Detroit Population
Source: U.S. Census 1940
Two company towns, two declines
A comparison I've heard since at least 2014 is that San Francisco is on an inevitable path to becoming the next Detroit. In each case, a small city with a bureaucratic government and an economic base rooted in a single industry experiences massive growth alongside that industry. Tensions grow between the industry and the city. When the industry outgrows the city, the city is left with less revenue and a mess to clean up. Detroit became the center of American industrial production after World War II — the result of government-funded, fully modernized supply chains and policies that made auto manufacturing impossible to lose money on. San Francisco became the center of the 21st-century economy through a similar set of conditions: massive capital flows, government policy that favored tech, and a labor market that drew talent from across the country. Both cities transformed quickly. Both grew unequal. Both, eventually, watched their dominant industry begin to outgrow them. The arcs rhyme. The question this page asks is how much, and where the rhyme breaks down.
The rise and fall of the home of America's auto industry
A chronological timeline of Detroit's transformation from wartime industrial capital to a city in crisis. Beginning with the bombing of Pearl Harbor and the federal mobilization that made Detroit the heart of American manufacturing, the sequence traces the auto boom, the seeds of its own undoing in suburbanization and racial exclusion, and the 1967 riots that marked the city's break with its postwar identity.
Source: U.S. Census 1940
Peak Detroit population
Source: US Census 1950
11% decline from the previous decade
Source: U.S. Census 1960
Source: U.S. Census 1970
From dotcom recovery to pandemic exodus
A chronological timeline of San Francisco's emergence as the capital of the 21st-century economy. Beginning with Google's IPO and the rise of Web 2.0, the sequence traces the iPhone era, the consolidation of Big Tech, the local backlash, and the pandemic-driven shift to remote work that called the city's centrality into question.
Detroit's wartime mobilization wasn't the result of entrepreneurial genius. The federal government banned civilian auto production in February 1942, then guaranteed manufacturers' costs and a percentage of profits on wartime goods. The auto industry literally couldn't lose money. By June 1942, 66% of Detroit's machine tools were producing military goods. San Francisco's tech economy emerged similarly downstream of capital flows the city itself didn't create — dotcom-era venture capital, post-2008 quantitative easing, and pandemic-era stimulus that flooded into tech companies whose products suddenly became essential. In both cases, the city benefited from policy and capital decisions made elsewhere.
Approximately 500,000 workers — many of them Black Southerners — flooded into Detroit for defense work between 1940 and 1950. The city's housing, services, and social structure were not prepared. The 1943 Detroit race riot, with 43 dead and 433 injured, was the bloodiest racial episode of the decade. San Francisco's tech-era migration was smaller in absolute numbers but transformative in its effect on local cost of living. Bay Area real estate values tracked the tech industry's growth. Long-time residents who didn't own property were pushed out. The city's homeless population grew alongside its wealth. The conflicts looked different — protests against tech buses rather than housing riots — but the underlying dynamic was the same: a new population economically incompatible with the existing one.
Detroit perfected the very technology its residents could now use to leave. Postwar highway construction, financed and politically supported by the auto industry, made suburban living possible. The Federal-Aid Highway Act of 1956 was the master stroke — federal money to build freeways into cities that primarily served as exit routes. Detroit's population began declining within a few years. San Francisco built the internet that made San Francisco optional. Zoom launched in 2012; by 2020 it had made physical proximity to a tech workplace a choice rather than a requirement. The city had perfected the technology its residents could now use to leave.
Detroit's tax base depended on auto manufacturing wages. When manufacturing left the city for suburbs and then for cheaper labor markets, the tax base collapsed. Even today, Detroit takes in less tax revenue than it did from property taxes alone in 1963. San Francisco's tax base depends heavily on tech companies and high-earning tech workers. The city's budget more than doubled from $6.4 billion in 2010 to $13.7 billion in 2020 — almost entirely on the back of one industry's success. If that industry's interests diverge from the city's, the brittleness becomes visible.
The single most important reason San Francisco didn't follow Detroit's exact path is housing supply. Detroit had cheap and available housing throughout its boom, which is how 500,000 workers could move in. San Francisco's byzantine housing policies — many of them enacted in response to the destruction caused by 1950s freeway construction — made it nearly impossible to build new housing. The result: San Francisco's population grew only 1.6% from 2003 to 2013, compared to Detroit's 14% growth from 1940 to 1950. The "exodus" risk that Detroit faced — millions of people leaving — was structurally impossible in San Francisco because there were never millions of people to leave.
Detroit's influx was largely poor rural Black Southerners with no economic option to return to where they came from. Their presence created a permanent population of city-dependent residents whose interests were tied to the city's services. San Francisco's tech-era influx was high-earning workers with abundant options. When the industry shifted to remote work, many of them left — for Austin, Miami, Bend, Nashville, or back to their hometowns. The city has shed high earners but retained its long-term residents, the opposite of Detroit's pattern.
By 1962, only Chrysler and Cadillac assembled vehicles in the actual city of Detroit. The factories moved. The physical economic activity moved. San Francisco's tech industry has not physically moved in the same way — even with remote work, most major companies still operate offices in San Francisco. Apple is in Cupertino, but Apple was always in Cupertino. The companies that defined Web 2.0 still have headquarters in or near the city. The exodus is of workers and personal residences, not of the underlying economic base.
Six years after the original version of this piece, San Francisco is changed but standing. The population is lower than its peak. Office occupancy has not recovered to pre-pandemic levels. Public services are visibly strained. But the city did not enter terminal decline, and it would be wrong to call it a failed comparison just because the worst-case prediction didn't land. What the comparison did get right is the structural lesson: cities whose tax base depends on a single industry are brittle in predictable ways, and the industry's success is not enough to save the city if the industry's interests diverge from the city's. Detroit's auto industry was profitable for decades after Detroit's collapse began. San Francisco's tech industry may continue to be profitable for decades after San Francisco's current strain. The relevant question isn't whether the industry succeeds — it's whether the city remains aligned with it. The Rhymes exploration on this site exists to surface these structural patterns. Not to predict outcomes, but to make patterns visible so the reader can form their own view. If the rhyme between Detroit and San Francisco still holds in another six years, more of the original prediction may come true. If it doesn't, the rhyme itself was still real — just not deterministic. History doesn't repeat. It rhymes. Sometimes the rhymes resolve into outcomes we expected. Often, they don't.