Last week, Portland City Council faced a major decision about the future of large-scale performing arts, one presented without a price tag. The Mayor’s office characterized the resolution as a planning step that commits no immediate funds. Under the resolution, the council would accept the steering committee’s recommendations and advance a Broadway-capable venue on a Portland State University parcel, envisioned as part of the proposed Performing Arts and Culture Center. At the same time, council would begin planning to reimagine Keller Auditorium as a smaller venue. As Councilor Candace Avalos observed, the discussion centered on money the city does not have, for a cost it cannot yet estimate.
The case for the new venue rests on a 2020 seismic report and what followed. That year, the city commissioned Miller Consulting Engineers to evaluate Keller Auditorium and found that the building would not withstand a major earthquake. The council then adopted a two-venue strategy in October 2024: renovate Keller and construct a new venue at PSU. A later market study found that Portland could support only one 3,000-seat venue. Consultants recommended the new build at PSU; the steering committee and the mayor’s ex officio group concurred. Staff informed council that the two-venue strategy was no longer viable. That left council one path forward: vote on whether to plan for the new venue.
Four councilors raised the same objection: no firm number for the city’s share. Ryan, Green, Smith, and Kanal each pressed for specifics about the city’s financial obligations following approval. Councilor Kanal referenced a cautionary precedent, recalling when the administration assured council that advancing a budget item for the Tennis Center would not commit the city to privatization, only for the contract transferring the facility to the USTA to never return to council. Councilor Avalos, who had introduced amendments in the City Life committee to require council approval of the financial plan, came into the session confident in those measures but grew concerned as the discussion went on.
In response, Deputy City Administrator Oliveira said that staff cannot give council a number until it authorizes the work. He clarified that the resolution would allow staff to collaborate with PSU and begin the process. The immediate next step would be assembling the capital stack, a compilation of all funding sources that must collectively cover the project cost. Oliveira noted that if staff had presented a complete capital stack upfront, councilors likely would have objected to work done without their approval. For the December 1 update, he offered a range: maybe $44 million, maybe $100 million, maybe the full capital stack “because we hit the lottery.” He reiterated that the information is unavailable because the process has not yet begun.
A range that wide means the city’s share is unknown. The councilors’ unease is not mere procedural squeamishness; rather, it is an appropriate response when council is asked to make a major decision without knowing the cost, especially in a city that has grappled with budget gaps in consecutive years.
The question of risk allocation emerged as PSU’s representative made clear: “This would be the city’s venue.” While the previous council had invited PSU to propose, a letter of intent signed in November 2024 assigned PSU the tasks of assembling the capital stack, securing state funding, and leading philanthropic efforts, all of which PSU has reportedly accomplished, along with completing market, economic, and traffic studies. The representative said the state bonding and philanthropy were secured, and indicated that the hotel, academic and community spaces, and parking structure either have their costs covered or are supported by other partners. Every element of this statement casts PSU as a partner that has met its obligations and places ultimate ownership, and thus risk, on the city. But no dollar figure was attached to any of these assurances.
Ownership decides who pays for cost overruns. State bonds, grants, and pledges are capped. One uncapped layer remains. That layer covers any difference between the project cost and the sum of capped sources, including all overruns. The building’s owner usually bears this responsibility. Three comparable projects show what ownership has cost in other cities.
Miami-Dade County owns the Adrienne Arsht Center, a Broadway-capable complex built through its largest public-private partnership. It opened in 2006, two years late and more than $218 million over budget, at a final cost of about $473 million. In Kansas City, the Kauffman Center’s guaranteed maximum price climbed by 20 percent to $393 million, $67 million above the 2004 estimate, during construction. Los Angeles’s Walt Disney Concert Hall was intended to open in 1997 for around $110 million but instead opened in 2003 at roughly $274 million. Los Angeles County bonded $110 million for the garage beneath it, counting on parking revenue to repay the debt, but the delay left the county servicing that debt without the expected revenue. In each case, the public owner paid the overrun.
Most large public projects run over budget. Research compiled by Bent Flyvbjerg at Oxford puts cultural buildings among the worst categories.1 This trend is partly because landmark projects often undergo design changes after budgets are approved. For a project of this scale in Portland, an overrun comparable to Kansas City’s would amount to tens of millions of dollars. The councilors who asked who would bear that cost were asking the right question.
Council can move the project forward without writing a blank check. The seismic risk is real, the single-venue market conclusion stands uncontested, and the process PSU describes has been thorough. Council should adopt Councilor Avalos’s amendments and begin the December 1 discussion by establishing ownership terms rather than focusing solely on headline figures. Before committing, council must determine whether the city’s contribution will be a fixed amount or a residual obligation. It should also clarify whether any overruns will fall to the city as owner or be contractually shifted to PSU and its partners. Council should also ask what “secured” means in dollars for state bonds and philanthropy, and should scrutinize assumptions about operating revenue that may factor into debt service.
The administration says it cannot answer until council votes, but council can and should attach clear conditions to its approval. The experiences in Miami, Kansas City, Los Angeles, and with the Tennis Center all point to the same lesson: permission must be contingent on the condition Councilor Avalos proposed, that the complete financial decision returns to council, in full, before the city signs. The first estimate is never the final cost. Council must decide now who will pay the difference.
Flyvbjerg maintains the largest database of its kind, more than 16,000 projects across some twenty fields and 136 countries, going back to 1910. Fewer than one in ten finished on budget and on schedule. His books on the pattern are Megaprojects and Risk (Cambridge University Press, 2003) and How Big Things Get Done (2023).