THE TEXTILE FINISHING INDUSTRY IN NEW ENGLAND: IS FLEXIBLE SPECIALIZATION OCCURRING?

This project is an in-depth study of flexible manufacturing in the New England textile finishing industry. Chapter 1 represents a literature review of the current debate surrounding the flexibility issue. Chapter 2 provides a brief outline of the methodology of the project. Chapter 3 provides a history of the textile industry, both in New England, and the United States as a whole. Chapter 4 zones in specifically on the textile finishing industry. It describes the production processes involved in finishing, and gives figures for employment within the industry at the national level. This chapter also identifies the ten key SIC codes that make up the finishing industry. Chapter 5 looks at industrial organization. It provides general information about the impact of imports on the industry as a whole, and the implications that this has for textile finishing. Further, it addresses the issues of profitability, productivity, supplier-customer relations, and firm size within the finishing industry. Chapter 6 addresses the issue of whether flexibility is increasing in textile finishing in New England. Chapter 7 provides a summary of the industry interviews that the study team for this project has undertaken at the


THE FLEXIBILITY DEBATE
The world we find ourselves in today is very different from that of the immediate post-war period. Economic downturns seem no longer to be short pauses in the progression to greater affluence. Rather, these downturns seem to pose a real threat to the economic system on which post-war growth is based. Piore and Sabel (1984) assert that the present deterioration in economic performance results from the limits of the model of industrial development founded on mass production. 1 Piore and Sabel question the viability of large, vertically integrated firms engaged in mass production, that once characterized an unprecedented period of growth, and suggest that today the most successful industries are increasingly characterized by "flexible specialization", a term indicating spatially dense networks of relatively small, vertically disintegrated firms.

The Flexibility Debate
Support for this hypothesis has been mixed. Some empirical studies strongly support the claims made by Piore and Sabel (1984), while others have found alternative forms 1 Mass production is defined as the use of special-purpose (product specific) machines and of semiskilled workers to produce large quantities of standardized goods. This chapter will review these divergent perspectives. Feldman and Yuan (1990:2) point to the fact that the debate over flexibility is ~ore and more being placed on theoretical ground demarcated by the French "Regulation School" of political economy.
In this approach, periods of stable growth in capitalist economies require a wide variety of non-market institutions, norms, and practices. These historical entities collectively constitute an "accumulation regime", determine the system's characteristics functioning, or "mode of regulation." They note that if it were not situated in this context, much debate over flexibility would not exist.
At one end of the spectrum, Piore and Sabel lay out the concept of flexible specialization and offer it as the more viable of two alternatives, the other being multinational Keynesianism. They do not, however, overlook its problems.
For example, they note the difficulties of envisaging a flexible economy that is successful, as there is no model on which to build such a notion, and this would require what they refer to as an exercise of the imagination.
Piore and Sabel offer four types of flexible specialization. The first is regional conglomerations. The second type of flexible specialization is federated enterprises. This form is associated with the "systeme Motte" and Japanese companies. Here the "federated enterprises hold one another's stocks and have interlocking boards of directors." (Piore, Sabel, 1984) The firm is not organized under an hierarchy structure, as with the mass production s y stem.
The third and fourth forms of flexible specialization are the firm with the solar-system model of orbiting suppliers and its close cousin, the workshop factory. Such firms are organized in nineteenth century style workshops.
Aqg1omeration Scott (1988) takes the concept of fle x ible specialization and adds a new dimensi o n, that of agglomeration. Agglomeration is the process whereby verticall y disi n tegrated firms cluster together to increase 3 the.ir combined total volume.
Localization economies are achieved for all firms in a single industry at a single location, consequent upon the enlargement of the total output of that industry at that location.
Scott observes that by the 1970s, in the face of the deepening unworkability of the fordist regime the governments of countries in North America and Western Europe started to dismantle the welfare-statist arrangements that had helped to regulate it but were now only becoming an additional element of the overall crisis.
As this one regime of accumulation began to collapse, in the vacuum that was created, another was beginning to take shape. In particular, Scott refers to the emergence of a number of new flexible forms of productive activity. He believes these are concentrated in three groups of industrial sectors: revivified artisanal and designintensive industries producing outputs largely but not exclusively for final consumption, high technology industries and their associated group of input suppliers and dependent subcontractors, and service functions, especially business services (Scott, 1988:175). Italy, rely on small interlinked establishments. Storper (1989)  Storper believes that what must be done to make a valid argument that a new, general period of capitalist economic development is occurring is to take what he refers to as the ''hard cases" which are the large, complex products in capital-intensive industries. Storper goes on to use the illustration of a study undertaken of the automobile industry in France. Amidst claims that the industry was actually a more automated form of mass production, he actually found a form of mass production operating alongside production units that were internally flexible. He further asserts that system logics are increasingly those of flexibility.

Flexibility and Labor
Schoenberger (1988b) accepts much of the theoretical underpinning of the flexibility theory, but questions the bright reality portrayed by its exponents. She points to the fact that flexible specialization promises little for those workers who have borne the brunt of industrial restructuring. She goes on to observe that those who survive the transition will have less to look forward to than under mass production. There may be gains in job stability and union relations, as fle x ibility requires more 7 participation of workers than was needed under fordism.
Schoenberger also sees flexible specialization as demanding extremely flexible work schedules, for example the growth of part-time work is seen as increasing flexibility.
The problems with this new organizational structure are not limited to the lower ranks of the workforce. With workers on the shop-floor having been made redundant, upper-echelon workers will increasingly become the major labor-cost concern for the firm. Further, with increasing pressure to reduce the cycle time from design to manufacturing, efforts will move apace to standardize and automate design and programming tasks.
Mass production was based on the broad expansion of incomes, yet writers have observed that flexible specialization has been characterized by a social polarization of incomes.
In addition, to this phenomenon of few rich people keeping the economy going, support from the welfare state to the poor is increasingly being withdrawn.
Schoenberger concludes that the transition to a flexible mode of accumulation will be fraught with troubles of its own. Indeed, the result may well be a crisis of effective demand. Sabel (1988) argues that until the mid-nineteenth century the region was the unit of economic activity. He 8 asserts that regions such as Sheffield, England, were both flexible and specialized. By the 1960s the giant corporation had replaced the region as being the main unit of economic activity, and with this organizational shift came all the traits of mass production. However, by the 1970s, when international competition rose, and world markets began to fragment, the same firms that had adopted mass production methods, became wary of investment in product-specific machinery due to the vulnerability of markets and rapidly changing demand. In response to this volatility, firms increasingly experimented with flexible organizational forms, which allowed for a more rapid response. Sabel asserts that this move to more flexible responses has led to the reconsolidation of the region as an integrated unit of production. Sabel also sees several problems. Even if regional 9 economies do emerge, they are unlikely to survive without the aid of national institutions, still undefined. Flexibly specialized firms may be better able to survive economic turbulence than mass production firms are. This does not make them immune to failure. Crisis regions must be assisted by buoyant areas if they are to survive. If this is to be successfully achieved a new national system of resource allocation may need to be created. Donaghu and Barff (1990) argue that they have found an alternative to flexible specialization. They argue that Nike's production system represents an alternative post-Fordist form of industrial organization to manufacturing based on flexible specialization. Donaghu and Barff (1990) show, "that within Nike, inflexible means of production coexist with highly flexible and dynamic forms of company organization. Third, with factories that only manufacture Nike products, the company places orders once a month, thus avoiding production varying by more than 20 percent each month. This acts as a stabilizer on production for factories that have been with Nike for many years as well as those that are newer.

A1ternative Models of Industrial Organization
The second focus of the paper supplements Schoenberger's work (1988a), that of a critical evaluation of the "New International Division of Labor" model.
According to Donaghu and Barff, (1990:3) Nike's organization of production is influenced by the attraction of low-wage labor in the periphery and only marginally by -the market potential such locations hold. The geography of the Nike production system however, is also determined by the nature and the stability of the long term partnerships it has established with its producers.
Since 1980, Nike's manufacturing organization has become progressively reconcentrated in South East Asia.
In terms of flexibility in machinery, the only elements of Nike's machinery found to be flexible is the computeraided design, computer-aided engineering, and some of the numerically controlled machines used by several South Korean subcontractors.
Thus the authors draw the conclusion that machinery within the firm must be considered infle x ible.
Donaghu and Barff see labor at Nike as numerically rather than functionally flexible. Numerical flexibility is the propensity of firms to adjust the quantity of labor used in production.
Functional flexibility is the case in which production workers are responsible for not only manufacturing but also process innovation, training, meeting with customers, and maintaining equipment.
Nike has a history of shifting production between regions and countries. internal integration of marketing, production, and R & D and the integration of product and process innovation; the use of flexible machines; the pursuit of economies of scope, internally, by multiple production, and externally, via strategic alliances; diagonally integrated interfirm relationships, wherein specialized production takes pla c e in-house and routine production is subcontra c ted; an occupational structure on a multiskilled group of engineers and nonspecialist workers in a core (i.e. functional flexibility) augmented by a casual, (numerically flexible) peripheral, workforce.
Donaghu and Barff assert that with the exception of the second criterion, Nike closely coincides with Cooke's definition.
They also observe however, that if they take Schoenberger's point that not every factor in a production system must convert to a new and discernible form, then Nike could be perceived as flexible. But if the move to flexibility is a process in which the core of the transformation is in the workplace, then Nike's industrial organization cannot be thought of as flexible.
In conclusion, Nike is fundamentally different from the mass production firms of the past. Nike is demand-driven, while mass production firms under Fordism were supplydriven. But Nike cannot be characterized as flexibly specialized. Sayer (1988) a rgues regardless of the state of mass production in the West, it is still widely in use in Japan. the central contrast between the alleged decline of mass production and the rise of small-batch production, and the implicit associations between vertical integration and mass production. (Sayer, 1988) Sayer concludes that superior economic performance by the Japanese and Germans over the United States and Britain has less to do with the choice of flexible specialization above mass production and more to do with broader environmental characteristics that cut across labor processes and production scale distinctions. Markusen (1991) Amin (1989) is also skeptical as to the validity of the flexible specialization theory. He points to the lack of attention paid to the widespread usage of batch and craft production under the Fordist regime. Amin also notes Williams et al (1987:405-439) point out that mass producers have been able to adapt to cope with greater consumer demand for differentiated goods. Amin goes on to quote Solo (1985) who observes that: No matter how flexible specialization is, it cannot wriggle out from the descending force of a decline in aggregate demand ... Nor would future uncertainty be less debilitating for the flexible specialization technology than for that of mass production.
Amin highlights, as have others, the evidence of little job and wage security, poor pay, and reduced formal protection of basic rights. He also argues that Sabel underplays the existence of differences in the characteristics and origins of various contemporary industrial districts, such as the key importance of large retailers and subcontractors as key actors in some areas, small firm networks in other areas, or even government expenditure especially in the defense sector, elsewhere (Amin et al 1986;Markusen 1991).
The main thrust of Amin's paper is an analysis of whether firms in the ''Third Italy" are flexible. As a result of his analysis he concludes there are severe problems with the flexibility thesis as defined by Piore and Sabel. He notes that in the "Third Italy": very few, and only the oldest of the industrial districts come close to resembling the "ideal type". The organizational structure in these areas appears to be changing, with new forces favoring industrial concentration emerging. Despite what Piore and Sabel claim to have found , some of the key elements of the successful industrial district are highly place specific.

Piore and Sabel use the textile industry in what has
the "Third Italy" as one of their key examples of the reemergence of craft principles, which led them to hypothesize about the emergence of flexible specialization on a more universal scale. Because much of the flexibility argument was based on the assertion that the phenomenon was occurring in the industrial districts in Italy, writers have returned to take a second look · at the Third Italy. Amin (1989) examined the Third Italy in light of the claims made by Piore and Sable (1984). He observes the rise in the number of small firms located in the region, observing the consensus that the initial rise in this number came about as a result of the extensive use of subcontracting by both medium and large firms in the area.
The combined growth of this sector, however, was attributed to the small firms ability to free themselves of the subcontracting relationships and sell specialist goods and services in the market and to other specialist buyers.
Amin notes that official statistics show a poorer level of innovation among small firms than their larger counterparts. They also show, conversely, that this lower innovative level does not translate into lower profits.
Smaller firms experienced lower wages and longer working hours, with lack of industry regulation to establish minimum working standards. In particular, the debate has not thoroughly addressed aspects of flexibility which apply to the textile finishing industry. Walker (1990) has begun to address aspects of the flexibility debate that are applicable to the textile finishing industry, but the argument has not yet been developed far enough. This problem will be addressed in chapter 6. 20

METHODOLOGY
One of the major problems with prior studies of flexible production is that they overwhelmingly rely on case studies, but little, if any systematic selection process was used to pick the industry. This study however, came about as the result of a systematic selection process.
In the first phase of the project, national data were were also computed using a regression-residential approach (Miller; Feldman, 1990).
The combination of these indicators identified which industries (analyzed at the four-digit SI C code l e vel) appeared to be e x periencing trends that implied increasing vertical disintegration and fle x ibility .
Based on analysis these indicators, te x tile finishing, aircraft parts, and medical and surgical instruments were selected for in-depth study.

The Transition to Mechanization
During the colonial period, the textile industry remained essentially a craft industry. The transition to mechanization began in 1790, when the firm of Almy and

Brown, tapping into the technical innovations of Samuel
Slater, spun cotton yarn with water-powered machinery. This led to the factory system of production being introduced.
Kennedy ( (Gordon et al, 1982). As Kennedy observes, the launching of this mill was, in many ways, to shape the course of how the industry would develop in at least four respects. According to Kennedy these were: the integration of power-driven spinning and weaving machinery within the same plant (spatial integration) , the f.inancing of a large-scale enterprise by the formation of a joint stock company and the sale of common stock, the employment of an independent selling house to distribute the product of the mill, and establishment of a separate textile machine industry. The Waltham system differed from the Rhode Island system. Under the Rhode Island structure, integration ran throughout the process from spinning machines to spinning to financing.
Weaving machines, weaving and finance were also integrated with distribution. This stands in contrast to the Waltham situation where spinning and weaving machines were disintegrated from spinning and weaving operations. The distribution process was also disintegrated. Financing under the Waltham system was partially disintegrated.
Up to 1850, labor for the mills came predominantly from New England farms. Mill owners lured farmers and their children, notably their daughters, into the mills, but many returned to farming after they had earned enough money to buy land. This was the case with much of the mill workforce around Lowell, Massachusetts, this being the single most important town in firms of textiles, in the country, meaning that there was no permanent mill population north of Boston.
In terms of the accommodations that mill workers occupied, it was a common practice in towns with less than 10,000 persons to find that the mill company would own a sizeable amount of the housing stock within the town. In 1924, for example, 62 out of 187 of the cotton-textile concerns in Massachusetts owned or controlled company houses (Burgy, 1932:153). However, this tendency was to change with improved transportation facilities in the form of better roads and automobiles, making for an increase in the independence and mobility of the mill worker. This factor led to a change in the relationship between management and worker due to the decreased reliance on the part of the 25 worker on management for housing.
The invention of the Draper automatic loom in 1895 and its distribution after 1900 had a very profound impact on the direction that the industry took for the following half century. Named after its inventor, this piece of machinery allowed for the changing of the bobbin without stopping the loom, the shuttle being refilled automatically with a fresh bobbin from a magazine on one side of the loom (Kennedy, 1936:8 (Kennedy, 1936:12).
In 1913  Labor proved to be a more significant portion of the total manufacturing cost in New England than in the South.
For example, in 1927, wages constituted 22.6 percent of value for fabrics produced in New England, but only 21.5 percent in the rest of the country. Further, the average annual wage paid to an operative in New England in 1927 was $944.00, while elsewhere this was only $724.00 (Burgy, 1932:145) One of the main reasons why wages were higher in the North was that the region was unionized, whereas in the South there was little, if any organized labor.
At the beginning of the emergence of the South, lack of skilled workers was a major inhibiting force. Wright

The Textile Industry Today
The entrance of the developing nations into the arena of competition in manufacturing added mounting pressure to New . England textile firms trying to find a lifeline in an increasingly hostile market.
In addition, demand began to stagnate. Since the 1970s demand in the United States has grown at the rate of less than one percent annually. The two forces increased competition and squeezed profits. The response has been to attempt to raise labor productivity by investing in automated machinery. There has also been a rise in mergers as firms scramble to gain economies of scale and scope.
The major response has been to move into specialty products. Firms in New England have sought to use existing machinery and technology combined in such a way as to yield them an advantage over the competition.
The growth of textile imports has had a profound effect on the textile industry as a whole. The trade deficit in textiles has risen dramatically in recent years. Net imports jumped to almost $1 billion in 1983 and increased to $2.5 billion in 1986 (See Table 2). This has been a growing trend from 1980 on (Kane, 1988:141).
Although the import share for textiles is not large relative to manufacturing as a whole, that share is rising.  1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

THE PRODUCTION PROCESS TODAY
This chapter looks at textile finishing as distinct from textiles as a whole. Stout defines fabric finishing as: those processes that fabric undergoes after it is fabricated by weaving, knitting, felting, or other methods in preparation for use by garment manufacturers or, in some instances, directly by the ultimate consumers ... Finishing becomes ever more complicated with the appearance of new fibers, new finishing processes and chemical compounds, and new ways of combining fibers into fabrics (Stout, 1970:329).
In 1970, when Stout was writing, firms tended to include both the finishing process and the actual making of cloth.
Stout goes on to note that despite this trend toward integration, in many large textile mills a large proportion of fabrics are still finished in mills other than the one where they are fabricated. Mills whose function is finishing are called converting mills, and persons who work there are known as convertors.
Fabrics as they come from the looms to the finishers are known as gray goods. This term has grown with increased usage to include any unfinished fabric and not just a fabric in need of bleaching. One would not except to find a gray good sold as the finished form, e x cept perhaps in the case of industrial usage.
Stout provides a detailed list of five reasons for 36 performing the task of finishing: to improve aesthetic qualities; to impart specific functional properties not inherent in fiber or fabric; to improve easy-care characteristics; to improve durability by improving certain service qualities; and to permit adaptation of present fabric constructions to changing market needs.

Fa.bric Finishing Processes
Fabric-finishing processes are categorized in different ways. These include: physical and chemical, wet and dry, for appearance and for function, and temporary and permanent. Stout (1970:331) divides finishes into three groups. These are: general processes used for many fabrics to improve appearance, serviceability, or both; processes designed primarily to change appearance or surface texture, and for which any change in function or serviceability is largely incidental; and processes designed to give special functional or serviceability properties not inherent in the fiber or fabric, and in which any change of appearance is incidental.  Singeing is the process by which fiber, projecting yarn ends, and other fuzz is burnt off to leave the fabric with a smooth, even surface. The mechanics of this operation involve the passing of the fabric over gas flames or hot plates at a speed fast enough so that the fabric itself is not burned but any protruding material is removed by burning. The fabric is then passed over a wet surface to insure the extinguishing o~ any smoldering fibers.
Shearing is the process of cutting pile to a uniform height, or in certain cases, eradicating all fuzz from the fabric. Shearing may occur several times for maximum evenness.
Weighting, softening, sizing and fulling, are processes for changing the feel of fabrics, to replace waxes, gums, or oils lost in finishing, and to increase flexibility and drape of the fabric. Weighting is the process by which silk is treated with metallic salts to give weight and body to the fabric. Lead, zinc, and tin have been used in this process, although today lead is no longer used because of the health hazards associated with its use.
Softening is the process by which one or more products is added to the fabric to make it softer. Softening agents used include dextrin, glycerin, sulfonated oils, waxes, and soaps.
Sizing is a yarn process in preparation for weaving.
Fulling agents, when applied, have the same effect on wool as sizing materials on other fabrics. This effect is used as a finishing process on wool products. Some of these products are removed from the cloth as a result of wear or washing, while others are durable.
Tentering, crabbing, and heat-setting are described as: the final processes by which warp and filling of woven fabrics are set at right angles to each other and pulled out to the dimensions at which they are to be dried as finished products.
Crabbing is the term applied to wet-tentering of wool; heat-setting is used for most of the thermoplastic fibers. Pins or clips grasp the selvages and hold the fabric in place as it travels through the machine, releasing it at the end where it is folded or rolled for packaging (Stout, 1970:335).
The processes described above are the general processes used in finishing of most fabrics.
Numerous other, more specialized finishing processes are also used. 5

Texti1e Finishing SIC Codes
The  In terms of the way in which the Census of Manufactures has reported the statistics used in this chapter, the statistics, including figures for employment, payroll, etc are reported for each establishment as a whole.
Aggregates of such data for an industry reflect not only the primary activities of the establishments, but also their activities in the manufacture of secondary products as well as their miscellaneous activities, including each firm's activities as contract work on materials owned by others and such work as repairs. thousand, an increase of 6.1 percent (See Table 3 ).
Further, the textile industry lost out in terms o f the percentage employed within total manufacturing. This share fell from 2.05 percent of total manufacturing in 197 2 , to 1.54 percent o f to t al manufacturing in 1987 (See Table 3 ).

INDUSTRIAL ORGANIZATION
This chapter profiles industrial organization in textile finishing.

Overview
The textile industry as a whole is suffering from import pressures, rising raw materials prices, and stagnating demand, and this has significant implications for textile finishing. When unfinished textile goods work their way down the line to the finishing stage, they too are confronted with reduced volume, higher prices passed on as a result of rising raw materials prices, and less work due to a rising share of the domestic market being held by cheaper foreign imports (See Table 5) .

Employment Trends in the Textile Industry
The national figures for total employment in textile finishing presented in Chapter 4 ( Standard and Poors (1990) discusses profitability in the entire textile indus t ry (SIC 22). Data on profitability in textile finishing alone are not available, so data may not apply exactly. 1988 1987 1986 1985 1984 1983 1982 1981 1980 1979  disposable personal income has increased an average of 9.7% annually since 1980, sales of U.S. textile mill products increased only 1.9% annually during that period. Failing to reap the implied benefits of growth in consumer spending, the industry turned to acquisitions (in textiles) to achieve that growth.

YEAR
The trend toward consolidation has also been spurred by a desire on the part of certain companies to become market leaders in core businesses. This allows firms to increase their shares of existing markets, and to acquire shares of new markets. Through consolidation firms also achieve economies of scale and scope.

Imports
The import threat has been one of the major forces that has led to the rising number of acquisitions. Most new investment by the industry is e x pected to be to enhance productivity, rather than increase capacity.
Standard and Poors observes that textile industry earnings as a percentage of sales have varied widely. This is very much a reflection of the tendency for variation in manufacturing as a whole, although textile earnings are generally seen to cover a wider range.
Despite improved productivity and profitability, profit margins within the industry remain low.
In 1988 and 1989, for example, net profits for all manufacturing companies were 5.9% and 5.0%, but in textiles were 3.1% and 2.5%, respectively (Standard and Poors:1990).
In addition to modernizing operations and reducing the size of the workforce while increasing productivity, textile firms have attempted a series of profit-enhancing activities.
They have also sought to achieve greater asset management and to increase inventory turnover. However, such factors as high real interest rates worked against the industry in the 1980s.

Textile Wages
Wages in the textile industry are lower than the wages  Table 6) YEAR 1989 1988 1987 1986 1985 1984 1983 1982 1981 1980   persons than employ below twenty persons, the average size of finishing firms is small.

State of the Industry
The discussion above provides a picture of an industry struggling to survive in the new global economy. The industry has restructured since imports began to have a strong impact in the beginning of the 1980s. However, despite discarding inefficient mills, laying off workers, and adopting automated machinery, the industry continues to struggle to survive.
The industry's strategy has been to invest in increased productivity, rather than increased capacity. This choice makes sense in the current economic climate, as demand is stagnant. Despite reductions in the workforce, the industry continues to be less profitable than manufacturing Criticism of flexible specialization theory as advocated by Fiore and Sabel (1984) centers around reliance on arbitrarily selected case studies of local areas that are then specialized without justification. Amin (1989) argues that unique circumstances in the chosen study areas were the reason flexible specialization thrived, rather than just flexible specialization itself being the cause of the areas' success. 8 As the flexibility debate has developed, some studies have found some, although not all, of the attributes of the classic definition of flexible specialization are to be present; enough indeed to support some of the assertions made by the theory, for example, the assertion that Nike is 8 See Fiore, Sabel (1984) for a discussion of industrial districts in the Third Italy, particularly the textile industry in Prato.
Also see Amin (1989) for a comprehensive criticism of flexible specialization in industrial districts and a critique of the theory's wider applicability.) flexibly integrated (Donaghu and Barff, 1990).

F1exibility and the Texti1e Finishing in New England
This chapter argues th~t the textile finishing industry in New England shows signs of some of the traits of flexibility.
However, the industry cannot be called flexible in the classic sense of the definition for several reasons.

Organizational Structure
The organizational structure of textile finishing firms in New England is much the same as it was fifty years ago.

Flexible Specialization
One must not overlook Gertler's (1988) argument that we are not experiencing a dramatic break from the past, but that batch production existed under fordism. This point is very relevant to the New England textile finishing industry, which does not fit neatly into the definition of mass production. Some of the traits associated with flexible specialization have been have been present for many years.
For example, many small firms with low and moderate numbers of workers e x ist in the industry. Walker (1990) argues extensively that the debate is not about a dichotomy. The choice is not simply one of whether an industry uses flexible specialization or mass production.
This simplifies a complex reality in which overlaps are plentiful. Walker (1990: 62-63) asserts that the source of confusion is: the misplaced emphasis on scale o f output (length of runs) instead of production methods (type of machinery) . One may get very long runs out of generalized machinery or short runs out of dedicated machinery, with corresponding effects on overall efficiency of labour and machine use, but the principles of mechanical improvement in each case are different.
Walker's argument with regard to dedicated machinery raises an important point, but he fails to take the argument far enough to emphasize the importance of looking more closely at dedicated machinery.

Flexibility in much of New
England's textile finishing industry rests on the firms' ability to use old dedicated machinery in flexible ways. Currently there is no well developed literature or ·theory to e x plain the form of fle x ibility which seems to be occurring in the New England tex tile industry.

Texti1e Finishing in New Eng1and
The textile finishing industry in New England has recognized the inevitability of large sections of the textile business being taken over by cheap foreign imports.
In many cases (usually in the cases of those who have survived) the industry's way of dealing with this threat has been to retreat into niche-oriented production areas, where superior use of finishing processes has allowed domestic firms a winning advantage over imports.
It is important to note that the niche-strategy is a defensive strategy rather than an offensive one.
This realization however does not seem to have stemmed the decline in employment in the industry. There has been a deliberate policy to increase productivity. The emphasis in the industry has shifted from increasing capacity to increasing efficiency.

Characterization of the New Eng1and Texti1e Finishing Industry
The New England textile finishing industry is characterized by: * Declining total employment * Rising import competition * Large numbers of small or medium sized establishments * Lower wages relative to manufacturing as a whole * A move to seek niches where firms can insulate themselves from foreign competition This characterization will be used t o argue that the industry cannot be called fle x ibl y spe cialized in the classic sense of the definition, but that, as discussed above it rather has "touches" of fle xi bility .  Piore and Sabel (1984) first found flexible specialization has had steadily rising numbers of jobs. Amin (1989:20) points to a 28.5 percent growth in employment in firms with less than 99 employees between 1971 and 1981. 9 Rising Import Competition   FOR 1972FOR -1985FOR (millions in current dollars) 1972FOR 1973FOR 1974FOR 1975FOR 1976FOR 1977FOR 1978FOR !979 1980FOR 1981FOR 1982FOR 1983FOR 1984FOR 1985   The percent increases for e x ports are somewhat misleading when compared with percent increases for imports as the import figures are much larger than the export figures. Table 9 contains the numbers of establishments in the textile finishing SIC codes, and the number of establishments with employees of 1-19, 20-99, and over 100.

Significant Numbers of Small or Medium Sized Establishments
If one defines "small and medium sized firms" as those employing under 100 persons, then in the ten textile finishing SIC codes the following percentages comprise small or medium sized firms (

Textile Firms Seek Niches
In the face of increased competition, many te x tile firms are seeking market niches to insulate themselves fr om foreign competitors.
Here they have advantages in the way they use their e x isting machinery to produce a superior product to the foreign import. The

Conc1usion
This chapter has drawn conclusions about the occurrence of flexibility in te x tile finishing and has defined the t ype and degree of flexible specialization, if any, that is occurring. It has also characterized the textile finishing industry in New England in terms of the key areas that prevent it from being termed as flexibly specialized under Piore and Sabel's (1984) definition.

FINDINGS OF THE INDUSTRY INTERVIEWS
This chapter summarizes the industry interviews that were conducted at three Rhode Island firms. The names of the companies have been changed to maintain confidentiality.
At each interview, a predetermined format was used.
Questions pertained to the company's history, production process, location, employees, customer/supplier relations, and perceptions of the future of the industry and the impact of imports.

Company History
At the X Company the owner stated that there are only three to four vertically integrated companies left in the United States that still produce fabric from the beginning of the process through to the end (ie animal or plant through to the finished cloth). Company X is one of these.
Thirty three years ago the father of the current owner bought one third of the company. The original equipment is still used.
The ABC Company is one of the few totally integrated textile firms left in the country. Aside from buying wool from New Zealand and Australia, all the production to the finished product is carried out by the company.
The company was founded in 1919 in Worcester, MA. The 71 founders came from Bradford, England to the United States.
In Worcester they merged their textiles specialties of yarnmaking and fabric manufacturing.
In the beginning they conducted their yarn and fabric business in two separate buildings in the Worcester area.
As the business grew , a new plant was needed for the growing business.
In 1929  Today the company is one of the leading producers of fancy wool and wool-blended worsted fabrics.
The Y Company used to have a plant in the South, but this closed down in 1978-79. Today the New England firm is more integrated than it has ever been.

Production Process
At the X firm, fabrics for home furnishings and wall coverings as well as apparel, are processed. The company processes materials such as wool, rayon, polyester, cotton, silk, and flax.
The mill performs sponging, whereby material is treated to prevent shrinkage. The processes used include high temperature steaming and cold water shrinking. The owner commented that the firm has the advantage over larger, more integrated companies because this is all it does, and therefore the company can provide a specialized service.
The cold water washing process used at the firm is continuous. Other processes carried out in the mill however, are batch operations.
At the ABC Company, processes include top dying, package dying, and piece dying, weaving and spinning.
The ABC Company is the leading U.S. firm in its industry and is known for its high quality fabrics. The company has gravitated increasingly to the production of fabric for industrial sporting goods and away from fashion goods due to the violability of the latter 74 market. For example the firm makes parachutes for the government which is a reliable product line.
Diversification is the key to the firm's strategy to survival.

Location
The X Company was dr~wn to its present location because of its loading area and size.

Future of the Industry and Imports
The owner of X This decline in textile jobs has severe implications for the workforce, much of which is being made redundant.
Manufacturing jobs have traditionally paid a good wage to workers who are unskilled or semi-skilled. Manufacturing jobs are being replaced by service sector jobs paying lower wages. The implications of this provides a real problem which must be addressed.
Union membership in the textile industry has continued to decline since 1970 (Statistical Abstracts, 1990) .
Further, the reason why wages in many service sector jobs are low is because they are non-unionized jobs.

Rising Import Competition and Firm Size
The textile industry is suffering as a result of rising import competition. Rising import shares in the domestic market have led to a period of rethinking on the part of the The finishing industry is thus highly disintegrated. Many of these small firms operate on a narrow profit margin, which puts them in a vulnerable position with rising import competition. At the interview with Company Y, this reality was brought home when the Plant Supervisor offered the opinion that at least one of the five major textile finishers in the area will go out of business in the next five years as a direct result of increasing competition from abroad.

Lower Average Wages Relative to Manufacturing as a Whole
Textile wages continued to be lower than manufacturing as a whole, although as a percentage of manufacturing wages the figure rose from 69.7 percent in 1980 to 73.3 percent in

1989.
In the interview at Company X, the President stated the company paid its workers more if they knew more jobs.
This allowed the firm greater flexibility with its workforce. If one worker did not come in on a given day, then another would be very available to fill in.

Niche Production
In an attempt to meet competition from imports, the textile finishing industry in New England is increasingly retreating from markets threatened imports, and seeking out niche markets where U.S. firms have an edge over imports.

Use of Machinery
In many cases the industry is using old dedicated machinery in innovative, flexible ways to produce different products. For example, at Company Y, one machine was used to dye two very different types of cloth, one plain, and one with a camouflage pattern. The literature on flexible specialization begins to address the issue of dedicated machinery, but does not go far enough. Walker (1990) points to the source of confusion over the misplaced emphasis on scale of output instead of production methods.
He points out that one may get very long runs out of generalized machinery or short runs out of dedicated machinery, with corresponding effects on overall efficiency of labor and machine use, but the principles of mechanical improvement in each case are differe nt. Walker raises an important issue 85 but does not go far enough to emphasize the importance of looking more closely at dedicated machinery.

Policy Implications
It appear s inevitable that jobs will continue to be lost in te xtile finishing in New England.
Yet the e x istence of these jobs is c rucial to many families who have worked in New England te x t il es for generations. Indeed, the reason many of t hese wo r kers ' families came to the United States 87 was to work in the mills.
Loss of these jobs for semiskilled and unskilled workers has severe repercussions both for the workers and the state. Education programs may have to be provided to retrain workers, as most jobs today outside manufacturing require a minimum of a high school diploma. The minimum qualification stated by each of the three firms interviewed was a high school diploma, but each company stated that it did hire workers without a high school diploma when necessary. Indeed, at Company X, the President stated that what he needed most was more high school dropouts.
The ABC Company stated that 54 percent of its workforce was comprised of minorities. Many of these workers speak little, if any English. If these workers are to be integrated into service sector jobs requiring use of English language skills then language training will also need to be provided.

Implications for Community Planning
The implications of flexible specialization are profound for the profession of planning. In order to plan well one must have a thorough knowledge of the developments that shape the landscape. If the perception used to plan is one which is outdated, the plan that results will lead to failure. It is essential that all planners are aware of the impact that economic change, flexibly specialized or otherwise, has had and continues to have on the landscape. where geographic industrialization is planned, cities will not be planned either.

Planning
In conclusion, it may take a national economic development strategy to provide the kind of planned economic development needed for the country to be able to continue to be competitive in the new global economy. A federally instigated economic development program might require overriding many of the powers given to states under the separation of powers. Loss of some control to the federal government may be a high price for the states to pay, but slipping into minor economic status in a New World Order is a higher one. The issue must be addressed before the worst case scenario becomes reality.