Costing-Out the Price of Education
Testimony of Nathan A. Benefield, Director of Policy Research, Commonwealth Foundation for Public Policy Alternatives to the Pennsylvania Senate Education Committee
December 8, 2009
Thank you, Sen. Piccola and members of the committee for the opportunity to testify on this important issue. I am Nathan Benefield, Director of Policy Research for the Commonwealth Foundation, a state public policy research and educational institution based in Harrisburg. I am joined today by Elizabeth Bryan, a Research Associate with the Foundation.
In the last two years, Gov. Rendell has used the “costing-out” study to justify proposed increases in state education subsidies. Yet while costing-out or “adequacy” studies have been conducted in over 35 states to quantify the amount of education funding needed, in no case did spending increases result in dramatic improvement of adequacy standards.
In this testimony I will focus on faulty assumptions that form the basis of costing out studies, the missing link between spending and performance, and conclude with some alternative education reforms that have produced significant results.
Flaws of Costing-Out Studies
Despite a cost to taxpayers of $650,000, the costing-out study’s results could have been predicted. The firm hired has produced costing-out or adequacy studies for dozens of states. The estimated “cost” to educate a child varies dramatically by state, not based on a cost of living index, but on what states were spending. One conclusion was consistent across the firm’s studies-each state needs to increase spending, typically by 20-40 percent!
These studies also offer a wide range of recommendations for the same function. For example, the firm recommended 63 instructional personnel per 1,000 pupils in Indiana, but nearly double that amount in Maryland. Regardless of what states were spending or their staff level, the Augenblick group’s recommendations are always the same-MORE!
The techniques used in costing-out studies suffer from methodological flaws. Augenblick, Palaich and Associates utilized all three of the most popular costing-out methodologies-“Successful Schools,” “Professional Judgment,” and “Evidence-Based” models.
The Successful Schools method evaluates school districts that currently have strong student outcomes. Many problems plague this approach-namely the inability to separate the success of the schools from the success of the students. Factors like family background have more of an effect on the achievement rates of individual students than school variables.
The costing-out study’s methods were based solely on what the state spent in traditional public school districts without first discounting wasteful, inefficient spending. More importantly, the study completely ignored charter and cyber schools (which receive only three-fourths of what school districts do) and private schools, which often spend half as much per-pupil as public schools.
This flaw alone should cause policymakers to question the study’s findings and recommendations. Indeed, no study based solely on a taxpayer-funded monopoly can truly measure what it should cost to educate a child.
Another method used in the Pennsylvania’s study is the Evidence-Based method, which relies solely on research chosen at the firm’s discretion. Often the information is handpicked to support the desired recommendation for increased funding. This method, like Professional Judgment, makes no use of actual district spending or performance.
Similarly, the Professional Judgment approach utilizes a panel of education experts to develop an ideal school and curriculum. There is an inherent conflict of interest in this approach, as those who stand to gain the most from additional funding are recommending spending increases. It’s not surprising that these panels often recommend cost estimates 30% above the Successful Schools method.
A major reason to question the costing-out study is that, not surprisingly, all of the spending recommendations are those lobbied for by the teachers’ unions and employees of the government school system (who sat on the study’s “professional judgment” panel).
These recommendations include smaller classes, government-run preschool, full-day kindergarten, and more teachers, counselors, support staff and school nurses (i.e., dues-paying union members). In reality, the overwhelming evidence suggests that large-scale preschool programs have a marginal effect on students’ academic performance, but an effect which disappears by 3rd grade-thus Head Start and “universal preschool” in states like Georgia and Oklahoma have had no measurable impact on performance. Similarly, the evidence indicates that smaller class sizes only have an impact on learning in the extremes.
At the same time, a number of reforms that would not benefit the special interests consulted during the study were left out. Ignored were reforms linked to academic improvement-merit pay for teachers, smaller schools, higher academic standards, and school choice.
In 1961, President Eisenhower cautioned against the “Military Industrial Complex”: special interests that profit from defense spending and yet have a disproportionate influence in setting defense policies. I would caution legislators against giving too much weight to the “educational industrial complex”-special interests that benefit from increased educational spending, regardless of the impact on student performance. I do not mean to denigrate teachers and education professionals engaged in a most admirable occupation. Rather, I wish to point out that teachers’ unions, school boards, school construction companies, and others have an interest in additional education spending, irrespective of the benefits for school children.
Rather than delineate this interest myself, I will quote outgoing National Education Association General Counsel Bob Chanin, who recently described the union thus:
And that brings me to my final, and most important, point. Which is why, at least in my opinion, NEA and its affiliates are such effective advocates. Despite what some among us would like to believe, it is not because of our creative ideas. It is not because of the merit of our positions. It is not because we care about children. And it is not because we have a vision of a great public school for every child. NEA and its affiliates are effective advocates because we have power. And we have power because there are more than 3.2 million people who are willing to pay us hundreds of millions of dollars in dues each year because they believe that we are the unions that can most effectively represent them, the unions that can protect their rights and advance their interests as education employees.
This is not to say that the concern of NEA and its affiliates with closing achievement gaps, reducing dropout rates, improving teacher quality, and the like are unimportant or inappropriate. To the contrary, these are the goals that guide the work we do. But they need not and must not be achieved at the expense of due process, employee rights, and collective bargaining. That simply is too high a price to pay!
When all is said and done, NEA and its affiliates must never lose sight of the fact that they are unions, and what unions do first and foremost is represent their members.
Or as Gov. Rendell aptly put it, while there are many lobbyists representing education organizations, “the one group that has no lobbyist here in Pennsylvania is our children.”
If you asked me how to improve the Commonwealth Foundation, I would naturally suggest more funding. My second suggestion would be to posit we need better pay for research staff. Of course, donors would be wise to question my recommendations, given my own self-interest.
However, when the Commonwealth Foundation seeks funding-as with any nonprofit organization or business-we must justify our value to those who have the option not to contribute (or buy a product), and we must compete with others also seeking funds (or making sales). This competition allows an evaluation of cost-effectiveness and value.
However, when school districts come to the General Assembly seeking more funding, it is often merely a question of how much more funding they should receive. And typically, this question is debated in a vacuum, as nothing outside the educational establishment is ever considered.
This is another major flaw of the costing-out study. By excluding any analysis of charter schools, public cyber schools, private and nonpublic schools, or homeschooling, the analysis misses out on what could be. The study ignores other options that could deliver better value and more “bang for the buck” in education and instead examines the status quo, and recommends more of the same.
Fallacy of More Spending
The myth put forward by the costing-out study and the educational industrial complex is that additional dollars spent on public school will increase performance. Yet the evidence indicates otherwise.
The 2007 costing-out study recommended increasing current expenditures from $9,512 to $12,057 per student. Overall, experts recommended an increase of $4.61 billion or a 26.8% increase. The number rises to $4.81 billion if those districts that already spend more than required by the costing-out estimates continue to do so.
Yet studies find no correlation between per-pupil expenditures and student achievement. A quick analysis of the latest per-pupil expenditures and PSSA scores by Pennsylvania school districts show almost zero correlation. Even for someone who thinks there is little relationship between spending and achievement, this complete lack of correlation is surprising, as students from wealthy households perform better academically, and the perception is wealthy districts spend more per pupil.
A recent study from the Math and Science Partnership of Greater Philadelphia on Pennsylvania school performance (specifically the 11th grade PSSAs) across school districts reports similar results. Namely, they find that demographic factors (residents with bachelor’s degrees, percentage of minority students, and percentage of low-income students) are correlated with academic performance. However, instructional expenditures per-pupil was not significantly correlated with academic performance and total spending per-pupil was negatively related to academic performance.
On a national level, there is also very little relationship between spending and related inputs and results. The American Legislative Exchange Council’s 2009 Report Card on American Education, comparing spending and performance across the 50 states, finds:
Surprisingly, the data show that academic achievement cannot be accounted for by any of the measures of public investment used in this study (pupil-teacher ratio, per pupil expenditures, teacher salaries, and funds received from the federal government), either singly or as a blend. This conclusion is borne out when variations in average SAT scores per state are tracked over the past two decades alongside changes in these measures of public investment. If anything, this statistical analysis demonstrates a positive, but weak, relationship between student success and percentage of federal funding, and pupil-teacher ratio – yet not in the manner one would anticipate.
The information shows that higher student scores on standardized tests correlate mildly with more pupils per teacher and less federal involvement with public school budgets.
This trend is evident in Pennsylvania. From 1980-81 to 2007-08, Pennsylvania taxpayers increased spending on public schools by $20 billion, or 400%. In per-pupil terms, after adjusting for inflation, Pennsylvania public school spending increased 125%, from $6,000 per pupil (in 2009 dollars) to almost $14,000. Yet performance has not doubled or even improved dramatically. SAT scores have remained stagnant over the past 30 years (see attached charts).
Let me diverge a bit from the main theme to address a related topic-Pennsylvania’s educational standards. Unfortunately, the PSSA exams fail to adequately inform parents, teachers, and the taxpaying public about the quality of their schools. Indeed, a comparison between Pennsylvania’s achievement test results with those reported by the National Assessment of Educational Progress (NAEP) shows that the PSSA significantly inflates the numbers of public school pupils who are deemed to be performing at or above grade level.
For instance, the PSSA found that 81% of 8th graders were proficient in reading, and 72% in math. But on the NAEP, only 36% and 40% reached proficiency, respectively.
In fact, the recent improvement in Pennsylvania school performance cited by the Department of Education is not reflected on the latest NAEP results. Much of the improvement in academic performance measured by the PSSA seems to be linked to the low-standards of the test, rather than real improvements in student learning, doing a disservice to both students and taxpayers.
Of course, the costing-out study acknowledges the lack of direct relationship between spending and performance, contending that it depends on how the money is spent. That is to say, additional funding must be spent “wisely.” Unfortunately, Pennsylvania school districts have a very poor record of spending money in such a manner.
A Commonwealth Foundation report, Edifice Complex: Where has all the Money Gone, notes that the fastest growing part of school district spending is in construction and debt. The latest Department of Education figures show that construction and debt spending grew 137% from 1996-97 to 2007-08, compared with 66% growth in instructional spending.
Furthermore, our analysis discovered that higher-spending districts spent a greater proportion of their funding on construction and debt. This relationship grows stronger when controlling for enrollment growth and district size. In other words, when given additional resources, many local school boards seem inclined to spend money on football fields and “Taj Mahal” buildings rather than student instruction.
Pennsylvania needs a better set of answers than simply suggesting more funding.
Weighted Student Funding
One positive aspect of the costing-out study is its emphasis on the per-pupil level of spending-weighing the needs of students-rather than funding for districts. Unfortunately, the state funding formula retains many of the protections for district interests. The funding formula developed was: (1) District funding in 2007-08 (no district will lose revenue, even if all its students leave) + (2) Four million dollars if the district is a Commonwealth Partnership District (this applies only to Pittsburgh Public Schools) + (3) A set percentage of State Target Funding, based on the “need” identified in the costing-out study, with higher weighting for districts with higher taxes; or a 1.5% increase on 2007-08 funding if the district spends more than the costing-out study said it needed to. This formula rewards lobbying of past years by protecting districts, while incentivizing higher local taxes.
A better formula would involve weighted student funding. This funding is awarded entirely on a per-pupil basis, with higher funding levels for low-income and special needs students, who cost more to educate. Weighted student funding is a way for school districts and state government to fund students instead of school buildings and staff. Nationally, at least fifteen large school districts use weighted student funding, and Philadelphia has been moving toward that model.
The Reason Foundation recently published an overview of weighted student funding, identifying benefits such as open enrollment/more choice for students, and greater autonomy for individual schools and principals and a guide to best practices.
Another way to improve Pennsylvania’s academic achievement lies in dealing with low performance throughout the educational system. Rather than simply testing students at the end of their academic career, Pennsylvania should end social promotion and ensure students learn the basics at earlier stages.
A good case study is the state of Florida, where an early detection system has been in effect since 1998. In Florida, students are tested on a yearly basis in grades 3-11. The state requires students to meet standards for promotion from grades 3 and 10, and some school districts use the tests as a requirement for promotion from other grades. The tests serve as a vital tool in measuring students’ achievement and recognizing those who need remedial assistance.
Pennsylvania also needs to do a better job of measuring, and reporting, student improvement, not just overall performance. Doing so will provide a better way to measure the quality of schools, teachers, and programs.
Pittsburgh City Schools recently received a grant from the Gates Foundation for implementation of an Empowering Effective Teachers Plan that contains a merit based pay provision. Merit pay has already been adopted in cities like Charlotte, Dallas, Denver, and Minneapolis.
Merit pay rewards high-quality teachers, helping to retain the best and the brightest. Merit pay also changes the incentives, motivating teachers to focus on improving student learning, rather than simply sticking around for annual bumps in pay, regardless of performance.
Repeal Prevailing Wage Laws
As mentioned previously, the fastest growing cost for school districts is construction. One of the cost drivers is the state’s Prevailing Wage law, which mandates state-set wages for all public construction contracts. These wages are, on average, 37% higher than the average market wages in the state and increase the cost of construction by 20% or more.
Sen. Mike Brubaker has introduced S.B. 695, proposing a three-year moratorium on the application of Pennsylvania’s prevailing wage law to school construction projects. As school districts spent $2.2 billion in construction in 2007, the savings would exceed $400 million-more when considering construction is usually debt financed, with costs paid with interest over many years.
If increasing spending will not improve education, what will? Greater school choice.
Since the mid 1990s, Pennsylvania has seen the introduction of several innovations that are making a real difference in how our students are educated. The growth of charter schools, and more recently of cyber charter schools, has allowed numerous families to find another non-traditional public school that better meets their educational needs.
And, since 2001, Pennsylvania’s Educational Improvement Tax Credit (EITC) has been providing scholarships to tens of thousands of students to attend a non-public school that is a better fit for them. We were greatly disappointed to see the EITC reduced in the budget, even as funding for public schools was increased by more than the entire cost of the EITC; this seems to indicate a higher value placed on some students than others.
A recent study by Stanford economist Caroline Hoxby found that charter school students outperform their peers, even those who unsuccessfully applied for admission, and that charter-school students tend to come from poor and disadvantaged families, yet scored almost as well as students in the affluent suburbs.
The same is true of students served by EITC scholarships; the average income of families served is $29,000, according to a Legislative Budget and Finance Committee report. Other recent studies show that vouchers in Milwaukee and Washington, DC have improved student learning, and a number of studies show that public schools improve with greater choice.
All the school choice options that exist in Pennsylvania-charter schools, cyber schools, and the EITC-serve a growing number of students, and most have long waiting lists.
Not only is school choice leading to real academic improvements, it’s doing so at a lower cost. Charter schools on average receive about 80% of what traditional schools spend. Per-pupil spending in district-run schools was over $13,000 in 2007-08, compared with under $11,000 in charter schools and $9,000 in cyber charter schools. The average scholarship in the Educational Improvement Tax Credit is just over $1,000.
Specific proposals we support to expand school choice include:
- Improving the state charter school law to allow alternative certifiers, such as universities.
- Allowing inter-district choice, with funding that follows the child, and eliminating the assignment of school by zip code.
- Expanding the EITC limit, and allowing individuals to take the credit.
- A parental choice tax credit-a tax credit toward educational expenses of parents, including private school tuition and homeschool costs.
- The Property Tax Relief Scholarship Act-the Commonwealth Foundation’s proposal to use a new funding stream (like gaming money) to provide scholarships to low-income families and property tax relief for homeowners.
While higher standards and an end to social promotion will help to improve the quality of education in Pennsylvania, real accountability comes when parents choose the best school for their children. Competition will provide schools with incentives to achieve higher standards. Implementing choice in the education system is the key to raising standards and bringing value back to high school diplomas. When parents choose, students win.
Pennsylvania has been trying to improve its education system with more spending for many decades. It’s time for another approach.
I thank you for the opportunity to present, and we look forward to any questions you may have.
# # #
EDITOR’S NOTE: For charts, see More Info Tab or PDF Version.