Deindustrialization and the American City

Deindustrialization has made it impossible for America to economically support a stable urbanized working class, resulting in the social and physical decay of cities that is only partially ameliorated by more recent urban gentrification and the accompanying service sector jobs. Fresh thinking is needed in everything from education to architecture to city planning to transportation, but it is so far lacking.

Part One

Industry and the City in America

The process of industrialization in the United States began to take off around the time of the Civil War. The rapid growth of the railroad across the nation, coupled with the increasing surpluses provided by mechanized agriculture, provided a base for the development of industry and the resultant rise of cities and the mass urban labor force. From around 1840 to the early 1950s, industry consistently grew as a portion of the American labor force, interrupted only by the Great Depression^[1]. To accommodate this industry, a wave of city building followed, adding growth to the old trading cities of the East such as Philadelphia and New York, and leading to the rise of new interior cities like Detroit, Chicago, and St. Louis. This process continued through the late 19th and early 20th centuries, stalling in the decade or so following the Second World War with the widespread growth of suburbs. City building in America has largely been in decline ever since.

The contours of the development of the US industrial base can be seen by tracing the development of the American machine tool industry. Machine tools are a good proxy for the overall robustness of an industrial ecosystem, as they form the base technical layer for much of heavy industry.^[2] In the United States, production of machine tools was scaled up during the Civil War, and by the 1880s high precision machine tools were being mass produced.^[3] Mass production continued to scale up through the Second World War, and technological innovation in the industry continued through the first decades of the Cold War—but by the 1970s the industry in the U.S. was stagnant.^[4] Competitors in Germany, Japan and China dominated. For the purposes of this investigation, this developmental course can be read as a rough periodization of the rise and fall of American industrial capacity in general.

The United States became majority urban sometime between 1910 and 1920—but American urbanization unfolded unevenly across space and time, generally spreading on the coattails of industry from the Northeast, then to the Midwest, then to the West, and finally to the South. Rail coverage followed this general pattern of spatial expansion, with mileage tripling between 1860 and 1880, and again between 1880 and 1920.^[5] The first states to become majority urban were Massachusetts and Rhode Island around 1850, and the Northeast as a whole became majority urban by 1880.^[6] Both the Midwest and the West became majority urban in the 1910s. As for the South, that region did not become majority urban until the 1950s, following its belated postwar industrialization (and the concomitant advent of air conditioning).^7]

In American cities, flourishing industries provided masses of American workers with social mobility on a scale never before seen. After World War II, deals negotiated between labor and management institutionalized collective bargaining within the industrial sector, giving workers an unprecedented degree of economic security. Walter Reuther's Treaty of Detroit, negotiated in 1949 between the United Auto Workers union and General Motors, is a typical example of these arrangements. In exchange for protecting GM from annual strikes, it gave workers health coverage, retirement benefits, cost-of-living adjustments for wages, and guaranteed vacation time.^[8] For millions of Americans at this time, the American dream was in fact achievable.

To a large extent, this system of political economy subsidized many other parts of city life. It supported entire local business ecosystems, with small businesses catering to local workers. It supported a strong local tax base, which went to funding public infrastructure such as schools, parks, firefighting, and policing. And most importantly, it supported a dense and complex network of social capital, which allowed families, groups of friends, religious institutions, and voluntary institutions all to flourish and mediate community life. The industrial city was bustling and vibrant—for example, in 1941, a streetcar ran every sixty seconds through the heart of Detroit.^[9]

This era is often romanticized, but its high level of social stability is indeed visible by certain measures: the homicide rate, for example, fell to its lowest-ever rate in the industrial era in 1935 as the system of political economy matured under New Deal reforms, and stayed there until the mid-1960s, when the dividends of deindustrialization started to come due en masse in America’s cities.^[10]

Industrial capital also drove the physical development of America’s cities to unforeseen heights, literally, financing the likes of the Chrysler Building even during the Great Depression. Every major U.S. city now has a cluster, sometimes a forest, of skyscrapers housing mostly service industries. Beyond cultural icons, though, the United States used its industrial surplus to invest in its infrastructure at the highest rate ever in its history, allowing cities to enjoy a well-functioning physical environment.^[11] In fact, America has not really built a city in this fashion since the beginning of deindustrialization. Newer cities such as Orlando or Las Vegas grew largely due to suburban sprawl—not organized around urban life in any communal sense—with large commuter populations who spend the vast majority of their lives outside of downtown.

The American urban landscape during peak industrialization, around the year 1950, reveals the context in which deindustrialization occurred. According to the 1950 Census, the top ten most populous cities in the U.S. at that time were, in order: New York, Chicago, Philadelphia, Los Angeles, Detroit, Baltimore, Cleveland, St. Louis, Washington D.C., and Boston. Since the previous census in 1940, all of these cities had registered consistent growth. The order of this list had shuffled around a bit in the intervening ten years, but its contents remained unchanged except for the replacement of Pittsburgh by Washington D.C., which had grown rapidly with the size of the federal government during the war.^[12]

This 1950 census is interesting as a snapshot of the American urban environment at its height, on the cusp of decline. Of these cities, all except Washington D.C. had grown on the back of industrialization. What’s more, the population of each of these cities save for Los Angeles and New York peaked in this census and as of 2020 have never come close to their 1950s-era populations (New York for its part began a decades-long decline but eventually recovered its earlier numbers). In all of the cities that witnessed population decline besides Washington D.C., deindustrialization played a large part in this process.

Part Two

Deindustrialization

The general trend of American deindustrialization is evident in basic national labor data. Total manufacturing jobs in total peaked in late 1943 at the height of World War II, then fell back to prewar levels as the country demobilized from war, and then continuously rising from there until the early 1970s, after which job growth stagnated and finally peaked in 1979, when it entered a slow decline lasting until 2000. At that point, the decline became precipitous, with the country losing nearly 36% of its manufacturing jobs during the decade between 2000 to 2010. Since then, aggregate manufacturing jobs have seen some growth, though the 2020 shock from the COVID-19 pandemic brought them back down to 2010 levels.

As stark as the absolute decline of manufacturing jobs is (especially after 2000), the raw numbers alone tell only part of the story—obscuring even more significant underlying trends. This is because while the 1970s saw stagnation—and post-1970s, decline—of manufacturing jobs, this happened while the national population, labor force, and economy were all growing. Even when manufacturing jobs were technically growing in aggregate before the 1970s, they were already failing to keep pace with the growth of the overall labor force.

From this perspective, it makes sense to examine the decline of the manufacturing sector as a proportion of the workforce as a whole. When we do so, we can see that the roots of American deindustrialization lie earlier in time than the post-1979 decline in the raw number of manufacturing jobs itself suggests. The data show that manufacturing jobs as a percentage of the labor force peaked at around 39% in 1943, and as the war effort wound down in 1945 they expectedly fell back to their prewar value of around 30% of the labor force. After this adjustment, the manufacturing sector held stable at between 30% to 33% of the American workforce, but around the years 1956-1957, this portion began to decline. The decline at first was slow but constant, and it began to accelerate slightly around 1970. This post-1970 rate of decline held mostly steady until 2010, with industry bottoming out at around 8% to 9% of the labor force. It has stayed there since.

Deindustrialization has unfolded very unevenly. As industry in America was a heavily regional—and often even a city-by-city—mode of economic organization, American deindustrialization was a regional process as well. It occurred at a level of granularity not captured by the national analysis above. Some areas of New England, for example, began to deindustrialize as early as the 1920s following the collapse of the local industrial base after World War I; whereas some areas of the South did not begin to lose factories until the mid-2000s.^[13]

This has been a very heterogeneous process. We can identify multiple distinct phases, such as the deindustrialization of cities like Detroit, St. Louis, and New York beginning in the mid-to-late 1950s; and the Federal Reserve policy of the period 1979-1984, when interest rates were raised to quash inflation. Among other things this led to the rise in international value of the dollar, eviscerating the American machine tool industry in the face of Japanese competition; and leading more recently to the deindustrialization of California’s low- and high-tech manufacturing industries starting in the 1980s and 1990s.^[14] The most recent phase has consisted of the most extreme decimation of manufacturing jobs in American history, which occurred between 2000 and 2010 thanks to such convergent blows as NAFTA, the global automotive industry crisis, and the “China shock”—the miscalculation that China would remain a mostly low-wage, low-skilled labor economy after U.S.-China trade relations were normalized in 2000.^[15]

[...] deindustrialization directly undermines the economic sustenance of entire communities, which in the absence of other economic opportunities naturally leads to impoverishment.

The societal effects of deindustrialization are in principle similar everywhere, regardless of local conditions. In areas that rely on industry, deindustrialization directly undermines the economic sustenance of entire communities, which in the absence of other economic opportunities naturally leads to impoverishment. This often leads to a vicious circle: the loss of homes, the loss of pensions, a reduction in the local tax base, a reduction in local public services provision, a loss of resources for local institutions, and physical urban decay. Moreover, deindustrialization in the United States almost always means a concomitant loss of employee healthcare benefits, as healthcare in America is very often tied to employment.^[16]

In addition to pulling the material base out from under a particular community, deindustrialization also deprives workers of access to the stable social role of “having a steady job” and participating in a public sphere of others with the same role. The strong social fabric that comes with an industrial sector is often irreplaceable by relatively precarious service sector work. This material and social erosion then begins to work its way up the deindustrialized community’s entire social and institutional stack: in addition to poverty and mass unemployment, deindustrialization has been shown to bring with it the collapse of social capital, and the collapse of social capital has in turn been demonstrated to lead to family breakdown, loss of public trust, the collapse of local institutions, and rising rates of violence, drug abuse, depression, and suicide.^[17]

In many cases in American history, deindustrialization has also stoked racial tension. Accelerated industrialization across the North in the first half of the 20th century induced millions of African-Americans to move out of the rural South and into northern cities, an event known as the Great Migration. They found mass employment in new industrial jobs—although their conditions were often poor relative to white workers.^[18] The stresses of deindustrialization have contributed to race riots, such as those of 1967 Detroit, and even long-term breakdowns of public order, a condition that can be found to this day in large swathes of many post-industrial cities.^[19]

The material and social collapse heralded by deindustrialization, coupled with intense racial conflicts, often led to the wholesale abandonment of urban centers. The most notorious example of this is White Flight, the mass abandonment of certain city neighborhoods by whites starting in the 1950s and 1960s, but often those with the means to leave collapsed urban centers will do so, regardless of race (see for example the continuous outflow of Baltimore’s black middle class). ^[20] Urban breakdown caused by the loss of industry has induced urban depopulation, and urban depopulation in turn further deprives cities of resources, leading to a feedback loop from which it is nearly impossible to escape under conditions of economic hollowing-out.

The extent of this hollowing-out is best shown by example, and examples are unfortunately easy to find. Most apparent is the degree of urban depopulation. Detroit is the only American city ever to grow above one million people and then contract back to below that number, having lost over 60% of its people since 1950—a time span during which the national population has nearly doubled.^[21] St. Louis lost 64% of its population, Baltimore lost nearly 40%, and Chicago lost over a quarter—all mostly to the new suburban developments springing up across the country. The vicious cycle of urban decay and increased suburbanization also often hits city finances particularly hard, with the most notorious example being New York City’s near-bankruptcy and resulting bailout from the federal government in 1975.^[22]

Such massive shifts aside, the severity of deindustrialization is best grasped at the local level. During an infamous event known as “Black Monday,” on September 19, 1977, in the city of Youngstown, Ohio, the steel manufacturer Youngstown Sheet & Tube announced the termination of 5,000 manufacturing jobs, effective immediately. Youngstown was heavily dependent on this industry for its basic existence as a center of economic and social life, and Black Monday heralded the city’s deindustrialization. In the few years after Black Monday, the city lost 40,000 manufacturing jobs, 400 businesses which the industrial sector had supported, over $400 million in personal income, and in some localities up to 75% of school tax revenue.^[23] Since then the city has lost nearly 60% of its population, and has not economically recovered.

This is not an isolated or temporary trend. In a study on the long-term nature of deindustrialization, researchers at Youngstown State University note that the effects of deindustrialization have been so pervasive as to permanently alter the characteristics of unemployment nationwide.^[24] In 1979, they note, 8.6% of the unemployed had been out of work for more than six months, while in 2005, this figure had more than doubled, coming in at 19.6%. Post-industrial impoverishment, too, is a long-term condition. The authors note that a few years after the major factory closures in Gary, Indiana, one fifth of that city’s households lived in poverty. Today, over a third of Gary residents remain impoverished.^[25]

The dynamics of urban collapse precipitated by deindustrialization are most infamously associated with Rust Belt cities such as Youngstown, but as deindustrialization has become a near universal trend across the American urban landscape, so too have its effects become universalized, as the Youngstown State researchers crucially noted. For example, they write that in 2009, the recently deindustrialized city of El Centro, California had an unemployment rate of 24.5%, comparable to the 25% rate recorded in Youngstown in 1980.

Part Three

The Post-Industrial City

As our labor force participation rate continues to sink, and our post-industrial class sorting continues apace, it is becoming increasingly difficult to make any class distinction in the American city outside of identifying winners (or more accurately, survivors) of deindustrialization, and the losers. There is largely nothing in between. Urban areas in the United States no longer have very much of a propertied bourgeois middle class, with that class largely relegated to small-business owners in exurban and rural areas.^[44]

It is true that American cities have experienced something of a renaissance since the 1990s, with the re-entrance of elites into many major cities, a trend known as “gentrification”.^[45] The most direct benefit of gentrification has been an increased tax base in many major American cities, particularly New York City and Washington, D.C. However, the benefits of this process have not accrued to existing post-industrial working-class communities, which despite building booms bringing new service sector businesses into their deindustrialized neighborhoods have largely remained mired in poverty.^[46]

From this we can conclude that gentrification has not meaningfully altered the political economy of the deindustrialized American city. Though increased demand for urban services may have greatly increased the number of urban gig workers doing things like driving for Uber Eats, this has not led to the sectoral stabilization of the urban working class, which has remained mostly dissolved since the flight of industry. Thanks to the precarity of the service sector and the hyper-precarity of ad hoc gig economy jobs, and in the absence of other stable options for mass employment, we can expect any advancement in the condition of urban service workers will not be permanent and structural, but rather temporary and highly contingent on the levels of consumption levels of urban elites, rather than structural. Raising the national minimum wage to $15, if it happens, won’t solve the problem, either.

What’s more, the shock of the COVID-19 pandemic may have caused a resumption of the post-war American tendency towards suburbanization. The pandemic threw class dynamics into sharp relief, with increasing numbers of knowledge workers able to work remotely and considering permanent escape from the crowded city, while “essential workers” remain tied to their service jobs or deal with unemployment. Elites can choose to leave the city at any time, and as skyrocketing real estate prices in the New York City's suburbs can attest, they may be gearing up to do so yet again.^[47]

Due to the fact that the vast majority of prosperity in our post-industrial age has redounded to the elite, much of what happens in America’s cities is contingent on the behavior of this class.^[48] Whether it’s in the realm of politics, the economy, or culture, the elite now wields a greater level of influence over society simply because many of the social classes below them have essentially collapsed.

If the portion of the elite that has urbanized since the 1990s once again flees the cities, the urban working classes may very well lose what little security the gentrification-era service boom provided to them, and without industry, there is nothing we can do to ensure their long-term economic well-being. We have been taught this lesson for seventy years, but we have yet to learn it. This is unsurprising, as the American elite is by its nature isolated from the immediate economic realities of deindustrialization. They do not bear the material or social costs of deindustrialization, so we should not expect to see a societal focus on these deep problems anytime soon.

To Think About

• Can the changes brought about by the lockdowns and disruptions of the COVID-19 pandemic contribute to the revitalization of American cities, or must they lead to further decline?
• If fewer people commute to city centers, is there an opportunity to re-purpose office buildings and even parking garages to say, housing, shopping and entertainment centers?
• Technology, especially in communications, has shifted factory jobs to knowledge jobs. How can we reform our educational system to embrace a significant number of working-class people, or their children, in these new jobs, enabling them to prosper in cities?