Money and Schools
Debating Ben Spielberg 50 Years After the Coleman Report
Is “more money” a vital education policy, when compared with other possible changes? Should taxpayers allocate significantly more money to existing K-12 public schools, without demanding structural reforms? On average, if existing K-12 public schools had more money, would students obtain significant or sustainable benefits?
Increasingly, conventional wisdom answers “yes:” many say that money alone, even without reform, helps students. Matt Barnum, policy editor of education website The 74 Million, posted in April: “At this point there’s a large body of evidence that more $ leads to better outcomes,” linking to a February summary of recent case study research. Nick Albares, a policy analyst with the Center on Budget and Policy Priorities, wrote in January’s Education Post: “common sense and research suggests [that] money matters for long-term outcomes.” In response to my 2015 column contra in Dropout Nation, the analyst Ben Spielberg wrote a sharp dispute.
In the spirit of the Education Post mantra of “better conversation, better education,” Spielberg and I dedicated several hours researching each other’s claims and meeting in person to develop a common fact base and framework. This column represents my reflections on that effort.
My claims and sources
I claim that America should not spend more money on public schools without significant structural reforms. On average, such extra money will not lead to significant or sustainable improvements for students. Instead, the extra money would mostly strengthen anti-reform forces, leading to ultimately worse results for students.
This claim relies on several primary sources.
First, the empirical link between money and school quality in the United States starts with The Coleman Report, a 737-page report published 50 years ago under the auspices of the 1964 Civil Rights Act. The Coleman Report was so dense that Daniel Patrick Moynihan convened a yearlong seminar of 80 scholars to study the report’s implications. One of those scholars was a graduate student named Eric Hanushek, who is now a Professor at Stanford University. Fifty years later, Hanushek wrote a recent review of the post-Coleman research. He notes that America is still failing African American students: “if we continue to close [racial achievement] gaps at the same rate in the future, it will be roughly two and a half centuries before the black-white math gap closes and over one and a half centuries until the reading gap closes.” Hanushek also posted a scatterplot of major funding surges, and writes that: “There is no correlation between changes in real per-pupil spending over the past quarter century and changes in 4th-grade student achievement in reading, providing no reason to believe that increasing school spending will by itself boost student achievement.” Hanushek also summarizes the “general consensus” that “how money is spent is much more important than how much is spent.” In his words, “the research does not show that money never matters or that money cannot matter. But, just providing more funds to a typical school district without any change in incentives and operating rules is unlikely to lead to systematic improvements in student outcomes. That is what Coleman found, and that is what recent research says.”
Based on this, education reformers argue that more resources should be allocated to schools only in concert with major improvements in parent choice and teacher professionalism, as happened in Washington, D.C., over the past decade.
Second, to better understand patterns in school spending in different countries and over time, a great starting point is “How the World’s Best-Performing School Systems Come Out On Top” by Dr. Mona Mourshed and Sir Michael Barber. Numerous researchers contributed to this report, using data from UNESCO and OECD, interviews with prominent sources throughout Asia and Europe, and scholarly inputs from Professor Ludger Wößmann of the University of Munich and Professor Lant Pritchett of Harvard. After reviewing high performing and rapidly improving countries, states, and cities, the report noted that although the United States and many other school systems have substantially increased public expenditures on K-12 schooling over the past several decades, student performance has generally remained flat or declined.
Third, to understand why the money has not helped students, it is important to study political science. The best sources include Bureaucracy by the late James Q. Wilson, former President of the American Political Science Association; Government’s End by Brookings Senior Fellow Jon Rauch; Special Interest by Stanford University Professor Terry Moe; and my own research on how teachers’ union leaders influence state legislatures. This research interest came from my personal history as the child of a public school teacher, which led me to spend several decades studying education politics and policy. The more money society adds to a system that does not respond to students, the more powerful vested interests become, damaging student interests in the long term.
Factual clarifications and corrections:
following up from Ben Spielberg
Conclusions about spending rest upon many baseline facts, and those facts merit a robust debate. To that end, after Spielberg posted his rebuttal of my Dropout Nation column, I responded, and then he responded back. We had several additional exchanges, and then sat down to discuss. To promote discussion, I agreed to post a review of several factual claims, offering concessions, stipulations, and clarifications.
(1) What spending counts? Spielberg is correct that the best benchmark for educational spending is public K-12 spending, rather than private spending or spending on tertiary education. Consistent with some of the popular press on cross-national comparisons, I had previously used broader statistics in my Dropout Nation columns, including private spending and tertiary spending in the United States vs. other nations. Since these categories are relatively larger in the United States than in other nations, I agree that this was inapt on my part. It would be more germane to exclude those categories. Going forward, I will use K-12 public expenditures in the United States where possible.
(2) How much does the United States spend? Since education has natural economies of scale (in innovation, purchasing, and content, for instance), it matters how much we spend in the aggregate. In other words, before we get to per capita spending, we have to look at the total budgets. According to the National Center for Education Statistics, public elementary and secondary school expenditures in the United States amounted to $621 billion in 2011–12. (This source gives several thousand dollars less per pupil than the fifteen thousand per pupil when private and tertiary budgets are included.) The aggregate $621 billion figure is the source of my Dropout Nation statement that: “America spends more than any society in history.”
Spielberg argues that the relevant question is how much the United States spends as a percentage of GDP. This metric is widely used, but silly. If the economy collapses, education spending as a proportion of GDP would go up because the denominator would fall. As I wrote earlier, by this argument, “the fastest way to get school spending right is to tank the economy.” To be sure, a higher GDP allows more spending, but the metric itself tells us nothing about whether more money is necessary or would be well spent.
To be fair, using public expenditures alone, the United States is not the global spending leader on a per pupil basis. The United States has nearly 50 million students, so the total education budget gets divided into roughly twelve thousand dollars per pupil. Three small countries — Luxembourg, Norway, and Switzerland — have higher per pupil spending, as shown here in Table B3.3. That said, these small and rich countries have fewer economies of scale, as their populations are 0.2%, 2.6%, and 1.6% of that of the United States. More to the point, the U.S.A. ranks fourth in per-pupil spending out of 34 industrialized countries, spending 23% more per pupil than top-performing Finland; 25% more than the OECD average; and 40% more than top-performing South Korea.
(3) How much has spending changed over time? Spielberg agrees that per-pupil spending in real dollars has increased over time, by large amounts (roughly four-fold since 1954, roughly 2.5x since 1970, roughly 75% since 1980).
Since these are real dollars, this means that resources per student have increased enormously over time. To this, Spielberg makes two responses.
First, Spielberg claims that “real spending should increase as the economy grows” and that “real spending on practically everything has increased in dollar terms since 1970.” I initially interpreted his statements as a claim that real spending necessarily increases as the economy grows, which would be, as I noted, “jarringly” false. Upon discussion with him, I learned that he meant only that as societies become wealthier, they spend more on most categories of goods. This argument is not jarringly false, but it strikes me as irrelevant to the question of whether the money is well spent. For instance, societies often spend more on meat as they become wealthy, but they become less rather than more healthy as a result. Also, although Spielberg’s broad point seems right (societies with more money spend more money on more things), the exceptions beg the question. In the private sector, whole categories of spending have declined, such as residential phone services. In the public sector, defense spending has remained roughly flat since 1970 despite a population growth of nearly 50%, and has declined significantly as a percent of GDP. Why couldn’t education be one of those services where innovation increased productivity per dollar over time?
Second, Spielberg claims a counterfactual — namely, student achievement might have declined if we hadn’t spent all this extra money. That claim defies belief. Since 1970, public budgets for schools have also skyrocketed in Japan, Australia, Italy, Britain, and France, without measurable improvements in student performance. Unless the world’s students suddenly became much harder to teach across diverse societies, the more likely explanation is that the public systems spent the money poorly.
(4) What is the evidence from charters? I note that charter schools spend less than traditional public schools on average, and deliver better results for students on average. To me, this serves as a third major data point showing that schools can do more with current budgets (the first two points were cross-country comparisons and historical comparisons). In response, Spielberg raises many issues about the diversity of the charter sector, the size of charter gains, and comparisons of charter gains vs. gains in some of the case studies of increased spending.
Since I spent thousands of words going back and forth with Mark Weber (a.k.a. “Jersey Jazzman”) about charter schools, I will be brief vis-à-vis Spielberg. The 2015 Stanford CREDO study of urban charter schools strikes me as robust and significant evidence that charters deliver better results for many sub-populations of students in need, without spending more money. Research such as that of Sue Dynarski in Boston shows that such superior results go beyond test scores, but it’s the national work of CREDO that conveys average sector-wide results. This is the major difference between the case study anecdotes that Spielberg, Barnum, and others report. As Hanushek writes, money sometimes matters, but sometimes it doesn’t. The aggregate experience of decades of spending increases across many different cities, states, and countries is that, on average, massive spending increases have not improved educational performance. Systematic improvements, such as we see in the charter sector, and as we’ve seen in some countries, provide clues that money must be coupled with real structural reforms, or be wasted (or worse) in terms of student results.