San Francisco's first official cost estimate for reopening the Upper Great Highway to weekday cars landed Monday at $9.8 million upfront plus up to $1.9 million a year — roughly $20 million over five years — and within hours both campaigns were selling it as vindication.

The number, from the Office of the Controller, is the city's first authoritative price tag on Proposition G, the "Great Highway for Everyone" measure headed to the November ballot. It comes in about $4 million below the upfront figure the city's own transportation and parks agencies floated last month — a gap the Yes campaign is treating as proof its opponents inflated the cost. But the controller also did something the agencies never did: it counted the recurring annual bill to run cars on the road, turning a one-time capital number into a multi-year obligation the No campaign now calls even more damning.

A letter signed by Deputy Controller ChiaYu Ma, released Monday, put the one-time cost of dismantling Sunset Dunes and reopening the road between Lincoln Way and Sloat Boulevard to weekday traffic at $9.8 million. After that, the analysis found, keeping cars on the stretch would cost between $500,000 and $1.9 million every year in added maintenance and operations.

The controller's office broke the upfront bill down: roughly $750,000 to strip out park features and restore the roadway, about $8.8 million to install traffic signals at eight intersections, and $200,000 for restriping and signage. The recurring annual costs would come largely from signal maintenance and from resuming the sand removal and dune reshaping that has been far cheaper since the road closed to cars in 2024. Notably, the office cautioned that the actual bill depends on future funding decisions by the mayor and Board of Supervisors, because an ordinance cannot bind future budgets.

According to the SF Standard, that $9.8 million upfront figure is about $4 million lower than the $13.95 million the San Francisco Municipal Transportation Agency and Recreation and Parks Department estimated last month — a number Mission Local reported after it was obtained by the No on G campaign. Those agencies never attached an annual cost to the reversal. The controller did, which is why a "cheaper" headline number is being read two very different ways.

Proposition G would amend the Park Code to allow cars on the Upper Great Highway on weekdays while keeping it closed from Friday evening through Monday morning and on holidays. It is the third time in four years voters have been asked to decide the fate of the two-mile coastal stretch. In 2024, Prop K converted the road into Sunset Dunes and passed 55 percent citywide — but 64 percent of Sunset voters opposed it, and the neighborhood went on to recall its supervisor, Joel Engardio, over the closure.

For the campaign fighting the reversal, the controller's math is a gift. "Proposition G will charge San Franciscans millions of dollars to close a beloved park," Lucas Lux, manager of the No on G campaign, said in a statement. "This expensive ballot measure would waste taxpayer money to replace a thriving park with a failed road, a section of which has closed anyway due to coastal erosion." Friends of Sunset Dunes called the official estimate more damning than the agencies' figure precisely because it exposes the ongoing annual costs the earlier number ignored.

The Yes campaign read the same document as a win on the top line. "We are pleased to see that the inflated number that our opponents are spreading has been cleared up," Jamie Hughes, an advocate for Prop. G, told the SF Standard. Hughes argued the recurring costs would be spent regardless of whether the road is open or closed and that the bulk of the money funds pedestrian safety: "Prop. G will make our streets safer and save lives," adding that residents "understand that these costs are a small price to pay for the safety of their families."

There are also complications neither side advertises. Reopening the road would require a coastal development permit, which Streetsblog San Francisco reported could trigger further studies or permit conditions, and new signals at Sloat Boulevard would force changes to the Ocean Beach Climate Change Adaptation Project — potentially a costly redesign of a planned public plaza. What the controller's letter cannot settle is the question voters have already answered twice. Multiple traffic studies of the closure have found no appreciable impact on congestion, and the park now draws up to 14,000 visitors on a weekend. The controller priced the road. Whether San Franciscans want to pay for it is back on the ballot Nov. 3.