First, an apology:
I was going to wait to get more information about square footages, capacities, and how they are calculated in DCPS before writing about them and utilization rates—but I was unable to resist delving into what recent data I have in my possession on this topic.
Second, another apology: What recent data I have in my possession on this topic is inconsistent and apparently contradictory. Here it is:
–A spreadsheet from DGS, showing square footages of DC-owned properties, was publicly posted and produced for the council in spring 2026 in response to oversight questions. It is unclear what is the date of this spreadsheet. For the sake of convenience, let us call this dataset Alpha.
–A spreadsheet from a recent FOIA production I received from the deputy mayor for education (DME) shows square footages for DCPS and charter schools. The metadata says this sheet was created in 2017, but the dates listed in it go to 2024, so it seems like it was last updated (by someone somewhere) in 2024. The source of measurements for DCPS and charter buildings owned by DC is noted with those agencies or as “FCA” (presumably as part of facilities condition assessments done by DGS), while the source of measurements for other charter buildings is noted as “PCSB” (presumably provided by those individual charter schools to the public charter school board). For the sake of convenience, let us call this dataset Beta.
–Another spreadsheet from a FOIA production I received from the DME shows square footages for DCPS buildings and charter school buildings owned by DC. The metadata says it was created in 2023, apparently by DGS (this supposition is supported by emails I received in the same FOIA production). For the sake of convenience, let us call this dataset Carla.
–A spreadsheet from the latest iteration of the master facilities plan (MFP) supplement (released in August 2026) was created by the DME and is publicly posted. Though it does not have square footages, it does have enrollments and capacities; similar datasets are created and posted by the DME annually. For the sake of convenience, let us call this 2026 dataset Delta.
(NB: Here is what I have received thus far in response to my FOIA requests for square footages, capacities, and their calculations, which includes datasets Beta and Carla and supporting emails. Although there are more spreadsheets of square footages than just Beta and Carla, those others seem to contain the same DCPS information and are not as comprehensive as Beta and Carla.)
Now let us look at what datasets Alpha, Beta, Carla, and Delta say about one school, Stuart-Hobson Middle School.
Recall that at a hearing in February 2026, Ward 6 council member Charles Allen disputed a DCPS estimate of Stuart-Hobson’s capacity. The DCPS representative said the school’s capacity was more than 600, while Allen (a parent there at that time) maintained that the school was bursting at the seams with an (audited) enrollment that year of 493.
The school’s renovation was completed in 2015. Per this DGS website (see here if that website is nonfunctional), it received a 13,918 square foot addition, the entirety of which has accommodated the school’s arts programming. The capacity for the school renovation that was used by designers and the SIT team (of which I was a member) was 450 (see here for a document from 2011 about it). In fact, that 450 capacity number was listed for Stuart-Hobson on every MFP document from 2015 to 2020.
But sometime in 2020, something changed—and it was not the footprint or use of Stuart-Hobson. Rather, starting in 2020 the capacity of the school was listed in MFP documentation as 676. And it stayed at that number in official MFP documents until sometime in 2023 or early 2024, when it was reduced to 612, where it has remained since. (See here for a listing of Stuart-Hobson’s capacity numbers over time as published by the DME.)
So it is that per official DC records, Stuart-Hobson’s capacity increased by 50% in just a single year between 2019 and 2020—without anything else changing!
Now, this might be understandable if each school’s capacity measurements were clearly explained. But to date, I have neither found publicly posted, nor received via FOIA, any document outlining that, except in the most general way (ie see p. 31ff of the most recent MFP here).
Perhaps unsurprisingly, datasets Alpha, Beta, and Carla have very different values for Stuart-Hobson’s gross square footage, despite all being promulgated relatively recently by the same agencies behind the MFP:
Dataset Alpha: 73,451 square feet
Dataset Beta: 99,325 square feet
Dataset Carla: 114,720 square feet
Such apparently contradictory and inconsistent data is bad enough with just one school—but it goes beyond Stuart-Hobson.
Number Games
Eliot-Hine Middle School sits about a mile east of Stuart-Hobson. Dataset Beta shows that Eliot-Hine, is 194K square feet, while dataset Carla shows that Eliot-Hine is 214K square feet. As with the variability in square footage for Stuart-Hobson, this variability in records for Eliot-Hine’s size appears to have happened without anything else changing.
The Delta dataset shows that the capacity of Eliot-Hine is 544, while it lists the capacity of Stuart-Hobson as 612. (Remember that in both the Beta and Carla datasets, Stuart-Hobson’s square footage appears to be almost 100,000 square feet less than that of Eliot-Hine.)
So let us divide the square footages in datasets Beta and Carla by the latest audited enrollments to get space per student in each of these schools:
Dataset Beta:
Eliot-Hine: 194,195 square feet, 499 students (SY25-26); space per student: 389 square feet
Stuart-Hobson: 99,325 square feet, 493 students (SY25-26); space per student: 201 square feet
Dataset Carla:
Eliot-Hine: 214,987 square feet, 499 students (SY25-26); space per student: 431 square feet
Stuart-Hobson: 114,720 square feet, 493 students (SY25-26); space per student: 233 square feet
Admittedly, this analysis is not really fair:
The square footages here are gross square feet, and schools are properly measured in ways that account for their programming and specialized spaces (like self-contained classrooms), not just gross square feet. But using this measure baldly with either set of data, Eliot-Hine appears to have almost twice the space per student than Stuart-Hobson.
(NB: The disparity here would be even worse with dataset Alpha, which shows Eliot-Hine with 233K square feet and Stuart-Hobson with only 73K square feet. Because dataset Alpha is undated and seems to not account for renovations in the same way that Beta and Carla do, I didn’t use it for this purpose.)
We can see this oddness at another angle, if we peer into the latest published utilization rates of the two schools from dataset Delta:
Eliot-Hine: 499 enrolled students (SY25-26); capacity (2025) 544; utilization rate: 91.7%
Stuart-Hobson: 493 enrolled students (SY25-26); capacity (2025) 612; utilization rate: 80.6%
Utilization rates here were determined by dividing audited enrollments by the listed capacities in dataset Delta. Every year, in fact, the DME publishes this data using this simple division (remember, without any public explanation of how individual school capacities are generated).
For SY25-26, Eliot-Hine appears to be slightly over the ideal utilization rate of 85%, while Stuart-Hobson appears to be slightly under. (Note that this is the same school year in which a parent publicly stated that Stuart-Hobson was overcrowded.)
If we extrapolate from this data to get Stuart-Hobson to the (ideal) utilization rate of 85%, then its enrollment would have to be 520. Using dataset Beta, that means Stuart-Hobson’s square feet per student would decrease to 191. Alternatively, using dataset Carla, Stuart-Hobson’s square feet per student would be 221.
Now, if Stuart-Hobson were filled to its current published capacity of 612, then the square feet per student would decrease to 162 (using dataset Beta) or 187 (using dataset Carla). By contrast, if Eliot-Hine were filled to its current published capacity of 544, then its square feet per student would be twice as much as that for Stuart-Hobson using either set of numbers: 357 square feet (using dataset Beta) or 395 (using dataset Carla).
So it is that with twice the gross square feet per student at full capacity as Stuart-Hobson per official records, Eliot-Hine was considered fully utilized in a year (SY25-26) when it had just SIX more students than Stuart-Hobson (499 versus 493).
While it is possible that Eliot-Hine has many self-contained classrooms that would affect its space per student considerably, it is also possible that the gross square footages in these datasets and/or published capacity numbers are wrong. But it is literally impossible for anyone outside DC government agencies to know, given that there are no publicly available outlines of how individual school capacities and square footages are calculated.
For weeks now, I have tried to get someone somewhere in DC government (via multiple emails, phone calls, and FOIA requests) to explicate this mystery. I have gotten no answers.
I do know that the DME’s office did not provide a full response to my FOIA request for square footages for DCPS schools starting in January 2020 through September 2026. The data they provided was only from 2023 on—notably, skipping over the time when the DME’s reporting changed Stuart-Hobson’s capacity from 450 to 676. The DME’s office also neglected to provide measurement data it contracted the Perkins Eastman firm to collect for each DCPS school for the MFP released in March 2024. (Separately, I asked a staffer at the DME’s office for that Perkins Eastman data—and got no response.) In addition, DCPS and DGS have delayed responding to my FOIA requests for the same materials.
Thus far, I have been avoiding the quiet part—so let us go there now.
Utilization Rates Determine Closures
Student populations in DC are decreasing. Given that we already have thousands of unfilled school seats, that decrease means that DC will likely face school closures again in the near future.
Every year, the DME publishes data on school utilization rates, calculated by dividing capacity numbers by enrollments. But while DCPS capacities are determined by someone somewhere in DC government (with the actual calculations apparently not available for public viewing), charters self-assess their capacities. The DME also outlines plans for use of excess space–but only in DCPS.
Dataset Beta underscores the deeply anti-public nature of all of that, as dataset Beta contains DC-government-generated FCA measurements for charter buildings owned by DC. That means that someone somewhere in DC government knows the size of these charter buildings and could figure out their usage and capacities.
But no one in DC government does that.
As it is, charters’ self-assessment of capacities and square footages can never be a disinterested analysis. For instance, a desire for more space can easily be represented as a need, while a desire to downplay excess space can be represented as a nonissue. These are not insignificant matters. Just a few years ago, for instance, I pointed out how four former DCPS school buildings owned by DC and occupied by charters could hold an additional 1500 students—or what amounts to several schools. But the DME’s MFP reporting obscured that.
Taken together, all of this means that DCPS buildings are held to a standard for capacity and utilization that no DC charter building is—all the while the DME blithely publishes utilization rates for both sectors as if those numbers were derived in the same manner (and while official datasets have different measurements for the same buildings at the same times).
Over the years, DCPS has offloaded most of its so-called excess buildings. The few that are left are currently being used for swing space as modernizations continue. The agency has refused to explain publicly what it plans to do with those buildings once this round of modernizations is done.
All of this is a potential boon for charters, which are entitled to DCPS excess buildings by law. Charters have craved these buildings for years–and for good reason. Having purpose-built school buildings purposefully located in residential areas with kids that are then leased for $0 is a much better deal than finding a building somewhere on the private (and more expensive) real estate market. But while DC charters have benefitted for decades from this public largesse, the offloading of DCPS buildings has slowed considerably in the last few years given a lack of DCPS closures (though not for lack of trying by this DME).
So the near future for DCPS is not only the likelihood of closures of neighborhood, by right schools that the DME labels as underutilized (“just doing the division, ma’am!”), but also the conversion of those buildings to charters.
All of which is much easier when official stats make a bursting-at-the-seams DCPS school appear underutilized.
For me, as a former parent who worked on its renovation, the current capacity number for Stuart-Hobson is difficult to reconcile with reality. For years running, there was a drop-down, drag-out fight to get Stuart-Hobson renovated. Eventually, the school community was able to wrest a renovation of sorts. In keeping with that reluctantly granted wish, DCPS and DGS proceeded without a master plan and (unsurprisingly) without full funding for the work ultimately done.
So it was that, starting in 2012 and going through 2015, Stuart-Hobson’s renovation proceeded with students, teachers, and staff in the building the entire time.
Now, if renovating a school while it is being used AS a school seems crazy to you, rest assured it was. My child (along with hundreds of other children and adults) endured horrible things for years at Stuart-Hobson as a result: gas leaks; utility failures; classes held outside when the inside was unbearable; incredible levels of dust and noise; and at least one evacuation.
But this happened not merely because there was no master plan or full funding from the jump. It happened also because one thing was made very clear: If Stuart-Hobson was closed for renovation and the school community re-located to a swing space, the building would likely not remain a DCPS school.
No one thought to put all that on the plaque in the school’s lobby commemorating the building’s renovation and thanking all the elected officials who (begrudgingly) funded it.
Nor did anyone put on that plaque that because of the funding limits and the reality that school was in session for its renovation, the size of hallways, stairwells, and classrooms was preserved exactly as built a century ago.
This in turn ensured that Stuart-Hobson’s renovation was not designed to accommodate more students than the school had historically had capacity for–which is why 450 was the capacity number we worked with for the renovation.
This school year, about 535 students are enrolled at Stuart-Hobson. Per parent reporting, there are not enough lockers for all of them. This is hardly surprising in a building designed for fewer students. But with hallway, classroom, and stairwell space also sized for fewer students, it could be construed as a possible fire hazard.
(Yeah, no one called me back about that, either.)
So What Is The End Game Here?
I don’t know.
I do know that the DME’s omission in my FOIA production of the period when Stuart-Hobson’s capacity was inexplicably increased to 676 coincides with the moment in February 2020 when DGS and DCPS personnel went through Duke Ellington HS (a purpose-built arts building completed in 2017) to measure whether another high school could fit into it. (No, I am not kidding.)
That suggests a concerted effort in that period by DC government actors to revise DCPS capacities—but why, where, and how I don’t know.
I do know that the agencies responsible are not easily (or at all) releasing information about it, while promulgating data that is, on its face, inconsistent and apparently contradictory when it should be neither.
And I also know that none of this can be good for DCPS.
That is because without clear, public explanations for calculations of individual school capacities and reliably and consistently reported square footages, someone somewhere in DC government could construe anything about any DCPS school for any purpose.
Which for all I know may be the end game.
