Excerpted from Chapter 1 of The Pre-K Debates edited by Edward Zigler, Ph.D., Walter S. Gilliam, Ph.D., and W. Steven Barnett, Ph.D. Copyright© 2011 by Paul H. Brookes Publishing Co. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Edward Zigler, Walter S. Gilliam, and W. Steven Barnett
Anything worth doing is worth doing well. Likewise, anything worth doing well is worthy of vigorous
debate regarding the goals of the effort and exactly how those goals should be achieved. Preschool education has grown into an idea that is worthy of debate. Indeed, vigorous debate has been an integral part of the evolution of the concept of early education. Since the 1960s, preschool education has grown from an experimental idea from researchers and scholars into one that is widely accepted by leaders in the worlds of policy, economics, and business. It is widely viewed as perhaps the best means for improving the educational and later-life outcomes of young children, addressing the racial and class gaps in educational achievement, and protecting our societal investment in education. The interested parties have grown from a handful of scholars to thought leaders across a myriad of ?elds, including business leaders, philanthropists, advocates, economists, lawyers, and public of?cials. This wide-ranging interest in early education is a strong indicator of the priority given to this topic and a societal acceptance of the importance of supporting our youngest and most vulnerable learners. Given this level of interest, debates are inevitable, and each issue of debate has its champions.
This book builds on the considerable evidence for the positive impact of high-quality early education, as well as the fact that many of our early education programs do not rise to the level of quality that research has consistently shown is possible to achieve and necessary for bene?cial impacts. Much has been learned about the essential ingredients of early education (e.g., Pianta, Barnett, Burchinal, & Thornburg, 2009; Zigler, Gilliam, & Jones, 2006a). The current debates are largely about the goals of public preschool education and how best to provide the ingredients needed to accomplish them on a large scale.
Rather than restating the rationale for early education, this book dives directly into the hottest
debates in the ?eld. For each issue of debate, the most vocal and erudite champions were identi?ed.
Each champion summarized his or her position and the evidentiary basis in a brief essay, and the essays were organized to place each debater in contrast to an opposing debater. In this way, the reader can
weigh the pros and cons for each position and emerge more informed as to what is known (and yet
unknown) about how best to provide early education in the United States. At the end of the book,
Martha Zaslow—one of the nation's most forward-thinking scholars on early education and child
care and Director of the Of?ce for Policy and Communication of the Society for Research in Child
Development—offers a summary and synthesis of the main topics of debates, highlighting areas of clear differences in opinion, as well as areas where some degree of consensus may be found.
To set the stage for the foray into these debates, Chapters 1 and 2 provide an overview of current
understanding of preschool education and the rationale for this intense interest in how best to provide it. Chapter 1, by Nobel Laureate economist James J. Heckman, provides an excellent discussion of many of the key concepts underpinning the importance of early education, whereas Chapter 2, by Sara D. Watson of the Pew Charitable Trusts, explains why these key concepts have so greatly captured the energy and enthusiasm of public of?cials and private philanthropy.
After this brief overview, Part I of the book moves directly into the issues of debate. Chapters 3–7 focus squarely on what is perhaps the biggest issue facing public-funded preschool: Should eligibility for publicly funded preschool be targeted speci?cally to low-income children or open to all children regardless of income? The debaters raise both practical and principled questions about alternative policies. Although the issue may appear to be one of polar-opposite opinions, a variety of middle-ground options
are also explored in the essays. Chapters 8–14 focus on what degrees or credentials preschool teachers
should possess. Again, the issue at hand is more nuanced than simply debating whether teachers should
possess a bachelor's degree or not. The debaters tackle this weighty issue from a variety of angles that
address both preservice training and ongoing teacher supports, as well as the ways in which the broader
contexts and conditions of early education may affect the answers. Chapters 15–20 address the goals of
preschool: Should public-supported preschool be focused primarily on cognitive or academic achievement, or should the mission be broader? Chapters 21–24 focus on discussing which locations and
administrative structures might best deliver preschool education. At present, public-funded preschool is provided across a wide array of provider organizations (e.g., public schools, private schools, Head Start
grantees, child care providers, other nonpro?t agencies). Which programs are best equipped to provide a
quality preschool experience to young children, and how can a system of early education best be created
from the programmatic building blocks that currently exist?
Parts II and III include issues that do not ? t neatly into debates, but they are important nonetheless. These topics address questions such as the following:
- How do we ensure quality and accountability in preschool programs? (Chapters 25–8)
- What should come before and after preschool? (Chapters 29–34)
- What lessons can be learned from state efforts to implement prekindergarten systems? (Chapters 35
- What might be the dangers in overselling preschool as a magical elixir to cure all of the ails in our
educational system? (Chapters 37 and 38)
The purpose of this book is not to settle these debates. Rather, it is to allow the interested reader an
opportunity to better appreciate the differing perspectives and opinions, understand the reach and limits of
the existing evidence in favor of each position, and form his or her own conclusion. As the French essayist
Joseph Joubert (1754–1824) opined, "It is better to debate a question without settling it than to settle a
question without debating it" (Lyttelton, 1899). With this in mind, this book is not intended to close any
debates, but rather to open them up for greater illumination and wider participation. Enter the debaters.
TARGETED VERSUS UNIVERSAL PRESCHOOL
- What are the three main criticisms of targeted preschool? What are the
- How does a universal program generate bene?ts that a targeted program
does not? Do these outcomes outweigh the added expense?
- Do you agree or disagree with the tactic of implementing universal pre-K and
then working to make it high quality? Conversely, what are the benefits of
starting with a high-quality, targeted pre-K program and then expanding it to
a universal program? (See examples in Chapter 6.)
The Economic Case for Targeted Preschool Programs
Art Rolnick and Rob Grunewald
As stewards of the public purse, policy makers are charged with allocating scarce resources to
investments that provide the greatest benefits relative to costs. Early childhood education (ECE)
targeted for at-risk children is such an investment. Economic research has made the case that investment in ECE for at-risk children provides extraordinary returns. Some of the bene?ts are private gains
for individuals in the form of higher wages later in life, but a majority of the bene?ts accrue to society
as a whole through reduced remedial education and crime costs and higher tax revenue.
However, returns to ECE investments are not all equal. The returns to universal programs (i.e.,
preschool programs available for all children) are much lower than returns to targeted programs.
Indeed, the returns to universal programs appear low even relative to other public investments. Consequently, based on the principle that resources should be invested in the highest return projects, ECE investments should be aimed at our most at-risk children.
In this chapter, we ?rst show that investments in at-risk children achieve substantially higher
returns than universal investments. We then respond to several criticisms of targeted preschool programs. We conclude with a discussion of the key features for successfully investing in targeted ECE
programs and advocate for a market-based approach.
ECONOMIC RESEARCH FAVORS TARGETED APPROACHES TO EARLY CHILDHOOD EDUCATION
The high return to early childhood investments targeted to at-risk children is well researched. Analyses of the Perry preschool program (Schweinhart et al., 2005), the Abecedarian project (Masse & Barnett, 2002), the Chicago Child-Parent Centers (Reynolds, Temple, Robertson, & Mann, 2002), and the Elmira Prenatal/Early Infancy Project (Karoly et al., 1998) show annual rates of return, adjusted for in?ation, ranging from 7% to just over 20% and benefit–cost ratios ranging from 4:1 to more than
10:1 (Heckman, Grunewald, & Reynolds, 2006). Researchers followed effects of these ECE programs
through adolescence and well into adulthood (Heckman et al., 2006).
These four longitudinal studies form the primary cost–benefit evidence in support of ECE investments targeted to at-risk children (i.e., from low-income households) and their families. Research also
suggests that early childhood investments targeted to at-risk children should start well before preschool.
Neuroscience shows that the ?rst few years of life are crucial to healthy brain development. Programs
that begin early are particularly important to children exposed to toxic stress (National Scientific
Council on the Developing Child, 2009).
RETURNS TO TARGETED PROGRAMS ARE HIGHER THAN UNIVERSAL
In this section, we show that returns to targeted preschool programs are higher than universal programs
through a critique of the case for universal preschool. Proponents for universal preschool claim that
the rate of return to middle-income children on a per-child basis is substantial—even though it is
smaller than the return to low-income children. They demonstrate the rate of return to middle-income
children in part by extrapolating from the longitudinal studies focused on low-income children. In
addition, proponents note that the total return to universal preschool is much larger than the total
return to targeted preschool because there are a large number of middle-income children relative to
In our response to the case for universal preschool, we begin by discussing three shortcomings
in universal proponents' claim that per-child returns to universal preschool are substantial. First,
extrapolating evidence from the longitudinal targeted studies to middle-income children is subject to a
fair degree of uncertainty. Second, benefits to low-income children from attending universal preschool
tend to drive the universal rate of return. Third, universal preschool supplants funds middle- and high income families would spend on preschool regardless of the availability of universal preschool.
In addition to these shortcomings, we argue that the per-child rate of return is a better criterion
for making funding decisions than the total rate of return. Moreover, we contend that the higher overall cost of universal preschool relative to targeted preschool could crowd out investment in potentially
higher return investments.
A targeted approach achieves a higher rate of return than a universal approach because low income children begin at a lower baseline than children from higher-income families. (Children from
higher-income families are more likely to start preschool closer to the developmental mean.) Hart and
Risley (1995) observed that by the age of 3 years, children who grew up in homes with parents on
welfare had only half the number of vocabulary words as children who grew up in homes with college educated parents. The gap observed by the researchers translates to high costs to society. We argue that
a preschool investment in the former child produces a high public return, whereas a preschool investment in the latter produces a modest public return. The longitudinal studies cited previously show that
high-quality targeted ECE programs can reduce costs for special education, grade retention, and the
criminal justice system and can increase tax revenue (Heckman et al., 2006).
Universal proponents point to studies of universal preschool programs to make the case that
children from all income levels benefit from attending. For example, a study of Oklahoma's universal
preschool program in Tulsa showed that children from higher income families posted test score gains.
However, children who quali?ed for free lunch and reduced-price lunch had higher test score gains in
letter-word identi?cation, spelling, and applied problems than children who paid full price for lunch
Some researchers have used the cost–benefit study results of the longitudinal studies to help estimate economic returns to universal programs. Examples of applying the results from studies of targeted
preschool programs to universal preschool are found in Bel?eld (2004), Karoly and Bigelow (2005),
and Lynch (2007). However, attempts to adjust the economic impact for middle-income and high income children are subject to a fair degree of uncertainty. For example, the Karoly and Bigelow study
of California universal preschool reported that preschool benefit–cost estimates range from roughly 2:1 to more than 4:1. Differences between the estimates are based on different assumptions regarding how
much middle-risk and low-risk children benefit from universal preschool relative to high-risk children.
If the per-child rate of return of middle-risk and low-risk children is relatively modest, the participation of high-risk children in universal programs would primarily drive the economic return. In
the Karoly and Bigelow study (2005), the most conservative estimate of 2:1 assumed that all benefits
to the universal program accrue to high-risk children, who comprise 25% of California's 4-year-old
population. The baseline benefit–cost ratio reported in the study was 2.6:1, in which middle-risk
children and low-risk children received 50% and 25% of benefits, respectively, relative to high-risk
children. (In the simulation, these impacts referred to children who would not have attended preschool without the implementation of the universal program.) The baseline benefit–cost ratio is
driven largely by benefits to high-risk children. Furthermore, if the California preschool program
was limited to the high-risk children and not offered to all children, overall costs would drop by 75%
(the proportion of middle- and low-risk children in the simulation), thus increasing the benefit–cost
ratio to closer to 8:1.
Although Karoly and Bigelow (2005) accounted for differences in preschool attendance by risk
level, their return to middle-income children may be overstated when existing parental investments are
fully taken into account. Offering free preschool to middle- and high-income families only marginally
increases the number of these children who enroll in preschool because many are already enrolled in
a tuition-based program. In 2005, approximately 47% of children ages 3– living below the poverty
line were enrolled in a center-based preschool nationally compared with 60% of children ages 3–5
living above the poverty line (National Center for Education Statistics, 2007). This disparity exists
despite the availability of Head Start, which is free for children living below the poverty line as de?ned
by the Of?ce of Management and Budget (Improving Head Start for School Readiness Act of 2007,
PL 110-134). In 2011, the poverty line for a family of four was $22,350 (U.S. Department of Health
and Human Services, 2011). Parents with higher levels of education, which is consistent with higher
levels of income, spend more resources on their children for child care than parents with lower levels
of education (Rosenbaum & Ruhm, 2007). For example, caregivers with education attainment levels
beyond high school spend twice as much per hour on child care as caregivers who did not graduate
high school. In addition, children in higher income families on average have a more enriched home
environment than children in low-income families. All of this evidence demonstrates that universal
preschool tends to supplant resources that higher income families already spend on supporting their
children's educational experience.
Universal proponents also point out that because there are a large number of middle-income
children relative to low-income children, the total net return to a universal program is substantially
higher than targeted programs (Barnett, Brown, & Shore, 2004). However, total net return is not
the best criterion for deciding whether to invest; a better criterion is the per-child rate of return.
The question should be, how should the next dollar be invested in ECE? The correct answer comes
from looking at marginal net benefits, not the total. On the margin, the next dollar spent on at-risk
children will produce a bigger bang for the buck than if the next dollar were spent on middle-income
However, even when a universal program demonstrates positive marginal net benefits, it is not a
suf?cient condition for funding. The rate of return or benefit–cost ratio must be compared with other
potential investments, including investments to develop human capital, particularly prenatal-to-age-3
programs for at-risk children. For example, the Nurse Family Partnership, which is modeled on the
nurse-based home visiting model studied in Elmira, New York, has demonstrated strong public rates of
return (Heckman et al., 2006). Other human capital investments may also prove to have higher rates
of return than universal preschool, such as after-school programs, teacher compensation reform, classroom size, and summer transition programs. Universal proponents must show that returns to universal
preschool trump these other investments.
To illustrate this issue, consider Georgia's universal preschool program and Michigan's targeted
program. The overall cost of universal preschool is generally more expensive than a targeted program
simply because it encompasses more children. For example, Georgia's universal preschool program
reaches 78,310 four-year-olds at an annual cost of $332 million. In contrast, Michigan's preschool program is targeted to 4-year-old children who have educational disadvantages; it reaches 24,091 children
for an annual cost of $103 million (Barnett, Epstein, et al., 2009). Compared with Georgia, Michigan
has more funds to spend on potentially higher return projects, such as prenatal-to-age-3 programs for
RESPONSE TO THE CRITICS OF A TARGETED APPROACH
In this section, we respond to several criticisms of targeted preschool programs.
Targeted Approach Will Not Garner Political Support
Proponents for universal preschool claim that programs with broad eligibility build a stronger base
of political support than targeted preschool. They point to the lack of political backing for programs
that support low-income families, including Head Start. However, Americans generally accept and
support targeting college scholarships and other ?nancial aid to students from low-income families.
The public understands that targeting college ?nancial aid helps support the country's work force and
democracy. Similarly, we argue th at the public will back targeted preschool as a pillar of human capital
Business leaders recognize the high rate of return to targeting at-risk children. For example, the
Minnesota Early Learning Foundation (2009), with a board of directors that includes key state business leaders, funds research to learn what works best and is cost effective to promote learning readiness
among at-risk children. Furthermore, universal preschool is not a silver bullet to gain political support.
The defeat of California's Proposition 82 in 2006 was an indication that voters are not necessarily
swayed by arguments for universal preschool (Institute of Government Studies, n.d.).
Targeted Approach Precludes Integrated Classrooms
Targeted approaches can provide integrated programs where children from diverse socioeconomic
backgrounds attend. A targeted program can be integrated into the existing system of public and private preschools and child care programs. For example, a scholarship program gives low-income families
the opportunity to enroll their children free or at a reduced rate with children from higher income
families who pay up to full price.
Targeted Approach Requires Means Testing, Which Is Dif?cult to Administer
Although there is a cost to means testing preschool applicants, the application process can be simpli?ed to make it easier for families to navigate. Quali?cation for the preschool program can be linked to
eligibility for other means-tested programs, such as temporary ?nancial assistance for families or child
care subsidies. We argue that the cost of means-testing families is small relative to the benefits achieved
from a targeted program.
Targeted Approach Creates Stigma
Universal advocates claim that targeted preschool stigmatizes low-income families who qualify for the
program. Stigma is affected by how targeted programs are presented to the public and eligible families. Stigma is not generally associated with scholarships and ?nancial aid for lower income children
to attend college. Instead, college scholarships and ?nancial aid are viewed as a key opportunity both
by the public and eligible students. Therefore, targeted preschool should be portrayed primarily as an
opportunity to invest in young children, not as a support to low-income families.
Targeted Approach Is Dif?cult to Bring to Scale and Reach All Eligible Children
Finally, we agree with universal proponents that reaching low-income families is a challenge, but we
disagree that means testing is the primary deterrent. A number of low-income families face several challenges, including housing and employment. Reaching these families takes focused resources, including exposure at a number of locations and outreach workers on the ground, regardless of whether the
program is targeted or universal. We argue that a targeted program frees up resources for intensive
outreach compared with an overall more expensive universal program.
HOW TO TARGET RESOURCES FOR PRESCHOOL: A MARKET-BASED APPROACH
To achieve the high rate of return discussed previously, we argue that preschool must not only focus
on at-risk children, but also emphasize parental involvement, start early, ensure high quality, and be
scalable. To achieve these characteristics, a targeted preschool program would benefit from the ?exibility, innovation, and incentives that are inherent in markets. To that end, we propose creating an ECE
endowment that would provide parent mentors and scholarships to families with at-risk children to
attend a high-quality preschool program (Grunewald & Rolnick, 2006).
Parental involvement is a crucial ingredient in the success of ECE programs. When parents receive
training in why and how to nurture their children, they are better able to support their children's development at home. Reaching children as early as possible is potentially most effective, as neuroscience
research shows that the ?rst few years are critical to brain development. As with universal preschool
models, targeted programs must ensure high quality. In addition, both models must be able to bring
their offerings to scale. Head Start and state child care subsidy waiting lists show a large number of
low-income children who could benefit from preschool.
A targeted market-based preschool system delivers these characteristics and addresses the concerns
raised by proponents of universal preschool. Investing in universal preschool can de?ect funding from
providing key parental outreach and education, such as home-visiting parent mentors, and from reaching children early. The model we propose begins with parent mentors with pregnant mothers; their
children become eligible for scholarships when they turn age 3. We contend that a system that provides
ECE from prenatal to kindergarten and includes a parent focus is more feasible and provides a higher
rate of return than universal preschool.
A targeted market-based approach also frees up resources to recruit low-income families into the
program, and by providing scholarships there is little stigma to participating in the program. In fact,
receiving a scholarship can be empowering to the family. The administration and means testing of
eligible children can be conducted ef?ciently, as discussed previously.
A pilot project of a parent mentor and scholarship model was implemented in St. Paul, Minnesota, with funding by the Minnesota Early Learning Foundation. Starting in January 2008, scholarships were provided to low-income families in two St. Paul neighborhoods with a high concentration
of families living below or just above the poverty line. Early observations by evaluators suggested
that participating parents were "enthusiastic about the scholarship program and reported how it benefited both their children and themselves" (Gaylor, Spiker, & Hebbeler, 2009, p. 21). Project data also
suggested that intensive outreach efforts have been more successful in recruiting families than passive approaches. The model was designed to be scalable and applied in other communities and states
around the country.
Preschool initiatives that provide resources to parents to enroll their children in preschool programs of their choice were also implemented in Denver, Colorado (http://www.dpp.org/), and Sioux
Falls, South Dakota (http://seuw.org/StartingStrong.aspx). Data from these initiatives help to determine the effectiveness of a market-based approach to investing in preschool. Just like scholarships for
higher education, scholarships for preschool can improve access to high-quality public and private
programs for low-income children.
Given that we have limited public resources, these resources should be allocated to initiatives with the
highest rate of return. Hence, funding programs targeted to at-risk families should have priority over
universal preschool programs. Furthermore, employing a market-based approach that targets resources
to at-risk families in the form of parent mentors and scholarships to high-quality programs could readily achieve scalability and ensure long-term, high-return outcomes.