How NOT to Regulate ISAs

Income-sharing agreements or ISAs are becoming an increasingly popular way to finance higher education. The key problem that ISAs face as a product is uncertainty about their regulatory status. Are ISAs “credit” for various statutory purposes? Or are they something else? Into the regulatory void comes a bipartisan bill, introduced by Senators Mark Warner, Marco Rubio, and Chris Coons, that would set forth a regulatory framework for ISAs. The problem is that the regulatory framework proposed is shamefully bad: it would give a green light to discriminatory financing terms and tie the CFPB’s hands from further regulating ISAs.

What the Bill Does.

So here’s what the bill does. First, it attempts to define ISAs. It doesn’t do this especially well because it defines an ISA as not being “a loan,” but the term “loan” isn’t defined anywhere. So we don’t really know what is an ISA. I guess anything that calls itself an ISA, it must be an ISA.

Second, the bill defines what are “qualified ISAs,” and provides that qualified ISAs are exempt from both state regulation, including usury laws and also from federal agency regulation by the CFPB, other than through enforcement actions. Qualification requires an ISA to have certain mandatory terms, and these terms offer some level of consumer protection. For example, the bill has a 360 month term cap and sets 200% of the poverty line as the trigger for repayment obligations. It also caps the effective interest rate on ISAs at 36% for consumers with income between 200% and 225% of the poverty line (that’s between about $25k and $28k for an individual). The bill also caps the aggregate percentage of borrower income for all ISAs at 20% of income, and has a formula capping the product of the number of payments and the income percentage required. (It’s not clear how this would work if the percentage steps up or down over time.) Assuming that ISA funders will all want ISAs to be qualified, the bill thus provides a standardized baseline of federal consumer protection lieu of preempted state law (but doesn’t allow states to opt-out of the preemption as other federal consumer protection statutes have done).

Inadequate Consumer Protections

One might reasonably wish for more stringent borrower protections for qualification. For example, qualification does not prohibit arbitration clauses in ISAs, nor is there anything in the bill equivalent to the borrower defenses that exist under the Higher Education, so there’s no discharge if the school closes or defrauds the student. Imagine if Corinthian students had ISAs. They’d still be paying for Corinthian for the next 30 years.

Likewise, there is no “error resolution” system required for ISAs under the bill because they are not subject to the Truth in Lending Act. Moreover, if you’re earning over 225% of the poverty line, there’s no usury limit in the bill, although the aggregate percentage of income cap still exists.

The bill does require a standardized disclosure regime for ISAs, which is a good thing (and it provides a safe harbor for ISAs that use model disclosure forms), but it comes up short on substantive regulation. For example, the bill requires ISAs to disclose consequences of default, but doesn’t regulate what those consequences are.

There’s also some double-speak in the bill. On the one hand, it caps the number of required payments, but then it provides that any period in which the borrower doesn’t have enough income to make a payment doesn’t count toward the limit. That’s setting up a lifetime of indebtedness. For example, if the ISA says 360 monthly payments, and the student isn’t earning enough for the first four years to have to make a payment, those four years get tacked on to the end. Instead of having 312 payments remaining, the student will still have 360 payments remaining. That means that the risk of the student not earning enough to repay isn’t really on the ISA funder. The ISA starts to function like cumulative preferred stock, where the dividends that aren’t paid just accrue. To me that starts to make ISAs look more like loans.

Greenlighting Discriminatory Lending

The real problem is that the bill expressly exempts ISAs from the anti-discrimination provisions of the Equal Credit Opportunity Act (even while applying the ECOA adverse action notice requirement). Specifically, section 501 of the bill provides that variation in terms among ISAs from the same funder for students in different majors or schools are not activities constituting discrimination.

This means that students at an HBCU or MSU (minority-serving institutions) or Gesthemane Christian College or the Ploni Almoni Rabbinical Academy could be offered different terms than the students at Big Time State or at Olde New England College, irrespective of any actuarial differences. Similarly, computer science majors could be offered a different rate than, say psychology majors or Chicano studies majors, irrespective of any actuarial differences. Given that school and major can have strong correlations with gender, race, and religion, the bill is granting an open license for discriminatory education financing. That’s outrageous.

A bill that will allow discriminatory education financing just enables existing inequalities to be compounded. That's part of the reason we prohibit credit discrimination. Graduates of certain schools have fewer opportunities than graduates of others in part because of the type of alumni networks that exist among graduates. We should be trying to level the playing field, not entrenching existing disparities.

We don’t tolerate discriminatory financing from loans, and we shouldn’t from ISAs. Shame on Senators Warner, Rubio, and Coons for indulging the ISA industry this way.

We need reforms in education financing, including more income-driven repayment (which gets to something similar to the ISA structure). But this bill is not the way to do it.

Comments

[FJP]

Wow. Welcome to 21st century indentured servitude. The bill proposes a maximum payment of 20% of income, but is there any concept in the bill of a graduated percentage based on income? I didn't see one in a quick scan; did I miss it?

So, someone making $30K has to pay $6000 a year on their ISA, reducing their income to $23K, is that right? Have these senators budgeted out living on $23K in any setting except Chez Mom and Dad? It appears they would only have to report the $23K net as gross income. That helps a little but not much.

And the max interest rate is 36% Do we expect that anyone will offer a significantly lower rate?

Perhaps the industry would say that the high rates are risk compensation for the fact that the bill does say that ISA obligations ARE dischargeable in bankruptcy. It would take more time than I have now to sketch out how this could distort consumer filings. Can someone give themselves a 20% raise by filing a Chapter 7 and reaffirming every debt except the ISA?

Other than dischargeability, why is one of these arrangements any better than a conventional student loan with income based repayment? It seems like an especially poor deal for anyone eligible for public service loan forgiveness.

[FJP]

Yes, my hypothetical graduate would be left with $24K, not 23K. Especially if they are a math major.

It's still not much to live on independently.

On the discrimination point, I agree that it is horrifying to imagine the Harvard borrower paying 9% and the Howard borrower being offered only the max 36%. A more interesting policy question is whether the education finance industry (loan or ISA) should be allowed to take into account the earning potential of the degree in deciding the AMOUNT of funding to provide (but not the repayment terms). This is particularly applicable to graduate and professional degrees.

On the one hand, that could help prevent students from getting into debt way over their heads. On the other hand, it could lead to certain professions like art and social work being accessible only to the relatively well-off.