Every finding, every week, counted in the open
Our data: Method · Track record · Data sources & updates
This page shows how we classified every building with a SWARMP or UNSAFE first report in cycle 10, how many cases are open, and how many dropped off the list each week and why. Counts only - no addresses. Updated with every weekly run - weekly runs so far: 4, the first on Oct 20, 2026.
How we classified all 1,439 findings
Out of 2,872 first cycle 10 reports, these buildings were rated SWARMP or UNSAFE. Each one is checked against every DOB job filing and permit (rules on the method page):
| Classification | Buildings | Share | On the list? |
|---|---|---|---|
| Open, no activity | 413 | 29% | yes |
| Open, firm known from earlier work | 181 | 13% | yes |
| Repair work under way since the report | 333 | 23% | no |
| Major project (demolition, new building, large alteration) | 6 | 0% | no |
| Amended to SAFE | 133 | 9% | no |
| Public owner (NYCHA incl. PACT/RAD, City, State) | 373 | 26% | no |
By borough
| Borough | SWARMP/UNSAFE reports | Open cases | no activity | firm known | with shed |
|---|---|---|---|---|---|
| Manhattan | 786 | 372 | 246 | 126 | 51 |
| Bronx | 226 | 82 | 57 | 25 | 8 |
| Brooklyn | 272 | 63 | 50 | 13 | 4 |
| Queens | 148 | 72 | 56 | 16 | 6 |
| Staten Island | 7 | 5 | 4 | 1 | 0 |
| Total | 1,439 | 594 | 413 | 181 | 69 |
Cases that dropped off, week by week
Every Tuesday each open case is checked again. A case leaves the list when a repair, contractor or scaffold permit for repairs is issued, a repair is filed, a major project starts, the report is amended to SAFE, or the owner turns out to be public. This table is the honest measure of the list: if many cases drop off soon after they appear, the list is late.
| Re-check | work under way | major project | corrected to SAFE | other / rule update | total |
|---|---|---|---|---|---|
| Oct 21, 2026 | 0 | 0 | 0 | 68 | 68 |
Run log
Every run of the list, with the DOB data it used: how many cases were open, how many were new against the previous run and how many dropped off. Several runs can use the same DOB data, because the City publishes some datasets less often than we check.
| Run | DOB data as of | open cases | new | dropped off |
|---|---|---|---|---|
| Oct 23, 2026 | Oct 19, 2026 | 594 | 3 | 4* |
| Oct 22, 2026 | Oct 19, 2026 | 595 | 0 | 0 |
| Oct 21, 2026 | Oct 19, 2026 | 595 | 0 | 68 |
| Oct 20, 2026 | Oct 19, 2026 | 669 | - | 0 |
The first run still contained six DOB test records (fictitious buildings the City keeps in its data); the rule update of Oct 21, 2026 removed them, so the open count fell by more than the dropped-off cases that day.
* Re-classified on the same DOB data after a rule update (for example a repair permit that was already in force when the report was filed): these cases are not in the removal log, so the number is the difference to the previous run.
New SWARMP and UNSAFE reports per month
First cycle 10 reports rated SWARMP or UNSAFE, by filing month, all owners. The Sub-Cycle 10A deadline on Feb 21, 2027 will bring a large wave; in cycle 9 about 45% of first reports were filed in the last three months before a deadline.
Hand-checked sample (Oct 20, 2026)
30 randomly drawn buildings (stratified by borough), checked filing by filing in nine datasets:
| Result | All 30 | Strict set (14) | Broad only (16) |
|---|---|---|---|
| Truly open, no facade activity | 16 (53%) | 9 (64%) | 7 (44%) |
| Open, contractor or facade work before the report | 6 (20%) | 4 | 2 |
| Work under way since the report | 6 (20%) | 0 | 6 |
| Completed | 0 | 0 | 0 |
| Unclear | 2 (7%) | 1 | 1 |
95% ranges (Wilson): strict set truly open 64% (39-84%), open including a known earlier firm 93% (69-99%). All six cases with work under way were in the broad set and are caught by today's rules (renewed permits of older jobs, a sub-filing of a large renovation, scaffolds for repairs, a demolition project). Small sample - wide ranges.
Second sample: the delivered list (Oct 21, 2026)
After an independent review we tightened the rules (shed age from the first permit of the shed job, Local Law 51 only for sheds first permitted on or after Jan 12, 2026, repair permits in force at the report count as work under way, DOB test records removed). Cases that left the list because of this re-classification are shown as "other / rule update" above, not as new repair work. Then 30 cases were drawn at random from the delivered list (stratified by borough and label) and checked permit by permit:
| Result | Cases | 95% range |
|---|---|---|
| No repair work on file since the report | 30 of 30 | 89-100% |
| No facade activity at all, also not in the three years before | 20 of 30 | 49-81% |
| Earlier facade work before the report (label "firm known") | 10 of 30 | - |
| Contractor named for the earlier work | 7 of 10 | - |
The check uses the same public DOB data as the list: work without a permit, and contracts that are signed but not filed, cannot be seen.
From report to repair: what cycle 9 shows
For 7,552 SWARMP or UNSAFE first reports in cycle 9 (filed July 2021 to June 2024), we looked for the first facade repair filing in DOB NOW:
Our own count of NYC Open Data (DOB NOW facade filings and job filings), research of Oct 20, 2026. The medians count only reports that got a filing or signal; across all 7,552 reports the median lies beyond 24 months (35% had a repair filing after 24 months). DOB NOW coverage starts in practice in 2021, so cycle 9 reports filed from July 2021 are used. Past cycles do not guarantee cycle 10.
Sources and caveats
NYC Open Data: xubg-57si, w9ak-ipjd, rbx6-tga4, ic3t-wcy2, ipu4-2q9a, 6bgk-3dad, 855j-jady, the DOB facade filing status list, HPD registrations and PLUTO. All counts are our own and updated weekly; see the method for the rules and their limits. Not affiliated with the NYC Department of Buildings.