Skip to main content
Back to Steel City Stories
When the Steel City Lost the Steel — Steel City Stories
HistoryChapter 31 of 4119 min read

When the Steel City Lost the Steel

Steel City Stories

Preparing the first passage.

Steel City Stories · Chapter 31

There is no morning in the records when Pittsburgh woke up and the steel industry was gone. That sentence is the correction this chapter is built to keep. Furnaces idled, mills went on short time, companies reopened contracts to cut wages, famous works rolled a last order and then spent weeks shipping what was left, and a regional workforce that had been counted in six figures became a workforce counted in the tens of thousands. The phrase people still use, that the Steel City lost the steel, is fair as a description of proportion and power. It is false as a date.

The figure this chapter will use, because a university research report printed it with a method and a region attached, comes from the University of Pittsburgh’s Center for Social and Urban Research. In a July 1996 study, Steven Rushen and Joe Giannamore wrote that the region’s steel industry employed about 25,000 workers in 1993, down from 105,000 in 1970. Most of that reduction, they wrote, took place during the 1980s. Between 1980 and 1987, employment in the region’s steel industry fell from 84,500 to 27,700. After 1987 the decline slowed and employment roughly stabilized. Those three pairs — 105,000 and about 25,000, then 84,500 and 27,700 — are the employment facts. Everything else in the chapter is an attempt to keep them from being collapsed into a funeral for a single plant.

What the number is, and what it is not

Rushen’s “region” is the Pittsburgh region as that research project defined it, not a casual synonym for the City of Pittsburgh. The same paper says that in 1993 steel-mill employment still represented 85.9 percent of the region’s primary-metals employment and 19.0 percent of its manufacturing employment. Regional steel mills produced $3.99 billion of output that year, 87.5 percent of a $4.57 billion primary-metals output, and the authors estimated steel exports at $3.354 billion, 10.4 percent of the value of the region’s exports. A workforce can be devastated and still be the core of what manufacturing remains. Both statements are in the same report. Using only the first is how a slogan gets written. Using only the second is how a chamber of commerce erases the people who were no longer on the payroll.

A second series, also published by the same university center, counts manufacturing jobs rather than steel jobs, and it should not be swapped in as if it were the same series. Drawing on the Bureau of Economic Analysis, Regional Economic Information System, Table CA25, the center’s table of manufacturing jobs shows the Pittsburgh metropolitan area with 317,142 manufacturing jobs in 1969, 283,255 in 1979, 218,380 in 1982, and 152,061 in 1987. Allegheny County’s manufacturing count in that table moves from 159,662 in 1979 to 84,975 in 1987. By 1995 the metropolitan manufacturing figure is 138,051 and the county figure is 71,212. The drop is enormous. It includes steel and it includes every other factory job the statisticians classified as manufacturing. It cannot be quoted as “steel employment” without changing the claim. It can be quoted as the wider industrial floor that fell while steel fell through it.

The Federal Reserve Bank of Cleveland, in its industrial-heartland study of Pittsburgh, supplies a third cut that is easy to over-read. Citing a Brookings Wharton paper, the bank notes that between 1977 and 1987 the United States shed about 500,000 auto-industry jobs and 350,000 steel-industry jobs. For this region it is more specific and more modest: during 1977 to 1982, the Pittsburgh metropolitan area’s employment declines, primarily related to the steel industry, amounted to about 23,000 jobs. Inside the first five years of that longer national window, the bank writes, Allegheny County lost nearly 16,000 steel-related jobs, Washington County nearly 2,000, and Beaver County nearly 5,000. “Nearly” is the bank’s word. This chapter will not sharpen it into a false precision. The value of the passage is geographic. The loss was not a Downtown story and not a single-valley story. It hit the central county and the river counties that had organized life around mill gates.

Put the three counts side by side and the shape appears. Steel employment in the university series falls by more than half from 1980 to 1987, from 84,500 to 27,700, on the way down from 105,000 in 1970 toward about 25,000 in 1993. Manufacturing employment in the government-derived table falls by more than 130,000 jobs in the metropolitan area between 1979 and 1987. The Cleveland Fed’s steel-related metropolitan loss for the sharper early window is about 23,000. None of these is “the day the industry died.” Together they are a decade in which the industry stopped being the region’s mass employer and did not stop existing.

The recession was real, and it was not the whole cause

The Cleveland Fed is useful because it refuses the comforting comparison with a bad national year. In the November 1973 to March 1975 recession, national employment declined 1.6 percent and Pittsburgh metropolitan employment fell 3.4 percent. In the twin recessions of the early 1980s, national employment declined 2.2 percent and Pittsburgh metropolitan employment fell 8.5 percent. Across the 1970s and 1980s, the bank writes, national employment expanded by more than 50 percent while the Pittsburgh metropolitan area’s employment expanded by only 4 percent. After the 1981–82 recession, job growth in the Pittsburgh area during the expansion through July 1990 was about 10 percent, against nearly 24 percent nationally. For the whole recession-and-recovery span the bank describes, cumulative employment gains here were about 2 percent while the nation grew about 20 percent. During the 1980s the Pittsburgh area posted a slight employment decline while a group of other industrial-heartland metropolitan areas grew about 10 percent.

Those percentages are not steel headcounts. They are the reason a plant-closing anniversary cannot carry the chapter alone. A region can lose a signature industry inside a national recession and still be experiencing something structural, which is what it looks like when the country gets the jobs back and the region mostly does not. The bank’s authors treat that gap as the signature of the steel shock, not as a mystery of local character. Workers in the Mon Valley were not uniquely unable to adapt. They were concentrated in an industry that shed capacity, shifted production, and stopped hiring at the scale that had built their towns.

The 1996 university report adds the industrial mechanism without romance. Domestic producers, it says, faced foreign mills, competition between integrated works and minimills, substitute materials such as plastics and aluminum, and a labor market in which other industries bid for educated workers. National producers raised productivity, cut capacity, and modernized. Open-hearth production in the United States ceased in 1992. By the middle 1990s about 90 percent of American steel was continuously cast. The region’s share of national capacity kept falling in the early 1990s even after the worst employment break: 8.10 million short tons in 1991, or 6.29 percent of national capacity, and 6.80 million short tons in 1994, or 5.50 percent. The report counted one integrated carbon-steel producer, accounting for 55 percent of regional capacity in tons, and six specialty-steel producers accounting for the rest. It counted no minimills in the region. Specialty steel was not a consolation prize invented by boosters after the fact. It was, by the university’s count, a large share of what capacity remained, and a different kind of employment from the mass integrated works that had organized homogeneous towns.

Homestead, Duquesne, and McKeesport were a sequence

If any single gate deserves the myth of a last day, it is the Homestead Works, and the published record will not give the myth a clean ending. The New York Times, in a story published July 27, 1986, described the last 23 workers at a plant where manufacturing had begun in the early 1880s. The works had rolled its last steel in late May. USX, formerly United States Steel, kept some production workers on to ship remaining steel, strip equipment for sale, and do the other chores of closing. The Times said those men and a few supervisors were the only ones left, and that their job was finished that day. It also gave the scale that makes Homestead more than a local obituary: 20,000 workers in the Second World War, and several thousand as recently as the late 1970s. A Battle of Homestead Foundation remembrance dates the final departure, a group of about two dozen including Mike Stout, as July 25, 1986, out the Amity Street gate. Two published dates, two days apart, describe the same ending crew. Neither date is the regional collapse. Homestead’s last steel was May. The shutdown chores ran into late July. Duquesne and McKeesport were on other calendars.

A Library of Congress historic-engineering survey of U.S. Steel’s National Works at McKeesport places that plant in the same Mon Valley system — Homestead, Edgar Thomson, Duquesne, Irvin, National, and Clairton — and says that until the collapse in 1982 those six plants comprised the company’s Mon Valley works. National had made pipe since the nineteenth century and had been, at its height, described by the survey as the world’s largest pipe producer. The survey says the decision to close National came from USX in February 1987, after a 184-day strike, the longest on record, and that August 28, 1987, was the last day of operations, with 22 maintenance workers left to punch out. A local chronology at Tube City Online dates the final day one day later, August 29, and counts 21 people. The one-day, one-worker difference is small enough to disclose and not large enough to change the meaning. Pipe production, in that local chronology, had already ended in October 1986 when the strike began, and steelmaking at National had ended in 1981 when the blast furnaces were permanently idled. A last punch-out in 1987 was the end of a remainder, not the start of the loss.

Duquesne’s “Dorothy Six” belongs in the same sequence and not as a mascot for the whole valley. The Tube City chronology, which is a compiled local timeline rather than a government statistical series, says that in May 1984 employees at the furnace received production awards for beating U.S. Steel quotas and were laid off the next day. In June 1984, with about 1,000 workers left making rolled steel and shapes, the company announced that the remaining mills would close and production would shift to Lorain, Ohio. In December 1984 steelworkers rallied to save the furnace and winterize it in hope of a buyer. In January 1985 chairman David Roderick told reporters he found it doubtful Dorothy Six would reopen. The engineering survey treats Duquesne, with Homestead and National, as shut down in the years after 1982. The award-and-layoff story, if the timeline holds, is the cruelest small fact in the sequence: a crew could set a company record inside a works the company had already decided not to keep. It is still one works. Edgar Thomson, Irvin, and Clairton are named in that same survey as parts of the Mon Valley system that existed until the 1982 collapse. The 1996 university report’s surviving integrated producer is the reason this chapter will not write “the last mill.” Survival at a smaller scale is part of the verified economic record.

What a payroll had been holding up

The employment figures are large because the wage was doing more than paying a household. Mill towns and city neighborhoods used steel pay to support stores, churches, school districts, municipal budgets, mortgages, and the ordinary expectation that a person without a college degree could raise a family within a bus ride of the gate. When the university report says primary-metals employment in the region fell by 1,170 jobs, or 3.8 percent, just between 1991 and 1994 — with blast-furnace and basic-steel employment at 25,052 in 1991, 23,866 in 1993, and 24,108 in 1994 — it is describing the late, slower movement. The social damage people remember was inflicted earlier, when the drop from 84,500 to 27,700 removed the margin that local institutions had treated as permanent. A church does not close on the day a furnace banks. It closes, if it closes, after years of collection plates that no longer match the building. A borough does not lose its tax base in a single ordinance. It loses it as assessments follow unemployment out of town.

The Cleveland Fed’s county split is the right scale for that claim. Nearly 16,000 steel-related jobs in Allegheny County, nearly 5,000 in Beaver County, nearly 2,000 in Washington County: those are different municipal maps, different school boards, different main streets. Beaver’s loss, the bank notes, was about 6 percent of that county’s overall employment in the period it measured, a staggering share next to Allegheny’s nearly 3 percent and Washington’s about 2 percent. Percentage of a county is not the same as suffering per family, but it explains why some valley towns experienced the 1980s as a civic emergency while Pittsburgh’s later publicity learned to talk about hospitals, universities, and finance. The region did not have one recovery. It had a metropolitan statistic that mixed a growing service economy with river towns whose largest private employer had left the property behind.

Workers were not a residue of nostalgia. The industry the reports describe was technically modernizing at the same time it was dismissing people. Continuous casting, the end of open hearths, higher national productivity, and the region’s later narrowing of a labor-productivity gap with the national primary-metals average are all in the 1996 study. By 1994, the authors write, regional primary-metals labor productivity had moved from $17,832 per worker below the 1989 national average to only $2,229 below it. Relative factor productivity moved from 3.7 percent below the national average in 1989 to 5.5 percent above it in 1994. Those are employer-side measures. They help explain how fewer people could make a still-serious tonnage. They do not pay the mortgage of the person who was no longer one of the fewer. A honest chapter has to say both: the work got more productive, and the community that had supplied the work was not the community that captured the productivity.

The name outlived the headcount

Pittsburgh kept the name Steel City because names are cheaper than payrolls and because the history was real. The 1996 report is the antidote to both boosterism and erasure. Steel-mill employment was still a fifth of regional manufacturing employment in 1993. Specialty producers headquartered here accounted, the authors wrote, for about a third of national specialty capacity, with six of the seven largest specialty companies keeping headquarters or some production in the region: Allegheny Ludlum, Armco, J&L Specialty, Latrobe, Washington Steel, and Teledyne Allvac. An industry that can still be listed that way has not vanished. An industry that has fallen from 105,000 workers to about 25,000 has lost the social contract that made the nickname true for ordinary households.

The plants that closed did not leave empty moral ground. Homestead’s works became, in the foundation’s later telling, a shopping complex on the site of a plant that had once produced a huge share of American steel. The engineering survey records that even before National closed, McKeesport’s mayor pressed the company to demolish an idled blast-furnace plant and convert acreage to an industrial park, and that in March 1987 USX conveyed the site to Allegheny County’s Regional Industrial Development Corporation. In December 1987 a successor firm leased an electric-resistance-weld mill and restarted a piece of it as the Camp Hill works. Restart is not restoration. A pipe mill with a new tenant is not 9,000 jobs, the peak employment Tube City associates with National. It is evidence that “closed” and “gone” are different words, and that public and private actors spent the late 1980s arguing over reuse while the employment statistic had already reset.

The political question that follows is the one the employment tables cannot answer by themselves. What does a region owe the towns that built the capacity it later wrote off? The record assembled here shows the write-off was a series of corporate decisions inside a national restructuring: shift production to Lorain, decline to reopen Dorothy Six, end Homestead after the last orders, convey McKeesport for an industrial park, keep a smaller integrated and specialty base. It also shows public agencies receiving sites and measuring manufacturing loss in official series. It does not show a single legislative moment in which the Commonwealth decided the Mon Valley should absorb the shock unaided, and this chapter will not invent one. The obligation is visible in the arithmetic. Eighty-four thousand five hundred to twenty-seven thousand seven hundred in seven years is not a business cycle. It is a transfer of risk from an industry to the households and boroughs that had organized themselves around it.

How to read a closing day without lying

A last crew at Homestead is worth remembering because people were there and because the Times put the number 23, or a few more with supervisors, on the page. It is a bad synecdoche. If the industry died on the day that crew drove out the gate, then the 27,700 steel workers still counted in 1987 are an embarrassment to the story, and the 25,000 in 1993 are a contradiction. If the industry never really declined, the drop from 105,000 is an embarrassment of the other kind. The accurate sentence is chronological. Employment had been falling for years before Homestead rolled its last steel in May 1986. Duquesne’s remaining mills had been slated for shutdown two years earlier. National’s blast furnaces had been idle since 1981, and its last maintenance crew punched out in 1987. Manufacturing employment in the metropolitan table was already in steep decline by 1982, when it stood at 218,380 against 283,255 in 1979. The famous gates closed inside that decline. They did not open it, and they did not close it.

There is a further discipline about heroism and blame. John Hoerr’s book on the industry’s decline is cited by the engineering survey as a leading narrative, and this chapter does not pretend to replace that book or to adjudicate the strike of 1986, the wage concessions of 1982, or the oil acquisitions that accompanied U.S. Steel’s renaming as USX. Those subjects are part of the history. They are also places where memoir, union argument, and corporate statement disagree about motive. Motive can be reported when a document states it, as when a company announces a shift of production to another plant. Motive should not be novelized. The verified core is smaller and harder: a regional steel workforce cut by roughly two-thirds in the 1980s on the university’s 1980–1987 series, inside a longer fall from 1970 to 1993, while steelmaking continued at a reduced scale and other manufacturing fell with it.

A public memory that prefers one photograph

Public memory likes a gate, a last shift, and a shopping center afterward, because a gate can be visited. The statistical record likes a table, because a table can show that the gate was one line in a regional subtraction. Both are necessary, and they answer different questions. The Times photograph of a nearly empty Homestead Works tells a reader what abandonment looked like at one address in the summer of 1986: a plant that had employed 20,000 people in wartime reduced to a closing crew. The university table tells a reader that two years before that summer the region’s steel employment was already on the way from 84,500 toward 27,700, and that seven years after Homestead’s last rolled steel the count was about 25,000 and no longer in free fall. A city that teaches only the photograph will think the industry died on a July day. A city that teaches only the later stabilization will think the 1980s were a normal modernization.

The manufacturing series is the check on both mistakes. Metropolitan manufacturing employment did not stabilize when steel employment did. It kept falling after 1987, from 152,061 that year to 138,051 in 1995, and Allegheny County manufacturing fell from 84,975 to 71,212 over the same years. Steel’s slower decline after 1987, which the university authors describe as basic stabilization, sat inside a manufacturing sector that was still shrinking. That is why “the steel collapse” and “deindustrialization” are related and not identical. Steel was the shock that made the region’s loss famous. Manufacturing was the broader count that shows the famous shock did not exhaust the loss. Service industries, hospitals, and universities grew into the story Pittsburgh later told about itself. They do not appear as a credit against the 1980–1987 steel drop, because a new job in Oakland is not, in these tables, a restored job in McKeesport.

The chapter also refuses a geography that stops at the city line. Pittsburgh’s municipal government did not employ the Homestead crew, the Duquesne furnace crew, or the National pipeworkers. Those plants stood in neighboring municipalities whose tax bases rose and fell with the works. The Cleveland Fed’s decision to print Beaver and Washington counties beside Allegheny is the corrective. A Steel City narrative that stays inside Pittsburgh’s corporate boundaries will undercount the loss and then act surprised that mill towns still measure time from the 1980s. The nickname was regional. The layoff was regional. The recovery narrative that followed was selective about which parts of the region got to be the evidence.

Capacity, tonnage, and the trick of a smaller success

By the early 1990s the university authors could describe a steel industry that was more productive and still shrinking in regional weight. Raw steelmaking capacity in the region fell from 8.10 million short tons in 1991 to 6.80 million in 1994, a 16.05 percent decline against a 3.96 percent decline in total United States capacity over the same years. The Pittsburgh and Youngstown districts had been combined in the statistical geography in 1989. That combined district produced 15.74 million tons of raw steel in 1989, 15.92 percent of national output, and 14.16 million tons in 1993, 14.48 percent. A drop of 1.58 million tons in the combined district is not the 1980s employment catastrophe repeated. It is the evidence that “stabilization” still meant a smaller share of the country. Specialty consumption nationally was rising faster than specialty production, the report notes, pulling imports up, while the Pittsburgh region held about a third of national specialty capacity. A region can be a specialty-steel capital and a shattered mass-employment region at the same time. The second fact is what the mill towns felt. The first is what later economic-development copy preferred to lead with.

Primary-metals employment from 1991 to 1994 fell by 1,170 jobs, 3.8 percent, close to the national 3.3 percent decline. Fabricated metals in the region fell by 192 jobs, 1.6 percent, also close to the national pace. The authors read those early-1990s movements as mostly national trends, unlike the regionally concentrated break of the previous decade. Relative profitability of the region’s primary-metals manufacturers had improved: 1.55 percent above the national average in 1989 and over 4 percent by 1991. Profitability recovered faster than the towns did, which is another way of saying that a firm’s income statement and a borough’s school budget are not the same document. The report is frank about the local business climate it considered unfavorable to manufacturing: high state business and local property taxes, slow regional growth, and poor infrastructure. Those are the authors’ judgments about conditions, not a finding that taxes caused the 1980–1987 employment drop. Using them that way would launder a decade of shutdowns into a single policy preference. They belong in the chapter as part of the mid-1990s argument about what the remaining industry needed, after the headcount had already been cut.

What remains disputed

The exact boundary of “the region” differs across studies, so 84,500 and the Cleveland Fed’s about 23,000 steel-related metropolitan jobs are not interchangeable. One is a steel-industry employment level at two dates. The other is a five-year loss described as primarily steel-related. Manufacturing totals from the Bureau of Economic Analysis are a third object. Readers who add them will invent a number nobody published. The Homestead final day is published as both July 25 and the day observed in a July 27 Times story. National’s last day is published as both August 28, with 22 workers, and August 29, with 21. Dorothy Six’s award-and-layoff sequence rests on a compiled local chronology and should be treated as that. Claims about how much of American steel Homestead once made vary by telling; this chapter does not adopt an unsourced fraction. What happened to household income, disability, and out-migration in each mill town requires local series this chapter does not pretend to have settled. The moral argument over concessions, trade, and corporate diversification remains a live historical fight. The employment break does not.

Evidence Ledger

  • Steven Rushen and Joe Giannamore, “Competitiveness of the Steel Industry in the Pittsburgh Region,” University of Pittsburgh Center for Social and Urban Research, July 1996, for the 1970, 1980, 1987, and 1993 steel-employment figures and the surviving capacity structure: Rushen steel report (UCSUR).
  • University Center for Social and Urban Research manufacturing table, 1969–1995, sourced to Bureau of Economic Analysis Regional Economic Information System Table CA25: UCSUR Table 36.
  • Federal Reserve Bank of Cleveland, industrial-heartland report on Pittsburgh, for recession comparisons and the steel-related county and metropolitan job losses: clevelandfed.org.
  • New York Times, July 27, 1986, on the Homestead Works closing crew and the late-May end of rolling: nytimes.com.
  • Historic American Engineering Record, Library of Congress, National Works survey, for the 1987 McKeesport closing and the Mon Valley plant system: loc.gov HAER.

This chapter is an original public expansion of PittsburghEverything’s Steel City Stories. It does not replace a Google Drive FULL MASTER.

Archival photographs relevant to this chapter. Sourced from Wikimedia Commons with attribution — not modern stock art.

Railroad tracks to nowhere, Carrie Furnaces in Rankin PA (8786570551)

Railroad tracks to nowhere, Carrie Furnaces in Rankin PA (8786570551)

Wikimedia Commons · Railroad tracks to nowhere, Carrie Furnaces in Rankin PA (8786570551).jpg · Roy Luck · CC BY 2.0

The blast furnaces and rolling mills of the Homestead Steel Works.Alternate Title: 'Perfec' stereograph. 13109.

The blast furnaces and rolling mills of the Homestead Steel Works.Alternate Title: 'Perfec' stereograph. 13109.

Wikimedia Commons · The blast furnaces and rolling mills of the Homestead Steel Works, by H.C. White Co..jpg · H.C. White Co. -- Photographer · Public domain

Carrie Furnace Hot Metal Bridge from Whitaker

Carrie Furnace Hot Metal Bridge from Whitaker

Wikimedia Commons · Carrie Furnace Hot Metal Bridge from Whitaker.jpg · Hhm8 · CC BY-SA 4.0

Related Stories