The Use and Misuse of Data on earnings by Undergraduate Major

From Inside Higher Education: Here’s my take on how students and families misunderstand and misuse the increasingly good information on lifetime earnings differences across academic majors.

New Book Announcement from Oxford University Press

Oxford has announced our new book, titled “The Road Ahead for America’s Colleges and Universities.” It’s available for pre-order.

New Book

Our new book will be titled, “The Road Ahead for America’s Colleges and Universities.” Oxford University Press is again the publisher. The book will be available in July of 2017. Here is the Table of Contents:

Part 1 Introduction

  1. The Rhetoric of Higher Education in Crisis
  2. The Diverse US Higher Education System
  3. The College Bundle

Part 2 Threats

  1. Internal Threat I – The Rising Costs of Higher Education
  2. Internal Threat II – Will Students Keep Coming
  3. Environmental Threat I – Stagnating Family Income
  4. Environmental Threat II – Public Disinvestment
  5. Technological Threat – The Online Revolution

Part 3 Summary and Policy

  1. The Enduring Bundle
  2. The Evolutionary Future
  3. The Road Ahead – Policy Options


Our View of what the Candidates are Saying about Higher Education

From The Conversation

Why Debt-Free College Will Not Solve the Problems of American Higher Education


Cato’s Regulation Magazine publishes our take on federal aid

Does Federal Aid Drive College Tuition?

Why Bernie Sanders’s Free College Plan Doesn’t Make Sense

We weigh in on the contentious issue of free in-state tuition in the Washington Post’s higher education blog:




Tuition-free public education is music to the ears of families who see state university sticker prices continually rising more rapidly than their income. Presidential candidate Sen. Bernie Sanders (Vt.) has tapped that sentiment with his proposal to use federal funds to make public college education free for in-state students.

His Senate bill 1373 makes for great sound bites on the campaign trail, but there are good reasons why none of his Senate colleagues have offered any support.

The political obstacles to rewriting the federal role in funding state-supported universities are formidable, and the social case for preferring public institutions to private nonprofit schools is not at all clear.

Sanders proposes using federal tax revenues to pay two-thirds of the current in-state tuition if the state will contribute the other third from increased state appropriations. A quick look at current in-state tuition across the United States shows why this part of the revolution is unlikely to win congressional approval.

In 2013-14, the average level of in-state tuition at the nation’s four-year public universities was $8,312. Some states have much higher average tuition because their higher education appropriation per student is low. Families in these high-tuition states stand to receive a benefit that is much larger than average while families in other states will see a much smaller benefit. In Vermont, the average tuition payer would get a $14,000 break. By contrast, a family in North Carolina only gets a $6,500 benefit.

Vermont is a net winner since the per-student benefit far exceeds Vermont’s per-student share of the national tax base needed to fund the program. North Carolina, by contrast, gets a benefit that is less than its per-student share of the cost.

To see which states are net winners per student and which are net losers, we subtract the average in-state tuition for the United States as a whole from each state’s average in-state tuition.

Vermont is indeed one of the big winners. The tuition break for in-state students is $5,640 larger than average, and two-thirds of that number ($3,760) is a net gain in federal revenue per in-state student enrolled in Vermont.

In North Carolina, by contrast, the difference is – $1,734.

On a per-student basis, states like Wyoming, New York and North Carolina would see part of their share of the national tax base go to pay the tuition for students in Vermont, Massachusetts and California.

Members of Congress from the losing states will not be amused.

The states that are big net winners at the start of the program are, in many cases, those that currently underfund higher education. The net losers are often states that currently do invest in higher education. In a sense, the Sanders policy rewards bad behavior at the state level.

Since states must contribute one-third of the revenue needed to make tuition disappear, all will have to prioritize higher education more than they currently do. This will require tax increases, or it will force states to move existing resources into higher education and away from other state priorities like health care, prisons, roads and K-12 education. Many states may be unwilling to do either of these, so a higher education makeover along the lines of what Sanders proposes will die a quick death in many state capitols.

The states that are the biggest net winners are the ones that will have to dig deepest into their own tax base to pump up higher education spending. Yet they are also the states that have historically chosen not to do this. Sanders’s proposal is an attempt to change state budget priorities, but the magnitude of the budget adjustment may induce many of these “winner” states to oppose the idea as well.
Net Winners and Losers

Net Winners and Losers

No state is required to participate, but that is unlikely to mute opposition to the idea in Congress. A state that opts out still sees its share of the federal revenue base redirected toward other states.

Existing federal aid like the Pell grant and Stafford loan programs are directed at individual students who are free to choose their path through the higher education system. The money follows them to large universities or small colleges, to public institutions or to private ones, to historically black colleges and universities or to colleges with a religious tradition.

The Sanders proposal leaves out roughly half of the nation’s colleges and universities while adding a complex new set of federal rules for micromanaging the affairs of states and universities alike.

There are simpler and less costly ways to improve affordability while preserving the virtues of the current student-centered system. Expanding the need-based Pell program would put more resources directly into the hands of students who need it the most. By contrast, cutting list-price tuition at public universities is an untargeted policy that benefits the well-to-do and the poor alike.

Some fear that putting more resources in the hands of students will encourage growth in the for-profit sector that siphons off a disproportionate fraction of federal student aid while offering little educational benefit in return. This is indeed a problem, but it’s one that can be handled by changing a single rule. At present, to participate in federal aid programs an institution must generate at least 10 percent of its revenues from students. At most 90 percent of a school’s revenues can come from government subsidy. This 90/10 rule could be tightened to eliminate the bulk of the worst offenders.

There is no compelling case for massively rewriting the way the federal government subsidizes higher education, and the political chances of a Sanders-style proposal moving through even a Democratic-controlled Congress are very low.

Pell Grants do not cause tuition inflation!

Bob Archibald and I have written a policy paper for the American Council on Education (ACE) about the ways that federal financial aid policy affects college pricing behavior. We take on the Bennett Hypothesis, which suggests a link between increases in federal aid and subsequent increases in college tuition.

You can find the paper on ACE’s website here: Federal Financial Aid Policy and College Behavior

Net Price vs. List Price: Why the BLS needs to change its measure of college cost

The Washington Post’s higher education blog (Grade Point) ran a short article of ours on the net cost of higher education and the Bureau of Labor Statistics’ flawed index of college cost.

The Cost of Private College isn’t Actually Skyrocketing



Getting the Facts Right about Higher Education Prices

by Robert B. Archibald and David H. Feldman

Everyone knows that college costs are soaring.One quick look at the Bureau of Labor Statistics’ price index for college tuition and fees shows that. According to these data tuition costs are rising faster than even medical care.

Although this is now the conventional wisdom, it’s quite wrong.

We are not the first to point out the problems with the BLS measure of college prices. Yet this price series continues to figure prominently in many discussions because it serves a vital role in a narrative about higher education that increasingly passes for simple common sense.

[See a list of the 57 U.S. colleges and universities that have a “sticker price” of more than $60,000 a year.]

The first part of the story is about university choices: Free-spending colleges have allowed programs to expand and tenured faculty to redefine their jobs toward unproductive research. They have also gold-plated the student experience with spa-like amenities. Rising state and federal grant aid encourages more attendance, and this demand allows schools to jack up the price.

Meanwhile, the easy availability of loans pushes an entire generation to take on an unwarranted and unpayable debt burden that will explode as the next financial bubble.

The logical end of this unfolding catastrophe is that the old university system will collapse within 20 years, disrupted by new technologies.

But a number of big facts don’t fit this story.

First, although the list price of a year in college is indeed rising rapidly, the price the average student has to pay is not rising nearly as rapidly. Secondly, much of the increase that students do pay is driven by decreases in state subsidies, not increases in the cost of providing the education.

Let’s start with the Bureau of Labor Statistics. In order to create the Consumer Price Index (CPI), the BLS gathers prices for hundreds of goods and services we buy. Those prices are supposed to capture what is actually paid, and for most items these “transaction prices” are exactly what the BLS uses.

For physician’s services and hospital services, the BLS does not use the “charge” you see on your doctor’s bill. If you have medical insurance, the “price” used by the BLS is the co-pay you put down plus any payment the physician receives from the insurance provider.  If you have looked at a medical bill lately you will see that the insurance provider has negotiated a considerable discount on the price listed on the bill. Few consumers of medical care actually pay the full listed charge.

For the college tuition and fees price index the BLS collects list prices published in college catalogs, not the prices students actually pay.

The average student pays considerably less because of institutional grants from the schools, federal grants, state grants, private scholarships and tuition tax credits.

And because much of the price cut comes from institutional grants (discounts), the price used by the BLS is much higher than the revenue the college gets from the average student.

Over time the list price and the net price have diverged quite a bit.

[Read more about tuition discounting here.]

This is because colleges and universities are increasingly using discounting to fill their seats and, for more selective schools, to seek out special talents and economic diversity in the incoming class.

In the figure below, we have plotted the real prices of college tuition and fees as measured by the BLS, the real price of medical care as measured by the BLS, and the real average net tuition at public and private nonprofit four-year institutions. (Anyone can access the college net price data from Table 7 in “Trends in College Pricing, 2014,” published by the College Board. All of the series use 1990 as the base. If the line slopes upward, then that price has risen more rapidly than the overall inflation rate. See their chart here.)

The two BLS indexes show the conventional wisdom:  List price tuition and fees have grown substantially more rapidly than the price of medical care.

The story for average net tuition is very different.

For students attending private four-year institutions the real net price has barely budged since 1990, which is in stark contrast to medical care and to the conventional story of soaring college cost.

For public four-year institutions the net price has not risen as much as the BLS numbers would suggest. In the recent recession, however, public university prices moved up substantially so that they now closely match the rise in medical care costs since 1990.

Does this tell us that public universities suffer from runaway cost?

If you see rising tuition, the cause has to be found in one of two places – increased spending by schools or decreased subsidies given to them.

Data from the College Board tells us that the real cost of providing the education for each full-time student rose at an annual rate of only 0.44 percent from 2001 to 2011. Over the same decade real income per person (gross domestic product) in the United States rose at almost twice that rate.

Over the same time period the real value of the state appropriation for public four-year schools fell by roughly four percent per year for each full-time student. The rise in the net price paid by students at public universities is overwhelmingly driven by state cuts, not rapidly rising spending by wasteful public universities.

There is indeed a crisis unfolding in American higher education today, but it has little to do with dysfunctional universities and excessive government handouts. The real crisis is much broader than anything decided by college administrators. You will find it in the hollowing out of the American middle class, the stagnation of family income for much of the population, and the 30-year retreat of state higher education appropriations.

Sound bites about fancy dorms are easy. Generating wage growth for the median-income family is much harder.

Preface to the Paperback edition

Since the preface to the paperback edition of Why Does College Cost so Much? is freely available on Google Books, I see no reason not to put it up here. This preface is a good introduction to why we are writing a new book on the challenges facing the American higher education system. The catastrophe-disruption narrative now widely entrenched to the point of being the common wisdom. Yet we think this narrative is based on fundamental misunderstandings of the higher education data and of the trends shaping the economic environment.

With that as introduction:

Preface for the Paperback Edition

The hardback version of this book appeared in late 2010. The U.S. economy was just beginning to recover from the financial meltdown, and median family income was substantially lower than it had been earlier in the decade. Rising list price tuition was beginning to bite into the pockets of people in the upper portion of the income spectrum who pay full price while putting a college education seemingly out of reach for many lower and middle-income families.

In the final chapter we called for a calmer and less politicized public debate about the complex U.S. higher education system. As time has passed, however, we haven’t seen much evidence that this call has penetrated the public discourse very deeply. If anything the rhetorical temperature has heated up.  The word “crisis” is increasingly appended to the term “higher education.” Google found 436,000 such citations when last we checked. Tuition and fees have reached seemingly “stratospheric” and “unsustainable” levels.  We hear of a higher education bubble that will presumably burst, causing colleges and universities to fail in large numbers. The burden of college debt is said to be crushing the next generation of American graduates.  And new technologies will soon disrupt higher education, completely changing what goes on at a university.

We stand by our earlier call.  The public conversation about higher education would improve if we could ratchet down the rhetoric.  We wrote this book to help people see how the century-long evolution of the entire economy helps explain the current circumstances of our higher education system.  Many of the cost, access, and quality problems we face are long in the making, and most of them have little to do with rising inefficiency or growing mismanagement in higher education. The real problems we face will not be cured overnight by simple policy changes or with magic technological bullets. Labeling them as crises is not helpful.

Our first task in this book is backward looking. We seek to explain why higher education costs and prices – and these are different things – have generally risen more rapidly than the inflation rate over the last sixty plus years.  This is inherently a comparative question. Are there other industries that have experienced something quite similar (yes), and do these industries share certain predictable traits (also yes)? Our work is also forward looking and policy-oriented, and the first step on the path to practical solutions is to know how we arrived at our current situation. Policies built on a misunderstanding of how we got here are not likely to produce good outcomes.

Context is very important for our analysis, and it is often lacking in discussions of higher education.  For example, “administrative bloat” is often tagged as an important driver of higher education costs.  The supposed “bloat” is the rising administrative fraction of the university workforce. This percentage has indeed grown, but so has the administrative fraction of the whole US workforce. Putting facts in context is essential if we are to separate between things that are problems and things that are not. The relevant question we should ask is whether or not added administrators – everything from IT professionals to psychological counselors – are worth the cost.

If context is a crucial framework for telling the story of rising college cost, data provides the real structure. Yet much of the public debate about higher education and its ills is built on unrepresentative examples and cherry-picked statistics. Given the wide variety of higher education options provided by thousands of institutions of all shapes and sizes, there is fertile ground for anecdotes that can highlight almost anything one wants to show. We make no apology for using a lot of data covering much of the past century. Some of the current crisis rhetoric should melt away when one takes a close look at the data for all of higher education.

The higher education bubble idea is a good example. It is based on the recent list price tuition increases combined with current labor market problems for some graduates. But the processes that lead to rising college costs (and price) have been at work for as long as we have good records.  And the economic return to higher education is growing, not shrinking. In recent years, much of that growth has come from falling wages for high school graduates, but the payoff to acquiring a degree is always relative to the alternative of entering the labor force without one.

Tuition and fees have indeed risen substantially over the past thirty-five years. Are they spiraling out of control? The College Board’s annual Trends in College Pricing is something that all commentators should be forced to read and use.  Recent issues show that there has been scant growth in what the average student actually pays (as opposed to the list price). Many students do face very high list-price tuition, but again the data in Trends in College Pricing show that a relatively small fraction of students attend institutions with the kind of tuition and fee bills that make headlines.  At present almost fifty percent of students at public four-year institutions face a list price of less than $9,000.  The large numbers of students who attend two-year schools pay even less.

Rising student debt is also an important national issue, but the crisis label serves tabloid journalism more than the public interest. Anecdotes and unrepresentative samples dominate the national conversation.  Debt grows when family incomes fail to increase sufficiently to cover the increases in college expenses, or if we succeed in getting students from poorer families into college in greater numbers. For all but the very top of the American earnings distribution family income has moved in reverse over the past decade, and as we demonstrate in this volume similar economy-wide forces have driven up the costs of providing a higher education.  Neither problem has a simple solution, and fixing the national income distribution goes well beyond higher education policy. In fact, the very practices that help colleges keep their programs affordable to lower income students (price discounting) have helped to push up list price tuition.

The growing debt problem, however, has been hyped by stories focused on unfortunate students who have debts that are often wildly unrepresentative. These students’ problems are real, but they are not typical. And in most cases, they were avoidable. Once again, the College Board’s Trends in Student Aid 2013 provides real data.  At public institutions the average debt per borrower (in 2012 dollars) increased from $20,800 in 1999-2000 to $25,000 in 2011-2012.  This is a problem, especially for students who do not finish a degree, but words like “skyrocketing” or “unsustainable” add nothing to the public debate about reforming how we aid families who want to finance long term educational investment.

Lastly, we remain agnostic about the potential for new technologies to revolutionize higher education. The fanfare that accompanied the advent of Massively Open Online Courses (MOOCs) suggested that big changes were in the offing. Instructional quality would rise while cost would plummet. This is the ideal reform. The effect thus far has been modest.  We have seen steady growth in the number of online courses and online degrees, but we are far from the demise of face-to-face learning. Courses that blend online features with traditional classes are also becoming popular. Evolution is a much better term than revolution to describe this process, and much of the impetus for reform is coming from within traditional higher education institutions as they try to preserve their mission while wrestling with the hard problem of rising cost.

We try to steer away from the politicization that characterizes much of the higher education debate. Our goal is to give the reader a measured analysis infused with data and grounded in a broad understanding of historical trends that have shaped both higher education and the economy as a whole. Use it to help sort through the hyperventilated rhetoric that no doubt will continue despite our calls for calm.

Robert Archibald

David Feldman

Williamsburg, VA


Unintended Consequences of “free” Community College.

Bob and I helped initiate the Washington Post’s new higher education blog today. Our piece is a short introduction to some of the real complexity of any substantial public policy idea, and today’s example is president Obama’s proposal to make community college free to students.

Here’s the link to the Post story:

Unintended Consequences of President Obama’s Proposal …


Two economics professors at the College of William & Mary, Robert B. Archibald and David H. Feldman, experts in college costs, question the idea of “free” community college tuition as proposed by President Obama — and crunched the numbers to find the potential fiscal impact on each state:

A week has passed since President Obama released his “free” community college proposal, and the broad outlines of the pro- and anti-positions are now discernible. Unemployment rates for people with more education than a high school degree are indeed lower, and their incomes are substantially higher. The job market of the future will undoubtedly shift toward jobs that require more than high school level skills. Free community college will indeed shift some students out of for-profit schools with low graduation rates that saddle graduates and non-graduates alike with very high levels of student debt. On the other side, we hear that students should have some skin in the game, for-profit schools are the innovative future, some college grads seem to be working in low-level jobs that don’t require fancy degrees, and more federal subsidy will just push up tuition. In addition, the proposal is supposedly dead on arrival anyway because it adds to the federal deficit and federalizes what should be a state decision. Anyone who is minimally aware should be able to see the left-right divide in this list of reactions.

In a short essay, we cannot separate the wheat arguments from the chaff. Instead, we offer two arguments and one bit of evidence to show the true complexity of this issue.

One of the program’s goals is to break down the financial barriers that keep many students from pursuing a useful certificate or degree. Two free years at a community college supposedly will make a four-year bachelor’s degree more affordable. Yet four-year colleges and universities depend on larger classes taught to first and second year students to keep cost down. These larger introductory classes are the flip side of the smaller and more teacher-intensive upper level classes of the final two years. The upper division courses are the ones that truly prepare students for a job market that prizes advanced training in technical and non-technical fields alike. If the proportion of freshmen and sophomores at four-year universities falls, this could push up the cost of a four-year degree for students who go directly to places like Ohio State or Oregon. Welcome to the “law of unintended consequences.”

In a similar vein, recent research shows that students who could do well at a four-year school but who choose to begin at a community college are less likely to graduate than their peers who go directly to a four-year school. Inducing students to go to college is a good thing, especially if their alternative is a low-paying dead-end job. But inducing students to switch from a four-year program to a “free” community college is a more nuanced outcome.

Lastly, the politics of this program are far from obvious. The chart below shows an estimate of the initial fiscal cost of making community college “free” in each state, measured as dollars per thousand of that state’s personal income. We say “initial” and “estimated” because we have no way to know precisely how many current students would qualify for aid (those in the appropriate programs with the requisite GPA) and how many more students will move into the system. Even without that extra knowledge, it’s clear that there is a lot of variation nationally. In Florida, for instance, the cost is 33 cents per thousand. In Kentucky, it’s three dollars and a penny.

Since the federal government is picking up ¾ of the tab, states like Iowa, South Carolina, and Arizona (the latter two quite red), would be clear net recipients of federal money even though they would have to kick in some extra cash themselves (the other ¼). How would South Carolina’s leadership react? Not sure.

 Feldman added this note about the chart:

“Fiscal cost” is the listed tuition times the number of students enrolled. That’s the cost that will have to be split between the feds (3/4) and the state (1/4) if you make CC free. This “fiscal cost” measure is really an “initial” cost, since it could grow if the number of enrolled students rises as a consequence of the price cut. But it’s a good starting place for the discussion. And this “fiscal cost” is based on an assumption that every student qualifies, which clearly is an overestimate. Some students wouldn’t qualify, either because their grades are too low or because they are not enrolled at least half-time. But we don’t think that students in Iowa, for instance, routinely get lower (or higher) grades than students in Florida. What we’re trying to show is the fact that there is a lot of variability among states, which comes out pretty clearly in that picture. The fiscal effects vary substantially.