First-Mover and Late-Developer Advantages Ch 1

“First-Mover” and “Late-Developer” Advantages: Institutions and Market Design in the New York and Los Angeles Garment Districts, 1900-1960 Abstract New York and Los Angeles are the largest garment districts in the United States and resemble one another in many ways today, but they have had very different industrial histories. New York City was the earliest “fashion” district in the United States and its “first mover” advantages during the 1880s and 1890s gave it a near monopoly of the U.S. women’s wear market by the start of the 20th century. Los Angeles was an insignificant garment district serving a relatively small regional market until the late 1920s and early 1930s when its pioneering sportswear designs contributed to its emergence as a “late developing” district. These historical differences in the timing of the development have endowed the New York and Los Angeles apparel industries with a distinctive set of economic institutions – trade unions, employer associations, contracting relationships, and manufacturing cultures. Differences in unionization have been particularly significant and have shaped the way that markets are organized in each district. The New York City garment district had highly-competitive labor and product markets during the late 19th century that were “unraveling” in a chaotic form of hyper-competition. Unions entered this market in the early 20th century with the goal of stabilizing the labor market and reached a series of collective bargaining agreements with employers that regulated both labor and product markets through during much of the 20th century. The Los Angeles district remained largely free of unions and its markets were governed by intense competition except for a brief period during the late 1930s to the mid 1940s. This natural experiment in alternative market “designs” has had path-dependent effects that have persisted into the current decade even though union strength had substantially eroded in New York City by the latter part of the 20th century. The latedeveloping Los Angeles industry opted for a set of competitive market arrangements that appear to be superior when compared the more-institutional market governance first adopted by New York. The Los Angeles apparel industry continued to grow long after New York’s began to decline from import competition and it has displaced New York as the largest U.S. garment district, with more than three times as many apparel industry employees, a larger share of domestic output of women’s wear, and a lower rate of job loss in recent years. However, New York is performing better than Los Angeles by measures such as productivity and pay. It remains to be seen which market design provides the better odds of survival in the future. “First-Mover” and “Late-Developer” Advantages: Institutions and Market Design in the New York and Los Angeles Garment Districts, 1900-1960 1 New York and Los Angeles are the largest garment districts in the United States and resemble one another in many ways. 2 Garment manufacturing in both districts is dominated by small firms using labor-intensive assembly technologies (principally sewing machines); most manufacturing jobs are low-wage and filled largely by first-generation immigrants; both districts specialize in women’s wear with many overlapping sub-specializations; and much of the local production is performed by contractors that are coordinated by “jobbers” and “manufacturers” who do limited garment assembly, but mainly serve as market intermediaries that receive orders from retailers and in turn commission contractors to assemble garments. New York and Los Angeles also have similar agglomeration economies -- large pools of flexible labor, proximity to buyers and suppliers, and ready-access to creative fashion knowledge – that contributed to the long-term growth of their apparel industries in the formative years of both districts, and both are now experiencing substantial job losses from import competition. These districts, however, have very different industrial histories. New York City was the earliest fashion district in the United States and its “first mover” advantages during the 1880s and 1890s gave it a near monopoly of the U.S. women’s wear market by the start of the 20th century. Los Angeles was an insignificant garment district serving a relatively small regional market until the late 1920s and early 1930s when its pioneering sportswear designs contributed to its emergence as a “late developing” district(see Gerschenkron, 1962). Population shifts to the West and Southwest and long-term consumer trends favoring lighter-weight and more casual clothing gave Los Angeles a growth advantage over the New York garment district in national markets for sportswear, but as a late developer, Los Angeles 1 This paper is part of a comparative study of the economic performance of garment districts in the United States, the UK, France, and Italy. Support for this research came from the National Science Foundation under Grant No. 0328635, the Fiscal Policy Institute, the Garment Industry Development Corporation, the Alfred P. Sloan Foundation, the International Labour Organization, and the Harvard University Center for Textile and Apparel Research. The paper has benefited from comments by Daniela Bigarelli, Bruno Courault, Paolo Crestanello, Lynn Oxborrow, David Terkla, and from presentations at the 2011 Industry Studies Conference, American University, and the International Labour Organization. 2 The New York garment district is defined as the five counties comprising New York City (Bronx, Kings, New York, Queens, and Richmond). The Los Angeles garment district is defined as the City of Los Angeles for periods prior to 1947 and Los Angeles County thereafter, except as otherwise noted. had the added advantage of being able to learn from the experiences of earlier U.S. garment districts, especially those of New York City. Nevertheless, catching up with New York was a slow process. By the end of World War II, Los Angeles was successfully attracting buyers from all regions of the country including New York (Lassiter, 1953, p. 21). However, it only accounted for 4.6% of total U.S. value-added in women’s wear compared to New York City’s 54.4%. It was not until the late 1980s that Los Angeles surpassed New York as the largest U.S. district in terms of employment and until the late 1990s in terms of value-added in women’s wear. The historical differences in the timing of development in the New York and Los Angeles have endowed each district with a distinctive set of economic institutions – trade unions, employer associations, contracting relationships, and district manufacturing cultures. The difference in unionization has been particularly significant and has shaped the way that each district has responded to structural changes in apparel markets and supply chains. The New York City garment district had highly-competitive labor and product markets in the late 19th century which were becoming increasingly chaotic as immigration sharply increased the supply of labor to the apparel industry and the number of immigrant-owned contracting shops competing for apparel production. Intensified competition brought falling wages, volatile employment, and sweatshop working conditions to the district, which sparked union organization. Growing union strength forced employers to bargain collectively on a district-wide basis beginning in 1909, first over wages and employment and subsequently over the regulation of product markets. The district’s apparel markets were heavily regulated by strong unions and collective bargaining during much of the 20th century, whereas the employers in the Los Angeles district successfully opposed unions and collective bargaining during most of the 20th century and its markets remained governed by intense competition. This paper examines the formative periods of the New York City and Los Angeles County garment districts in order to understand how and why they experienced such different growth paths. The basic argument is that that both districts had predominately craft-based industrial structures in the late 19th and early 20th centuries, but that New York had substantial first-mover advantages from the externalities of fashion knowledge, proximity to a large and affluent market, and the availability of abundant immigrant labor with apparel skills. These agglomeration economies were reinforced by intense competition in product and labor markets that wages low relative to other districts in the late 19th and early 20th centuries, by scale economies that became available during the districts transition from craft production to mass production in the 1900s, and by economies of scope and specialization as new jobber-contractor supply chains emerged in the 1920s. Beginning in 1909, the New York garment district a new “institutional” externality was also emerging as unions and employers “created a system of market regulation through collective bargaining that brought both monopoly protections and efficiency advantages the district. When Los Angeles embarked on a path to become the United States’ most successful “late-developing” district in the late 1920s and 1930s, it too had a full complement of traditional agglomeration economies and was quickly able to emulate New York’s most efficient methods of work organization and contracting. However, Los Angeles elected to avoid the New York industry’s extensive engagement with unions and market regulation through collective bargaining. These choices between union regulation and market competition represent a natural experiment in alternative market “designs”. 3 The choices made by each district during their formative periods have had path-dependent effects that have persisted into the current decade, even though union strength had substantially eroded in New York City by the latter part of the 20th century. 3 While the market failure issues of U.S. garment districts bear little resemblance to those of two sided matching markets that are featured prominently in the “market design” literature (Roth, 1990; Roth and Xing, 1994), there is th a common element in that New York City’s atomistic apparel industry of the late 19 century was developing inefficiencies from hyper-competition in much the same way that the market for matching recent medical school graduates with intern and resident positions in U.S. hospitals “unraveled” between 1940 and 1952 (Roth, 1990). Addressing the market failures in the early New York City garment district by establishing a centralized regime for regulating the garment industry through collective bargaining has similarities to the adoption of a centralized National Resident Matching Program to replace the chaotic job matching of the decentralized market for recent medical school graduates (ibid.). One possible distinction between the design of an efficient market matching arrangement and the negotiated regulatory system of the New York City garment district is that the re-design of the apparel industry’s markets included anti-competitive regulations as well as those that reduced the inefficiencies of chaotic markets. However, a recent article by Priest (2010) points out that the matching arrangements for medical school graduates also contribute to monopsony pricing. Building District Advantage: Agglomeration Economies in New York and Los Angeles Fashion apparel production in the United States has always clustered in and around urban areas where agglomeration externalities from large pools of labor, proximity to suppliers and customers, and shared industry knowledge are present. The important labor pools for apparel manufacturing initially consisted of workers with craft skills of cutting and tailoring and the know-how to capture the hang and drape of the fabric anticipated in clothing designs and entrepreneurs with the skills and experience needed to start new manufacturing and contracting firms. Later on, pools of unskilled labor that was willing to learn sewing skills, work hard for low wages, and accept seasonal employment became important. The important types of shared knowledge are about changing tastes in fashion, color palettes, fabric finishes and new product designs. The important suppliers are those that sell fabric, buttons and other trims, provide dyeing and finishing services and lend short-term operating capital. The important buyers are retailers, wholesalers, jobbers, and manufacturers that sell branded clothing. The Roots of Agglomeration Economies New York became the nation’s largest city in the late 18th century. With its large and growing regional market, influx of immigrant labor, excellent port facilities, and location at the hub of a transportation network served by the Hudson River, the Erie Canal, and a network of roads, New York became a well-established commercial and manufacturing center for many different industries by the early 19th century. Its commercial and tourism relationships with Europe made New York City the entry point for European fashion designs in the early 19th century and it was recognized as the fashion center of America as early as the 1840s (Levine, 1969, pp. 2-5). By 1880, the City had a population of over 1.2 million and the affluence of its fashion-oriented consumers provided a prime retail market for women’s wear as large department stores such as Bloomingdales and Macy’s expanded their ready-to-wear fashion clothing departments and specialty shops selling fashionable clothing became more numerous. Initially these apparel stores sold clothing made by New York City’s small seamstress and tailor shops as well as imported fashions from Europe. As demand grew for ready-made clothing in the 1880s and 1890s, however, retailers turned to emerging factory production in New York City as a source of supply and the combination of agglomeration economies and emerging scale economies soon made the New York garment district the principle source of supply for this expanding market. Los Angeles had many of the same agglomeration economies as New York in the late 19th and early 20th centuries and it had tailors’ and seamstress’ shops that made tailored men’s and women’s wear in the 19th century, some coat and suit manufacturing in the early 1900s, and it was manufacturing work clothes and dresses prior to 1920. However, its apparel production was dwarfed by New York City. For example, the entire state of California had fewer than 200 garment firms and fewer than 4,000 employees as late as 1921 compared to New York State with almost 5,000 firms and over 90,000 workers (Teper, 1937, Appendix C), and it accounted for less than 1.5% of U.S. women’s wear sales in 1925 compared to New York State’s market share of almost 79%.. What Los Angeles’ smaller and less-fashionable apparel market lacked was New York’s first-hand knowledge about European fashion and design. It was not until the late 1920s that Los Angeles acquired a key district advantage in fashion knowledge and design when it developed its own sportswear fashion “look” that was distinctive from New York’s more-European fashion tradition. Los Angeles took advantage of the growing demand for mass produced clothing in the 1920s to expand its sportswear market regionally and then nationally, and to launch itself as a late-developing garment district with access to the full range of agglomeration economies available in New York City. Pools of Labor Recent immigrants have always been the major source of garment industry labor and New York and Los Angeles have long had the highest percentage of new immigrants among major American cities. Even after immigration was restricted, first by travel limitations during World War I and then by national quotas in the mid-1920s, immigrants remained the primary labor pool for the industry in both districts. Recent German immigrants, as well as American workers, were the original source of skilled tailors and seamstresses for the artisan shops that produced clothing in the New York district for much of the 19th century. Immigrants from Eastern Europe became the major labor supply in the 1890s and 1900s (Levine, 1969, Ch. III). An observer of the industry at that time specifically noted the rapid growth in Russian (and mostly Jewish) immigrants arriving in New York during that period -- 89,097 between 1882 and 1887, 48,000 a year from 1987 to 1900, and 107,347 in 1902 – as an important factor contributing to the growth of the apparel industry (Pope, 1970, Part II, Ch. I). Subsequent waves of immigrants in the first half of the 20th century came from Italy, various Caribbean nations, and Puerto Rico, along with African-Americans migrating from the South. Chinese immigrants came to New York (from California) as early as the middle of the 19th century, but were not an important source of immigrant labor for the apparel industry until the mid-1960s. Many of these early immigrants had tailoring skills or other apparel industry experience that were critical during the craft-like stage of early apparel manufacturing while later immigrants without craft skills facilitated the introduction of mass production work methods as the industry moved towards a more industrial mode of manufacturing. Tailoring and cutting skills were particularly important in New York’s cloak and coat industry, the sector that launched the earliest craft-based apparel factories. As the demand for ready-made cloaks and coats increased, manufacturers began to contract production to smaller and more specialized sewing firms that were operated largely by immigrant entrepreneurs and staffed by semi-skilled immigrant workers (Levine, 1969, Ch. III; Waldinger, 1986). Low overhead, relatively inexpensive labor, and the willingness of immigrants to work hard and accept intermittent jobs allowed these small contracting shops to undercut costs in the larger and more-established “inside-shops” that employed skilled sewers (Pope, 1970, pp. 62-65). The largest immigrant group in the Los Angeles garment industry has always come from Mexico, but the ethnic mix has become more diversified over time. For example, Los Angeles began to draw immigrant labor from China as the demand for mining and railroad labor grew during the 19th century (Cleland, 1922). Mexican and Chinese immigrants continued to provide an important source of low wage labor to the Los Angeles garment industry through the 1960s, when they were joined by immigrants from other parts of Latin America and other Asian countries. Proximity to Suppliers Both the New York and Los Angeles garment districts have long benefited from the externalities of proximity to suppliers of fabric and other inputs. At the turn of the century, New York’s emerging apparel manufacturing drew upon suppliers that had been serving skilled tailors, seamstresses, and small artisan tailoring shops. As the manufacturers and jobbers of fashionable apparel products moved from the congested lower East Side garment district to more modern and spacious accommodations in New York’s mid-town area, a diversified fashion cluster was formed in a roughly thirty-five block area (bounded approximately by 5th and 9th Avenues and 34th and 42nd streets) that included both contractors and a variety of specialized suppliers – fabric and clothing wholesalers, button shops, pattern-makers, cutting shops, sample makers, embroiderers, dyeing and finishing firms, and equipment mechanics. The downtown Los Angeles garment district contains a similar mix of factories and suppliers, but on a less-dense scale, now sprawling over almost 100 blocks from 7th Street to the Santa Monica Freeway and from Main to San Pedro Streets. The distances are small enough in both districts to allow manufacturers to have face-to-face communications with their contractors and other suppliers. Proximity to Buyers New York is the oldest buying center in the country, and the first district to serve a national market, while Los Angeles remained a major sales center for the West Coast region until the 1940s. Large New York City manufacturers initially hired salesmen to visit buyers while smaller manufacturers sold through wholesalers who served small retailers. When national markets and large retailers became more important after 1914, and as styles, fabrics, and grades of clothing began to proliferate, New York City manufacturers began to operate showrooms for buyers who travelled to New York in advance of each season to view the full range of products available in the district. The proliferation of styles, fabrics and grades of clothing also contributed to the growth of a special kind of wholesaler in New York, known as a “jobber”, who served as a one-stop commercial intermediary by acquiring fabric, organizing production, and ensuring the rapid delivery of products to retail buyers. By 1919, the combination of year-round showrooms and the prevalence of jobbers who were highly adept in sales virtually eliminated the need for travelling salesmen as buyers either located offices in New York or came to New York for the buying season (Levine, 1969, pp. 413-415). Los Angeles initially depended on travelling salesmen and buying offices when it principally served a regional market in the 1920s and 1930s and buying offices continued to be important for reaching national, as well as regional, markets in the immediate postwar period (Lassiter, 1953, pp. 18-22). Thereafter, Los Angeles followed a similar pattern to New York as it gradually evolved into a national buying center replete with seasonal “fashion weeks” and large showroom facilities such as the California Market Center and the New Mart where industry trade shows are also organized (Bonacich and Appelbaum, 2000, pp. 34-35). Los Angeles, however, has always remained well behind New York as a headquarters location for fashion buyers. Proximity to Knowledge about Fashion and Design New York has been at the forefront of fashion and product innovation since its beginning as a fashion center. Its seamstresses and tailors benefited from the flow of ideas about fashion trends in Europe reported by affluent tourists returning from Europe during the 19th century or copied from clothing imported from France and England (Green, 1997, p. 112-114). This success as a first-mover fashion center allowed it to support a wide portfolio of design traditions and multiple product specializations. The district manufactured military uniforms in the middle of the 19th century that prepared the way for its first fashion specialization in tailored men’s and women’s wear products in the 1880s. It added specializations in shirtwaists in the early 1900s and dresses became New York’s largest specialization in the 1920s, with 125,000 different dress styles being produced in New York by 1939 (Hochman, 1941). New York began to develop its own sources of design ideas in the early twentieth century. New York-based Vogue Magazine, for example, switched from featuring articles on cultural topics of broad interest to the affluent to a focus on fashion and design beginning in 1909, the Women’s Wear Daily newspaper was founded in 1910, and district-wide fashion shows became an important public venue for showcasing fashion ideas beginning in 1943. While most of New York’s clothing designers during this period were self-taught and relied heavily on Europe for inspiration, New York also began to develop a training infrastructure for producing its own fashion designers. A small art and design school in New York, which later became the Parsons School of Design, launched a specialized fashion design program in 1904 that initially served the market for theatrical costumes. By the start of the boom period of the New York City garment industry in the 1920s, the Parsons School had become a general school for fashion and design with links to the European fashion industry through its pioneering program in fashion and design located in Paris. The 1920s was also the beginning of a long period of intense product differentiation (National Retail Dry Goods Association, 1936), which further increased the demand for both design ideas and trained designers and led to the founding a second fashion and design school, New York’s Fashion Institute of Technology, in 1936. In contrast to the rich fashion and design infrastructure that was well-established in New York by the 1920s, Los Angeles relied on a group of self-taught designers during the 1950s, most of whom were inspired by the casual western lifestyle and proximity to Hollywood. These designers began to develop sportswear and swimwear designs with a California “look” and became the backbone of the growing Los Angeles garment district by launching their own manufacturing businesses during the 1930s. The California “look” was described by one observer as clothes with “a groomed casualness, distinctive for their individuality and beauty of simplicity” (Lassiter, 1953, p. 17). Their design innovations included the dirndl dress, peddlepushers, and the backless bathing suit, and they also pioneered in the use of casual fabrics like faded denim and burlap (Lassiter, 1953, pp. 6-8). However, Los Angeles did not organize its first district-wide fashion show featuring local designers until mid-century and these fashion events have been overshadowed by their counterparts in New York (Community Redevelopment Agency of the City of Los Angeles, mimeo, April 2011). Los Angeles has also lagged behind New York in the training and employment of fashion designers (Community Redevelopment Agency of the City of Los Angeles, mimeo, April 2011). There were no schools dedicated to fashion and design in Los Angeles until 1969 when the Fashion Institute of Design and Merchandising (FIDM) was founded in a neighborhood adjacent to the Los Angeles garment district. The FIDM was followed by a second apparel design program in 1978 when The Otis Institute of Art affiliated with New York’s Parsons School. 4 4 The Otis Institute later separated from Parsons and was renamed the Otis College of Art and Design. The widely-held view in the industry is that New York’s long-term leadership in design, coupled with its skilled work force, have allowed its various women’s wear specializations to have higher fashion content than those of Los Angeles. This is confirmed by the findings of an industry survey in 1946 that the largest fraction of women’s suit and cloak firms in New York were producing better-priced garments while its closest rival district, Los Angeles, was mostly producing medium-priced products (Lassiter, 1953, pp. 23-24). More recent studies suggest that New York continues to have a fashion advantage, although Los Angeles is considered the leading design center for sportswear (Community Redevelopment Agency of the City of Los Angeles, mimeo, April 2011). District Development: Industry Structure and Market Competition As the oldest and for many years the largest of the U.S. garment districts, New York City represents a rich example of the evolution of the apparel industry’s market structure, production methods, contracting arrangements, and competitive conduct. By 1880, the agglomeration economies needed to reinforce the New York City garment district’s first-mover advantages were falling into place. The district began to adopt industrial methods of production during the late 1880s and the modern-day structure of the district was largely established by the 1920s, allowing it to serve both mass and customized product markets through a wide variety of factory sizes and contracting arrangements. The district operated under a regime of intense competition during the late 19th and early 20th centuries. Such competition contributed to growth by holding down production costs, but it also created inefficiencies from high rates of labor turnover, frequent business failures, and considerable labor unrest. A wave of large strikes beginning in 1910 forced the industry to recognize and bargain with unions and collective bargaining gradually became a vehicle for regulating inefficient competition to the benefit of both employers and workers in the district. The Emergence of the Women’s Wear Specialization Almost ninety percent of apparel manufactured in New York City at the middle of the 19th century consisted of custom-made men’s wear (Pope, 1970, Appendix A). This market was supplied by independent tailors who could cut fabric, sew garments, and perform finishing and pressing operations, and by small artisan shops that handled a wide range of designs and fabrics and performed the full range of entrepreneurial and assembly tasks needed to produce a final product. There were 12,319 garment workers employed in New York state’s artisan shops in 1880, most of whom were in New York City, and this number about doubled by 1890 (Levine, 1969, pp. 13-17) . As demand for ready-made apparel continued to grow during this period, New York City became established as the pre-eminent “first mover” garment manufacturing district. Artisan shops were the major source of supply and women’s wear was becoming an increasingly important product. Women’s wear accounted for about one quarter of New York City’s apparel production in 1880 and almost 40% a decade later (Pope, 1970, Appendix A). The tailored Cloak and Coat sector was initially the largest women’s wear specialization in New York (Helfgott, 1959, p. 52) was the main employer of highly-skilled craft labor in the industry. Major structural changes in apparel manufacturing occurred in New York City between the 1890s and the 1920s as demand for women’s wear flourished. By the turn of the century, women’s wear represented half of all apparel production in New York City, about 45,000 apparel workers were employed in women’s wear factories (Pope, 1970, Appendix A, p. 64), and New York City produced 64.5% of all clothing manufactured in the United States (Levine, 1969, p. 385). The Transition to Industrial Production The growth of apparel manufacturing in New York further accelerated after 1900 as garment making evolved from highly-skilled craft production in vertically-integrated artisan factories to industrial production with large and often mechanized firms, specialized labor, and a heavy reliance on contracting networks (Levine, 1969, pp. 2-17). Increasing scale permitted greater occupational specialization with separate crafts of cutters, pressers, and sewers (both hand sewers and sewing machine operators), and a variety of smaller crafts like embroidery. These new factories became known as “inside shops” because they incorporated the full set of skills and financial assets needed to design products, purchase and cut fabric, assemble entire garments, and market the finished product. A highly-competitive sector of small and often immigrant-owned contracting shops were also emerging during this period to perform the less-skilled sewing tasks at lower wages than were paid to the more-skilled workers in the inside shops. The growth of contractors allowed manufacturers to concentrate on the most skilled activities -- design, fabric acquisition, cutting of fabric, preparation of samples, and selling to wholesalers and retailers. At the beginning of the 20th century it was not uncommon for large manufacturers to have shifted to being “outside shops” doing some internal production while employing as many as thirty or more contractors. These contractors ranged from small shops using craft-like production methods to large factories using mass production methods with narrowly-defined job classifications (Levine, 1969, pp. 1317, 144-45). The tailored cloak and coat sector was among the first to use contractors for mass assembly, followed in later years by skirt manufacturing and then by a series of new manufactured products -- shirtwaists, dresses, and blouses. Mass production initially organized as “section work” in which components of parts of garments were assembled in stages and where the division of labor depended on the number of garments being produced. For the largest production runs, an extreme form of labor specialization was developed known as the progressive bundle system (PBS) in which sewing machine operators repeatedly performed the same narrow sewing operation in a time cycle measured in seconds. Workers received the materials to be sewn in large bundles so that they could sew continuously without interruption, the bundles moved from work station to work station until the final garment was assembled, and work flow along the PBS assembly line was carefully engineered so that each operation took approximately the same time (Doeringer and Watson, 1999). The higher productivity and lower unit costs of industrial production, along with the growth in the range of products, styles, levels of quality and price offered by manufacturers in the New York City (Levine, 1969, pp.411-13) allowed the district to further increase its national market share to 69.3% by 1909 (Levine, 1969). The Establishment of Modern Contracting Arrangements During the 1920s, the New York City garment district developed what would become its modern organizational structure. Manufacturing Census data for 1921 shows that New York’s garment employment stood at 86,540, almost two-fold growth since 1900 (Pope, 1970, Appendix A), and its market share had risen further to 74.2%. Despite its large market share, the growth of small contracting shops meant that New York City remained a district of relatively small and competitive firms, with an average employment of 18, well below the national average of about 28 (Levine, 1969, Table III; Atack and Passell, 1994, Table 17.5). Contracting was also becoming more systematic with the rapid expansion of supply chains organized by “jobbers”. Jobbers were originally wholesalers that bought apparel from manufacturers in large “job lots”, which were then re-sold in smaller in smaller quantities to retailers. Jobbers had sales experience and a range of retail contacts that manufacturers often lacked and they gradually assumed many of the functions of manufacturers -- buying fabric, contracting for apparel production, and selling finished garments to manufacturers or directly to retailers. During the 1920s, a three-tier contracting chain gradually emerged that allowed manufacturers, jobbers, and contractors to become more specialized. Specialization allowed the district to develop a manufacturing model that combined the flexibility of small scale craft production with the efficiencies of specialization and the scale economies from the division of labor. Jobbers specialized in securing materials, organizing production by contractors, and marketing clothing to retailers. Contractors specialized in fast, flexible, and low cost sewing and final assembly. The presence of the jobber-contractor system allowed large manufacturers to abandon production and to concentrate instead on building their scope economies of design, brand development, and marketing, while using jobbers and contractors to capture scale economies of manufacturing and provide quick and flexible supplies of many different products. Small “inside-shop” manufacturers with the flexibility to switch quickly from one product to another served a high-end market niche by designing and producing smaller orders of highquality products at relatively quick speeds. Jobbers became the focal point for women’s wear manufacturing in New York City, particularly in the dress sector, as styles proliferated and mass market demand increased during the 1920s. However, some jobbers also had design capabilities and some manufacturers also managed their own contracting so that the distinction between jobbers and manufacturers became blurred and these terms are now used interchangeably by the industry. Los Angeles resembled New York in the late 19th and early 20th centuries, but on a far smaller scale and with less-fashionable products. It also experienced similar structural changes to those in New York, although they were more gradual and spread out over a longer period of time. Like New York, Los Angeles initially adopted the “inside shop” model for both women’s and men’s wear. But unlike New York, it largely continued this model through the mid-1940s, using contractors mainly during periods of peak demand and for specialty work like embroidery (Lassiter, 1953, p. 41). This reflected the greater importance of products like work clothes (which could be mass produced) and sportswear (where at least the more-basic products had longer production runs that permitted greater specialization and division of labor). However, even the small women’s wear manufacturers that were introducing fashionable clothing styles continued to use the inside-shop model through the 1930s (Lassiter, 1953, p. 41). The late 1920s was also the period when Los Angeles first aspired to become a significant late-developing district that could escape the confines of its regional market and compete with New York City in the emerging market for sportswear. When Los Angeles began to realize this goal in the mid-1940s, manufacturers had already started to contract out some of their production and contractors were gradually becoming an important source of lower cost production in the district. Although some of the largest Los Angeles manufacturers have retained a manufacturing capacity for samples, higher-end products where quick production is needed, and products where there were scale economies to be realized, the general trend since World War II has been to contract more and more production. As in New York City, all but the largest manufacturers have now mostly abandoned in-house garment production to their contractors (Bonacich and Appelbaum, pp. 11-14, 27-29). Cut-Throat Competition and First Mover Advantage With its prevalence of small firms, specialized supply chains, mobile capital and labor, dense information networks, and piece-rate pay incentives, the apparel industry has many earmarks of a competitive industry. The combination of a relatively atomistic market structure, easy entry by firms, quasi vertically-integrated contracting chains, and substantial agglomeration economies found in New York City garment district has often led it to be singled out as a model for the efficient production of fashion apparel (Lassiter, 1953 Ch. VI; Bonacich and Appelbaum, 2000; Kranton and Minehart, 2001; Wintrobe and Breton, 1986). Competition in New York City’s garment district peaked during the late 19th and early 20th centuries when the apparel industry benefited from a form of hyper-competition, known in the vernacular of the industry as “cut-throat” competition, which was fueled by high levels of excess capacity. Excess capacity is a normal part of the apparel industry because of the seasonality of apparel demand and the volatility of fashion markets. However, additional excess capacity was also introduced during this period by the increasing returns to scale that accompanied the transition to mass production and from the easy entry of immigrant-owned contracting firms that rapidly expanded sewing capacity under the emerging contracting system in New York. Excess capacity triggers an adjustment response where prices can be driven below average costs and firms must either lower their production costs or leave the industry. In a laborintensive industry like apparel with relatively little technological change, reducing costs often leads to “sweatshop” employment practices – cutting wages, extending daily and weekly work schedules without extra compensation, using coercion and the threat of job loss to spur greater labor productivity, and allowing working conditions to deteriorate – as manufacturers and contractors compete for scarce work. Poorly-paid labor, ten hour work days, and frequent after-hours work at home were common features of the apparel industry in the New York City garment industry of the late 19th century (Levine, 1969, pp. 168-173). Working conditions were also unsafe and unhealthy. A typical New York apparel factory, for example, had crowded production areas that were “packed so densely that the girls could scarcely move from their chairs”, poor ventilation, few toilet facilities, and little drinking water (Levine, 1969, p. 10.) Conditions improved somewhat during the 1890s as manufacturers with sweatshops on the lower East Side were moving to larger and more modern premises in mid-town Manhattan. However, cut-throat competition, long hours, and poor factory conditions continued as the new wave of immigrant contractors located in the lower East Side lofts vacated by manufacturers (op cit., pp. 168-173). Between the rapid growth in the supply of immigrant labor and the presence of cut-throat competition, wages were also lower in the New York garment district than in other major districts during the late 19th and early 20th centuries, particularly for less-skilled labor. Wages of factory-based sewing machine operators in 1906, for example, were 49% higher in Chicago than in New York, 34% higher in Boston, and 23% higher in Baltimore, with only Philadelphia (with a wage premium of only 2%) coming close to New York’s wage level (Segal, 1960, p. 94). New York manufacturers deliberately maintained a larger contracting capacity than was needed either to meet seasonal demand or to achieve minimum efficient scale of production (Teper, 1937, pp. 13). A study of the New York apparel industry at the turn of the century concluded that this excess capacity was intended to enhance manufacturers’ bargaining power over contractors. Manufacturers regularly took advantage of excess capacity to negotiate lower prices with contractors who regularly came “hat in hand like a beggar” to their offices looking for work, and that contractors regularly pressed their workers to accept lower wages and sometime bear the costs of equipment and supplies (U.S. Industrial Commission 1901-02, vol. XV p. 332, cited in Levine, 1969, pp. 16-17). As the jobber-contractor system spread during the late 1910s and early 1920s, jobbers replaced manufacturers as the drivers of excess capacity. They conducted auction-like bargaining sessions with groups of contractors in which the number and production capacity of bidders substantially exceeded the volume of orders being placed and competition for business in the early 1920s was seen to be as “chaotic” as in the 1900s (Levine, 1969, pp. 403-407, 409). A study in 1925 by the Governor’s Advisory Committee on the New York Cloak Industry reported that 18.5% of New York cloak contractors accounted for 86.3% of contracted cloak output while “the remaining 81.5% of contractors …. were retained purely for bargaining purposes” (Cited in Teper, 1937, pp. 13-14). The Committee also noted that excess contractor capacity had spillover effects on the wages paid by manufacturers in their inside shops: “When work is scarce, as it usually is, except for a few weeks in the season, the workers are told that in order to meet the exigencies of price competition and to bring some work into the [inside] shop they must enter into secret arrangements contrary to the minimum labor standards which have been agreed upon.” (ibid, p. 14; Levine, 1969, p. 16). The combination of strong agglomeration economies, growing economies of scale, and falling production costs from cut-throat competition gave the New York City garment industry an edge over other districts in the late 19th and early 20th centuries that allowed it to secure a commanding share of the women’s wear market. Nevertheless, New York City’s apparel markets remained contestable by other garment districts because fashion designs are easily copied, skilled labor can be trained, jobs can be simplified, labor pools can migrate, regional wage differentials can change, and national buyers can quickly learn where similar products can be obtained at lower prices. The risk of competition from outside the district was also rising because cut-throat competition had its own inefficiencies in the use of labor and capital. It spawned revolving pools of marginal firms charging prices below average costs of production, high rates of labor turnover and frequent business failures inhibited training, and the low wages and poor working conditions that accompanied cut-throat competition caused spontaneous strikes and labor violence. Strikes in particular were becoming a negative externality for the entire district because they brought widespread losses of time-sensitive production, encouraged buyers to shift orders to other districts with less labor unrest, and resulted in some manufacturers relocating to areas where strikes were less likely. The inefficiencies of strikes became an even more serious concern in the 1880s as garment workers sought to address problems of low wages and sweatshops by forming trade unions. Skilled tailors and cutters employed in the cloak and coat sector, New York’s largest and most skilled specialization at the time, were the driving force behind these early unionization efforts (Levine, 1969, pp, 92-98). Strike activity tended to be concentrated at the start of each manufacturing season when the costs of lost production were greatest and it was most difficult for employers to replace the skilled workers who were on strike. These strikes were initially of short duration and there was high turnover among the early unions at the end of each season (Levine, 1969, p. 42-43). By the 1890s, however, New York garment unions were becoming larger, more numerous, and more militant. Militancy took the form of demonstrations and marches as well as mass picketing and there was increasing solidarity among the members of these new union locals that were forming around relatively homogeneous work “communities” defined by craft and type of product, religion, nationality, and neighborhood. Opposition to unions became correspondingly more intense union militancy increased. Apparel employers discharged employees who joined unions, lockouts were used more frequently when union workers went on strike, and larger strikes and demonstrations often led to arrests and sometimes prison terms for strikers. Because concerted employer resistance to unions was difficult in an industry dominated by small firms and seasonal demand, employers also began to form industry associations to coordinate district-wide strategies for defeating unionization (Levine, 1969, pp. 56-57). Smaller union movements were also emerging in most of the other major garment districts – Chicago, Philadelphia, Cleveland, and Baltimore – and there were various attempts during the 1890s to bolster local bargaining power by bringing local unions under a national union umbrella (Ulman, 1955), principally through affiliations with the eclectic Knights of Labor, the radical IWW, or the craft-based American Federation of Labor (AFL). National affiliations, however, proved impermanent for local garment unions in New York and elsewhere until 1900 when New York’s largest craft workers union, the “United Brotherhood of Cloak Makers’ Union No. 1 of New York and Vicinity”, convened a national convention of “all workers in the trade, namely of those working in cloaks and suits, or at custom ladies tailoring, or in skirt shops” for purposes of establishing a nation-wide, craft-based union for skilled workers in the women’s clothing industry (Levine, 1969, pp. 102-103). Although only seven locals from four cities (including New York) were represented at this convention, the AFL quickly granted an official charter to the International Ladies’ Garment Workers’ Union (ILGWU), which was headquartered in New York City where roughly half its membership was located (Levine, 1969, p. 103). Counterpart national unions in the men’s wear industry (the United Garment Workers Union and its successor the Amalgamated Clothing Workers of America) were also active in New York during this period, but the ILGWU provides the earliest and richest example of how unions and collective bargaining could “create” new agglomeration externalities in the New York City garment district by establishing a collective mechanism for re-designing New York’s apparel markets.