According to the Bay Area Council Economic Institute (BACEI), there are roughly 28,200 homeless people in the California’s nine-county Bay Area, which includes San Francisco, Oakland, and San Jose. Extrapolating from previous research, I’m guessing about a third of these individuals are chronically homeless, defined as being without housing for at least a year. This is a tough bunch to help: between mental illness, physical disability, substance abuse, lack of social skills, a fierce independent streak, and/or neurocognitive disorganization, the chronically homeless are often unable to live normal, productive lives. No, most of these folks can’t “just get a job”.

In olden times, residential hotels and mental institutions housed similar individuals. For instance, San Francisco had 65,000 residential hotel units in 1910, compared to around 19,000 units today. These were teeny rooms with barely enough space for a bed and a dresser (bathroom down the hall) but at least they offered shelter. and safety from the streets. And, adjusting for the state’s population growth, California’s psychiatric hospitals house 93 percent fewer patients today than at their peak in 1959 (37,000 patients in 1959 versus 6,800 in 2017). If the number of residential hotel units had kept up with population growth and there were fewer legal barriers to institutionalizing individuals with severe mental illness, chronic homelessness wouldn’t be such an issue in the Bay Area. I imagine the same would be true for the country as a whole.

But we can’t go back to the good ol’ days. For one thing, communal showers and toilets wouldn’t work for the same reason public restrooms haven’t worked in San Francisco. This headline explains why: Sex, drugs and filth plague city-sponsored public restrooms. For another, construction costs are crazy in the big coastal cities. BACEI estimated it would cost $450,000 per unit to house the homeless in the Bay Area - and that estimate assumed units would be partially factory-assembled. On top of that are miscellaneous development fees, which easily add another $50,000 give-or-take per unit. And then there are endless delays while waiting for approvals and lining up contractors. There is, after all, a huge construction labor shortage in the US.

Housing the chronically homeless boils down to time and money, which boils down to politics and economic incentives. There’s no need for it to be so expensive and time consuming to build housing for the homeless. The approval process could be streamlined, development fees waived, and tax breaks given. Units could be assembled in efficient and low-cost factories* and transported by ship, rail and truck. This is doable.

Next: Converting dream into reality. How to build 10,000 housing units for the chronically homeless.

* One of the reasons the BACEI estimate was so high is that it assumed factory assembly of units would take place locally, given Bay Area political dynamics. But the requirement that units be assembled in or near the Bay Area adds to costs. “Local” can sometimes be worse for the environment and much more expensive than non-local alternatives. For instance, shipping from China leaves both a smaller carbon footprint and is less expensive than transport by truck. To illustrate: the units for a recently opened four-story “prefab” apartment building in Berkeley were shipped from China to the Port of Oakland and then trucked to the construction site less than 20 miles away. The developer noted that the cost of trucking to Berkeley from the Port of Oakland was “more expensive than the cost of shipping from Hong Kong”. Still, in the long run the US will need domestic housing factories, ideally located in less densely populated areas near rail transport, which also leaves a much smaller carbon footprint than truck transport.