Search

Background: In a heavily cited paper, Finer (2007) asserted that by age 30, 82Š of US women born 1939-1948 engaged in premarital sex, increasing to 94Š for those born 1969-1978. Using the same data, our age 30 estimates are 55Š and 87Š for women born 1939-1948 and 1969-1978. Our analyses thus document strikingly different levels and trends. Methods: We replicate Finer's single-decrement Kaplan-Meier estimates of premarital sex using Cycles 3-6 of the National Survey of Family Growth, the same data as analyzed by him. We then contrast such single-decrement estimates for both premarital sex and first marriage with estimates of the simple percentages in three states: an origin state in which women begin life as never-married virgins and two destination states for first sex and for first marriage, depending on which occurs first. These analyses provide an empirical illustration of the fact that single-decrement estimates cannot be interpreted as simple percentages for demographic processes involving multiple decrements. Results: Our cohort estimates document increases in the percent of US women who had premarital sex by age 25, rising from 53Š to 75Š, 83Š, and 87Š for those born 1939-1948, 1949-1958, 1959-1968, and 1969-1978, respectively. Contribution: Our cohort analyses reveal sharp increases in premarital sex for US women born between 1939 and 1968, with increases most rapid for those born in the 1940s and 1950s. Our findings also reemphasize a standard lesson from formal demography - that single-decrement life table estimates cannot be interpreted as simple percentages for a multiple-decrement demographic process.

BACKGROUND In the U.S. today, premarital first births occur disproportionately to women with low education and income. We lack studies of whether this education gradient was present in cohorts born earlier. OBJECTIVE We examine education differences in the proportion of U.S. white and black women who: (a) experienced a premarital conception taken to term resulting in a first birth, and (b) had a premarital first birth by age 35. Among those experiencing a premarital conception, we examine the association between education and whether women married before the birth. We examine these patterns for birth cohorts born between 1925 and 1959. METHODS We use the 1980, 1985, 1990, and 1995 June Fertility Supplements from the U.S. Current Population Survey to examine cohorts of women born between 1925 and 1959. The survey asked women the dates of their first marriage and their first birth, allowing us to determine premarital conceptions taken to term, and whether the resulting first births occurred within or outside of first marriage. We present descriptive information on the proportion of black and white women in each cohort who experienced the events of interest by age 35. RESULTS For all cohorts, women with low education were generally more likely than their more educated counterparts to experience premarital conceptions and premarital first births. For blacks, but not whites, who experienced a premarital conception that was taken to term, those with more education were more likely to marry before the birth. CONCLUSIONS In the U.S., the concentration of premarital conceptions and premarital first births among less educated women was present for cohorts extending back to those born in 1925. 1. Introduction It is well known that, in the U.S., nonmarital births have increased for both whites and blacks (Bachu 1999; Wu 2008; Ventura 2009), and that, in recent decades, women who have nonmarital births are disproportionately of low education and income (Upchurch et al. 2002; Ellwood and Jencks 2004). This paper examines whether the relationship between education and having a premarital conception or first birth existed for earlier U.S. birth cohorts. 2. Data and methods We pool data from the retrospective marital and fertility histories from the June 1980, 1985, 1990, and 1995 U.S. Current Population Surveys (CPS). These June CPS fertility and marital supplements were administered to married women aged 15 or older and never-married women aged 18 or older. We use these data to identify the date (to the nearest month) of first births and first marriages. We assume that all conceptions occurred nine months prior to the birth, and employ a definition of a premaritally conceived but postmarital first birth as a first birth occurring less than seven months after a first marriage.4 This in effect assumes that first births that occur seven or eight months after a first marriage are premature births conceived within marriage. This procedure clearly introduces some classification errors, but provides a standard operational definition of a premarital conception for those conceptions taken to term. Because these data contain no information on pregnancies ending in miscarriage or abortion, our analyses are limited to conceptions taken to term. A further limitation of these data is that they lack information on cohabitation. However, the vast majority of premarital conceptions taken to term and resulting in a first birth to women born between 1925 and 1959 occurred in periods when abortion and cohabitation were relatively uncommon. Unweighted estimates are presented throughout. We limit our analysis to first marriages and first births. For the birth cohorts we examine, about three out of four premarital conceptions taken to term involved first births, and about nine out of ten marriages occurring between a premarital conception and a first birth were first marriages (results not shown). …