ABC, It’s as Easy as 1 2 3: Parents as Early Teachers

Since 1983, with the publication of A Nation at Risk, citizens and the government alike have been acutely aware of what has become known as the “achievement gap.” This gap refers to differences in student achievement along racial and socioeconomic lines, and has spurred numerous interventions, from the No Child Left Behind legislation to the “No Excuses” charter school movement and the Thirty Million Words Initiative.

Previous research has shown that the achievement gap can “open” at as early as nine months of age. With this research, it has become more evident that learning disparities between low-income children and their more affluent counterparts begin in the home. This proves to be an interesting public policy challenge, as preschool services for low-income children typically begin at age three and can only serve a fraction of the population. Head Start, one of the largest and most well known programs for low-income children, only serves 35 percent of those in need and spends around $7,600 per child to do so, costing taxpayers close to $8 billion per year.

Debates around the impact of Head Start and other similar programs aside, the sheer cost alone should make us wonder if there could be a more effective way to provide early education services to a larger proportion of low-income children. New research from Roland Fryer, John List, and Steven Levitt may help us uncover a way to do just that. Their NBER working paper, Parental Incentives and Early Childhood Achievement: A Field Experiment in Chicago Heights, details a randomized field experiment in which they offer parents incentives to engage in activities and behaviors to increase their children’s cognitive and non-cognitive development.

For the study, there were three groups of four-year-old children and their parents, recruited through a large marketing campaign in Chicago Heights. One of these groups served as the control group, and the other two as treatment groups. Biweekly “Parent Academy” sessions served as the focal point of the study. The Parent Academy taught parents how to help their children develop crucial skills such as spelling, counting, memory, and self control. Participating children were also assigned periodic “homework” assignments, in addition to some cognitive/non-cognitive tests. Families in the treatment groups were paid for attending the biweekly Parent Academy sessions and for the children’s performance on the homework and cognitive/non-cognitive tests. In one treatment group, families received this payment immediately, in the form of cash. In the other group, the payment was placed in a college trust account that could be accessed when the children enrolled in college.

The researchers used student test scores at the end of the program to estimate two different measures of the impact of the Parent Academy. Intent-to-Treat (ITT) estimates showed the impact of being offered a chance to participate in the program, regardless of the families’ compliance with the program. The Local Average Treatment Effect (LATE) estimates showed the impact of actually participating in the program. The aggregated estimates for non-cognitive development, for example memory and self-control, were statistically significant. LATE estimates for the cash and college groups were 0.225 standard deviations (SDs) above the mean and 0.217 SDs above the mean, respectively. All scores were made to fit a normal distribution with a mean of 0 and standard deviation of 1. Reporting impact as a standard deviation allows for an easier comparison across tests within the study, and even across programs from other studies. A 1 SD improvement would put students in the 84^th percentile of performance, meaning they perform better than 84 percent of four-year-olds. Here, the SDs correlate with falling roughly into the 58^th percentile (at the mean, students are in the 50^th percentile).

The graphic above depicts the relative size of the impact on students by race, in addition to showing a comparison between the impacts generated by this program and those seen by other major, early education programs. Interesting impacts emerge when students are disaggregated by race and skill level upon entering the program. Gains for Hispanic students are quite large, 0.367 SDs on cognitive skills and 0.428 on non-cognitive skills. Gains for White students are even larger, 0.932 and 0.821 SDs, respectively. For comparison, the impact of Head Start on test scores is 0.145 SDs, and the impact of Perry Preschool on achievement at age 14 years is 0.203 SDs. Unfortunately, the impact on Black students is negative but not statistically significant. The researchers believe that the difference between Black and Hispanic students on cognitive domains may be partially explained by the “Home Language Hypothesis.” This hypothesis essentially states that Hispanic students who speak Spanish at home will gain the most by attending English-speaking preschools.

Additionally, the researchers found that students who entered the program above the median on non-cognitive measures experienced greater success, seeing a 0.3 SD impact across both domains. Those students who were above the median on non-cognitive skills, but below the median on cognitive skills, gained the most of all. The researchers hypothesize that these differences are due to the fact that non-cognitive skills are a “necessary input for learning.”

This research has some interesting policy implications. The first is that a program that requires much less input and utilizes parents as teachers can have significantly larger impacts on student achievement than some of the most heralded, early education programs. This program, which requires less space and fewer teachers, could allow for more children to be served, and at a younger age than traditional preschool. While we may not be comfortable with the idea of putting children in school at the age of one, many people may agree that empowering parents to teach their children at that young age is perfectly acceptable. The second implication is related to cost. This program was run at a cost of $3,600 per child. At that rate, this program could serve over twice as many children as Head Start with the same budget and with larger impacts for White and Hispanic children. These implications are something to consider as policymakers decide how to provide high-quality, early education experiences to more children and for less money.

Article Source: Fryer Jr, Roland G., Steven D. Levitt, and John A. List. “Parental Incentives and Early Childhood Achievement: A Field Experiment in Chicago Heights.” No. 21477. National Bureau of Economic Research, 2015

Featured Photo: cc/(Governor Kate Brown)