Part 3 of 6 in The Good Side of 311

The Community Life Improvement Program closes 91.3% of its tickets within the city's own published SLA — on 40,043 closed tickets filed since 2024. The next agency on the list is Parks and Recreation at 74.6% on 13,065 tickets. The gap between those two is seventeen percentage points.

Where the 91.3% sits among the others

The Office of Homeless Services posts a higher met-rate than CLIP, 93.3%, on 3,973 closed tickets. That is the highest rate on the chart above. Its work is narrower than CLIP's — heavily weighted toward Opioid Response Unit calls, where the city's published SLA is short and the response is fast. Parks and Recreation is the next agency after that on the list at 74.6% on 13,065 tickets. CLIP, at 91.3% on 40,043, sits between them on rate and well above both on workload.

What CLIP actually runs

Two high-volume work types dominate the CLIP closed-with-SLA universe. The first is the named code SR-CL01 — Graffiti Removal — at 21,533 tickets with a 90.0% met-rate and a median close of about 4.9 calendar days on a 7-business-day SLA. The second is a catch-all that shows up in the dataset under a nullservice_code field but with a recognizable service name — "Maintenance Complaint" — at 17,012 tickets with a 93.1% met-rate.

The Graffiti Removal number is the easier one to make sense of. CLIP runs a dedicated graffiti-removal program with its own crews, on a published 7-business-day window, and the median ticket closes inside a calendar week. That is exactly the shape of a service the city has chosen to staff. It works.

The "Maintenance Complaint" rows are messier. They have no service code; they enter CLIP's queue under a service-name string only, and the published SLAs on those rows are a mix of 20, 70, 90, and 120 days. I am reporting the aggregate 93.1% met-rate but I am not collapsing those heterogeneous SLAs into a single commitment number. The high met-rate is real in aggregate; the per-promise story is heterogeneous, and the city is the only one that knows the per-promise detail.

When I asked CLIP, they know how good they are — and right to be proud

When I shared the finding with the department, Tom Conway, the Deputy Managing Director and Executive Director of CLIP, pointed to the approach he says produces it.

"CLIP has always taken a proactive approach to tackling quality-of-life issues by inspecting entire blocks rather than simply reacting to isolated complaints." — Tom Conway

On the 91.3% itself, Conway's response was direct.

"…the cultural connection to communities throughout Philadelphia makes CLIP employees strive to be the most efficient and effective as it is their neighborhood they are cleaning. This dedication is clear and shines through in the data highlighted in the article written." — Tom Conway

I also wanted to give a personal shout out to Tom. He's been with CLIP and its precedent agencies for nearly 30 years. I don't know what the title progression is like, but that's an impressive history of public service and helping us keep Philly as one of the lowest graffiti cities I've experienced. In addition, on a personal note, I could hear the pride in his voice about the quality of work his agency is able to deliver. It's good to see pride in one's work!

Why this counts as the headline

Of the seven named 311 agencies, only one — the Office of Homeless Services — closes a higher share of its tickets on time. Its operation is narrow and small, weighted toward high-priority response codes that are not the kind of ticket most residents will ever file. CLIP is the highest-met-rate big operation in the city's 311 system: the one a Philadelphia resident is realistically going to interact with, and the one most likely to close their ticket on time. That is the headline that holds.

A quick aside on the other departments to contextualize their jobs

The other agencies' lower met-rates on the chart should not be read as a uniform verdict on those departments. Some of the published SLAs in the city's 311 dataset are unrealistic for the work they govern. A 40% met-rate on an unrealistic SLA is not the same failure as a 40% met-rate on a realistic one — it means the published window was set to a fiction and the agency never caught up to it.

The clearest case is snow shoveling. The published SLA on city-wide sidewalk clearing after a storm is 1 business day. The median close time on that code runs about six times longer than the window. The city set a standard without a plausible way to meet it. CLIP's 91.3% is the headline because CLIP's published SLAs are the kind of windows a real operation can actually beat; that is part of why their rank holds. The other agencies' lower rates reflect a mix of workload, SLA design, and execution — not a uniform failure mode.

What 91.3% actually delivers

A wall tagged on Tuesday in Philadelphia will, on the median, be painted over by Friday. That is the 4.9-day median close on the SR-CL01 Graffiti Removal code, on a 7-business-day SLA, on 21,533 closed tickets in the analysis window. The 91.3% rate is the rate of that operation at scale — 40,043 closed tickets, those two codes plus CLIP's smaller work, on a workload big enough that the rank is anchored to a real agency rather than a small-sample artifact.

Tom told me the precursor to CLIP was a graffiti-cleanup initiative — a Mayor Rendell pet project from the late 1990s. He was a little rueful that I had not heard of CLIP before this research; when I mentioned it, he chuckled. After three decades of being in business, CLIP achieves a 91.3% performance score on about 20000 311 complaints per year... that's a lot of grafitti to clean up (amongst other things). And yet they do. That's a scale example of government agency success that deserves to be noticed, understood and replicated.

Notes, Sources, and Methodology

All data is the OpenDataPhilly 311 Service Requests table public_cases_fc, closed-with-SLA cohort of 481,010 tickets across the seven named 311 agencies filed 2024-01-01 through mid-2026. Information Request rows are excluded. service_code may be null on CLIP rows; null-code rows are included in the per-agency met-rate but never collapsed into a fabricated code.

Methodology. The met-rate is the share of an agency's closed-with-SLA tickets whose closed timestamp falls within the published SLA window under the city's own unit arithmetic (business-day math for Business Days SLAs, calendar-day math otherwise). Ranking is by met-rate descending. The next-best comparator for CLIP is Parks and Recreation; the two-proportion z-test uses a pooled-p denominator and a two-sided alternative.

Limitations. The Office of Homeless Services met-rate (93.3%) is higher than CLIP's (91.3%) but on 3,973 tickets, not 40,043; the "best big agency" framing is more useful for residents, but a different volume cutoff would change which agency holds rank 1. CLIP's "Maintenance Complaint" rows carry heterogeneous published SLAs (20/70/90/120 days); I report the aggregate met-rate but do not collapse those SLAs into a single commitment number.