If you don’t care about my editorialization, and just care about the charts and map alone, you’re wrong (*unless you’ve already read this)! I can’t stop you from skipping context and interpretation, but I think it’s important.
In early September 2026, the DC Department of Public Works (DPW) launched a Status Collection Map, so that residents serviced by DPW could see when they could expect their trash or recycling to be picked up. Like the actual solid waste service, as of the end of September, that website was still not working as intended.
I happen to moonlight as an Advisory Neighborhood Commissioner (ANC), and as part of that, I get some visibility into what happens when resident interactions with city services go sideways. I've seen numerous neighbors' complaints to DPW in that capacity, when their solid waste pickups get missed, and their subsequent begs for help (which I often join in to prod for updates and visibility) are either slowly or insufficiently addressed. Inspired by some of fellow ANC Josh Jacobson’s work creating dashboards to supplant inadequate trackers for kiddie pools and rat treatment, I was motivated to make something better myself. I have lately been a little GIS-curious after being out of that game for about a decade now, and I recently got access to the paid version of Google Gemini, courtesy of a Google Drive family plan upgrade.
In short, people deserve a functioning local government with services they’re paying for through tax revenue. If they are not receiving those services, they deserve to know why, with a plan to rectify the problem.
I built, with substantial coding assistance from my robot, my own trash and recycling pickup dashboard. Unlike DPW’s status dashboard, the goal is not to show a live map of where the waste has been picked up within the day (I hypothesize few care about the exact moment of pickup). My goal is to identify where there are misses, especially repeat misses, and map the actual geometries (trash and recycling routes and city divisions like Ward or ANC). I have layered into those geometries:
I like putting the limitations before the interpretations so that people can keep them in mind while reading the analytic conclusions. Not to get too into theoryposting here, but it’s incumbent upon me to present you with enough evidence that things are going wrong in the city that you are sufficiently convinced to doubt things are going right.
The outcome we are trying to identify is missed collections, but our key measurement of that outcome is 311 requests requesting service for a missed collection. Assuming minimal to no false or erroneous requests, we are getting an underestimate of the problem. That assumption seems reasonable because it’s a pain to report on 311, which has had a smattering of issues this year and many have started using Snap311 instead of the official app. It’s also not hard to assume that the underreporting is data missing not at random. Further analysis attempting to predict pickups using sociodemographic factors would be meaningful for assuring an equitable policy intervention.
The outcome mapped is also a count of 311 requests rather than a rate. Showing a percent of serviced households (or individuals) within the geometry reporting missed service over the time period would adjust for the concentration of serviceable addresses. Including both would be the most informative, but calculating an accurate intersection of routes within political representation boundaries (+ vice versa) and the tickets that fall within those intersections, requires very careful analytic oversight (and the time to provide it). Mark that down for a future feature.
DPW services >100k households (~105k at last count). Of them, >6k missed trash or recycling pickup service in the last 6 months, a total of >17k times (>12k trash and >5k recycling). Nearly 1 in 3 households who reported a missed collection in the last 6 months reported another (and if you look at an earlier cut of this data, that's increasing!, up from below 1 in 4). Many reported numerous times. Wards 7, 5, and especially 4, had the most tickets filed. All Wards, by law, have about the same number of people, though they don’t have the same number of DPW-served addresses. Ward 1 has the highest rate of misses per service address count over the time period. Ward 3 the lowest (~1/5th the rate of Ward 1). Ward 4 has the most service addresses and Ward 2 has the fewest.
Even with Ward 2’s vanishingly few service addresses, ANC 2B appears to have the highest repeat request rate of any ANC for DPW missed collection tickets. SMD 2B08 has the highest repeat request rate within that ANC. You can see how this landed on my radar, and it’s both reassuring and frustrating to see a problem validated by the data.
This will sound like a digression, but I think some of the process difficulties can and should inform the conclusions (see: POSIWID, form follows function). When I initially wrote this up, among other concerns, I was trying to get confirmation of some unusual routes’ existence. If you look at some of the smaller alleys around the city, you’ll find some funny cutouts like recycling routes RSP20_2 or RSP50_5 in ANC SMD 2B08, which contain a high concentration of 311 missed recycling pickup complaints. But searching for information on DPW service routes is remarkably difficult: I couldn’t find these on the Open Data DC site, nor on basic Google searches. As it turns out, Gemini/Antigravity pulled data from sites I cannot directly access on a web browser using DPW’s Routes Assignment Experience Builder and the DC Office of the Chief Technology Officer’s (OCTO’s) REST Services Directory. I was able to eventually confirm these routes exist, but it was not easy.
For these little routes, I couldn’t figure out the exact expansion of the initialism RSP, but it seemingly stands for Recycling Special/Service/Single-stream Program/Packer. For RSP50_5, the 50 is the route group identifier, and the _5 redundantly identifies the day of the week (Friday). For recycling, 6 of these are identified: RSP10_1, RSP20_2, RSP21_2, RSP30_3, RSP40_4, RSP50_5. For trash, 7 of these are identified: IC10, IC11, IC20, IC21, OR11, OR21, and OR30. They don’t denote the special case in the initialisms, which seemingly stands for Inner City/Circle and Outer Ring, but use a similar two digit route group identifier which implies the day of the week.
The special routes don’t cover a lot of service addresses, but they do seem to have a lot of missed collections. Consider OR30, whose fractional area combined with ~50 requests saw a density of ~1k/sqmi (OR21 saw ~2k/sqmi on just 2 requests, but the request count and route are too small to see as anything but noise).
A look at the charts will show you that Monday routes were the least reliable and Thursday routes were the most reliable. I don’t have much to say about this other than it looks like some of you have lucky pickup days assigned and that perhaps sanitation workers take inspiration from Garfield.
The clearest overall theme is that DPW is ignoring their massive service problem. Service across all 8 Wards, all 46 ANCs, all 345 SMDs, and all 223 (103 trash and 120 recycling) routes deserve scrutiny. Ask DPW and they might tell you that there are frequent call-outs, which were more intense in the hottest summer on record.
Looking at some of these smaller routes, it’s hard not to think that they’re less resilient to call-outs. If a single person can’t make their shift, does it mean that the route performance is degraded? Some sites skipped? The route skipped entirely that day? To investigate this hypothesis, I FOIA’d (R014873-093026, September 30th, 2026) employee counts and leave per route per day since the beginning of the year. I’ll add any updates here.
I’ve at least seen incidental sick reporting and route delays, so I know DPW has the historical data to predict this, and regardless of the analysis. Even on its face, they could, should, must hire more people. Less staff leads to more overtime, which leads to more expenditure, which leads to less staff. It’s a vicious cycle.
It’s also possible that DPW’s community relations team may have inaccurate information on routes. If they don’t share clear and accurate data with the public, it’s certainly possible they don’t share it internally either. It would not surprise me if the reason for repeat misses on some of these routes is that a follow-up collection from a 311 report is routed to the wrong place or the wrong team. They wouldn’t be the only DC agency with that problem.
DPW could investigate their data and make these changes. DPW could provide updated information to all residents without a FOIA. DPW could improve their infrastructure, physical and digital, instead of allowing it to break and stay broken.
DPW provided responsive records to my FOIA about a month later, on November 2nd, 2026. I first converted these data from PDF (yes really lol) to CSV and then imported and tested the hypotheses against them.
The biggest problem with Monday is that it's a backup from the weekend - there's both more solid waste because people do more things over the weekends and they constitute a multi-day period where the solid waste is not being serviced. It makes sense that more routes run Mondays than Thursdays.
But my hypothesis about small routes was absolutely validated. If your route is staffed normally by 1-2 people instead of 5+, it makes sense that one person calling out leads to total degredation in service for the route, and that hits the alleyways (like my nearby neighbors) with the little trucks and little crews extra hard!
Furthermore, and stop me if this is a complete shocker to you, it's exacerbated by bad weather. There are special rules about when it's >95F out for shorter hours and extra breaks, that really mitigate the crappy working conditions - but when it's 85-95, aka still hot and awful, people call out. When there's heavy rain, people call out. These things are obvious, they're now validated by the data, and they're solveable! How? Improved equipment (fans, cooler clothes, better rain jackets and boots) and, most importantly, more staff. Not expecting callouts is willfully ignoring human behavior, it's bad planning. You can't reliably run even the most staffed route without backup staff. DPW clearly needs more field staff, and yet, right now, it doesn't look like they're hiring any.
DPW has not had a good year. It's been unlucky, but it’s hard to look at the FY2026 Performance Plan and think key performance indicators have been met. Sure, residential trash and recycling pickup under the Solid Waste Management Administration isn’t all of what DPW does, but it certainly is one of the most salient touchpoints to the District government. If people don’t get routine service—and worse, if people can no longer depend on reporting missed service—faith in basic functions of local government start to erode pretty quickly. My neighbors have had repeat misses and have been frankly gaslit by absurd responses from DPW and their community relations team, and something’s got to give. We need a real plan, and solid, material change, sooner rather than later.
If you’re reading this and you’re a resident who’s missed service: keep entering 311 requests, you’re not alone. Write to your ANC, Councilmember, and DPW.
If you’re reading this and you’re in Council: demand answers from DPW and budget them accordingly for fixes. Make sure leadership is held accountable.
If you’re reading this and you’re in DPW leadership: figure out route issues (you can use my tool to find the problems!) and adjust accordingly. Hire more people. Get it right, the city depends on you.