Areas of Expertise


Achievement Gaps
Literature Library


What are the Achievement Gaps?

  1. Achievement Gap in Income

Heredity and Environment: What Role Do Family Characteristics Play in Explaining the Black-White Test Score Gap?

A group of authors explores the effects of family environment on the black-white test score gap.

This reports some of the ideas and findings from the following source:

Phillips, M., Brooks-Gunn, J., Duncan, G. J., Klebanov P., & Crane, J. (1998). Family background, parenting practices, and the black-white test score gap. In C. Jencks and M. Phillips (Eds.),The black-white test score gap(pp. 103-145). Washington, DC: Brookings Institution. Pp. 111-140

To see other reports that originated from this same citation, go to the bibliography.

Richard Herrnstein's and Charles Murray's 1994 book,The Bell Curve, received widespread attention by arguing that social and economic inequalities between black and white Americans were weak predictors of children's test scores. The alternative, these authors argued, was that the black-white test score gap resulted, to some significant degree, from hereditary differences between black and white Americans.

This heredity-versus-environment controversy is what motivates scholars Meredith Phillips, Jeanne Brooks-Gunn, Greg J. Duncan, Pamela Klebanov, and Jonathan Crane to revisit the influence of family characteristics on young children's intelligence test scores.

Defining and Measuring Family Characteristics Is a Critical Task

Phillips and her colleagues say that many studies that look at the influence of socioeconomic factors on children's test scores find that these factors account for no more than about one-third of the black-white test score gap. However, family situations and influences are exceedingly complex. Phillips and her colleagues argue that by defining family characteristics more broadly, we can explain substantially more than one-third of the test score gapperhaps as much as two-thirds of the test score gap.

But measuring family characteristics is a tricky business if the goal is to disentangle hereditary from environmental influences. After all, heredity itself is a family characteristic. The challenge, then, is to try to tease out genetic,as opposed to social, influences that families have on their children's test performance. Herrnstein and Murray point to this difficulty when they argue that some apparently "environmental influences" may really be masking underlying genetic effects.

Phillips and her colleagues approach this problem in three steps:

  1. They look at how much impact different family characteristics have on the black-white test score gap.
  2. They try to estimate how much genetic "contamination" is present in their measures of family characteristics.
  3. They present reasons why their estimates may exaggerate the effects of either hereditary or environmental influences.

The Effect of Family Characteristics on the Black-White Test Score Gap

The authors begin by comparing the difference in the performance of black and white 5- and 6-year-olds on standardized preschool vocabulary tests. They then estimate the relative effects of different family characteristics by seeing how much the black-white gap decreased when different characteristics were taken into account.

Table 1 presents a summary of the effects that Phillips and her coauthors found by studying two large samples of children and their mothers. The row at the top of Table 1 lists the difference between black and white children's test scores (white children scored on average 16.27 points higher than black children).

Each other row in the table identifies "how much" of the 16.27 point difference was due to different family characteristics. For instance, if black and white parents had the same number of years of education, the difference between their children's scores would decrease by 0.2. This indicates that when black and white parents spend equal amounts of time in school, their children's scores become slightly more similar.

Table 1. Summary of the Effects of Family Characteristics on Children's Test Scores

Initial Black-White Test Score Gap


Family Characteristic


Amount of Gap Explained

Years of parents' education

Do black and white parents have the same number of years of education?


Mother's family background

Do the parents of black and white mothers have equal levels of education and occupational status?


Quality of mother's schooling

Did white and black mothers go to schools of equal quality? For instance, were their schools similar in terms of the number of economically disadvantaged students, racial integration, teacher quality, student-teacher ratio, and teacher turnover? Were their schools equally safe?


Educational outcomes

Do black and white parents know the same amount when they leave school (even if they have similar schooling)? For instance, do they show similar scores on the Armed Forces Qualification Test?


Family income

Do black and white parents have equal income and wealth (controlling for things such as level of education)?


(The authors note that this effect is larger for other data sets using different standardized tests.)

Parent attitudes and values

Do black and white parents have the same educational expectations for their children? Do they show the same amount of self-efficacy and self-esteem?


Grandparents' influence

Did black and white children's maternal grandparents live in the same region? Did they receive the same amount of schooling? Did they have the same occupational status? Did they have the same reading material in the home? Did they have the same number of children?


(This is not a direct influence. The authors say that grandparent influences affect children, for the most part, through the mother's parenting practices)

Parenting practices

Do black and white parents provide the same quality home environment for their children? Do they take them on learning experiences outside the home? Do they participate in literary experiences with their child? Do they have cognitively stimulating experiences in the home? Do they use the same punishment methods? Are they equally warm with their children? Is the physical environment of the home equally clean and safe?


It is important to note that the numbers in the right-hand column of Table 1 do not "add up" in a straightforward manner. That is, when all of the influences are taken into account in the authors' statistical models, the "importance" of some characteristics drops. This is because Table 1 does not give the reader a sense of how the different family characteristics interact with each other. What Table 1 provides is a perspective on the relative importance of different family characteristics.

The authors conclude that the racial difference between black and white families is larger than Herrnstein and Murray recognized. They argue that when more family factors are taken into account, the environment accounts for two-thirds of the black-white test score gap.

Genetic Contamination in Measures of Family Characteristics

Phillips and her colleagues note, however, that family characteristics may themselves have a genetic base.

How can we tell how much hereditary influences "contaminate" environmental explanations?

Phillips and her coauthors create two proxies: one for environmental influences and one for hereditary influences. The environmental influences include the range of different family characteristics described above. The hereditary influences include rough measurements of the mother's genetic cognitive contribution to her children, including:

  • her AFQT scores
  • her class rank in high school
  • the interviewer's assessment of her understanding of the interview

The authors then estimate how much the genetic influence proxy reduces the explanatory importance of the environmental influence proxy. They find that about 26% of the apparent effect of family environment is really a genetic effect.

Why Estimates of Environmental Influences May Either Be Higher or Lower

Even after these two steps, Phillips and her colleagues say that they still cannot accurately measure how much effect family environmental characteristics have on the black-white test score gap. The reason for this is that the way that they measured environmental influences may have given either too much or too little weight to the real effects of family environment.

They list factors that might cause their estimates of the relative importance of heredity and environment to be off.

Factors that might overestimate of the importance of environment:

  • Father's genetic contribution was not taken into account. This might cause the environmental factors to have too much influence.
  • The ways that a mother's genetic contribution to her children was measured were imperfect.Armed Forces Qualification Test scores, class rank, and interview assessment of the mother's understanding are all estimates of a mother's hereditary gift to her children. Until cognitively relevant genes can be identified and mapped, we cannot be sure exactly what a mother passes to her children.
  • Genes for non-cognitive traits were not controlled. The authors say that some non-cognitive hereditary dispositions (e.g., toward depression) can actually hamper or facilitate cognitive development.

Factors that underestimate the importance of environment:

  • A mother's cognitive skills may increase the effectiveness of environmental factors. Very smart mothers may provide very stimulating environments for their children. Not only would her children get the advantage of her genes, but they would be additionally helped by the rich environment she creates for them. This would underestimate the influence of the family environment.
  • There are still many aspects of a child's environment that were not measured.For instance, whether a child was exposed to lead poisoning or whether they were exposed to standard English are environmental factors that could influence scores on vocabulary tests.

The Bottom Line

Phillips and her colleagues conclude that we simply do not have the ability yet to disentangle hereditary from environmental influences on achievement. While many studies find that socioeconomic factors account for only about one-third of the black-white test score gap, Phillips and her colleagues believe that broader measures of children's home environments may account for up to two-thirds of the gap.

Their research does not settle the question of whether the black-white test score gap is partially or entirely environmental. However, say Phillips and her co-authors:

"[I]t does help identify the family characteristics that matter most for the gap. It also suggests that eliminating environmental differences between black and white families could go a long way toward eliminating the test score gap." (p. 138)

In short, eliminating social and economic inequities between black and white Americans may not get rid of the test score gap, but it would most likely lessen the gap significantly.

Research Design:

Research Question

What is the effect of family characteristics on the black-white test score gap?

What are the relative effects of small children's heredity versus their environment on their vocabulary scores?


Phillips and her colleagues draw their data from two sources:

  1. Children of the National Longitudinal Survey of Youth: 1,626 African-American and European-American five- and six-year-olds who took the Peabody Picture Vocabulary Test-Revised (PPVT-R).
  2. Infant Health and Development Program (IHDP): 315 black or white children who also took the PPVT-R. The IHDP children also took another aptitude test: the Wechsler Preschool and Primary Scale of Intelligence (WPPSI).

Phillips and her colleagues sought to determine the relative importance of a wide range of family characteristics for children's vocabulary test scores. They did this by running statistical models in which they would factor in different influences and examine how the included variables changed the differences in black and white children's test scores.


This research was funded, in part, by the Mellon Foundation, the Spencer Foundation, and the National Institute for Child Health and Human Development's Research Network on Child and Family Well-being.


Copyright © 2008 Learning Point Associates. All rights reserved.