4.15 Fertility

Every round of the NLSY79 has included a section on fertility.  In each survey year, both men and women are asked if they have had children.  The NLSY79 is designed so that it is possible to construct a detailed history of each respondent’s fertility. 

The first three rounds of the NLSY79 (1979, 1980, 1981) have very short fertility sections.  In 1979, respondents were asked if they had ever had any children.  For those individuals who answered “yes,” the number of children as well as their birth dates were recorded.  In addition, youths were asked about the total number of children they desired and expected to have.  The 1980 and 1981 surveys updated information for respondents who had any children since the last survey.

In 1982 the fertility data collection was greatly expanded due to additional funding provided by the National Institute of Child Health and Human Development.  During this survey, full retrospective information about the respondent’s fertility history was collected.  Men and women were sent through separate sections.  Men were asked not only the child’s birth date but also the child’s sex, where the child lived, and, if the child was deceased, the date of death.  Women were asked the same information as the men in addition to detailed questions about each pregnancy, enabling researchers to track the wantedness, length, and outcome of each pregnancy.  Information about the interval between each live-birth pregnancy was also recorded, such as whether the couple was using birth control prior to the pregnancy.  Combining these data with questions about whether the pregnancy was wanted provides researchers with the ability to distinguish between planned and unplanned pregnancies.  Male respondents were asked about the wantedness of the first child only.  Finally, all respondents were questioned about current contraceptive practices and their expectations about future fertility.

Beginning in 1983, an even longer “Fertility” section was fielded.  In this larger section, additional questions were asked of all respondents regarding sexual activity and contraceptive use.  Men provided details about children who resided outside the home, such as how far away they lived and how often the respondent visited with them. Women responded to additional questions about their health care during pregnancy, for example, if they smoked or drank during pregnancy.  Additionally, females were asked how often they visited a doctor for prenatal care.  Detailed questions also recorded the time, number, and variety of prenatal procedures such as sonograms and ultrasounds.  And whether an amniocentesis was performed.  This expanded “Fertility” section asked details about the birth, such as length of hospital stay, child’s birth size, and each baby’s immunization record.  Lastly, in 1983, respondents who were noninterviews in 1982 were administered a fertility supplement that mimicked the retrospective fertility section collected at the 1982 interview. 

In 1987 the fertility section began a new pattern.  In odd years, such as 1987, 1989, 1991, and so forth, only a sub-section of the fertility questions was fielded.  In even years, such as 1988, 1990, 1992, and so on, the full set of fertility questions similar to those asked in 1986 was fielded.  With the advent of biennial administration after the 1994 survey, the expanded fertility section has been included in each survey round since 1994.  The odd year fertility sections verified information about previously reported children and asked about the current residence of each child.  Additionally, respondents were asked if they had given birth to any more children since the last survey.  If births had occurred, the name, sex, date of birth, and current living arrangements of the child were recorded.

Since 1986, even year “Fertility” sections have collected detailed information in conjunction with the child assessments and interviews.  To provide users with a detailed understanding of the long fertility section, Table 4.15.1 presents a general outline of the major components of this section of the NLSY79 questionnaire.  Because there are minor variations on content and universes over time, users are advised to review the questionnaires for the survey years of interest prior to undertaking analyses.

Table 4.15.1 Questions Asked of Respondents in Long Fertility Section during Even Numbered Survey Years

In 1984, the NLSY79 began collecting information on abortions through a self-reported confidential card.  This method of collecting sensitive information significantly improved the reporting of this type of pregnancy outcome.  In 2002, these questions were added to the survey itself.  For an evaluation of the abortion information, users should consult Mott (1985).

Questions on nonbiological children were first asked in 1982, were repeated in 1984 and 1985, and have been regularly included in all even year surveys since 1986.  With some variations between survey years, researchers can identify whether a child is step or adopted, whether they are deceased, the sex and birth date of each child, and the child’s usual place of residence.  A detailed written and statistical description of these data can be found in Center for Human Resource Research (1991).

Fertility History

Researchers can create fertility event histories in a number of ways.  One method is to extract the variables from each year’s survey data which record when each child was born.  However, NLSY79 data show that this is not an accurate method for creating an event history.  In each survey, respondents are asked to correct information in the fertility roster.  (Prior to 1993, this was the Children’s Record Form or CRF.  Beginning in 1993, this is the BIO/NBIO Child Roster).  Each year, numerous changes are made.  For example, in the 1994 survey, parents changed some portion of the birth record for 548 children.  While this number appears high, the vast majority of changes are to the spelling of children’s names.

Because the raw recorded data on dates of birth, sex, and status (adopted, deceased) entered on the Children’s Record Form (CRF) are subject to interviewer as well as respondent error, NLSY79 staff created a fertility event history beginning in 1982.  This was done to aid users and, at the same time, evaluate the quality of the NLSY79 fertility data.  This series, found in the “Fertility and Relationship History/Created” area of interest, lists a number of variables including the birth day, month, and year of every child born to NLSY79 respondents; numbers and outcomes of pregnancies; ages of respondents at the birth of selected children; dates of death for children who have died; and usual residence of children.

Users should understand that the “Fertility and Relationship History/Created” area of interest has been part of the NLSY79 data set for many years and consists of edited and constructed variables that incorporate the results of a significant cleaning and editing process begun in the early 1980s.  This effort began as part of an evaluation of the retrospective fertility data collection in 1982/1983 in comparison to base year and updated collections that took place in 1979, 1980, and 1981.  Additional evaluations have been conducted periodically since then.  For further information on the quality of NLSY79 fertility data, users should see Mott et al. (1983) and Mott (1985, 1998). A complete description of the contents of the “Fertility and Relationship History/Created” area of interest is provided in Appendix 5 of the NLSY79 Codebook Supplement and also describes how the data were checked, lists research reports that investigate the quality of the data, and explains special coding and edit flags. 

Age at First Birth

Beginning in 1982, every NLSY79 data release has included a created variable that tracks the age of respondents when they first give birth.  The graph of age at first birth as of the 2002 survey is shown in Figure 4.15.1.  Readers can also find created variables in the “Fertility and Relationship History” area of interest that track how old the respondent was when the second and third births occurred.

Figure 4.15.1 Age at First Birth for NLSY79 Respondents as of the 2002 Interview

Note: Graph created from 2002 variable R76880.00.

Fertility Expectations and Desires

The fertility section provides researchers with an opportunity to examine whether expectations about the future compare accurately with what actually happens in respondents’ lives.  In 1979 and 1982, respondents were asked how many children they desired. 

In 1979, 1982, 1983, 1984, 1985, 1986, and every even year since 1986, respondents have also reported on the number of children they actually expect to have.  Table 4.15.2 compares the number of children desired by the youth at the time the surveys began in 1979 with the number of children born to each respondent by 2006.  The table suggests that typical respondents have had fewer children by 2006 than they desired when they were ages 14–21.

Table 4.15.2 Number of Children Born to NLSY79 Respondents by 2006 vs. Number of Children Desired in 1979

Fertility and Relationship History

The 2000 release of the NLSY79 data introduced a new series of constructed variables for each survey year that provide information about the respondent’s relationship status.  All surveys were examined to match names of spouses or partners for the entire administration of the NLSY79.  The first variable provides information about the respondent’s total number of spouses or partners since 1979 and can range from 0 to 9.  The percentage of respondents who have never reported having a spouse or partner can be found in Table 4.15.3.

Second, a unique code is assigned to each new spouse/partner at the interview date.  If the same spouse or partner resides with the respondent during the next survey round, the spouse or partner code remains the same.  If the respondent has a new spouse or partner, the next available number is given to that person.  If in a later survey round the respondent is reunited with a previous spouse or partner, the number of spouse/partners is not increased, and the code for that year reflects that spouse/partner’s original number . 

Table 4.15.3 Percentage of Respondents Who Have
Never Reported Having a Spouse or Partner

The second created variable reveals whether the person living with the respondent is a spouse, coded 1, or a partner, coded 33.  In some instances, the exact relationship is indeterminate in a given round; in these cases, a code of 36 is used. If no spouse or partner is present at a given survey point, but the respondent has reported a spouse/partner in the past, their relationship code will be zero.  Respondents with no known spouses/partners receive a code of –999. 

There is no guarantee that the same individual(s) were present between survey rounds or that all partners are accounted for, as a spouse or partner may have appeared between survey rounds but not have been present at any survey point.  Early examination suggests that this applies to only a modest proportion of cases. In some instances, identification of spouses who only were present between rounds may be possible by using the NLSY79 marriage history, as well as the marriage transition information available at each survey point.

Survey Instruments and Documentation:  Information is found in the “Fertility” section of each questionnaire.  For additional details on the edited and cleaned fertility data found in area of interest “Fertility and Relationship History/Created,” see Appendix 5 in the NLSY79 Codebook Supplement.

Data Files:  Raw unedited data on fertility, pregnancy, sexual activity, and contraception can be found in areas of interest “Children,” “Child Record Form/Nonbiological,” “Birth Record,” and “Birth Record xxxx.”  Area of interest “Fertility and Relationship History/Created” contains edited and created variables on fertility, pregnancy, and marriage.

Related Topics: See the "Sexual Activity & Contraception" and "Child Care" sections of this guide.  For information on nonbiological children, see the "Household Composition" or "Gender" sections of this guide and areas of interest “Child Record Form/Nonbiological” and “Birth Record xxxx.”

Comparison to Other NLS Cohorts:  Maternal prenatal care information and  health-related characteristics are provided on the NLSY79 Child and Young Adult file and are linked to individual children.

Information on fertility and on the status of children has been regularly collected from the NLSY97 and the Original Cohorts.  Data include number of children, dates of birth, sex, and life status. Consult the BLS website at www.bls.gov/nls or each cohort’s User’s Guide for exact survey years and the types of information available.

References

Mott, Frank L., Paula Baker, R. Jean Haurin, and William Marsiglio, “Fertility Related Data in the 1982 National Longitudinal Survey of Work Experience of Youth:  An Evaluation of Data Quality and Preliminary Analytical Results.” Columbus, OH:  CHRR, The Ohio State University, 1983.

Mott, Frank L. “Evaluation of Fertility Data and Preliminary Analytical Results from the 1983 (Round 5) Survey of the National Longitudinal Surveys of Work Experience of Youth.” Columbus, OH:  CHRR, The Ohio State University, 1985.

Mott, Frank L. “Male Data Collection:  Inferences from the National Longitudinal Surveys.” Columbus, OH:  CHRR, The Ohio State University, 1998.