Stealth Edits and Clarity on Funding of New Harvard and MIT Nonprofit

A lot of the online discussion about the 2U acquisition of edX over the past few days has centered on the new nonprofit that will be created and governed by Harvard and MIT out of the proceeds of the sale. “How I’d Spend $800 Million” and “Some Ideas for the New Nonprofit Funded by the Sale of edX” are two examples from IHE. I think it’s worth understanding some of the details about this nonprofit’s funding.

Some of the confusion comes from Harvard and MIT messaging, as many of the details on the nonprofit have yet to be decided and the public messaging is changing.

I tweeted (still not sure if that should be a noun or verb) on June 30th that in a Harvard Magazine interview with provost Alan Garber, it was not clear how much of the $800 million would go into the nonprofit’s operating funds. The key paragraph:

Asked how much of proceeds from sale would be used as endowment, and how much allocated to current use in support of the stated aims of the new nonprofit, Garber said that a final decision had not been made, but he expected that “a large component of it will be endowment.”

MIT’s FAQ was a little more clear on the subject but not on the question of endowments vs. operating funds.

10. Will MIT profit from the transaction?

No. Except for funds edX will use to repay a recent line of credit from MIT and Harvard, the full proceeds from the transaction will flow directly to the nonprofit the universities will govern together, and not to either university. Over the years, MIT and Harvard have contributed $80 million total ($40 million each) to edX; they will not recoup those funds from the sale.

Because edX is a public charity, the proceeds from its sale can only be distributed for a purpose consistent with edX’s mission, not to compensate those who contributed to the nonprofit.

Clarity Needed

These statements were concerning, but David desJardins noted in reply that this would be “‘endowment of the nonprofit’, not Harvard’s endowment.” That was an important clarification, but initially I couldn’t figure out where he got this different reading of the interview until I saw that Harvard Magazine edited the article in question without any record of the changes. That same paragraph now reads [changes in bold]:

Asked how much of the proceeds from the sale would be used as endowment for the new nonprofit, and how much allocated to current use in support of its goals, Garber said that a final decision had not been made, but he expected that “a large component of it will be endowment.”

The MIT FAQ has also been updated on this subject, with additional information thrown in [changes in bold]:

10. Will MIT profit from the transaction?

No. Except for funds edX will use to repay a recent line of credit from MIT and Harvard ($15 million from MIT and $15 million from Harvard), the full proceeds from the transaction will flow directly to the nonprofit the universities will govern together, and not to either university. Over the years, MIT and Harvard have contributed $80 million total ($40 million each) to edX; they will not recoup those funds from the sale.

Because edX is a public charity, the proceeds from its sale can only be distributed for a purpose consistent with edX’s mission, not to compensate those who contributed to the nonprofit.

It’s good to see that the message is coming together with more information, but it would help if Harvard Magazine in particular would note any material edits to such a newsworthy document.

Two Other Notes

  • Those twin $15 million lines of credit are a significant part of the story. As many have noted, edX was losing money every year. It is clear now that Harvard and MIT agreed to keep edX alive with new cash infusions, but they were not willing to keep contributing donations, thus the lines of credit. I strongly suspect the search for a future home for edX was driven by the need for the lines of credit and the doubts that edX end the operating losses, and when the pandemic hit, it was clear that the timing was right to make a change.

  • Another key item to note, as Michael Feldstein did in a LinkedIn post, is that Open edX will be maintained by the new nonprofit, not by the edX public benefit corporation to be owned by 2U.

What Does the Nonprofit Have to Spend?

While the details have not been finalized, the funding for the nonprofit appears to be as follows (and credit again to David desJardin’s estimates as key source, along with a few other anonymous sources):

  • $800 million will go from 2U to Harvard and MIT;

  • $30 million will be used to pay off the two lines of credit;

  • No money will be sent to the edX institutional partners that provided seed funding;

  • A large majority of the funds will be placed in the nonprofit’s new endowment, and invested by money managers;

  • Perhaps $50 million to be allocated for annual operations of the nonprofit;

  • Even without investment returns, this should fund the nonprofit for at least 15 years if the operating estimates are close, but if the investment returns are close to 7 or 8% per year, those returns could fund most of the operations without drawing down the endowment very much.

Treat this is best guesses at this stage and not as definitive.

The Open edX platform itself needs heavy redesign to get up to speed in terms of usability and useful data to support the stated research mission. This Class Central post by Pat Bowden compared Chuck Severance’s Programming for Everybody MOOC on Coursera, FutureLearn, and edX (platform monogamy is not his thing, apparently) and described some of the edX user interface challenges.

Comparing the same course on different platforms emphasized the learning experience of each provider and helped me pinpoint what I like and dislike in each.

On Coursera, the left navigation sidebar lists each week’s worth of course material, allowing you to keep track of your progress. On FutureLearn, a To Do page serves the same purpose. Both also include video durations, allowing learners to estimate the remaining course load.

By contrast, edX’s interface offers much less visibility into your progress. Each week’s work is divided into sections and each section into pages. You can’t tell at a glance how many pages are left to complete a section nor how many sections are left to complete the week.

Taking this course reminded me why I prefer Coursera and FutureLearn over the edX format.

Course design and scalability have also proven to be cumbersome on edX.

Michael pointed out another improvement needed for the Open edX platform in his blog post.

Despite the excellent work of a few researchers, and despite the rhetoric of the institutions at the time that edX was launched, one reason we have not gotten more and better research out of edX is that the platform, incredibly, was poorly designed for educational research. How did MIT build a platform for massive-scale learning in 2012 and fail to think about what sorts of educational data and metadata they would need to facilitate research? What does that tell us about the real priorities behind the initial push to production? It’s a mystery.

The point is that much of the nonprofit’s operating budget should go to rearchitecting and improving the underlying Open edX platform that will enable deeper research as well as support edX itself.

Overall, there is some real potential for this Harvard and MIT nonprofit, but I hope we get better and more transparent descriptions moving forward.

Update 7/6: Added bullet point on institutional partners and seed funding. Added point on returns on investment funding significant part of the operations.