Protectionism, Economic Monoculture, and the Real Lessons of the Rust Belt

Executive Summary

In 1950, Detroit and Pittsburgh were among the most prosperous industrial cities in the world. Detroit led the world in automotive production, and Pittsburgh’s steel output was in many respects unmatched. However, both faced similar forces in the 1970s: foreign competition, technological change, and rising labor costs driven by union lock-in. Both turned to government for protection. By 2013, the differences between the two cities were stark. Detroit had filed for the largest municipal bankruptcy in American history, while Pittsburgh had remade itself as a center for healthcare, education, and technology.

This paper argues that the divergence between these two was not due to geography, luck, or the arrival of foreign competition. Rather, Detroit — through a series of decisions at the state and local levels — created an industrial monoculture. Pittsburgh did not.

An industrial monoculture is an economy whose fiscal base, labor market, and political institutions become tightly coupled to a single industry. While this industry thrives, the broader economy appears strong. However, industrial monocultures carry structural fragility, as any shock to the dominant industry spreads through the entire local economy. Worse, there are few, if any, alternative sectors that can absorb displaced labor and capital. Detroit’s monoculture was not a natural outcome of free markets. It was constructed and sustained through a series of protectionist policies. Union contracts in the Treaty of Detroit (1950) shielded the industry from domestic labor market discipline. Michigan’s Public Act 198, passed in 1974, was used by the auto industry to reduce its property tax burden. The Poletown eminent domain seizure in 1981 resulted in the government buying $200 million worth of land only to sell it to General Motors for $8 million. The Michigan Economic Growth Authority tax credits, first passed in 1995, were later expanded to subsidize the Big Three automakers. At the federal level, the auto industry benefited from voluntary export restraints in the 1980s and an $80 billion bailout in 2009.

Each layer of protection reduced the competitive pressures that compel firms to adapt. As a result, adaptation largely stopped. By the time Japanese automakers arrived in the American South, Detroit’s industry had spent 25 years insulated from the discipline that would have kept its costs, productivity, and product competitive. Foreign competition did not cause Detroit’s collapse. It was a series of policy choices, each built on the last, that calcified Detroit’s auto industry around expensive labor contracts and products that were simultaneously more expensive and less reliable than those of its foreign counterparts.

Pittsburgh, by contrast, did not do this. It did not rewrite its eminent domain laws to save blast furnaces or create a credit program to subsidize steelworker wages into the 2030s. Instead, Pittsburgh’s economy retained universities, hospitals, and research institutions that predated and operated independently of the steel industry. When steel collapsed, capital and labor in and around Pittsburgh had alternative sectors to absorb them.

The lesson here is not that government can reliably create economic diversity. Markets do this far more efficiently and with fewer opportunities for rent-seeking. The lesson is that single-industry protection is one of the most dangerous forms of intervention a community can choose, precisely because it creates the very fragility that it claims to prevent. Today’s industrial policy, which includes tariffs, subsidies, mandates, and “Buy American” requirements, applies the same concentrating logic but at a national scale.

The difficulties many Rust Belt communities face today are real. They are not the result of foreign competition or free markets, but of past protectionism. The protectionism of the present, in turn, proposes many of the same tools that helped produce the problems these communities now confront.

Key Points:

• The difference between Detroit and Pittsburgh is not explained by geography, luck, or foreign competition, but by economic structure. Detroit became an industrial monoculture, while Pittsburgh retained a more diverse economic base.
  
• Industrial monocultures are highly productive in good times but structurally fragile. When an economy’s tax base, labor market, and political institutions are all tied to a single industry, any shock to that industry cascades through the entire local economy.
  
• Detroit’s industrial monoculture was not the product of free markets. It was constructed, layer by layer, through a sequence of protectionist policies.
  
• Pittsburgh avoided industry-specific protectionism. It did not rewrite eminent domain law or create targeted subsidies to preserve a declining industry.
  
• Protectionism weakened Detroit by removing the competitive pressure that forces adaptation.
  
• The Rust Belt’s challenges today are real. They are also the result of protectionism, not free markets or foreign competition.
  
• Today’s industrial policy risks recreating the same dynamics that caused Detroit’s decline, but at a national scale.
  
• The protectionism of the past created many of the very problems that present-day protectionists claim their policies will solve.

1. The Political Temptation of Protection

Imagine that you are a politician. One of the largest employers in your district employs tens of thousands of workers, generates millions of dollars in local tax revenues, and supports a dozen or more ancillary industries. Its workers are organized, politically active, and vocal. The solution seems obvious: implement policies designed to keep that industry afloat and the workers employed. Unlike the workers and firms that stand to benefit, these costs are spread across millions of people who are unorganized, politically disengaged on the issue, and often unaware of how much the policies affect their wallets.

Mancur Olson identified this dynamic in 1965 (Olson, 1965). Concentrated interests have powerful incentives to lobby for protection from competition. The costs of protection, however, are spread thinly across a much larger population that has little incentive to organize against it. The nearlyinevitable result is that policy bends toward the vocal and organized minority at the expense of the silent and unorganized majority. Over time, this tendency only bends further as the protected industry grows more entrenched and economically dependent on protection for survival.

This dynamic is especially powerful when an industry is concentrated in a single city. Here, the politics are especially fierce, and the long-run consequences are especially devastating. A protected industry does not just distort prices or misallocate resources from an economic efficiency standpoint. It shapes the city’s economy, institutions, and identity. Entrepreneurs build businesses that serve the protected industry because that is where opportunities are most promising. Capital flows toward the protected industry under the presumption that the protection will continue. The city stops diversifying, resulting in an industrial monoculture, where the local economy is so thoroughly organized around a single industry that workers have few alternatives should that industry eventually decline.

No industry dominates forever, however. Technological progress marches on, foreign competitors improve, and consumer preferences change. The question that matters for communities is whether they are prepared to absorb these pressures when they arrive — either by innovating to remain competitive or by pivoting to new opportunities. Protectionist policies attempt to postpone that reckoning by shielding industries from competitive pressures. In doing so, however, they often allow underlying problems to compound. When market reality finally arrives, it does not find a community ready to adjust. It finds one that has been falling behind decade after decade while becoming increasingly dependent on a single industry. What could have been an unpleasant but manageable process of industrial transition becomes a catastrophic collapse.

The economic histories of Detroit and Pittsburgh are instructive examples of the problems of industrial monocultures and what economic diversity prevents. Both rose to industrial dominance in the first half of the twentieth century. Both faced foreign competition and technological disruption beginning in the 1970s. Both had powerful unions that lobbied aggressively for protection from these pressures. Yet their trajectories ultimately diverged. One city has filed for the largest municipal bankruptcy in American history. The other became a widely-cited model of post-industrial revival.

The story of why one city failed while the other thrived is not one of geography, demographics, or luck. It is a tale of two different responses to industrial decline. Detroit organized its economic and political life around the automotive industry through an escalating suite of interventions and protections. These ranged from relatively modest measures, such as state tax credits for the Big Three to keep workers in Michigan, to extraordinary actions, including rewriting eminent domain law to seize an entire residential neighborhood to build a factory. Pittsburgh, by comparison, pursued no comparable strategy. It did not underwrite steelworker wages with state tax credits, rewrite their eminent domain laws to save a blast furnace, or bet the state treasury on successfully keeping one industry alive at all costs.

Detroit’s ongoing recovery proves the point from the other direction. As the city gradually allowed its economic base to diversify rather than embracing an automotive monoculture, it has begun growing again for the first time in generations.

Foreign competition and technological change posed genuine economic challenges, but they were not what turned Detroit’s decline into a catastrophe. That outcome was largely the result of policies that discouraged diversification and deepened the city’s dependence on a single industry.

2. Detroit: Building and Defending a Monoculture

According to the 1950 Census, Detroit was home to 1.85 million residents and 757,722 workers. The Big Three automakers — Ford, General Motors, and Chrysler (now Stellantis) — were booming. Of further benefit were the ancillary industries that had set up shop nearby. Steel mills, metal fabrication shops, tire manufacturers, not to mention the marketing, telecommunications, and construction companies, were all opening at breakneck pace. According to the 1950 Census, of the 757,722 people employed in Detroit, 210,747 (27.8 percent) worked directly for the automotive industry. Another 187,770 worked in related occupations, bringing the total to 398,517 workers, 52.5 percent of the city’s workforce.

In 1950, the United States produced 8.0 million vehicles, while global production totaled 10.6 million. Detroit alone produced 5.3 million vehicles — roughly half of the entire world’s automotive output. Each day, some 14,520 vehicles rolled off Detroit assembly lines and into driveways, garages, and business fleets around the world.

The economic concentration was unmistakable. Detroit earned its nickname, “The Motor City,” because its identity was inseparable from the automotive industry. Its economy, politics, and civic culture all revolved around automotive manufacturing. Detroit was, for all practical purposes, an industrial monoculture. When the industry thrived, Detroit thrived. And when it faltered, so too did the city.

2.1 Locking in the Structure

The United Auto Workers transformed Detroit’s economic concentration into political power. Founded in 1935 with 25,769 members (Fine, 1958), the union initially struggled to organize automobile plants. The passage of the National Labor Relations Act of 1935, combined with the 1937 Battle of the Overpass increased support for the UAW. By 1955, UAW membership exceeded 1.5 million. By 1970, approximately 95 percent of Detroit’s automotive workforce was unionized, compared to a national private-sector unionization rate of roughly 35 percent.

Where Detroit was once controlled by The Big Three, the Big Three were now squarely controlled by the UAW, giving the union significant control over the economic policies of one of America’s wealthiest cities.

In 1950, UAW President Walter Reuther (later profiled in death as “the most dangerous man in Detroit”) secured what became known as the Treaty of Detroit with General Motors through collective bargaining (Lichtenstein, 1995). Similar deals soon followed with Ford and Chrysler. The agreement guaranteed annual raises that were greater than any cost-of-living adjustments, pensions of up to $117 ($1,600 in today’s money) per month, and 50 percent coverage of any hospital and medical insurance costs for union members. It also mandated that all new hires at the automotive company became members of the UAW for the first year of their employment but could then quit the union if they so desired. Over the coming decades, more provisions were added to the Treaty of Detroit.

In 1955, the UAW negotiated the Supplemental Unemployment Benefits (SUB) with Ford Motor Company. Under the arrangement, Ford agreed to pay five cents for every man-hour worked into a dedicated account. Workers who were laid off — provided the layoff was not the result of misconduct — could then receive up to $25 per week (about $300 in 2025 dollars) per week from the fund in addition to state unemployment benefits. The program quickly became a model for other Detroit automakers by the early 1960s. Even in 1956, however, the problems with this plan were beginning to show, as labor costs were skyrocketing (Problems of the Ford Plan, 1956). These benefits were extended in 1967 to cover a longer time period and to offer even more money (Linder, 1968). Combined with state unemployment benefits, senior workers were able to take home up to 92 percent of their pay when laid off.

In 1970, the UAW negotiated a “30 and out” retirement plan, whereby any UAW member who had worked for 30 years at an automotive plant could retire and receive full benefits (Lichtenstein, 1995). Someone who had gone to work at 18 could, under this new plan, retire at 48 and receive a full pension. The pension would be divided into two parts: the basic benefit, which guarantees about $19,000 per year, and a supplemental benefit, which matches what a worker will receive from Social Security once they retire. By 2007, total hourly labor costs at the Big Three, which included wages paid to workers plus legacy pensions and healthcare obligations, were just under $73 per hour compared to $48 per hour at non-union Japanese factories in southern US states (Wyman, 2007).

All of this led to substantial increases in autoworker compensation relative to the national average. By 1970, autoworkers were paid upward of 40 percent more than manufacturing workers in non-automotive sectors. Detroit remained buoyed by the postwar boom and low gas prices, which sustained demand for bigger and faster cars (Rae, 1984). Unfortunately for Detroit, this period would not last.

These arrangements carried significant and direct financial costs on the Big Three. They also included detailed work rules governing how many employees were to be assigned to particular tasks. Seniority systems constrained the ability of management to redeploy labor in response to changing conditions. A firm facing intense competition cannot sustain a cost structure that renders it uncompetitive indefinitely; it must either adapt or exit the market.

2.2 Protection without Adaptation

By the time Japanese automakers arrived in American markets in the late 1970s, Detroit’s automotive sector had faced limited competitive pressure to adopt new production technologies for decades. Japanese manufacturers entered the market with several structural advantages over their Detroit based counterparts. Japan paid far less in labor per hour ($6.20 per hour compared to the Big Three’s $16.80 per hour). They were also further along in automating production processes. By 1980, Japanese auto plants could produce 22 cars per worker per year compared to about 15 for Detroit’s Big Three. Each vehicle required roughly 100 hours of labor in Japan versus about 150 in Detroit.

The differences were not limited to production efficiency. Japanese vehicles were generally less expensive, more fuel efficient, and more reliable, requiring fewer repairs over time (Lincicome, 2022; Crandall, 1987). These advantages were especially consequential in the American market, where consumers were facing rising fuel prices in the wake of the 1973 and 1978 oil shocks.

Federal policy played a significant role in shaping the competitive environment of the automotive industry. The Reagan administration, seeing the protectionist impulses of Congressional Democrats, negotiated the voluntary export restraints (VER) with Japan beginning in 1981 (Reagan, 1990). The VERs limited Japanese auto imports and provided Detroit with a reprieve at the federal level. The International Trade Commission estimated that 44,000 manufacturing jobs were preserved (US International Trade Commission, 1982).

Michigan had been layering state and local protections on top of federal trade policies for years. In 1974, the state enacted Public Act 198, the Plant Rehabilitation and Industrial Developments Act, which allowed local governments to grant manufacturers property tax abatements of up to 50 percent for as long as 12 years on new or rehabilitated facilities. This funneled more capital investments into the industry and away from alternative options.

In 1980, Michigan rewrote its eminent domain law through the Uniform Condemnation Procedures Act, opening the door for the use of eminent domain for commercial development projects. The following year, Detroit Mayor Coleman Young used the law to seize 465 acres in the Poletown neighborhood for a General Motors plant. The project displaced 4,200 residents and led to the demolition of 1,300 homes and 140 businesses. The city acquired the land for roughly $200 million and sold it to General Motors for $8 million, along with a 12-year property tax abatement estimated to be worth another $60 million. In return, GM promised to create 6,000 jobs. The plant opened in 1985, employed some 3,000 workers, and closed in 2019.

In 1995, Governor John Engler created the Michigan Economic Growth Authority (MEGA), a refundable tax-credit program designed to attract new jobs to the state. Over time, however, MEGA increasingly became a tool for retaining existing jobs that were threatening to leave. In 2009, Governor Granholm expanded the number of MEGA tax credits to help keep the Big Three anchored in Michigan. The program stopped issuing new credits in 2011, but by then Michigan had already authorized some $12 billion in MEGA credits, including about $4.5 billion promised to the Big Three in exchange for retaining roughly 86,000 jobs through 2032.

These federal and state policies did exactly what they were designed to do: remove competitive pressures. But industries insulated from competition have little reason to bear the costly adjustment process that adaptation and innovation require. As a result, the quality gap between American and Japanese vehicles widened through the 1980s (Crandall, 1987) and 1990s. Every new layer of protection reinforced Detroit’s industrial monoculture and narrowed economic alternatives.

2.3 The Collapse

These struggles eventually reached Detroit’s residents. By 1990, the city’s population had fallen to just over one million, while the number of people employed in the automotive sector stood at roughly 280,000 workers (US Census Bureau, 1990). In 1950, by comparison, Detroit was home to 1.85 million people and 338,000 autoworkers. The consequences were significant. Detroit’s tax base eroded along two dimensions. First, fewer residents meant fewer people living, working, and spending money in the city, reducing tax revenues. Second, declining property values sharply reduced property tax collections. This led to schools becoming underfunded, which exacerbated middle-class flight from Detroit, leaving behind a larger share of lowerincome households to support the city’s finances. Detroit’s debt burden consequently grew from relatively modest levels in 1990 to $18 billion in 2008, while its bond rating deteriorated, limiting access to capital markets (City of Detroit, 2013).

Although the financial crisis of 2007-2009 affected the entire nation, Detroit was hit especially hard. By 2008, the city had more than 67,000 foreclosed properties — the highest foreclosure rate in the nation — and median home values had fallen below $10,000 in many neighborhoods (City of Detroit, 2009). National unemployment stood at 7.3 percent in 2008. Detroit’s unemployment rate reached 22 percent. At the same time, the city’s poverty rate climbed to 35 percent.

The automotive sector also suffered during the financial crisis. In an effort to restore profitability, Detroit’s Big Three increasingly relied on high-margin vehicles such as SUVs and pickup trucks rather than smaller sedans. This strategy proved costly when fuel prices surged. Between 2000 and 2010, average gasoline prices more than doubled, undermining demand for the vehicles on which Detroit had become most dependent.

Figure 1: US Regular All Formulations Gas Price, 2000-01-01=100

Source: US Energy Information Administration via FRED®, Federal Reserve Bank of St. Louis. Shaded areas indicate US recessions.

From 2000 to 2008, domestic auto sales fell by 30 percent, declining further once the recession began.

Figure 2: Motor Vehicle Retail Sales: Domestic Autos, Jan 2000=100

Source: US Bureau of Economic Analysis via FRED®, Federal Reserve Bank of St. Louis. Shaded areas indicate US recessions.

The effects of these long-running trends became most visible in the aftermath of the financial crisis. In 2010, Detroit was a city whose infrastructure and footprint were designed for a population of 1.85 million, despite having only 713,000 residents (US Census Bureau, 2010). In some areas, 80 percent of the houses stood vacant. Houses could be purchased for as low as $100 provided, the buyer assumed the outstanding property taxes. That same year, thenMayor, Dave Bing announced that “his administration cannot afford to go on providing services such as schools, firefighters, buses and rubbish collection to large areas of the city where the population has dropped sharply,” (McGreal, 2010). He noted that “fewer people paying property taxes has left a $300m hole in the budget.” While no residents would be forced to move, those who did not move to the still-covered areas of Detroit were told they would “need to understand that they’re not going to get the kind of services they require.”

In 2013, Detroit filed for Chapter 9 bankruptcy, the largest municipal bankruptcy filing in US history. The extent of this was so severe that Detroit considered liquidating the art collection housed at the Detroit Institute of Art. This was ultimately averted when businesses, foundations, and the state of Michigan agreed to donate more than $800 million as a part of the city’s debt restructuring plan (Smith, 2013). By the end of 2014, Detroit had emerged from bankruptcy with $7 billion of its total $12 billion in unsecured debt either restructured or written off, along with about $1.7 billion set aside for improvements to city services (City of Detroit, 2013).

The bankruptcy marked the culmination of long-term structural decline. Sixty years of protectionist policies at the local, state, and federal levels had shielded the Big Three from competitive pressures to innovate, to control costs, and to adapt. The policy environment intended to stabilize the industry and protect it from foreign rivals had instead contributed to its longrun fragility, which eventually left Detroit more vulnerable to the collapse of a single-industry monoculture.

3. Pittsburgh: Economic Diversity

Pittsburgh in 1950 was the undisputed steel capital of the world. Allegheny County had a population of 1.5 million, with the city of Pittsburgh itself being home to 676,000 residents (US Census Bureau, 1950). Hundreds of thousands of jobs were supported by industry giants such as US Steel, Duquesne Steel Works, National Tube Works, and Allegheny Ludlum (Hoerr, 1988). US Steel alone employed over 300,000 workers and produced 35 million tons of steel annually. At its peak, a single Pittsburgh furnace could outproduce Great Britain in steel production, while the region’s industry as a whole exceeded the combined steel output of the Axis powers during World War II.

Like Detroit, Pittsburgh’s industry was shaped by powerful unions and government protection. The United Steelworkers of America (USW) represented roughly 650,000 workers by the early 1950s. In 1952, amid the Korean War, national demand for steel remained elevated for weaponry, vehicles, and infrastructure. The USW had been negotiating for months with steel companies over wages and working conditions, citing the increased demands placed on workers during the war. However, the executives at the steel mill pointed out that they could not afford to meet the demands of the union without raising steel prices, something the Truman Administration had explicitly forbidden in 1951 with price controls.

In response, the USW threatened a nationwide strike to begin on April 9, 1952 (Hoerr, 1988). President Truman responded with Executive Order 10340, which directed Secretary of Commerce Charles Sawyer to seize and operate the steel mills (Truman, 1952). The order was wide-reaching, covering 88 steel companies operating more than 500 plants nationwide and representing about 90 percent of the US steel production capacity. The seizure resulted in Youngstown Sheet & Tube Co. v. Sawyer, which the Supreme Court decided on June 2, 1952, in a 6-3 ruling against the administration, requiring the immediate return of steel mills to private ownership.

Amid the Youngstown legal battle, the planned strike went into effect. It lasted a total of 53 days. The economic consequences were staggering: steel production declined by about 21 million tons, while workers lost an estimated $400 million in wages, and defense production significantly slowed. The dispute ultimately concluded after the USW secured modest increases in wages and fringe benefits for workers.

Then, in 1959, there was another strike at steel plants nationwide. This time, steel executives sought a change in the union’s contract that would allow the company to change crew sizes, revise work rules, and implement new machinery in order to reduce the amount of labor used in the production of steel. The union opposed these changes, and on July 15, 1959, the 519,000 members of the USW went on a strike.

After 116 days, President Eisenhower invoked the back-to-work provisions found in Section 206 of the Taft-Hartley Act. This was upheld by the Supreme Court in an 8-1 decision in Steelworkers v. United States, and workers were ordered to return to work. Although the strike ended, productivity slowed in its aftermath, presumably due to the poor relationship between workers and management and low worker morale (United Steelworkers of America v. United States, 1959).

The same year, cracks in the steel industry began to show. For the first time, the US imported more steel than it exported (Hoerr, 1988). The rest of the world was beginning to catch up in terms of steel production. Because other countries were free to use new production techniques and technologies, foreign steel producers had acquired an edge over their US counterparts.

The 1960s saw rising foreign competition in the domestic steel industry, especially from Japan, which was undergoing rapid industrialization. According to a report from the US International Trade Commission, by 1967, “imports [of steel] had grown to the point that there was congressional interest in establishing quotas on imports of iron and steel products,” (US International Trade Commission, 1982). Seeing this, in 1968 “both West Germany and Japan proposed to place voluntary restrictions on their steel exports to the United States in order to forestall the imposition of quotas” (McClenahan, 1991). In 1969, the agreement was put in place, limiting imports from the two countries to 5.75 million net tons. Armed with newfound protection from foreign competition, the steel industry was poised to make a comeback.

Up to this point, the parallels between Detroit and Pittsburgh hold almost perfectly. Both were dominant in their respective industries on the national and world stages. Both had heavy concentrations of employment in their particular sector. Both had strong unions to contend with. And both had received substantial protection from the federal government in the form of voluntary export restraints and related trade measures. Yet the trajectories of the two cities would ultimately diverge.

3.1 The Difference That Mattered

There was, however, one crucial difference: Pittsburgh had an economic foundation that was not solely dependent on its dominant industry. The University of Pittsburgh and Carnegie Mellon University, for example, were mature institutions with national reputations. A network of hospitals existed, serving the community and employing thousands of people. These sectors were not part of the steel economy. They didn’t depend on steel revenues or sales. They were not organized around steel labor contracts, and they did not collapse alongside the steel industry.

This was not the result of a deliberate diversification strategy. Pittsburgh’s universities and hospitals were not a hedge against the decline of steel. Those institutions developed independently. While they may have benefited from subsidies, both at the federal and state levels, they nonetheless offered robust alternative destinations for capital, labor, and entrepreneurial activity.

Because Pittsburgh possessed economic assets beyond steel, lawmakers faced less pressure to preserve the industry at all costs. Pittsburgh did not layer state and local protections on top of federal protections in a bid to retain the steel industry. There was no MEGA-style tax credit program. There were no eminent domain seizures to give the steel industry additional land at taxpayer expense. In fact, quite the opposite: where Detroit used eminent domain to seize land from residents in Poletown in 1981, Pittsburgh went against the community’s desire to save the famed “Dorothy Six” blast furnace when it closed in 1984. Known around the world for its sheer size and productivity, the furnace had become a symbol of the city’s industrial strength. The city rejected several proposals to save the furnace, which was finally torn down in 1988.

With the increasing viability of these alternatives, the USW’s bargaining strength waned considerably over the 1980s. In 1983, after months of negotiations, the USW agreed to concessions (Serrin, 1983). Some of the specific concessions were a wage cut of $1.25 per hour, reduced vacation time, and limiting automatic cost of living adjustments only to years where inflation exceeded three percent.

In exchange for these concessions, the USW gained early retirement incentives, increased corporate funding for the Supplemental Unemployment Benefits program, and a “dignity and justice” provision whereby a worker had to be proven guilty of wrongdoing before they could be suspended or fired.

Armed with cost savings measures, the steel industry in Pittsburgh was poised to start making a comeback. Unfortunately, like the auto industry in Detroit, the steel industry in Pittsburgh failed to capitalize on this opportunity. Tornell (1997) describes this as “rational atrophy.” Briefly, management in the steel industry can allocate profits in three ways: reinvesting in steel operations, investing in other sectors, or distributing profits to shareholders. As Tornell notes, “the steel firm’s reaction to the excessive wage increases the unions pushed for was to reduce the share of profits they reinvested in steel.” In other words, rather than use the cost savings they had secured, management decided to let the steel industry atrophy instead of using that money to reinvigorate it, concluding that any investments in steel would ultimately fuel further labor disputes.

Tornell points to US Steel’s acquisition of Marathon Oil, Husky Oil, and Texas Oil and Gas in 1982, 1984, and 1985 as examples of this strategy. US Steel claimed that they did so because they were unable to secure financing to invest in steel. However, as Tornell notes, “this explanation is not fully convincing because US Steel used $1.4 billion of its own cash to buy Marathon Oil. In principle, it instead could have used this cash to invest in, for example, [new technologies like] continuous casting.”

This sent a clear message to the USW. While they had made many concessions to try to preserve jobs for their members, management had already begun shifting their focus and investments away from steel and toward new opportunities. When invited for subsequent rounds of negotiations, USW leaders largely declined. The result was a wave of steel plant closures throughout the 1980s. In early 1983, US Steel began shutting down its operations in Homestead and Rankin. Bethlehem Steel also announced that it would close its Lackawanna plant and reduce operations in Johnstown, eliminating 7,300 jobs. Later that year, US Steel announced the permanent shutdown of part or all of 28 plants and mines.

The contraction continued in 1984. The Duquesne Works facility, which housed the famed “Dorothy Six,” closed in 1984. Jones & Laughlin Aliquippa Works, once one of the largest steel mills in the entire world, also shut down. All told, between 1981 and 1986, over 150,000 steelworkers in the Pittsburgh area alone lost their jobs with an additional 95,000 manufacturing jobs in downstream industries vanishing (Hoerr 1988). Suburban towns, such as Homestead and McKeesport, both less than ten miles outside of Pittsburgh’s city center, saw their populations decline sharply. Unemployment in the region rose to 27 percent and only really came down when workers left Pittsburgh to find work elsewhere, with many settling in Birmingham, AL, then known as the “Pittsburgh of the South.”

While the steel industry never died off, it was clear that it was not going to return to its heyday of massive numbers of employment. The postwar boom the industry experienced, with the domestic car industry and the installation of railroads across the country, had ended and with it, the demand for new steel had seriously diminished. At the same time, minimills, which could recycle scrap steel and were not bound by union contracts, along with foreign producers, stepped in to serve much of the remaining market.

It was not until around 2000 that global demand for steel began to grow again, driven largely by the rise of China as an economic superpower. Unfortunately for Pittsburgh, China was able to fill most of their demand with Chinese-made steel.

3.2 Reluctant Diversification

Pittsburgh’s transition was neither clean nor market-driven in any simple sense. The latter half of the 1980s saw significant growth. In 1985, the University of Pittsburgh Cancer Institute was founded, which would eventually become part of UPMC’s Hillman Cancer Center. To support the expansion, the institutions drew on federal NIH money to build research capacity and attract top talent to the area. In 1986, UPMC began consolidating with other university-affiliated hospitals, laying the groundwork for a larger, integrated health care and research system.

At the same time, Carnegie Mellon University became one of the first six colleges to register a .edu domain on “the Internet.” With this, researchers at Carnegie Mellon were able to communicate not just internally, but externally with researchers at other schools such as Berkeley, Columbia, Purdue, Rice, and UCLA. Their ability for collaboration was leaps and bounds ahead of most peers and they were able to attract some of the top talent in the world to join their faculty, transforming CMU from, in the words of the Pittsburgh Post-Gazette, “a regional technical school into a national engineering powerhouse.” With this, their ability to attract research funding exploded from roughly $12 million per year in the 1970s to over $110 million by the late 1980s. Much of this money was used to build infrastructure and to attract talent, particularly in the field of robotics, with CMU launching the world’s first robotics PhD program in 1988.

The 1990s continued this trend, with the healthcare sector growing and expanding steadily year after year. In 1994, the UPMC formed the Tri-State Health System Network, extending its medical footprint into the surrounding states. Increased private and public investments in healthcare fueled economic growth. By the early 2000s, the healthcare sector had become Pittsburgh’s largest employer.

Figure 3: All Employees: Education and Health Services: Nursing and Residential Care Facilities in Pittsburgh, PA (MSA)

Sources: Federal Reserve Bank of St. Louis; US Bureau of Labor Statistics via FRED®. Shaded areas indicate US recessions.

Pittsburgh continued to innovate and attract more new businesses and industries. In 1999, Pennsylvania established Keystone Opportunity Zones, “defined-parcel-specific areas with greatly reduced or no tax burden for property owners, residents and businesses.” More broadly known as special economic zones, these areas provide powerful incentives for economic development and can help change local politics toward more market-friendly alternatives (Moberg, 2017). Importantly, these zones were not directed toward any industry in particular.

In 2006, Google opened a Pittsburgh office, putting Pittsburgh on the map as a tech center (Carter, 2016). In 2009, Duolingo, the language learning app, was founded as a part of Carnegie Mellon University’s Olympus incubator. In 2015, Uber and Carnegie Mellon University formed the Uber Strategic Partnership and Advanced Technologies Center. That same year, Facebook opened its Oculus VR office in Pittsburgh. The following year, Amazon opened a Pittsburgh office focused on machine translation, Alexa integrations, and Amazon Web Services.

While not a laissez-faire story, the critical difference between Detroit and Pittsburgh is one of diversification. Pittsburgh’s subsidies and government investments pointed in multiple directions at the same time. Healthcare, research and development, robotics, and business services operate in distinct markets, draw on different skills and, importantly, are largely insulated from one another. Healthcare demand does not depend on the investment cycles of the robotics sector, and vice versa. The portfolio of opportunity is the opposite of a monoculture.

4. Detroit’s Comeback

Fortunately, Detroit’s story has a final chapter. Even before the city filed for bankruptcy, the seeds of recovery were in the ground. In 2010, Rock Ventures, led by its CEO, Dan Gilbert, had begun investing in the city. Between 2010 and 2024, Gilbert’s company had invested $7.5 billion in real estate, which led to the creation of more than 17,000 new jobs and making Rock Ventures the city’s largest private employer. For the first time in Detroit’s history, an automaker was not the city’s top employer (Feloni & Lee, 2018).

In the process, Gilbert helped break the industrial monoculture that had defined Detroit for decades. The city was becoming a technology hub with companies like Detroit Labs and Newlab. Financial services followed, with companies such as Rocket Mortgage, Rocket Money, and Detroit Venture Partners expanding downtown. Today, all four of the Big Four accounting firms — Deloitte, EY, KPMG, PwC — maintain offices in downtown Detroit. Attracted by low property values and growing economic opportunity, other companies came as well, like StockX, which sells high-end consumer goods, and Fathead, maker of “officially licensed and custom wall decals.”

All of this spurred a resurgence in population; in 2023 Detroit experienced its first population growth since 1957 and has continued to grow (City of Detroit, 2024). The downtown population is younger (57 percent are between the ages of 25-34) and more educated (45 percent have a bachelor’s degree and 34 percent have a master’s or professional degree).

Detroit’s recovery did not come from saving its automotive monoculture. It came from allowing new industries to fill the space that six decades of concentrated automotive protections, subsidies, and political incentives had inculcated. The same city, geography, and infrastructure that produced economic hardship under the monoculture are producing genuine recovery today through diversification.

5. Implications for Today

The story of Detroit and Pittsburgh is, in part, a story of how different cities respond to the same political and economic pressures. Those same pressures are alive today.

There is an irony at the center of today’s industrial policy debate. The communities that advocates of protectionism and industrial policy want to help the most are struggling in large part because of protectionism and industrial policies of the past. The decline was not caused simply by “free markets” or “foreign competition.” It was shaped by decades of tariffs, quotas, export restraints, union contracts backed by political guarantees, tax abatements, and direct subsidies. Each of these insulated incumbent industries from the competitive pressures that would have forced adaptation and diversification. The result was not resilience, but dependence on a narrow industrial base.

A market-based approach points in the opposite direction. Prices, not politics, should guide capital and labor to where they are most productive. Allowing failing firms and industries to contract frees resources so they can be redeployed elsewhere. This process is disruptive and often painful for workers and communities in the short run. But as Mokyr (2016) argues, with the right institutional and cultural settings, the process of adaptation rather than calcification leads to greater growth and continued prosperity.

The current wave of industrial policy in 2025 and 2026 is animated by the same logic that guided earlier protectionists. Calls for “supply chain security” have been replaced by “national security” and “good jobs” has shifted toward “putting American workers first,” but the underlying mechanism is familiar. Concentrated industries with organized lobbying power capture diffuse costs borne by unorganized consumers and workers in other sectors. Over time, the protected industry stops adapting and the gap between what it produces and what the market wants widens. When protections either stop or prove insufficient, the reckoning is more severe than it would have been had adaptation occurred earlier.

Politically, the difficulties of a free-market approach are real. Markets allow industries to decline. These declines are visible, concentrated, and potentially devastating to specific places. Organized workers in a declining industry form exactly the concentrated interest group that Olson described. Even elected officials who understand that protection is economically harmful often support it because allowing large job losses is politically perilous.

This does not rescue industrial policy from its fundamental flaws. Governments cannot identify tomorrow’s winners in a reliable way. Subsidies create political dependencies that outlast their rationale and become a national security risk in and of themselves. Concentrated intervention leads to the same dynamics Olson warned about. The first-best answer of letting markets determine industrial structure, allowing prices to direct resources, and exposing firms to competition remains.

The lesson of Detroit and Pittsburgh is that single-industry protection is the most damaging form of intervention a community can choose. It undermines the diversity that would allow the community to survive the decline of a dominant industry.

6. The Real Costs of Protection: Stagnation, Fragility, and Dependence

Detroit and Pittsburgh rose together on the same postwar tide and saw their dominant industries decline for largely identifiable reasons. Detroit organized its economic and political life around a single industry and spent decades defending that concentration against the competitive pressures that might have forced adaptation. When that industry finally began to decline, the city had little else to fall back on. Pittsburgh, by contrast, maintained a more diverse economic base and survived the collapse of its dominant sector because it had other sources of growth.

Detroit’s bankruptcy was not a random act of fate. It was the predictable outcome of an industrial monoculture rooted in sixty years of concentrated economic and political investment in a single industry. The city’s recovery, which began only after the economic base diversified, underscores the dangers of industrial monoculture. Today, Detroit is rebuilding itself around technology, finance, and business services rather than automotive assembly, and is experiencing growth for the first time in generations.

Left to their own devices, markets tend to produce diversity. No single industry dominates indefinitely, and rational actors spread their bets rather than concentrating them. Detroit’s monoculture did not emerge from markets. It was constructed through politics. Union contracts that locked in labor rigidities, tariffs, and voluntary export restraints insulated incumbents from competition, and government subsidies favored existing industries rather than enabling new ones. Every act of protection deepened the monoculture and narrowed the viable alternatives within the community.

The first-best prescription remains the same: let markets work. Prices, not politics, are better suited to directing capital and labor toward their most productive uses. But the second-best prescription, that takes seriously the political realities that policymakers face, is equally clear: policymakers should avoid concentrating support on a single industry. Economic resilience comes from economic diversification.

References

Bureau of Economic Analysis. Regional Economic Accounts: Per Capita Personal Income by Metropolitan Statistical Area. Washington, DC: US Department of Commerce. https://www.bea.gov/data/incomesaving/personal-income-county-metro-and-other-areas

Bureau of Labor Statistics (2025). Quarterly Census of Employment and Wages (QCEW). Washington, DC: US Department of Labor. https://www.bls.gov/cew/

Carter, D.K. (Ed.). (2016). Remaking Post-Industrial Cities: Lessons from North America and Europe (1st ed.). Routledge. https://doi.org/10.4324/9781315707990

City of Detroit (2009). Neighborhood Stabilization Program Plan. Planning and Development Department. Retrieved January 12, 2026 from https://detroitmi.gov/Portals/0/docs/Planning/PDF/NSP/detroitNSP_R31 _29_09_2.pdf

City of Detroit (2013). Chapter 9 Petition for Municipal Bankruptcy, Case No. 13-53846. US Bankruptcy Court, Eastern District of Michigan.

City of Detroit (2024). Detroit Grows in Population for the First Time in Decades. May 16, 2024. Retrieved January 12, 2026 from https://detroitmi.gov/news/detroit-grows-population-first-time-decades

Crandall, Robert (1987). The Effects of US Trade Protection for Autos and Steel. Brookings Papers on Economic Activity, vol. 1, pp. 271-288.

Feloni, Richard and Samantha Lee (2018). Billionaire Dan Gilbert has invested $5.6 billion in nearly 100 properties in Detroit. Business Insider. Retrieved January 12, 2026 from https://www.businessinsider.com/dangilbert-detroit-properties-bedrock-map-2018-8

Fine, Sidney (1958). “The Origins of the United Automobile Workers”. The Journal of Economic History. 18 (3): 249–282. 27Why Detroit Failed and Pittsburgh Recovered

Hoerr, John (1988). And the Wolf Finally Came: The Decline of the American Steel Industry. University of Pittsburgh Press.

Lichtenstein, Nelson. 1995. The Most Dangerous Man in Detroit: Walter Reuther and the Fate of American Labor. New York: Basic Books.

Lincicome, Scott (2022). Rewriting Protectionist History. CATO At Liberty. https://www.cato.org/blog/rewriting-protectionist-history

Linder, Walter (1968). Aftermath of the 1967 U.A.W. Strike: Sellout and Insurgency in the Auto Industry. PL Magazine, March-April. https://www.marxists.org/history/erol/1960-1970/pl-auto-67.pdf

McClenahan, William. (1991). The Growth of voluntary export restraints and American foreign economic policy, 1956-1969. Business and Economic History, 180-190.

McGreal, Chris (2010). Detroit mayor plans to shrink city by cutting services to some areas. The Guardian, December 17, 2010. https://www.theguardian.com/world/2010/dec/17/detroit-shrinking

Moberg, Lotta. (2017). The political economy of special economic zones: Concentrating economic development. Routledge of London.

Olson, Mancur (1965). The Logic of Collective Action. Harvard University Press.

Problems of the Ford Plan (1956) “Supplemental Unemployment Benefits: Problems of the Ford Plan,” Indiana Law Journal: Vol. 31(3), Article 7.

Rae, John B. (1984). The American Automobile: A Brief History. Chicago: University of Chicago Press.

Reagan, Ronald (1990). An American Life: The Autobiography. Simon & Schuster.

Serrin, William (1983, March 2). Steel Union Leaders Ratify Concessions. The New York Times. https://www.nytimes.com/1983/03/02/us/steel-unionleaders-ratify-concessions.html

Smith, Roberta (2013). In Detroit, a Case of Selling Art and Selling Out. The New York Times. https://www.nytimes.com/2013/09/11/arts/design/indetroit-a-case-of-selling-art-and-selling-out.html

Tornell, Aaron (1997). “Rational Atrophy: The US Steel Industry,” NBER Working Paper 6084 (1997), https://doi.org/10.3386/w6084

Truman, H. S. (1952). Executive Order 10340: Directing the Secretary of Commerce to Take Possession of and Operate the Plants and Facilities of Certain Steel Companies. Federal Register.

US Census Bureau (1950). 1950 United States Census of Population. US Department of Commerce. Retrieved January 12, 2026 from https://1950census.archives.gov/howto/statistical-data.html

US Census Bureau (1990). General Population Characteristics – Michigan. US Department of Commerce. Retrieved January 12, 2026 from https://www2.census.gov/library/publications/decennial/1990/cp-1/cp1-24.pdf

US Census Bureau (1990). Summary Population and Housing Characteristics. US Department of Commerce. Retrieved January 12, 2026 from https://www2.census.gov/library/publications/2012/dec/cph-1-24.pdf

US International Trade Commission, (1982). Economic Effects of Export Restraints. USITC Publication 1256. Washington, DC.

United Steelworkers of America v. United States, 361 US 39 (1959). https://supreme.justia.com/cases/federal/us/361/39/

---

Friday’s July employment report offered an unusually good occasion to revisit something I wrote about almost exactly a year ago: the widespread misunderstanding of revisions to the monthly jobs numbers. On August 7, the Bureau of Labor Statistics reported that nonfarm payroll employment fell by 23,000 jobs in July, while the unemployment rate edged down to 4.1 percent. At the same time, May payroll growth was revised from +129,000 to +63,000 and June from +57,000 to +20,000: a combined downward revision of 103,000 jobs.

Those numbers certainly suggest that the labor market has weakened. The three-month average payroll gain is now only about 20,000. But before turning a disappointing jobs report into either a recession call or another round of accusations that government statisticians are cooking the books, it is worth remembering what these numbers actually are.

They are estimates. The headline payroll number comes from the Current Employment Statistics survey, in which BLS collects payroll information from roughly 119,000 businesses and government agencies representing about 622,000 worksites. That is an enormous sample, covering about one-quarter of American payroll employment. It is nevertheless still a sample. Not every establishment responds immediately, seasonal factors are continually recalculated, businesses are born and disappear, and additional information arrives after the first release. Consequently, revisions are not corrections of “mistakes” in the ordinary sense. They are updates to an estimate as information improves. BLS explicitly notes that the two preceding months are routinely revised as additional survey responses arrive and seasonal factors are recalculated. After two revisions, when nearly all reports have been received, the estimate is considered final.

This distinction matters particularly for July’s negative first print. The reported loss of 23,000 jobs sounds very precise. It is not. BLS estimates that the 90-percent confidence interval surrounding a monthly payroll change is roughly plus or minus 122,000 jobs. Applied mechanically to July, that means a first estimate of −23,000 is consistent with an underlying change somewhere in the neighborhood of −145,000 to +99,000. That does not make the estimate useless. It means that the proper way to read it is as one observation in an evolving stream of evidence rather than as a perfectly measured head count.

That was also the point of my analysis following the extraordinary revisions in the July 2025 report. At the time, May and June 2025 were revised downward by a combined 258,000 jobs. I examined the historical distribution of payroll revisions and found that it was decidedly unlike the neat bell-shaped distribution people often have in mind when they hear that something is “three standard deviations” from normal. Payroll revisions have fat tails: unusually large observations occur considerably more often than a normal distribution would imply.

Since then, I have substantially expanded that dataset and asked a more practical question: do revisions themselves tell us something about the business cycle? The answer is yes — but considerably less than headlines sometimes imply.

Among 484 observations in my sample where the economy remained in a National Bureau for Economic Research (NBER) expansion from the initial estimate through the final release, nearly 40 percent were revised downward. Downward revisions, in other words, are perfectly ordinary during good economic times. And during stable contractions, upward revisions were actually much more common than downward ones. A single negative revision is correspondingly weak as a recession signal. When an expansion is underway, the historical probability of entering recession within the next 12 months in my sample is about 12.2 percent. After one negative payroll revision, it is 11.0 percent: essentially no warning at all.

Persistence is somewhat more informative. After two consecutive negative revisions, the 12-month recession probability rises to 17.3 percent; after three, 16.7 percent; and after four, 20 percent. The strongest signal appears when repeated downward revisions also acquire meaningful size. Once an ongoing string of negative revisions accumulates to approximately 30,000 to 40,000 jobs, subsequent recession risk rises noticeably. At a cumulative −40,000, the historical 12-month probability reaches roughly 26 percent, more than twice the expansion baseline.

That is interesting, but it is not a recession alarm. 

The samples become small quickly, the relationship is not perfectly linear, and extremely large revisions are not automatically more informative than moderately large ones. The sensible conclusion is narrower: one disappointing number tells us little; persistent deterioration across successive estimates deserves considerably more attention.

(The above chart should be read as a measure of how the warning signal changes as downward revisions accumulate. During an ordinary expansion, the historical probability of entering recession within the next 12 months is about 12.2 percent. But among the 27 observations in which an ongoing sequence of downward payroll revisions accumulated to at least 40,000 jobs, 25.9 percent were followed by recession within a year: roughly twice the baseline rate. That does not mean a cumulative 40,000 job revision predicts recession, much less causes one; roughly three-quarters of those observations were not followed by recession. Rather, it suggests that once downward revisions become both persistent and sizable, they contain more information about deteriorating economic conditions than a single negative revision does.) 

I also subjected the data to some simple forensic tests because accusations that BLS numbers are politically manipulated have become routine. I looked for suspicious rounding and unusual final-digit patterns, calendar effects, election-year behavior, campaign-season anomalies, and partisan differences in revisions. The initial estimates and total revisions showed no suspicious digit patterns. Average revisions did not vary meaningfully by calendar month. Presidential election years showed no statistically meaningful revision bias, and during August through October of presidential-election years — the period when political incentives ought to be greatest if manipulation were occurring — average revisions were virtually identical to those in other months.

(The forensic tests look for statistical fingerprints that might suggest systematic distortion. Terminal-digit tests find no suspicious heaping in initial prints, first revisions, or total revisions; the unusual pattern in second revisions is consistent with later adjustments clustering near zero. A runs test finds mild persistence in the direction of revisions, but nothing inconsistent with changing economic conditions, while calendar-month tests find no recurring seasonal pattern. Most importantly, revisions in presidential election years and during the politically sensitive August–October campaign window are statistically indistinguishable from other periods. None of these tests can rule out manipulation categorically, but together they provide no statistical evidence of systematic political interference.)

There are patterns in the data. Economic statistics are not random numbers generated by a roulette wheel. Revisions cluster, business conditions change, reporting arrives unevenly, and turning points are particularly difficult to measure. Non-randomness is not evidence of fraud.

July’s −23,000 nonfarm payrolls number deserves attention, particularly alongside the weaker May and June estimates. But the lesson from nearly half a century of revisions is not that the latest number should be ignored. It is that we should resist giving any single first estimate more authority than it demonstrably possesses. The monthly jobs report is best understood not as a final measurement of the economy, but as a successive attempt to see something enormous, complicated, and constantly changing with necessarily incomplete information.

You’ve no doubt heard the horrifying stories of dictators who abused their people for personal gain or sheer cruelty. Stalin’s Great Purge, Mao’s engineered famine, and Pinochet’s brutal torture of dissidents are just a few notorious examples. If all dictatorships are hell, a difficult question remains: which offers the better odds of survival, or even prosperity? Is it a dictatorship that tolerates some free-market capitalism, or one built on socialism?

A disclaimer is in order: This is not a defense of tyranny. We should resist it, whatever the form — absolute monarchies, military juntas, or proletarian dictatorships. Whether it’s the few controlling the many or the many terrorizing the few, all policies repressing freedom, silencing dissent, or justifying violence against innocents are inconsistent with our natural rights. They are morally evil and unjustifiable.

“The more the state ‘plans’ the more difficult planning becomes for the individual.”

— Friedrich A. Hayek

Socialism has appeared in many forms — Revolutionary (USSR), nationalist (Germany), agrarian (Cambodia), on and on. Wherever it does not collapse quickly due to its economic defects, it inevitably destroys all political freedoms. Two types of tyrannical regimes emerge: one type allows private property and business initiative, the other squelches them.

Under a regime where the rule of law is replaced by the arbitrary rule of men, the economic system imposed still makes a world of difference. And the contrast is nowhere more vivid than in the struggles of Chile under Augusto Pinochet and Venezuela under Chávez and Maduro. 

Let’s start with the fundamental difference: Who controls the economy? In reality, none of the regimes provided consistent legal protection for property owners. Pinochet, Chávez, and Maduro all routinely violated these fundamental rights. But citizens and private entrepreneurs under Chile’s junta had much more control over their personal and economic decisions than those trapped in Venezuela’s socialist economy.

What might seem trivial is, in reality, the difference between meeting your basic needs as a human being and literal starvation. Capitalism means private ownership of the means of production. Under socialism, the state owns everything. Even if political rights are restricted, private ownership is much more desirable for the average person. That’s why life in authoritarian Chile was more bearable than in Venezuela after it embraced democratic socialism.

“Private property creates for the individual a sphere in which he is free of the state.”

— Ludwig von Mises

At the start of Pinochet’s rule, the Chileans faced tough economic challenges. None of their difficulties can compare to the mass food and basic supply shortages, hyperinflation, and large-scale emigration experienced by the Venezuelans under socialism. While Chile struggled with poverty and unemployment during parts of the regime, the economy improved significantly following pro-market reforms in the 1980s.

Private ownership of the means of production means that it is possible to start a new business and to compete. This has a positive impact on growth, living standards, and poverty rates. Private business ownership is limited or in name only, if it is possible at all, in socialist countries.

Following advice from Milton Friedman and his “Chicago boys,” Pinochet’s government reduced inflation, slashed tariffs, deregulated the economy, privatized pensions, and introduced private competition into education and healthcare. The economy experienced a deep recession in the early 1980s, but it rebounded and, by the end of the century, it had a solid foundation for sustainable growth. 

“Socialism only works in two places: Heaven, where they don’t need it, and hell, where they already have it.”

— Ronald Reagan

Even though Chile was ruled by a ruthless dictator, free-market reforms generated an entrepreneurial boom. The result for ordinary citizens was unprecedented upward mobility. Capitalism, even under a military junta, expanded middle-class prosperity.

Contrast Pinochet’s regime with the damage done by socialism in Venezuela. Chávez took power in 1999 in one of the richest countries in Latin America, blessed with vast oil reserves and a relatively educated population. He inherited a middle-income economy with immense potential. Then Chávez found a solution to a nonexistent problem.

The state expropriated and nationalized thousands of private businesses, including oil, electricity, telecommunications, and agriculture. Chávez demonized private enterprise and implemented sweeping price controls that destroyed incentives to produce. The government managed to purchase political support while oil prices were high. Like many other resource-rich countries with interventionist policies, Venezuela failed to diversify its economy.

When oil prices inevitably collapsed, so did Venezuela. Hyperinflation destroyed prosperity. The economy shrank by around 75 percent between 2014 and 2021. More than seven million people fled the country, most of them college-educated, creating the largest refugee crisis in the Western Hemisphere.

Private property also creates a path to restoring other kinds of liberties. When individuals, not the government, own and control the land and the capital, it is possible to restore political freedoms. Recent research suggests that capitalism has historically been the ordinary citizens’ shield against tyrannical abuses. The study “You Have Nothing to Lose but Your Chains?” published in Public Choice shows that socialism inexorably leads to authoritarian rule and the violation of human rights. Capitalism provides the kind of opportunities that socialist regimes inevitably extinguish — including the opportunity to regain the consent of the governed after a dictatorship.

“Economic freedom is an essential requisite for political freedom.”

— Milton Friedman

Pinochet stepped aside in 1988 after a voter referendum destroyed his claims to legitimacy. Chileans won back their democracy. Chávez and Maduro ignored attempts to unseat them and entrenched their rule by eliminating checks and balances and weaponizing the state apparatus against their opponents. Controlling the productive assets and eliminating alternatives gives a dictator even more power and resources to control the population. Venezuelans now face a failed state with no prosperity and no freedom on the horizon.

The lesson here is not that dictatorship is good. Pinochet is no hero and committed countless acts of brutality. But if you find yourself under the control of a tyrant, you’d better hope he leans toward free enterprise. Capitalism is the greatest poverty-fighting tool ever tried because it’s based on incentives, not ideology. Unlike socialism, private property also props open an escape hatch, offering the hope that the people can someday regain their dignity and autonomy.

With gasoline prices back above $4 a gallon, many people are once again asking: Are we running out of oil? At around $90 a barrel, crude oil has climbed 50 percent from its January low of $60. That sounds dramatic, but history provides a useful perspective. The highest annual average oil price was $111.67 in 2012. The real outlier was the 1974 OPEC oil embargo, when crude prices exploded 252 percent in a single year, from $3.29 to $11.58 a barrel. A comparable shock today would send oil to roughly $295 a barrel. Could that happen? Yes. Is it likely? No. The future is shaped not by worst-case scenarios, but by probabilities, incentives, and human ingenuity.

But dollar prices are only half the equation. The real question is not “What does oil cost?” but “How much of my time does a barrel require?” To answer this question we must take a look at hourly wages. For example, blue-collar compensation (wages and benefits) has increased 326 percent since 1980.

Once we divide the money price by hourly compensation, we obtain the time price—the number of hours required to earn one barrel of oil.

The true price of oil is measured in time, not dollars. In 1900 oil was only $1.19 a barrel, but blue-collar workers were only earning 14 cents an hour, putting the time price at 8.5 hours. In 1900 oil cost less in dollars, but much more in hours. The time price eventually fell to just 0.46 hours in 1970. Then OPEC showed up and pushed the price to over four hours by 1980. The price fell back to 0.7 hours in 1998 and then back up to 4.14 hours in 2011. Today the time price is barely over two hours, nearly half the 2011 peak.

Even more revealing than today’s price is the futures market. Today’s price tells us where oil is. Futures prices tell us where the market thinks it is going. Unlike television pundits, futures traders back their forecasts with their own money.

The market is signaling that oil prices are likely to decline over time. If you think they’re wrong, the market invites you to prove it and profit from your insight. Why does the market expect lower prices? Because history shows that high prices create powerful incentives to discover new supplies, substitutes, and innovations.

Political shocks, wars, sanctions, and OPEC decisions can temporarily disrupt oil supplies, but knowledge keeps expanding them. Horizontal drilling, hydraulic fracturing, and other innovations have unlocked vast new reserves once thought unreachable. The story of oil is not one of depletion, but of discovery.

Human ingenuity creates abundance in two ways. First, it discovers more oil. Second, it helps us accomplish more with every gallon we consume. In 1980, America’s best-selling car was the Oldsmobile Cutlass, which averaged about 20 miles per gallon: 17 in the city and 23 on the highway. By 2025, the Honda CR-V had become the most popular two-wheel-drive vehicle. Its gasoline model delivers about 31 miles per gallon, while the hybrid reaches roughly 40 miles per gallon. That represents an improvement of 55 to 100 percent over 45 years.

The hybrid performs especially well in city driving because it relies more heavily on its electric motor, captures energy through regenerative braking, and shuts off the gasoline engine while stopped.

In 1980, a blue-collar worker had to work over four hours to buy a barrel of oil, and the typical car traveled about 20 miles per gallon. Today, that same barrel costs just 2.15 hours of work, while modern hybrids travel about 40 miles per gallon. Put those gains together, and each hour of work now buys 3.72 times more transportation than it did in 1980.

Better engines are only part of the story. Cars themselves have become more affordable as well. The surprise isn’t that today’s cars cost more dollars. It’s that a blue-collar worker today needs 41 fewer hours to earn a new Honda CR-V than a worker in 1980 needed to earn a new Oldsmobile Cutlass. According to J.D. Power, the Cutlass sold for $6,735 in 1980. With the BLS reporting blue-collar workers earning $6.82 an hour, its time price was 988 hours. Today, a Honda CR-V starts at about $31,500. At current blue-collar earnings of $33.28 an hour, its time price is 947 hours. Despite being vastly safer, more reliable, more fuel-efficient, and packed with technologies unimaginable in 1980, the modern CR-V costs 4 percent less time to earn than the Cutlass did.

America has helped energize the world by giving its citizens the freedom and property rights to discover new knowledge. That freedom has unlocked vast new supplies of oil, not because the Earth created more petroleum, but because human ingenuity learned how to find and extract what was once beyond reach. The relationship is a virtuous circle. More knowledge gives us access to more energy, and more energy empowers us to create even more knowledge. Every new oil well is also a new lesson in geology, engineering, materials science, and entrepreneurship. The ultimate resource is not oil, but human freedom. Free people create new knowledge, and new knowledge transforms finite physical atoms into ever greater resource abundance.