What are the costs of a full-scale commitment to early childhood education?

In my new book, From Preschool to Prosperity, I outline a specific proposal, backed by research, for a full-scale early childhood program. This program consists of three components: (1) universal full-day pre-K for all 4-year olds; (2) income-targeted full-time, full-year child care and pre-K from birth to age 5 for all disadvantaged children, similar to today’s Educare program; (3) income-targeted Nurse Family Partnership home-visiting/parenting services for all first-time disadvantaged mothers.

After allowing for pre-K and child care services we are already providing through the federal government and state governments, the net annual cost of each of these components would be: (1) universal pre-K, $25 billion per year; (2) targeted child-care, birth to age 5, $58 billion per year; (3) NFP, $4 billion per year. Although these numbers add up to $87 billion, there is some overlap if all three components were funded, for example the pre-K at age 4 for disadvantaged children is part of both component (1) and component (2). Adjusting for this overlap, the net cost of the entire package would be $79 billion.

$79 billion is obviously a lot of money. But the amount represents only 2 percent of total federal, state and local taxes. Alternatively, this amount represents about 13 percent of what is currently spent on public K-12 education. Therefore, this $79 billion in annual spending represents a considerable but not outlandish increase in our investment in education, at a tax cost that is a modest increase in the overall tax burden.

The costs of this $79 billion in annual investment would be considerably outweighed by the benefits. This proposal would annually provide pre-K to 3 million 4-year-olds, and child care services to an additional 3 million children ages 0-3, along with NFP services to about 1 million families. The future earnings benefits for former child participants would have a present value of between 2 and 3 times the costs of this investment, so the package would be worthwhile for these direct benefits alone. Earnings benefits for parents from the free child care and the home-visiting services would by itself have a present value of about one-and-a-half times this investment, so the package would even be worthwhile if there were no benefits for child participants. Finally, the evidence suggests that the spillover benefits of the greater skills of former child participants and their parents on the wages of other workers would more than double overall benefits, as argued in a previous blog post. Overall, the package’s benefits should exceed 10 times its annual costs, simply in terms of increased earnings for the economy. And this does not count any benefits from reduced crime, which would be quite likely to be large.

Early childhood education at full-scale is not cheap, but is affordable. Among policy alternatives for boosting the long-run growth of the economy, early childhood education has some of the best evidence of high benefit-cost ratios that more than justify full-scale investment.

Posted in Early childhood program design issues

Fiscal benefits: pre-K pays for itself in the long-run

As I discuss in my new book, From Preschool to Prosperity, pre-K and other early childhood programs provide important “fiscal benefits”. By “fiscal benefits”, I mean increases in tax revenue or reductions in needed spending, even at the same tax rates and with the same spending policies. These fiscal benefits allow governments to lower tax rates without cutting public services, or to increase public services without increasing tax rates, either of which will benefit taxpayers.

Fiscal benefits of early childhood programs occur through several mechanisms. In the short run, the most important fiscal benefit is the reduction in special education costs. Special education is very expensive, costing as much as $10,000 per year per child. High-quality pre-K leads to significant reductions in special ed assignments. However, these reductions vary greatly from study to study, ranging from 23 to 86 percent. Although all such reductions could yield significant cost savings, especially given that special ed assignments tend to persist over most of a child’s K-12 career, the wide range means there is some uncertainty about the exact magnitude of these shorter-term cost savings. (See ReadyNation for more exploration for the potential for special ed cost-savings for financing pre-K programs.)

In the longer run, a wider variety of fiscal benefits become important. These fiscal benefits include: savings in prison costs and criminal justice system costs due to the effects of early childhood programs in reducing crime; increases in tax revenue at the same tax rates due to higher earnings; reductions in welfare costs due to higher earnings. These longer-run fiscal benefits also vary greatly from study to study. For example, in the Abecedarian study of child care and pre-K from birth to age 5, there were no significant reductions in crime. This may reflect that the control group in this study also had lower crime rates. Therefore, the fiscal benefits of early childhood programs may be greater if the programs are targeted at groups that are likely to have more substantial problems, such as involvement with the criminal justice system.

A variety of simulation studies find that in the long-run, the fiscal benefits of pre-K are sufficient for such programs to pay for themselves. However, the studies vary in how long it takes for pre-K programs to become self-financing. Lynch’s work suggests that pre-K breaks even fiscally about 9 years after the program is begun. In contrast, Dickens and Baschnagel conclude that pre-K breaks even fiscally about 49 years after the program is begun. These different break-even periods reflect different assumptions about program costs and program effects.

The challenge for public policy is that policymakers make decisions in the short-run and often take a short-run perspective. Early childhood programs require short-run investments, and aren’t self-financing immediately. However, if policymakers can be encouraged to take a longer-run perspective, early childhood programs can help improve the long-run budget position of all levels of government, federal, state and local.

Government’s goal is not to make money. However, government policymakers and the public should be concerned about fiscal commitments that might yield escalating costs over time. In the case of early childhood programs, over time the fiscal resources needed to finance these investments will diminish, and eventually the programs will help either reduce taxes or finance other public services. In the meantime, these programs will provide important benefits, both to participating children and families, and to the national economy.

Posted in Distribution of benefits, Timing of benefits

Anti-crime benefits of pre-K are another good rationale for broad public support for pre-K

Another important spillover benefit of high-quality pre-K is the effect of these programs on lowering subsequent crime rates of participants. As discussed in my new book, From Preschool to Prosperity, these anti-crime effects provide broad public benefits by reducing the odds of becoming a crime victim, and by lowering prison costs.

Benefit-cost studies of both the Perry Preschool Program and the Chicago Child-Parent Center pre-K program have found significant anti-crime benefits. When economists put a dollar value on these anti-crime benefits (that is, when they add up the reduced criminal justice system costs plus how much people are willing to pay to reduce their odds of being victimized by crime), the results frequently show that anti-crime benefits are of similar order of magnitude to the future earnings benefits of pre-K.

Although I think anti-crime benefits are important, and appeal to some audiences, I think it is important to also emphasize positive benefits such as the broad increase in wages and prosperity due to broad access to quality pre-K. A focus on anti-crime benefits always runs the risk among some voters of stigmatizing the disadvantaged, which may reduce voter sympathy and support for educational and other programs to help the disadvantaged. This emotional reaction may not make rational sense, but there is a natural human reaction among some people to respond to threats of violence with solutions that rely on retribution and punishment.

However, from a rational analytic perspective, high-quality pre-K can be a cost-effective way of reducing the costs of crime. These anti-crime benefits of pre-K are why lobby groups such as “Fight Crime: Invest in Kids” have found broad support among police chiefs and sheriffs for crime prevention via early childhood education.

Posted in Uncategorized

Peer effects in K-12 education are important spillover benefits of early childhood education

One reason why all families benefit from publicly-funded preschool, including families who don’t enroll their children in public preschool, is the increased educational achievement due to peer effects in K-12 schools.  As discussed in my new book, From Preschool to Prosperity, these peer effects provide an important “spillover” rationale for why preschool might attract broad voter support.

Empirical evidence finds that everyone in a kindergarten class benefits if a higher percentage of the kindergarten class enters with better skills from high-quality pre-K. The effects on an entire class’s end-of-kindergarten test scores, due to having one additional child who enrolled in high-quality pre-K, are 16% to 50% greater than what would be expected from the direct effects on that individual child’s achievement. Other studies find similar peer effects in subsequent grades, with achievement gains of children during a given grade being higher if the entering skills of their peers in the same class are higher.

Peer effects make intuitive sense. If my child’s peers have better cognitive skills, the teacher doesn’t need to spend as much time remediating these peers’ skills, and can spend more time enhancing my child’s skills. If my child’s peers have better social skills and behavior, less class time is lost to classroom disruption, and my child will learn more. In addition, my child may directly learn better cognitive and social skills from his or her peers.

Peer effects provide an argument from enlightened self-interest for why all families with children in public school should support government efforts to broaden access to quality pre-K. Broader access to pre-K will improve overall achievement gains in the K-12 system.

Posted in Distribution of benefits

Spillover benefits of early childhood education may more than double economic benefits

In my new book, From Preschool to Prosperity, I devote an entire chapter to discussing “spillover benefits” of early childhood education, that is benefits that spillover from the families participating in early childhood programs to other people. Evidence suggests that spillover benefits are large enough to more than double economic benefits.

Spillover benefits are important because they provide an argument, from the perspective of enlightened self-interest, for why all voters should support early childhood education. Some voters do not have children. Some voters with children may believe that their families do not need any government assistance with early children education. Why should such voters be willing to pay taxes to provide early childhood education for “other people’s children”? Compassion is one rationale, but enlightened self-interest provides a complementary rationale: early childhood education can be argued to provide economic benefits for all voters, including those who do not have children in these programs.

One important spillover benefit of early childhood education is the effects of broader development of skills on overall business competitiveness and hence worker wages. Business productivity and competitiveness depends on overall worker skills, not just one individual worker’s skills. I can be the most skilled worker in the world, but my employer will find it more difficult to introduce new technologies unless my co-workers are also skilled. Greater skills by my co-workers will thus enable my employer to be more competitive and increase my wages. In addition, my employer’s competitiveness and ability to pay high wages may also depend on the skills of the suppliers to my employer. Finally, my employer’s competitiveness, in high tech clusters such as Silicon Valley, may depend on the flow of ideas and skilled workers across competing firms in related industries, and therefore greater skills in these clusters will increase overall cluster wages.

These skill spillovers explain the empirical finding that an individual’s wages depend not only on his or her own educational level, but also on the average education level of all other workers in the same metropolitan level. These empirical estimates of wage determinants find that the spillover effects on others’ wages of an individual getting more education actually are greater, summed over all workers, than the direct effects on the individual’s own wages.

Therefore, I benefit, and my children benefit, if “other people’s children” get more skills, through early childhood education or other policies. This goes against the intuitive notion that some people have that there is a fixed number of good jobs that pit my children’s interests against the interests of other children. This intuition is wrong because when we increase everyone’s skills, the number of good jobs will expand in response, and will expand enough that everyone’s opportunities are enhanced.

We’re all in this economy together. If policies broadly raise skills, this will increase overall economic productivity and prosperity. Early childhood education is one research-proven strategy for broadly increasing skills in a cost-effective manner.

Posted in Economic development

My article with Elaine Weiss in Huffington Post on pre-K as part of pro-children strategy

In a recent article in Huffington Post, Elaine Weiss and I discuss pre-K’s potential as part of a pro-children strategy. Elaine Weiss is National Coordinator of the Broader, Bolder Approach to Education at the Economic Policy Institute.

The article describes some of the evidence from my new book showing that pre-K works, and in particular highlights the research evidence from Boston. The article argues that pre-K can overcome fade-out of effects as children age, and that this works particularly well when pre-K is accompanied by smart investments in K-12 education. Massachusetts again provides a good example of such smart investments.

As both my book and the article argue, pre-K and other early childhood education investments can be even more effective as part of more comprehensive strategies. Early childhood education investments are not the only investments in skills that can work. And policies to invest in skills should be accompanied by policies to create more demand for high-skill, high-wage jobs.

Posted in Early childhood program design issues, Economic development

My article for Bridge magazine on Michigan’s pre-K program published September 30

My article for Bridge Magazine on Michigan’s pre-K program was published on September 30.  Bridge Magazine is a publication of the Center for Michigan, a non-partisan organization that examines policy issues connected with Michigan’s future.

To sum up the article: Michigan has significantly improved pre-K access in the last two years. But Michigan’s pre-K program may need additional funding, intelligently spent, to maintain its past record of getting high-quality results.

Posted in Early childhood program design issues

The pre-K controversy over teacher credentials: mixed evidence, but successful programs tend to use certified teachers paid public school wages

My new book, From Preschool to Prosperity, includes a discussion of the controversy over whether pre-K programs get better results if they use credentialed teachers. The controversy arises because the research on this topic is mixed.

However, even though the research is mixed, the pre-K programs with the best evidence of success tend to use credentialed teachers paid public school wages. This is true for Perry Preschool, the Chicago Child-Parent Center program, and the universal pre-K programs in Boston and Tulsa.

Obviously more research on the impact of different teacher credential requirements is warranted. But if one wants greater assurance that a pre-K program will be effective, it seems prudent to follow the example of programs with demonstrated success, and seek to hire credentialed teachers and pay wages close to the public school standard.

It seems likely that paying higher wages will reduce teacher turnover. Most studies of teaching suggest that new teachers are less effective than teachers with some experience, so higher teacher turnover would be expected to lower program quality.

It is possible that the mixed research on teacher credentials is related to the interaction of credential requirements with salary standards. Imposing credential requirements, but then paying salaries below public school levels, may increase teacher turnover by restricting the pool of eligible teacher candidates to those who have the alternative of teaching in the K-12 system, but then underpaying them relative to their alternative opportunities.

In addition, the impact of teacher credentials and salaries may depend on the size of the pre-K program. A small program may be able to find some dedicated staff who will work at the program despite the low pay. At a large scale, this strategy is harder to carry out.

Given the importance of quality in pre-K, a wise strategy is to err on the side of overinvesting in quality, until further research clarifies what program features are essential for quality.

Posted in Uncategorized

Pre-K quality: even slight improvements have large economic benefits, which justify even large costs to improve quality

One argument made in my new book, From Preschool to Prosperity, is that even slight improvements in preschool quality can be shown by the economic logic of educational investments to have large benefits. This justifies spending quite a bit of money to improve quality even slightly.

Over a career, the average American has large earnings. Even for low-income families, the average child can expect to have future earnings, over their entire lifetime, whose “present value”, as of age 4, sums to over one-half million dollars. (“Present value” adjusts future dollars down in value to reflect that dollars today are more valuable because they could be invested and earn interest.)

Suppose there is some quality improvement in a pre-K classroom. Even if this quality improvement only improves average future earnings by a very small percent, when we sum this quality improvement over an entire career, and over all the children in the classroom, we get a very large economic benefit. If the investments in training or teacher selection or curriculum that lead to this quality improvement lead to quality improvements for more than one year, the sum of benefits over multiple subsequent years will be even greater.

For example, suppose some quality improvement in pre-K classrooms increases average future earnings of children in that classroom by only one-half-of-one percent. (In the real world, this “average effect” might in many cases consist of helping some children, but not all children, by much more.) The benefits per child of this improvement will be about $2,500 in increased present value of earnings (1/2 of 1% times one-half million dollars in expected future earnings). If there are 15 children in the classroom, the total benefit for the entire class is $37,500 ($2,500 times 15). If the investment in quality improvement has benefits that last for three years, then total benefits of the quality improvement will exceed $100,000 (adding the $37,500 over three years, with some discounting for the last two years). Spending up to $100,000 to achieve such a quality improvement would be worth it from an economic benefit to cost ratio. This could pay for quite a bit of curriculum improvement, teacher training, higher teacher salaries, or teacher hiring incentives, if these measures actually improved classroom quality.

Based on studies of the relationship between kindergarten entrance test scores and future earnings, achieving a future earnings improvement of one-half of one percent would only require average kindergarten entrance test scores to improve by about 0.7 percentiles (for example, moving a student from the median to the 50.7th percentile, with their test scores greater than 50.7% of all entering kindergartners, or from the 30th percentile to the 30.7th percentile). It seems plausible that investments in pre-K quality improvements might be able to achieve such test score improvements, or even greater improvements.

We don’t know as much as we would like about how pre-K quality is affected by different investments. But we do know that even slight increases in quality have large effects. This justifies great effort to discover what quality investments work. And it also poses a challenge for researchers: even slight improvements that are hard to detect empirically may have benefits greater than costs.

From an economic perspective, quality in pre-K matters a lot.

Posted in Uncategorized

Pre-K benefits both the poor and the middle class, but child care and parenting programs’ benefits are more targeted

In my new book, From Preschool to Prosperity, I review the research evidence on whether the benefits of early childhood programs go beyond children from low-income families to include middle-class children. This research evidence suggests that preschool at age 4 has similar-sized benefits for low-income and middle-class children, but that child care and parenting programs seem to have benefits only for children from low-income families.

The research evidence on middle-class benefits of preschool is sparse, but the best evidence suggests that pre-K causes similar-sized increases in test score percentiles at kindergarten entrance for middle-class and low-income children. In Tulsa, the pre-K impact for middle-class children in increased test score percentiles is about 90% of the impact for lower-income children. In Boston, the pre-K impact for middle-class children in increased test score percentiles is about 70% of the impact for lower-income children.

The available evidence on how early test scores are related to adult earnings suggests that these test score percentile effects predict similar ratios of dollar effects on adult earnings. Thus, the Tulsa results predict that the lifetime dollar increase in adult earnings for middle-class children will be about 90% of the increase for low-income children.

However, even though the dollar earnings effects are similar, universal pre-K would still have significant income redistribution effects. Children from low-income families would be expected to have lower future earnings than children from middle-class families. Therefore, even though the dollar earnings effect is similar, the percentage boost to the future standard of living will be much higher for children from low-income families.

For child care and parenting programs, the evidence on how effects differ for children from different income groups is also sparse. For parenting programs, several studies of the Nurse Family Partnership program suggest that benefits are restricted to low-income families. For child care programs, research on the Infant Health and Development Program also suggest that the benefits of high-quality child care are restricted to low-income children.

Why is the available research evidence sparse? In part, because there are simply more research funds available to study the effects of early childhood programs on low-income families. This research funding situation is likely to persist. So policymakers trying to determine the design of early childhood programs will have to make do with sparse information on the universality of these programs’ benefits.

Why might early childhood programs have this pattern of effects, with benefits for middle-class children more apparent for pre-K than for child care and parenting programs? Perhaps the benefits for pre-K in developing social skills are hard for middle-class parents to duplicate on their own, especially given the high price of high-quality private pre-K. In contrast, perhaps parenting and child care is somewhat easier for many middle-class parents to provide on their own.

If the goal of early childhood program design is to maximize the benefits for improving children’s future life course, this research suggests that there is support for universal pre-K, and support for income-targeted child care and parenting programs. One important caveat: this only counts benefits for children. Child care programs may have important benefits for parents that may be more broadly shared across income groups, and which also deserve consideration by policymakers.

Universality of pre-K programs can be backed by good evidence of broad benefits for diverse groups of child participants. In contrast, universality of child care programs would need to be justified by important benefits for diverse groups of parents.

Posted in Distribution of benefits