AMERICAN.COM

While for-profit educational institutions have come under intense scrutiny from the Senate for their recruiting practices and the amount of money they spend on advertising, private not-for-profit and public institutions have so far escaped such scrutiny. One reason is that the Congress has legislative authority to regulate “career colleges” and this has been interpreted to cover all for-profits, even if many now offer a wide range of majors not tied specifically to careers, and even as many public institutions, especially community colleges, are becoming ever more career-oriented (see here).

Setting that issue aside, when comparisons are made of the costs of recruiting and advertising practices, one large expense incurred by the private not-for-profits is usually left out of the calculation: “institutional aid,” the discounts that students get from the posted tuition (or “sticker price”) a campus nominally charges (institutional aid by publics is relatively small since they don’t have large endowments and their tuition is relatively low). There have long been fights over the disclosure of the details of how these institutional grants are awarded and to whom, but colleges do report to the Department of Education the percent of students who receive institutional grants and the average award. For students, the problem is evident: while averages are reported, a prospective student has no way of knowing what students like her receive. This information asymmetry clearly works to the advantage of the colleges, who use institutional aid to create an arcane system of differential pricing and yield management that would make airlines proud. But with available data we can revisit the costs that not-for-profits spend on recruiting students.

Let’s begin with estimates from the National Association for College Admission Counseling (NACAC) estimates what it calls “costs to recruit.” According to its 2009 report, these costs for private and for public colleges are:

However, this estimate for private colleges and universities is far too low, since it does not include institutional aid—a major recruiting tool.

For the 900 or so four-year, degree-granting, private not-for-profit institutions that reported data to the Integrated Postsecondary Education Data System (IPEDS), over the last five years 77 percent of students, on average, received aid from the college they attended. Moreover, the average level of this aid has increased dramatically: from around $8,000 per student in Academic Year 2003-2004 to over $10,000 in 2007-2008, the last year for which IPEDS data are currently available.

Depending on how you consider institutional aid—particularly how much of this aid you count as recruiting expenses—the costs to recruit in private not-for-profits could be quite steep—the $3,000 counted by NACAC plus the $10,000 per student in institutional aid.

There is another cost that these schools incur that is not often considered.

Across these not-for-profit colleges, more than 25 percent of students who enroll for a first year do not, on average, return for as second. The amount of money each school spends on students who do not return after year one can be quite high—and as the size of institutional grants grow, unless the retention rate goes up (which it hasn’t), the losses campuses incur will continue to grow. As is evident in the figure below, the average not for-profit institution of higher education is now spending about $750,000 per year on institutional grants to first-year students who do not return after a year. (The societal losses for first-year attrition are even higher, since these students have also received federal grants and often state grants, and they and their families have also incurred remaining tuition and related expenses.)

Our state of knowledge about what works to improve student success in college is pathetic and the federal government, which makes a huge investment in student success, especially through Pell Grants, has not stepped up to the plate on this (for example, the Institute for Education Sciences in the Department of Education funds 16 R&D centers, only one of which has a focus on postsecondary education).

But we do know that students, who spend an incredible amount of time applying for college, are often unprepared for what happens when they enroll. Among the major reasons for student failure during the first year are: having too much fun (but also its obverse, social isolation), poor academic preparation, and running out of money (see here or here). While we don’t know much about improving student success, one strategy that seems successful, especially for low-income, first-generation students (both risk factors for dropping out), is a combination of summer boot camp and “university 101”—usually a semester-long “frosh” seminar that covers the basics of campus life and allows students to develop the skills and friendships that are necessary for successfully negotiating campus life.

Seems like running these kinds of courses would be a wiser use of institutional money than burning $750,000 per year on students who don’t make it through their first year.

In my first three posts, I noted the growing importance of the Programme for the International Assessment of Student Achievement (PISA) and the Organization for Economic Cooperation and Development (OECD) in discussions of American education policy. The stakes for how seriously we take PISA and OECD’s policy pronouncements are high and, for a number of reasons, growing.

Most importantly, as the Obama administration and the Common Core Standards Initiative continue to seek to internationally benchmark our standards and assessments, PISA will become an even more prominent metric. And since OECD uses PISA as more than an assessment, its policy recommendations will become even more visible.

Moreover, OECD’s education advice is about to become even harder to avoid. OECD has just announced it is establishing a new program called “Leveraging Knowledge for Better Education Policy,” directed by the ubiquitous PISA head, Andreas Schleicher. This project is also referred to as the “GPS” project: yes, a “global positioning system” for education.

Continue reading this post here.

Image by ccarlstead.

In my last post, I noted how Andreas Schleicher, the head of the Programme for the OECD’s International Assessment of Student Achievement (PISA) project, has highlighted just how much the United States would gain if only it matched Poland’s growth on PISA. He testified before the Senate HELP committee that if the United States matched Poland’s performance on PISA, the United States would have gained “over $40 trillion in additional national income.” This ignored the gains our students have made on the National Assessment of Educational Progress (NAEP), often exceeding the gain in performance among Polish students—but since NAEP is our national assessment and not OECD’s, these improved skills selectively escaped notice.

Embedded in OECD’s argument is the assertion that countries that perform well on the PISA assessment will grow faster economically than low-performing ones. In a recent OECD report on “The High Cost of Low Educational Performance,” we are told that a “modest goal of having all OECD countries boost their average PISA scores by 25 points over the next 20 years … implies an aggregate gain of OECD GDP of USD 115 trillion over the lifetime of the generation born in 2010.” And if we can only bring all countries up the level of Finland, OECD’s best-performing education system in PISA, the world would gain USD 260 trillion.

Note that one of the funny traps that OECD falls into when extolling the virtues of PISA is conflating high performance on PISA with the cognitive skills that PISA supposedly measures: it’s as if just scoring higher on PISA will generate magical returns to a nation and the world. Indeed, the subtitle to the OECD report, “the long-run economic impact of improving PISA outcomes,” makes this point crystal clear (and helps explain why our students’ gains on NAEP don’t deserve mention).

Continue reading this post here.

Every author dreams of coining a phrase that becomes so much a part of accepted wisdom that it drives the way people think and the way governments shape policies. Among the most powerful is Tom Friedman’s “The World is Flat.” While some argue that the world is far from flat (see here and here), the term has clearly reshaped people’s consciousness. In the field of education, it has helped feed an emphasis on “international benchmarking” and the rising prominence of the Organization for Economic Development and Cooperation (OECD) in discussions of American education policy.

A key example is the “Common Core Standards Initiative” (CCSI) that governors and school chiefs in almost every state and the District of Columbia have signed on to. This initiative is creating a set of common curriculum standards for math and English language arts in grades K-12 that states can voluntarily adopt. Led by the National Governors Association and the Council of Chief State School Officers (CCSSO), the CCSI has the clear support of the Obama administration, which has made state sign-on to these standards a part of any successful strategy for winning Race to the Top money and possibly even future Title I money.

Continue reading this post here.

Yesterday afternoon, the American Enterprise Institute hosted a lively panel in which well-respected education historian Diane Ravitch discussed her new book, The Death and Life of the Great American School System.

Diane is arguably the best educational historian working today and one of the best the nation has ever produced. Chapter after chapter she confirms what we all know about education policy and practice—it is relentlessly based on fads built on the flimsiest of evidence. Diane shows that good ideas are often taken to scale without any thought about how any of reforms might work in a larger venue. She shows that ideas often become invested with magic properties so that people see them as a silver bullet that will cure all our ills.

But while her analysis is often spot-on, she also makes mistakes. And perhaps her most consistent ones pertain to choice and charter schools.

(Continue reading this post.)

As the Obama administration gears up for reauthorizing the Elementary and Secondary Education Act (ESEA), the administration is making its case for an array of major revisions. We already see Secretary of Education Arne Duncan marshaling evidence to make the case for change. Thus far, he’s earning wildly mixed grades for the quality of his team’s homework. Secretary Duncan deserves a thumbs down for his efforts to placate No Child Left Behind critics by blaming the current version of ESEA (a.k.a. NCLB) for a supposed narrowing of the K-12 curriculum. He deserves much higher marks for his forthright charge that states are “dumbing down” their proficiency standards.

Claims about the narrowing of the curriculum need to specify exactly which students have supposedly experienced this change. NCLB was focused mostly on the earlier grades, but it also mandates that students be tested in high school. While there is some evidence about curriculum narrowing in elementary schools (see here, but also see here), the evidence about high school curricula is quite clear and quite to the contrary.

In contrast to the earlier grade levels, which often focus on basic skills, high school years provide students with the opportunity to develop more deeply their skills and knowledge of subjects such as music, arts, and social studies, all supposed victims of a narrower curriculum. Fortunately, we have good data on what subjects high school students study—data gathered as part of the High School Transcript Study conducted by National Assessment of Educational Progress (NAEP). These data document which courses students actually took, as compared to what parents, school administrators, and others believe happened to the curriculum. Secretary Duncan has used these statements of belief as support for his concern about narrowing the curriculum (see here); but do these beliefs jibe with reality?

Transcript data show that, rather than taking a narrower curriculum, America’s high school students are taking more courses across a broader array of subject areas than ever before. As evident in Figure 1 above, the number of English credits high school graduates completed was unchanged between 2000 (pre-NCLB) and 2005, while the number of math credits increased. Meanwhile, student exposure to science (not mandated under NCLB until 2007) and social studies increased, and there was no change in the average number of foreign language and fine arts credits completed. Things may have radically changed since 2005 when the data were collected, but transcripts show that student exposure to arts, science, social studies, and foreign language, subjects supposedly being crowded out, either went up or remained the same.

Above, Figure 2 displays a longer view of high school graduates’ exposure to social studies. Here we see that at each point over 15 years, high school graduates have had the same exposure to U.S. history and government/civics. In addition, we see a substantial increase in the percentage of high school graduates completing credits in world history and world geography. Here too there is little evidence of crowding out—in fact, we see an increase in exposure to the world beyond the borders of the United States.

In short, there were increases or at least stability across a wide range of subject areas, with no signs of narrowing of the high school curriculum. This was due, in part, to high school graduates increasing the number of credits they have completed (from 25 in 1998 to 26 in 2000 and 27 in 2005) and by taking more academic courses.

Claims about narrowing of the curriculum are already common and will become more so as the reauthorization debates proceed. When these discussions move into high gear, it is important to specify which level of schooling the speaker has in mind, and it is even more important to have better and more recent empirical evidence to support such claims.

It would also be far more productive if the discussion focused on why, given this increase in academic course taking by high school students, we haven’t seen any improvements in student achievement as measured by assessments such as NAEP’s Long Term Trends or the Program for International Student Assessment. Similarly, in light of the recent disappointing fourth and eighth grade NAEP math results, a better question would be: if our elementary schools have indeed focused so much more time on math, why haven’t our students gained more in return? In short, the flimsy evidence regarding narrowing of the curriculum and the continuing poor performance of our students does not justify rethinking the law’s focus on math and reading.

Instead of spending time on much-discussed but poorly documented complaints, like curriculum-narrowing, Secretary Duncan’s reauthorization agenda should focus on better-documented issues such as the laughable “proficiency” standards that states have developed to fulfill NCLB’s accountability mandate. Fortunately, the secretary has acknowledged that low state standards are a major flaw, arguing that “the biggest problem with NCLB is that it doesn’t encourage high learning standards. In fact, it inadvertently encourages states to lower them. The net effect is that we are lying to children and parents by telling kids they are succeeding when, in fact, they are not.”

The evidence behind this claim, especially National Center for Education Statistics studies that show many states set their proficiency standards at or below NAEP’s basic standard, is far more compelling than the evidence about narrowing the curriculum. Indeed, this evidence justifies wiping clean NCLB’s compromise that set national proficiency goals but allowed each state to define its own level of proficiency and to write their own tests to judge progress.

On November 23, President Obama announced a new initiative called “Educate to Innovate,” which is designed to increase the interest of “young people” in Science, Technology, Engineering, and Math (STEM). The president described his plan to “harness the power of media, interactive games, hands-on learning” and to host high-visibility events, such as a White House science fair. Linked to the “Race to the Top” (RTT) initiative, which is aimed more broadly at improving education, this new initiative creates a two-pronged effort for building the future STEM workforce.

The new initiative is a good idea, and linking it to RTT is essential. To the extent that Educate to Innovate works, more American students will enter the STEM pipeline while RTT seeks to fix the now-broken pipeline. In theory, these two initiatives will work together to ensure that as more students become interested in STEM they can develop the needed skills.

While we can imagine that the high visibility of the Educate to Innovate program might be effective, it is more difficult to imagine that the RTT program, which after all is a short-term infusion of money, will lead to systematic, long-term, and sustained improvements in the STEM pipeline.

There are many reasons for concern:

First, as many have noted, the actual math and science skills of American K-12 students are lower than those of almost all our competitors. This is evident in how poorly our students fare in TIMSS and PISA, the two largest international math and science assessments. While our fourth-grade students have made strong gains over the last 15 years on the National Assessment of Education Progress (NAEP), in the most recent assessment, in 2009, scores were flat compared to 2007. Gains by our eighth-grade students have been far slower. On Long Term NAEP, our 17-year-old students score the same as their counterparts of the early and mid-1970s.

The lack of student skills traces back, at least in part, to a uniquely American (and uniquely bad) math curriculum, which differs from the curricula of almost every other country that outperforms us. Our curriculum is “a mile wide and an inch deep”—contributing to our students’ failure to develop cumulative math skills. In addition, math and science teachers in earlier grades far too often have deficient skills and lack the knowledge and love of these subjects.

If we don’t fix the curriculum and upgrade the teacher workforce, any increase in student interest in STEM becomes meaningless. Educate to Innovate could simply produce more students who think they can do STEM and then discover that mastering a video game (“Oh my god, I’m at level 27 of that cool STEM video game, so I must be a scientist!”) does not translate into mastering the universe of calculus or biology.

Finally, both RTT and the Educate to Innovate initiative miss the mark on one basic point: if our goal is to increase the STEM workforce, then we must remember that a STEM worker is not a 17-year-old high-school student who has taken an additional math or science course and professes an interest in science or math.

STEM workers usually hold degrees or certificates in science, technology, engineering, or math from post-secondary institutions. Again the signs portend ill. Large numbers of American students who enter community colleges must take remedial math, and a smaller but not insignificant number of students entering four-year institutions start with similar deficiencies. Many students with these deficiencies never complete their programs.

Further, far too many post-secondary institutions are not in the business of “growing” STEM students, but instead are in the business of “sorting and discarding” students who enter with an interest in STEM. In many institutions, calculus, organic chemistry, and similar courses are designed as “wash courses,” with the explicit task of ending the science or math careers of large numbers of students. These courses are taught in large lecture halls and coupled with recitation sections often lead by foreign graduate students in a system designed to discourage, rather than encourage, students. Many who start off with an interest in STEM get grades of C or below in these courses, effectively ending their STEM studies and careers. Far too many would-be physical scientists end up as political scientists.

Certainly, we don’t want engineers unable to perform calculus, but our colleges and universities have the responsibility to teach students who are interested in STEM, using better teaching techniques. If Educate to Innovate works as intended and more students actually start college interested in STEM, colleges and universities must be ready to properly educate them.

The administration’s new initiative may fatten the STEM pipeline, but today that pipe is broken in many places. The Race to the Top initiative is a short-term funding stream that is likely to be forgotten before today’s young people who may be exposed to the new Educate to Innovate initiative have completed the five, ten, or more years it takes to turn a young student into a STEM worker. We need more consistent long-term commitment to improving the quality of America’s teachers and curriculum rather than new flashy initiatives.

This morning the National Center for Education Statistics released the 2009 math scores from the National Assessment of Educational Progress (NAEP). The results are bad news for the nation, but even worse for those who want to hold firm in the looming reauthorization debate over the No Child Left Behind Act (NCLB).

Signed into law in January 2002, NCLB formed the foundation of the Bush administration’s education policy. In the face of bitter resistance from teachers and from critics who thought the law was poorly designed and unrealistic, the Bush administration mounted one consistent defense: math scores on the NAEP, the nation’s report card, were increasing, especially among black and Hispanic students and among the nation’s lowest performing students. While critics questioned whether the post-NCLB gains actually marked an improvement over the pre-NCLB trend, defenders responded that it would take time for NCLB’s reforms to gain traction, and that gains would accelerate over time. This is why the results of the 2009 assessment are so important. How did the nation’s students do in the NCLB era?

At the fourth grade level, student scores were unchanged since 2007. This is unprecedented—the NAEP math assessment has been given eight times since 1990, and this is the first time that scores did not increase. At the eighth grade level, scores continued their trend of slowly increasing, up 2 points since 2007. (Between 2003 and 2005, scores increased by 1 point, and between 2005 and 2007 by 2 points.)

This certainly is not good news for the proponents of NCLB. But if one takes a longer historical view, the news buried in the report is even worse.

The 2003 NAEP math assessment was the first assessment after NCLB’s enactment. The six-year post-NCLB period between 2003 and 2009 can be matched almost exactly in length by the seven years between the 1996 NAEP math test and 2003. The simple comparison of pre- and post-NCLB scores is bad for NCLB as shown in Figures 1 and 2.

Figure 1 shows the pattern for fourth grade students, graphing the size of the gains overall and for each of the student groups that NCLB was specifically designed to help: low-performing students, black students, and Hispanic students.

In each case, we see that the pre-NCLB gains were greater than the post-NCLB gains, sometimes substantially. For example, among the lowest-performing students in the nation (those scoring in the bottom 10 percent), scores between 1996 and 2003 increased by 15 points. In the NCLB years, they increased by only 5 points. Gains among Hispanic and black students were also far lower during the post-NCLB period.

Figure 2 shows the results for eighth grade. The gaps between the gains in the pre-NCLB versus post-NCLB period are much smaller than for fourth grade, but for each group the gains were lower after 2003.

There are many possible explanations for this pattern—and unfortunately it’s impossible to determine which one is correct. I believe that one of the most obvious flaws in the law, which contributed substantially toward this disappointing result, was calling for all students to be proficient in reading and math by the year 2014, while allowing states to develop their own proficiency standards and assessments. The results, in hindsight, were predictable yet unintended—states had strong incentives to adopt low standards, maybe even lower than they would have adopted without NCLB.

In 2007, when I was commissioner of the National Center for Education Statistics, we showed that most states were setting their proficiency standards at NAEP’s basic level and some states even lower than that. Secretary of Education Arne Duncan has called this “lying to children and their parents because states have dumbed down their standards.” The irony here is that NCLB was built on a strong state standards and accountability movement but may have actually served to undermine the movement’s goals. The work of the National Governor’s Association and the Council of Chief State School Officers on common core state standards is particularly important in rectifying this mistake.

Others offer their own reasons for the failure of NCLB—ranging from underfunding, to maligning teachers, to offering too much choice, to… The list goes on and on. Unfortunately, we have too many possible explanations for far too little data—but the bottom line is clear: NCLB has not worked the way it was intended and the nation is worse off because of it.

The reauthorization of NCLB was going to be contentious to begin with, now these newest math results will serve to intensify the debate.

On average, four-year colleges graduate fewer than 60 percent of their freshmen within six years, and at many institutions, graduation rates are far worse. This is one of the findings of a new AEI study released this week that I coauthored, found here.

In the report, we rank colleges and universities according to their graduation rates, as reported to the U.S. Education Department, and their admissions selectivity, as reported in Barron’s Profiles of American Colleges. (This method ensures comparisons among like institutions.) And the disparities in graduation rates among institutions with similar admission standards are shocking. For example, James Madison University and the University of Louisville are both public schools ranked “very competitive,” but James Madison manages to graduate 81 percent of its students within six years, while Louisville graduates 44 percent. That is, a student at James Madison has double the odds of graduating than a student at the University of Louisville. This is just one example of many.

My coauthors and I are not arguing that colleges should simply give diplomas to students indiscriminately. However, few parents and even fewer students are aware of these statistics. High school guidance counselors most of all—those who guide student decisions—should make use of these data to help turn high school graduates into college graduates. You can find my presentation and the related report materials here.