What schools gain and lose people mid-year?

And are charters pushing students onto neighborhood schools?

Families move in or out of DC. They get in off a wait list into a school they wanted. They decide a school isn’t good for their child and switch. Or maybe some schools push (“counsel”) out problem students. What is happening inside our school system?

One narrative that comes up often in DC education debates goes like this: some charter schools enroll a lot of kids at the start of the year and get funding based on that, but then many of the kids leave and go back to their neighborhood school. The neighborhood school does the work of educating the student, but either doesn’t get a bigger budget or, if they do, it’s hard to suddenly hire more teachers, special education teachers, etc. once the year has started. Is that right?

DC does budget an “enrollment reserve” for DCPS schools that exceed their budgeted enrollment, but last year it was $1 million (enough for about 8.7 teachers across the whole system). For the coming year, the DC Council temporarily doubled this reserve and also created a new, $2 million “Mid-Year Mobility Fund” to “support schools with significant enrollment increases or significant turnover in enrolled students after the SY23-24 count day.”

Ward 3 Councilmember Matt Frumin has been talking about the mid-year mobility issue since long before he was elected to the council. Meanwhile, Abe Clayman, the Director of Education Initiatives for the Federal City Council, crunched some mobility numbers last year with 2020 data, and shared his thoughts on this question with me. I decided to pull up the data from 2021-22, which is available from the OSSE School Report Card’s data page. You can access my calculations here.

The bottom line

For the 2021-22 school year, DCPS schools (excluding alternative schools¹) had about 3,600 entries (people enrolling at a school) and 2,600 exits (people disenrolling), for a net gain of 1,000 students between October (when the official enrollment count happens) and the following May.

Public charter schools (PCS), meanwhile, had about 400 entries and 1,900 exits, for a net loss of about 1,500 students.

In total, the public school system lost about 500 students after October, who didn’t just switch to another DCPS or charter school but either left DC, went to a private school, or just stopped going to school.

Charters have slightly fewer exits, many fewer entries

The charters saw about 4.9% of students exit a school mid-year versus 5.4% for DCPS schools. That’s not much of a difference. The big difference is: few charter schools accept new students mid-year — only about 10% of entries were into charters.

38 charter schools enrolled zero new students mid-year, while only one DCPS school did that. 94 charters (78% of them) had five or fewer new mid-year students; only seven DCPS schools did the same, and they are all application-only or lottery-only.² Basically, schools that aren’t neighborhood schools (and thus not required to take new students within a boundary) and are fully enrolled, didn’t take anyone new.

We can differentiate DCPS schools into the neighborhood schools, citywide lottery-based schools, and selective (application-only) schools (also alternative schools, which I’m not counting.)

As you can see, lottery and selective groups look a lot more like charters. Lottery schools take more kids, but much less than neighborhood schools do; lottery and selective schools shrink a bit, on average, over the year, but less than the average charter does.

When do people leave/arrive during the year?

We can plot the numbers of students gained or lost for each month of the year, for every school:

The lines at the fringe are (some of) the alternative schools. Let’s exclude those and look at just the neighborhood schools:

Most neighborhood schools gained students over the year, some considerably.³ A smattering lost students (11 schools lost five or more on net). Also, something was going on at Stoddert Elementary School in the fall of 2021 and especially between December 2021 and January 2022, when 24 students left.

Here are the charter schools:

Only three schools gained students over the year.⁴ A handful had large attrition, while the main mass of schools all had a steady attrition ending them anywhere from zero to 20 students smaller by the end of the year.

The same goes for the DCPS lottery and selective schools. Except one,⁵ they all stayed flat or shrank up to 17 students, comparable to the typical charters.

Are charters pushing people onto neighborhood schools?

I don’t see evidence here to suggest some kind of nefarious scheme to push kids from charter schools. They do leave charters to the tune of about 5% of the student body over the year, but lottery DCPS schools are about the same.

It would be smart for both the charter board and DCPS to look at the outlier schools that have much larger attrition than is typical, especially if it persists year to year.

One caveat is that this data doesn’t tell us whether the students who move mid-year are more (or less) likely to be at-risk, English language learners, or receive special education services. If schools are seeing greater attrition among those groups than in general, that would also be cause for concern, and regulators should watch that. (OSSE’s data set does break out these subgroups for year-to-year re-enrollment per school; maybe I can analyze that.)

Basically, though, my analysis reinforces that neighborhood schools are simply different from other schools. Neighborhood schools can gain students any time someone moves into the neighborhood (or is settled into the neighborhood by social services or as a refugee or such), or chooses to leave another school. That school has to make room immediately, and meet that child’s specific needs if they have disabilities, or don’t yet speak English, or something else.

Neighborhood schools need to be funded to account for this, and can’t simply add staff at the drop of a hat. That means they may end up with a higher per-pupil budget than other schools, or more empty capacity in the building, because slack just needs to be part of the equation. Principals don’t like that this happens, but they know it’s part of the job, and so should we.

What follow-up questions do you have about this? Does anything seem wrong? If you play with the data, do you notice anything else? I’d love to hear what you think — drop me a line or post a comment here!

1. I’m excluding a few types of DCPS and charter schools: alternative schools (schools that primarily serve students older than high school age, or who have been suspended or expelled from other schools, are under court supervision, or various other criteria, often helping them to earn a GED degree); schools at the DC Jail; or programs that help adults get their high school degrees. (I also exclude Briya PCS, which has both adult education and pre-K education for the adults’ children.)

2. The DCPS application-only schools that added fewer than five new students were Benjamin Banneker Academic High School (the one that added zero); Duke Ellington School of the Arts; School Without Walls High School; McKinley Technology High School; and Bard High School Early College DC. The lottery-only schools were School-Within-School @ Goding; Capitol Hill Montessori @ Logan; and Dorothy Height Elementary School.

3. The biggest gainers among DCPS neighborhood schools were Hendley Elementary School, Cardozo Education Campus, MacFarland Middle School, Roosevelt High School, Browne Education Campus, and Brookland Middle School.

4. Social Justice PCS and Two Rivers PCS - Young Elementary School gained a net of two students, and The Children’s Guild DC PCS gained three.

5. Thaddeus Stevens Early Learning Center grew by 12 over the year; it was in its second year of operation and its first full year open in person.