The Impending Crisis in Special Education
A flood of diagnoses that threatens to drown SFUSD's Special Education programs
As most readers will know, the district finances are under adult supervision by the Department of Education’s Fiscal Crisis and Management Assistance Team (FCMAT). Separately, the district chose to engage FCMAT to analyze its Special Education program. The FCMAT team wrote a detailed report and made a presentation to the Board back in April.
It’s great that the district commissioned the report. All the publicly available special education data concerns the demographics and disabilities of the students and the type of educational setting in which they are placed. The report filled a gap by comparing San Francisco’s staffing and management practices with “industry norms” i.e. state averages. It went through each type of special education role and benchmarked SFUSD’s staffing guidelines for that role with state averages. It found that, for most job roles, staff in San Francisco had to support fewer students than similar staff in other districts.
The report also called out some areas where SFUSD’s management practices did not follow best practice. For example, individual sites were allowed a lot of leeway in deciding who to put in Special Education. In the presentation, the FMCAT analyst said:
what we learned from interviews that I think is really critically important for your
district to consider is that certain students have a disability and the school teams were concerned that they really didn't need special education but there wasn't other support available at the school. So those students were placed in special education when really if there were other interventions and supports available at the school they could have access to those instead.
The report had this to say about Resource Specialist Program (RSP) staffing:
Many district/county staff do not recognize that the resource program is overstaffed or the consequences of this problem. In addition, staff indicated that the Special Education Department has proposed RSP teacher reductions and attempted to move RSP teachers from schools with lower RSP student-to-teacher ratios to schools with higher RSP student-to-teacher ratios in the past, but these changes were not approved because of resistance from school staff or because the board of trustees did not approve the plan.
The industry-standard practice is to share a resource teacher between two schools when caseloads are low at both schools, but staff reported the district/county rarely does this.
The report also criticized SFUSD’s method for calculating paraeducator staffing levels (basically, SFUSD ignores 1:1 paraeducators in their formula but shouldn’t) and called out the overuse of 1:1 paraeducators.
“Many [districts] throughout the state have taken steps to remove the designation of 1-to-1 support because it unintentionally reinforces the concept of one adult assigned to one student…In 2024-25, the district/county has 416.44 FTE one-to-one paraeducators…. Although there is no industry standard for comparison, this is high for a district/county of this size ”
The report also highlighted problems with the district’s data management. A question as straightforward as “how many students attend non-public schools” (students attend non-public schools at public expense if the district/county cannot provide the special education services that the student requires) produces a count of 80 in one system and 141 in another.
This is all good stuff. If the district wants to improve, and commissioning the report indicates that it does, having a report that points out areas to improve is a helpful first step. The key is whether the district will take action on the recommendations.
The Latest Data
None of that is why I’m writing this post. The report also quoted some numbers and whenever I see numbers I get excited. My first instinct is to double-check them, if possible. FCMAT had data on the numbers of students in special education so I went to the same CDE source they used. FCMAT’s report only went up to 2023-24 but the CDE site had the newly published 2024-25 numbers and those were eye-popping. Here is how the proportion of TK-12 students in special education has evolved over the last few years1.
Although enrollment has fallen everywhere, the absolute number of students with disabilities has increased. San Francisco added 685 new students with disabilities last year alone. That’s a 9.2% increase, far more than the 3.4% average across the state. It comes on top of the 650 it had added over the two prior years2.
Types of Disability
There are 14 categories of disability, but four of them account for 90% of cases, both here and in the state as a whole. The distribution in San Francisco is shown below. The statewide distribution is broadly similar. San Francisco has significantly fewer Speech or Language Impairment (17% vs 21%) but it has more Autism (24% vs 22%) and Other Health Impairment (17% vs 15%) cases.
Different disabilities have different age progressions. Speech or Language Impairment cases peak in kindergarten or 1st grade, then decline rapidly and almost vanish by high school. Autism cases peak also peak early but decline little through later grades. Cases of Specific Learning Disability and Other Health Impairment grow steadily through elementary school, peak in middle school, and decline only a little thereafter3.
These categories are not all growing at the same rate. Autism cases are exploding, up 40% in San Francisco in just two years. Children in California are apparently five times as likely to be diagnosed with autism as children in Texas. Some of the rise in autism cases is attributed to more screening. The more you look for something, the more you find it. Cases of other disabilities are also growing faster here than elsewhere in California. Some of that may be due to earlier and more systematic screening for dyslexia.
If we look at the increase in cases by grade, preschool to 3rd grade cases have jumped by 36%. Most of these new cases are probably Autism and Speech or Language Impairment, which tend to get diagnosed early. San Francisco has seen 10%+ increases in all the other grades too. Although the rest of the state saw fairly small increases in cases in middle and high school, San Francisco has seen 10%+ increases in those grades too. Those are probably Specific Learning Disability and Other Health Impairment cases.
Disproportionality in Special Education Identification
Disability diagnoses also vary by race. In 2018-19, the last year for which we have this data, there were roughly the same number of Asian students with Autism and Specific Learning Disability but Latino and Black students were more than four times as likely to have Specific Learning Disability than Autism.
If we look at the propensity to be in special education by racial group, we see that Black, Latino, and even Asian students are significantly more likely to be in Special Education in San Francisco than elsewhere.
Black students are more likely to be in Special Education in San Francisco than in any other district with at least 200 Black students (San Francisco has 2900). Nearly one in every three Black students is in Special Education. Since boys are nearly twice as likely as girls to be placed in Special Education (in San Francisco and across the state), the share of Black boys in Special Education in the city might be over 40%.
Budgetary Concerns
Special education has to strike a balance. The students have to receive the particular help outlined in their Individual Education Plans (IEPs). At the same time, it is desirable to keep them integrated with the general population as much as possible in order to foster social interactions with peers and to receive regular classroom instruction which is more rigorous than they’d receive in special education. The term of art is that students should be in the Least Restrictive Environment (LRE) possible, given their disability. The CDE reports on how much time each Special Education students spends in a regular classroom setting. The reporting buckets are:
80% or more of the time in general education,
40-79% of the time in general education,
39% or less of the time in general education,
separate school or other setting.
Conveniently, what’s good for the student is also good for the budget. Student-adult4 ratios are lower in special education settings. The less time a student spends in general education the more expensive it is to provide that student’s education.
Although there is some dedicated special education funding from state and federal sources, the FCMAT report says 74% of total special education costs come from SFUSD’s unrestricted general fund. Money spent on special education is money that is no longer available for other students. This makes it vital to ensure that only those who truly have a disability receive special education. As the report says:
Identifying students for special education when they do not require it is not a best practice and increases special education costs. When they are not enrolled in special education, students experience less stigma, increased access to rigorous instruction given in general education, and more interactions with their typically developing peers.
The increase in autism cases is particularly bad news for the budget. They tend to require more intensive help. Fewer than 40% of autistic students spend at least 80% of their time in general education settings. For students with the other common disabilities (Specific Learning Disability, Speech or Language Impairment, Other Health Impairment), the range is between 70% to 80%.
Moreover, unlike students with Speech or Language Impairments, students with autism tend not to exit Special Education after a few years. They’re likely to be in it until they graduate.
Conclusions
In a time of dwindling resources and increasing numbers of diagnoses, the district has to do less of *something*. That means some combination of:
using staff more efficiently, by adjusting caseload guidelines or sharing staff between sites or adopting adult-student staffing guidelines instead of ignoring 1:1 paraeducators;
reducing the number of special education students by applying consistent diagnostic criteria across the district instead of allowing each site discretion to do its own diagnoses.
taking resources away from general education students in order to serve special education students better;
paying lip service to everyone while failing to meet the obligations to anyone.
The FCMAT report and presentation said that the proportion of students with disabilities was slightly lower in San Francisco than in California (13.6% to 13.7%). In all my analyses, I always exclude the two Five Keys schools when talking about San Francisco because, although SFUSD gave them their charter, their students are primarily adults and primarily in LA County jails. Including them in any San Francisco count feels wrong. In this case, excluding them raises San Francisco’s special education numbers because only 3.2% of the 3,447 students at Five Keys Independence High had a disability.
The drop measured in 2021-22 is probably due to the schools having been closed in 2020-21. When schools were closed fewer students could be diagnosed with disabilities. This wouldn’t show up in the enrollment numbers until the Census Day in October 2021. Logically, some of the increase in 2022-23 was probably because there was a backlog of undiagnosed cases from when the schools were closed.
CDE stopped publishing this type of data after 2018-19 but I presume the same is true today.
The adults may be teachers, speech therapists, paraeducators, or other professionals
Federal and state laws mandate special services for students with disabilities. The costs of these services have grown much faster than the funding the state and federal government provide.
The result: Local school districts are picking up the bill and paying more and more special education costs.
State-wide data shows the local school district share of these special education costs grew dramatically from 8 percent in 2004–05 to 61 percent in 2018–19.
Even with declining enrollment, more and more children are found eligible for special education, particularly children with more severe and costly disabilities.
In 2023–24 in San Francisco, the school district paid $196,336,787, or a whopping 74.45 percent, of total mandated special education costs.
Here is the real kicker: These increased special education costs result in less services for all the other students.
That’s right. The less money in the general fund, the less money left for services for all the other students. Not only does the increased cost of special education create a budget imbalance, it denies other children the education services they need.
FACT: The combined special education state and federal financial funding is insufficient to pay for even the most efficient special education programs. This is a statewide problem. All school districts struggle to pay for these higher special education costs. Blaming school districts for these rising costs is a total distraction.
Federal government shortchanges kids
When the federal government created special education laws, there was a pledge by the feds to pay 40 percent of the excess costs. It never came close.
According to the Congressional Research Service, current funding is less than 12 percent, and the shortfall in the 2024–25 school year nationwide was $38.66 billion.
San Francisco flubbed an opportunity to fix this when it adopted the student success fund. Read this to find out why: https://thevoicesf.org/san-francisco-can-do-a-better-job-of-helping-the-school-district/
Bellevue (WA) recently saw a bump in student achievement. However in the years prior, the district argued that the public school enrollment losses it was seeing (>10%) were the result of high local living costs. Maybe, but poor folk are exiting public ed, the students that remain are likely those of richer families. And if there's one thing that's well studied in education, it's that family income is one of the best predictors of education attainment. As the low income families are exiting the local public system, praise is given for the boost in achievement. That blows.
I wonder (and expect) that the difference in the number of SF SPED kids (vs the state) is also consequence of selection bias ... school choice. And in particular, the choice to choose private school, and the choice for private schools to reject, select out, and otherwise not cater to SPED kids. Why in SF? Many SF families can afford the private choice AND that's likely correlating with the level of education of their parents, another predictor of student success.