Meet The Teachers

What is known about educators in San Francisco and how they compare to other districts

The California Department of Education (CDE) publishes surprisingly detailed data about staffing in schools and districts across the state. No names are public but it is possible to see demographic data, education level, and years of experience for over 360,000 employees across the state. We can also see all their assignments e.g. which courses they are teaching at which schools.

Unfortunately, CDE has not updated this data since 2018-19. For once, the pandemic is not the excuse. CDE has an ongoing project to update how all this data is collected and reported and the project is way behind schedule. Some data that they told me would be available by end of 2021 is still not available. When I last checked, a couple of months ago, they said they hope to update the staff demographic data by later this year but there’s still no target date for updated course enrollment information. Since I’ve already been waiting for more than one year, I decided to stop holding my breath and plough ahead with the older data. Although the precise numbers will be a little out of data, I believe the broad conclusions will still be accurate because everything in education changes slowly.

One note: all enrollment and staffing numbers shown below are for district-operated schools only. If the numbers differ from what you may see on other sites, it’s probably because I went to the trouble of excluding data on charter school employment and enrollment from the numbers in order to ensure the validity of the comparisons.

The data covers three types of employee:

• Teacher: “An employee of the school district who holds a position requiring certification and whose duties require direct instruction to the pupils in the school(s) of that district.”

• Pupil Services: “An employee of the district in a position requiring a standard designated services credential, health, development credential, or a librarian credential and who performs direct services to pupils (e.g. counselors, guidance and welfare personnel, librarians, psychologists, etc.).”

• Administrator: “An employee of the district in a position requiring certification but who is not required to provide direct instruction to pupils or direct services to pupils.”

Other district employees (IT staff, custodians etc.) are thus not covered in the data. 

Staffing Levels

To compare staffing levels across districts, we need a metric that adjusts for the enormous differences in enrollment across districts. My first choice was a conventional student-teacher ratio but I rejected it because ratios are not additive which makes them hard to use (SFUSD has a 15.4:1 student-teacher ratio, a 97:1 student-pupil services ratio, and a 169:1 student-administrator ratio which combine, very unobviously, to an aggregate 12.3:1 student-staff ratio). I opted instead to measure staffing in terms of employees per 1000 students. SFUSD employs 65 teachers, 10 pupil services staff, and 6 administrators per 1000 students for a total of 81 staff per thousand students.

SFUSD employs more staff per student than any other district in the Bay Area. Most districts are under 60 staff per thousand pupils compared to SFUSD’s 81. SFUSD even employs more than Palo Alto which has the most income of any school district in the state.

If we extend the comparison set to the largest school districts in California, we can see that SFUSD employs the most teachers per thousand pupils, the most pupil services staff per thousand, and the third most administrators (behind San Bernardino City and Oakland) per thousand pupils.

SFUSD likes to benchmark itself against Long Beach Unified because Long Beach is a big urban district whose majority Latino student population scores highly on standardized tests. Long Beach employs 51.1 staff (43.1 teachers, 4.6 pupil services, and 3.4 administrators) per thousand pupils. That’s nearly 40% fewer than SFUSD’s 81.5 per thousand. In fact, SFUSD employs more teachers than Long Beach (3424 vs 3141 teacher FTEs) even though it has 20000 fewer students (52468 vs. 72935).

Some readers may be thinking that this comparison is a bit unfair to SFUSD. In most counties, there is a clear distinction between the school districts and the County Office of Education (COE). In Alameda, there is one COE and around 20 separate school districts. In San Mateo, there is one COE and 23 separate school districts. The COEs have completely separate staffs and budgets. In San Francisco county, there’s only one school district and the boundary between SFUSD and SFCOE is blurry. The COE is run by the district superintendent and supervised by the district board of education.

A COE has clearly defined responsibilities that include educating students who are incarcerated or on probation or who have been expelled from other schools. COEs thus serve only a tiny fraction of students: 0.1% in Alameda, 0.2% in Contra Costa, 0.3% in San Mateo, 0.6% in San Francisco and Santa Clara and 1.0% in Marin. [An interesting question about which I have no data: why do these rates vary so much from one county to another?] But COE students are expensive to serve, requiring far more staff than normal school districts. San Francisco COE is at 151 staff per thousand students, far below the others which range from Alameda’s 209 per thousand students to San Mateo’s 524 staff per thousand. Why is SFCOE so low? It’s possible that SFCOE is able to run super-efficiently because it’s so closely associated with SFUSD. It’s also possible that there are people included in the SFUSD headcount who should be in the SFCOE one. But even if SFCOE had twice the staff it does, with the extras coming from SFUSD, this would only reduce the SFUSD headcount from 81.5 to 80.6

A better reason to be skeptical about this analysis is that I’m not convinced of the accuracy of SFUSD’s data. Read on to find out why.

Average Age

The average age of SFUSD’s teachers lies smack in the middle of the distribution of average ages across all districts.

But averages sometimes conceal interesting data. 9.9% of all SFUSD’s teachers were aged over 65, far more than the 0-3% range that is more typical.

9.9% is such a startlingly high figure that I dug into it a bit. It translates into 337.8 teacher FTEs coming from teachers over 65 but 258 of those 337.8 came from teachers assigned to the District Office rather to school sites and 90% of those were given the assignment code 6014 which means “a staff member permanently employed by a local educational agency that is temporarily assigned to teach classes while a permanent teacher is either out on leave or is yet to be hired.” (emphasis added)

I don’t know enough about the inner workings of SFUSD to know whether the district does indeed maintain a legion of centrally employed geriatric substitute teachers or whether this is some strange error in the district’s reporting. If the latter is true, and the legion of geriatric permanent substitutes does not exist, then this would of course reduce the staffing differences between SFUSD and other districts.

It wouldn’t eliminate the gap, however. There is other evidence that SFUSD employs more teachers than most other districts. In a post a few months ago, I looked at actual class sizes in California schools and found that SFUSD’s average elementary school class sizes (particularly in grades K-3) are, next to Palo Alto’s, the lowest in the state. To have small class sizes, a district must employ more teachers.

Years of Experience

The average years of experience of SFUSD’s teachers lies smack in the middle of the distribution of all districts.

The legion of centrally employed geriatric substitute teachers, if it exists, is boosting SFUSD’s average because some of them apparently have more than 50 years of experience.

Gender Mix

This chart shows the percentage of teachers who are female for all school districts with more than 2000 students. The districts are colored according to the grade span of the school. It’s obvious that, on average, elementary school districts have a higher percentage of female teachers than high school districts with unified districts sitting between the two. That probably matches your intuition but it is still interesting to see it borne out so starkly in the data.

SFUSD (67% female) is a bit below the median of all unified school districts but not so far as to raise any flags.

Race and Ethnicity

Here’s the racial and ethnic composition of teachers in each of the Bay Area Unified School Districts with more than 5000 students.

About half of SFUSD’s teachers are White, which is a lower percentage than in every other district bar Oakland. Over a quarter (27%) are Asian or Filipino, higher than in any other district. About 15% are Latino, a higher percentage than in 19 of the other 26 districts shown. 5.5% are Black, a higher percentage than in 18 of the other 26 districts shown.

Evaluating the racial and ethnic makeup of educators is tricky, partly because of the nature of the topic but also because it’s unclear what the goal should be:

• one stream of thinking is that students need to see people like themselves among the teaching body in order to reach their potential. Such thinking will lead people to compare the composition of the workforce with the composition of the schools.

• another stream of thinking is that hiring should show no evidence of bias. This will lead people to compare the workforce composition with that of the adult population. But should the comparison be with the city of San Francisco or the whole Bay Area from which people might commute? And should the comparison be to the adult population as a whole or to the adult population with college degrees, the minimum typically required for a teaching job?

Here’s how the composition of SFUSD teachers compares with the SFUSD student body, the adult population of San Francisco, and the population eligible to be teachers, namely adults with a Bachelor’s Degree or higher¹.

Whatever point of comparison you prefer, SFUSD has fewer Asian teachers than would be expected. The share of teachers who are Latino and Black roughly matches the share of San Francisco adults who are Latino and Black, which is significantly higher than their share of the college-educated adults in the city and significantly lower than their share of the student population.

Up Next

In coming posts, I’ll look at how teachers are distributed around the schools in the city and delve further into teachers employed out of the District Office.

1. Strictly speaking, this is the population of adults 25 and over with a Bachelor’s degree, which is the only level that the Census Bureau publishes an ethnic education breakdown (see ACS Table S1501). If I were not so lazy I would make some estimate that would allow me to exclude graduates who are past retirement age. Doing so would lower the proportion who are White to something closer to 50%.