1976 and 1981 averaged 2.1 to 2.2 per woman. This stability in cohort birth expectations, especially for married women, suggests that at the aggregate level birth expectations data are reasonable predictors of completed cohort fertility.'

Another way to assess the quality of birth expectations data from the Current Population Survey is to compare for different survey years the percentages of women in the survey either stating that they were uncertain as to their future birth expectations or not responding at all to the question. Table A shows that a decline occurred between 1971 and 1981 in the proportion of married women 18 to 34 years old who stated that they were uncertain about the number of future births they expected. This "uncertainty" factor is an important component in the overall decline in nonreporting status over the 10-year period.

Nearly 10 percent of wives 18 to 24 years old indicated "uncertainty" in 1971, declining to 7 percent in 1981 (table A). Among both 25-to-29-year-olds and 30-to-34-yearolds, the percent reporting "uncertain" declined between 1971 and 1976 and remained essentially unchanged up to 1981. The decline in the percentage of wives reporting

'See Martin O'Connell and Carolyn C. Rogers, "Assessing Cohort Birth Expectations Data from the Current Population Survey: 19711981," Demography, Vol. 20, No. 3 (August 1983).

"The numbers of women in table A are weighted numbers so that different survey years can be compared, thus, they differ from those shown in appendix table A-1.

"uncertain" as to their future birth expectations in the decade may be the result of the increasing availability and use of contraception, thereby giving women more confidence in their ability to plan completed family size.

Table A also indicates that, in 1981, the highest proportion of uncertain respondents was found among 25-to-29year-old women (nearly 10 percent in 1981). In all likelihood, women approaching age 30 may be facing the most challenging childbearing decisions: a childless woman may be deciding about whether or not to have any children at all, and a mother may be deciding whether or not to have another child to expand an already existing family.

The overall stability in birth expectations and the decline in the proportion of women stating uncertainty about their lifetime birth expectations suggests that these data are useful as predictors of completed cohort fertility for married women. The birth expectations of married women are not, however, representative of the entire cohort of women since, in 1981, only 58 percent of 18-to-34-year-old women were currently married. The lifetime expectations of single women, who must consider the likelihood of both marriage and childbearing, have not been found to approximate their completed fertility as closely as that of married women. Thus, the proportion of women in a birth cohort who are married is crucial in assessing the future value of birth expectations for the entire cohort.

'See Martin O'Connell and Carolyn C. Rogers, op. cit.

(Numbers in thousands)

SOCIOECONOMIC DIFFERENTIALS

IN FERTILITY

Table B summarizes the socioeconomic differences in childbearing in the year ending June 1981 among all women 18 to 44 years old. Data are also shown in the table for two age groups of women, 18 to 29 and 30 to 44 years old, in order to examine childbearing among women over age 30, a subject of recent interest. Additionally, since certain characteristics tend to be age-related (e.g., education and family income), this dichotomy by age 30 also helps to control for the confounding effects of age on fertility. The fertility rate shown in table B is defined as the number of women who had a birth during the 12-month period pre

age group. In 1981, the fertility rate for the Nation was estimated to be 70.9 per 1,000 women 18 to 44 years old, not significantly different from the 1980 rate (71.1 per 1,000). The 1981 rate for women 18 to 29 years old was 100.4 per 1,000 women, considerably higher than the rate for women 30 years of age and older (38.7 per 1,000). In order to gain a better understanding of the timing of childbearing among American women, first-birth rates are also presented in table B. Thirty-eight percent of all women 18 to 44 years old who had borne a child in the 12 months preceding the June 1981 survey gave birth to their first child, resulting in a first-birth rate of 27.1 per 1,000 women. Women under 30 years of age recorded a first-birth rate of 44.6 per 1,000 women, many times that of women age 30 and above (7.9 per 1,000 women) who are now approach

Of all births to women 18 to 29 years old in 1981, 44 percent were first births; among 30-to-44-year-old women, only 20 percent of all births were first births. A previous study, using data from the June 1976 CPS, showed that among 30-to-44-year-old women, 14 percent of all births to these women in 1976 were first births, resulting in a rate of 4.3 first births per 1,000 women, compared with 7.9 per 1,000 women in 1981. Data from the National Center for Health Statistics also show an increase since 1971 in firstbirth rates for women age 30 and older.

In 1981, the fertility rate for married women, husband present, was 98.7 per 1,000 women 18 to 44 years old. A significantly lower rate (70.8 per 1,000 women) was recorded for other married women, husband absent (including separated). Although not statistically significant, a higher proportion of births to married women with husband present are first births (37 percent) than for married women with their husband absent (30 percent). Part of this apparent difference may be due to differences in age and duration of marriage. A lower rate of childbearing is found among single (never-married) women (27.0 per 1,000 women) than among

*See Martin O'Connell and Carolyn C. Rogers, "Differential Fertility in the United States: 1976-1980," Family Planning Perspectives, Vol. 14, No. 5 (September/October), 1982, pp. 281-286.

See National Center for Health Statistics, Monthly Vital Statistics Report, Vol. 31, No. 2, Supplement (2), "Trends in First Births to Older Mothers, 1970-79," May 27, 1982.

Women Who Have Had a Child in the Last Year, by Age, Marital Status and Labor Force Status

currently married women (96.8 per 1,000 women). Fiftythree percent of all births to single women were first births, which is the highest proportion among all marital classes. A woman's labor force status is one of the most transitory characteristics shown in table B, since it is affected by the woman's pregnancy status via exit from the labor force during the pregnancy and postpartum period. Ideally, data should be presented by the woman's labor force "commitment" or past work history in order to obtain a better picture of fertility differences among women. However, an analysis of recent childbearing by labor force status may aid in the planning of maternal and child social service programs that assist women in re-entering the labor force after the birth of a child. Of all women 18 to 44 years old in June 1981 who had had a child in the preceding 12-month period, 42 percent were in the labor force as of June 1981. The fertility rate of women in the labor force (44.0 per 1,000 women 18 to 44 years old) was significantly lower than the fertility rate of women who were not in the labor force as of June 1981 (126.1 per 1,000 women). The low fertility rate of employed women reflects the time constraints of coordinating motherhood and employment. The high fertility rate found among women not in the labor force does not necessarily imply that these women had never been in the labor force; this category also includes women who have only temporarily dropped out of the labor force to have and care for a child.

Employed women 18 to 29 years old in 1981 had a fertility rate which was substantially lower (54.2 per 1,000 women) than that of their counterparts who were not in the labor force (193.1 per 1,000 women). Although a higher proportion of women not in the labor force are married (67 percent) than employed women (50 percent), this difference in the proportion married does not account for the labor force differential in period fertility rates. The bar graph shows that the fertility differential between employed women versus those not in the labor force persists among both never-married and ever-married women, with higher rates to ever-married women.

A similar childbearing pattern is evident among 30-to44-year-olds. A lower fertility rate was recorded among employed women (25.9 per 1,000) than among those not in the labor force (60.2 per 1,000). Once again, the proportion of employed women who are married (90 percent) differs significantly from the proportion married among women who are not in the labor force (95 percent). The bar graph also shows that after controlling for the woman's marital status, fertility rates are still highest among women not in the labor force.

Among employed women 30 and over, nearly 20 percent are currently childless and are more at risk of having a first birth than are women who are not in the labor force (7.1 percent). Concomitantly, a higher proportion of employed women 30 to 44 years old who had a birth in the last year had their first birth (24.9 percent), compared with 17.1 percent of women not in the labor force. These data indicate a greater tendency among employed women to delay motherhood in order to pursue a career. OUT-OF-WEDLOCK CHILDBEARING

Data in table C from the June 1981 CPS show that relatively high levels of out-of-wedlock childbearing occurred

'Percentages of childless women are based upon unpublished CPS data.

among Black women, women with low levels of family income, and women with less than a high school education. The nonmarital or out-of-wedlock fertility rate is defined as the number of out-of-wedlock births per 1,000 unmarried women" 18 to 44 years old. The rate of out-of-wedlock childbearing for Black women 18 to 44 years old in 1981 was 67.2 births per 1,000 unmarried women, a level nearly four times higher than that for White women (17.9 per 1,000).

For both races, higher nonmarital fertility rates are associated with lower educational attainment and lower family income. Among Black women, the highest rate was recorded for those with family incomes under $5,000 (117.1 per 1,000) as compared with Black women with a family income of $20,000 or more (48.3 per 1,000); among White women a higher rate is also found among women with a family income below $5,000 (44.1 per 1,000), five times the level for White women with family incomes of $20,000 and above (8.6 per 1,000). Table C also shows that the nonmarital fertility rate for White women who were not high school graduates (51.4 per 1,000 women) was nearly nine times that of women who had completed at least 1 year of college (5.8 per 1,000 women). Similarly, Black women who were not high school graduates had higher nonmarital fertility (92.7 per 1,000) than women who had completed at least 1 year of college (34.0 per 1,000). Women who become pregnant while in school often withdraw from school in order to have and care for their child. This places them at a considerable disadvantage in terms of completing their education and also, in the longer run, in terms of securing employment.

Nonmarital fertility rates do not differ consistently by race for the residential characteristics in table C. For White women, there is some evidence that the nonmarital fertility rate was lower in the South than in the rest of the Nation. Although the pattern appears to be the same for Black women, the difference is not statistically significant. Fresh

"For these calculations, unmarried women are either single (nevermarried) or widowed and divorced; separated women are excluded.

High school, 4 years..

5,764

17.6

1,367

72.8

nock and Cutright's study of out-of-wedlock childbearing found lower nonmarital fertility rates among both White and Black women residing in the South. Table C also shows that White women residing in poverty areas had a higher nonmarital fertility rate (28.6 per 1,000) than women residing outside poverty areas (16.5 per 1,000). This may be partially due to the noted difference in out-of-wedlock childbearing by family income.

RELATED REPORTS

Advance data on birth expectations and fertility for June 1981 were published in Current Population Reports, Series P-20, No. 369. Data from the June surveys of 1972 through 1980 appeared in Series P-20, Nos. 248, 265, 277, 301, 308, 325, 341, 358, and 375. The results of the June 1971 survey, which include data on childspacing and cumulative fertility

See Larry Freshnock and Phillips Cutright, "Models of Illegitimacy, United States, 1969," Demography, Vol. 16, No. 1 (February 1979), pp. 37-47.

'Data from the CPS are not tabulated by the woman's poverty status but rather by poverty area residence. Poverty areas are defined in terms of census tracts or minor civil divisions in which 20 percent or more of the population was below the 1969 poverty level.

by age for birth cohorts and by interval since first marriage for marriage cohorts, were presented in P-20, No. 263. Similar data from the June 1975 survey are shown in Series P-20, No. 315. Birth expectations and childbearing data from the 1967 Survey of Economic Opportunity were published in P-20, No. 211.

Statistics on fertility by social and economic characteristics from the 1970 census were published in Final Report PC(2)-3A, Women by Number of Children Ever Born, and PC(2)-3B, Childspacing and Current Fertility. A detailed study entitled Perspectives on American Fertility was published in Current Population Reports, Series P-23, No. 70.

For references to census and survey data on fertility for earlier periods, consult the Bureau of the Census catalog. For further information on the availability of these reports, write to:

Data User Services Division Customer Services (Publications) Bureau of the Census Washington, D.C. 20233