Portland rent-board presentation: covered-unit medians under $1,600; zeros and exempt registrations distort dataset

Matt Walker, a Portland Rent Board member and the meeting’s data presenter, reviewed the city’s 2023–2024 rental-registration data and methodology at a February special meeting, saying the dataset should be interpreted only for covered units because many exempt units are entered as zeros in the registration software.

The presentation matters because the board must rely on the registration dataset to produce the annual rent report, to identify possible violations, and to set or recommend changes to board procedures. Board members and staff said the software’s requirement that a numeric value be entered forces many exempt units to show a $0 rent, which distorts simple counts and some summary statistics.

Walker told the board that the dataset snapshot used for the analysis compares “current” rent (November 2023) to “previous” rent (November 2022) because the full 2024 registration year was not yet available. He said he did not remove outliers except for $0 rents tied to exempt units and relied on medians rather than averages to reduce the influence of high-valued outliers.

Dylan, a Licensing & Housing Safety staff member who supplied and explained the registration data, summarized staff guidance on the zeros: “I have run into things where we thought there was a violation and turns out it was an exempt property and they just gave us all the data and did not select the exemption,” and added that some property owners register an otherwise-unoccupied unit with a 0 rent to preserve the record.

Key counts and findings Walker presented (as summarized at the meeting): 17,513 units were registered in the snapshot; about 11,000 of those were classified as covered units and roughly 6,500 as exempt. Licensing staff’s proactive auditing and enforcement covered about 3,900 units across roughly 550 properties during the reporting period, a figure the presenter said represented roughly 22% of units in the city.

On application volume, Walker read material from a quarterly report showing the board had received about 43 applications during the year; Walker said the annual report counted 14 approvals, 1 denial and 28 tabling actions. Dylan clarified to the board that the “28” figure represents the number of tabling actions logged during the year and can include multiple tablings of the same application, which inflates that count relative to the number of distinct properties still pending.

Walker presented median rents by unit type (covered units, after removing zero entries): studios roughly $1,100; one‑bedrooms about $1,400; two‑bedrooms about $1,700; three‑bedrooms about $2,000. He said the overall median rent for covered units—calculated after excluding zero entries tied to exempt registrations—was about $1,588. He emphasized those medians are medians of reported, covered-unit rents and cautioned against drawing conclusions from the exempt-unit rows, which he said “are not reliable for analysis.”

The presentation also covered rent-increase distributions. Walker reported most covered-unit increases clustered in the 5–10% band (the allowable annual-increase practices include AIP/CPI and turnover exceptions). He said about 6% of covered units reported increases greater than the allowable 10% ceiling, and that those registrations were targets for Licensing & Housing Safety investigations.

Board members and staff discussed presentation refinements requested before publication: (1) add a short executive-summary slide listing three key takeaways for council, (2) clarify the color-key and labels on district-by-district medians, (3) remove or footnote exempt-unit charts so readers do not misinterpret zeros as actual $0 rents, and (4) flag records reporting $0 for covered units for audit. Kristen, a Rent Board member, specifically suggested a brief year‑over‑year comparison on the executive summary so council members can see trends.

Recommendations Walker offered for the annual report and registry practice included: explicitly excluding exempt-unit zero entries from median and distribution calculations; a software-side “flag” or completeness check that prevents or flags clearly inconsistent registrations (for example, registrations that claim a $0 rent for a covered unit or increases over 10% without a supporting board approval); producing a “rental history card” for each registered unit that shows base rent and all allowable increases (to clarify banked increases); and giving applicants a link to the live spreadsheet with plots so other board members can reproduce or replot the analysis.

Board members thanked Walker and staff for the analysis and asked staff to correct labeling and presentation issues before the report goes to council. There were no formal votes on data-reporting practices during the workshop; the board directed staff to take the suggested edits and return a revised draft in follow-up meetings.

Walker closed by reiterating that the board should treat the covered-unit dataset as the authoritative source for rent-control analysis and that exempt-unit entries should be either reported in a clearly separate table or require legislative change to capture the same fields as covered units.

Next steps: staff will prepare a revised draft of the annual report and the board will consider the draft and the proposed completeness-check edits at a future meeting. No formal policy changes were adopted at the workshop.