In what one might interpret as a sign, the October 29 council hearing on teacher retention started half an hour late and went past midnight. Dozens of public witnesses testified, while the three government witnesses were called up at 10 pm. They included no one from DC’s charter sector.

The only council member by that time (and really for most of the hearing) was the council chair, Phil Mendelson. Repeatedly with the government witnesses, Mendelson referenced the late hour and ensuing difficulty of discussing data. It may be that the council chair is suffering from a serious irony deficiency, given not only that he scheduled this hearing in this precise manner, but also that the council has held repeated hearings on this subject for years without DC’s high teacher turnover changing much.

Nonetheless, there were several points of that October hearing (see the video here) that call for closer attention:

1. At various moments, Mendelson referenced high principal turnover in DCPS (ie at Miner, at the 7 hour,10 minute mark) and the toll of stress on DCPS teachers (ie the 6 hour 29 minute mark). But at no point do I recall anyone saying much, if anything, about either in DC’s charter sector.

2. At the 6 hour 34 minute mark, Mendelson noted that getting retention data from the office of the state superintendent of education (OSSE) and DCPS is difficult. As far as I know, nothing was said about getting data from DC’s charter sector on this subject.

3. For the better part of an hour, Mendelson questioned the DCPS chancellor about working with kids with academic challenges. As far as I know, no one said much of anything about this topic with respect to the charter sector.

4. During the section of the hearing where a few charter schools were featured, Mendelson noted (at the 2 hour 24 minute mark) that one KIPP DC campus (Connect) has low attrition while WILL has high attrition—to which the KIPP DC representative said the schools “are not the same.” (Indeed!) A few minutes before, the DC Bilingual representative noted that charters need more resources. (No word on whether that would be revenue bonds, like those Sojourner Truth was just awarded for its expansion in the wake of its reported (and incredible) 35% teacher retention rate in SY23-24.) Perhaps unsurprisingly, the head of Lee brought angst and agita (starting at the 4 hour 34 minute mark) to the idea of public accountability in DC charter schools. (Lee had a 53% teacher retention rate in SY23-24.)

See a pattern here?

The charter sector section of the hearing was represented by leaders from Yu Ying; DC Bilingual; Monument Academy; and KIPP DC. The idea appeared to be that those charter schools were chosen for their excellent teacher retention rates. And that is (kind of) correct:

Yu Ying teacher attrition

SY23-24 9%
SY22-23 14%
SY21-22 12.7%
SY20-21 15.5%

DC Bilingual teacher attrition

SY23-24 17%
SY22-23 14%
SY21-22 15%
SY20-21 5.9%

Monument Academy teacher attrition

SY23-24 11%
SY22-23 5.6%
SY21-22 8.7%
SY20-21 0%

The exception to this is KIPP DC. And when I say “exception,” I mean that at some KIPP DC campuses teacher attrition is exceptionally bad, descending from a high of 68% (at WILL in SY23-24), with multiple campuses in multiple years reporting 50% or greater attrition rates. The data below for KIPP DC’s 20 campuses is going back in time (from left to right) for four school years, from SY23-24 to SY20-21:

AIM:                            42%; 31%; 41%; 11%
DC Arts & Tech:         26%; 36%; 21% 7%
College Prep:              31%; 22%; 19%; 15%
Connect:                     5%; 13%; 26%; 14%
Discover:                     11%; 17%; 27%; 3%
Grow:                          39%; 23%; 13% 11%
Heights:                      63%; 51%; 26%; 10%
Honor:                        44%; 39%; 23%; 0%
KEY:                            31%; 26%; 50%; 10%
Inspire:                       32%; 24%; 25%; no data
Pride:                         11%; 27%; 17%; no data
Lead:                          30%; 14%; 24%; 15%
Legacy:                        33%; 30%; 23%; 7%
LEAP:                          53%; 39%; 12%; 19%
Northeast:                   38%; 34%; 48%; 11%
Promise:                     16%; 18%; 14%; 7%
Quest:                        31%; 24%; 19%; 13%
Spring:                       28%; 28%; 26%; 9%
Valor:                         32%; 29%; 43%; 17%
WILL:                         68%; 56%; 55%; 24%

While something appeared to have happened between SY20-21 and SY21-22 that caused most KIPP DC campuses to have greater teacher attrition rates (pandemic? high principal turnover? data snafu? who knows?), overall at KIPP DC the teacher attrition rate was 34% in SY23-24 and 48% in SY22-23, per OSSE reporting.

By comparison, DCPS’s overall teacher attrition rate for SY23-24 was 21% per OSSE reporting. This is not to say that everything is rosy at DCPS. For instance, here are DCPS schools with the highest reported attrition rates for teachers that year:

Excel: 57%
Ketcham 48%
Wheatley: 44%
Kelly Miller: 43%
Kramer, Walker-Jones: 41%
MacFarland, Brookland: 39%
Savoy: 37%
Garfield: 36%
John Lewis: 35%
Hart: 34%
Plummer, Bruce-Monroe, Hendley: 33%
Anacostia, Hardy, Moten, Cleveland: 32%
Kimball, Johnson, Hearst, Simon: 31%

None of that is good, of course–and it’s not as if DCPS schools with lower teacher attrition rates are all far behind those above. But the greatest number of DC schools with the highest teacher attrition rates have been DC charters (something that the charter board write-up of the hearing (#7 on p. 4) conveniently doesn’t mention).

Indeed, averaged across both sectors, the teacher attrition rate for SY23-24 at all DC’s publicly funded schools was 26%, dwarfing national rates.

Now, if you think my data journey above is not easy to follow, you are right!

That is because nowhere in DC does this data exist in this form in a handy spreadsheet across multiple years. I got the numbers for those four charter schools in the hearing from two different sources: individual charter LEA annual reports and OSSE’s recording of teacher attrition for SY23-24. (See here for the OSSE data and here for posted charter school annual reports.)

DC’s charter board posts annual reports for all charter LEAs on its website (see link immediately above). Each charter is mandated to report teacher attrition, high/low teacher pay, and executive pay. Most do—but some simply don’t. (Looking at you for SY22-23 Appletree, Basis, Bethune, Capital Village, CC Prep, Digital Pioneers, Eagle, and Children’s Guild.)

Then, too, the last of these annual reports currently posted is for SY22-23—almost 2 YEARS ago. It’s not like charter schools don’t have this information—they have to report recent teacher retention annually to OSSE.

And that’s not even getting to what is relayed in those charter school annual reports. For instance, there were multiple charter LEAs I saw for SY22-23 that reported executive salaries that were inaccurate because what was being reported was only ONE person’s salary—and not always the highest paid person. (Looking at you Appletree, Capital City, DC Prep, Eagle, and Friendship.)

Then there’s interesting math—such as Creative Minds reporting a 0.31% teacher attrition rate in their SY22-23 annual report. In SY20-21, the school reported having 41 teachers, whereas the year before it reported having 48 teachers. The difference between 48 and 41 is 7, which happens to be 14% of 48. Since the school’s SY20-21 annual report noted a teacher attrition rate of 0.14%, it appears that both a decimal and a percentage were used for the school’s SY22-23 number for teacher attrition.

Yet DC charter school annual reports are only part of DC’s teacher attrition data problem.

OSSE’s datasets for teacher retention appear to be only for SY22-23 and SY23-24. And OSSE’s workforce reports do not seem to precede SY18-19. Even then, those workforce reports are issued only every other year (because yeah, teacher retention in DC hasn’t been a four-alarm fire for years).

As it is, starting on p. 32 of the most recent workforce report, we can infer not only that schools with wealthier student bodies appear to have the most stable employment, but also that charters overall have much less stability than DCPS schools. Indeed, that page alone makes it hard to understand why the council had no charter board personnel as a government witness on October 29.

Helpfully, EmpowerEd created a comprehensive spreadsheet this year on this subject. It doesn’t just list charter teacher retention for SY23-24 using data from OSSE—it also lists teacher pay scales (which are now mandated to be public in DC); expulsion and suspension rates; and executive compensation so you can see clear (and not always pretty) trends.

More specifically, this spreadsheet highlights what we in DC need and deserve more of on this subject:

Annual Robust Analysis Of Data

This could show, among other things, how student demographics are correlated with teacher attrition and how principal retention is correlated with both. We could also see if there is any correlation between higher suspension rates and teacher attrition. (That may better illuminate why, for instance, 25% of KIPP DC’s campuses in SY22-23 had >20% suspension rates.)

Such analyses could also highlight which campuses and LEAs are struggling both to retain teachers and keep students in school—both of which have a profound effect on student achievement and well-being. This gets to the heart of what many teachers testified about on October 29: supports for students and teachers IN our schools, not just allowing our schools to become a pass-through for both.

(And as long as we’re doing data analysis, how about calculating a return on investment for all those wildly paid executives at charter schools with not wildly paid teaching staff and high attrition and/or suspension and expulsion rates?)

OSSE is the natural partner to do such analysis on the regular. But OSSE couldn’t even be bothered with calculating average teacher attrition across DC’s charter sector for SY23-24 on its data chart. (I got 32% FWIW.)

Ensuring Reliable Data

There appears to be actually THREE datasets for teacher retention for two separate years. That is because the OSSE dataset for SY22-23 doesn’t seem to match the data from the posted SY22-23 charter board annual reports.

Consider that in its SY22-23 annual report, Achievement Prep reported that its attrition rate for teachers was 71%–or what would be a retention rate of 29%.

But in its SY22-23 data set, OSSE reported that Achievement Prep’s retention rate for teachers was 32%–or what would be an attrition rate of 68%.

Now, if Achievement Prep’s annual report that year was incorrectly confusing attrition with retention, it *could be* that its attrition rate for that year was only 29%, not (an eye-popping) 71%.

So here are the data possibilities for SY22-23 for just this one LEA, Achievement Prep:

71% attrition/29% retention
68% attrition/32% retention
29% attrition/71% retention
None of the above

There may be a little difference between OSSE reporting and charter school annual reports IFF there is a difference in accounting for the (presumably very small) numbers of teachers who leave teaching at the school but nonetheless remain at the same school the next year in another capacity. But that isn’t delineated anywhere I could see. Moreover, we do not (yet) have posted SY23-24 charter school annual reports, so there is no way to have another data point for clarity–not to mention that there is also no way to know why there are different values between data for SY22-23 reported in those annual reports and data reported for that same school year by OSSE.

Then there are OSSE’s data inconsistencies. For instance, OSSE reported in column G on the overall LEA sheet for SY23-24 that Appletree had a 68% teacher retention rate—or what one could construe as a 32% teacher attrition rate. But in the campus-specific data from OSSE for Appletree for that year (which is the sheet marked retention overall school), if you average the retention rates for individual Appletree campuses in column J, you get 62% retention, or what one could construe as a 38% attrition rate.

So: Does Appletree have a teacher attrition rate of 32% or an attrition rate of 38%?

There may be ways to account for this difference, such that one shouldn’t simply average individual campus retention or attrition rates to come up with a number for an entire LEA. But the data I am referencing comes from two different sheets of the same spreadsheet in a column denoted on both sheets the same way: “Percent of teachers who remained employed as teachers at the same school.” And since there is no posted annual report for Appletree for SY23-24, we do not have a cross-check (not that this would necessarily provide clarity—see above).

This data stuff is very low-hanging fruit that the council could actually do something about—like having a comprehensive database for all our publicly funded schools. (You know, like the one OSSE was supposed to have at some point called SLDS, for statewide longitudinal data system–except, well, who knows where that is?) Short of the council chair being testy about parsing data at a late hour, it is unclear what will be done by that body toward pushing OSSE and LEAs toward more robust data aside from scheduling another hearing in 2025, which after all these years seems performative at best.

In the end, DC doesn’t lack for people to speak about our poor teacher retention and what needs to happen. (If you have only a few minutes, for instance, listen to the October 29 testimony of teachers Laura Fuchs starting at the 4:36:20 mark and K.C. Boyd starting at the 5:13:00 mark and the testimony of Ward 5 state board of education rep. Robert Henderson starting at 2:43:25.)

And we certainly don’t lack for evidence of problems around DC teacher attrition (i.e. here’s stuff from 2016 and 2017 and 2018 and 2019). In addition to dwarfing national averages, DC also has every year many schools with teacher attrition rates (much) greater than 50%, which is an incredible burden on everyone both inside and out of those schools.

To illustrate first-hand how confusing the accounting gets, below are recent teacher attrition percentages in DC charter schools:

–The first number is for SY23-24, using OSSE reporting of retention converted to attrition.
–The second number is for SY22-23, using charter school annual reports. For LEAs with multiple campuses (and thus multiple attrition campus attrition numbers), I averaged the reported numbers for SY22-23.
–The third number is also for SY22-23, but using OSSE reporting of retention converted to attrition.

DC students and teachers deserve much better.

Teacher Attrition Percentages:

Academy of Hope:      12%; 4.76%; 22%
Achievement Prep:     47%; 71%; 58%
Appletree:                   32%; 21%; 49%
Basis:                          27%; 38%; 46%
Breakthrough:             19%; 29%; 40%
Bridges:                       51%; 56.8%; 44%
Briya:                           17%; 7%; 14%
Capital City:                34%; 21%; 25%
Capital Village:           50%; no data; 57%
Carlos Rosario:           19%; 15%; 17%
Cedar Tree:                16%; 10%; 25%
Center City:                22%; 19.9%; 37%
Chavez:                       21%; 18.4%; 21%
CC Prep:                     47%; no data; 67%
Creative Minds:         33%; 0.31%; 35%
DC Bilingual:             17%; 14%; 16%
Digital Pioneers:       46%; 42%; 70%
DCI:                             21%; 21%; 33%
DC Prep:                     32%; 26.7%; 40%
DC Scholars:              65%; 39%; 49%
DC Wildflower:          0%; 0%; no data
EL Haynes:                 25%; 19.7%; 31%
Eagle:                          32%; 18.5%; 47%
Early Childhood:       41%; 35%; 35%
Elsie Whitlow Stokes: 31%; 31%; 33%
Friendship:                29%; 18.5%; 39%
Girls Global:              71%; 58%; 44%
Global Citizens:        37%; 23.5%; 57%
Goodwill:                   39%; 33%; 29%
Harmony:                   22%; 87.5%;47%
Hope:                          64%; 39%; 52%
Howard:                     28%; 2%; 52%
IDEA:                          45%; 42%; 27%
I Dream:                     73%; 63.6%; 45%
Ingenuity Prep:         34%; 29%; 30%
Inspired:                     27%; 24%; 38%
Kingsman:                  40%; 40%; 71%
KIPP:                           34%; 29%; 48%
LAMB:                         18%; 13%; 17%
LAYC:                         0%; 0%; 33%
LEARN:                       54%; 54%; 53%
Lee:                             47%; 27%; 32%
MM Bethune:            25%; no data; 44%
Maya Angelou:          24%; 11%; 41%
Meridian:                   31%; 18%; 34%
Monument:                11%; 5.6%; 15%
Mundo Verde:            29%; 23%; 45%
Paul:                            36%; 29%; 35%
Perry Street:              38%; 6%; 47%
Richard Wright:        43%; 9%; 43%
Rocketship:               51%; 60%; 52%
Roots:                         12%; 12.5%; 33%
SEED:                         34%; 31.5%; 44%
SELA:                         24%; 6.2%; 10%
Shining Stars:          50%; 50%; 53%
Social Justice:          42%; 77%; 23%
St. Coletta:               43%; 42%; 62%
Statesmen:               21%; 13.6%; 56%
Family Place:            0%; 18%; 0%
Children’s Guild:      19%; 14%; 43%
Next Step:                  23%; 23%; 38%
Sojourner Truth:       65%; 16%; 24%
Thurgood Marshall: 30%; 29.7%; 34%
Two Rivers:                32%; 34.6%; 24%
Washington Global:  41%; 23%; 41%
Latin:                           21%; 12%; 28%
Wash. Leadership:     30%; 22%; 24%
Yu Ying:                      9%; 14%; 9%
Youthbuild:                25%; 12.5%; 25%