Why earnings benefits of pre-K can be so large for the middle-class

Posted on August 22, 2011 by timbartik

My recent working paper on Tulsa’s pre-K program predicts that pre-K will increase the future earnings of both the poor and the middle class by similar dollar amounts.  (This paper, co-authored with Bill Gormley and Shirley Adelstein of Georgetown, can be found at the Upjohn Institute website or at the website of the Georgetown Center for Research on Children in the U.S.)

Why are predicted earnings effects for the middle-class so high? Our intuition might suggest that pre-K will have larger effects for children from low-income families.

The working paper does not explain the result, but there are some good reasons why pre-K might have large effects on children from middle-class families. First, even for middle-class families, a good pre-K program might enhance so-called “soft skills” or social skills. Many of these soft skills are important for future labor market success. Such soft skills might be difficult to enhance in early childhood through good parenting alone, without some social skills development in group settings such as a pre-K program.

Second, given the nature of the income distribution, even modest increases in educational attainment may have large effects for children from middle-class families. For example, as I explored in my book “Investing in Kids”, the dollar increase in earnings from attaining a college degree (versus a high school degree) is over twice the dollar increase in earnings from attaining a high school degree (versus being a high school dropout).  (See p. 229 of my book.) Therefore, even if pre-K affects a fewer percentage of middle class kids (compared to low-income kids) by increasing their educational attainment, it may have large dollar effects if it has larger effects on college degrees for the middle class group.

From an economic development standpoint, both investments in kids from low-income families, and investments in kids from middle-class families, can pay off.

Posted in Distribution of benefits, Early childhood programs | Comments Off on Why earnings benefits of pre-K can be so large for the middle-class

Presentation to American Chamber of Commerce Executives

Posted on August 18, 2011 by timbartik

My August  4th speech to the American Chamber of Commerce Executives’ convention, and the accompanying powerpoint, have now been posted at the Upjohn Institute’s website.

This speech integrates much of the material from this blog, and my book “Investing in Kids”, into a relatively short (20 minutes, 5 pages, 11 slides) presentation.  In particular, it incorporates my recent research on the benefits of preschool for children from middle-class families, and the spillover benefits of education for everyone in local economies.

Posted in Early childhood programs, Economic development | Comments Off on Presentation to American Chamber of Commerce Executives

Even slight increases in kindergarten readiness yield benefits that exceed costs for pre-K

Posted on August 16, 2011 by timbartik

One implication of my recent paper on Tulsa’s pre-K program is that even modest increases in kindergarten readiness would be predicted to have large future earnings benefits for former participants in pre-K programs.  (This paper, co-authored with Bill Gormley and Shirley Adelstein of Georgetown, can be found at the Upjohn Institute website or at the website of the Georgetown Center for Research on Children in the U.S.)

Using research by Chetty et al, we conclude that even a one percentile increase in a child’s test scores at kindergarten entrance is associated with an increase in the present value of the child’s future earnings of around $1500. A one percentile increase in test scores is a slight increase.  For example, such an increase would move a child from the median of test scores – half the children below in test scores, half the children above – to the 51st percentile, with test scores that exceed 51% of all children entering kindergarten.

A quality half-day pre-K program for one school year might cost around $4500 (see the research on the cost of preschool quality by the Institute for Women’s Policy Research). Therefore, such a half-day program will have future earnings benefits exceeding costs even if their effect is only to raise kindergarten readiness by 3 percentiles. Other benefits of pre-K programs, such as lower special education costs or lower crime rates, would imply that pre-K programs could have benefits exceeding costs with even lower effects on test scores of kindergarten entrants.

Furthermore, even slight increases in pre-K quality are likely to have benefits exceeding costs. If we can improve the pre-K curriculum or teacher quality sufficiently to increase average test score effects by even one percentile, the overall benefits for an entire pre-K class are enormous. For example, a 1 percentile increase in average test scores for a pre-K class of 15 children is estimated to increase the present value of future earnings benefits by $22,500 (=$1500 times 15 students).  This benefit would be sufficient to justify considerable spending on improving teacher or curriculum quality, even if that spending only has modest effects in improving quality.

The broader point is that early learning improvements have large long-run benefits. This point has previously been made in the debate over K-12 teacher quality by Chetty et al, and in the debate over K-12 class size reductions by Krueger.

Because early learning improvements have such large long-run benefits, even sizable investments in improving early learning can make good economic sense.

Posted in Early childhood program design issues, Early childhood programs, Economic development | Comments Off on Even slight increases in kindergarten readiness yield benefits that exceed costs for pre-K

Talk tomorrow at American Chamber of Commerce Executives convention

Posted on August 3, 2011 by timbartik

I will be speaking tomorrow at the annual convention of the American Chamber of Commerce Executives. As the organization’s name suggests, ACCE includes executives from over 1400 chambers of commerce in the U.S. and Canada.  I will be speaking on the business case for early childhood programs, in a session that includes Sara Watson of Pew and David Rattray, Vice President for Education and Workforce Development of the L.A. Chamber of Commerce. After the convention, I will be out of town for a few days. I may not resume regular posting until August 22nd.

Posted in Uncategorized | Comments Off on Talk tomorrow at American Chamber of Commerce Executives convention

Does preschool benefit middle-class kids? Yes

Posted on August 1, 2011 by timbartik

A just-released working paper by me, along with my co-authors William Gormley and Shirley Adelstein at Georgetown, considers, among other things, how Tulsa’s pre-K program affects middle-class children.  For middle-class children, we project that both the half-day and full-day versions of Tulsa’s pre-K programs have large future benefits in increased earnings, with these benefits exceeding costs. (We also find benefits exceeding costs for children from low-income families, but this is a more expected finding, as it is consistent with extensive previous research.)

The paper includes estimates of how Tulsa’s pre-K program affects test scores of students who are NOT eligible for a free or reduced-price lunch. We estimate that the average earnings of parents of these students are 135% of average earnings for the Tulsa metropolitan area. Therefore, this is by most definitions a “middle class” group of students.

We estimate that Tulsa’s half-day pre-K program, conducted for one school year at age 4, will increase the test scores of these middle-class students by 11 percentiles. This increase is over and above the increase that would occur without the program during that year, due to normal learning due to aging or participation in other programs.

What does an 11 percentile improvement mean? Consider a student who without Tulsa’s pre-K program would have scored at the 40th percentile on some test at kindergarten entrance. This 40th percentile performance means that the student’s test score level would have put them at a higher level than 40 percent of all entering kindergartners, and below 60 percent of all entering kindergartners.  An 11 percentile boost means that the student would instead score at the 51 percentile level – above 51 percent of all entering kindergartners. I think most parents would consider this a significant improvement.

We also find that a full-day pre-K program improves test scores of these middle-class students by 16 percentiles.  While doubling the pre-K day does not double student learning, it does bring about a significant increase.

Using estimates from Chetty et al, we project that these test score boosts will lead to considerable increases in future earnings. The 11 percentile boost is predicted to increase the present value of future adult earnings by over $16,000. The 16 percentile boost is predicted to increase the present value of future earnings by over $24,000.

As explained in the paper, these projected earnings increases assume that the increased skills will persist into adulthood to an extent similar to what is observed in Chetty et al’s study.  Although this projection requires this assumption, we present some evidence that this assumption is roughly accurate for previous high-quality preschool programs, such as the Perry Preschool program and the Chicago Child-Parent Center program.

These future earnings boosts have a present value that is many times these programs’ costs. The ratio of earnings benefits to costs for middle-class kids for the half-day program is 3.76. (The ratio for low-income kids is somewhat higher, at 4.01) The ratio of earnings benefits to costs for middle-class kids for the full-day program is 2.79. (For low-income kids, the full-day program has an earnings benefits to cost ratio of 3.09.) These “benefit-cost ratios” only consider one possible benefit of these programs, the effect on boosting future earnings. Benefits would be higher if we considered other benefits such as reduced crime or reduced need for special education costs.

The implication is that high-quality pre-K programs can produce net economic benefits by expanding their scope to include middle-class as well as low-income students.  Many middle-class parents will find it difficult to provide high-quality preschool with their own resources.  A high-quality publicly-supported preschool program can produce social benefits for the middle-class as well as for the poor.

Posted in Early childhood program design issues, Early childhood programs | 2 Comments

Debt, health care, and educational investments

Posted on July 29, 2011 by timbartik

What is really the issue with the national debt? The real issue is not the short-run deficit. The deficit should be high right now to deal with the needs caused by the recession, and to stimulate the economy. The problem is the long-run deficit.

The key problem with the long-run deficit is health care spending.  Unless we figure out how we will deal with growing government spending on health care relative to the size of our economy, we will have a severe long-run deficit problem and debt problem.

Any good solution to this long-run deficit problem requires: (1) some type of controls over health care costs and some priority setting in health care spending; (2) some good long-run source of relatively economically efficient tax support for health care spending, such as a national value-added tax, as even with good health care policy, we are likely to want to spend more on needed health care; (3) making sure we free up money for high-quality investments that will increase economic growth.

As Ezra Klein and Ethan Pollack have recently pointed out, most of the current budget cut proposals focus on cutting the domestic non-security related discretionary federal budget. Over one-half of this so-called “discretionary” budget is investment in infrastructure, education and research, and over one-sixth is in education, child care, job training, and other human capital related investments.  These cuts don’t make sense because these investment funds, if they are well-spent, can increase the size of the economy, which makes it easier to afford to meet our many national needs, including health care, and to deal with our forecast debt burden.

We already see in recent state and local budget trends that state and local governments, squeezed by the elimination of federal fiscal stimulus funds, along with rising health care costs and the continued recession, have significantly cut back on education. These cut backs in education spending have the potential for adversely affecting both short-run job growth and long-run economic growth. For example, a recent analysis by Lucy Dadayan and Robert Ward of the Nelson Rockefeller Institute of Government at SUNY-Albany shows that local government education employment is down 231,000 jobs since its peak in September 2008.

The big long-run issue facing the U.S. is whether we can figure out policies that will allow us to make high-return investments at a sufficient scale to boost U.S. economic growth rates in per capita income.  These high-return investments include but are not limited to high-quality early childhood programs. This will require us to deal with our fiscal problems by creatively dealing with health care policy, and by squarely facing up to the need for efficient sources of future revenue for health care. Otherwise, health care costs in the government sector will put a squeeze on everything else government does, and health care costs in the private sector will put a squeeze on private sector real wages, which will make it more difficult to afford to meet many of our national needs, including our national need for better child development.

Substituting an “investment deficit” for a “federal budget deficit” makes no sense as a long-run strategy.

Posted in Economic development | Comments Off on Debt, health care, and educational investments

Child tax benefits’ effects on hard skills and soft skills

Posted on July 28, 2011 by timbartik

Early childhood policy is more than preschool, child care and parenting assistance programs. It also includes broader policies that may affect the quality of child development, including policies that affect parental income. Such policies include macroeconomic policies to reduce unemployment, job training policies targeted at parents, and social welfare policies that provide cash or in-kind assistance to parents.

A good recent paper by Kevin Milligan and Mark Stabile provides evidence on how government “child tax benefits” affect children’s achievement and social skills. (The paper is being published in August 2011 in the American Economic Journal: Economic Policy. A free working paper version of the paper is available here. ) The particular program they examine is in Canada, but it has some similar features to the Earned Income Tax Credit in the U.S. Under such programs, the government provides some type of cash assistance to families based on income and number of children. The Canadian program is more generous to families with very low income, whereas the U.S. EITC requires some earnings. The program benefits phase out as family income increases.

The paper finds that higher levels of child benefits increase literacy test scores from ages 4 to 6, and math test scores from ages 6 to 10, for boys whose mothers have high school education or less. The paper also finds that higher child benefits reduce physical and social aggressive behavior among girls whose mothers have a high school education or less. The estimated effects are often large.

Broader income support policies to help parents with children can complement preschool programs and other programs that provide families with services. Higher family incomes may reduce various stresses, such as being forced to move, that negatively affect child development. Higher incomes may provide parents with greater peace of mind or time to do a better job of parenting. Better early childhood services such as preschool may then build on this improved child development.

Posted in Early childhood program design issues, Early childhood programs | Comments Off on Child tax benefits’ effects on hard skills and soft skills

Pew issue brief for my book “Investing in Kids”

Posted on July 27, 2011 by timbartik

The Partnership for America’s Economic Success, a project of the Pew Center on the States, has written a short (8-page) issue brief summarizing my book “Investing in Kids”. This might be useful to those who want a short summary of my book which is deliberately targeted to address business issues.

The policy brief frames early childhood programs as part of a “new model” for economic development that focuses on “attracting, developing and maintaining human capital”.  As the policy brief points out,

“In the short run…, early education and care programs help secure skilled workers and raise their productivity by…attracting highly skilled workers with young children…Over the longer term, effective early childhood programs enable children to become more creative, adaptable, team-ready employees. [Therefore,] a comprehensive economic development strategy must include early learning programs, which [will]…retain high-paying businesses…[and] attract…sophisticated industries. ..{Economic] development authorities should collaborate with policy makers to promote the targeting of public funds toward high-quality, evidence-based early childhood programs, and consider investing economic development funds directly in quality early childhood programs.”

Posted in Early childhood programs, Economic development | Comments Off on Pew issue brief for my book “Investing in Kids”

Solving the jobs deficit while lowering the government debt burden

Posted on July 26, 2011 by timbartik

Well-known Yale economist Robert Shiller had an opinion piece in the Sunday New York Times arguing for stimulating the job market through balanced budget increases in both taxes and spending on productive public services. (Professor Shiller is known for a number of areas of research, but in particular deserves a great deal of respect for having argued back in 2003 that U.S. housing prices were overvalued.)

The basic idea of Shiller’s piece is familiar to many economists, but perhaps not so much to many non-economists. A balanced budget increase in taxes and spending will tend to increase job creation. The spending will directly increase jobs, with consequent multiplier effects on other jobs as this income is respent. The increase in taxes will have some negative effects on jobs by reducing disposable income. However, these negative effects will be less as some of this disposable income would have otherwise been saved.

Because the increase in public spending is paid for fully by increased taxes, the public debt will not increase. Furthermore, because the increased jobs will increase GDP, the ratio of public debt to GDP will decrease, which should decrease the real burden of the public debt.

What could be added to Shiller’s piece is that the increased GDP will tend to reduce future deficits. Increased GDP and jobs will reduce unemployment compensation payments and welfare payments, while increasing tax revenue. Research suggests that for each $1 increase in GDP in the short-run, the budget deficit will decline by 38 cents.  Therefore, an initially planned balanced increase in both taxes and spending will in a year or so actually reduce the budget deficit.

Shiller sensibly argues that we should focus this tax-financed increased public spending on the most useful government projects. This means that such increased public spending will have a dual benefit. In the short-run, the projects will provide much-needed jobs. In the long-run, the public services will increase economic growth and/or the quality of life.

Among the most useful possible government spending projects are early childhood programs. As I pointed out in a previous post, a balanced budget increase in public spending on preschool would cost about $175,000 per job created in the short-run.  Although this is a relatively high cost per job created, because we are paying for it with taxes, which has some negative effect, in the long-run this spending creates economic and other benefits far greater than its tax costs.  That is, the present value of the social and economic benefits from increased preschool spending far exceeds $175,000. The present value of earnings benefits alone will be 3 or 4 times these costs. The job creation benefits are bonus benefits.

Our political system is currently focused on the budget deficit. Although the budget deficit is a problem, it is mostly a long-run problem. The more acute short-run problem is the jobs deficit. We need to aggressively address the jobs deficit with programs that will also improve the long-run budget deficit problem, while growing the economy. Early childhood programs could be part of a needed policy package to address all these goals.

Posted in Early childhood programs, Economic development, National vs. state vs. local | Comments Off on Solving the jobs deficit while lowering the government debt burden

Politically attractive financing for early childhood programs

Posted on July 25, 2011 by timbartik

The most straightforward way to finance early childhood programs is to have the government pay for these programs upfront. However, this may not be politically attractive at certain places and times.

I’ve mentioned previously the idea of financing early childhood programs through their effects in reducing later special education costs. Some philanthropists or private investors would provide the money upfront for an early childhood program. Their investments would be repaid by local school districts and state governments through savings in special education costs. Excess savings, if any, would be plowed back into sustaining the financing for early childhood programs. As I mentioned, as of right now, we are uncertain whether this model would work. Are the savings in special education costs sufficient to make this model sustainable? Some experiments or demonstration projects are needed.

However, there are other options for financing early childhood programs through what are sometimes called “social impact bonds” or “pay for success contracts”. (See the work of Jeff Liebman, a professor at the Kennedy School, for a discussion of a variety of possible policy areas in which such bonds/contracts might be feasible.) Specifically, it would be possible for a state government to finance early childhood programs through paying for these programs’ effects on kindergarten readiness.

To give a specific example, a state government could say it would be willing to pay $2,000 upfront per child for a preschool program. But the state government would also contractually commit to paying a certain amount per improvement in test scores at kindergarten entrance. For example, the payment could be $500 for each 1 percentile improvement in average test scores at kindergarten entrance.  A good one-year half-day pre-K program can raise test scores by 10 percentiles or more. (For example, this is implied by Reynolds’s results for the Chicago Child-Parent Center program). A program that raised test scores by 10 percentiles would receive a bonus payment of $5,000 per child. The upfront $2,000 per child plus the bonus payment of $5,000 would be more than enough to pay for one year of a high quality half-day preschool program, which might cost around $5,000, based on data from the Institute for Women’s Policy Research’s publication, “Meaningful Investments in Pre-K”.

This financing scheme might have two advantages. The first is political. In some states’ political situations, tying preschool financing to program performance might be reassuring to some policymakers.  The state only has to make large payments if preschool programs prove to be effective.

The second advantage is substantive.  This payment structure provides strong incentives for good performance.  Programs will only be created and survive if they are able to deliver good performance.  If better results can be obtained from a particular set of program features, such as a more learning-rich curriculum or higher teacher qualifications, there are some positive incentives to adopt those program features. If full-day programs produce much greater results than half-day programs, there are incentives to offer full-day programs.

One possible problem with this “pay for success” plan is whether success can be accurately measured.  As I have previously discussed, for one-year preschool programs for 4 year olds, effects on kindergarten readiness can be accurately measured using “regression discontinuity” evaluation designs. (For pre-K, such evaluation designs in essence  compare test scores of children who are just starting pre-K and just missed the age cutoff for pre-K the previous year, with similar children who completed pre-K the previous year, are just starting kindergarten, and just made the age cutoff for pre-K the previous year. )

A second problem is making sure that the tests used to measure “success” in fact measure what we want to encourage pre-K to do. For example, if we want pre-K to include some emphasis on “soft skills/social skills”, then such tests need to be included in the formula for determining bonus payments for pre-K providers.

A third problem is that although large for-profit providers and public school districts might be able to provide the funding upfront to pay for program costs, before any bonus payments are received, obtaining upfront funds might be more difficult for smaller private pre-K providers. This problem might be in part addressed by providing some state, philanthropic, or private investor financing for smaller pre-K providers.

Given the advantages but possible problems in “paying for kindergarten readiness” contracts for financing pre-K, it would seem appropriate for government policymakers to do demonstration projects to explore this approach’s potential. Perhaps some state will explore doing so as part of their application for federal funds under the competition for the Race to the Top –Early Learning Challenge program.

Posted in Early childhood program design issues, Early childhood programs | Comments Off on Politically attractive financing for early childhood programs