VoxEU Column Education

Long-term effects of investments in universal early education: Evidence from the American kindergarten expansion

Proponents of universal early education hope that such investments will yield long-term socioeconomic benefits. This column presents evidence that state governments funding public kindergartens in the US actually widened socioeconomic disparities across adults, as whites made gains that blacks did not. That result highlights the sensitivity of policy interventions to their means of funding and the structure of existing alternatives.

There are perennial efforts to expand public schooling to children under age five in the US, and universal early education is already a mainstay in many European countries. Such programs come at a considerable cost to taxpayers, but they have a number of theoretical justifications. For example, early education may be difficult for many families to fund, and in addition to their private returns, investments in early education might benefit society at large. Social spending earlier in the lifecycle might also have more “bang for the buck,” given that children may be most susceptible to intervention when young, and given the potential complementarities between earlier and later investments. This argument has been forwarded to justify expansions of means-tested early intervention (Cunha, et al. 2008). And while the same arguments could justify a means-tested program, a universal program may better promote equal access.

Still, there is little empirical evidence to date of the benefits of universal early education, particularly over the long term.1 Does it raise productivity, eventually paying for itself through higher tax contributions? Does it ultimately lower expenditures on social programs like cash welfare? Does it help to close skill gaps between the rich and poor? In a recent paper (Cascio 2009a), I address these questions by examining the long-term consequences of a series of state initiatives to fund public school kindergartens in the US. The US kindergarten expansion highlights how the effects of universal early education are potentially shaped by the funding mechanism and the existing policy landscape.

The kindergarten expansion and its long-term impacts

State funding of kindergartens in the US is a surprisingly recent development. When the modern era of federal intervention into early education began in the mid-1960s with the introduction of Head Start, about only half of state governments made contributions to school districts offering educational programs for five year olds. Over the two decades that followed, the remaining states began supporting kindergarten on an equal basis with later grades, closing this funding gap. State policymakers argued that the funds would expand access to kindergartens, and in doing so, reduce school failure rates and lower eventual state expenditure on social programs.

The state initiatives did meet their first-order goal of increasing the number of seats in kindergarten programs. Figure 1 shows that kindergarten enrolment as a share of first-grade enrolment in states passing funding initiatives began a dramatic rise starting in the mid-1960s, converging to that in the rest of the country in the 1990s. Figure 2 gives more direct evidence on the contribution of state funding to this trend, plotting the ratio of kindergarten to first grade enrolment, on average, in the years around passage of a state funding initiative. These averages are normalised to zero in the year prior to the initiative and adjust for secular trends at the state level and shocks to enrolment among comparable states.2 The kindergarten enrolment rate was on average almost 30 percentage points higher than would have otherwise been expected after a funding initiative was passed, and most of the gain was achieved quickly, within only two years. Kindergarten enrolment did not deviate from trend in the years prior to the initiative, suggesting that these gains are truly driven by state funding, not some other unobserved factor or policy.

Figure 1. Trends in public school kindergarten enrolment

Figure 2. Effect of state funding on kindergarten availability

State funding thus generated a sharp difference in access to kindergartens among children born only a few years apart. Did the investment pay off? Did it narrow socio-economic gaps in well-being? To address these questions, I estimate the effects of a funding initiative in one’s state of birth on several outcomes targeted by state policymakers, de-trending the data as described above.3 I present separate estimates by race, an available proxy for socio-economic status during childhood.

Kindergarten programmes widened racial disparities

My findings are consistent with long-term benefits that potentially justify the investment in kindergarten programs, but they point to increased – not reduced – socioeconomic disparities in adult well being for cohorts entering school after kindergarten was state-funded. For example, white high school dropout rates were lower than expected among cohorts who turned five after an initiative was passed, but black high school dropout rates were not. I also find a significant reduction in institutionalisation among whites that is unlikely to have occurred for blacks.

The finding that kindergartens widened racial disparities is surprising. If anything, we would have expected kindergartens to have generated a relative rise in human capital investments in black five year olds. Why did this not occur?

One possible explanation is that high quality education was already widely available to blacks through existing public programs. Along these lines, I find that the state-funded kindergartens drew black five year olds out of Head Start, which has been found to have yielded significant long-term benefits to participants in the cohorts of interest (Garces, Thomas, and Currie, 2002; Ludwig and Miller, 2007). Another possible explanation is that there was a decline in the relative quality of schools attended by blacks turning five after the initiatives. I lack the data to show this directly, but suggestively, school districts responding to the state grants do not appear to have raised enough additional revenue to cover the costs of operating kindergarten programs on par with later grades. This failure to augment state funds may have resulted in lower-quality kindergartens than anticipated across the board, and blacks may have been more likely to be assigned to the lowest-quality programs. Even if the quality of kindergartens was not affected, the necessary funds would have needed to come from somewhere, and may have been diverted from other, later school programs from which blacks may have benefited disproportionately.


It would be hasty to generalise these findings to another time or place. Nevertheless, the US experience with kindergartens offers lessons for policymakers today, particularly if an interest in universal preschool stems from a desire to “level the playing field” for kids from disadvantaged backgrounds. First, it is important to consider the availability and quality of existing alternatives. At least some disadvantaged children might already have access to preschool education through other government programs. Second, it potentially matters how an expansion to universal programs is funded. If subsidies don’t fully cover the costs of offering the program, providers might make choices about resource allocation that continue to disadvantage those children most in need.


1 Short-term effects for children of an expansion in universal childcare in the province of Québec, Canada were recently explored by Baker, Gruber, and Milligan (2008).

2 More formally, Figure 2 plots the coefficients on indicators for year relative to the initiative from a regression that includes as controls state dummies, linear time trends by state, and dummies for year interacted with dummies for Southern region and dummies for prior state expenditure on education (medium and high). Some of these trends are likely due to the initiatives, so these estimates likely provide a lower bound on the effects of state funding.

3 The findings are based on Census data from 1980, 1990, and 2000. Underlying regressions also adjust for age and Census year.


Baker, Michael, Jonathan Gruber, and Kevin Milligan. 2008. “Universal Childcare, Maternal Labour Supply, and Family Well-Being.” Journal of Political Economy 116(4): 709-745.

Cascio, Elizabeth U. 2009a. “Do Investments in Universal Early Education Pay Off? Long-term Effects of Introducing Kindergartens into Public Schools.” NBER Working Paper 14951, May 2009.

Cascio, Elizabeth U. 2009b. “Maternal Labour Supply and the Introduction of Kindergartens into American Public Schools.” Journal of Human Resources 44(1): 140-170.

Cunha, F., J. J. Heckman, L. J. Lochner and D. V. Masterov. “Interpreting the Evidence on Life Cycle Skill Formation,” eds. E.A. Hanushek and F. Welch, 2006. Handbook of the Economics of Education, 12, 697-812, Amsterdam: North-Holland.

Garces, Eliana, Duncan Thomas, and Janet Currie. 2002. “Longer Term Effects of Head Start.” American Economic Review 92(4): 999-1012.

Ludwig, Jens and Douglas Miller. “Does Head Start Improve Children’s Life Chances? Evidence from a Regression Discontinuity Design.” Quarterly Journal of Economics 122(1): 159-208.

2,205 Reads