«Abstract This study analyzes how parental investment responds to a low birth weight (LBW) outcome and finds important differences in investment ...»
Parental Investment Responses to a Low Birth Weight Outcome
Who Compensates and Who Reinforces?
Brandon J. Restrepo
U.S. Food and Drug Administration
February 24, 2016
This study analyzes how parental investment responds to a low birth weight
(LBW) outcome and finds important differences in investment responses by maternal
education. High school dropouts reinforce a LBW outcome by providing less investment
in the human capital of their LBW children relative to their normal birth weight children whereas higher-educated mothers compensate by investing more in their LBW children.
In addition, an increase in the number of LBW siblings present in the home raises investment in a child, which is consistent with reinforcement, but this positive effect tends to be concentrated among high school dropouts. These results suggest that studies analyzing the effects of LBW on child outcomes that do not account for heterogeneity in investment responses to a LBW outcome by maternal education may overestimate effects of LBW on child outcomes for those born to low-educated mothers and underestimate such effects for those born to high-educated mothers.
J.E.L. Classification D13, I14, I24, J13 Keywords Parental investment; human capital; low birth weight; education; income
1. Introduction There is strong evidence that the endowment of human capital at birth affects a child’s longrun health, education, and labor market outcomes.1 A question of long-standing interest to economists is whether parents allocate resources among their children so as to reinforce or compensate for differences in their children’s endowments. If parents reinforce differences in child endowments by investing more in the human capital of their better-endowed children, as predicted by the theoretical model of Becker and Tomes (1976), then inequality resulting from differences in their children’s endowments will be accentuated. However, if parental preferences for equity outweigh the concern for efficiency, then parents will compensate for differences in child endowments in order to reduce inequality among their children (Behrman et al. 1982).
There are numerous studies that estimate the effect of child endowments on parental investment.2 However, estimation of the average effect of child endowments on parental investment behavior may mask important heterogeneity across families, such as variation in the direction and magnitude of investment responses by a family’s socioeconomic status (SES).
Parental investment responses to an adverse endowment shock may vary by a family’s SES because expectations for child quality, parenting knowledge and skills, and the availability of resources to respond to such an endowment shock may vary by SES. For example, parental education may affect (a) preferences for equity versus efficiency, (b) the productivity of investments in poorly endowed children, and (c) knowledge about potential ways to compensate for an adverse endowment shock, leading some parents to reinforce and others to compensate.
And since poor families are more likely to face credit constraints, they may be unable to compensate for an adverse endowment shock even if they have strong preferences for equity among their children.
A small but growing literature has examined whether there is heterogeneity in the impact of child endowments on parental investment by SES. Datar et al. (2010) found that, while on average early-life health and educational investments reinforce a low birth weight (LBW) outcome, they found no evidence of heterogeneity in investment behavior by income or education. In contrast, Hsin (2012) found that on average parents neither compensate nor reinforce a LBW outcome, but lower-educated mothers reinforce a LBW outcome by spending less total and developmental time with their LBW children relative to their NBW children while higher-educated mothers compensate for a LBW outcome by spending more time with their LBW children. A possible explanation for the divergent findings regarding heterogeneity in investment For example, LBW children have been shown to achieve poor health, cognitive and behavioral development, education, employment, and earnings outcomes relative to NBW children (Currie and Hyson 1999; Case et al. 2005;
Black et al. 2007; Currie and Moretti 2007; Oreopoulos et al. 2008; Fletcher 2011; Datta Gupta et al. 2013; Chatterji et al. 2014a,b; Figlio et al. 2014; Cook and Fletcher 2015).
Most studies have found evidence that parents reinforce endowment differences among their children (Rosenzweig and Schultz 1982; Rosenzweig and Wolpin 1988; Behrman et al. 1994; Behrman and Rosenzweig 2004; Rosenzweig and Zhang 2009; Aizer and Cunha 2012; Datar et al. 2010; Frijters et al. 2013; Parman 2015). Several studies have found that parents compensate for differences in their children’s endowments (Griliches 1979; Behrman et al. 1982;
Pitt et al. 1990; Ashenfelter and Rouse 1998; Ermisch and Francesconi 2000; Del Bono et al 2012; Bharadwaj et al.
2014). And a few studies have found that parents neither compensate nor reinforce endowment differences (Royer 2009; Almond and Currie 2011a; Lynch and Brooks 2013). See Almond and Mazumder (2013) for an excellent review of the literature.
behavior by SES is that, because human capital is multidimensional (Cunha and Heckman 2008;
Cunha et al. 2010), parents may simultaneously reinforce and compensate for endowment differences along different dimensions of human capital and such multipronged within-family investment strategies may vary by SES.3 I contribute to the literature by analyzing the impact of LBW—and heterogeneity in its impact by education and income—on parental investment measures that have been shown to affect both cognitive and noncognitive child development outcomes. Birth weight is easily observed by parents, has been shown to influence parental investment behavior and, while a child’s endowment may have multiple dimensions, features of the endowment other than birth weight are less easily observed by both parents and researchers. The Home Observation for Measurement of the Environment (HOME) assessment, which captures a wide variety of inputs that influence the quality of the home environment, is used as a measure of the time and goods investments provided by parents to promote two important dimensions of child human capital.
The cognitive stimulation subscale measures investments that promote the cognitive development of children such as the frequency with which parents read stories to children and help children with numbers and the alphabet. The emotional support subscale measures behavior such as how parents discipline their children and the amount of affection directed towards children, which promotes the social and emotional development of children. Previous studies have found that HOME scores increase cognitive and noncognitive skills,4 both of which shape education and labor market trajectories.5 And the cognitive stimulation and emotional support HOME scores are particularly useful measures of investments in the human capital of LBW children, since these children have been shown to be at increased risk of developing cognitive deficits, as well as emotional and behavioral problems (McCarton 1998).
Using data on children born to female respondents in the 1979 cohort of the National Survey of Longitudinal Youth and a within-family estimation approach to identify the effect of child endowments on parental investment, I find that on average mothers reinforce a LBW outcome and that there are important differences in investment behavior by maternal education and family income. The investment response to a LBW outcome is strongly compensatory in families with high-educated mothers and strongly reinforcing in families with low-educated mothers. Mothers Using data on siblings from rural Ethiopia, Ayalew (2005) found evidence that parents reinforce endowment differences with respect to educational investments, but compensate with respect to health investments. And using data on Chinese twins, Conti et al. (2015) found that the twin who experienced a negative early-life health shock receives more health investment later in life relative to her twin sibling who did not but receives less educational investment; they conclude that families behave so as to equalize investments across children who are differentially affected by early-life health shocks.
Recent studies find evidence that the quality of the home environment is important for the cognitive and noncognitive development of children (Blau 1999; Guo and Harris 2000; Brooks-Gunn et al. 2002; Aughinbaugh and Gittleman 2003; Todd and Wolpin 2007; Cunha and Heckman 2008; Cunha et al. 2010). For example, Todd and Wolpin (2007) find that 10-20% of the math and reading test score gaps between whites and nonwhites can be closed if home inputs (as measured by the overall HOME score) were equalized at the average level observed for white children. Blau (1999) estimates the elasticities of child development outcomes with respect to cognitive stimulation and emotional support to be in the 0.07-0.14 range, which are much larger than the estimated elasticities with respect to other inputs such as child care.
A growing body of work shows that both cognitive and noncognitive skills are important for socioeconomic success (Heckman et al. 2006; Carneiro et al. 2007).
who are high school dropouts reinforce a LBW outcome by providing about 12% of a standard deviation (SD) less emotional support to their LBW children relative to their NBW children, while mothers with 4 years of college or more compensate by providing about 16.5% of a SD more cognitive stimulation and 7.5% of a SD more emotional support to their LBW children. In addition, an increase in the number of LBW siblings has a positive effect on investment in a child, which is consistent with reinforcement, but the magnitude of the reinforcing investment response is much larger among low-educated mothers. The investment differences by family income are qualitatively similar to those by maternal education, but are generally smaller in magnitude.
I make two important contributions to the literature. First, I show that the response of parental investment to a LBW outcome differs across families by maternal education and family income.
If variation in a LBW outcome within disadvantaged families were inconsequential, then my findings would be of little interest. However, I find that, among high school dropouts, nearly onequarter of multiple-child families have at least one LBW and one NBW child. Differential investment responses to a LBW outcome may limit upward economic mobility for LBW children born to low-educated mothers, since they reinforce the adverse consequences of LBW.
Second, my findings have important implications for research on the consequences of a LBW outcome. The well-known fetal origins hypothesis states that the intrauterine environment programs adult disease (Barker 1990) and there is a wealth of evidence that economic outcomes are also influenced by fetal and infant health (Almond and Currie 2011a,b). However, the observed association between infant health and child outcomes may capture the effects of both biological factors and parental investment responses to a child’s infant health outcome. The investment patterns by maternal education that I observe suggest that, for example, the biological effects of LBW emphasized by Barker (1990) may be mitigated by high-educated mothers and aggravated by low-educated mothers.
The paper proceeds as follows. First, I describe the data used in the analysis. Second, I outline the empirical approach I employ. Third, I present and discuss the results. And finally, I conclude.
The data used in this study come from the National Longitudinal Survey of Youth-Child and Young Adult (C-NLSY) file, which contains a rich array of information about the children born to female respondents of the National Longitudinal Survey of Youth 1979 (NLSY79). The NLSY79 began in 1979 with a sample of 12,686 individuals between the ages of 14 and 21.
Detailed information about children born to women in the NLSY79 sample has been collected biennially since 1986 and I use data through 2012 in the analysis. Over this period, a total of 11,512 children have been born to 4,932 women in the NLSY79 sample and over 75% (3,757) of these women have had two or more children.