Heather O'Donnell, senior planner for the City of Eugene, told the Envision Eugene technical advisory committee on Oct. 16 that staff has revised the methodology used to calculate housing density for the city's growth monitoring and Buildable Lands Inventory work.

O'Donnell said the new approach sums the total number of units and the total number of buildable acres across development sites, then divides units by acres, rather than averaging per-project densities. "We have come up with a better way to refine the actual densities that we're seeing from development permits," she said, adding that the change is intended to reflect the scale of projects more accurately.

The change matters for Eugene's long-range planning because density assumptions feed capacity estimates used in Urban Growth Boundary (UGB) and housing-supply work. "If you give equal weight to a single-unit lot and a 300-unit project, you misrepresent what the city is building," O'Donnell said.

Using the updated method, staff reported 4,570 units across 606 acres in land designated low-density residential, yielding an average of 6.7 units per acre. Staff also reported average densities of 5.6 units per acre for single-unit detached projects and 11.8 units per acre for townhouse projects. Developments on vacant land averaged 8.9 units per acre under the new calculation, O'Donnell said.

O'Donnell and Elena (associate planner) walked committee members through examples of how the data are cleaned and aggregated. One simple example showed a development site with duplicate records in the raw monitoring data because tax-lot and Buildable Lands Inventory (BLI) boundaries had changed since the original BLI run; staff remove duplicate rows and use the most recent entry to represent the site. In a more complex example, a multi-building project recorded four permit records was split between a general-office (GO) plan designation and medium-density residential (R-2); staff allocate the project's units to each plan designation and to the buildable acres associated with those parts of the site.

Committee members asked how nonresidential elements such as parking or protected natural areas affect density calculations. O'Donnell said density uses "buildable acres," which includes the acreage actually available to be developed on the site and therefore includes parking and other on-site uses when they are part of the buildable area, and that protected acres (for example, an adopted local protection area) are removed when they meet the criteria for protected land. "We're trying to get the number of units over the buildable acres," she said.

Staff described several category clarifications shown in the monitoring tables: group quarters (nursing homes, jails, public dormitories) are shown for transparency but are typically analyzed separately from conventional housing-unit counts; single-room-occupancy units are tracked as a separate category so they can be classified consistently with state definitions (less than six bedrooms as single-unit, six or more as multiunit); and assisted-living or similar specialized housing is flagged separately.

Committee members and staff discussed data revisions that can occur after permits are approved. In one example staff found a project whose permit records were later revised downward by four units; staff said updated numbers will be pushed into the reporting system as part of the year-end data push.

Elena and O'Donnell said the city has been working with consultants and the information-technology team to automate deduplication and other data-cleaning steps so the monitoring tables can be reproduced reliably. O'Donnell said the automated method preferentially uses the most recent row of data where duplicates exist so that permit revisions and staged building permits resolve to the current total for a site.

The committee asked about future data additions staff expect to track: tenure or condominium status is not currently captured at the time of building permit unless the site is platted as condo lots, but staff said they can query the regional land information database to identify condo lots after the fact and plan to refine monitoring to reflect Oregon's new housing-needs rules. Members also asked about floodplain mapping; staff said the group is tracking federal floodplain map changes and will treat land as protected for capacity calculations only if it is formally designated as protected under local rules or by federal designation.

O'Donnell said staff will send the revised, quality-controlled data set to consultants and to the committee for a final review; she expected to push the updated numbers to the reporting system by the end of the year and to incorporate the changes into the next published growth-monitoring package. The committee did not take a formal vote on the methodology at the Oct. 16 meeting and asked staff to return with any needed clarifications.

Staff and multiple committee members praised the level of detail in the new tables and the effort to document the workflow. "Having new people forces you to rethink how you share information and what people might not know," said Tiffany Edwards, Etech chair and vice president of the Eugene Area Chamber of Commerce. Longtime member Sue (listed in committee records as Sue) said the revised method and the documentation give her confidence in the data used for long-range decisions.

The next steps staff described include a final consultant review of the methodology and a continuation of the Buildable Lands Inventory and urban-growth-strategy work later this year; staff said they will notify the committee when the consultants' review is complete and when the updated reporting tables are published.