Equity financed human capital

When setting up or scaling up a business, an entrepreneur has a choice of raising capital through either equity or debt. But when it comes to human capital, investments are almost exclusively financed with debt.

Almost.

In 2016, Purdue University, a public university in Indiana, became one of the first universities to start offering a type of equity investment to students as a means of financing higher education. Purdue’s Income Share Agreements (ISA) provide students with up front funds, in exchange for a proportion of their earnings after graduation. The typical ISA would see a student pay 3-4 per cent of her income for about 8-10 years, with the total amount paid capped at 2.5x the initial loan.

The appeal of ISAs is clear, and explains why they’re popping up all over the place (eg University of Utah and Clarkson University). They provide universities with a portfolio of assets that will generate a steady income stream for years to come. Furthermore, ISAs have an additional benefit of aligning outcomes of institutions to their students: only those universities that are confident of their courses’ value on the jobs market will be interested in ISAs; those that know their tuition is overpriced and unconnected to the economic value of their degrees will steer clear of them.

Sure, there’s a risk that they could lose money. Underperforming students won’t pay back the investment in full. And some degrees aren’t meant to produce high income returns in the workforce.

But to the extent that universities are able to invest in the services provided and altering the student mix, they’ll be able to influence that income stream. It really becomes an equity portfolio that has to be properly managed.

Underpinning the ISA is a moderately hefty data exercise. Data from the US Census Bureau is used to estimate the expected income trajectory of a student that studies economics or science or arts, and that then sets rates of repayment and the term of the agreement. An economics graduate for example, would be required to pay 3.38 per cent of her income for 100 months; a science graduate 3.68 per cent, for 104 months.

That seems a little rudimentary. For one, the Census Bureau is forced to categorise subject matters in a way that makes sense for their purposes, but may not make sense for Purdue’s. Two, the data collected by the Census Bureau is self-reported. And three, there are about 60 universities and colleges in Indiana alone. Even if the data is able to identify the average earnings of an English Lit major coming out of Indiana, individual earnings for the Purdue graduate could be quite different from someone graduating from Crossroads Bible College.

The frustrating thing is that the better data needed to achieve a more refined solution exists! Personal income tax data overcomes all of the pitfalls above, providing an accurate account of each individual graduate’s lifetime earnings, by institution and specific programme (not just broad discipline).

The even more frustrating thing is that it doesn’t really exist at all. It’s out there in chunks; unmatched, and collected in a way that can’t necessarily be utilised for an exercise like this.

Researchers at University of Ottawa’s Education Research Policy Initiative (ERPI) have shown how powerful such a data set could be. ERPI has built a data set that tracks the individual tax records of some 620 thousand graduates, from across six Ontarian higher education institutions and 48 disciplines over a seven year period (somewhat similar to BLADE). And from this they are able to extract much more meaningful — and specific — income profiles and distributions.

The findings aren’t revolutionary per se:

Consistent with the existing literature, the current analysis shows that STEM fields generally earn higher post-graduation earnings than non-STEM fields. However, within each field of study, the trajectory of post-graduation earnings can vary widely across departments. While some departments within the same field have similar outcomes, which justify classifying them under a single field of study, others do not…

Moreover, some non-STEM fields such as economics and political science have comparable earnings with STEM fields, while for some STEM fields such as biology, the post-graduation earnings are more similar to social sciences departments than to other science departments.

But from an asset manager’s perspective, data like this would be invaluable when trying to optimise a portfolio.

It’s worth noting that that it hasn’t been all smooth sailing for ISAs, with some institutions attracting quite a bit of heat for misinformation (a problem that could also be solved with better data). Nonetheless, Purdue’s ISAs at least, have been quite successful and are already providing a return of between 5 and 7 per cent. And as more flexible and online learning options enter the market, ISAs could become a real game changer for higher education finance.