of the relationship between edu­ cation and religion and certain aspects of family life· The study is being conducted by the Office of Population Research at Princeton University and the Center for Population Research at Georgetown University· A large probability sample of all American institutions of higher education admitting women was drawn and this particular college was included* Most of the questions are factual and background in nature, dealing primarily with education history, religious background and present attitudes, and attldues toward family size· There are no liTightn or "wrong" answers to the attitudinal questions· The questionnaire should take only 15 to 20 minutes to complete· Your replies will be held completely confidential and analyzed in a statistical way only· Only the research team and no one at the college will read the questionnaire· Thank you very much for your time and cooperation· All you need to do is to circle the number corresponding to your answer·

COLLEGE WOMEN QUESTIONNAIRE

APPENDIX A

FOR OFFICE USE OKLY

1

IsD

4

2

3

4

5

Small city

Suburban

Small town

Rural area

2 3 4

Catholic school

Other denominational school

Preparatory (non«denominational school)

1

Public high school

5

4

3

2

1

As a child

5· From what kind of secondary school did you graduate?

1

Large city (over 100,000)

1 At present

community at present and where you lived as a child?

4· Which of the following categories best describes your home

Other (please describe):

3

With parents

2

In rental unit not with girls from college

With husband

1

In group housing with other college girls

3. What is your present housing arrangement?

7

More than 4 years of college

7

β

5

4

3

2

1

Father

2 3

Undecided No

2

1

θ

Father

1

2

1

0

Mother

Yes

11. Do you intend to work after college?

All

Some

None

Amount

in Catholic schools? (Circle one number for each parent)

χα How much, if any, of this education did your mother and father receive

6

5

4

3

2

1

Mother

College graduate

Some college

High school graduate

Some high school

Grammar school graduate

Some grammar school

9. How much formal education did each of your parents receive?

4

Already married

2 3

Undecided

No

2

No

1 2

Yes

No

17, Will you try to plan the number of children in your family?

1

Yea

16> Will you try to plan the spacing of your children?

IS, How many children do you think you will probably have?

like to hare altogether?

14( If you marry, how many children do you think you would

1

Yea

of your married life?

13, Do you intend to have a career outside the home during most

2 3

Yes

1 1

e. Mother's father

1

22

22

22

22

22

22

22

22

22

33

33

33

33

33

33

33

33

33

44

44

44

44

44

44

44

44

44

55

55

55

55

55

55

55

55

5 S

1

d. Best girl friend

1

1

c. Father

e. Closest college friend

1

1

>a

b. Mother

a. Yourself

to religion?

22

22

22

22

22

33

33

33

33

33

44

44

44

44

44

_

55

55

55

5 S

5 S

1s S,1 >A K1

How mach much importance doea does (did) each of the following peraons attttck 21. Hon

Host Most recent dates

i.

1 1 1

h. Host girl fHends

g. Father's father

1

1

d. Mother's mother d
How many children would there have to be in a family:

3

3

3 3

3

3

3

3 3

4

4

4 4

4

4

4

4 4

A few times a year or less About twice a month Every week More than once a week

O

3

3

i· If people do not intend to have children, they ought not get married

j· The ideal family has at least five or six children

1· A married couple is obligated to want a large family

k· It is morally wrong to have more children than one can afford

3

h· Having children is the most important function of marriage

3 3

A wife has a responsibility to keep children from bothering her husband in his leisure time

g· A woman should derive her main satisfactions in life from a family rather than from a career, intellectual, or community activities

ft

e· Some equality in marriage is a good thing but, by and large, the husband ought to have the main say in family matters

d· It is somehow unnatural to place women in positions of authority over men

c· Boys should be given more freedom than girls

bt A mother with young children should not work if she does not have to

a· It is perfectly all right for a woman to have a career and a family at the same time

5

5

5

5

5

5

23» Please circle whether you (1) agree strongly, (2) agree somewhat, (3) are undecided or don't know, (4) disagree somewhat, or (S) dis­ agree strongly, with each of the following statements* Agree Agree Undecided or Disagree Disagree Strongly Somewhat Don't Know Somewhat Strongly

y, The family should take priority over religion

x« Religion is a purely private matter

w« Religious organizations should let people think for themselves

v· A truly religious person should go to church or synagogue

tu The Catholic interpretation of family planning is presently undergoing change

t· Catholics are morally bound to support laws which outlaw the sale of birth control devices

desirable than small families

B· The Catholic Church regards large families as more morally

r« All major religions in the U.S. are in favor of family planning

q» Voluntary restriction of family size (by morally acceptable means) in order to improve one's standard of living is morally wrong·

p· It is a perfectly legitimate function of religion to determine the proper means of family limitation

o· The repeated use of methods of birth control is physically injurious to the woman

n· It is important for young people to marry a member of their own religion

nit A family should have as many children as possible and God will provide for them

K>

Last

First

Father Mother 11 2 2

1 2

Middle Initial

012345678 9+

SUters

29. Number of grandparents born in the U. S.

0 12 3 4

28. Father's main occupation (his past occupation if deceased) Please be specific:

012345678 9+

Brothers

27· Number of brothers and sisters you have had:

U. S. Other

26. Country of birth of parents:

U. S. Other

25· Country of birth;

24. Year of birth:.

Mama:

BACKGROUND INFORMATION

Try to picture the kind of life you will be leading about 10 year· after you leave college· Please try to imagine and write a brief essay describing what you hope to be doing, the kind of family and work your husband might be doing if you marry, and the home and standard of living you might enjoy.

233

2

2

Irish

4

4

5

Scandanavian

Italian

9 10

8

9

10

Russian, othet Slavic or Hungariaa

Spanish, Portuguese, Latin American, Puerto Rican

Don't know

Greek, Armenian, Turkish

7

7

Polish

11 12

11

12

8

6

6

French, Freoch Canadias, Belgian

5

3

Swiss Gennan

3

German, Austrian, Dutch

1

Father

1

Mother

English, Scotch, Welsh Eng. Canadian

Nationality Background

30. Which of the following best describes your main nationality background on your father's side and your mother's side? (Circle only one number in each colamni)'

AUTHOR Ballachey, E., 125 Barrett, D., 106 Barton, A. H., 14, 121 Blood, R. O., Jr., 110 CampbeU, A. A., 50, 110, 131, 182 Carlson, E. R., 125 Clemens, A., 110 Cohen, A. R., 124 Cole, W., 88 Corey, S. M., 16 Crutchfield, R., 125 Davis, H., 97 Duncan, O. D., 140 Dunsford, J., 106 Fagley, R. M., 88, 91 Festinger, L., 124 Fichter, J., 222 Ford, J. C., 89 Freedman, R., 110, 140, 182 Clock, C., 87, 213, 222 Greeley, A., 71, 86, 87 Hoffman, L. W., 110 Hoveland, C. I., 125 Jacob, P. E., 28, 87, 89, 121-22 Kelley, G., 89 Krech, D., 125 Lauriat, P., 30, 45

INDEX Lenski, G., 213 Mishler, E. G., 5, 35, 44, 88, 140 Myers, G. C., 110 Noonan, J. T., 88 Nye, F. I., 110 O'Dea, T. F., 92, 98, 222 Osgood, C., 191, 194 Patterson, J. E., 50, 110, 131, 182 Potter, R. J., 4-7, 13, 35, 44-45, 50, 88, 102, 104, 140, 202 Rosenberg, M. J., 125 Rossi, P., 71, 86, 87 Sagi, P. C., 4-7, 13, 35, 44-45, 50, 88, 102, 104, 126, 140, 202 Shuster, G., 106 Slesinger, D., 140 Stark, R., 87, 213, 222 Suci, G., 191, 194 Suits, D. B., 216 Szykman, M., 49 Tannenbaum, P., 191, 194 Webster, H., 28 Westoff, C. F., 4-7, 13, 35, 44-45, 50, 88, 102, 104, 126, 140, 202 Whelpton, P. K., 30, 45, 50, 110, 131, 182 Widtsoe, J., 88, 92

235

SUBJECT INDEX attrition, problem of, 16-17, 60-61, 82-83 belief'dissonance, 165 beliefs, 87-119, distance measures (D scores), 194-97; factor analysis, 197-202; group mean scores, 192-94; as predictors, 213-15. See also fertility values birth control and: college year, 108-09; health, 103, 107, 175-76; images of the church, 102-03; motivation, 97-101; re­ ligion, 101-08, 173-74; responsi­ bility, 100; sale of devices, 106 career and family, 112-14, 182-86, 201 career intentions, 110-11, 180. See also fertility values Catholic Church, and change, 15-16, 102, 104-05, 179-80; images of, 175, 177—80 Catholic colleges, 23-25 Catholic education, 5-7. See also fertility values Catholic sample, 9 Catholic textbooks, 88—89 Catholics, and number of children desired, 39-46. See also fertility values children, as function of marriage, 92-97; responsibility for, 98-101. See also family size Christianity, 88 coeducational schools, 21-22 cognitive dissonance, 124 college, choice of, 47-48 consensus, and fertility values, 27-28 contraception, methods and moti­ vation, 50-51 desired family size, 52-54; realism of, 44, 130, 132. See also fer­ tility values education of parents, 61-62, 145-47 expected family size, 132

factor analysis, 201-02, 217-18 family ideology, 220 family limitation, see birth control family norms, consensus, 137; in­ fluence of, 138 family planning, intentions, 50-59. See also fertility values family size, images and norms of, 135-38; intergenerational asso­ ciation, 142-44; preference, 21-49. See also fertility values family values, and religion, 112 fertility and: Catholic education, 4; higher education, 3; national­ ity, 146; religion, 4; working wives, 182 fertility values and: beliefs, 124-26, 163-69, 213-15; career intentions, 33, 181-86, 206, 210; college year, 33-46, 53-55, 57; consensus, 27-28; economic re­ sponsibility, 169-70; education of parents, 61-62, 145-47; factor analysis of correlates, 201-02, 217-18; family-size images, 206, 211; family-size norms, 137—39; high school, 42-46; higher edu­ cation, 33-55; ideal family, 168-69; images of the church, 177-80; motivation, 55-56; mul­ tivariate analysis of correlates, 204-13; nationality, 64-65, 67-70, 146-49; occupation of father, 64—66, 144—45; reference groups, 132-35; religion, 33-46, 53-55, 130-32, 215, 218; re­ ligious attitude, 155-58; re­ ligious practice, 152-56; reli­ giousness of self, 77, 82, 84-85, 160-61, 207; religiousness of sig­ nificant others, 71-77, 81, 159, 208, 210-11; selectivity, 57-58; siblings, 140—44, 207, 223; stan­ dard of living, 172-73; type of college, 33-46, 53-55; variance explained, 206, 208, 210, 212

SUBJECT INDEX generations, and family size, 139-44 higher education, and changing values, 27-28; and consensus, 27-28; effects of, 120-22; and fertility, 3. See also fertility val­ ues, beliefs hypotheses, 29-33, 89 ideal family, 92, 94; and familysize preference, 124—26, 168-69; images of, 135-38; theory of, 122-24 importance of religion, see reli­ giousness intercourse, right to, 96-97 Jews, and number of children de­ sired, 35-36. See also fertility values large families, ambivalence toward, 91; images of, 135-139; obliga­ tion to have, 163-64; orientation to, 4 Latter-Day Saints, see Mormons marriage, and children, 91-101; and religion, 88-89 methodology, 13-17 Mormon Church, 88, 98 Mormons and: fertility, 30; num­ ber of children desired, 38-39; sex, 88, 97-98. See also fertility values motherhood, see career and career intentions multivariate analysis, 204—13 nationality, and religious practice, 150-57. See also fertility values no religious preference, and num­ ber of children desired, 33-34; sample, 34. See also fertility values nonsectarian schools; 21-23 number of children desired, see family size, fertility values occupation of father, 64—66, 144-45 Old Testament, 88

procreation, duty of, 89-91 ; primacy of, 97. See also mar­ riage Protestant colleges, 23, 36-37 Protestant sample, 14, 34^-36 Protestants, and number of chil­ dren desired, 34—37. See also fertility values questionnaire, 226-33; administra­ tion of, 12 rationality, 115, 125 reference groups, 132-34 religion, commitment to, 223. See also fertility values religious attitude, 155-57 religious practice, 152—56 religiousness, as a control, 71-82; measurement of, 155-56. See also fertility values research design, 5-7, 14-17 response rate, 12-13, 18 sample, 8-12 secularization, 34, 57, 96, 125, 221 selectivity into college, 6, 42-43, 46-48, 95-96, 106 sex, and Mormon Church, 88, 92, 97-98 siblings and: family planning, 142; family-size preference, 141-43, 223; religion, 140; religious practice, 142-44 small families, images of, 135-38 social class, 64. See also occupation of father social origins, 144-52 socioeconomic controls, 60—71 standard of living, 172-73 synthetic cohort technique, 14-16 Talmud, 88 women's colleges, 21, 23-25 woman's role, 222 work, and children, 188-89; and fertility, 182 work intentions, 110-11, 180. See also career intentions