Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Dec 15.
Published in final edited form as: Youth Soc. 2021 Jul 7;54(8):1377–1401. doi: 10.1177/0044118x211026941

Post-Pregnancy Factors Predicting Teen Mothers’ Educational Attainment by Age 30 in Two National Cohorts

Julie Maslowsky 1, Haley Stritzel 1, Elizabeth T Gershoff 1
PMCID: PMC10723653  NIHMSID: NIHMS1903903  PMID: 38107471

Abstract

Women who begin childbearing as teenagers attain lower levels of education than women who delay childbearing until age 20 and later. Little is known about post-pregnancy factors that predict educational attainment among teen mothers. The current study examined whether teen mothers’ environment and experiences 2 years after their first birth contribute to their educational outcomes by age 30, net of selection factors associated with teenage childbearing. Data were from two cohorts, the National Longitudinal Surveys of Youth 1979 (N = 241) and 1997 (N = 378). Multinomial logistic regression modeling was used to assess associations of post-pregnancy factors with teen mothers’ educational attainment. Having child care was associated with increased odds of attaining a high school diploma and of attending college in both cohorts. Providing regular and subsidized child care for teen mothers is an opportunity to support teen mothers in achieving higher levels of educational attainment.

Keywords: teenage childbearing, educational attainment

Introduction

Half of women who begin childbearing as teenagers graduate from high school, compared to 90% of women who do not have a teen birth (Manlove & Lantos, 2018). Women’s educational attainment and degree type (e.g., high school diploma vs. GED) are strong predictors of health and wellbeing across the life course for themselves and their children (Hendrick & Maslowsky, 2019; Zajacova & Everett, 2014). Hence, supporting teen mothers’ educational attainment may be a potential mechanism for breaking the intergenerational cycle of poor health and social outcomes commonly associated with adolescent childbearing. Understanding influences of teen mothers’ educational success is essential for informing effective programming and policies to support teen mothers’ educational pursuits. In the present study, we sought to identify modifiable determinants of educational attainment among two cohorts of U.S. women who first gave birth prior to age 20.

The Educational Penalty of Teenage Childbearing

Teen mothers attain fewer years of education and are less likely to obtain a high school diploma than women who delay childbearing beyond the teenage years (Mollborn, 2007). Based on data from the National Survey of Family Growth, 53% of women ages 20 to 29 who began childbearing before age 18 earned a high school diploma compared to 90% who delayed childbearing until age 20 or later. Although some teen mothers do earn a GED, the combined diploma and GED rate for teen mothers (70%) is below that for nonparent teen women (94%) (Manlove & Lantos, 2018). Teen mothers also attain fewer overall years of education, including college, than women who delay childbearing (Kane et al., 2013). After accounting for differences in background characteristics, mothers who gave birth before age 19 in the National Longitudinal Study of Adolescent Health completed nearly two fewer years of education by ages 24 to 34 than mothers who delayed childbearing until age 20 (Kane et al., 2013).

Pre-Pregnancy Selection Factors and Teen Mothers’ Educational Attainment

The association of teenage childbearing with adverse social outcomes later in life, including lower educational attainment, is well-established. However, debate remains as to whether this is a result of parenting as an adolescent (post-pregnancy factors) or pre-pregnancy selection factors, that is, background characteristics that are predictive of teen childbearing (Jaffee et al., 2001). Selection hypotheses attribute the association to pre-existing characteristics of teen mothers and their environments, rather than the experience and sequelae of teen motherhood itself. Conversely, social influence hypotheses posit that post-pregnancy experiences stemming directly from parenting as a teenager primarily explain associations with adverse outcomes.

Pre-pregnancy selection factors associated with both teenage childbearing and lower educational attainment include academic, family, residential, and geographic factors. Academic factors include the teen’s academic aptitude (Lou & Thomas, 2015), attitude toward school (Coleman & DeLeire, 2003), years of education completed prior to giving birth, and her parents’ educational attainment (Ou & Reynolds, 2008). Family and residential factors include the number of siblings the teen has and whether she resides with her parents or her partner prior to and after giving birth (Mollborn, 2010). Geographic factors include living in the Southern United States and living in a rural area (Maslowsky et al., 2019). Household poverty status and racial/ethnic minority status are also associated with higher rates of teenage childbearing and lower educational attainment (Upchurch, 1993).

Some evidence supports the hypothesis that women’s background characteristics primarily explain the adverse outcomes associated with teenage childbearing (Geronimus et al., 1994). Other evidence shows that even after accounting for background characteristics, teen mothers are more likely to be unemployed, live in poverty, and receive public assistance than their peers who delay childbearing until their 20s (Assini-Meytin & Green, 2015; Kane et al., 2013; Lee, 2010; Mollborn, 2007).

Post-Pregnancy Factors and Teen Mothers’ Educational Attainment

While the educational penalty for teen motherhood is well documented (Manlove & Lantos, 2018; Kane et al., 2013; Perper et al., 2010), effective strategies for mitigating this penalty are less understood. The variation in educational outcomes among teen mothers of similar backgrounds suggests that their life contexts after birth have implications for their educational attainment (Mollborn, 2010, 2007; Manlove et al., 2000; SmithBattle, 2007). However, previous research identifying specific, modifiable factors for teen mothers’ educational experiences is limited. Three main post-pregnancy influences of teen mothers’ education beyond selection factors have been identified in prior studies: (1) household composition; (2) material resources; and (3) whether subsequent childbearing is delayed.

Regarding household composition, teen mothers who live with their spouse or romantic partner are less likely to graduate high school than teen mothers living with their parents (Mollborn, 2010). This may be due to practical, emotional, and financial support received from parents and other family members in navigating the dual role of adolescent student and new parent. Practical support from family may come in the form of providing child care while a parenting teen attends school or completes homework, helping a teen mother learn how to care for an infant, or providing transportation to school and healthcare appointments. Conversely, teen mothers who cohabit with their romantic partner may be simultaneously juggling the responsibilities of caring and providing for a household, in addition to the role of new parent. Over and above household composition, access to financial and material resources (including child care, transportation, and housing) facilitates teen mothers’ educational persistence during pregnancy and after birth (Mollborn, 2007; SmithBattle, 2007). Lastly, likely in part due to the additional resources required for caring for multiple children, teen mothers who have “rapid repeat” births (subsequent births occurring within 2 years) attain less education than teen mothers who delay subsequent childbearing (Manlove et al., 2000).

Educational Attainment across Multiple Cohorts of Teen Mothers

The current study examined what factors predict educational attainment among two cohorts of teen mothers nearly 20 years apart—one that was initially assessed in 1979 and a second that was initially assessed in 1997. In replicating our model across two cohorts, we are able to determine whether the post-pregnancy factors that predict attainment among this population of youth have changed over historical time, and thus may be cohort- or generation-specific, or whether they are consistently predictive over time, which would indicate factors that are more universally predictive of attainment and thus potentially more reliable targets for intervention. We know that rates of teen childbearing in the U.S. have declined over time and, with those changes, the demographic make-up of teen mothers has changed (Child Trends, 2019). The most dramatic change is in marital status: The percentage of teen mothers who were unmarried increased from around 30% in the 1970s to 89% in 2013 (Ventura et al., 2014). Further, in the 1970s and 1980s rates of teenage childbearing were highest among Non-Hispanic Black teens; beginning in the 1990s, rates have been highest among Hispanic teens (Kearney & Levine, 2015).

Social service availability and eligibility for teen mothers has also shifted historically. Of particular note, the welfare reform of 1996 resulted in stricter eligibility requirements for receiving cash aid public assistance (Duffy & Levin-Epstein, 2002). In order for minors who were parents to receive the newly created Temporary Assistance for Needy Families (TANF), they had to live with a parent or adult guardian and participate in a school or training program; the previous income assistant program, known as Aid to Families with Dependent Children (AFDC), had not had such a requirement. As a result, the proportion of teen mothers receiving welfare dropped from 25% in 1997 to 5% in 1999 (Acs & Koball, 2003). The two samples of teen mothers examined in this study thus had their first children under different policy regimes—the 1979 cohort was eligible for AFDC while the 1997 cohort was eligible for TANF.

Considering historical changes in demographic characteristics and social services available to teen mothers in the U.S., identifying post-pregnancy factors consistently associated with increased educational attainment across cohorts and across time may shed light on important points of intervention that can effectively support teen mothers’ educational pursuits.

The Current Study

The current study addressed two research questions: (1) To what extent do post-pregnancy factors contribute to teen mothers’ degree type attained by age 30 net of pre-pregnancy selection factors?, and (2) What are the consistent post-pregnancy factors predicting teen mothers’ educational attainment across two national cohorts? We expected that post-pregnancy factors (i.e., receiving public assistance, household poverty status, living with a romantic partner, employment status, use of child care, and giving birth to an additional child) would predict educational outcomes at age 30 net of selection factors (i.e., background characteristics and educational experiences prior to giving birth) for teen mothers in both cohorts. We assessed post-pregnancy factors 2 years after birth for both theoretical reasons and data availability. Specifically, the first year after a first birth is a time of great transition and adjustment, so we aimed to capture teen mothers’ contextual environments after the initial adjustments associated with new motherhood had passed. Finally, due to the timing of interviews and the fact that some data were only collected every 2 years in NLSY, some data on post-pregnancy factors were first available 2 years after women’s first births.

Figure 1 illustrates the conceptual model of the associations of post-pregnancy factors with teen mothers’ educational attainment by age 30, net of pre-pregnancy background characteristics and educational experiences that are associated with both early childbearing and women’s educational attainment. We examined how post-pregnancy factors are associated with teen mothers’ educational attainment across two national cohorts of women: women who were 30 years old in the 1990s (National Longitudinal Survey of Youth (NLSY) 1979), and women who were 30 years old in the 2010s (NLSY 1997 cohort). NLSY79 and NLSY97 are ideal data to test the research questions as they contain rich contextual information about women’s lives before and after their first births. These surveys were specifically designed to support cross-cohort research: both use a similar sampling design and similar questionnaires, particularly for questions related to socioeconomic attainment (Bureau of Labor Statistics, n.d.-a).

Figure 1.

Figure 1.

Open in a new tab

Conceptual model of combined contribution of selection and post-pregnancy on teen mothers’ educational attainment by age 30.

The current study addresses limitations of past research in several ways. First, some research noting the role of post-pregnancy factors in teen mothers’ educational outcomes neglects to account for selection factors that may explain teen mothers’ lower educational attainment compared to other women (Perper et al., 2010). Further, in some previous research examining teen mothers’ educational outcomes, all women who first give birth before age 20 are considered teen mothers (Mollborn, 2007; Mollborn & Blalock, 2012). However, including women who already graduated or dropped out of high school prior to giving birth when examining influences of teen mothers’ educational attainment potentially neglects to account for the inherent endogenous relationship of high school educational experiences and timing of childbearing. Many women who give birth as teenagers had poor academic achievement and/or dropped out of high school prior to pregnancy (Lou & Thomas, 2015; Marcotte, 2013; Mahler, 1999; Manlove, 1998) and thus, their dropout was not due to their experience as a pregnant or parenting teen. The current study excludes women who dropped out or had already graduated high school during the interview year prior to the year of their first birth.

Previous research examining teen mothers’ educational outcomes only assesses outcomes through their early 20s (Mollborn & Blalock, 2012; Perper et al., 2010). However, teen mothers’ formal schooling may be interrupted around the time of their birth and therefore not completed until later in the life course than women who delay their childbearing until their 20s and beyond. Therefore, our study assesses educational outcomes through approximately age 30.

Finally, much of the research examining teen mothers’ educational outcomes compares their experiences to those of non-teen mothers. Another important question concerns the predictors of educational experiences within samples of teen mothers. Some predictors of educational persistence and success for teen mothers are unique from those of non-parent teens, such as the need for child care. Examining the predictors of educational attainment among teen mothers specifically is key to informing policies and programming aimed toward pregnant and parenting teens.

Methods

Participants

Data are from the National Longitudinal Surveys of Youth 1979 (NLSY79) and 1997 (NLSY97) (US Bureau of Labor Statistics, n.d.-b). The NLSY79 is a national survey of 12,686 youth ages 14 to 22 living in the United States in 1979. Participants were interviewed annually from 1979 to 1994 and then biennially through 2012. The NLSY97 began in 1997 and was designed to continue the work of the NLSY79 by studying the labor market and educational experiences of a more recent cohort; thus the sampling designs and questionnaires between the two surveys are similar. The 8,984 respondents in the NLSY97 were born between 1980 and 1984 in and were first interviewed in 1997 (when they were approximately ages 12–18). Follow-up surveys were conducted approximately every year through 2015 to 2016.

Our analytic sample included 241 women from NLSY79 and 378 women from NLSY97 who had their first birth prior to age 20 and after their Wave 1 interview, who were enrolled in high school in the year prior to the year of their first birth, who reported educational outcome information at the interview conducted closest to when they were age 30, and whose child lived with them 2 years after birth (see Figure 2 for sample selection information). Only teen mothers who reported a first birth after their Wave 1 interview were included in order to include selection factors prior to the mothers’ first births in analyses. Only teen mothers who were enrolled in school the year prior to their first birth were included in order to address potential endogeneity in the association between educational experiences during high school, teen childbearing, and long-term educational achievement.

Figure 2.

Figure 2.

Open in a new tab

Selection of analytic sample of NLSY79 and NLSY97 participants.

As NLSY sampling procedures were not intended to create a representative sample of teen mothers in the United States, NLSY sampling weights were not used in the current study per NLSY recommendations (US Bureau of Labor Statistics, n.d.-c). As such, the current study does not represent a nationally representative sample of teen mothers, but rather a geographically, racially, and ethnically diverse sample of women in the United States who gave birth as adolescents between 1979 and 1984 or between 1997 and 2004. The current study was deemed not to be human subjects research by the Institutional Review Board of the sponsoring university.

Measures

Dependent Variable: Degree Type Attained by Age 30

The dependent variable of interest was the degree type teen mothers attained by approximately age 30. Due to the timing of interviews, not all teen mothers reported educational outcomes at exactly age 30; some completed the target interview when they were slightly older or younger. Thus, education was measured at the interview year in which the respondent was 30 or at the previous interview before turning 30.

In each interview, participants reported the highest grade they had completed, as well as whether they earned a GED or a high school diploma. Combining these two variables created a categorical overall educational attainment variable with four mutually exclusive categories: no degree, GED, high school diploma, or at least some college (at least 13 years of education).

Selection factors: Background characteristics and educational experiences prior to first birth.

Unless otherwise specified, all selection factors come from reports at Wave 1 and data collection procedures were the same for the 1979 and 1997 cohorts.

Age at first birth and at educational attainment report.

Respondent’s age at first birth and their age at reporting educational outcomes was controlled for in all models.

Race and ethnicity.

Racial and ethnic categories in NLSY surveys are “Hispanic,” “Black,” and “non-Black/non-Hispanic.” The “non-Black/non-Hispanic” category included those not identified as “Hispanic” who were identified as “White” or “Japanese, Chinese, Vietnamese, Asian Indian, Native American, Korean, Eskimo, Pacific Islander, or of another race besides black or white.” In addition, the NLSY97 includes a category for multiple races.

Household poverty.

NLSY created a dichotomous variable capturing whether participants lived below the federal poverty line in the calendar year prior to the Wave 1 interview.

Geographic region.

NLSY coded the county participants lived in at age 14 as in a “Southern” or “Non-Southern” region of the country. Similarly, NLSY97 coded the Census region in which a participant lived in each year, allowing for the identification of residence in the South for respondents who were age 14 or younger at the start of the survey. For participants who were older than 14 at the start of the survey, Southern residence was identified retrospectively.

Lived with parents at age 14.

Participants described the adults they lived with at age 14 as their parent, stepparent, other relative, or other man or woman not related. This variable was dichotomized into two categories: living with both biological parents and all other arrangements, as the absence of the biological father in particular has been consistently linked with a higher risk of teenage pregnancy (Coyne & D’Onofrio, 2012).

Lived in rural community.

NLSY coded the county and state of participants’ residences at the Wave 1 interviews as rural or non-rural according to the percent urban population of the county.

Maternal education.

Participants reported the highest grade or year of regular school their own mothers had completed, ranging from 0 to 20 years of education.

Number of siblings.

In NLSY79, participants reported the number of living brothers and sisters they had in their Wave 1 interview. As a proxy for number of siblings, participants in NLSY97 reported the number of children in the household under age 18 at Wave 1.

Household composition at year of first birth.

Two dichotomous variables captured whether the mother lived with her parents and whether she lived with a spouse or romantic partner at the year of her first birth.

Schooling completed by year of first birth.

In both surveys, participants reported how many years of school they had completed. As teen mothers in the present study were ages 14 to 17 at their first birth, their age at first birth is controlled for in all models to account for the expected differences in years of schooling completed between ages 14 and 17.

Attitudes about school.

Participants enrolled in school at Wave 1 in the NLSY79 or in the NLSY97 were each asked a comparable set of four questions regarding how they felt about their school; both sets of items were on a 4-point scale with low values meaning more agreement with the statement (79: 1 (very true) to 4 (not true at all); 97: 1 (strongly agree) to 4 (strongly disagree). Each cohort was asked about the quality of their teachers (79: “the teachers know their subjects well”; 97: “the teachers are good”), the engagement of their teachers (79: “the teachers were willing to help personally”; 97: “the teachers are interested in the students”), the extent of student misbehavior (79: “the students can ‘get away with’ anything” (reverse scored); 97: “there is a lot of cheating on tests and assignments” (reversed scored)), and whether they feel safe at the school (79: “did not feel safe” (reverse scored); 97: “felt safe at this school”). We calculated the attitudes toward school variable within each cohort as the mean of these four items; higher scores indicated more positive attitudes.

Academic aptitude.

The aptitude test scores are from the unofficial Armed Forces Qualifications Test (AFQT) administered to all NLSY79 participants in 1980 assessing arithmetic reasoning and reading comprehension. Instead of the AFQT, NLSY97 participants competed the Armed Services Vocational Aptitude Battery (ASVAB) during the first wave of the survey. NLS staff adjusted both AFQT and ASVAB scores to account for age at the time of the test and converted the scores into percentiles.

Independent Variables of Interest: Post-Pregnancy Factors

The post-pregnancy variables of interest included contextual factors and life events occurring approximately 2 years after the mothers’ first births that have previously been theoretically and empirically linked to teen mothers’ educational outcomes. These factors were drawn from the first available interview at least 2 years after the participant’s first birth.

Household poverty.

As described above with pre-pregnancy household poverty, NLSY reported whether participants lived below the federal poverty line in the calendar year before the interview.

Public assistance.

A dichotomous variable indicated whether the respondent received any public assistance from any government program in the calendar year before the interview (e.g., AFDC, Supplemental Security Income, or any “other public assistance”) and/or unemployment compensation received by herself and her opposite sex partner (if applicable).

Household composition.

Two dichotomous variables captured whether the respondent lived with one or both of her parents and whether she lived with a spouse or romantic partner.

Employment status.

Participants described their current employment status at each wave and NLS investigators coded this status according to standardized categories of labor force participation created by the U.S. Census Bureau: employed, unemployed, out of the labor force, or in active forces (US Bureau of Labor Statistics, n.d.-d). According to these definitions, teen mothers in the NLSY79 were considered “employed” if they reported having a job (including full or part-time) or being in the armed forces and “not employed” if they were unemployed or out of the labor force (i.e., keeping house, going to school, or unable to work) in the past week. Participants who had a job but were temporarily absent (e.g., due to illness) were also considered “currently employed.” In the NLSY97, a comparable labor force participation variable was not available for all waves. Instead, in order to be consistent with the NLSY79, we considered respondents in the NLSY97 to be employed each year if they reported that they were currently working (part- or full-time) with a valid employer; otherwise they were considered not employed.

Use of child care.

In the NLSY79, questions regarding child care arrangements were asked retrospectively starting in 1986 about participants’ children during their first 3 years of life. The question used in the current study was, “In the 2nd year of (child’s) life, was (he/she) cared for in any regular arrangement such as a baby sitter, relative, day care center, nursery school, play group, or some other regular arrangement?.” Child care arrangements were measured slightly differently in NLSY97. If a respondent was in school or working, she was asked in what type of child care arrangement each child spent the most time in 2001 and all of the types of child care arrangements she used in a typical week starting in 2002. No questions about child care were asked prior to 2001. Respondents were coded as having child care if they reported using any child care arrangement (e.g., relatives or child care center) other than in their own care. Note that this operationalization differs slightly from NLSY79 variable, which captures regular child care as opposed to any child care. If a respondent was not in school or working, she was coded as not having child care if she reported that the main reason she was not looking for a job was due to a lack of child care.

Additional births.

Teen mothers reported the month and year of all births, allowing for the creation of a dichotomous variable indicating whether or not the mother gave birth again within 2 years after her first birth.

Analysis

The primary analysis consisted of multinomial logistic regression models predicting degree type attained by age 30 were conducted separately by cohort with odds ratios presented for ease of interpretation. Prior to conducting the multinomial logistic regressions, a principal component analysis (PCA) of all selection variables was conducted to reduce the number of estimated parameters (Coley et al., 2016). Specifically, a PCA of a polychoric correlation matrix was used, as the covariates included categorical as well as continuous variables (Kolenikov & Angeles, 2009). PCA with an orthogonal varimax rotation of component loadings resulted in three principal components in each sample that explained 51% of the variance in selection variables in NLSY79 and 49% of variance in NLSY97. Principal component scores for the three retained components were generated for each participant and used in analytic models. All data preparation and analyses were conducted in Stata 14.0 (StataCorp, 2015).

Missing data.

No data were missing for the educational outcome variables as all participants missing educational outcome information were dropped from the analytical sample (see Figure 2). In NLSY79, missing data ranged from 0% to 15% for selection and post-pregnancy variables. Pre-Wave 1 poverty variables and poverty 2 years post-birth had 15% and 10% of data missing, respectively. All other variables had less than 7% of data missing. In NLSY97, missing data ranged from 0-31% for selection and post-pregnancy variables. Family background variables such as poverty status and household composition at age 14 had the highest rates of missing data, at 31% and 27% respectively. About 22% of respondents had missing data on the ASVAB because they chose not to participate. All other variables had less than 12% of data missing. To address missing data, 25 multiply imputed datasets were created for NLSY79 and NLSY97 samples separately. Multiple imputation addresses missing data by creating copies of the dataset with missing values imputed based on selected covariates; regression analyses are then run separately by dataset and parameter estimates of interest are combined together to give a single estimate (Royston, 2004). All analyses were run with Stata’s mi estimate command to compute pooled coefficient and variance estimates with information from the 25 multiply imputed datasets. This method of addressing missing data leads to less bias in parameter estimates compared to list-wise deletion or mean imputation (Enders, 2013).

Sensitivity analysis.

Due to the wording of the NLSY employment items, which do not separate part time and full time employment, we were not able to distinguish part time and full time work in our analysis. To test for differences associated with heterogeneity of employment arrangements, we conducted sensitivity analyses in which we replaced the dichotomous employment variable with the percentage of weeks employed (either full or part time) during the past year. The results of those analyses were substantively the same as those reported below.

Results

Differences in Demographic Characteristics of Cohorts

Table 1 presents descriptive characteristics of the NLSY79 (N = 241) and NLSY97 (N = 378) samples and the results from statistical comparisons of these characteristics across the cohorts using t-tests or chi squares. The analytic samples of teen mothers in NLSY79 and NLSY97 cohorts significantly differed on our outcome of educational attainment by age 30 (X2 [3, N = 619] 9.445, p = .024). The overall difference in attainment appears to be driven by differences in the percent that have no degree (79: 15.4%; 97: 21.4%) and the percent that have at least some college (79: 16.6%; 97: 9.3)

Table 1.

Unweighted Descriptive Characteristics of Analytic Samples from the National Longitudinal Surveys of Youth 1979 and 1997.

NLSY79 (N = 241)
 NLSY97 (N = 378)
Mean or proportion SD Mean or proportion SD t or chi square p Value
Degree type at age 30 9.45 .024*
 No degree 15.4 21.4
 GED 19.9 19.8
 HS diploma 48.1 49.5
 At least some college 16.6 9.3
Selection (background characteristics prior to first birth)
 Age at first birth 17.4 1.1 17.1 1.3 −3.01 .002**
 Age reported degree attainment 29.4 1.2 29.4 0.7 0.12 .908
 Race/ethnicity 10.78 .013*
  Hispanic 18.3 26.5
  Black/non-Hispanic 53.5 44.2
  Non-Black/non-Hispanic 28.2 27.8
  Multiple races – – 1.6
Household under federal poverty level prior to Wave 1 46.8 44.0 0.37 .545
Lived in south at age 14 50.2 47.5 0.42 .518
Lived with both parents age at age 14 50.2 28.0 26.79 <.001***
Lived in rural community at Wave 1 24.8 21.2 1.00 .318
Maternal education (0–20) 9.7 2.7 11.3 2.6 6.82 <.001***
Number of siblings at Wave 1 4.6 3.1 2.8 1.6 −9.93 <.001***
Lived with parent(s) at year of first birth 76.8 67.7 5.87 .015*
Lived with spouse/partner at year of first birth 17.4 26.6 6.77 .009**
Highest grade completed by year of first birth (1–12) 10.2 1.1 10.2 1.5 −0.29 .775
Attitudes about school (1–4) 3.2 0.5 2.7 0.5 −12.39 <.001***
AFQT aptitude percentile score in 1980 (1–100) 26.6 21.4 – –
ASVAB percentile score in 1997 (1–100) – – 27.4 21.4
Post-pregnancy (context 2 years after birth)
Household under federal poverty level 43.5 48.0 1.10 .294
Receiving public assistance 49.0 76.2 47.95 <.001***
Living with spouse/partner 43.3 35.7 3.59 .058
Living with parent(s) 37.1 42.9 2.03 .154
Currently employed 30.7 65.5 68.19 <.001***
First child in child care 36.9 83.7 134.48 <.001***
Have given birth to additional child(ren) 22.8 20.9 0.32 .571
Open in a new tab

Note. Significant differences across cohorts based on t-tests and chi-square tests are reported for every variable present in both samples.

*

p < .05.

**

p < .01.

***

p < .001.

There were also statistically significant differences between the cohorts in selection factors. Teen mothers in NLSY97 were significantly younger at their first births (17.1 compared to 17.4 years old). The NLSY97 sample consisted of a higher proportion of Hispanic teen mothers (26.5% vs. 18.3%) and lower proportion of Black/Non-Hispanic teen mothers (44.2% vs. 53.5%) than the NLSY79 analytic sample. Significantly fewer of the NLSY97 cohort than the NLSY79 cohort lived with both parents at age 14 (28.0% vs. 50.2%). They had significantly fewer siblings (2.8 vs. 4.6) and their own mothers had significantly more years of education (11.3 years vs. 9.7 years). Teen mothers of the NLSY97 cohort were significantly more likely than the NLSY79 cohort to live with their romantic partner (26.6% vs. 17.4%) and less likely to live with their parents (67.7% vs. 76.8%) the year when they first gave birth. There was no significant difference in years of school completed by year of first birth, with teen mothers from both cohorts having completed 10.2 years of school on average prior to their first births. Teen mothers from NLSY97 also had significantly less positive attitudes about their schools than those from NSLY79. Age at reporting degree attainment, rural residence, family poverty, southern residence, and highest grade completed by year of first birth did not differ significantly by cohort.

In terms of post-pregnancy factors, the NLSY97 cohort was significantly more likely to have received public assistance 2 years after their first births (76.2% vs. 49.0%) and to have been currently employed (65.5% vs. 30.7%). Compared to 36.9% of teen mothers in the NLSY79, 83.7% of teen mothers in the NLSY97 reported that their first child was in child care 2 years after their birth, although this difference could be due in part to how the two surveys asked about child care. Teen mothers’ household poverty status, whether they were living with any parent or spouse/partner, and whether they had given birth to additional children did not vary across cohorts.

Degree Type Attained by Age 30 in the NLSY79

Table 2 presents results from multinomial logistic regression models predicting teen mothers’ degree type. Given that “no degree” was assigned as the reference category for all models, results indicate the odds of attaining a GED, high school diploma, or completing at least some college compared to not receiving any degree by age 30. After accounting for other selection and post-pregnancy factors, teen mothers who had their first child in child care (vs. those who did not) had over six times the odds of receiving a high school degree (OR = 6.74, 95% CI [2.00, 22.69], p = 0.002) and over four times the odds of completing at least some college (OR = 4.90, 95% CI [1.29, 18.59], p = 0.02). Teen mothers who had a rapid repeat birth had an 80% reduction in odds of completing a high school diploma (OR = 0.20, 95% CI [0.07, 0.57], p = 0.003) and a 70% reduction in odds of completing at least some college (OR = 0.30, 95% CI [0.18, 1.00], p = 0.050), compared to those who did not have a subsequent birth.

Table 2.

Multinomial Logistic Regression Models Predicting Teen Mothers’ Degree Type Attained by Age 30.

NLSY79 (N = 241)
GED
 High school diploma
 At least some college
Selection factors RRR SE 95% CI RRR SE 95% CI RRR SE 95% CI
 Principal component #1 0.96 0.28 0.54 1.70 2.65*** 0.76 1.52 4.64 1.89* 0.60 1.02 3.51
 Principal component #2 1.98** 0.49 1.22 3.22 1.91** 0.44 1.22 3.00 1.68* 0.42 1.03 2.76
 Principal component #3 1.03 0.41 0.48 2.24 1.69 0.62 0.83 3.47 1.19 0.49 0.53 2.68
Post-pregnancy factors
 Household in poverty 2.68 1.75 0.75 9.61 1.36 0.82 0.42 4.44 0.97 0.66 0.25 3.69
 Receiving public assistance 0.58 0.34 0.18 1.85 0.74 0.42 0.25 2.23 1.76 1.11 0.51 6.07
 Living with romantic partner 1.26 0.87 0.32 4.89 1.01 0.67 0.27 3.72 0.92 0.68 0.21 3.95
 Living with parents 0.41 0.26 0.12 1.43 0.85 0.52 0.26 2.79 1.11 0.74 0.30 4.13
 Currently employed 0.51 0.34 0.14 1.85 0.37 0.22 0.11 1.21 0.92 0.61 0.25 3.39
 First child in child care 2.34 1.59 0.62 8.89 6.74** 4.17 2.00 22.69 4.90* 3.33 1.29 18.59
 Gave birth to additional child 0.91 0.48 0.33 2.53 0.20** 0.11 0.07 0.57 0.30* 0.18 0.09 1.00
NLSY97 (N = 378)
 Principal component #1 1.31 0.25 0.89 1.91 0.96 0.16 0.69 1.33 1.23 0.32 0.74 2.06
 Principal component #2 1.44 0.42 0.81 2.56 1.80* 0.46 1.09 2.96 1.40 0.53 0.67 2.92
 Principal component #3 1.06 0.16 0.79 1.43 1.97*** 0.28 1.49 2.60 2.43*** 0.55 1.56 3.79
Post-pregnancy factors
 Household in poverty 1.05 0.44 0.46 2.36 0.60 0.22 0.29 1.23 0.35 0.19 0.12 1.03
 Receiving public assistance 0.67 0.30 0.29 1.59 1.16 0.46 0.54 2.51 1.58 0.90 0.52 4.80
 Living with romantic partner 0.60 0.28 0.24 1.47 0.54 0.22 0.25 1.19 0.19** 0.12 0.05 0.67
 Living with parents 0.92 0.37 0.42 2.04 0.68 0.25 0.33 1.40 1.02 0.56 0.35 2.98
 Currently employed 0.76 0.29 0.36 1.62 1.73 0.61 0.87 3.44 1.27 0.69 0.44 3.66
 First child in child care 1.52 0.66 0.65 3.54 2.67* 1.11 1.19 6.03 10.21* 11.22 1.18 87.98
 Gave birth to additional child 0.70 0.29 0.32 1.57 0.57 0.21 0.28 1.17 0.34 0.22 0.09 1.22
Open in a new tab

Note: “No degree” is the reference category. NLSY79 = National Longitudinal Survey of Youth 1979; NLSY97 = National Longitudinal Survey of Youth 1997;

GED = General Education Development/High School Equivalency Certificate; RRR = relative risk ratios; SE = standard error; 95% CI = lower and upper bounds of 95% confidence intervals shown. All models control for age when reporting educational outcomes.

*

p < .05.

**

p < .01.

***

p < .001.

Degree Type Attained by Age 30 in NLSY97

Similarly to the NLSY79 results, having some form of child care was associated with over twice the odds of competing a high school degree (OR = 2.67, 95% CI [1.19, 6.03], p = 0.02) and over ten times the odds of completing at least some college (OR = 10.21, 95% CI [1.18, 87.98], p = 0.04). Additionally, living with a romantic partner 2 years after birth was associated with an 81% reduction in odds of completing some college (OR = 0.19, 95% CI [0.05, 0.67], p = 0.01).

Discussion

The current study examined the extent to which teen mothers’ environment and experiences 2 years after their first birth (post-pregnancy factors) were associated with their educational outcomes by age 30 above and beyond factors demonstrated to select women into teenage childbearing (selection factors). This study also sought to identify whether any such factors were consistent across two national cohorts of teen mothers. Our results show significant associations between some post-pregnancy factors, access to child care in particular, and educational outcomes.

Despite differences in demographic characteristics across the cohorts, one post-pregnancy factor significantly predicted the lifetime educational attainment of women across both cohorts: use of child care. Child care was associated with higher odds of graduating high school and completing at least some college for teen mothers of both the NLSY79 and NLSY97 cohorts. Additional post-pregnancy factors related to educational attainment in one of the two cohorts. In NLSY79, having an additional child within 2 years of the first birth was associated with significantly reduced odds of earning a high school diploma or completing at least some college. In NLSY97, living with a romantic partner was associated with lower odds of completing at least some college. The findings related to the educational benefits of child care and educational penalties of rapid repeat childbearing demonstrate that some post-pregnancy experiences directly related to motherhood continue to predict teen mothers’ educational outcomes after considering their backgrounds and prior educational experiences.

These findings are consistent with previous research noting the importance of some post-pregnancy factors on health and social outcomes associated with teenage motherhood (Jaffee et al., 2001). In particular, access to nonparental child care early in a child’s life has specifically been associated with improved educational outcomes for mothers from disadvantaged backgrounds, including teen mothers (Gordon et al., 2004; Mollborn & Blalock, 2012; Ramey et al., 2000). Findings from the present study are also consistent with previous research demonstrating rapid repeat childbearing among teens to be associated with adverse outcomes, including educational outcomes (Manlove et al., 2000).

A good deal of previous research has focused either on pre-pregnancy or post-pregnancy factors as predictors of teen mothers’ educational attainment or has tested either/or hypotheses regarding pre- and post-pregnancy factors (e.g., Jaffee et al., 2001). Our results show that post-pregnancy factors predict teen mothers’ later educational attainment after adjusting for known association between pre-pregnancy factors and educational attainment. Thus, it is clear that both pre- and post-pregnancy factors are important and should be factored into our understanding of educational attainment in teenage mothers and our efforts to intervene. Pre-pregnancy factors can be targeted in prevention programs, while post-pregnancy factors can be incorporated into programming to support educational attainment among women who had their first child as a teenager.

Although the two NLSY studies were not designed to be nationally representative samples of teen mothers, the overall studies were in fact nationally representative (US Bureau of Labor Statistics, n.d.-a). Thus, our teen subsamples reflect the country’s geographic, racial, and ethnic diversity and increase our confidence that our findings are not specific to one demographic group or region. Our study attempted to identify factors that are broadly influential in teenage mothers’ educational attainment in order to suggest potential intervention strategies in this diverse population.

Further, the consistent association of access to child care with higher educational attainment across two historical cohorts demonstrates the robustness of child care as a support for teen mothers’ education. During the two decades separating the NLSY79 and NLSY97 cohorts, the demographic landscape of teenage motherhood shifted. Although teenage birth rates decreased from 52.3 births per 1,000 women ages 15 to 19 in 1979 to 51.3 births in 1997 (Ventura et al., 2014), the demographic composition of teenage mothers shifted as more births occurred to Hispanic and unmarried mothers and fewer to Black and married mothers (Kearney & Levine, 2015; Ventura et al., 2014). The policy environment also shifted, as the predominant welfare policy shifted from AFDC to TANF and resulted in fewer teenage mothers receiving welfare support. Across these shifts, teenage mothers continued to have lower educational attainment than their peers who began childbearing later and to experience the lifelong socioeconomic consequences associated with lower educational attainment. There is great value in understanding the outcomes of a high-risk population under a variety of historical and policy contexts in order to inform future policy decisions. A factor that is continuously predictive of positive outcomes despite demographic and policy shifts is a strong candidate for future policies to target. We have identified such a factor: access to child care after a teenager gives birth.

Strengths and Limitations

The present study has many strengths. It draws from two geographically, racially, and ethnically diverse national samples of teen mothers and contains rich information on the selection and post-pregnancy factors potentially impacting their educational attainment. It also follows teen mothers through approximately age 30 to capture the educational attainment of women who may have had an interruption in their schooling due to pregnancy, childbearing, and/or childrearing, but returned to school and continued their education later in adulthood.

Further, the findings are noteworthy as they demonstrate the variations in degree type associated with teen mothers’ experiences with childrearing not detectable in studies that solely examine number of years of schooling attained. Our finding that providing child care and delaying subsequent births is associated with higher likelihood that teen mothers complete high school or some college, but not a GED, is especially important. Teen mothers who are able to complete high school or some college instead of a GED are likely to have better health and socioeconomic outcomes compared to those who completed only a GED ((Liu et al., 2013; Zajacova, 2012; Zajacova & Everett, 2014). Access to child care and resources for avoiding rapid repeat pregnancy are potentially salient points of intervention.

This study also has limitations and presents opportunities for future research. First, some variables were measured differently across NLSY79 and NLSY97. It is possible that variations across cohorts may be due in part to differences in construct measurement. Different measurement across cohorts also made it impossible to compare the associations of different types of child care (e.g., relative vs. center-based care) with teen mothers’ educational attainment. Further, NLSY lacked information on resources available for pregnant and parenting students at their most recent schools. Such information would help to improve understanding of the effectiveness of existing resources for promoting long-term educational outcomes. Second, our inclusion criteria included that teen mothers had not dropped out or completed high school prior to pregnancy. This allowed us to avoid potential endogeneity among pregnancy and educational attainment but did exclude many teenage mothers who had already dropped out of or completed high school by the time they became pregnant. Thus, our findings apply only to teenage mothers who become pregnant while enrolled in school. Finally, our measure of household composition was chosen to capture the teen mother’s most proximal social environment, given past research showing that living with one’s parents versus a romantic partner is associated with greater likelihood of high school graduation (Mollborn, 2010). However, this measure does not capture additional supportive relationships she may have with people outside her home. In future research, it would be beneficial to examine the role of specific supportive relationships (e.g., with mentors, peers, teachers, extended family and other supportive persons) in teenage mothers’ long-term educational achievement.

Conclusion

While our findings are noteworthy, they are not surprising. There are many existing policies and programs that support teen mothers with access to child care and resources to prevent rapid repeat pregnancy (for a review, see SmithBattle et al., 2017). However, these policies and programs serve very few of the mothers of the over 180,000 children born to women ages 15 to 19 in the United States annually. As maternal education is strongly linked to socioeconomic well being for both women and their children, focusing resources towards effectively supporting teen mothers with the post-pregnancy factors demonstrated to support high school diploma attainment may be a key step in helping to break the cycle of social disadvantage associated with teenage motherhood in the United States.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by grants from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (K01HD091416 and P2CHD042849), the National Science Foundation (1519686), and the William T. Grant Foundation Scholars Program. The funders had no role in the study design, collection, data analysis, interpretation of the data, writing the manuscript, or the decision to submit the paper for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders.

Biographies

Julie Maslowsky is an associate professor of Community Health Sciences at the University of Illinois at Chicago.

Haley Striztel is a postdoctoral fellow at the Carolina Population Center, University of North Carolina at Chapel Hill.

Elizabeth T. Gershoff is a professor of Human Development and Family Sciences at the University of Texas at Austin.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

  1. Acs G, & Koball H (2003). TANF and the status of teen mothers under age 18 (No. A-62). The Urban Institute. [Google Scholar]
  2. Assini-Meytin LC, & Green KM (2015). Long-term consequences of adolescent parenthood among African–American urban youth: A propensity score matching approach. Journal of Adolescent Health, 56(5), 529–535. 10.1016/j.jadohealth.2015.01.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Child Trends. (2019). Teen births. https://www.childtrends.org/indicators/teen-births
  4. Coleman M, & DeLeire T (2003). An economic model of locus of control and the human capital investment decision. Journal of Human Resources, 38(3), 701–721. 10.3368/jhr.XXXVIII.3.701 [DOI] [Google Scholar]
  5. Coyne CA, & D’Onofrio BM (2012). Some (but not much) progress toward understanding teenage childbearing: A review of research from the past decade. Advances in Child Development and Behavior, 42, 113–152. 10.1016/B978-0-12-394388-0.00004-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Duffy J, & Levin-Epstein J (2002). Add it up. Teen parents and welfare…undercounted, oversanctioned, underserved. Center for Law and Social Policy. https://www.clasp.org/sites/default/files/public/resources-and-publications/publication-1/0090.pdf [Google Scholar]
  7. Enders CK (2013). Dealing with missing data in developmental research. Child Development Perspectives, 7(1), 27–31. 10.1111/cdep.12008Geronimus [DOI] [Google Scholar]
  8. Geronimus AT, Korenman S, & Hillemeier MM (1994). Does young maternal age adversely affect child development? Evidence from cousin comparisons in the United States. Population and Development Review, 20, 585–609. 10.2307/2137602 [DOI] [Google Scholar]
  9. Gordon RA, Lindsay Chase-Lansdale P, & Brooks-Gunn J (2004). Extended households and the life course of young mothers: Understanding the associations using a sample of mothers with premature, low birth weight babies. Child Development, 75(4), 1013–1038. [DOI] [PubMed] [Google Scholar]
  10. Hendrick CE, & Maslowsky J (2019). Teen mothers’ educational attainment and their children’s risk for teenage childbearing. Developmental Psychology, 55(6), 1259–1273. 10.1037/dev0000705 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Jaffee S, Caspi A, Moffitt TE, Belsky J, & Silva P (2001). Why are children born to teen mothers at risk for adverse outcomes in young adulthood? Results from a 20-year longitudinal study. Development and Psychopathology, 13(2), 377–397. 10.1017/s0954579401002103 [DOI] [PubMed] [Google Scholar]
  12. Kane JB, Philip Morgan S, Harris KM, & Guilkey DK (2013). The educational consequences of teen childbearing. Demography, 50(6), 2129–2150. 10.1007/s13524-013-0238-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kearney MS, & Levine PB (2015). Investigating recent trends in the US teen birth rate. Journal of Health Economics, 41, 15–29. 10.1016/j.jhealeco.2015.01.003 [DOI] [PubMed] [Google Scholar]
  14. Lee D (2010). The early socioeconomic effects of teenage childbearing: A propensity score matching approach. Demographic Research, 23, 697–736. 10.4054/demres.2010.23.25 [DOI] [Google Scholar]
  15. Liu SY, Chavan NR, & Glymour MM (2013). Type of high-school credentials and older age ADL and IADL limitations: Is the GED credential equivalent to a diploma? The Gerontologist, 53(2), 326–333. 10.1093/geront/gns077 [DOI] [PubMed] [Google Scholar]
  16. Lou C, & Thomas A (2015). The relationship between academic achievement and nonmarital teenage childbearing: Evidence from the Panel Study of Income Dynamics. Perspectives on Sexual and Reproductive Health, 47(2), 91–98. 10.1363/47e2115 [DOI] [PubMed] [Google Scholar]
  17. Mahler K (1999). Dropping out of school increases chance of birth for Whites and Hispanics. Family Planning Perspectives, 31(3), 153. 10.2307/2991703 [DOI] [Google Scholar]
  18. Manlove J (1998). The influence of high school dropout and school disengagement on the risk of school-age pregnancy. Journal of Research on Adolescence, 8(2), 187–220. 10.1207/s15327795jra0802_2 [DOI] [PubMed] [Google Scholar]
  19. Manlove J, & Lantos H (2018). Data point: Half of 20-to-29-year-old women who gave birth in their teens have a high school diploma. Child Trends. [Google Scholar]
  20. Manlove J, Mariner C, & Papillo AR (2000). Subsequent fertility among teen mothers: Longitudinal analyses of recent national data. Journal of Marriage and Family, 62(2), 430–448. 10.1111/j.1741-3737.2000.00430.x [DOI] [Google Scholar]
  21. Marcotte DE (2013). High school dropout and teen childbearing. Economics of Education Review, 34, 258–268. 10.1016/j.econedurev.2013.01.002 [DOI] [Google Scholar]
  22. Maslowsky J, Powers D, Hendrick CE, & Al-Hamoodah L (2019). County-level clustering and characteristics of repeat versus first teen births in the United States, 2015–2017. Journal of Adolescent Health, 65(5), 674–680. 10.1016/j.jadohealth.2019.05.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mollborn S (2007). Making the best of a bad situation: Material resources and teenage parenthood. Journal of Marriage and Family, 69(1), 92–104. 10.1111/j.1741-3737.2006.00347.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mollborn S (2010). Exploring variation in teenage mothers’ and fathers’ educational attainment. Perspectives on Sexual and Reproductive Health, 42(3), 152–159. 10.1363/4215210 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Mollborn S, & Blalock C (2012). Consequences of teen parents’ childcare arrangements for mothers and children. Journal of Marriage and the Family, 74(4), 846–865. 10.1111/j.1741-3737.2012.00988.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Ou S-R, & Reynolds AJ (2008). Predictors of educational attainment in the Chicago Longitudinal Study. School Psychology Quarterly, 23(2), 199–229. 10.1037/1045-3830.23.2.199 [DOI] [Google Scholar]
  27. Perper K, Peterson K, & Manlove J (2010). Diploma attainment among teen mothers. Child Trends. https://www.childtrends.org/wp-content/uploads/2010/01/child_trends-2010_01_22_FS_diplomaattainment.pdf [Google Scholar]
  28. Ramey CT, Campbell FA, Burchinal M, Skinner ML, Gardner DM, & Ramey SL (2000). Persistent effects of early childhood education on high-risk children and their mothers. Applied Developmental Science, 4(1), 2–14. [Google Scholar]
  29. Royston P (2004). Multiple imputation of missing values. The Stata Journal, 4(3), 227–241. 10.1177/1536867X0400400301 [DOI] [Google Scholar]
  30. SmithBattle L (2007). Legacies of advantage and disadvantage: The case of teen mothers. Public Health Nursing, 24(5), 409–420. 10.1111/j.1525-1446.2007.00651.x [DOI] [PubMed] [Google Scholar]
  31. SmithBattle L, Loman DG, Schneider JK, & Chantamit-O-Pas C (2017). An umbrella review of meta-analyses of interventions to improve maternal outcomes for teen mothers. Journal of Adolescence, 59, 97–111. 10.1016/j.adolescence.2017.05.022 [DOI] [PubMed] [Google Scholar]
  32. StataCorp. (2015). Stata statistical software: Release 14. StataCorp LP. [Google Scholar]
  33. Upchurch DM (1993). Early schooling and childbearing experiences: Implications for postsecondary school attendance. Journal of Research on Adolescence, 3(4), 423–443. 10.1207/s15327795jra0304_6 [DOI] [Google Scholar]
  34. US Bureau of Labor Statistics. (n.d.-b). The NLSY79. https://www.bls.gov/nls/nlsy79.htm
  35. US Bureau of Labor Statistics. (n.d.-c). National longitudinal survey of youth 1979 sample weights & clustering adjustments: Practical usage of weights. https://www.nlsinfo.org/content/cohorts/nlsy79/using-and-understanding-the-data/sample-weights-clustering-adjustments
  36. US Bureau of Labor Statistics. (n.d.-d). National longitudinal survey of youth 1979: Labor force status. https://www.nlsinfo.org/content/cohorts/nlsy79/topical-guide/employment/labor-force-status
  37. Ventura SJ, Hamilton BE, & Mathews TJ (2014). National and state patterns of teen births in the United States, 1940–2013. National Vital Statistics Reports, 63(4), 1–34. [PubMed] [Google Scholar]
  38. Zajacova A (2012). Health in working-aged Americans: Adults with high school equivalency diploma are similar to dropouts, not high school graduates. American Journal of Public Health, 102(S2), S284–S290. 10.2105/ajph.2011.300524 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zajacova A, & Everett BG (2014). The nonequivalent health of high school equivalents. Social Science Quarterly, 95(1), 221–238. 10.1111/ssqu.12039 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES