Most neighborhood schools grow 10-20% after modernization

When DCPS neighborhood schools start being modernized, families have to travel farther—often much farther—to the school’s swing space. That often precipitates a decision to switch schools. But then, when the school comes back to its regular location and a beautiful new building, it’s more attractive than ever and many families want to enroll.

That’s been my anecdotal observation, and I wanted to better understand it with data. The DCPS enrollment team kindly gave me a list of which schools were in swing spaces over the last ten years, when, and where. I correlated it with publicly available enrollment data to see how schools’ enrollment changes when they go into swing space and when they return to a modernized building.

The bottom line: Schools do see a jump in enrollment when they get modernized, though not every school has. Some schools lose enrollment when they go to swing space, but not as many as I’d thought.

Let’s look at the data!

Here is a graph of schools’ enrollment shifts:

This graph takes the enrollment in the year before going to swing space and scales that to the arbitrary value of 100. Then, the next data point—“Swing First”— is the relative enrollment; basically it’s a percentage of the enrollment from the Pre-Swing year. Swing Last is the last year of being in swing space, which is usually year 2 but occasionally year 3. Post-swing 1 and 2 are then the two years following, in the new modernized building.

As you can see, many schools see a drop from Pre-Swing to Swing First but not all. Then almost all schools see a big jump in Post-Swing 1 and 2. (Among other reasons, people deciding on schools for year 2 get to actually see the modernized building when considering the school, versus anyone who goes in year 1.)

The dashed lines, for Jefferson, WatkinsCW Harris, and Houston, reflect that those schools stayed on-site during their modernization.

Here’s a table of schools with their relative change going into and out of swing (sorted by the change when they come out):

(The numbers used for swing and post average the years involved, such as averaging the 2 years post-swing. The four at the bottom are still in their swing spaces now.)

As you can see, all but a few schools grow after modernization and most grow to 110-120% of the average while in swing space.

A few of these are outliers for some reason: Roosevelt and Cardozo high schools and Eliot-Hine saw really big jumps at the start and/or end. I think there were other factors going on for those (for instance, the Shaw middle school grades moving into Cardozo increased its enrollment), so I omitted them from the graph above.

Things I’d still like to analyze

I’m not really sure why Watkins, and also Eaton, didn’t grow while others did. I could look at lottery data, because for schools like those, whether or how many they take from the lottery also has an effect on enrollment.

I’d like to look at how far the swing spaces are from the school. Some are very close (Roosevelt’s swing space was “offsite” but it was at MacFarland, next door, for instance) while others are quite far.

Garfield Elementary, in Ward 8, is in a swing space 3 miles and on the other side of a huge park from the school. Parents at Amidon-Bowen Elementary in southwest DC, and Burroughs Elementary in Brookland, are unhappy they will have to each travel 3.5 miles to Columbia Heights for their swing space. Do (or will) these long distances harm a school’s enrollment more? Can we expect it to rebound afterward?

What else do you notice in the data? What factors do you think affect a school’s swing space enrollment changes that we might be able to quantify? You can look at the full data spreadsheet here.