Congress is Coming for Colleges

Culling crappy colleges by compelling compensation

As promised, I’m back with part two of my read of the House reconciliation bill. As not promised, it’s late because I’ve been spending all of my time getting my house in Asheville, NC ready to Airbnb. As I worked on the house, I listened to Abundance by Ezra Klein and Derek Thompson and The Debt Trap by Josh Mitchell. I’ve been thinking about this bill a lot in general, but also in the context of what the authors share in these books, and there are a few things that have stuck in my brain:

1. Policies should be designed to directly address the problem they’re trying to solve. We’ve tried to solve too many problems through the student aid system, in particular through repayment. Whether the issue be program quality, college costs, insufficient access to the social safety net, poor completion rates, or too-low incomes, repayment shouldn’t be the mechanism we try to use to solve all these problems. If we are trying to ensure people have enough money to pay their student loans, we need to ensure that people are able to complete school in a reasonable timeframe, earn a credential of value, earn good wages, and not get slammed by other expenses (e.g. healthcare, living expenses, freaking groceries, etc.). Policies intended to impact these things should directly call them out - not be a secondary effect.

2. Too many policies assume perfect (and economically rational) human behavior. Humans are messy. College-going and repayment is especially messy. It’s rife with irrational choices (e.g. I will pay much more to go to my “dream” school rather than a comparable affordable option) and lack of perfect knowledge (e.g. I’m just gonna not pay, what’s the worst that could happen?). The pipeline isn’t some straightforward thing where a student enrolls totally prepared for the coursework, finishes in two or four years, then graduates, gets a job, and stays in repayment. Sure, that is what happens with some folks, but it’s pretty rare. People enroll, need remediation, drop out, re-enroll, transfer, maybe graduate in 6 or more years, start repaying, change plans, consolidate, stop paying, only pay part of their bill, get transferred between servicers, on and on and on. We’d like to think of these things as linear but they’re not. My graduate school advisor, Steve Desjardins, has done a ton of work on this, and recently reminded me of this paper that I think is a must-read for anyone doing higher education policy. A lot of policies (and parts of this reconciliation bill) take a far too academic approach to the messy world of higher education, and it would do folks some good to leave DC and talk to some real, on-the-ground people every once in awhile.

3. Writing a policy is only the first step. I know that a ton of people have worked tirelessly crafting the language around statute and regulations over the years. But that work is only the first step of getting things done. A policy is only as good as its ability to be implemented, and I have so, so many concerns about how difficult some of the things in this bill will be to implement (the freaking median program cost of attendance thing continues to make me wither whenever I think of it). I’m also worried that the Hill might be making some… problematic assumptions when it comes to how implemented certain existing policies (e.g. the FUTURE Act) are in reality, and how quickly things can be implemented going forward, especially with FSA sitting at half its normal headcount and having been flat funded since FY2022.

ANYWHO. Let’s talk college accountability.

There a few big changes when it comes to college accountability in this bill. I will be the first to tell you that accountability for colleges is SORELY needed. Right now, the accreditation process dictates a college’s operating status (and this is a whole can of worms I won’t go into), and a handful of policies hold colleges accountable for their financial solvency and borrowers’ repayment of loans. The primary mechanism here is the cohort default rate (CDR), which requires schools to keep their default rates under 30 percent over the course of a 2-3 years (and under 40 percent for one year) if they want to participate in the federal student loan program. Ignoring the fact that these are ridiculously high thresholds (and the colleges that have hit them in the last 10 years haven’t experienced any consequences), there’s also a catch: the loans with the largest balances aren’t included in the calculation.

This means that graduate programs can charge as much as they want and give borrowers a crappy credential and barely worry about repercussions because PLUS loans aren’t included in the CDR. And because PLUS isn’t included in the CDR, schools can also encourage parent borrowers to take on way more debt than they can afford and not worry about them defaulting because they aren’t held accountable for the repayment of those loans. While these loans generally have better repayment outcomes than undergraduate debt, excluding them allows colleges to ignore the outcomes of those borrowers outright.

This bill changes that by creating what I’ll call Gainful Employment (GE) for All. It holds colleges - ALL colleges - responsible for repaying a portion of the unpaid balances of their borrowers. While this bill kills GE, which only held programs at for-profits and certificate programs anywhere accountable for borrowers’ repayment, it applies the principles of GE to all programs. Colleges would get measured on three cohorts: completing students, non-completing undergraduates, and non-completing graduates. This is a pretty good way to portion out the cohorts, as completers have better repayment outcomes than noncompleters and undergraduate noncompleters will have less debt than graduate noncompleters. It includes PLUS loans (with Parent loans being included in the undergraduate cohorts) and address the trickiness associated with consolidation loans (though I feel bad for whoever at FSA is going to have to figure out how to implement this). It would also be nice to see two completion categories (undergrad and grad) so that the cohorts can be directly compared and colleges that award a lot of graduate degrees (and debt) don’t mask the undergraduate outcomes.

Starting in 2028-29, colleges would be required to make a reimbursement payment annually, which is 1 minus the reimbursement percentage divided by the non-repayment balance, times 100.

To calculate reimbursement percentage:

• For completing cohorts, the reimbursement percentage is the median value-added earnings for each program divided by the median total price charged to students. Median earnings is pretty straightforward, and total price is defined as the amount charged to borrowers minus any non-Federal grant or scholarship aid.

• For non-completing undergraduates, the reimbursement percentage is the share of students who received aid who didn’t complete within 150 percent of program time (e.g. 3 years for 2-year programs, 6 years for 4-year programs). In the case of community college transfers, the reimbursement percentage is the share of borrowers who didn’t complete their program at a four-year college. This is totally insane, as community colleges have zero control over what a student does when they’re enrolled at another institution.

• For non-completing graduates, the percentage is just the share of borrowers who did not complete their program within 150 percent of program time.

  To calculate the

The non-repayment balance is:

• The difference between the amount owed by borrowers and the amount paid, including the amount of interest waived (such as by the RAP) and any amounts forgiven or discharged by the Secretary.

I’d provide examples, but unfortunately, these data aren’t currently available, particularly when it comes to amount due vs. amount paid. There are some proxies available, and the House Ed and Workforce Committee published some data in the fall when they published the College Cost Reduction Act. The biggest losers when it comes to the risk-sharing proposal are large for-profits, but there are also a few surprises in there, like Liberty, a conservative darling with a troubled past; NYU, which is well-known to drown students in debt; and Georgia State, a college often lauded for its efforts to innovate in its ability to serve its students (thankfully for them, their large risk-sharing payment is offset by the PROMISE Grant, which I’ll cover next).

While I appreciate the estimates put out by the Committee, I worry that we can’t run the exact numbers on the risk-sharing proposal. This is because FSA doesn’t collect amount paid by borrowers in NSLDS, which is a primary component of the calculation. Putting forward such a significant risk sharing policy without having an exact understanding of the reimbursement amounts is very concerning. Implementation will end up being kind of a nightmare, with FSA then 5,500 colleges working through the data reporting schema and ensuring it is accurate. Then when the rates are finally calculated, some colleges end up losing their shit at their Members of Congress because the cost impacts are way higher than estimated. We then end up with changes to the policy or exemptions, which leads to all of the work that went into the original schema being scrapped or revised. Then we start the implementation cycle all over again, and no college ends up being held accountable. It’s not a good path to go down, and if Congress wants to keep something like this, they should know exactly the penalties before they sign it into law so we don’t end up wasting time and money with all of the back and forth that will inevitably happen.

PROMISE Grants

This new campus-based program (called Promoting Real Opportunities to Maximize Investments and Savings in Education) provides colleges who meet certain criteria with an annual, noncompetitive grant. It has three components where schools demonstrate how they are:

1. Increasing affordability by providing a price guarantee or minimizing the maximum price of completion;

2. Increasing access, including participating in a variety of existing information disclosures; and

3. Increasing student success, including reducing time to completion and increasing value-added earnings.

Part of the affordability aspect of the PROMISE grant is providing students with a price guarantee, wherein they disclose to students the maximum they will pay to complete their program. While this is a good thing, I wonder how it will work practically given the issues that I pointed out last week with the program-based cost of attendance, where most students don’t know their major when they enroll or switch majors while enrolled. I also have questions as to how this will be enforced and what will happen if an institution does not fulfill its obligations, which as far as I can decifer, isn’t outlined in the bill.

The PROMISE grant amount equals

• the lesser of

  • the difference determined by subtracting 1 from the quotient of the 3-year average of the median value-added earnings of completers,

  • or the number 2

• times the 3-year average of the total number of Pell Grants received by enrolled students

• times the 3-year average of the share of low-income students who completed their program within 100 percent of the program time (or who transfer from a two-year college and complete a program at a four-year college).

Just typing that out nearly gave me a seizure, so I’m not going to provide an example, but this doesn’t have quite the same problem as the risk-sharing proposals because the data largely exists from other reporting or policies.¹ There are also a few additional details here, but the general thrust of the program is: get more money if you’re providing solid on-time completion and earnings, but also, get some money just by enrolling a lot of low-income students.

By the Committee’s estimates, the biggest winners here are regional comprehensives, which makes sense, as they’re selective enough to not have a ton of students needing so much remediation so that they fail to complete, are lower cost so as to enroll a good number of low-income students, and tend to be more focused on providing programs more tied to the workforce (as opposed to say, programs in the humanities, though most offer those as well). What’s interesting is that states with generous aid programs win out here as well, with California, New York, Florida, and Texas dominating the list, so the Committee may be considering this an indirect way to encourage state reinvestment in higher education.

The way that I see it, the PROMISE grant would help offset the risk-sharing proposal for community colleges and minority-serving institutions in particular. While this is a nice incentive for some colleges, I’m skeptical of how this will meaningfully improve outcomes for schools in the long run, and how the heck a gutted ED is going to manage this program.

What this means for colleges

The higher ed lobby is one of the strongest in the nation. Big College(TM) is represented by hundreds of colleges who pay their own lobbyists as well as dozens (hundreds?) of other interest groups that represent their members, from college presidents to campus bookstores to professional associations across a variety of professions. There are a whole lot of financial interests tied up in colleges, and every single state and district has at least one college in it. This means that every House member has a college knocking on their door talking about how the risk-sharing provisions will put them out of business, reducing the job market and economic stimulus that comes with having a college in that district. The deluge is greater for those with a lot of colleges in their district (or for senators, who have all different types of colleges in their state), because most colleges will fight this policy, with four-year colleges having different complaints than community colleges.

According to data, the biggest losers on net would be most of the large, for-profit colleges that show up on the risk-sharing list. Perhaps this is why the bill also proposes getting rid of the 90/10 rule, which holds for-profit colleges accountable if more than 90 percent of their revenues are made up of federal sources.

There is no way a lot of colleges would be able to sustain these loses, so they’d either have to shut down or opt out of the federal aid programs. Some of the for-profit ones are big (and rich) enough to potentially start their own lending programs, which would present a whole slew of other issues for the students that enroll there and would have to be (/should be) dealt with by the consumer finance space (like the CFPB’s Private Student Loan Ombudsman, should the CFPB come back…).

Congress to colleges.

On net, I really don’t hate a lot of this, though I do think that these are massive changes to make very quickly should this reconciliation bill pass, especially with a (let me mention it again) totally gutted Department of Education. I would love to see a cost breakdown by section, and I’d love to see groups engaging with this part of the bill productively, because think this is actually something Democrats can get behind.

I’ll be back to more regular posting, though down to one post a week for all due to other work that has come across my desk. And in a couple days I’ll have the next paywalled post up breaking down more of my analysis of this bill.

I appreciate y’all.

1. Earnings isn’t the most reliable, but the proxies are better than what exists for dollars repaid.