Spatial Equity at the GSD: Balancing Budget, Space, and People

We started this project with one core question: in light of many students not having access to desks, how do we make space at the GSD, and Gund Hall in particular, work for everyone? Based on research, interviews, and a close examination of the GSD's resources, we realized that at its core the problem we were grappling with is a resource problem; and the question morphed into a much more general one about the financial health of the institution as a whole.

The core issue is described by Wyatt Roy's analogy below:

Gund Hall, the GSD's home, was designed in the 1960s, with construction beginning in 1969 and the building completed in 1972, around an intended population of 400 students and 70 faculty. The spatial constraints we are feeling today are the result of growth that has far exceeded the building's original intent. This project uses financial modeling to demonstrate how right-sizing the institution, by bringing it closer to its original intent, would both alleviate space scarcity and balance the budget in a manner that can be sustained well into the future.

Against a design intent of 400 students and 70 faculty, today's GSD runs at roughly twice and three times those numbers. The school has more than tripled its teaching corps relative to the original plan while enrollment only doubled. While the GSD has added square footage to its total footprint by acquiring additional buildings, the fact is that most academic energy is still concentrated in the main building. Furthermore, there's a case to be made that this expansion has only exacerbated budgetary constraints.

Being oversized is not only a matter of crowding; it is a structural cost problem. Each student costs the school about $87,000 a year to educate but pays about $39,000 in net tuition, so every seat is subsidized roughly $48,000 from the endowment and gifts.

Adding students widens that gap rather than closing it. From AY15‑16 to AY24‑25 expenses and revenue have grown nearly in lockstep, at about 3.4% a year. Where has this additional revenue come from? Over the past decade almost all of the GSD's revenue growth came from just two sources, net tuition and the endowment draw, about $11M each; gifts and sponsored research actually shrank.

Critically, tuition did not grow because of enrolling additional students. It was price: net tuition per student rose about 3.5% a year, well ahead of inflation.

With this in mind, holding the population steady is not an option either: if the GSD stays at its AY24‑25 population of 888 students while expenses continue to inflate at 3-4% a year, the budget will either slip into deficit within a few years or tuition will have to keep rising, putting it even further out of reach of many students. Hold tuition affordable and the school runs deficits; raise it to keep pace and it prices people out.

We already see this squeeze playing out in the data over the past decade, with the school paying for its overgrowth by making each seat steadily more expensive. The problem is therefore the structure of the institution, the ratio of people to resources, not the pace of growth, and shrinking is the one move that raises resources per student instead of chasing costs.

The counterfactual is a small financial model built from primary sources: Harvard GSD Fact Books and financial reports. A large share of the work happened before any modeling at all, in the calculation and reconciliation needed to turn messy, inconsistent, year-to-year sources into the single consolidated table it all rests on.

Each step is a committed script that writes a SQL table and a report; the whole suite regenerates from the database with one command. Nothing is hand-tuned.

The GSD's endowment sits around $675M and has grown roughly 5% a year net of payout, and that compounding is the whole game. What share of the budget its payout can cover depends on how big that budget is, and the budget shrinks not by enrolling fewer students but by employing the smaller faculty and staff a 400-student school actually needs. Two bounding cases fix the range, both inflating about 4% a year (HEPI): held at today's size, the payout climbs only from about 39% to 43% of the budget; cut to the design-intent size, it covers 39% rising to about 95%. By the end of that glide the endowment payout and gifts alone exceed the smaller budget, so the school would run a surplus even if net tuition held flat or fell to nothing. Any realistic path sits between the two bounds.

Faculty and staff salaries are the bulk of the budget, so the number that actually moves the finances is how many people the school employs, not how many students it enrolls. Shrink faculty and staff in proportion as the school right-sizes, hold the building and core admin, and the operating surplus climbs as payroll falls faster than the tuition a smaller student body brings in. At today's 401 faculty and staff a conservative 3% endowment decade leaves the school about $20M underwater; the surplus does not clear zero until faculty and staff fall to roughly 141 and 108 (a student body near 550). And this is not a problem the endowment can grow its way out of: even at the expected 5%, keeping today's 401 faculty and staff runs about $11M underwater, and headcount has to fall to about 315 (roughly 700 students) just to break even.

The recommendation is therefore a staffing target, not an enrollment one. Breaking even under the pessimistic 3% case takes faculty down to about 141 and staff to about 108, but sitting exactly on breakeven leaves no cushion for a bad decade. So the plan cuts a step deeper, to about 131 faculty and 101 staff, which clears roughly +$2.3M even at 3% and about +$11M at the expected 5%. That surplus is both the margin of safety and the room to hire some positions back, which is what keeps the cut generous rather than austere. Going further, toward the 400-student design intent and its 102 faculty, would throw off more still, but every step beyond this point comes out of faculty positions; protecting those is why the recommendation stops here rather than at the floor.

Students come down in proportion too, and that is what protects the experience rather than eroding it. Holding the faculty-to-student ratio keeps each student's share of teaching attention constant, and thinning the student body de-crowds Gund Hall, so the students who remain get more space and resources, not less. Matching about 131 faculty to roughly 515 students holds that ratio at a school the institution can actually sustain. Right-sizing is not a budget trick played on students; it keeps the education intact and makes it affordable.

Put revenue and expenses for the recommended plan on the same ten-year glide. As enrollment falls to ~515, net tuition shrinks, but the endowment keeps growing, so its draw rises as a share of expenses and overtakes tuition early, in AY29-30. That share keeps climbing, and against ~4%/yr cost inflation it reaches about 64% of the budget by year 10. The school runs a small, widening surplus the whole way, room it can put toward key programs rather than just staying solvent.

That shift is what buys the school room to choose. The more of the budget the endowment covers, the less it depends on revenue it has to chase, and the fewer reactive decisions that dependence forces: enrolling more students than Gund Hall was built for to book their tuition, shaping research around what sponsors will fund, or deferring upkeep when cash is tight. The more of the budget the endowment covers the more pressure is taken off and the more freedom is gained to set admissions, research, and facilities by mission instead of by cash flow.

Cost per student actually rises, from about $87k to $149k, because the fixed costs (Gund Hall, core admin) would now be spread over fewer students; right-sizing is not a per-student efficiency play.

The operating surplus per student does climb as the school shrinks, and at the recommended population of ~515 students it reaches about $21,000 by year 10, roughly 55% of today's net tuition per student (~$38,964).

Right-sizing generates solvency and a real but modest cushion per student, not a windfall. Shrinking the school even more would actually throw off more surplus and free up even more space per student, but this is a reasonable middle ground since the school has added additional programs and capabilities since Gund was built.