Levi Strauss & Co.: a jeans maker that closed its own factories and managed suppliers instead

The situation

On 25 September 2003 Levi Strauss & Co. announced that it would close the last factories it owned in North America. Two sewing and finishing operations in San Antonio, Texas, would shut by the end of the year, with about 800 jobs lost. Three Canadian sites would follow in March 2004: sewing plants in Edmonton, Alberta, and Stoney Creek, Ontario, and a finishing centre in Brantford, Ontario, employing about 1,180 people between them. All of their output would move to contract factories in the company’s global sourcing network.

The chief executive, Phil Marineau, gave the reason in one line. Levi’s, he said, was in “a highly competitive industry where few apparel brands own and operate manufacturing facilities in North America.” To stay competitive the company needed to put its money into product design and development, sales and marketing, and its relationships with retailers. “Regrettably, these closures will affect workers who have done a tremendous job for the company over the years,” he added.

The announcement was the last step in a retreat that had been going on for six years, and it came at a bad moment for the business. Net sales had peaked at US$7.1 billion in 1996. By 2003 they were US$4.09 billion, and the company would report a net loss of US$349 million for that year. Total debt stood at US$2.3 billion. The global workforce, more than 37,000 in 1996, had fallen to about 12,300 by November 2003, of whom about 4,360 were in the United States. In the early 1980s Levi’s had run 63 manufacturing plants in the US alone. After March 2004 it would run none in North America.

The case asks whether a brand built on an American work garment could hand all of its production to other people’s factories and still control its costs, its quality and its reputation for treating workers well. It also asks what the company owed to the people whose jobs went.

Background

From dry goods to a family-owned brand

Levi Strauss arrived in San Francisco in 1853 to open a West Coast branch of his brothers’ dry goods business in New York. In 1873 he and a Reno tailor, Jacob Davis, were granted a US patent for work trousers reinforced with copper rivets, the product that became the blue jean. The company stayed in the hands of Strauss’s relatives after his death in 1902, and the Haas branch of the family ran it for most of the twentieth century.

Levi’s sold shares to the public in 1971. The listing did not last. In the mid-1980s the Haas family, led by Robert (Bob) Haas, a fifth-generation member of the founder’s family, bought out public shareholders in a buyout financed largely with debt. Sources differ on the details: a Motley Fool profile from 1999 gives the year as 1985 and the price as US$1.6 billion, while Family Business Magazine dates it to 1984 at US$1.7 billion. In 1996 a second transaction, which the company’s filings call a recapitalisation, bought out employee and other shareholders and placed control in a voting trust held by family members. The company’s 2003 annual report says this deal was funded partly by cash and partly by about US$3.3 billion of bank borrowing. The debt from 1996 shaped every decision that followed, because interest had to be paid whether sales rose or fell.

How Levi’s made money

Levi’s earned most of its revenue selling jeans and casual trousers at wholesale to department stores, chains and specialty shops. The Levi’s brand was its largest business. Dockers, a line of khaki trousers launched in 1986, became the leading men’s trouser brand in the US by unit sales, according to the company’s 2003 filing. In July 2003 the company added a cheaper line, Levi Strauss Signature, sold through nearly 3,000 Wal-Mart stores in the US and also through Wal-Mart in Canada, Germany and the UK.

The brand carried a price premium, and the company paid to maintain it. In an October 2002 presentation to investors, Levi’s said it spent about 7.7% of revenue on advertising and promotion, against 3.8% for apparel peers. Marketing on that scale needed a gross margin to match, and that margin was being squeezed by two things: cheaper rival products and the cost of running its own factories in high-wage countries.

The slide after 1996

The 1990s were a decade in which Levi’s lost its hold on the US jeans market. Figures published by the Motley Fool in 1999 put the company’s share of US jeans sales by retail value at 31.0% in 1990 and 16.9% in 1998. Over the same period VF Corporation, owner of Lee and Wrangler, rose from 17.9% to 25.3%, and retailers’ private labels, such as J.C. Penney’s Arizona brand, rose from 3.2% to 20.2%. In men’s jeans, Levi’s share was said to have fallen from 48.2% to 25%.

Sales followed share down. After the 1996 peak, revenue fell to US$5.96 billion in 1998 and to about US$5.1 billion in 1999. The company began closing plants in 1997. NPR reported in November 1997 that Levi’s would shut 11 plants and lay off 6,400 manufacturing workers, more than a third of its North American factory workforce, citing “excess capacity”. In February 1999 it said it would close 11 of its 22 remaining North American plants and cut about 5,900 jobs, about 30% of its US and Canadian workforce, according to the Christian Science Monitor. By the end of 2002 the company’s annual report counted 37 owned production and finishing facilities closed worldwide since 1997 and about 23,600 jobs removed.

Leadership changed during the slide. Bob Haas was chairman and chief executive in 1999; by 2003 the chief executive was Phil Marineau. In December 2003, as losses mounted, Levi’s brought in managing directors from the restructuring firm Alvarez & Marsal as a senior adviser to the chief executive and as chief financial officer.

The industry

A large market with low growth

Jeans were a big but slow-growing category. Using NPD data, Levi’s told investors in 2002 that Americans bought 518 million pairs of jeans in 1997 and 529 million in 2001, a rise of 2% in four years. The channels through which those jeans were sold were changing faster than the total. Sales through mass merchants such as Wal-Mart and Target rose from 151 million to 165 million pairs. Sales through chain stores, a traditional Levi’s outlet, fell from 120 million to 101 million. Levi’s estimated that entering the mass channel would raise the share of the US market it could reach from 53% to 84%, which was the case for Signature.

Rivals on two sides

Competition came from below and above. Below, VF Corporation’s Lee, Wrangler and Rustler brands and retailers’ own labels sold jeans for less. Levi’s 2003 annual report named VF as its chief competitor across several channels and listed Arizona, Sears’ Canyon River Blues, Wal-Mart’s Faded Glory and Target’s Mossimo and Cherokee among the private and exclusive labels it faced. Above, designer and fashion brands charged more and set the styles young buyers wanted. Gap, which designed its own jeans and sold them only in its own shops, grew its share of the US market from 2.7% to 4.9% in the 1990s on the Motley Fool figures.

Forces at work

The economics of sewing jeans favoured the buyer of production, not the owner of it. Cutting, sewing and finishing denim used widely available skills and machines. Thousands of contract factories in Mexico, Central America and Asia could do the work, so no single supplier had much bargaining power over a large brand. Retail buyers had more: the growth of mass merchants and private labels gave large retailers the ability to source similar jeans themselves and to press brands on price. Entry barriers into jeans were low at the product level and high only at the level of a national brand, which is where Levi’s still held an advantage.

Labour costs made the gap between owning and buying very wide. US apparel manufacturing employed about 924,000 people in January 1990, according to Bureau of Labor Statistics data. By January 2000 the figure was about 491,000, and by January 2004 about 281,000. When Levi’s closed a San Antonio plant in 1990, the former workers’ leader said she had earned US$9.10 an hour as a sewing machine operator, while the Costa Rican contractor that took over the work paid US$1.08 an hour, according to the Christian Science Monitor.

Trade rules

Trade policy was about to widen that gap further. Since the 1970s, quotas under the Multi-Fibre Arrangement, and from 1995 the WTO’s Agreement on Textiles and Clothing, had limited how much clothing each developing country could ship to the US and Europe. Those quotas were phased out over ten years and ended on 1 January 2005. Levi’s own 2003 filing warned that the end of quotas could increase competition from lower-cost countries such as China. The US Department of Agriculture later reported that imports of cotton trousers from China rose by 1,094% in the first nine months of 2005, before new safeguard limits were imposed. A brand still sewing in Texas and Ontario would face rivals sourcing from the cheapest places in the world, with no quantity limits.

The decision

By 2003 the choice facing Marineau and the board was narrower than it had been in 1997. Levi’s had already moved most of its production out. In 1997 about 40% of its products came from contract manufacturers. By 2002, owned plants made only 19% of its goods, and the company had told investors it expected contractors to supply 90% in 2003. The question was whether to keep a last core of owned production in North America or to finish the job.

Keep some owned plants

There were arguments for staying. Owned plants gave the company direct control of quality and of how workers were treated. They provided a link to the brand’s American story, which still mattered in marketing. The San Antonio site had at its peak made more than 4 million pairs of jeans a year, and the company had a long relationship with its workforce there. A company spokesman told the Christian Science Monitor in 1992 that American workers often produced better garments more quickly, and that “You can’t only look at labor costs.”

Finish the move to contract production

The arguments for closing were mostly financial. The remaining North American plants made a small share of the company’s goods; the San Antonio operation produced less than 5% of the products Levi’s sold in the US, the company said in 2003. They carried fixed costs that did not fall when sales did, at a time when revenue had dropped for seven years in a row and lenders were pressing for results. Levi’s estimated that plant closures since 1997 had produced cumulative annual savings of US$390 million by 2002, with US$530 million forecast for 2003, money it said it would put back into product and customer economics. Contract production also offered flexibility: the company could move orders between countries as costs and trade rules changed, instead of being tied to a few sites.

The company was also changing how it bought. Under cut-make-trim arrangements, Levi’s supplied fabric and the contractor sewed it. Under package production, the contractor bought the materials and delivered finished goods. In 2003, 53% of Levi’s products came through cut-make-trim and 36% through package deals, and the company said it expected to source essentially all US products on a package basis by the third quarter of 2004. That meant passing more of the supply chain, and more of its risks, to suppliers.

The standards question

The harder question was whether Levi’s could keep its standards in factories it did not own. Here it had a head start. In 1991 it had adopted what it and later commentators describe as the first global code of conduct for suppliers drawn up by a multinational, known as the Terms of Engagement. In 1993 it stopped sourcing from China over human rights concerns, returning in 1998. A Stanford case written by David Baron in 1994 describes how the company built audit forms and trained staff to check contractors against the code. When auditors found children under the minimum age in two Bangladeshi factories in the early 1990s, Levi’s agreed with the contractors that the children would keep being paid while they attended school, with Levi’s paying for tuition and books, and would be offered jobs when they reached working age. Critics and supporters both held the company up as an example of what a code could do. The 2003 decision rested on the belief that this system could cover the whole of its production.

What they did

Closing the plants

Sewing in San Antonio stopped around Thanksgiving 2003, and laundering and finishing ended the following month. The Associated Press reported in January 2004 that the two plants had operated for 26 years and that workers had earned on average US$10 to US$12 an hour. A company spokesman, Jeff Beckman, called the decision delayed but unavoidable: “we have to be competitive to survive as a company.” He said headquarters, design, sales staff and several distribution centres would stay in the US. The Canadian plants closed in March 2004, ending more than 130 years of the company making its own jeans in North America.

Severance for the San Antonio workers comprised two weeks’ pay for each full year of service, an enhanced pension, a period of subsidised medical cover and a payment of US$2,450 each, according to a company spokeswoman. The Levi Strauss Foundation set up a Community Transition Fund to make grants in San Antonio and the affected Canadian towns. The company did not disclose the cost of the closures separately; its restructuring charges, net of reversals, were US$115.5 million in 2002 and US$89 million in 2003 across all actions.

The closures did not end the cost cutting. In 2004 Levi’s shut plants in Spain and Australia, reduced head office staff in the US and Europe, suspended a global software installation and moved its European Dockers office from Amsterdam to Brussels. Restructuring charges that year were US$134 million. The company kept eight owned plants outside North America at the end of 2003, in Europe, Asia and South Africa. Over the following years it closed almost all of them.

Making the supply chain visible

With production in other people’s factories, the company turned to disclosure as a way of keeping control. On 12 October 2005 it published a list of the active owned and contract factories making Levi’s, Dockers and Signature products, along with a 70-page Terms of Engagement guidebook setting out factory-level practices, examples of violations and timeframes for fixing them. David Love, senior vice president for global sourcing, said greater transparency would “provide additional momentum” for improving conditions, and hoped other brands would follow so that companies sharing factories could cooperate on monitoring. Michael Kobori, who ran the code programme, said the list recognised suppliers that met the code and created accountability.

Publishing a full list was unusual in the industry at the time, and Love said he hoped it would become standard practice. A transparency report card from the period, archived by the Maquila Solidarity Network, gave Levi’s full marks for disclosing factory names and addresses and 71 out of 100 overall.

Water and chemicals

The environmental side of the supply chain came later. In 2011 the company’s designers introduced Water<Less finishing techniques, which the company said could cut the water used in finishing a garment by up to 96%. By March 2016 it said the techniques had saved more than 1 billion litres, and it published 21 of them as open-source methods for other brands. It set a target for 80% of its products to be made with Water<Less methods by 2020. In March 2019 it reported more than 3 billion litres saved and more than 67% of products made with the techniques.

Spreading the sources

The sourcing map kept shifting after 2003. The end of quotas in 2005 made China more attractive for a time. Later, as Chinese wages rose and US trade policy hardened, the company moved away again. Its chief financial officer, Harmit Singh, told Fortune in April 2025: “Eight or 10 years ago, we were sourcing into the U.S. 15-16% from China. Today, it’s approximately 1%.” In its annual report for the 2024 financial year, Levi’s said it sourced from independent contractors in about 28 countries, with no more than 30% from any single country, and that the only manufacturing plant it still operated was in South Africa. Its top vendor countries in 2025 were Bangladesh, Cambodia, Egypt, Pakistan, Sri Lanka and Vietnam, according to a report carried by Yahoo Finance in April 2025.

What happened

The short run

The first results after the closures were mixed. Net revenues in 2004 were US$4.07 billion, flat as reported and down 4% at constant exchange rates. But gross margin rose from 38.5% to 43.8%, operating income rose 15% to US$361 million, and the company returned to a net profit of US$30 million after the loss of US$349 million in 2003. “We begin 2005 with a healthier and more competitive business,” Marineau said. Debt net of cash fell by about US$151 million to US$2.02 billion. The company did not separate the effect of the North American closures from its other actions.

Exhibit 1: Levi Strauss & Co. financial results, 1996 to 2004 (US$ millions)

Sources: Associated Press via NBC News (1996); Levi Strauss & Co. Form 10-K for fiscal 2002 (1998) and fiscal 2003 (1999 to 2003); fiscal 2004 results release. The company restated several years in its 2003 filing; for example, net income for 2002 was first reported as US$25 million and restated to US$7 million.

Exhibit 2: How Levi’s made its products

The long run

Sales stayed around US$4 billion for several years. Family Business Magazine noted that revenue was about US$4 billion in 2011, when Chip Bergh became chief executive. Under Bergh the company returned to growth, opened more of its own shops and went back to the stock market. Its shares were listed on the New York Stock Exchange in March 2019 at US$17 each, raising about US$623 million and valuing the company at about US$6.6 billion, according to CNBC. Revenue in 2018 had been US$5.6 billion.

By the 2024 financial year, net revenues were US$6.36 billion and net income US$211 million. The gross margin was 60.0%, far above the 38.5% of 2003, and the shape of the business had changed. Company-operated stores accounted for 31% of revenues and the company’s own websites for 10%. Levi’s ran 1,276 stores of its own in 39 countries. It employed about 18,700 people, more than in 2004, but only about 800 of them worked in manufacturing and procurement, while about 10,800 worked in retail. The company had moved out of making things and into selling them directly.

The spread of sources became an asset in 2025, when the US imposed new tariffs. On 2 April 2025 the US placed duties of 10% to 49% on the countries where Levi’s main vendors were based. Because China supplied only about 1% of its US-bound goods, Singh described the company’s position as manageable and said tariffs would have “a minimal impact to our margin structure” in the following quarter, since most spring and summer goods were already in the country. Later in the year the company said tariffs reduced its fourth-quarter gross margin, offset in part by price rises.

Other views

Workers and their advocates

The closures looked different from the factory floor. Viola Casares, who had helped organise boycotts and pickets after the 1990 San Antonio closure, told the Associated Press in 2004 that she felt “in mourning”. She said the 2004 benefits were far better than what she had received 13 years earlier as a single mother of four, but that it would be hard for displaced workers, especially those over 50, to find jobs paying as well. The group she co-founded, Fuerza Unida, had argued in 1992 that the 1990 closure was not an economic necessity, pointing to Levi’s 1991 profits of US$357 million on nearly US$5 billion of sales. It sued the company alleging discrimination and lost in the state courts. The company denied its claims and said it had given 90 days’ notice where 60 were required, extended medical benefits and paid for retraining.

Analysts and investors

Market analysts were less troubled. Walter Loeb, a New York retail analyst, told the Associated Press in 2004 that the profit from moving production overseas outweighed the symbolic value of American manufacturing, and that investors were not sentimental. He added a warning that cheaper production would not by itself win back customers, because the brand’s exclusivity was “no longer as important to customers.” Closing plants cut costs, but the share lost in the 1990s had gone to cheaper labels below and fashion brands above, and cost cuts alone did not win it back.

The limits of codes

Labour rights groups gave the company credit for its code and its transparency while pointing to gaps. The report card archived by the Maquila Solidarity Network noted that Levi’s code did not meet International Labour Organization standards on working hours, had no living wage requirement, and that the company did not publish audit results. Kobori himself later told Reuters Events that when the company released its code, “our competitors laughed at us”, and that it had moved 80% of its monitoring resources from audits to working with factories on improvements. That shift is an admission, of a kind, that audits alone had not delivered the change the code promised. Economists of the period, such as Richard Rothstein of the Economic Policy Institute, quoted in 1992, saw Levi’s as “no better or worse” than the rest of the industry in moving jobs abroad, and argued that the answer lay in trade policy rather than company codes.

Questions of fact

The public record leaves some things unclear. The company never published the cost of the North American closures separately, nor the cost per pair of jeans in its own plants against contractors. Estimates of its market share in the 1990s come from secondary sources rather than from the company. Its claims about Water<Less savings are its own figures and have not been independently audited in any source this case draws on.

Reading it through a framework

Make or buy, through transaction costs

Transaction cost economics, associated with Oliver Williamson, says a firm should do an activity itself when buying it in the market would be costly or risky. The main test is asset specificity: if a supplier would have to make investments useful only to one customer, or if the product depends on knowledge that is hard to write into a contract, owning the activity makes sense. If the activity uses general skills and many suppliers can do it, buying is cheaper.

Jeans sewing in 2003 failed the test for ownership on almost every count. The machines and skills were general. Thousands of factories could do the work, so Levi’s faced little risk of being held to ransom by one supplier. The product could be specified in detail through patterns, wash standards and inspection. The cost of owning was high: US wages several times those of contractors abroad, and fixed costs in a business whose sales had fallen 42% from their peak. The company’s own figures put cumulative annual savings from plant closures at US$530 million by 2003, more than the US$310 million operating income it earned that year.

The framework also explains what Levi’s kept. Design, brand marketing and, later, retail are activities where knowledge is specific to the company and hard to contract out. The growth of owned stores to 31% of revenue by 2024 is the other half of the same logic: Levi’s bought the generic middle of the value chain and took more control of the ends.

The framework points to one cost that buying does not remove. When production moves to suppliers, the firm still carries the reputational risk of what happens in their factories. Transaction cost theory calls the costs of writing, monitoring and enforcing contracts “governance costs”. The Terms of Engagement, the audit staff, the 2005 factory list and the 70-page guidebook were Levi’s governance costs. Whether they were enough is the question critics raised.

The value chain

Michael Porter’s value chain splits a firm into primary activities (inbound logistics, operations, outbound logistics, marketing and sales, service) and support activities. A firm should put its resources into the activities where it can create value buyers will pay for. Levi’s in 2002 was spending about 7.7% of revenue on advertising, twice its peers, while also running plants that added cost without adding anything a shopper could see. Moving operations to suppliers turned a fixed cost into a variable one and freed money for marketing and, later, stores. The gross margin path, from 38.5% in 2003 to 43.8% in 2004 and 60.0% by 2024, reflects both cheaper production and a shift towards selling at full price in the company’s own channels, so it overstates the effect of outsourcing alone.

Since then

Levi’s has kept narrowing its focus. In late 2024 it began reviewing options for Dockers, the trouser brand launched in 1986 that had been a mainstay of its business in the 1990s. On 20 May 2025 it agreed to sell Dockers to Authentic Brands Group for an initial US$311 million, with up to US$80 million more depending on performance. The chief executive, Michelle Gass, said the company had “maximized the value of the business”. The US and Canadian part of the sale closed on 31 July 2025 and the rest was expected to close around the end of February 2026.

For the year to 30 November 2025, Levi’s reported net revenues from continuing operations of US$6.28 billion, up 4%, and net income from continuing operations of US$502 million. The gross margin rose to 61.7%. In the fourth quarter its own stores and websites accounted for 49% of revenues. Gass described the strategy as becoming a “DTC-first, head-to-toe denim lifestyle brand”. The company’s guidance for 2026 assumed US tariffs of 30% on goods from China and 20% on goods from other countries.

The US apparel manufacturing workforce has kept shrinking. BLS data show about 87,000 people employed in January 2024 and about 71,000 in September 2026, against 924,000 in January 1990. The jobs Levi’s closed in San Antonio and Ontario have not come back in any form, and the company has not suggested they will. Its sourcing policy, with a cap of 25% to 30% on any one country, is now presented as its main defence against trade shocks.

What to take from it

• Owning production makes sense when the work is specialised or hard to specify, and Levi’s sewing and finishing had become neither by the late 1990s.
• Outsourcing removes fixed costs but not responsibility, so the money saved has to pay in part for codes, audits and disclosure that keep suppliers to the brand’s standards.
• Cutting production costs did not by itself reverse a loss of market share that came from rivals’ prices and fashion, and Levi’s sales stayed near US$4 billion for years after the closures.
• A sourcing base spread across many countries, with a cap on each, gave Levi’s room to move when US tariffs rose in 2025.
• How a company treats workers when it closes plants is remembered, and the gap between the 1990 and 2004 severance terms shows the company had learned from the earlier protests.

Questions to think about

1. Levi’s said plant closures saved US$530 million a year by 2003, against operating income of US$310 million that year. What other information would you need to judge whether closing the last North American plants was worth the restructuring charges and the reputational cost?
2. San Antonio workers earned US$10 to US$12 an hour in 2003, and a Costa Rican contractor was reported in 1992 to pay about US$1 an hour. Assume, for the sake of the exercise, that a pair of jeans needs 30 minutes of sewing and finishing labour and sells for US$40. What is the labour saving per pair, and how large is it as a share of the retail price? What does your answer leave out?
3. Was the company’s market share loss in the 1990s caused by its cost base, by its products and marketing, or by changes in where Americans shopped? What evidence in the case supports each answer?
4. How far is a brand responsible for working conditions in factories it does not own, and does publishing a factory list change that responsibility?
5. What are the risks of a cap of 25% to 30% on any one sourcing country, as well as the benefits?
6. Should Levi’s have kept a small owned plant in North America for product development and marketing purposes, as some brands later did? Make the case for and against.

Concepts at work

Make-or-buy analysis, transaction cost economics, value chain, supplier code of conduct, triple bottom line. Explore them in Insights.

Further reading

Oxford University Press covers this story as a teaching case in Cases in International Business on Oxford Business Trove. The sources below are the public record this page was written from.

Written from the public sources listed below. Figures are as those sources report them.