Why Earnings Alone Cannot Define Higher Education Accountability
Why Earnings Alone Cannot Define Higher Education Accountability
Why the accountability debate is more complicated than it looks
Higher education accountability is increasingly centered on earnings outcomes. The assumption is straightforward: students earn a credential, enter the workforce, and their salaries reflect institutional quality.
But Glenda Morgan argues the reality is far more complex.
Earnings are not produced by institutions alone. They are shaped by geography, labor markets, career pathways, industry structures, and personal choices. Treating salary as a direct institutional output ignores the broader systems that influence economic outcomes.
That distinction matters because accountability systems shape policy, funding, and which programs institutions choose to sustain.
Why earnings are not a clean institutional metric
A graduate’s salary reflects more than where they studied.
Regional differences play a major role. Urban and rural labor markets produce different wage outcomes, even for students with similar credentials. Cost of living also affects salary structures. The same graduate may earn dramatically different wages depending on location.
Career pathways matter too. Some professions have highly structured salary trajectories, while others develop more gradually over time.
Morgan’s argument is that earnings are a systems-level outcome, not a simple cause-and-effect institutional measure.
Why median earnings can distort accountability
Median earnings simplify complexity into a single number.
That can obscure important differences between programs and professions. High-variance programs may produce both very high and very low earners. Low-floor professions may provide critical public value despite lower salaries.
Morgan also argues that earnings snapshots fail to account for long-term trajectories. Some fields produce immediate returns, while others develop more slowly over the course of a career.
Research shows that liberal arts graduates, for example, may initially earn less than engineering graduates but eventually narrow or surpass those gaps over time.
What accountability systems should measure instead
Morgan argues for a more nuanced accountability framework.
Completion rates should play a larger role, particularly given the scale of students with some college but no credential. Time to degree also matters because delays increase cost and debt burdens.
Geography, labor markets, and career variation should be incorporated into outcome measures. Accountability systems should recognize that different programs produce different types of value and different earning trajectories.
Most importantly, institutions should be evaluated using multiple measures rather than a single earnings metric.
Why this matters for public policy
The design of accountability systems influences institutional behavior.
If metrics are too narrow, institutions may reduce investment in socially valuable professions with lower earnings outcomes. That could worsen shortages in fields like teaching, counseling, and social work.
The challenge for policymakers is to build systems that value outcomes without oversimplifying how education, labor markets, and society actually interact.
Read Glenda Morgan’s article “Earnings Data Are Driving Policy—and Misleading It” for more insights.
Transcript
Wes (00:26.786) Morgan, thank you for joining us today and welcome to the President’s Forum Podcast.
Glenda Morgan (00:47.604) Thanks and it’s a pleasure to be here.
Wes (00:50.488) Hey, your article argues that it isn’t just a measurement of earnings that’s the problem. It’s actually a causality problem. So it’s very detailed in laying that out for us, but earnings are being attributed to institutions when they’re actually produced by systems. Can you explain that to our listeners and tell us a little bit about why that distinction matters?
for how we design accountability in public policy for higher education.
Glenda Morgan (01:26.25) Sure, yeah, you know, in a lot of the accountability discourse that’s going on, earnings are often treated like a clean institutional output. know, somebody goes to college or university, they graduate, they have earnings and they’re seen as a, you you’ve got cause and effect. But actually what happens is much more complex than that, is that somebody goes to university, they take one of a variety of different kinds of programs.
and then they graduate. But what they actually earn is a product of all different kinds of things. It is a product of where they graduate, are they going to be living in urban or rural kind of setting, but also what kind of a job they’re going into. Some jobs have very determined pathways, others are much more flexible.
And so you’ve got these multiple causality things going on and so what people are actually earning after they graduate is the result of multiple factors all acting together. So it’s not just cause and effect. It’s a highly complex kind of a system. So holding one aspect of that responsible for the outcome is just a crazy sort of setup.
you know, because what’s actually happening is you’ve got all kinds of things interacting to produce a highly variable.
Glenda Morgan (03:21.268) It makes sense to everybody, you know, where you live is going to determine what your costs of living are. And it also sort of determines what you’re paid. I mean, it’s so ingrained in us to understand that, but somehow it hasn’t made its way into the metrics yet. You know, it’s not just urban and rural. It’s also, I mean, there’s a regional aspect that I didn’t write about because my colleague Phil has written about that. But where you live determines a lot of
your costs but it also determines where you’re paid. I used to work for Gardner and they actually you know it was a fully remote company but they actually linked your your salary to where you were living. There were high cost places and low cost places.
Wes (04:05.432) Yeah, that makes sense. Well, in this paper, you also mentioned you described three types of programs that have very different earning structures. And the three programs that you lay out are pipeline programs, high-variance programs, and low-floor programs. First, can you just describe what each those are, each program is for our listeners? And then…
I’d love to get into some of the details of measuring those and why one single median measurement doesn’t quite work.
Glenda Morgan (04:43.114) Sure, as we go on, just want to be sure to call out Ithaca, which my little article was based on their research. Ithaca SNR did some great research on South Carolina, but it’s broadly applicable. So much depends on the kind of the program and then the pathway out of that program for graduates out of there. And they identified three. So the first one are pipeline programs. This is where
You graduate from a program and your pathway is pretty determined. You’re something like nursing where, you know, there are a couple of different paths you can take, but it’s pretty set. And your salaries are in some ways determined by that pathway. And so they’re somewhat predictable. Another one is engineering, you know, how you progress and where you go. You you’ve got certifications and things like that that you do, but it’s certainly set.
And then you’ve got much more flexible kinds of programs. Sorry. High variance programs. this, you know, with a pipeline program, your career and what you’re going to do after you graduate are are largely determined by the program that you’ve done.
Wes (05:58.563) high variance programs.
Glenda Morgan (06:15.136) With high variance programs, it’s less a profession than a set of opportunities. So something like business and even computer science, I would argue, are high variance programs. So they’re not only in terms of what you’re actually going to do is going to vary a lot. You can go to lots of different kinds of places and it’s really up to you in terms of what you’re going to do and what you’re going to make of that, but also your salary, what you’re actually paid.
is going to determine is going to vary a lot. So you’re to have a huge variation in terms of earnings and pathways and occupations. It’s really not determined by the actual degree. It’s determined by what your interests are and how you progress in that. I, for example, I have a PhD in political science, you know, and
you could have become, I could have become a professor or I chose to become an industry analyst and it’s the ultimate high variance kind of programs. And then you’ve got low floor programs and these are sort of, they’ve got elements of both of those in that there’s a big variation in terms of what people do, but earnings are traditionally fairly low. So things like social work, counseling,
often the arts as well. So there’s a lot of variation in terms of what people do, but the floor tends to be pretty low as well in terms of what they make.
Wes (07:49.358) Could we lump in like teaching, mental health programs? Yeah, okay. So these are programs that we actually really do need.
Glenda Morgan (08:03.59) Absolutely, yes. You know, as a society, we rely on those kinds of things. But they have traditionally been paid less. In part, you know, there’s somebody who writes about librarians, for example, who talks about vocational awe, you know, where everybody really admires what they do, but they aren’t prepared to pay for it. And so you’ve got these low-floor kinds of things.
Wes (08:31.79) Okay, so when you take a median, when you just break that down and take one number out, how does that not yield the accountability that we’re actually looking for?
Glenda Morgan (08:47.914) So, you know, people often think about medians as being better than averages and they are, but, you know, they aren’t accounting for the variation across that. Particularly, I think the most egregious example is the high variance programs because a median is just telling you, you know, the middle of between the bottom and the end. And it’s not sort of really telling you in general how people are going to do there, but they’re certainly not capturing
the value of the input as well. There’s a logic breakdown there because what people are earning is determined by the system, not by the actual input of the beginning. It’s just the beginning point that we’re putting a lot of emphasis on and it’s not really a valid measure of anything.
Wes (09:44.674) Well, it just seems that those three different types of programs could create a little bit of a problem having, just evaluating that one number, particularly at the end of the day, when you’re looking at social value of some of these low floor careers and the credentials that are required for that.
Glenda Morgan (10:10.014) Yeah.
Wes (10:14.146) We have, you can’t get rid of all of these credentials because they don’t provide you the economic return that some other careers might because you need them for society. How do you deal with that?
Glenda Morgan (10:28.82) Yeah, you know, that’s a slightly different thing than I argued in the piece, but I think, you know, we have to think about what we need as a society. I remember, as it happens, I’m South African originally. And there was this sort of amazing moment where I sort of understood things in a much deeper kind of way. I was just before I came to the US, it was the end of apartheid.
And as it happened, I went to the University of Cape Town, one of the best universities in the continent of Africa. And I remember hearing a conversation and it was a time of rapid change. There was this guy who was on the Board of Governors, the Board of Regents of the University of Cape Town. He was a businessman, very successful. He said,
My job is to understand the role of the university. And so, for example, in the College of Medicine, we have to provide doctors to the whole of the society. And, you know, as a businessman, I understand inputs and I understand outputs. And if we only get one kind of input, we’re only going to have one kind of output.
So we need multiple kinds of inputs in order to provide doctors for all the different parts, know, for rural, for plastic surgeons, for orthopedic surgeons, for all these different kinds of things. And so I think in terms of our accountability, we need to think of the same sort of thing, inputs and outputs, you know, we need social workers, we need teachers, we need these kinds of things. So we need to make sure that we produce them because we’re going to hurt if we don’t.
Wes (12:23.086) Right, right. Well, you know, that’s clearly the the when you’re talking about we don’t just measure inputs. We do want to look to outcomes. You’re I mean, that’s speaking President’s forum language. We’ve been talking about that for a long, long time. But look, we can’t just we can’t measure accountability by, you know, the way that education is provided, whether that’s in person or online or.
Glenda Morgan (12:34.208) Yeah.
Wes (12:51.16) We can’t just look to the inputs, but inputs and outputs can both be important. Boiling it down to one specific earning number is more complicated than it seems, but let’s get to the, if we’re redesigning this system, tell us what you would build if it were a ground up build on accountability. Well, how would you do it?
Glenda Morgan (13:16.734) We’ve got 43 million Americans with some college no credential. And I think…
Wes (13:49.538) Ha
Glenda Morgan (14:14.472) you know, you can have the best earning credential in the business, but if you’re not actually getting the credential, it’s not going to help you. So I think, you know, including more metrics there, including completion, time to degree, those kinds of things, you know, is sort of is part of that. And really developing a more nuanced measure of that. So including regionality.
including urban versus rural, those kinds of things. So that’s sort of how I would start to design it more from the ground up. But I would put heavily an emphasis on if somebody actually is going to college that they’re coming out of it with a degree or a credential of some sort.
Wes (15:01.878) I love that thinking and that does get forgotten when it’s just one metric after, if you’re just looking at earnings, you’re not seeing all of the non-completers and the cost to the system that that is.
Glenda Morgan (15:15.455). Yeah, no, absolutely. And then they’re stuck with the debt often. And it’s just a sort of nightmare. So I want that to be part of the part of that sort of calculation, but also, you know, thinking also in terms of where people going and how they’re doing. The other thing we haven’t talked about is also time, which I wrote about in the in the article is that, you know, a snapshot in time is not going to give you a
a great measure because some of these professions, for example, the pipeline things are relatively high earning right out the gate, whereas other ones are slow brewing. So there are studies that show that right out the gate engineering graduates earn much more than say, science people. But in the long term, the liberal arts actually catch up and overtake.
I think just looking at snapshots in time is problematic. You need a longer term measure.
Wes (16:26.22) I’m glad you brought that up because that’s a huge variance and it’s really important to capture. It’s hard to capture. It’s very difficult. I don’t know if there’s a clean way that you can do that, but your point is some of these take a much longer time than five years out your credential. They brew over a career.
Glenda Morgan (16:44.768) Yeah, absolutely. Yeah, no, absolutely. And, you know, going back to the median issue, I’ve just been rereading Todd Rose’s The End of Average. And a lot of people have some issues with the book, but I sort of really like it. It’s that, you know, when you’ve got things that don’t correlate, you’ve got multiple measures that don’t correlate, just using an average really gives you a bad result. You know, he uses the example of
Wes (16:55.086) Mm-hmm.
Glenda Morgan (17:10.096) airplane cockpits. Originally they were designed for the average person but turns out nobody’s actually average. Because you’ve got these multiple measures, know, and so we need to sort of bring multiple measures into things instead of using that median of just the earnings.
Wes (17:28.398) Right, well this has been a very interesting conversation Morgan. We will direct our listeners to your piece on this so they can read all the details and we would love to continue this conversation as things move forward with accountability during this administration and future administrations. We really appreciate your thinking about this.
Glenda Morgan (17:37.269) Bye.
Glenda Morgan (17:51.134) my absolute pleasure and lovelies to speak with you. Okay, thanks.
Wes (17:54.616) Thanks, Morgan.
Recent Comments