COLLEGE ASPIRATIONS AND EXPECTATIONS AMONG NEW AFRICAN AMERICAN MOTHERS IN LATE ADOLESCENCE

Abstract

It is a generally accepted finding in the sociological literature as well as in public discourse that adolescent mothers are less likely than their non-parenting counterparts to graduate high school and to attend college. For several decades, however, researchers have pointed out that the implied causal process from teen motherhood to academic failure has been largely unsupported by empirical research. In fact, scholars have recently argued that motherhood may actually serve as a positive turning point in the lives of young women. Using a sample of young African American women, the present study assesses the degree to which teen motherhood not only affects college aspirations but also expectations. Further, it tests the ability of these effects to explain the well-known educational attainment gap between teen mothers and their non-childbearing peers. Results indicate that, in general, young mothers’ college aspirations are similar to those of non-mothers, but that their generally high aspirations for academic success appear to be effectively countered by their decreased educational expectations.

Introduction

It is a generally accepted finding in the sociological literature as well as in public discourse that adolescent mothers are less likely than their non-parenting counterparts to graduate high school and to attend college (Anderson 1993; Escamilla and Santhiveeran 2005; 2001). Further, in the United States, the pregnancy rate among African American teens is nearly twice that of Caucasians, and African American teenage pregnancies are less likely than those of their white counterparts to end in abortion (AGI 2006; South and Baumer 2001). Hence, just as the phenomenon of teenage childbearing is racialized, so too are the gross disparities in income and general family well-being that result, at least in part, from the limited education typically associated with such childbearing.

Recent research, however, has documented that there is little evidence for the causal model implied above, a model which marks adolescent pregnancy as a trigger for negative life outcomes. In fact, as noted by Furstenberg (2003; 2007), SmithBattle (2007b), and Mollborn and Morningstar (2009), many of the correlated “outcomes” of teenage childbearing, including psychological distress, limited educational achievement, low socio-economic status, and poor child well-being, can be attributed to the relatively high level of disadvantage experienced by teen mothers before they give birth. Prior to becoming pregnant, teen mothers are more likely to come from underprivileged households and neighborhoods and thereby are more likely to have little attachment to school and to have few recognizable opportunities for educational and economic mobility (Furstenberg, Brooks-Gunn, and Chase-Lansdale 1989; Hoffman, Foster, and Furstenberg 1993; SmithBattle 1995; 2007b). In other words, the popular image of teen motherhood as the “beginning of a downward spiral that includes dropping out of school, depending on welfare or minimum wage jobs, and remaining single, poor mothers” is largely inaccurate (SmithBattle 2007b, 409), as the conditions often thought to result from teen pregnancy are also likely to be the conditions under which a teen pregnancy takes place.

In a recent qualitative study of teen mothers’ and their experiences as students, SmithBattle (2007a) reiterates this selection effect but goes one step further than her colleagues by arguing that, contrary to popular opinion, teen childbearing may actually cause young women to reevaluate their goals and priorities, resulting in a newfound desire to improve their life chances. More specifically, SmithBattle (2007a) shows that, following childbirth, young women often experience an increase in educational aspirations but that their competing demands and responsibilities and limited support often prevent such aspirations from translating into educational achievement. In other words, SmithBattle (2007a) argues that teen motherhood has the potential to serve as a positive turning point in the life trajectories of disadvantaged youth if these youth are provided with the necessary support. Given that education remains the single most important pathway to breaking the cycle of poverty and to enjoying upward mobility (Hout and DiPrete 2006), claims such as SmithBattle’s (2007a) cannot and should not be taken lightly. Using data from a longitudinal study of African American youth, the current study expands upon SmithBattle’s findings by quantitatively exploring changes in both the educational aspirations and expectations, particularly as they relate to college graduation, of new mothers in late adolescence.¹

Although SmithBattle (2007a) does not frame her argument in terms of aspirations versus expectations, there is reason to believe that the suspected increase in educational aspirations following a teen birth might not be complemented by a comparable increase in educational expectations. Instead, the added responsibility, limited support, and social stigma perceived by teen mothers may actually generate a decrease in educational expectations, resulting in a sizeable gap between aspirations and expectations. Said differently, young mothers may enhance their educational goals but may feel that such goals are out of their reach, presumably due to both their relatively higher level of disadvantage pre-pregnancy and the added responsibilities associated with motherhood. Past research has revealed that the educational aspirations and expectations of disadvantaged youth in the U.S. are often mismatched in this way (Hanson 1994). That is, expectations often fall short of aspirations. Hanson (1994) refers to these mismatched youth as one form of “lost talent,” seemingly because such a discrepancy between aspirations and expectations is projected to be important for predicting life outcomes, particularly educational attainment/achievement. Hence, despite the potential significance of simply supporting SmithBattle’s (2007a) claims in a different context, , this study also seeks to understand if the suspected increase in teen mothers’ educational aspirations following the birth of a child is countered by a decrease in educational expectations and what this potential mismatch might mean for later educational achievement.

In sum, this study addresses the following research questions: First, how does the birth of a child predict changes in young women’s educational aspirations and expectations in late adolescence? Second, how does the discrepancy between aspirations and expectations differ between young mothers and non-mothers? Third, controlling for aspirations, do educational expectations explain the education gap between teen mothers and their non-parenting counterparts? Answers to such questions will expand upon SmithBattle’s (2007a) claims that the birth of a child can serve as a positive turning point in the lives of disadvantaged youth and may provide insight into the possibility that young mothers’ newfound educational aspirations may be effectively countered by simultaneously decreased expectations.

Academic and Popular Conceptions of Teen Childbearing

As Furstenberg (2007) notes, teen childbearing has never been uncommon in the United States. Due to several demographic and social changes in the mid-to-late twentieth century, changes which include relaxed standards with regard to premarital sex, the rising age of marriage, and the swelling number of teenagers in the population (a result of the baby boom of the 1950s), adolescent childbearing caught the public’s attention and was suddenly identified as a major social problem. Moreover, given that such demographic and social changes were occurring more rapidly among African Americans and, perhaps, that the Civil Rights Movement was threatening white supremacy, Black teens were singled out as the source of the problem (Furstenberg 2007; 2009; Luker 1996; Nathanson 1991). This focus on African Americans was exemplified in Daniel Patrick Moynihan’s 1965 report on the state of the Black family, in which he wrote of Blacks, “A cycle is at work; too many children too early make it most difficult for parents to finish school…Low education levels in turn produce low income levels, which deprive children of many opportunities, and so the cycle repeats itself” (Moynihan 1965, 27).

Although Moynihan’s report was condemned by many for its specific focus on the Black family, even his immediate critics (e.g. Rainwater and Yancey 1967) accepted the premise that teen childbearing “came at a great cost to adolescent parents, their children, and society at large” (Furstenberg 2007:15). As one prominent demographer put it, “The girl who has an illegitimate child at the age of 16 suddenly has 90 percent of her life’s script written for her…Her life choices are few and most of them are bad” (Campbell 1968, 238).

Given this assumption about the devastating effects of teen parenthood, an assumption that was common not only among the general public but also among academics, social science researchers were mobilized to study and quantify the causes and consequences of adolescent childbearing. Underlying this research was the belief, both by funding agencies and by many researchers themselves, that the prevention of teen childbearing would bring about a drastic reduction in poverty (Furstenberg 2007). Not surprisingly, given such assumptions, evidence from early social science research seemed to support the notion that teen childbearing was destructive for young mothers and their families, as these mothers were consistently shown to have lower education levels, lower incomes, and more disadvantaged offspring than their counterparts who postponed childbearing into adulthood (e.g. Campbell 1968; Lim 1977).

Since the early 1990s, however, researchers, many of them inspired by their own research subjects, have shown earlier studies on the consequences of teen childbearing to be severely flawed. In particular, these later researchers (e.g. Cherlin 2001; Geronimus and Korenman 1992) pointed to the use of short-term studies to draw conclusions about long-term life chances and the failure, and sometimes inability, to take into account the fact that “teenage pregnancy does not occur randomly,” or that “early parenthood happens disproportionately to adolescents (and their partners) who are different in a variety of ways from the teens who typically delay parenthood” (Furstenberg 2007:47). That is not to say that poverty causes teen childbearing but simply that “poverty is thus the context within which the overwhelming majority of instances of motherhood under 20 occur” (Pheonix 1991, 90).

Those who were able, through advanced statistical tests and clever study designs, to account for such issues as best as possible found that the effect of teen childbearing on young women’s lives has been grossly overestimated (Geronimus and Korenman 1992; Hoffman, Foster, and Furstenberg 1993; Hotz, McElroy, and Sanders 1996; Levine and Painter 2003). What these more recent studies tend to show is that “teenage childbearing is more a consequence than a cause of economic and social disadvantage” and that for many young mothers, waiting to bear children until adulthood would not greatly improve their life chances unless such a change was accompanied by a dramatic transformation of life circumstances (Furstenberg 2007, 51–52). Unfortunately, notions of teen motherhood in the popular press have failed to incorporate such findings.

Since the identification of adolescent childbearing as a significant social problem in the early-to-mid 1960s, teen mothers have been represented in the news media through multiple, sometimes competing, discourses. As Kelly (1996) notes, however, “the discourses do not compete as equals; some carry little weight and are marginalized, whereas others are considered authoritative and dominant” (423). While academics were once considered the authoritative experts on the topic of teenage childbearing, they are increasingly no longer so, as their research, if presented at all, is often juxtaposed by conservative pundits or researchers who continue to claim the devastating effects of teen childbearing on young women, their families, and society at large (Furstenberg 2003; 2007; Kelly 1996). In fact, as Furstenberg (2007) documents, the press often actively ignore the most significant conclusions of academic reports if those conclusions present a different narrative than the traditional one which insists that teen childbearing ruins lives and communities. For instance, one of the first reports to summarize the growing amount of research on the not-so-devastating effects of adolescent parenthood concluded that “the popular image of severe and life-long social and economic disadvantage” that is thought to result from teen childbearing has been “exaggerated” (Hayes 1987, 132). This conclusion, however, was “largely ignored in the press coverage of the National Academy Report” (Furstenberg 2007, 47), as such coverage focused instead on the report’s finding that teen mothers, in general, tend to be at a greater disadvantage than their non-parenting counterparts.

Dominant discourses of teen childbearing as personally and socially destructive rely upon several interrelated stereotypes of young mothers. Because the recognition of teen childbearing as a significant social problem was heavily influenced by the notion that African American women were the ones choosing to have children at young ages, such stereotypes are heavily intertwined with the common stereotypes of African American women articulated by Patricia Hill Collins (1990) nearly two decades ago. This is most clearly portrayed in Furstenberg’s book (2007), Destinies of the Disadvantaged, when he entitles one of his chapters “From Teenage Mother to Midlife Matriarch” in order to acknowledge that public outcry over teen motherhood is due, in part, to the stereotype that such mothers will grow into matriarchs, albeit welfare-dependent, who purportedly eschew the “family values” of White, middle-class America.

In an analysis of over 700 news articles, Kelly (1996) points out that these stereotypes of young mothers are woven into two stigmatizing discourses in the mass news media. The first of these discourses is framed in terms of “the wrong family” and is usually propagated by economic, political, and social conservatives (Kelly 1996, 430). What defines the “wrong family” discourse is its “concern over the fate of the traditional family…as well as the changing relationship between families and the state” (Kelly 1996, :431). This discourse stigmatizes teen mothers for choosing the wrong (read: not two-heterosexual-parent) family model. The second most common discourse identified by Kelly (1996) in the mass news media is framed not in terms of “the wrong family” but in terms of the “the wrong girl”. Such discourse tends to be a prominent one among liberal activists in the news media, and it is one in which teen childbearing is not seen as a moral failure (as it is in the “wrong family” discourse) but as a product of dire circumstances, typically poverty, abuse, or both. Hence, this discourse amounts to “the analysis that the wrong girls are having and keeping babies” (Kelly 1996, :429). Although this discourse appears to acknowledge the shared conditions of young mothers prior to becoming pregnant, it fails to identify these conditions as structural and institutionalized and, instead, identifies them as personal tragedies. As Kelly (1996) explains,

“Rather than attend to the conditions of poverty and frame the social psychology as rooted in the problems of the powerless, however, experts have tended to focus attention on the psychological motivations of girls who become pregnant and decide to keep their babies” (428).

In this discourse, the reason that teen childbearing is viewed as tragic is because their early entrance into parenthood continues to be interpreted as damning them to a life of further despair, as if they would have broken the cycle of poverty if it were not for having a child. Hence, although this latter discourse does not view teen mothers as morally deviant and acknowledges that their childbearing is often a condition of their (individual) circumstances, it clings to the notion of teen motherhood as detrimental to the lives of young women and their children. In both the “wrong girl” and the “wrong family” discourses, teen childbearing, especially among Black women, becomes that which is stigmatized and that which is to be prevented.

Motherhood as a Turning Point

Contrary to popular and early academic conceptions of teen childbearing as destructive, the notion that teen childbearing may serve to redirect disadvantaged and/or deviant youth toward more conventional values and activities has been noted and supported in in a wide range of literatures. The criminological literature, for instance, it has been shown that “motherhood exerts a trajectory-modifying effect on problem behavior— that it serves as a form of social control” (Hope, Wilder, and Watt 2003, 568; see also Hunt, Joe-Laidler, and MacKenzie 2005). That is, although adolescent girls who become pregnant generally exhibit higher levels of delinquency and/or antisocial behavior than their never-pregnant peers, those who transition into motherhood tend to reduce their involvement in such behavior. More recent research has found similar effects for young men who transition to fatherhood (Moloney, MacKenzie, and Hunt 2009).

Instead of focusing on motherhood as simply a deterrent for engaging in deviant behavior, SmithBattle (2007a) argues that motherhood often serves to motivate young women academically. In doing so, she contests the popular image of young mothers as unmotivated drop outs content to living lives of poverty and hardship. Although some of the young women in SmithBattle’s (2007a) study had indeed dropped out of school, SmithBattle notes that almost all of her respondents, regardless of school status, “described the emergence of new priorities and concerns for their future” including a “renewed commitment to schooling” (355). This renewed commitment to schooling often entailed a desire to improve grades and to consider, often for the first time, attending college. Importantly, SmithBattle’s (2007a) respondents attributed their newfound motivation for academic success to their new role as mothers and the desire to provide “a good life” for their children.

Other qualitative studies, both in the United States and elsewhere, corroborate such findings regarding the motivating effects of teen motherhood. For instance, Rolfe (2008) and Kirkman, Harrison, Hillier, & Pyett (2001) reveal that young mothers in England and Australia, respectively, often discuss their role as mothers in terms of “growing up” and becoming more responsible. Likewise, in a review of qualitative studies on adolescent motherhood in the United States, Clemmens (2003) notes that a common theme running through past research, yet not always explicitly addressed, has been the perception of motherhood as a positively transforming event in the lives of young women. Nevertheless, as observed by Furstenberg (2007), SmithBattle (2007a; 2007b; 2009), Burdell (1995), and Wilson and Huntington (2006), the conception of teen motherhood as a potential turning point in the lives of young women has been trivialized, if not ignored, in both popular and academic discourse.

Constraints and Lowered Expectations

Perhaps the tendency to trivialize the potentially beneficial role of teen motherhood can be at least partially attributed to the high likelihood that young mothers will face certain constraints not faced (or faced to a much lesser degree) by their non-parenting counterparts. In accordance with this possibility, SmithBattle (2007a) contends that such constraints, specifically work demands, family and child care responsibilities, and lack of school support, may help to explain why teen mothers’ increased educational aspirations may not translate into increased educational attainment and achievement. For example, although most of the young mothers in SmithBattle’s (2007a) study reported a newfound desire to go to college, the author notes that “only one teen received college advisement from school personnel” (356) . Furthermore, the teen mothers also reported that work or child care responsibilities often interfered or competed with their ability to attend and succeed in school.

SmithBattle (2007a) frames these constraints as structural factors that thwart her respondents’ chances of realizing their educational aspirations. Research in the status attainment literature, however, provides a slightly different yet still consistent way of framing SmithBattle’s (2007a) findings in terms of young mothers’ mismatched aspirations and expectations. In this literature, scholars have established both the theoretical and empirical need to differentiate between aspirations and expectations, or similarly between abstract and concrete attitudes toward education (Kerckhoff 1976; Mickelson 1990). Although abstract and concrete attitudes toward education are not synonymous with aspirations and expectations, respectively, the rationale for differentiating between both abstract/concrete attitudes and aspirations/expectations lies in the realization that people understand the limitations that their circumstances place upon their ambitions and goals. Hence, as Kerckhoff (1976) explains, distinguishing between aspirations and expectations acknowledges the “the distinction between ‘wanting’ something and ‘expecting’ something” (371), the latter of which is anticipated to exert a greater influence on actual achievement.

It has been argued, for instance, that young people’s aspirations tend “to transcend their awareness of limitations” (Han 1969, 679) and, hence, to represent broader cultural values irrespective of social class. Educational expectations, on the other hand, are thought to be highly contingent upon social circumstances, as “adolescents become sufficiently knowledgeable to be able to estimate the probabilities of various outcomes” (Kerckhoff 1976, 371). Perceived structural constraints, then, are thought to influence educational expectations much more so than aspirations, as “everyone wants the same outcomes, but people at different levels of the stratification system quite understandably expect different outcomes” (Kerckhoff 1976, 371). Further, as mentioned above, because expectations are often grounded in beliefs about what is possible given one’s circumstances, expectations are thought to be more strongly related to actual attainment than aspirations. Hence, to the degree that competing demands often experienced by young mothers and persistent cultural stereotypes are successful in reducing their perceived chances of academic success, teen mothers may experience a decrease in educational expectations despite their hypothesized increase in aspirations.

Given the simultaneous increase in aspirations and decrease in expectations that is expected to occur for young mothers, it is anticipated that, relative to their non-parenting peers, teen mothers will have educational expectations that fall short of their aspirations. Hanson (1994) categorizes such youth as “lost talent” due to the likelihood that their aspirations will be unrealized. This resulting gap between aspirations and expectations may help to account for the disparity in educational attainment between teen mothers and non-mothers. The pathway proposed here, then, differs only slightly from that proposed by SmithBattle (2007a). Whereas SmithBattle (2007a) argues that young mothers’ rise in aspirations is offset by the structural constraints of teen motherhood, the present study proposes that such constraints are internalized by young mothers and expressed in the form of decreased expectations for success, expectations with help to explain the relationship between teen motherhood and educational outcomes.

Hypotheses

Based on the above literature review, we expect the following relationships to emerge among teen motherhood and educational aspirations, expectations, and attainment. In line with SmithBattle (2007a), we expect that young mothers will experience a rise in educational aspirations following the birth of their child (hypothesis 1). Given the many structural and cultural constraints associated with teen parenthood, however, we also anticipate that young mothers will experience a simultaneous reduction in educational expectations (hypothesis 2). Hence, we predict that young mothers will be more likely than non-mothers to have aspirations that exceed expectations, or in line with Hanson, “lost talent” will be much more likely among the former than among the latter group of respondents (hypothesis 3). Lastly, we anticipate that, independent of aspirations, expectations will partially explain the well-known educational disparity between teen mothers and their non-parenting counterparts (hypothesis 4).

Data and Methods

Sample

The current study is part of a broader, ongoing longitudinal research project, the Family and Community Health Study (FACHS), which examines the social, psychological, and contextual risk and protective factors associated with African American families’ health and wellbeing (see Gibbons, Gerrard, Cleveland, Wills, and Brody 2004; Simons, Lin, Gordon, Brody, Murry, and Conger 2002 for a detailed description of sampling procedures). In brief, a total of 867 African American families from rural and suburban areas in Iowa and Georgia participated in the first wave of data collection in 1997 (when target youth were 10 to 11 years old). Data from subsequent waves were collected every 2 to 3 years thereafter. Because the current study focuses on childbearing, it utilizes target data from the most recent three waves (waves 3 through 5) of FACHS, when the target youth were in late-adolescence and early adulthood. At all waves of data collection utilized here, data were collected via a computerized survey that was implemented in a single visit to the respondents’ homes.

Of the 867 target youth interviewed at wave 1, 689 remained in the study at wave 5 (79.47%). Of those remaining in the study at wave 5, 397 were young women, who serve as the focus for the current study. In an effort to limit the variability in the timing of respondents’ childbirths and to ensure the proper time-ordering of events, those respondents who reported having had multiple children and those whose childbirth occurred prior to wave 3 were dropped from the sample (N=24). Other respondents were excluded if they failed to provide complete information on all variables of interest (N=20). The final sample, then, consists of 307 young women who, if they have a child, had that child between waves 3 and 4 of the FACHS study (i.e. during late adolescence). Although this reduction in sample size may introduce some degree of selection bias and necessarily restricts the number of mothers in the sample, it is a necessary sacrifice in order to ensure, as best as possible, the accuracy and comparability of respondents’ experiences of becoming mothers.

Variables

Below, we describe the operational definitions of all variables used in subsequent analyses, beginning with the dependent variables: college aspirations, expectations, and attendance. Although the above literature review refers to educational attainment in its most general form, the dependent variables in the present study are all concerned with aspirations, expectations, and attainment at the college level. Although such measures allow for a lesser degree of nuance than some readers may want, this focus on college is both theoretically important and empirically necessary. Recent research, for instance, has indicated that the effects of teen childbearing on high school graduation have declined over the past several decades, while the gap between teen mothers and their non-parenting counterparts in postsecondary school attendance has widened (Hofferth, Reid, and Mott 2001). Hence, it is primarily college attendance that differentiates young mothers from their non-childbearing peers. Further, stratification research indicates that college and, more frequently, advanced degrees are more important than high school graduation in predicting life outcomes (Fairweather 2006; Policy 2004). Consistent with this research, very few FACHS respondents (7%) have obtained less than a high school diploma by wave 5 of the study (early adulthood). Hence, the dependent variables pertaining to educational aspirations, expectations, and attainment are as follows:

College Aspirations

Aspirations were assessed at wave 3 and 4 via the following question: “If you could go as far as you wanted to in school, how much education would you like to have?” At wave 3, possible responses included (1) less than high school, (2) graduate from high school, and (3) more than high school. At wave 4, possible responses included (1) less than high school, (2) graduate from high school, (3) some college, and (4) college degree or higher. Due to insufficient cell sizes across response categories, both wave 3 and wave 4 responses were collapsed into a 2-level dummy variable. At wave 3, respondents who reported wanting “more than high school” were coded 1 for having college aspirations. At wave 4, respondents who reported wanting a “college degree or higher” were coded 1 for having college aspirations. Because of the lack of nuance in the wave 3 measure, academic orientation was also assessed at wave 3 and controlled for in all analyses. This variable is further described below.

College Expectations

Expectations were assessed via the following question: “How far do you think you will actually go in school?” This question was asked immediately after the question assessing aspirations. Possible response categories were the same as those noted for aspirations, and a similar coding scheme was followed. At wave 3, respondents who reported expecting an educational level of “more than high school” were coded 1 for having college expectations. At wave 4, respondents who reported expecting a “college degree or higher” were coded 1 for having college expectations.

Aspirations > Expectations

A dummy variable was created to differentiate those respondents whose college aspirations exceeded their expectations from those whose aspirations and expectations were better matched. Those respondents who, at wave 4, indicated a desire to go to college but who expected to achieve less than a college degree were coded 1 for having aspirations that surpassed expectations. All others were coded 0. Given the lack of nuance in the educational aspirations and expectations questions in the FACHS data, a continuous discrepancy measure was impossible to construct. Although previous studies have used a binary measure of mismatched aspirations and expectations similar to the one used here (e.g. Hanson 1994), a continuous discrepancy measure would be useful for future studies so that the magnitude of the discrepancy between aspirations and expectations could be assessed.

College Attendance

At wave 5, respondents reported their level of education to date. Those who reported having had any college education were coded 1, while those who reported a high school diploma or less were coded 0. Although this simplistic measure of college attendance does not capture achievement in terms of academic success or graduation, the relatively young age of respondents in this sample precludes such an analysis. That is, with a mean age of roughly 22 at wave 5, college graduation cannot yet be expected from the majority of respondents. Nevertheless, this measure provides a nice indication of whether or not the respondent followed the traditional path to higher education by early adulthood.

Mother Status

The primary independent variable of interest was the respondents’ parental status. More specifically, of interest was whether or not a respondent transitioned into motherhood during the time period under investigation. The following categorization, then, was based on wave 4 responses after those who reported having had multiple children or a child prior to wave 3 were excluded from the sample. Respondents who indicated that they had been pregnant were asked “Do you currently have any biological children?” Those who indicated that they have a child were coded 1 for transitioning into motherhood, while those who indicated that they had never been pregnant or who had been pregnant but did not currently have a child were coded 0 for not transitioning into motherhood.

In addition to the variables described thus far, we control for several factors that may be selective of teen motherhood status. Respondents’ general academic orientation was assessed via 6 survey questions at wave 3. Respondents were asked how much they agreed or disagreed with statements like: “You try hard at school,” “Grades are very important to you,” and “You do not feel like you really belong at school” (reverse coded). Responses were averaged across items to form the index. To assess respondents’ general commitment to conventionality (conventional values) at wave 3, they were asked 7 questions about the importance of such things as having a secure income and stable job, to save money for the future, to be successful in work, and to have a good marriage. Responses ranged from (1) not at all important to (5) extremely important and were averaged across items to form the index. Community disorder was assessed via 8 survey items at wave 3. These questions asked respondents to report on the frequency of certain neighborhood events in the past 6 months, including a fight in which a weapon like a gun or knife was used, a violent argument between neighbors, a gang fight, a robbery or mugging, and people selling or using drugs. After reverse coding, responses ranged from (1) never to (3) often and were averaged across items to form the index. Friends’ reinforcement of deviance was assessed via 14 survey items at wave 3. Such items asked respondents to report on what their friends would do if the respondent were to engage in certain deviant behaviors, like lying to parents or teachers, drinking a lot of alcohol, beating up on someone, and skipping school without an excuse. Respondents indicated whether their close friends would (1) tell them to stop, (2) do nothing, or (3) encourage them to do it again. These responses were averaged across items to form the index. Family income, as reported by respondents’ primary caregivers at wave 3, and target’s age were also controlled in all analyses. Income was reported in whole dollars but was coded in tens of thousands of dollars.

Analytic Notes

Given the dichotomous nature of the three dependent variables of interest, we use logistic regression with robust standard errors in each of the following regression analyses.² In addition to logistic regression, we use a simple Chi-square test of independence to assess whether “lost talent” was more common among young mothers than among non-mothers. Prior to discussing the results of these analyses, however, we briefly present some descriptive statistics for the current sample.

Results

Descriptive Results

Table 1 displays the means, standard deviations, and intercorrelation coefficients for all variables of interest. On average, sample respondents were 16, 19, and 22 years of age across the respective three waves of the study utilized here. Roughly 18% of these respondents reported having had their first child between waves 3 and 4, typically during late adolescence (average age was 17). Respondents’ family incomes ranged from $900 to $175,000 and averaged about $33,000 at wave 3. By wave 4 of the study, roughly 87% of respondents aspired toward a college degree, while only 76% of the sample expected to earn one. Further, 13% of respondents held expectations that fell short of their aspirations. By wave 5 of the study, 52% of respondents had attended some form of postsecondary education. These statistics, which are slightly higher than other estimates for African American youth (Hanson 1994), are indicative of the fact that the current sample probably does not include the most disadvantaged youth. For instance, the current study excludes youth who experienced multiple pregnancies or a pregnancy prior to wave 3. Consistent with this logic is the preliminary finding that, prior to their pregnancies, young mothers did not significantly differ from their non-parenting peers in their college aspirations and expectations.

Table 1.

Intercorrelations and descriptive statistics for study variables

	1	2	3	4	5	6	7	8	9	10	11	12	13
1College Aspirations (w4)	––
2College Aspirations (w3)	0.28	––
3College Expectations (w4)	0.56	0.14	––
4College Expectations (w3)	0.35	0.47	0.23	––
5Academic Orientation (w3)	0.23	0.11	0.25	0.23	––
6College Attendance (w5)	0.28	0.16	0.38	0.20	0.08	––
7Aspirations > Expectations (w4)	0.15	0.01	−0.69	0.00	−0.14	−0.25	––
8Family Income (10K)	0.13	0.12	0.15	0.11	−0.08	0.27	−0.05	––
9Age (w4)	0.08	0.07	0.04	0.05	−0.06	0.00	0.05	0.09	––
10Community Disorder	−0.09	−0.11	−0.14	−0.08	−0.10	−0.08	0.10	−0.11	0.05	––
11Conventional Values	0.18	0.15	0.17	0.23	0.32	0.18	−0.06	0.02	−0.06	−0.04	––
12Friends' Deviance	−0.09	0.02	−0.13	−0.03	−0.36	0.06	0.07	0.09	0.16	0.13	−0.15	––
13Mother Status	0.01	0.10	−0.17	0.03	0.01	−0.17	0.19	−0.02	0.22	−0.09	−0.04	0.04	––
Mean	0.87	0.92	0.76	0.90	3.12	0.52	0.13	3.26	18.76	0.43	4.58	1.40	0.18
Std. Deviation	0.34	0.28	0.43	0.30	0.48	0.50	0.34	2.70	0.91	0.39	0.45	0.34	0.39
Range	0–1	0–1	0–1	0–1	1.5–4	0–1	0–1	0–17.5	17–21	0–1.75	2.43−5	1–2.5	0–1
Cronbach's Alpha	––	––	––	––	0.76	––	––	––	––	0.80	0.82	0.90	––

Note: Highlighted cells are significant at p < .05; N = 307

Analytic Results

Table 2 presents the logit results for the models predicting college aspirations and expectations at wave 4 while controlling for wave 3 responses. In predicting aspirations, only three variables reached statistical significance. As expected, college aspirations and academic orientation at wave 3 significantly predicted aspirations at wave 4. More specifically, those respondents who previously aspired toward a college degree were nearly 5 times more likely than those who did not to aspire toward college in late adolescence/early adulthood. Likewise, a 1-unit increase in respondents’ academic orientation (at wave 3) increased the odds of aspiring toward college at wave 4 by roughly 230%, controlling for all other variables in the model. Also consistent with past research is the finding that family income had a significant and positive association with college aspirations, as each $10,000 increase in family income was associated with a 20% increase in the likelihood of aspiring toward a college degree. These same patterns were present in the model predicting college expectations. That is, those who expected college, who were more academically oriented at wave 3, and who came from more affluent families had a higher predicted probability of expecting a college degree at wave 4. The primary difference between the two models, then, lies in the effect of teen childbearing.

Table 2.

Exponentiated Coefficients (Odds Ratios) for Logit Models Predicting Educational Aspirations and Expectations

Predictors	Aspire College		Expect College
Aspirations (w3)	4.748	**
Expectations (w3)			2.514	*
Academic Orientation (w3)	3.312	*	2.934	**
Age	1.474		1.362
Family Income	1.197	*	1.178	**
Community Disorder	0.769		0.494
Conventional Values	1.372		1.266
Friends' Reinforcement of Deviance	0.581		0.576
Mother Status	0.681		0.264	***
N	307
ll	−99.021		−143.204
Chi-square	32.376		37.335
Pseudo-R2	0.166		0.155

Note: Robust standard errors used.

^*p<.05,

^**p<.01,

^***p<.001 (two-tailed); N = 307

Contrary to hypothesis 1, respondents’ parental status was negatively associated with college aspirations, after controlling for previous aspirations and other relevant selection variables. Notably, however, this association failed to reach statistical significance. Hence, teen childbearing did not significantly predict a change in college aspirations for the young women in the current sample. In contrast, mother status was significantly and negatively associated with college expectations, a finding which supports hypothesis 2. In particular, after controlling for prior expectations and other relevant selection variables, respondents who entered motherhood were roughly 74% less likely than their non-parenting counterparts to expect a college degree. Hence, although young mothers were no less likely than their childless peers to aspire toward college graduation, they were significantly less likely to expect to achieve it.

To test hypothesis 3, a Chi-square test of independence was performed to examine the relation between parental status and mismatched aspirations and expectations (in that aspirations exceeded expectations). Supportive of hypothesis 3, the association between these variables was significant (X² = 11.44, p ≤ .001), indicating that, by wave 4, young mothers were significantly more likely than their non-parenting counterparts to have college expectations that fell short of their aspirations. Such findings were consistent with the aforementioned regression results indicating that young mothers were just as likely as non-mothers to have high aspirations but were much less likely to have high expectations. Further exploration of this relationship revealed that, among those with college aspirations, 69% of teen mothers and 89% of non-mothers expected to earn a college degree. Interestingly, 18% of non-mothers who did not aspire toward college still expected to obtain a college degree. This was not the case for young mothers, as none of those in the sample held expectations that exceeded aspirations.

Table 3 presents the logit regression results predicting any college attendance by wave 5. Ordinal regression (with the following categorizations: less than high school, high school only, more than high school) was also used when predicting educational attainment, but substantive results did not change from those found with the logistic models. Given the small number (7%) of respondents who indicated having less than a high school degree and the theoretical importance of the college/no college distinction, results from the logit model are presented.

Table 3.

Exponentiated Coefficients (Odds Ratios) for Logit Models Predicting College Attendance

	1		2		3		4		5
Aspirations (w3)	1.231		1.390		1.075		1.115		1.391
Expectations (w3)	2.792		2.720		2.204		2.357		3.062
Academic Orientation (w3)	1.448		1.512		1.275		1.064		1.300
Age	0.783		0.829		0.785		0.768		0.825
Family Income	1.233	***	1.237	***	1.219	***	1.205	***	1.235	***
Community Disorder	0.835		0.749		0.776		0.887		0.841
Conventional Values	1.786		1.731		1.562		1.493		1.731
Friends' Deviance	1.775		1.939		2.300	*	2.620	*	1.987
Mother Status			0.361	**	0.351	**	0.436	*	0.431	*
Aspirations (w4)					6.243	***	2.422
Expectations (w4)							5.168	***
Aspirations > Expectations									0.213	**
N	307
ll	−191.540		−186.681		−178.784		−169.801		−179.707
Chi-square	26.159		34.982		44.399		51.976		45.474
Pseudo-R2	0.099		0.122		0.159		0.201		0.154

Note: Robust standard errors used.

^*p<.05,

^**p<.01,

^***p<.001 (two-tailed); N = 307

In the control-only model (model 1), family income was the only significant predictor of respondents’ educational attainment. More specifically, a $10,000 increase in family income was associated with a 23% increase in the likelihood of having any college education by early adulthood. As expected, when parental status was included in the equation (model 2), it too was significant in predicting college attendance, such that young mothers were roughly 64% less likely than their non-parenting peers to have any college education by early adulthood.

Model 3 added in the effect of college aspirations, which was significantly and positively associated with college attendance. Given the aforementioned finding that motherhood did not significantly affect college aspirations, the inclusion of aspirations into the model did not significantly change the relationship between mother status and educational attainment. As shown in model 4, however, respondents’ college expectations were significantly and positively related to college attendance, and such expectations significantly attenuated the effect of teen childbearing and reduced to nonsignificance the effect of aspirations. More specifically, Sobel-Goodman test results indicated that, independent of all control variables, college expectations significantly mediated 28.8% of the relationship between teen childbearing and college attendance (p < .01)³. These findings were highly supportive of hypothesis 4.

Model 5 tested the effect of the aspiration/expectation mismatch on college attendance. In line with the results thus far, model 5 revealed that respondents whose aspirations exceeded their expectations were roughly 79% less likely than those without such a mismatch to have any college-level education by early adulthood. By focusing on this mismatch as a predictor, we aligned ourselves with Hanson (1994) and her conceptualization of “lost talent.” Post-hoc tests revealed that the mediation effect of this gap between aspirations and expectations was highly significant (p < .01), with about 24% of the total effect of teen motherhood on college attendance having been mediated.

Discussion

The evidence presented here revealed that teen motherhood, among the current sample of African American mothers in late adolescence, was not significantly associated with an increase in educational aspirations, at least as they relate to college graduation. Encouragingly, it was not associated with a decrease in such aspirations either. Given that the large majority of young mothers already aspired toward a college degree pre-pregnancy and, hence, had nowhere to go but down, findings with regard to aspirations should be considered with caution. Further, as mentioned earlier, the current study’s relatively narrow focus on college aspirations made it an incomplete test of SmithBattle’s (2007a) claims and, hence, the current findings are not wholly inconsistent with Smithbattle’s.

Findings with regard to changing expectations, however, were just as hypothesized. Despite similar pre-pregnancy levels of college expectations between mothers and non-mothers, respondents who entered into motherhood were much less likely than their non-parenting counterparts to expect a college degree in early adulthood. Consistent with these results, “lost talent” was much more common among young mothers than among non-mothers. Such findings were supportive of the possibility that the structural constraints noted by SmithBattle (2007a) to be associated with teen motherhood and, perhaps, the cultural constraints of popular stereotypes are internalized by young mothers as insurmountable barriers to achieving their educational aspirations. Taken together, findings for both aspirations and expectations were consistent with the notion that “expectations get adjusted to the real world” (Kerckhoff 1976, :371), while aspirations are often derived from dominant cultural norms and, thus, are much less dependent on social circumstance.

Not only did adolescent mothers in this sample experience a decrease in educational expectations following the birth of their child, but it appears that this change in expectations partially explained the gap between teen mothers and non-mothers in educational attainment. By early adulthood, young mothers were significantly less likely than their non-childbearing peers to have any college education. This finding, along with the finding that college aspirations had little effect on college attendance after considering the effect of expectations, calls into question the importance of high aspirations among teen mothers since such aspirations appear to be effectively countered by decreased expectations.

Prior to discussing the potential implications of these findings for guiding future research and, perhaps, for informing the development of effective intervention practices, we first address the three major limitations of this study. First and foremost, although the current study reduced sample selection bias by drawing data from a probability sample, it may have introduced a different form of selection bias by dropping respondents who experienced multiple pregnancies or who experienced a pregnancy outside of the selected time frame. Although this bias was necessary to ensure the proper time-ordering of events and to ensure comparability of respondents’ motherhood experiences, it produced a very homogenous sample in terms of age at first pregnancy. This homogeneity necessarily restricts the study’s generalizeability, and thus, future studies are needed to test its generalizeability to a more diverse population of adolescents.

Secondly, the present study lacked nuance in the measurement of educational aspirations, expectations, and attainment. The dichotomous nature of the dependent variables and their narrow focus on college provided a very crude indicator of changing aspirations and expectations and their ability to predict educational attainment. Given this lack of nuance, this study was unable (1) to assess academic aspirations and expectations other than those related to college graduation, (2) to explore differences in the magnitude of young mothers’ aspiration/expectation mismatch, and (3) to examine the influence of such a mismatch on academic success (e.g. GPA, graduation, etc.). Future studies, then, should explore the production of aspiration/expectation gap and its implications with more subtlety.

Lastly, the length of times between study waves was arbitrary and relatively short. Because past work (e.g. Furstenberg 2007) has shown that some teenage mothers continue their educational pursuits well into adulthood, the educational disparities shown here between young mothers and their nonparenting peers in young adulthood should not be considered absolute and enduring.

Despite these limitations, the present study revealed that the majority of teen mothers did indeed aspire toward a college degree both before and after childbirth, yet after childbirth many of these mothers no longer expected to earn one. Such decreased expectations partially explained the lack of college attendance among young mothers in early adulthood. Hence, it appears that the cultivation of teen mothers’ educational ambitions by school officials, policy makers, and other individuals who shape these women’s life chances may not be enough to allow these young mothers to fulfill their educational goals and to close the achievement gap between themselves and their non-childbearing peers. Although such efforts are laudable and probably necessary, the present findings indicate that simultaneous efforts must be made to prevent the expectations of teen mother’s from dropping. In order to determine how this could best be done, much more research that pays particular attention to the distinction between aspirations and expectations is needed.

Future research might seek to determine, for instance, the process through which teen motherhood gives rise to decreased educational expectations and when in this process expectations begin to fall. More specifically, do structural constraints experienced by young mothers slowly erode expectations? Or, do expectations begin to deteriorate immediately upon knowledge of a pregnancy, perhaps due simply to cultural discourses andthe anticipation of future constraints? No matter the process, however, education expectations, unlike aspirations, are linked primarily to barriers to educational success, whether real or perceived. Hence, the minimization of both cultural and structural barriers experienced or anticipated by teen mothers appears to be a worthwhile means to prevent falling expectations. The work of Klaw and colleagues (Klaw 2008; Klaw, Rhodes, and Fitzgerald 2003) supports this notion, as such work indicates that factors like social support and access to child and health care can enhance young mothers’ educational persistence. Further, the work of Hellenga, Aber, and Rhodes (2002) suggests that the aspirations-expectations mismatch among young mothers, at least with respect to vocation, is not universal, as those with greater resources (e.g. access to nonrelative child care) did not report such a mismatch. Hence, by beginning to contest existing cultural discourses that paint young mothers as destined for disadvantage (Furstenberg, 2007) and to dismantle the accompanying stratification system that confines their educational opportunities, young mothers might not have to readjust their expectations to a reality full of constraints, and teen motherhood will no longer be tantamount to “lost talent.”