Which High Schools Would You Close?

Why closing even six schools may be too few

Next month, SFUSD will announce which schools it wants to close. The process so far has been very deliberate. The public face of it has been focused around the selection of the ten criteria that will be used, and the precise weight that will be given to each criterion, in the calculation of a school’s Composite Score. Unfortunately for my purposes, most of the criteria are poorly defined and/or rely on internal data that is not publicly available¹. Since it is not possible for anyone outside the district to reproduce even the inputs to the district’s calculations, it is not possible to predict which schools will have high Composite Scores and which will have low ones.²

Not having the crutch of a Composite Score, I’ll attempt to construct a list of schools which might be vulnerable using available data and hints about the district’s thinking process. My list will have some obvious problems that I will point out. It is not intended to be either a recommendation for what SFUSD should do or a prediction for what SFUSD will do. My goal instead is to give you a sense of how difficult it’s going to be for the district to make its decisions.

I’m going to start with the high schools for two reasons:

1. There are fewer of them, which makes it easier to write about the case for or against an individual school

2. There is more overcapacity in high schools than in elementary or middle schools.

Let’s start with a review of the facts.

Enrollment

In 2023-24, the district had 15,269 students distributed across 17 high schools. The largest school, Lowell, has more students than the eight smallest ones combined. The four largest schools contain more than half (55%) of all students. The seven largest schools contain over 75% of all students.

And this is the case even though “SFUSD is limiting the number of seats at larger, highly-requested schools to ensure enrollment at smaller schools”³. If they were not limiting seats, the four biggest schools would have up to 65% of students.

Geography

Ideally, I’d start here with a map of where students actually live. I have been very surprised that a map showing the number of students per census tract was neither produced by the district, nor sought by the District Advisory Committee, at any point during the whole process. It would seem to be relevant. Below is a map showing the distribution of children in the city, which is the closest approximation I could produce.

A map that excluded the under-fives and private school students would probably show an even greater skew towards the south of the city.

The four largest schools are all near either the western or northern edges of the city. Schools in the center, east, and south are all smaller and contain higher percentages of socio-economically disadvantaged students, students with disabilities, and English Learners. The exception is SOTA, which has by far the fewest disadvantaged students, and practically no English learners.

Building Condition and Capacity

For the District Advisory Committee, the district estimated the facility capacity for each of its high school buildings by counting the number of classrooms, assuming 30 students per class, and assuming that 15% of classrooms would be unoccupied at any time. The district acknowledged that this was “an imperfect method for assessing facility capacity.”

One reason it was imperfect is that it often resulted in estimated capacities that were lower than the actual number of students enrolled⁴. Lowell has had more students than its stated capacity of 2,448 every year for which I have data. Wallenberg exceeded its stated capacity of 587 for forty consecutive years from 1982 to 2022. Galileo’s stated capacity was 1,709 but it had over 2,000 students every year from 2005-2012. Lincoln averaged nearly 2,450 students for a decade in the 2000s, compared to its stated capacity of 2,219. The stated capacities are, however, useful for the chronically under-enrolled schools. Historical enrollment records would underestimate their capacity. Here’s my calculation of the actual and spare capacity at each school.

By my calculation, the actual capacity of SFUSD’s high schools is 23,911 students, substantially above SFUSD’s own estimate of 22,211. The map below shows the spare capacity at each location. You can mouseover to see the details for each school.

Facilities Condition

The disc color in the map above indicates the facilities condition. A couple of years ago, the district undertook an extensive assessment of the condition of its facilities and calculated a Facilities Condition Index that measured how much work each building required. The lower the FCI, the less work is required on a building.

Building Condition, as measured by the FCI, is one of the ten criteria SFUSD will use but it has a weight of only 5.2% in the Composite Score calculation. Nevertheless, I doubt SFUSD will propose closing a building that is in excellent condition while keeping open a similar-sized building that is is poor condition. Such wanton waste of public money might attract negative publicity and damage the chances of the Facilities Bond measure that is on the ballot in November.

Balboa and Mission are in the worst condition of any schools in the district. Each requires about $80m of work. Galileo requires about the same investment in dollar terms but it’s a much bigger campus so the cost per student there is much less. If all three survive the impending closures, it’ll indicate that the district has not learned any lessons about focusing its investments.

The campuses in the best condition are two of the smallest: Independence and Downtown. I must confess that I do not understand why Independence, a school where students are on campus only 5-15 hours per week, needs its own building nor why its facilities needs were addressed before those of schools that are filled every day with students.

The Target

The district has been vague so far about the size of the planned reductions. Let’s do a bit of Math. The district has said its goal is to operate its buildings at 90-95% utilization⁵. Given the current enrollment of 15,269, that means it needs space for between 16,000 and 17,000 high school students. It has, depending on which capacity estimate you use, space for either 22,211 or 23,911 students, which means it needs to reduce capacity by somewhere between 6,000 and about 7,500 students. That’s a lot. You could close the nine smallest schools and still have ample room for everybody in the eight that remain.

It won’t be that simple. The district has promised, and board members have frequently said, that they are going to ensure that no community is disproportionately affected by the school closures. Of course, no one has defined in advance what disproportionate is which gives everyone the freedom to claim that any plan they dislike is disproportionate. If you closed the nine smallest schools (i.e. O’Connell, Marshall, Wallenberg, International, Independence, Academy, Jordan, Downtown, and Wells), you’d be affecting 7% of Asian students but 32% of Latino students. That might struggle to pass. It might also be illegal: every district of any size has at least one continuation school so I suspect having one is a requirement.

Some Speculation

Let’s start from scratch and see if we can imagine⁶ the debate that’s going on within SFUSD about which schools to close. I’m going to come up with a list of schools to close and then point out the problems with my list. Remember: this list is neither a recommendation nor a prediction. My intention is not to advocate for or against any particular school but to point out how difficult it’s going to be for the district to balance its constraints and make its decisions.

I’m not going to refer to any academic measures in producing this list. We have no data about how good or bad a job the high schools are doing about teaching their students. It’s true that we know the percentage of 11th graders at each school who pass SBAC tests and the percentage who pass AP tests and we also know how many are chronically absent, and how many graduate on time. The problem is that we have no way of adjusting these figures for the quality of the incoming students. Students who attend Lowell excel - but they also excelled in middle school. There’s no way to calculate how much value Lowell added. Conversely, students who choose to attend Jordan are doing so precisely because they don’t think a conventional high school would suit them. Few of them meet standards in 11th grade but I’m willing to bet that few of them met standards three years before in 8th grade. If they were thriving in 8th grade, they probably wouldn’t have chosen Jordan. The only way to determine if Jordan is a good school is to evaluate if the students who go there do better than similar students who attend other schools. We don’t have the information to do this.

The district has given some clues to its thinking. It has pointed out that “when we assign families to schools lower on their [enrollment application] lists, they are more likely to leave SFUSD⁷”. 85% of those who receive their first choice school ultimately enroll in SFUSD compared to only 61% of those who receive their fourth choice. Giving more families their top choice school appears to be a goal.

It has also pointed out that “in every zip code, at least 40% of applicants list Lincoln, Lowell, or Washington as their top choice” (emphasis added). This would seem to imply that those three schools are safe.

Let’s look at the next largest. Galileo is an interesting case. It has nearly as many students as Lincoln and Washington and it’s the fifth most popular among families applying for 9th grade. It’s also a bit of a distance from any other high school. Those are strong arguments for keeping it. On the other hand, it does need $80m worth of work. Let’s keep it for now and see where we end up.

The McAteer campus⁸ was only supposed to be a temporary home for SOTA until a new home was built on district-owned land near the Civic Center. A new Arts school was the flagship project of the district’s last Facilities Bond but that bond money was squandered on…stuff. To announce the closure of the Arts school right before the next Facilities Bond vote might scupper its chances. Let’s let sleeping dogs lie and keep it where it is.

The Academy - San Francisco@McAteer, to give it its awful mouthful of a name, was created in order to occupy the unused portion of the McAteer campus. Enrollment was stable in the range of 320-360 every year until 2023-24 when it suddenly dropped to 265. It’s unclear what caused that decline, but since we’re looking for schools to close, that might provide a reason. However, if we’re keeping SOTA open, it makes sense to continue to use the rest of the campus. Independence could be moved in to the space Academy vacates, thereby freeing up Independence’s building in the Sunset.

Balboa, Jordan, Burton, and Marshall form a little cluster.

• Balboa has the most students of the four and the most convenient location for student commutes. Among families applying to 9th grade, only Lowell and Lincoln receive more 1st choice applicants. But the campus requires a huge amount of work.

• Jordan is on the site of the former Luther Burbank Middle School that was closed in 2006. Jordan is intended to be a small school so it can’t really be faulted for only occupying 1/6 of its campus. But it can be faulted for never approaching its intended capacity of 400 students. It hasn’t reached 300 since 2007 and it’s been below 240 the last six years. It’s impossible to justify keeping such a small school in such a big campus. The only question is whether to close it or colocate it with another school.

• Burton is substantially bigger than Marshall but both have campuses in good condition that they occupy only a fraction of. Burton used to be twice the size of Balboa but Balboa has been the bigger every year since 2006. Marshall used to have over 1,000 students but it has averaged around its current 475 for the last decade.

Do you keep open the school that is doing best (Balboa) or the school with the most space that needs less work (Burton)? For now, let’s prioritize the money aspect and keep Burton open. We’ll co-locate Marshall and Jordan on Marshall’s campus⁹. Balboa students get to choose whether to attend Burton, Marshall, or Lincoln.

There’s another cluster of schools in the center / east of the city.

• Mission, like Balboa, requires a huge amount of work but closing it would set off a firestorm. It has the most Latino students and the most Black students of any school in the city and it’s been amazingly successful at getting its students into Berkeley. However, because it’s in such bad physical condition, SFUSD won’t want to send students from any closed school there.

• SF International and O’Connell are quite different: one focuses on newcomer students while the other stresses technical education. But there’s more than enough room in O’Connell’s campus to accommodate both. Let’s have them co-locate.

• Neither of the two continuation schools, Downtown and Wells, has had more than 200 students since 2012 and both buildings have a capacity of 434 so the district only really needs one. Downtown’s building is in better condition but I don’t know which location would be more convenient for the students. Continuation schools are for kids who’ve already dropped out once so making it easy for them to show up is critical.

• Wallenberg has by far the most consistent enrollment of any high school in the city. Its enrollment was in the mid 600s literally every year from the 1980s until 2022. It’s been 550 the last two years. It’s a perfectly fine school but, a bit like Academy, I need to find places to close. Let’s add this to the list.

So far, my list includes the closing of Balboa, Wallenberg, Academy, and one of the continuation schools and the moving of Independence, Jordan, and SF International to unused spaces in other schools. That’s six buildings the district would no longer have to maintain and upgrade.

You may think that my arguments for closing some schools and retaining others have been weak. And you’d be right. But let me point out some concrete issues with my list:

• The list may not be bold enough. It only reduces the capacity by about 5,200-5,300 which is less than the 6,000-7,500 I calculated earlier was needed.

• One of SFUSD’s self-imposed guidelines for how it’s going to handle school closures is that students will not be transferred to a school that has a worse FCI than the school they’re leaving. If you close Wallenberg, you can’t transfer the students to either Mission or Galileo because both are in much poorer condition than Wallenberg. Washington is in better condition but they won’t all fit there. Moving Independence to the McAteer campus might also be problematic for the same reason.

• The schools on my list serve 3,041 students, which is 20% of all students. But those affected are 24% of Latino students and 25% of Black students, compared with 16% of Asian students and 15% of White students. It affects 23% of disadvantaged students, 26% of English learners and 26% of disabled students. Is that inequitable? If so, it will be hard to fix. It is only as close as it is because I had Balboa and Wallenberg (which have more Asian and White students) on my list instead of, say, Burton or Marshall or O’Connell.

Recall from the pie chart earlier that the four biggest schools (viz. Lowell, Lincoln, Washington, and Galileo) account for 55% of all students. 74% of all Asian students but only 36% of all Latino students are enrolled in one of those four schools. If those four schools are safe, then the schools that are vulnerable contain just 26% of Asian students but 64% of Latino students. This make it mathematically impossible for them to be equally affected by the closures.

Suppose we bite the bullet and decide to close Galileo instead of Wallenberg. That’ll bring us within our capacity target and we’ll save $80m in facilities costs, which is not a small thing. But where can those 1,700 Galileo students be accommodated? Given the map of where children live in the city, it’s possible that relatively few of Galileo’s students actually live near Galileo. The reprieved Wallenberg would only have room for about 100. Another few hundred could go to Washington but that’s literally across the city. Mission has space but a parent there told me they offer no Honors/AP classes before 12th grade so it would be seen as an inferior alternative by many Galileo parents. And it’s in even worse condition than Galileo so the district wouldn’t want to transfer students there anyway. O’Connell is academically weaker again. Lowell, Lincoln, and Burton are a very long way away.

It’s not easy, is it? I don’t envy the people who have to make the decisions.

Questions to Ponder

My list cut capacity by over 5,200 and affected 3,000 students. Is that too much or not enough?

If you could only keep one of Balboa or Burton, would you keep the one that is the more successful (Balboa) or the one that is larger and needs the least work (Burton)?

What is equitable? What would make any proposed list more equitable?

1. It’s no longer possible to calculate something as simple as the distance to the three nearest schools. It used to be straightforward but now “density” of students living close to a school is going to be included in the calculation in some completely unspecified way. Is it the density of students currently attending that school? Or density of students attending any SFUSD school who happen to live close to the school in question? How will density be measured? How will it be factored in to a distance calculation?

2. I wonder if the district will release enough information to make it possible to audit the Composite Score calculations or if it will remain a black box.

3. Slide 48 of this Enrollment Center presentation to the District Advisory Committee

4. A major reason is the 30 students per class element. All the larger high schools average more than 30 students per class, sometimes as high as 35-37. A student could go through four years at some schools and never encounter a class smaller than 30.

5. See slide 29 of June 25th board presentation.

6. I have not spoken with anyone in SFUSD in over a year and the district rejected my application to be a member of the District Advisory Committee. I am thus not privy to any inside information about which schools might be on the district’s list.

7. Slide 41 of this Enrollment Center presentation to the District Advisory Committee

8. Eugene McAteer High School closed after the 2001-02 academic year. In its heyday in the mid-1980s, it had over 2,100 students.

9. A co-founder of Jordan serves on the Board of Education, which might make the district reluctant to close it.