How to report a broken street light in Kansas City

The street light on your block has been out for a week. Kansas City has a number for that, and the number works.

Dial 311, or (816) 513-1313, or open the myKCMO app. The request goes to Public Works, to a crew whose work group is literally called PW Streetlights, and if the city finishes it, it finishes it fast: an average of 2.8 days across the 18,831 streetlight requests it has resolved since February 2021.

The part that does not work is anything you might want to check for yourself afterwards. The city publishes a file of every street light in Kansas City. It has 92,536 of them, it has been downloaded more than fifteen hundred times, and it has not been updated since August 2021.

The thirty-second version

• To report it: dial 311 inside Kansas City, Missouri, or (816) 513-1313. The city’s 311 page gives the call center’s hours as Monday to Friday, 8 a.m. to 7 p.m. The myKCMO app on iOS and Android and the city’s web intake file the same request at any hour.
• Where it goes: Public Works, work group PW Streetlights — 20,315 of the 20,332 streetlight requests in the city’s file.
• How long it takes: the 18,831 that resolved averaged 2.8 days. Forty-one are still open.
• To check on it: data.kcmo.org, dataset d4px-6rwg, filter issue_type to Streetlights. Free, no account.
• Do not trust days_to_close on its own. It is filled in on requests that never closed, including all 41 that are open right now.
• Do not trust the street light map. Dataset n2wy-vqag, five years stale, no addresses, and 3,938 of its lights carry a retirement date.

Route one: filing it

All four routes land in the same system and produce the same kind of record, so pick whichever is in front of you. What the data does show is which one people actually reach for. Of the 20,332 streetlight requests filed between February 27, 2021 and September 14, 2026:

Phone is still the single biggest channel at 43.5 percent, twenty-five years into the internet. The three app and web routes together come to 10,999, or 54.1 percent.

Wherever it comes in from, it goes to the same place. 20,315 of the 20,332 requests are routed to the work group PW Streetlights; 16 went to PW Signs and one to Parks and Recreation Maintenance. They arrive from across the city fairly evenly — 3,766 from the 4th council district, 3,499 from the 2nd, 3,419 from the 3rd, 3,328 from the 6th, 3,310 from the 1st and 2,904 from the 5th, with 106 carrying no district at all.

Tell them what is wrong as precisely as you can, because the system files by sub-type and the sub-type decides who goes out. The categories it uses, and how often each has been used:

The catch, part one: the portal cannot rank its own categories

That third row is not how the city’s portal shows it, and the reason is a single character.

The label for a damaged pole is stored two different ways. On 2,551 requests it reads:

Pole Damaged| Down or Leaning

— with a pipe character where the comma belongs. On another 470 it reads the ordinary way, with a comma. Group the file by sub-type, as the portal does, and the two halves are counted as two unrelated categories: the pipe version lands in third place with 2,551, below “2 to 4 lights out” at 2,689, and the comma version lands in ninth with 470.

Add them and downed poles are 3,021 — the second most common streetlight complaint in Kansas City, ahead of multiple lights out. A resident, a council office or a reporter reading the ranking as the portal presents it gets the order of the city’s own streetlight problems wrong.

This paper has seen this exact defect in this exact file before. When The Stare pulled the whole 311 dataset on September 10, 2026, the largest issue category overall turned out to be written two ways and to outrank the label the portal ranks first once summed. What is new here is that the split falls on the one streetlight category that involves a pole in the road rather than a bulb.

The catch, part two: the clock runs on requests that never close

days_to_close is the column anyone reading this data reaches for, and it is not what it appears to be.

All 41 streetlight requests that are still open carry a value in it. Not one of them has a resolved date. Their days_to_close figures average 8.1 days and run up to 59.

The oldest of the 41 makes the point on its own. Work order 273915 was opened on June 29, 2024, is still marked received, and prints a days_to_close of 2. As of September 18, 2026 it had been open 811 days.

The same trap changes the headline number. Average the column across everything and you get 10.6 days, which is wrong in both directions: the 18,831 resolved requests averaged 2.8 days and the 1,460 canceled ones averaged 111, and only 747 of those canceled requests have a resolved date at all. The slowest request the city did finish took 565 days; the slowest canceled one carries 854. The three-second version: read current_status before you read days_to_close.

This too is the streetlight slice of something already published. The September 10 piece found days_to_close filled in on 46,572 requests across the whole file that never resolved. It is the same defect, and it has not been fixed.

The catch, part three: the map of the lights themselves

Here is where it stops being a data-quality note and starts being a gap.

Kansas City publishes a dataset at data.kcmo.org called, simply, Street Lights — n2wy-vqag. It holds 92,536 points. It has been viewed 1,859 times and downloaded 1,527 times, which is a lot of people looking for something.

Its description field is empty. The city says nothing at all about what the file is.

It has five columns:

There is no street address. No pole number of the kind stencilled on a pole. No lamp type, no wattage, no owner, and — the one a resident would want most — no status. Nothing in the file says whether a given light works, or is the city’s responsibility, or is still there.

Its rows were last updated in August 2021, five years ago. The rowsUpdatedAt stamp Socrata keeps on the file reads 1628990442, which is 01:20 on August 15, 2021 in UTC, the evening of August 14 in Kansas City. It has not moved since.

Then there is the last column. 3,938 of the 92,536 — 4.3 percent — carry a retirement date. The file the city publishes as its street lights includes thousands of lights it had already marked retired. Whether those poles are still standing is not something this file will tell you; what it records is that the city took them off the books.

The retirement column has its own problem, and it is the one that makes the whole file hard to use. Its dates run like this:

Then it stops. The newest retirement in the file is January 30, 2017 — four and a half years before the file itself stopped moving. Either Kansas City retired no street light between February 2017 and August 2021, or the column quietly stopped being filled in. The file cannot tell you which, and the practical consequence is the same either way: a blank retirement date here does not mean the light is in service.

So can you find out whether there is supposed to be a light on your block?

Not from what the city publishes, no.

The asset file gives you a coordinate but no address, so matching it to your street means working in latitude and longitude; it is five years old; and 4.3 percent of what it shows you was already retired when it was published. The 311 file is current and has a street address on all but three of its 20,332 streetlight requests, but it is a record of what people complained about, not of what the city owns.

You can learn that four neighbours have reported the same pole. You cannot learn whether the pole is the city’s.

Check us

# every streetlight request, by sub-type -- watch the pipe
curl -s "https://data.kcmo.org/resource/d4px-6rwg.json?\$select=issue_sub_type,count(1)&\$where=issue_type='Streetlights'&\$group=issue_sub_type&\$order=count_1%20DESC"

# the clock, split by status
curl -s "https://data.kcmo.org/resource/d4px-6rwg.json?\$select=current_status,count(1),avg(days_to_close),count(resolved_date)&\$where=issue_type='Streetlights'&\$group=current_status"

# the 41 still open, oldest first
curl -s "https://data.kcmo.org/resource/d4px-6rwg.json?\$select=workorder_,open_date_time,days_to_close&\$where=issue_type='Streetlights'%20AND%20current_status%20in('received','new')&\$order=open_date_time"

# the street light file: how many, how many retired
curl -s 'https://data.kcmo.org/resource/n2wy-vqag.json?$select=count(1),count(retired_date),max(retired_date)'

# when that file last moved
curl -s 'https://data.kcmo.org/api/views/n2wy-vqag.json'

Sourcing: all counts are from the City of Kansas City's open-data portal — dataset d4px-6rwg, "311 Call Center Reported Issues," and dataset n2wy-vqag, "Street Lights" — queried by The Kansas City Stare on September 18, 2026, together with each dataset's own published metadata. The 311 file grows daily, so its totals will be larger when you read this; the ratios and the label defect will not have changed unless the city fixes them. The captures and the queries behind every number are archived at editorial/_sources/street-lights-2026-09/. The 311 phone number, the call center's stated hours and the myKCMO routes are carried from this paper's September 10, 2026 guide to filing a 311 request, where they were sourced to a city department's own page; kcmo.gov refuses automated requests and did so again this week, so the hours are quoted rather than independently archived. No private individual, address, or request is identified in this piece. Where this story reports an inconsistency in the city's data, it reports it as a measured fact and does not assert a cause; The Stare has not established one.