The political economy of early childhood programs

Posted on February 2, 2011 by timbartik

This lengthy series of posts on the income distribution effects of early childhood programs illustrates why some foundations, such as Pew, have been so interested in “universal” pre-k. The attraction is that universal pre-k potentially helps overcome some perceived limitations in macro effects and political sustainability of other initiatives in early childhood intervention.

Some highly targeted early childhood programs, such as the Nurse Family Partnership, may have modest costs. Unless the benefit-cost ratio for these programs is extraordinarily high, this also implies that the “macro” effects of these programs on the poor will be modest as well. Of course, we could combine a number of these programs to get larger effects. But then the combination of these programs will suffer some of the problems of more intensive targeted programs.

Targeted early childhood programs that are more intensive, such as the Abecedarian program, can have large effects in helping the poor.  Such programs may have benefits exceeding costs. However, these programs may have costs for the non-poor that are relatively large, but deliver relatively few direct economic benefits to the non-poor. (Of course, the programs may deliver more indirect benefits to the non-poor, such as reduced costs of crime and welfare.)

The political dilemma for targeted programs is that they either are too costly to be politically attractive, or insufficiently intense to have large effects on the problem. This dilemma could be overcome if we magically conjure up a program with extraordinary ratios of benefits to costs.  We could have large effects at modest costs if we could find an early childhood program with a benefit-cost ratio of 100 to 1. But although high-quality early childhood programs have benefits that are many multiples of costs, they are unlikely to reach such extraordinarily high ratios.

More universal programs, such as universal pre-k, have the political attraction that they may plausibly deliver some direct benefits to the non-poor.  In the case of universal pre-k, the data seem to suggest that the resulting distribution of benefits is still targeted on the poor. However, the benefits of universal pre-k are broadly enough spread that benefits exceed costs for the middle class. This provides economic as well as political reasons for broadening the program. Finally, because the poor still benefit the most, the resulting program is still reasonably highly targeted. The benefits are certainly much more targeted than policies, such as business incentives, that simply increase overall job growth.

The political and economic appeal of broadening program services still leaves room for much debate over exactly how to broaden services. Broadening services to the middle class may provide most of the extra political and economic support that is promised by “universality”.  Whether services should also be broadened to the upper class may depend more on the political context.

One issue is whether we can identify other early childhood interventions with sufficiently broad benefits that a well-designed program might want to serve the middle-class as well as the poor. For example, one wonders whether such broad benefits might be feasible for some parenting education programs.  My book, Investing in Kids, focuses on early childhood programs for which we have good evidence of long-term effects.  We should certainly continue to explore the potential long-term effects of other early childhood programs.  At the same time, we should explore how to change the political climate to develop greater support for more targeted approaches, even if these are expensive for the non-poor.

Posted in Business incentives, Distribution of benefits, Early childhood programs | 8 Comments

How does U.S. investment in preschool compare to other developed countries?

Posted on February 1, 2011 by timbartik

Laura Bornfreund at Early Ed Watch has a useful post comparing U.S. investment in preschool with other developed countries. The other developed countries are those in the Organization for Economic Co-operation and Development (OECD). OECD includes such countries as the United Kingdom, Germany, France, etc.  Ms. Bornfreund uses official data from the OECD’s annual Education at a Glance, 2010.

What she finds is that for U.S. students enrolled in preschool, the class sizes and spending per student are comparable or even “better” (greater spending per student, smaller class sizes) than the OECD average.  However, the U.S. has a lower percentage of 3 and 4 year olds enrolled in preschool than the OECD average. As she points out, among 3 and 4 year olds, several European countries have near 100 percent enrollment in preschool.

Of course, we don’t know the quality of this preschool in all these countries.  However, at the least this illustrates a political difference between the U.S. and many other developed countries.

I suspect that for many of these countries, there is very little rigorous social science evidence of the long-term effectiveness of the preschool program. Yet many of these countries invest a great deal in near-universal preschool. The political climate is such that these governments are willing to invest in preschool simply because the concept makes sense. Rigorous empirical evidence is not required.

As a good policy wonk, I of course prefer to have research findings of program effectiveness. Even if we agree that the program is worth funding, research can point to ways to improve the program. But, among developed countries, the U.S. is somewhat unusual in demanding exhaustive evidence of program effectiveness for many social programs, including early childhood programs. This reflects a political climate that is suspicious of government intervention. No one can wave a magic wand and change this political climate, and in any event, there are pros and cons of being suspicious of government intervention.  The need for demonstrated research results for government investments in early childhood programs is simply a reality of the U.S. political context.

Posted in Early childhood programs | 1 Comment

Income distribution effects of the Nurse Family Partnership

Posted on February 1, 2011 by timbartik

Chapter 8 of Investing in Kids also estimates the effects on different groups of “full-scale” implementation of the Nurse Family Partnership program. The NFP provides nurse home visitation services to first time mothers, from disadvantaged families, during the pre-natal period until age 2.

NFP has been studied by several random assignment experiments, so we have rigorous evidence on its effects on the participant mothers and children. These experiments establish that the program is much more effective for disadvantaged families. The experiments also establish that the program works better if the services are delivered by nurses rather than delivered by staff with lower professional credentials (and lower pay).

I studied the income distribution effects of a “full-scale” NFP. Full-scale means that the program can serve all first-time disadvantaged mothers. This includes roughly 9% of all children.

A full-scale NFP has economic development benefits for state economies that significantly exceed costs. The ratio of state economic development benefits to costs is about two.

However, the net earnings benefits for the lowest income quintile are modest. Earnings for the lowest income quintile would increase by less than 3% of income. This is less than one-half of the earnings benefits for the lowest income quintile from universal preschool, and less than one-tenth of the earnings benefits for the lowest income quintile from the Abecedarian program.

The NFP’s benefits are limited in part because, by design, it only serves a particular segment of the low income population. In addition, the hours of service per participant household are more limited than for preschool or full-time child care.

On the other hand, the NFP’s costs are more modest.  Even at full-scale, the program’s annual costs are less than $4 billion for the entire U.S.  The program has costs for the upper 80% of the income distribution, but these costs are modest. The average cost of the taxes to pay for the program amount to one-20th of 1% of income, or one part in 2000 of income.

Therefore, even though the NFP program is highly redistributional, with its earnings benefits focused on low income households, it is perhaps politically feasible due to its modest costs. But its modest costs may also limit the size of NFP’s benefits.

Of course, it should be noted that the NFP is not really intended to have large “macro” effects on the income distribution by itself. NFP just provides one kind of valuable service for a segment of the low income population. It is unreasonable to expect it to have large income distributional effects.  NFP when combined  with many other effective anti-poverty programs might have large effects on the income distribution.

Posted in Distribution of benefits, Early childhood programs | 2 Comments

Income distribution effects of high-quality full-time child-care and preschool, birth to age 5

Posted on January 31, 2011 by timbartik

Chapter 8 of Investing in Kids also considers the effects on different income groups of more intense early childhood programs. I consider the effects of a program similar to the Abecedarian program.

The Abecedarian experiment provided full-time child care and preschool, from shortly after birth to age 5, to children from disadvantaged families. The program was run as a random assignment experiment, so we have rigorous evidence on its effects. We have follow-up information on Abecedarian participants through age 21. The Abecedarian program is a past experiment. However, it is quite similar to the current Educare demonstration program sponsored by the Ounce of Prevention Fund and the Buffett Early Childhood Fund.

I consider how an Abecedarian program targeted at the disadvantaged would affect the income distribution. I assumed that such a program would be restricted to families in the lowest 15 percent of the income distribution, but would be provided to all such families. One reasons for targeting this program is that it is so expensive. The program’s gross costs, over the five-years of service, are almost $80,000 per child. Even if the program was targeted at low income families, its annual net costs would be about $40 billion.

An Abecedarian-style program for every child whose family’s income was in the lowest 15% of the income distribution would have large effects on average incomes in the lowest income quintile. These lowest income quintile households have less than about $20,000 in annual income. An Abecedarian-style program would boost the earnings of the lowest income quintile by over 35% of this quintile’s income.

The Abecedarian boost to the lowest income quintile is considerably more than is provided by one year of half-day school year preschool. As detailed in a previous post, such a preschool program does provide a significant boost to the earnings of the lowest income quintile, of over 6% of quintile income. But the Abecedarian boost is over five times as great.

The Abecedarian program’s benefits to the lowest income quintile are so great that, despite its high costs, such a program provides economic development benefits to a state economy that exceed program costs.  Each dollar of program costs increases the present value of state residents’ earnings by over $2.

On the other hand, the Abecedarian program has large costs for the remaining 80% of all households, those with annual household income over $20,000. Due to the taxes needed to pay for a full-scale Abecedarian program, these remaining four income quintiles would suffer income losses of about one-half of one percent of income. This is a considerable cost to pay for one government program. It could be argued to be affordable, but it might be a hard sell.

These large costs for most households contrast with universal pre-k. As detailed in a previous post, universal pre-k has net benefits for about 60% of all households. Universal pre-k’s net excess of taxes over benefits for the remaining 40% of upper income households are only one-fifth of the net losses caused by the Abecedarian program.

Of course, an Abecedarian-style program does not need to serve all children from low-income families.  But a more limited Abecedarian program would face the challenge of deciding which low-income households to choose from among those eligible. This raises questions of equity among low-income households. In addition, scaling back the program would limit the benefits as well as costs of the Abecedarian program.

A full-scale Abecedarian –style program could have dramatic effects on the income distribution. But it is so targeted and so expensive that it is reasonable to doubt whether such programs could be politically expanded to a large-scale.  Such large-scale expansion might require a considerably different political and economic climate.

Posted in Distribution of benefits, Early childhood programs | 2 Comments

Amazon copies of Investing in Kids now in stock

Posted on January 28, 2011 by timbartik

Amazon now lists my book Investing in Kids as in stock now, both in paperback and hardback, and therefore available for immediate shipping. Before, Amazon for some reason had the book listed as “available in one to two months”, which is not exactly the most compelling sales pitch. For those of you who prefer buying from Amazon rather than directly from the Upjohn Institute, buying from Amazon will now be quicker than it was before, although I doubt if it took 1 to 2 months before. From my perspective as an author, a higher sales rank from Amazon would be a plus. If you’ve read the book and wish to review it at Amazon, that would be very helpful as well.

Posted in Uncategorized | Comments Off on Amazon copies of Investing in Kids now in stock

Sliding scale fees for pre-k

Posted on January 28, 2011 by timbartik

An expanded pre-k program may provide both greater benefits for a state’s economy, and be more politically sustainable, if this expanded program provides services not only to the poor but to the middle class. On the other hand, free preschool for children from upper income families may have lower economic development benefits, as many of these children are already in high-quality preschool programs paid for by their parents. Free preschool for children from upper income families may also have political costs, as some voters may resent such subsidies.

One compromise solution is to expand pre-k as a universal program, but to charge some fees to parents from upper income families. This will reduce the net taxpayer cost of the universal pre-k program, and will reduce subsidies to upper income families that may be politically damaging.

I should mention that this issue of “fees” is separate from the issue of whether pre-k is best delivered by public schools or by private preschools. Both a “free universal” program and a “fee-charging universal” program could have varying combinations of public versus private delivery.

Charging fees faces some issues. First, any fee system will have some additional administrative costs. A fee that varies with family income will have even more administrative costs. The extent of these extra administrative costs depends upon whether the program insists that the fee always reflect an accurate measure of the family’s current economic status.  For example, how is family income verified? How often must family income be re-verified?

Fees may also have some stigma costs.  Depending upon how fees are administered, some families may feel that the sliding scale fees identify the program as being only for disadvantaged families.  This might reduce enrollment by some disadvantaged and middle class families, even if these families do not pay any fees. If these fees reduce enrollment by disadvantaged and middle class families, then the fees will reduce the net benefits of the expanded preschool program for a state’s economy.

In chapter 8 of Investing in Kids, I did some simulations of the effects of charging some fees for a universal pre-k program. The fees are assumed to be charged only to children from families in the top 40% of the U.S. income distribution, in households with over $62,000 in income. The assumed fee is half the average costs per child of the preschool program, which still makes the program quite competitive with typical preschool and child care costs.

Compared to a free universal pre-k program, this sliding scale fee universal pre-k program reduces overall program costs by about 10%. The fees cause some upper class families to not use the universal program. The fees also reduce costs to taxpayers of the program by about one-fourth compared to a free universal pre-k program.

On the other hand, the fees do have significant administrative costs.  Overall, the net economic development benefits for a state economy from a sliding scale fee universal pre-k program, versus a free universal pre-k program, are about the same.

The main effect of sliding scale fees is to redistribute the net benefits of a universal pre-k program within the top 40% of the U.S. income distribution. Upper income households without children benefit more from a sliding scale fee vs. free universal program, because of the lower taxpayer costs of the sliding scale fee program.  Upper income households with children benefit more from a free universal pre-k program than from a sliding scale fee program.

Overall, it seems unclear whether sliding-scale fees make a universal pre-k program easier to enact and sustain. Whether sliding scale fees are a political plus depend on political attitudes among both pre-k users and non-users in upper-class income groups.  Fees are not an obvious magical solution to making universal pre-k more politically feasible. But sliding scale fees may be politically helpful in some circumstances. From an economic perspective, charging sliding-scale fees for the upper class is unlikely to have big enough net effects in reducing costs, once one accounts for administrative costs, for there to be big economic gains from charging fees.

Posted in Distribution of benefits, Early childhood program design issues, Early childhood programs | Comments Off on Sliding scale fees for pre-k

The politics of universal vs. targeted pre-k

Posted on January 27, 2011 by timbartik

As a previous post has outlined, there are some economic tradeoffs in deciding between a pre-k program with “universal access” for all 4-year-olds, versus a pre-k program that is more narrowly targeted at children from low-income families.  Targeting children from low-income families for pre-k costs less, and has higher benefits per dollar invested. On the other hand, targeted pre-k only benefits a small minority of all income groups. Universal pre-k benefits a majority of all income groups. Universal pre-k also increases the net economic development benefits of preschool, by expanding services to children from middle-class households who will have significant earnings benefits from expanded access to high-quality preschool.

Therefore, the economic arguments are mixed. What about the political merits of universal vs. targeted preschool? Which approach to preschool policy is likely to be more successful in sustaining high-quality preschool programs over time? Chapter 8 of my book Investing in Kids addresses this issue.

An argument for universal pre-k is that targeted programs in the U.S. tend not to sustain broad political support. According to Harvard sociologist Theda Skocpol, “When U.S. antipoverty efforts have featured policies targeted on the poor alone, they have not been politically sustainable.”

On the other hand, an argument against universal pre-k is that it subsidizes upper class families who do not need government aid. As Novel prize-winning economist James Heckman has argued, universal pre-k may create “an opposition group saying “Why should we subsidize affluent working women?””.

An “in-between” political argument is provided by Robert Greenstein, executive director of the Center on Budget and Policy Priorities. According to Greenstein, “targeted programs…are more likely to be strong politically when they serve low-income and moderate-income working families as well as the very poor….Advocates of new universal programs need to acknowledge the political difficulties posed by the large costs of such programs…”

I think it is apparent that the political strength of universal versus targeted pre-k will vary from state to state and over time. However, on the whole, I think the arguments for universal pre-k are more compelling. The political and economic benefits of expanding pre-k to the middle class seem to argue in favor of universal pre-k.

A possible fallback position is to have an “in-between” pre-k program that could be described as “broadly targeted” or “near universal”. Such a program would include the middle class as well as the poor, but take some steps to reduce program costs by reducing subsidies for the upper class. One way to try to do this is through sliding scale fees for preschool. I will consider the issue of preschool fees in a future post.

Posted in Distribution of benefits, Early childhood programs | 1 Comment

The State of the Union and the States

Posted on January 26, 2011 by timbartik

President Obama’s State of the Union address focused on “winning the future”. This strategy included winning the future through public investments, including educational investments.

For advocates of early childhood programs, it is noteworthy that the educational investments that were mentioned in the SOTU speech were in K-12 and higher education. Early childhood investments were not mentioned in the speech. Perhaps some initiatives will appear when the federal budget is announced in a few weeks. However, based on recent reports from Lisa Guernsey, of Early Ed Watch at the New America Foundation, this appears doubtful. If there are any early education initiatives in the budget, they appear likely to be modest, given the announced freeze on federal discretionary spending.

The SOTU address also extensively discussed competitiveness of American business. But other noteworthy omissions from the SOTU speech were proposals to increase the competitiveness of U.S. manufacturers through an expansion of the federal Manufacturing Extension Partnership, or to increase the competitiveness of U.S. manufacturers and other businesses through an expansion of customized job training. I have advocated for expanding these programs in a recent paper for the Hamilton Project.  Chapter 5 of Investing in Kids includes a discussion of improving the efficiency of business incentive design. Chapter 5 discusses the evidence for why manufacturing extension and customized job training are far more efficient business incentives than typical tax incentives, such as property tax abatements. I know that proposals to expand such services have been discussed in the Administration. But they did not make the cutoff for the SOTU speech, which does not bode well for such proposals being included in the budget.

We can debate why early childhood programs and effective services to manufacturers were omitted from the speech.  Why were they less politically attractive in the current circumstances than what was mentioned in the speech? I suspect it has something to do with political resistance to expanded federal intervention in both early childhood programs and assisting individual manufacturers.

But the bottom-line in this: if investment in early childhood programs is to be maintained and expanded over the next several years, this is largely going to be up to state and local government decisions.  If business incentives are going to de-emphasize tax incentives, and put more emphasis on efficient business services, this is largely going to be up to state and local government decisions.

This is one reason why, as events would it, Investing in Kids is turning out to be a timely book. The book focuses on the state and local returns, in economic development benefits, from early childhood programs and reformed business incentives. It is state and local policymakers who over the next several years will determine the future investment in such programs.

Posted in Business incentives, Early childhood programs, Economic development, National vs. state vs. local | Comments Off on The State of the Union and the States

Targeted pre-k vs. universal pre-k: overall effects and distributional effects

Posted on January 26, 2011 by timbartik

In a previous post, I showed that under plausible assumptions, universal pre-k would have broad benefits for the middle class as well as for the poor. But how does universal pre-k compare with a preschool program that is more tightly targeted on the poor?  After all, it seems reasonable to assume that preschool’s effects on the poor are higher per child. Therefore, targeting would seem to have some merit.

In chapter 8 of my book Investing in Kids, I include some comparisons of universal and targeted pre-k. I consider the benefits of a targeted preschool program that is restricted to families with less than around $34,000 of income. This is about one-third of all families.

Compared to universal pre-k, such a targeted preschool program has significantly lower costs. I estimate that this targeted program only costs one-fourth as much. Furthermore, this targeted program still makes preschool accessible to all poor children, who have higher benefits from preschool.

This targeted pre-k program has higher overall benefits per dollar invested. The overall benefit to cost ratio for universal pre-k is 2.78. In other words, for each dollar invested in universal pre-k, there is an extra $2.78 for the state economy in economic development benefits. These “economic development benefits” are higher per capita earnings for state residents.  In contrast, for targeted pre-k, the ratio of state economic development benefits to costs is 7.16. This is over twice the ratio for universal pre-k.

However, the net benefits for state economies from universal pre-k, compared to targeted pre-k, are about 10% greater. Expanding pre-k to middle income families has lower returns per dollar than the returns per dollar from targeting the poor. These lower returns per dollar reduce the average returns per dollar of universal pre-k compared to targeted pre-k. But the net returns from expanding pre-k to the middle class are still positive.  These positive net returns to the middle class from universal pre-k are sufficient to offset the excess of taxes over benefits for upper income households.

One could consider a pre-k program that is less tightly targeted. For example, based on these distributional assumptions, there would be even greater net benefits from a program that was “targeted”, but very broadly, at the bottom 60% of all households, with less than $62,000 in income. However, I don’t think this is what most people mean by a “targeted” preschool program. Usually “targeting” means focusing on the poor and near-poor, not including the middle class. A future blog post will consider whether charging sliding scale fees for preschool to children from upper-class families could help shape a preschool program that is more broadly targeted to include the middle class.

From a pure economic perspective, going beyond targeted pre-k to universal pre-k pays off modestly in producing net benefits.  However, recall that these estimates are based on the assumption of a quite drastic tail-off in benefits from preschool as we go from the poor to the middle class. If this tail-off is less drastic, then the economic benefits from going beyond targeted pre-k to universal pre-k would be greater.

However, any analysis of targeted vs. universal pre-k must also consider the politics.  I will address this in a future blog post.

Posted in Distribution of benefits, Early childhood programs | 1 Comment

Preschool and state economic development: employers speak out

Posted on January 25, 2011 by timbartik

One of my Google News alerts came across this interesting story from Iowa.  Iowa is currently debating the future of its “Statewide Voluntary Preschool Program”, which currently enrolls over half of all Iowa four-year olds. The program is designed to move towards universal access for all four-year-olds, although school districts are allowed to have sliding scale fees for children from families who are ineligible for free or reduced price lunch.

The Iowa House of Representatives recently voted to eliminate funding for the program, and the debate has now moved to the State Senate. The Governor and some state legislators have discussed possibly replacing this universal program with a voucher system for preschool that would be more targeted to assist low-income children.

As part of this debate, on Monday January 24th the Iowa Senate Education Committee heard testimony from Elliott Smith, the Executive Director of the Iowa Business Council. The Iowa Business Council represents the state’s 20 largest employers.

According to a report from Radio Iowa, Elliott Smith, on behalf of the Iowa Business Council, took a strong position in favor of the state’s current universal pre-k program. He argued that Iowa’s current program was important for the state’s economic development, both long-term and short-term:

“The quality of Iowa’s education system has been a top priority for the Business Council for many years. It’s the graduates of this system from which our labor pool is stocked and replenished,” Smith said.

“Access to quality early childhood education should be available to parents of all Iowa children regardless of socioeconomic status — with no gaps.” According to Smith, state-funded preschool is a selling point when Iowa businesses are trying to lure employees to move here from other states.

These comments on behalf of the Iowa Business Council demonstrate that there is support from the business community for the notion of early childhood education as a key part of economic development strategies. Furthermore, this testimony provides some evidence that a state’s investment in universal pre-k can help tip location decisions in a state’s favor.

These arguments are consistent with the main point of my book Investing in Kids : a state or local government’s investment in high-quality early childhood programs can pay off in economic development benefits for a state’s economy, both in the long-run and short-run.

Posted in Early childhood programs, Economic development, Timing of benefits | 1 Comment