Skip over navigation

Journal Issue: The Next Generation of Antipoverty Policies Volume 17 Number 2 Fall 2007

Improving the Education of Children Living in Poverty
Richard J. Murnane


Implementing these recommendations will entail significant costs, though the costs are relatively modest given the importance of the problems that the recommendations address and the social payoff to solving them. Moreover, because the proposals do not involve entitlements, the size of the initiatives could be tailored to federal budget realities. I consider the cost of each proposal briefly, in turn.

It would cost little to revise adequate yearly progress requirements along the lines suggested by Robert Linn as part of the reauthorization of NCLB. The same would be true for making improved high school graduation rates one of the criteria for satisfying those requirements. But to estimate graduation rates accurately, most states need dramatically improved data systems. It is thus important to continue the Department of Education grant program that helps states develop systems to track individual students over time. Increasing annual funding for those grants from $25 million to $50 million would be a good investment, because high-quality long-term data on students are essential both to good educational policymaking and to evaluating the effects of innovations and new investments. Moreover, because good data systems are low on the education priority list for many state legislatures, substantial grants may be necessary to catalyze progress.

Congress might appropriate another $60 million for grants to states interested in aligning high school graduation requirements with the demands of postsecondary education and work. And it might appropriate another $20 million to push forward research on how to make the twelfth-grade NAEP examinations serve as a benchmark for varying state graduation requirements.

The annual budget of the Massachusetts METCO program, which serves 3,300 students, is roughly $20 million, or $6,000 per student. Using these figures as a base, I estimate that a one-to-one matching grant awarded to states to create interdistrict choice programs might cost the federal government $3,000 per student, on average, including funding for external evaluations. A $120 million annual investment in competitive matching could provide new educational options for approximately 40,000 low-income students attending poorly performing schools. Although this initiative is expensive on a cost-per-student basis, evaluations could provide extremely valuable information about how to design interdistrict choice programs.

The most costly recommendation is the third: targeted competitive matching grants for state and district initiatives to improve teaching in high-poverty schools and to tackle the secondary school problem. To interest districts and states in applying for the matching grants, the federal contributions would have to be large enough to fund the required evaluations and provide significant program money as well. Congress might invite proposals for projects with a total cost, including evaluations, of up to $20 million a year for up to five years, with a one-to-one matching rate. Thus, the annual federal cost for each project would be a maximum of $10 million. Funding 100 such projects would cost $1 billion.

Competitive matching grants to address the secondary school problem would be somewhat more costly. The annual federal cost for grants to states to reduce dropout rates might be $500 million. Another $800 million a year could usefully go to school districts to develop and test new educational models for increasing high school graduation rates while retaining high standards.40

The total annual cost to the federal government of my proposals would be $2.525 billion. 41 That is roughly 20 percent of the $12.7 billion that Congress now allocates under Title I, Part A, of No Child Left Behind for locally designed programs to improve the skills of students at risk of academic failure, especially those attending high-poverty schools.42 Should the funds for the proposed programs come out of this appropriation? There are arguments on both sides.

The main reason to use existing Title I funds to pay for these programs is that Title I funding has not improved the achievement of the target population of students.43 Why not reallocate the money to more promising uses? The complication is that, contrary to the specific provisions of the legislation, federal Title I funding may not have brought about a longterm increase in the resources used to educate disadvantaged children. Although an increase in Title I funds allocated to a district does initially raise the district’s instructional spending, the increase almost entirely disappears after three years.44 In other words, the Title I funds end up paying for core expenses, such as teacher salaries and professional development, that otherwise would have been paid for with local or state tax revenue.

How can federal Title I funds end up replacing local and state education funding when the ESEA legislation explicitly prohibits such substitution? Let us consider the example of Central City, an urban district in which all schools are eligible for Title I funds because they serve high concentrations of low-income students. Suppose in 1992 the average spending per student in Central City’s schools is $6,000, of which $1,000 is Title I funds. In 1993 Central City receives an increase in Title I funding of $500 per student because the newly available 1990 census shows that the number of low-income students in Central City has grown markedly. In accordance with the law, Central City increases per student funding for the 1993 school year by $500, to a total of $6,500. Over the next several years, however, the purchasing power of the $6,500 is eroded by inflation. To maintain the ability of its schools to purchase the goods and services that they had purchased in 1993, the city council would have had to increase local funding. But facing strong pressures from voters to keep property taxes in check, it does not do so. So by 1995, the real purchasing power of the $6,500 is no greater than that of the $6,000 per student that the district spent in 1992. In effect, the increase in Title I funding has allowed the district over a several-year period to avoid the unpopular tax increase that would have been necessary to keep real per student spending (that is, net of inflation) constant.45

If Title I funds used by a district to fund core educational activities were withdrawn, the district would face a fiscal crisis. It would need either to raise taxes to generate more revenue or to reduce spending by eliminating professional development or by laying off teachers and increasing class sizes. Districts under such fiscal pressure would hardly respond favorably to invitations to compete for matching grants for interdistrict choice programs or programs to increase high school graduation rates.

For these reasons, I recommend that federal funds for these programs be new money, added to the federal education budget. Additional funding would address the frequent criticism that NCLB is a laundry list of new unfunded mandates. Districts could use their Title I money to pay for their share of the cost of competitive matching grants.