Their parents are generally well-educated and affluent. They receive more time and educational resources from those parents than the average child gets from theirs. Yet they get into more conflicts with their classmates at school, display relative little interest and enthusiasm about learning tasks, and register only middling academic performance. About whom are we talking? Adopted children. This is the paradox of adoption in America.

Reliable estimates on the school adjustment and academic performance of adopted children have been hard to come by. State laws bar children from being identified as adopted in school records or vital statistics. Children living with two adoptive parents represent only one percent of all children in the U.S., so there are relatively few of them in national sample surveys of children and families. When special surveys on adopted children have been done, they have relied on parents as informants.1

To expand what we know about adopted students, for this Institute for Family Studies research brief, I carried out a fresh analysis of data from a large longitudinal study of 19,000 kindergarten students that was conducted by the National Center for Education Statistics beginning in 1998. This study, known as the ECLS-K, contains enough cases of adoption (160 children and families) for some reasonably robust estimates of its effects, at least in the early grades. Because the study began in kindergarten, we know the adopted children were all adopted early in their lives, if not at birth. The ECLS-K has the further advantage of being based on direct assessment of student knowledge and skills, and teachers’ rather than parents’ reports about classroom conduct.2 This is the first study of adopted children’s school behavior that is based on independent teacher reports and makes use of a representative national sample of students from adoptive families.

Previous analyses of federal survey data indicate that adoptive families tend to be better off financially than other families with children in the United States.3 This is partly due to self-selection and partly because of the screening that adoptive parents must go through before they are allowed to adopt. In addition, adoptive parents have higher levels of education and put more effort into caring for their children than biological parents do.4 Yet my analysis shows that adoptees do not do as well in school as one would expect from their highly advantaged home environments. The results call into question the widely held assumption that larger investments of money and time in children can overcome the effects of early stress and deprivation and genetic risk factors.

Classroom Conduct

Kindergarten and first-grade teachers were asked to rate the classroom behavior of children in the ECLS-K sample—how well they got along with other children in a group situation. In both the fall of kindergarten and the spring of first grade, adopted children were more likely than biological ones to be reported to get angry easily and often argue or fight with other students. In kindergarten, the average problem behavior rating for the adopted group was at the 64th percentile, whereas for children with both birth parents, it was at the 44th percentile. (Higher percentile ranks indicate more problem behavior. The 50th percentile represents the average rating score for all students.) Children in single-parent, step, and foster families were at the 58th percentile as a group, not significantly different from the adopted group.

These gaps in problem behavior scores in kindergarten widened slightly after adjusting for differences across groups in average child age, parent education level, and racial-ethnic and gender composition, as shown in Figure 1. The adjusted average problem behavior rating for the adopted group was at the 66th percentile; for children with both birth parents, it was at the 45th percentile. After adjustment, adopted children showed significantly more problem behavior than children in single-parent, step, and foster families, who were at the 56th percentile as a group.

zill draft fig1

A similar pattern was evident in the behavior ratings provided by first-grade teachers. Adopted students were at the 63rd percentile in raw teacher ratings of problem behavior, vs. the 43rd percentile for students from families with two biological parents. Students from other types of families were at the 60th percentile, not significantly worse than adopted students. After adjustment, the adopted group was at the 64th percentile, significantly worse than the both birth-parent group (42nd percentile), and not significantly different from students from single-parent, step, and foster families (56th percentile).

Positive Approach to Learning Activities

Children in the ECLS-K were also rated by their teachers on how well they paid attention in class, whether they seemed eager to learn new things, and whether they persisted at challenging learning tasks. Scores on these measures have proven to be predictive of later academic performance and career success beyond elementary school.5 Adopted children were rated less highly with respect to such positive approaches to learning than were children being raised by both birth parents. In kindergarten, the adopted group’s average rating score was at the 45th percentile, whereas for the two-biological-parent group, the average was at the 57th percentile. The learning-approach ratings for adopted students were not significantly different from those for students living with single or remarried birth parents or foster parents, who were at the 43rd percentile. (Higher percentile ranks indicate a more positive approach to learning activities; the 50th percentile again represents the average rating score for all students.)

After adjustment for age, sex, parent education, and racial-ethnic differences, the adopted group—now at the 42nd percentile—showed a significantly less positive approach to learning than children living with two biological parents, whose adjusted ratings were at the 56th percentile. The adopted group was also lower than the single-, step-, and foster-parent group, now at the 47th percentile, although this difference was only marginally significant.

zill draft fig2

The approach-to-learning ratings of first-grade teachers revealed a similar pattern. Prior to adjustment, adopted children were at the 44th percentile, significantly lower than students with both birth parents (56th percentile) and not significantly different from students from other family types (39th percentile). After adjustment, the adopted group was at the 39th percentile, significantly lower than the two-birth-parent group (53rd percentile) and not significantly different from the students from other family types (42nd percentile). The adjusted percentiles for both kindergarten and first-grade teachers’ ratings are shown in Figure 2.

Early Reading Skills

In addition to gathering teachers’ reports of children’s behavior and attitudes, the ECLS-K tested children’s academic skills. Here, too, disparities emerge between students from different family types, though they are generally smaller in magnitude than on the measures reported above.

As the participating children began kindergarten, the ECLS-K assessed their pre-reading skills, such as recognizing letters by name, associating sounds with letters, identifying simple words by sight. Adopted children seemed to perform these tasks fairly well: their  average score on the reading assessment reached the 53rd percentile. This was not significantly different from the average score of children being raised by both birth parents (55th percentile). And the adoptees outperformed children from other family types (single-parent, step, or foster families), whose observed average score was at the 38th percentile.

When the early reading scores were adjusted for differences across groups in age, sex, parent education, and racial-ethnic composition, a different picture emerged, however. As shown in Figure 3, the adjusted average score of the adopted group fell to the 44th percentile, primarily because of the high educational attainment of their adoptive parents. The adopted group was now not significantly different from the “other family types” group, whose adjusted score rose to the 45th percentile. The adjusted score for children with both birth parents also went down—to the 52nd percentile—but remained significantly higher than that of “other family types” group.

zill draft fig3

The reading skills of nearly all students in the ECLS-K sample were assessed again in the spring of first grade. Children were tested on grade-appropriate skills like reading and understanding words in context. Once again, the reading skills of children with both birth parents (56th percentile) exceeded those of children in single-, step-, and foster-parent families, who were at the 41st percentile. Adopted children, at the 49th percentile, were not significantly more advanced than the latter group.

The picture was similar after adjustment for demographic and socioeconomic differences across groups. Children with both birth parents, now at the 53rd percentile, were significantly more advanced than children in single-, step-, and foster-parent families, now at the 46th percentile. Adopted children, now at the 41st percentile, were not significantly behind the latter group (the apparent difference was not statistically significant due to a larger margin of error in this comparison).

Early Mathematics Skills

In the fall of their kindergarten year, the ECLS-K assessed children’s pre-arithmetic skills like counting by rote, recognizing written numerals, and understanding greater, lesser, and equal relationships. Again, the adopted children did fairly well. Their average raw score was about the same as that for all U.S. kindergartners, that is, at the 50th percentile. However, they did less well than children with both birth parents, whose average math skills were at the 59th percentile. And their performance was not significantly better than that of children from single-parent, step, and foster families, who were at the 41st percentile (again, the apparent difference was not statistically significant due to a larger margin of error).

When children’s math scores were adjusted for the same demographic and socioeconomic factors mentioned above, the performance of the adopted group was significantly worse than that of both the birth-parent group and the “other family types” group. As shown in Figure 4, the adopted group’s adjusted math score was at the 39th percentile, while the adjusted score of the both-birth-parent group reached the 52nd percentile and that of children from other family types was at the 47th percentile.

zill draft fig4

In first grade, ECLS-K children were tested on grade-appropriate math skills like carrying out simple arithmetic operations. Students with both birth parents were at the 56th percentile, significantly more advanced than those from single-parent, step, and foster families, who were at the 38th percentile. Adopted first graders, at the 40th percentile, were not significantly ahead of that group. After adjustments for demographic and socioeconomic disparities, the percentile rank for adopted children fell to the 32nd percentile. Their adjusted math performance was significantly inferior to that of biological children in two-parent families (52nd percentile) and children in other family types (46th percentile).

Why Don’t Adopted Children Do Better?

The data presented in this research brief show that adopted children in kindergarten and first grade display above-average levels of problem behavior, exhibit below-average levels of positive learning attitudes, and score below average on reading and math assessments, despite their advantaged family background. Why don’t the plentiful resources and strenuous nurturing efforts of adoptive parents lead to better classroom conduct and higher achievement by their adopted children? Possible reasons why family resources do not always produce great outcomes may be found in attachment theory, traumatic stress theory, and behavior genetics.

Attachment theory holds that a warm, intimate, and continuous relationship with at least one adult, usually the mother, is essential for the mental health of infants and young children. Children who do not develop a stable and secure bond during early childhood, or have the bond disrupted, are subject to both short-term distress reactions and longer-term abnormalities in their feelings and behavior toward other people. Not having a stable maternal bond is apt to produce long-lasting deficits in the child’s social development, deficiencies that are not easily remedied by a new home environment, no matter how favorable.

Some adopted children experienced neglect, abuse, or other stressful events prior to their adoption. According to traumatic stress theory, the likelihood of long-term emotional scars depends on the intensity and duration of the stress. Severe or prolonged early stress can have long-lasting effects on a child’s development, effects that a supportive adoptive family may only partly ameliorate.

Behavior genetics is relevant because adoptive parents usually cannot choose or control the genetic endowment of the children they adopt. Indeed, often they are not even fully informed about the educational attainments or occupations of the child’s birth parents, nor do they learn much about their family histories. In other instances, adoptive parents consciously select children with known disabilities, hoping that the parental care they provide will enable the children to thrive.

Because the educational attainments of adoptive parents are exceptionally high, the genetic endowment of most children available for adoption is likely to be less favorable to intellectual accomplishment than the endowments of their adoptive parents. No matter how much intellectual stimulation and encouragement the parents provide the child, they may not be able to overcome the limitations of the child’s genetic heritage.

Probably all three of these theoretical perspectives have something to offer in accounting for the paradox of adoption in America. However, none of the findings presented here is meant to minimize the tremendous contribution that adoptive parents make to the children they take in or to society in general. Many adopted children do reasonably well in school and enjoy lives that are far better than they would have experienced had they not been adopted. And they do so at less cost and burden to the public than if the children were raised in foster homes or institutions.

At the same time, it is important for both prospective adoptive parents and policy-makers to be realistic about what adoption can and cannot accomplish. The availability of the “adoption option” does not do away with the need for better prevention of unplanned and unwanted conceptions, so that fewer children are born into high-risk situations where they are likely to experience neglect or abuse and become in need of adoption.

Nicholas Zill is a psychologist and survey researcher who has written on indicators of family and child wellbeing for four decades. Prior to his retirement, he was the head of the Child and Family Study Area at Westat, a social science research corporation in the Washington, D.C., area.

1. Bramlett, M.D., Radel, L. F., & Blumberg, S. J. (2007). The health and well-being of adopted children. Pediatrics, 119, S54–S60 (Supplement).
Zill, N. & Bramlett, M.D. (2014). Health and well-being of children adopted from foster care. Children and Youth Services Review, 40, 29–40.

2. West, J., Denton, K., & Germino-Hausken, E. (2000). America’s Kindergartners: Findings from the Early Childhood Longitudinal Study, Kindergarten Class of 1998-99, Fall 1998. U.S. Department of Education: NCES Report 2000-070.
Zill, N. & West, J. (2000). Entering Kindergarten: A Portrait of American Children When They Begin School. U.S. Department of Education: NCES Report 2001-035.

3. Bachrach, C. A., London, K., & Maza, P. L. (1991). On the path to adoption: Adoption seeking in the United States, 1988. Journal of Marriage and the Family 53: 705–18. Fisher, A.P. (2003). Still ‘not quite as good as having your own’? Toward a sociology of adoption. Annual Review of Sociology 29: 335–61.
Vandivere, S., Malm, K., & Radel, L. (2009). Adoption USA: A Chartbook Based on the 2007 National Survey of Adoptive Parents. Washington, D.C.: The U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation.

4. Hamilton, L., Cheng, S., & Powell, B. (2007). Adoptive parents, adaptive parents: Evaluating the importance of biological ties for parental involvement. American Sociological Review 72: 95–116.
Case, A. & Paxson, C. (2001). Mothers and others: Who invests in children’s health? Journal of Health Economics 20: 301–328.

5. Duncan, G., et al. (2007). School readiness and later achievement. Developmental Psychology 43: 1428–1446.