Restructuring, Gender & Employment In Flux:

A Geography of Regional Change

in Cornwall Ontario


Abstract. For the past two decades, uneven development in general, and the changing structure of the labour market in particular, have figured prominently in geographic literature aimed at identifying and explaining the causes, consequences and possible trajectories of restructuring events in industrial countries. Where several studies have explored the national dimensions of aforenamed work force in Canada, less work has been done on the regional nature of the socioeconomic processes shaping gendered employment patterns. We argue that exploring the dimensions of 'feminisation' demands a critical reexamination of gender-neutral regional and economic geographies in an effort to understand how gendered divisions of labour and local identity formation are in flux. Critical feminist geographies of regional change are well positioned to expand theoretical debates concerning restructuring and, in turn, contribute to better informed regional redevelopment strategies. Our research in Cornwall, Ontario is used to illustrate why the inclusion of gender as a necessary analytical category is important to developing new geographies of regional change, and how restructuring in an 'old order 'manufacturing centre is linked to fluctuating gendered divisions of labour and concepts of class and identity formation.

1 Introduction - A Critical and Gendered Approach

This paper critically explores the dynamics of industrial restructuring by linking structural employment change within Canada to local processes of deindustrialization, employment adjustment and labour force feminisation in Cornwall, Ontario. We draw on feminist critique asap way of reformulating our understanding of regional change and the ways that it can be investigated. Specifically, claims are made for including 'gender' as a central analytical constructing geographies of restructuring by arguing that 'economic' production is fundamentally social and gendered in nature. We argue that alternative views on uneven regional development have much to offer theoretical economic geography - especially by challenging hegemonic views on the Economy' and opening up new spaces for social, spatial and economic diversity.

Drawing on notions of 'spatial divisions of labour' (Massey, 1984; 1994; 1995) our research highlights how spatially specific dynamics of economic change in Canada have differential impacts on women and men as they negotiate employment conditions in regional labour markets. Here, we look at notions of 'feminisation' and post-Fordist employment trends in the context of 'locality' as a way to critically explore the gender neutral content of analytical categories used in economic geography as well as the importance of place in restructuring theory.

1

2 Restructuring and Deindustrialisation - Starting Points for Debate

To the degree that we are witnessing a historical transition,still far from complete and in any case, like Fordism, bound to be partial in certain important respects, so we have encountered a series of theoretical dilemmas. Can we grasp theoretically the logic, if not the necessity, of the transition? (Harvey, 1990 F. 173).

The acts of theorising and explaining in geography, particularly in this era of widespread socioeconomic change, demands that we rethink 'essential' relations within, and 'ways of knowing' our social reality. This section revisits, from a critical feminist point of view, how the 'problems' and 'processes' of restructuring are defined and explained starting with a reevaluation of 'spatial divisions of labour' (Massey, 1984). While Massey's work has been instrumental in linking the uneven dynamics of industrial restructuring, gender and the social relations of production in the United Kingdom, we argue that: (a) gendered class-based analysis still presents fundamental theoretical limitations to feminist geographers - especially those doing research i n'other' socioeconomic contexts not adequately explained by dominant class and 'Fordist ' economic discourse and (b) locality research is fundamental to informing the conceptual complexity of regional labour market restructuring and 'feminised' employment patterns. Our review is not intensive, but instead focuses on how theorising in economic geography in general,and restructuring studies in particular, might better account for social and economic diversity by including gender as a necessary relation in social and spatial divisions of labour.

Although there is a growing and diverse feminist literature theorising key processes in industrial restructuring, we suggest that this work has remained marginalised within (masculine)geographic disciplines - including regional science and industrial geography - which remain largely gender-neutral. Emphasis added to suggest that masculine and male-dominated paradigms and disciplines in geography are 'real' science, whereas gender-aware, feminist paradigms are positioned to be 'less real'. It is often the case that feminist geographic studies tend to be included in the 'social' and 'cultural' elements of geographic literature (read: 'soft 'science' ) as opposed to the 'harder' (masculine) research areas of resource management,regional science, economic geography and industrial geography. Within industrial geography a swell, there is a tendency to focus on secondary manufacturing as the engine of 'productivity 'which assumes a highly restricted (masculine) definition of work and social formations.

As a 'place', Cornwall gives our problematising focus and context. It also allows us to isolate aspects of the restructuring process - in this case patterns of deindustrialisation and labour force feminisation - in situating Canadian experiences of socioeconomic change in wider theoretical debates. Further, locality provides a 'space' where global-local processes can be conceptualised and critically evaluated. In other words, it allows us to concentrate on 'process dynamics' (Harvey, 1990) and move beyond the limitations of spatial fetishism and class analysis in trying to better understand the present climate of accelerated social, economic andpolitical change.

From the outset it is necessary to state there are several (unresolved) theoretical tensions associated with our contribution to a gendered geography of restructuring. To begin with, our knowledge is partial and situated, and as such does not seek to make totalising claims about gender as an analytical construct or, about 'restructuring theory'. In fact later in the paper, some critical attention is directed towards the tendency of economic geography to use monolithic representations of socioeconomic processes based on hegemonic portrayals of what a restructuring experience 'is' that may have limited relevance to the cases of Canada and cornwall. Secondly, several strands of geographic and other literature are brought together in our research and as such, their treatment is by no means exhaustive. It is hoped however that our analysis will help highlight the underlying assumptions and silences about 'gender' in geographies of restructuring that will contribute to conceptualisations of regional change based on diversity and histories of functional specialisation. Finally, an obvious paradox exists without approach which is both critical of the content of selected 'traditional' analytical categories used in mainstream economic analysis (like 'class'), and at the same time, uses these very categories to explore possible alternatives to conceptualising regional change. We recognised this internal contradiction and use it to develop a reflexive approach to theorising. 'Critique' in this paper involves a renegotiation of key analytical constructs (social divisions of labour, class)within mainstream economic geography, rather than a call for 'throwing them away'. Despite the limitations cited here, it is hoped that these will be understood in the context of our overall objectives concerning critical discussions in economic and regional geographies of change.

specific in the use of the term 'gender neutral' in our discussion instead of the more politically charged 'gender blind' for reasons that will be developed more fully later in the paper. In short,the former type of analysis (reflected in the much contemporary regional geography) includes women by counting them in studies of employment change while at the same time remaining neutral about the politics of theorising and researching. The latter type of analysis neglects the gendered element of 'human' agency and assumes a masculine position in researching and theorising. We contend that both of these types of approaches to restructuring theory are theoretically limited in explaining the dynamics of regional change.

2.1 Rethinking 'the Social' in Spatial Divisions of Labour

In the past decade, economic geographers have learned much from social theorists and the idea that 'economic space' is the product of the differentiated and intersecting social relations of the economy (Massey, 1984, 1995; Massey, 1994). The work of Massey (1984, 1985,1995) and Massey and Meegan (1979, 1982, 1985) has been central in challenging geographers to rethink the key processes and relations involved in economic change, and therefore represents the starting point for our theoretical review of restructuring events.

Massey's principal contributions to developing a 'gendered geography of restructuring 'stem from her notions of 'spatial divisions of labour' and the fundamentally gendered nature of the 'social processes of production'. 'Spatial divisions of labour' has generated extensive debate which will not be reviewed here. However, particularly useful in the context of developing a 'new' regional geography is the recognition of: (1) patterns of economic change as linked to the historically-specific nature of economic cycles in each region, nation or locality; (2)mutually recursive relationships between corporations and the locale(s) in which they operate;and (3) the fundamentally social nature of production as comprised of sets of relations (between capital and labour) which are stretched out over space.

Drawing on Massey's framework, it is logical that the historical and spatial links creating uneven layers of development, and which underlie the concentration and restructuring of capital,interact with preexisting investment and production patterns. Although this structuralist approach has been criticised as a form of 'top down' analysis which ignores specific local experience, it

does allow Massey to evaluate 'types' of local structures in the context of change initiated at more aggregated economic levels. In our research context, it also allows us to establish initial 'globallocal' links and helps inform our theoretical work on Cornwall, Ontario as in some ways' typical' of changes to older manufacturing communities in Central Canada since the late 1980s.

Furthermore, in connecting economic, social and spatial processes, Massey's spatial divisions of labour provide a powerful conceptual basis for arguing that the social relations of production are not a spatial. In other words, the way our process under capitalism is not the sameeverywhere, but differs between 'regimes' in ways that reflect their particular histories and influence their futures (Johnston, Hauer and Hoekveld, 1990). Thus, adjustments to the nature and level of employment within in industry may mean plant closures for some localities, andprosperity for others.

The concept of uneven development, if it is to have any purchase on the structure and dynamics of economy and society more widely, must refer to more than the fact that there are more jobs in some places than others. Such measures...do noting themselves link that inequality to its causes in the deeper structures of the organisation of society. In order to do this,uneven development must be conceptualised in terms of the basic building blocks of (in this case) capitalist society(Massey, 1984 p. 86).

The pivotal importance of Massey's spatial divisions of labour for our research lies in its movement away from positivistic and predictive assumptions of economic behaviour concerning investment decisions and class relations by arguing that changes in industrial structure are butane aspect of the construction of local class characteristics and the social relations of production.Other important 'non-economic' factors in local variation include: the wider social structures of the community; alternations in the nature of consumption; changes to the spatial organisation of production; national ideology and political climate; and the restructuring of state finances. At the heart of Massey's 'spatial' framework therefore are complex social relations between capital,labour and the state.

At a general, descriptive level, statistical analysis of employment feminisation and

regional functional specialisation in Canada and Cornwall can be explained in terms of spatial divisions of labour (McKenna, 1996; Roberge and McKenna, 1995). For example, we compared unemployment rates, participation rates and occupational data by gender at a number of levels of statistical aggregation (national, provincial, regional and subregional) and discovered that consistently high rates of male and female unemployment in Cornwall, and declining overall rates of labour force participation over time (relative to other regions and aggregate trends)substantiated notions of Cornwall as a declining old order manufacturing centre, vulnerable to thenegative impacts of globalisation and liberalising trade. However, in terms of how Cornwall's regional labour market was 'feminising' and adjusting to drastically altered employment and social conditions, 'spatial divisions of labour' provide limited theoretical potential for explaining the particular dynamics of socioeconomic transformation The main reason for this stems from the lack of theoretical options offered by spatial division of labour for including social and economic processes that are not class-based, capitalist or readily included in globalisation discourses involving new international divisions of labour.

In fact, Massey offers few alternatives to the primacy of capital in initiating change in the location of economic activity. Examinations of the potential of the state as a policy maker,influences on the behaviour of capital, and elevating the importance of 'social concerns' in restructuring analyses (unemployment, labour dislocation, wage levels, the exploitation of marginal groups - women, youth, immigrants, - working conditions and community development programmes), are difficult to incorporate into restructuring frameworks based on spatial divisions of labour.

Part of the reason for this stems from the fact that Massey's 'building blocks' of society draw heavily on class analysis, and the dynamics of British social formation. We assert that the application of this framework cannot be wholly parachuted on to other industrial landscapes because the historical, cultural and social conditions of economic development will differ, for example, between Central Canada and Southeastern England. Concerns of this nature have been raised in particular by Fagan (1989) in work on the Australian steel industry and through Fincher's (1989) research into labour markets in specific areas of Melbourne, Australia.

Massey's emphasis on class relations as constituting social processes, also tends to reduce the various practices of other spheres of daily life to reflections of work place activities(Fincher, 1989). Not only does this assume that work place activities are always more important to individuals than 'other' spheres of involvement, but it tends to: (a) impose the dichotomy of 'public and private' activities and spaces on Massey's economic landscapes, and (b) obscure the importance of gender constructs in shaping individual behaviour in either sets of these activities.Indeed, the very term 'economic landscapes' implies a theoretical construction of the lived world as made up of 'economic relations' that constitute 'other relations'. Feminist geographers have argued that these types of dichotomies permeate mainstream gender-neutral (and gender-blind)restructuring discourse in a way which implicitly assigns a less important position to 'other'(usually social) relations (Gibson-Graham, 1992, 1995a, 1995b; Rose, 1995; McDowell, 1992a,1992b).

Following from Massey's work on the importance of 'place' in restructuring studies, was a renewed focus on localities research in the United Kingdom (The Lancaster Regionalism group, 1985; Bagguley et. al., 1990) which also explored the relationship between class politics, gender and the politics of place by creating images of classes defined first and foremost by the economy where individuals are located in the capitalist relations of production. Where locality studies further expanded notions of socio-historical diversity and the importance of region in the context of spatial divisions of labour, notions of economic class not necessarily relevant to Canada remain the centre of social theorising.

The use of 'class' in most restructuring analyses draws on dominant discursive conceptions of the economy as constituting society and therefore presents a narrow and static view of the subjective ways individuals and groups relate to one another. In so doing, it limits interpretations of the dynamics of restructuring in our markets, 'feminisation' and the social relations of production to those that are captured by capitalist-proletariat relations. However,rather than throw the notion of 'class' away, it is more useful to revise its analytical potential to represent the intersection of multiple social dimensions or processes - economic, gender, political and cultural (Gibson and Graham, 1992) - which allow us to view 'social processes' as fundamentally gendered, dynamic and mutually interactive.

' Other examples of dichotomies in economic geography that assume this theoretical construction (of weaker and stronger, or feminine and masculine) include: developed-underdeveloped regions; public-private spaces; productive reproductive activities; industrial-non industrial areas/countries; and urban-rural economies.

7

Further, this revised understanding of class might go some way to clarifying how themechanisms of change which shape the social relations of production and are fundamentally tied to regional socio-spatial organisation - a point not explored in Massey's work. Left at the level of conjecture in 'spatial divisions of labour' is how industrial decline and changing employment conditions will inevitably produce social and spatial 'effects'. This lack of clarity may stem frogman implicit acceptance of the new international division of labour (NIDL) in Massey's arguments which does not capture the complexity of changes in employment or investment patterns in advanced capitalist countries.

A&reenter to the NIDL as an explanatory framework overstates the role of technological change as that of simply mobilising productive capital from one location to another, an approach that does not facilitate analysis of regional labour markets or socio-spatial processes (Fagan,1989). Using examples from the restructuring of Australia's steel industry trough the 1970s and1980s, Fagan starts with the local scale in understanding the firm and its environment and the various restructuring processes which determine how investment occurs within localities. What we find theoretically instructive in terms of our Cornwall research, is the way Fagan attempts to contextualise the abstract laws concerning the behaviour of capital, labour market dynamics andthe role of the state in explaining how the benefits and costs of industrial change are shared unequally between regions.

However, common to Massey's and Fagan's theoretical approach is the assumption that women are as a reserve army of labour with universal characteristics (immobility, inexperience,conservatism, flexibility) that determine the options for women and men in paid work (Johnson,1989). What we have learned from our review of Massey's work is that the inclusion of gender in restructuring studies demands a better understanding of diversity in defining the processes of production (paid and unpaid) and in the dynamics of labour market formation and adjustment.Gender cannot simply be 'tacked on' to class analysis, nor should it be viewed as something that complicates the study of social processes. Rather, gender transects, structures and permeates all aspects of human agency and shapes inequality in the work force and definitions of 'work 'through, among other things, the reproduction of sexual divisions of labour. The following section explores in more detail how we develop 'gender' in the social processes of production and spatial divisions of labour.

3 Gender and Restructuring - Fluctuating Conditions of Employment

Since the late 1940s, several prevailing employment trends (inseparable from social and political change) in most Western industrialised countries have led to a greater theoretical and practical concern with gender in 'new' regional geography. First, there have been dramatic increases in the numbers of women going out to work. Second, there remains a significant separation between the types of work done by women and that done by men. Third, and representing a focal point of this research, there are intra- and inter-regional distributions of 'women's work' in industries and industrial sectors (this has become more evident since the1960s).

In relation to this last point, gendered divisions of labour in waged work are not only visible between industries, but also within them. It is often the case that women occupy the lowest skilled jobs which require little of ficial training and lower pay (Women and the geography Study Group of the IBG, 1984). In discussions of spatial divisions of labour,Massey (1984, 1994) places gender firmly in the restructuring literature by claiming that without an understanding of the gender roles which underlie the workings of society and formation of economic classes, we cannot hope to present a reasonable analysis of the spatial behaviour of women and men, nor of the institutions both dependent on and influencing that behaviour.Theorising about restructuring and 'place', she states that the "...geography of gender relations has been an important element in British industry's attempts to reorganise geographically; to restructure itself out of crisis" (Massey, 1994 p. 82).

Changes in the patterns of women's employment in Canada and elsewhere have taken place alongside increased state intervention into the economy and most recently in the context of a'rolled-back' welfare state and the rapid international redeployment of capital. In the past two decades, considerable empirical research on industrial location and regional inequality in the united Kingdom, Canada and the United States has pointed to the links between women's changing employment patterns and the wider processes of economic change.

[there is a] whole complex of social causes through which different parts of the population are allocated to different kinds of jobs within society as a whole. Without some grasp of the gendered-construction of the...labour market it is quite impossible to understand the recent locational contortions of...capital as it has tried to restructure, and relocate itself out of crisis. A feminist perspective, the, has a lot to contribute to better explanation within geography(Massey, 1984 p.l2).

For example, insights into job skill and gender differentiation have been particularly~constructive in linking the processes of restructuring (including technological and production change) to employment, gender and the uneven dynamics of development (Massey, 1984;massey and Meegan, 1982; McDowell, 1992a; Roos, 1985; Rowthorn, 1986; inter alia). Theclassification of jobs as 'unskilled', 'semi-skilled' or 'skilled' often shows a gender bias that can determine who gets laid off in the case of employment cutbacks, or, creates an emerging,choice]L elite within the production regime. Here we see the crucial intersection of employment,ender and the social]L processes of production. Women's perceived status as secondary income earners (working for 'pin money') and assumed role as primary reproductive workers can shape behaviours in the 'productive' sphere. Indeed, the main contributions of Massey's theoretical/work on restructuring and gender concern how the processes of production fundamentally socialnd display gendered patterns of participation which allow us to link regionally differentiated economic change to how gender influences the ways women and men experience that change.

However, a few theoretical limitations to Massey's work on geographies of gender and employment need to be addressed in the context of our work. First, we have already argued that Massey's discussion of gender relations often takes second place to her discussion of class relations. In terms of constituting gendered identities and constructing categories of 'work',~Massey's ana]Lysis implies that people only become 'feminine' or 'masculine' within the narrow context of capitalist productive relations, which is clearly not the case. Second, in an otherwise detailed anatomy of industrial restructuring, the analytical value of 'gender relations' in making concrete connections between productive life and social;L life in restructuring communities is not explored. Theoretically, there is no sense that gender and social process are dynamic, and that both respond to, and shape, the economic and social processes of regional change. This may be~wing to an unstated acceptance of dominant discourses about the 'Economy' which marginalises (or excludes) from analysis alternatives to capitalist social relations. Further, dominant discursive notions of 'economy' which imply that industrial society is a centred singularity. We argue against this type of 'organic functionalism' and contend that there is more room for explicitly gendered approaches to restructuring theory that include diversity of economic and social forms which function 'inter effectively' (Gibson-Graham, 1995a) with capitalism.

In this sense, we suggest that Massey's the theorising is 'gender neutral' because gender is presented as a monolithic socio-analytical category which shapes the economic imperatives ofrestructuring theory but does not challenge the way those economic imperatives are conceptualised]Lised or explained. Based on our research, we have found that it is difficult to reconcile the social, spatial and economic diversity associated with restructuring experiences with totalising representations of capital, class and gender. This last point is particularly relevant to Canadian restructuring experiences which, it has been argued (Jenson, 1989), are not class-based or the product of more 'classic' forms of Fordism .

Assuming that different localities may be characterised by a certain industrial mix and labour market, it is logical to suppose that gendered relations and identities will in some way be shaped by these (among other historical) conditions. The importance of a locality in reproducing the social relations of production is not aligned to its location per se, as much as its existence as a setting for capitalist development. Therefore, when linking gender, employment and production in discussions of differential reproduction over space greater attention needs to be given to the role of locality in the articulation, juxtaposition and intersection of multiple and gendered social relations. That these relations will be charged with internal tensions and contradictions is part of the uneven nature of restructuring. Therefore, we need to clarify our understanding of gender constructs not as 'things' so much as processes wrapped up in the dynamics of regional change.Here, it may prove useful tO explore arguments made by Thrift and Olds (1996), for a more 'polycentric'' approach to economic geography in developing gendered perspectives to what is conventionally regarded as 'the economic'.

In arguing for explicitly gendered geographies of employment and restructuring,conceptualising locality as the focus of identity formation and gendered divisions of labour is both a crucial a difficult task. Theorising about identity formation demands some knowledge of how material, lived relations translate into individual and collective political identities in flux, and how these in turn help reinforce, challenge or transform lived relations of power (Chouinard,1996). In the context of gendering 'necessary relations' and revising our understanding of 'the processes of production' in economic geography, we argue that new discursive spaces concerning diversity, conflict and partial positioning are key tasks. For instance, Chouinard's discussion of processes as 'in place' and 'in flux' as they give rise to conflictual and fragmented ways of being and becoming in patriarchal, capitalist societies was helpful in broadening our theoretical understanding of the social processes of production as representing capitalist andnon-capitalist relations.

Gender and class formation within local states involve processes of conflict, accommodation, co-optation, and transformation, which are simultaneously conceptual,discursive, practical and material, and hence subjective as well as objective (Chouinard, 1996 p.1485).

An expanded approach to thinking about the 'social processes of production' is also instrumental in explaining how individuals may simultaneously occupy multiple class position sand have complex gendered identities in their daily (paid and unpaid) labour process.

Considerable feminist geographic literature exists that investigates the ways gender relations are reproduced in the daily activities of people's lives - both in the productive and reproductive spheres (Mackenzie, 1989; McDowell, 1991a, 1991b; Women and the Geography study Group of the 1BG, 1984; Massey, 1994; Rose, 1985, 1995; Roos, 1985; Saunder, 1989;McDowell 1992a, Pratt and Hanson, 1995; Gibson and Graham, 1992 inter alia) . Work by mackenzie (1989), Mackenzie and Rose (1983), Mackenzie and Kobayashi (1989) and Milroy and Wismer (1994), has demonstrated that the manner in which women use their time, resources and functional environments at individual levels plays an integral role in shaping and reproducing the social relations of production and organisation at more aggregate levels:

Seeing the articulation of gender relations with environmental relations means focusing on the activities involved in the production of goods and services, with reference to or simultaneously with those involved in the biological and social reproduction of people (Mackenzie,1989 p. 55).

12

Further to clarifying how to view gender as a dynamic construct in the processes of industrial restructuring, we define gender as an analytical category which regulates and structures identity around sexuality and the social defined differences between women and men.Knowledge that gender constructs change over time, are inclusive of various productive regimes(not only capitalist) and permeate all aspects of social life helps us broaden our theoretical understanding of how 'economic' process are regionally differentiated.

This is important for a couple of reasons. First, it expands our view of the labour process and allows us to consider how different types of work are socially constructed. It facilitates research into the arena of (capitalist) production and the work force as inclusive of activities (not necessarily capitalist) and unpaid labour our that shape social formations. In turn, thisallows us to challenge the validity of masculine polemics in economic theorising and notions of separate and static 'spheres' {ie. private vs. public) in gender-neutral restructuring literature. It also helps us move beyond the confines of reductionist class-based analysis in reformulating the connections between the necessary social and economic relations of a geography of restructuring.

Building on this first point is a second issue concerning the conditions under which women and men enter the labour market and how they see themselves as gendered and classed subjects in the labour processes. Understanding how people define their own identities and place themselves in the social hierarchy in relation to others helps us reflect on the subjective and objective elements of class and identity formation. For example, much has been written about the fact that women end up in 'job ghettos' because their assumed dual productive/reproductive responsibilities restrict the types of waged labour they can perform and the locations they find accessible (Pratt and Hanson, 1995) .

Consistent with recent feminist literature (Hanson and Pratt, 1995, 1994; Rose, 1995;McDowell, 1991, 1992a, 1992b; Gibson and Graham, 1992; Gibson-Graham, 1995a, 1995b),our research experience indicates that explanations of feminised employment patterns based largely on the 'dual role' argument offers only partial insight into the dynamics of gendered divisions of labour - particularly in small to medium sized communities with relatively little industrial diversity that have undergone a period of profound restructuring of the productive

13

base. The reasons for this are twofold. First, they draw on dominant discourses of The economy and reinforce masculine theoretical constructions of the lived world as made up of economic relations that constitute other relations. And second, they tend to exclude men's restructuring labour identities and roles from theoretical arguments about changing patterns of work (paid and unpaid). The following section explores these concerns in more detail.

4 Feminisation - Concept and Contradictions

Increasingly, 'new' regional geographies of development and employment have begun to look at labour force feminisation and how is it connected to the feminisation of work (McDowell,1991, 1992a, 1992b; MacDonald, 1994; MacFarland, 1993; Cohen, 1991; Mackenzie, 1989,1988; Andrew and Milroy, 1988; Massey 1994; 1989; 1984; Massey and Meegan 1985, 1982,inter alia). Given the limitations of this paper, simplified explanations of these terms are used to clarify how they may relate to restructuring debates. In general, labour force feminisation refers to the increasing participation of women in the paid work force relative to men in the paid workforce. This notion is sometimes used interchangeably (which causes confusion) with references to 'the feminisation of work'. Here, feminine attributes and status are assigned to work that i slower-wage, part time or non-standard arrangement (flex time, piece work, contract, on-callwork), usually non-unionised, routine and broadly associated with post-Fordist and/or 'globalised' production regimes (see for example McDowell, 1991).

A brief overview of Canadian labour market trends highlights the complexities of 'feminisation' - both in terms of overall employment trends and the type of work performed by women and men. To begin with, despite the feminisation of the work force, census data suggest that greater relative female participation in the lab ~labour force over the past two and a half decades has not necessarily led to their increased representation in all sectors of the economy. For instance, in 1971, 85.7% of women in Canada were employed in 131 of 484 occupations recorded by labour force survey data (Hughes, 1990). Two thirds of these jobs fell into five categories: clerical, machining, services, processing and health. The 353 occupations dominated by men were much more diversified and employed just over 80% of the work force. A comparison with census data from 1986 and 1991 shows surprising stability in the number of traditional and non-traditional occupations for women and men. According to the 1991 Survey of Work Arrangements, work patterns and employment opportunities for women continue to be strongly influenced by the presence and age of children (Statistics Canada, 1991). The Survey also showed that the location (for example, 'home-based' or 'external' work place), and quality(part-time, full-time) of employment is strongly influenced by gender and family status. This may in part be owing to the way in which questions in the Survev were structured (around a limited range of 'gender issues') but likely reflects prevailing patterns relating to female and male employment.

Numerous studies on women in the Canadian work force (Marshall, 1994; Hughes,1990; Chawla, 1990; Devereaux and Lindsay, 1993; Statistics Canada Labour Force Survey data, 1961 - 1991) point to patterns of gender difference describing women's labour force participation since the early 1970s: (1) women earn less money than men in the same or similar occupations (although the gap is decreasing - see Table 1 below); (2) more women than men a reemployed in part-time, service sector jobs; (3) in family situations with two earners, women's income is becoming proportionately more important to family income; and, (4) women with young children continue to have the lowest labour participation rates.

Table 1 Average Annual Earnings by Occupation and Sex,

Canada 1992

source: Statistics Canada, 1994 p. 34

This indicates that the Canadian labour force is not feminising in a way which equates with women improving their overall labour position through, for example,expanded participation in full-time employment in 'non-traditional' occupations. In turn, this raises further questions about the nature, conditions and availability of jobs in the restructuring labour market. Are 'good jobs' (unionised, well-paid, full-time) disappearing or not available to new labour force entrants (youth, women, immigrants)? If these 'good jobs' - traditionally dominated by men - are disappearing how does our concern with gender inform on mens' experiences with employment restructuring? What other factors are structuring the labour market to reproduce sexual divisions of labour? Or, is it more accurate to reflect on a 'core' of good jobs and a large majority of (less desirable) 'periphery' jobs as McDowell (1992a) suggests?

Feminist economists argue that Canada's renewed commitment to trade liberalisation and export-led growth involve several 'gendered assumptions' which have the effect of deepening,rather than reducing, gendered divisions of labour (Armstrong, 1993; Brodie, 1994; Skrypnek and Fast, 1993; McFarland, 1993; Cohen, 1991; MacDonald, 1994). Since the early 1980s, and through the establishment of the FTA (1984), NAFTA (1993) and the continuing General agreement on Trade and Tariffs (GATT) Canada's macro-economic policy has encouraged a type of shock treatment aimed at improving efficiency and competitiveness. As a part of this drive for efficiency, government supply management schemes (marketing boards, protective tariffs, subsidy programmes) and social welfare spending have been reduced in light of arguments that they 'distort' market functions in an era of globalised competitiveness. The inference is that market distortions are the root cause of problematic economic performance relating to productivity, competitiveness and employment. In terms of labour market adjustment, and the effects of economic liberalisation on women the following assumptions a remade: (1) that the demand for female labour will increase through competition with low-wage countries; (2) as low wage employment spreads throughout the economy, the labour force will be 'feminised' as women are substituted for higher-wage males; and (3) men's jobs will be 'deskilled' and converted into the kinds of jobs traditionally done by women (Cohen, 1991;Armstrong, 1993; MacDonald, 1994).

The reality of the labour market however, does not support these assumptions. In 1993,the median total income of women and men in Canada fell 3.2% to $18,000 from the previous year (Statistics Canada, 1995). The decline for women was 3.5%, compared to 2.4% for men.Between 1989 and 1993 (adjusting for inflation), median employment income for both sexes dropped 6.1% - men's income dropping slightly more than women's income.2 Analyses of the wage and skill structure of Canada's labour market in the past decade indicates it is becoming increasingly polarised in terms of gender and income divisions (Drache, 1993; Mahon, 1993;Myles; 1993).

2 In a bulletin released by Statistics Canada in December 1995, the government organisation indicated that for the first time since 1967, the 'wage gap' separating men and women increased in 1994 (cited in the Ottawa Citizen,Wednesday December 20, 1995, A4). For every dollar earned by men in 1994, women earned only 70 cents which was down from 72 cents in 1993. The wider earnings gap was thought to be a result of the largest annual advance in men's earnings in almost two decades, as opposed to a decrease in women's earnings.

16

One way to explain the combined phenomena of a feminised work force and the feminisation of work is through a process called 'I harmonising down' (Armstrong, 1993). This accounts for, among other things, greater statistical regularity between male and female participation rates, and declining male incomes. In the face of a shrinking labour market, and fewer full-time 'good' jobs, it is suggest that men are increasingly moving into areas where women have traditionally been employed (lower skill, part time, service sector). While most of the 'good' (full-time, well-paid) jobs are still held by men, larger numbers of male labourers are adjusting down in terms of their status and pay to take traditionally female-dominated work. The implication is that restructuring and globalisation has created more 'women's work' in the labour market for which women and men are competing.

While there is a certain descriptive utility associated with the concept of harmonising down, advocates of this argument fall into the same macro-statistical trap as those gender neutral economists with whom they find fault. To start with, advocates of the harmonising down argument display little analytical concern with the importance of regional variation in how social relations and gendered identities might be linked to functional specialisation. Further,'harmonising down' focuses on wide scale female experiences of employment (making homogenising assumptions about 'female labour') and leaves unexplored the complexities of male employment change which presumably contribute to the half of the process. And finally,we content that a gendered theoretical reformulation of restructuring employment patterns should account for diversity of socioeconomic forms and processes over space. Despite the feminist critique of neo-liberal macro-economic policies, the feminist economists exploring the dynamics of harmonising down seem to accept patriarchal discursive positions regarding 'the economy' (as constituting of paid work) and class/identity formation (as capitalist processes operating in the 'formal' economy). A brief look at some post-Fordist literature may help clarify some of the mechanisms shaping 'feminisation' by looking at how jobs are being divided into a core/periphery relationship.

4.1 Feminisation in a post-Fordist Labour Market

Numerous variants of post-Fordist theory exist (the most influential being the French regulationist school and the flexible specialisation approach) which share assumptions about women's secondary position in the labour market (McDowell, 1991a, 1991b, 1992; Duncan,1992; Myles 1993; Mahon, 1993; Drache, 1993). It is beyond the scope of this section to provide an extensive review of this literature, however, post-Fordist insights into how emerging production regimes exacerbate existing rigidities in gendered wage and skill structures contributeto understandings of heightened polarisation between employed individuals.

Extensive feminist literature exists on: (1) how women's entry into the labour force was not on equal terms with men owing 'traditional' ideas about gender roles, and; (2) how women's paid labour is neither properly valued, or recognised in terms of its contribution to social and economic structures.

Even in the newly expanding public sector, women were concentrated at the bottom of the occupational hierarchy,trapped in the ghettos of 'female' jobs where caring and servicing were seen as desirable but poorly rewarded attributes. In the manufacturing sector women were also concentrated in less skilled and low-paid jobs...[P]art time employment with less security and fewer occupational rights and benefits was a key strategy in expanding female employment. Thus traditional ideas about gender roles combined with labour market regulation created a labour force that was highly segmented by gender (McDowell, 1992a p.181).

In contrast, the post-Fordist regime is char ~characterised by new types of industrial organisation as capital restructures itself to take advantage of international markets and more profitable production locations. The development of 'just-in-time' production and delivery systems, greater industrial decentralization and job 'de-skilling' (Massey, 1982, 1984) have created an entirely new labour market demand and dynamic. With greater capital mobility and production flexibility, comes the demand for new 'forms' of labour including temporary work,discontinuous patterns of employment, part-time contracts and fewer fringe benefits (Mackenzie,1989; McDowell, 1991a, 1992). Given their unequal conditions of entry into the labour force during the Fordist regime, there is a certain continuity in women's labour position in the current period of economic transformation. McDowell (1992a) suggests that changing industrial labour requirements are creating a 'core' of skilled, permanent workers and a 'periphery' of less-skilled,temporary workers. The minority of women in the 'core' are becoming increasingly polarised from the much larger numbers of women in the 'periphery'.

McDowell's assertion that post-Fordism has reinforced, not restructured, women's exploitation by creating a more complex and divisive employment structure is important for a number of reasons. First, it exposes the lack of debate in the post-Fordist thesis over thecontinuity of women's labour patterns (McDowell, 1992). Second, it lays to rest once and for all the idea that women's entry into waged labour (or 'feminisation' of the labour force)automatically has an emancipatory effect on women. And third, it highlights the deepening contradiction between economic and social restructuring - particularly in terms of women's labour exploitation in the productive and reproductive spheres.

However, our positions as Canadian researchers, highlight certain limitations with McDowell's post-Fordist analysis which expose implicit 'universal' assumptions associated with dominant (British and American) restructuring discourses. Generalisations are made about gender and social formation that draw (implicitly) on Britain's class structure and experiences of fordist production. Drawing on Jenson's (1989) work, we suggest that the transition from fordism to post-Fordism is specific to nation and place, often partial at best and as a result indeed of theoretical 'contextualisation'. It does not necessarily follow that industrial and labour market restructuring in Canada is best explained using unexamined categories of socio-spatial and economic relations (including gender and class). For instance, the term 'permeable'(Jenson, 1988) has been used to characterise Canada's Fordist order, in the sense that much of its resource and manufacturing sector has been infiltrated by American capital and thus is strongly influenced by policies of the American state and multinationals. This term was developed to describe how Canada's post war economic development history differed from other advanced industrial societies. Jenson argued that the crisis of Fordism in Canada (traced to the 1970s) is a crisis of the political arrangements of federalism (as it shaped economic policy after 1945) more than it is a crisis of the party system. While the dynamics of 'permeability' of the Canadian economy are difficult to characterise, few would argue that reliance on foreign investment and regionally-based disputes over the control of natural resource revenue have compounded problems of fragmenting collective identities based on language, culture, political power and economic change. Jenson (1989) suggests that since 1945, Canadian federal governments have pursued a type of
Fordism that "was designed domestically but always within eye to the continental economy" (p. 78). While we might dispute how much of the attention to the continental economy was a matter of choice or necessity (given the heavily branch-plant nature of our manufacturing sector for example) it is clear that the permeable nature of the canadian economy has meant the development of tighter and tighter links with the American economy. That the Fordist paradigm in Canada was not organised around class-based collective identities (Jenson, 1989) has led to some important questions about how we view the processes of social formation and the restructuring dynamics of gendered divisions of labour in different places . It may prove instructive to reevaluate class concepts in locality research as a means of describing the gendered dynamics of regional labour market restructuring in Canada.

Finally, there is little (or no) indication in McDowell's theorising of the spatial differentiation of the social relations of production, or, acknowledgement that the content and gendered characteristics of core-periphery jobs may reflect regional functional specialisation.This is a potentially important oversight given that regions with relatively undiversified industrial structures (for example, historically, Cornwall's economy has been characterised by textiles,apparel and pulp manufacturing) with have different 'job-wage-skill-gender' structures than industrially diverse areas with a range of activities including manufacturing, research and development, high technology and information based industries.

As a way to work through the strengths and limitations of our theoretical perspectives on gender, restructuring and employment, the following sections discuss some findings from our research in Cornwall, Ontario which begins with a brief overview of employment restructuring in canada.

5 Placing Our Research - Restructuring in Cornwall

Profound socioeconomic restructuring in Canada (particularly throughout the 1980s and early 1990s) has lead to considerable research among public policy makers and academics concerning the 'causes' of economic change and implications of labour force feminisation. The working population in Canada has shifted dramatically, from 60% of employed people working in the goods sector in the late 1940s to over 70% presently working in service industries. and, while the service sector is far from recession proof, it had the lowest rates of female and male unemployment during the economic downturn of the late 1980s and early 1990s (Statistics Canada, 1995). The sector hardest hit by aggregate job loss between 1989 and 1994 was that of goods-producing. Referring to the unemployment rends by industry between 1981 and 1993shown in Table 2, in 1993 female and male unemployment rates in goods producing were 12.7%and 14.0% respectively, compared with about 9% for both sexes in the service sector (Statistics canada, 1994). In fact, throughout the 1980s, nearly all of the net job creation took place in the service sector (Economic Council of Canada, 1990) and similar trends are projected for thel990s.

Table 2 Unemployment Rates by Industry in Canada, 1981-1993source: Statistics Canada, 1994 p. 27

A significant part of this growth has been caused by increasing female participation in the work force. National statistics indicate that in the past twenty years alone, female participation rates have increased from 44% in 1975 to 58% in 1994 (Statistics Canada, 1994). In addition,unemployment has risen less among women in recent years than it has for men. Between 1989and 1993, the number of unemployed women rose 42% compared with 64% among men(Statistics Canada, 1994). According to some researchers, the links between rising male structural unemployment, work force feminisation and the feminisation of work in Canada are wrapped up in the processes of freer trade and global competitiveness which has contributed to deindustrialisation in certain vulnerable sectors of the economy.

The empirics of Canada's service economy with respect to wages and job skills are now fairly well understood. There is indeed evidence of growing dualization in the Canadian labour market, but this is not a result of the shift to services as such.Rather it is an economy-wide trend that represents one of several 'branching points' in the transition to a post industrial economy (Myles, 1993 p. 351).

Although we have argued against simplistic theoretical assumptions concerning the transition in Canada from Fordism to post-Fordism i, it is the case that significant structural changes have occurred (and continue to do so) to the manufacturing base, particularly in parts of Central Canada.

The problem of 'runaway' manufacturing plants and deindustrialisation has become a major economic concern in canada. Critics of the FTA point to a recent sharp decline in manufacturing employment as evidence that unregulated free trade with the U.S. has already sparked a turn around in the direction of Canada's economic development -undermining our more diversified industrial sectors, and reinforcing our reliance on resource extraction and other 'peripheral' or 'dependent' types of employment. In may high-profile instances, the job loss in Canada resulted from a decision by employers to transfer production to alternative facilities in the U.S., often located in the deep South...To be sure, the dire statistics of job loss and disinvestmentmust be interpreted with caution...Nevertheless, the scale of the job loss and the accompanying hardship in Canada's industrial heartland, together with the numerous reported cases of relocation to the U.S., suggest that this is a matter for considerable concern (Stanford, 1991 p. 3-4).

A variety of sources (government, public interest groups, industry) have documented job loss in Canada since the recession in the early 1980s, most particularly in the context of the FTA with the United States (Canadian Centre for Policy Alternatives, 1991; Statistics Canada's canadian Economic Observer; Survey of Employment; Labour Force Review; and, Perspectives,inter alia). The magnitude of job loss, and rapidity of employment change has received much attention. For example, in studies of the short-run impacts of economic restructuring thought toke linked to the FTA, the Canadian Economic Observer claimed that between January 1989 and february 1991, 289,000 jobs in manufacturing were lost, whereas the Survey of Employment reported that up to 435,000 jobs disappeared over roughly the same period.

Following Stanford's advice (above) to interpret job loss with caution, it is generally agreed upon that Canada is losing large numbers of manufacturing jobs in particular, and 'spinoff' jobs as a consequence. Referring to Table 3, we see that over a period of two decades the distribution of employment by industry has been away from the goods producing sector and towards the service sector. Examining the trends by sex, women's representation in services has increased by about four and a half per cent, and dropped approximately the same amount in goods producing. Similar trends are apparent for men, but the magnitudes are larger - an increase of about six percent in service jobs and a similar decrease in goods producing. The data also reveal that the relative representation of women (considered as a per cent of total sectoral employment) has increased by eight percent in service industries and decreased by just over four percent in goods producing industries.

Table 3 Distribution of Employment By Industry and Sex, Canada1975- 1993

source: Statistics Canada (1994) p. 19

Whether the continuing trends of decreased participation in goods producing jobs is a permanent result of continental free trade, the recessions of the early 1980s and early 1990s, or global restructuring trends, is not entirely clear. Likely, it is a combination of all three sets of factors. It is the scale of job loss and the accompanying hardship in Canada's industrial heartland, which helps focus attention on the complex problems of 'restructuring' and rapid deindustrialisation in places like Cornwall.

The fact that Canada lost a much larger share of its manufacturing jobs in the post-FTA recession than did the United States (where manufacturing employment fell by 7% between january 1989 and February 1991, compared to Canada's 15%), suggests that the economic downturn of the early 1990s was not solely to blame for Canada's struggling manufacturing performance and labour market restructuring (Stanford, 1991). The issues posed by capital mobility, wage competition and inter-regional competition (within Canada; between Canadian regions and areas of the continental trade zone; and, in the context of Canada's global trade relations) need to be examined more closely in ongoing attempts to define the impacts of restructuring on labour markets, new work relations and the issue of 'feminisation'. It is important to restate the fact that the feminisation of labour is not simply a 'women's problem'.Rather it is indicative of a profound transformation and polarisation of the labour market (Drache,1993; Mahon, 1993; McDowell, 1992; Myles, 1993; Armstrong, 1993; inter alia) . The extent and scope of the changes argues against those who would claim that Canada's present unstable economic performance is indicative of a period of 'adjustment'. An examination of various aspects of restructuring and the labour experience demands an understanding of economic 'regions' in Canada.

5.1 Restructuring in Cornwall

Communities like Cornwall, in Eastern Ontario are in some respects 'typical' of classic deindustrialisation scenarios in vulnerable industries. Situated on the north shore of the St.Lawrence River (about 100 km south-east of Ottawa and 115 km west of Montreal), Cornwall has a population of approximately 47,000. It is the County seat of the United Counties of foSterling, Dundas and Glengarry - a larger, primarily agricultural regional which as population in excess of 105,000 people. Located on the United States border, it provided an excellent sitefor American branch plants seeking to avoid tariffs by producing in Canada for Canadian and international markets. Paper products, textiles and apparel and chemicals have been the traditional mainstays of the City's industrial base.

Figure 1 Cornwall in Its Regional Contextsource: City of Cornwall, Economic Development Office (1995)

However, with the implementation of free trade with the United States Cornwall lost its 'old order' locational advantage on the international border virtually overnight. In the period between June of 1989 and January of 1992, the City of Cornwall (about 47,000 inhabitants) lost over 2,500 jobs in manufacturing alone. Of the five major plant shut downs in this period(where 200 or more employees lost their jobs), four of them involved textile or apparelproduction3 which had been identified as old order industrial sectors highly vulnerable to international competition (Porter, 1991; Economic Council of Canada, 1978).

By March of 1993, the City of Cornwall deemed their economic position to be so tenuous that they applied to the Provincial Government of Ontario for Emergency Crisis Funding. At that time, the unemployment rate was estimated to be anywhere from 26% to 40% and over 50% of the adult population was receiving some form of social assistance (City of Cornwall, 1993).

3 These five plants included: Courtaulds Textiles (360 employees, closed December 1992); Dominion Textiles(350 employees, closed June 1992); Amoco Fabrics (200 employees, closed July 1990); Courtaulds Films (240employees, closed October 1990) and Combustion Engineering (200 employees, closed October 1990).

24

Social welfare, employment and counselling agencies in the City4 drew explicit connections between restructured labour force conditions and increased incidence of family violence, poor nutrition, school drop out rates, stress, teenage pregnancies and substance abuse (Figure 2).

Figure 2 Restructuring in Cornwall - Linking Economic, Social and Community Changesource: Proposal for Emergency Crisis Funding, City of Cornwall (1993).

Clearly Cornwall was experiencing the complex structural impacts of the changing conditions of international capital investment in regions of Central Canada's traditional manufacturing core. In terms of developing gendered geographies of employment and some understanding of 'feminisation' in this old order manufacturing centre, life history interviews with displaced male and female workers were instructive in: (1) highlighting complex and fluctuating gendered divisions of labour in the context of job loss or job change; (2) seeing how people defined their own gendered identities and 'class' given changing structural conditions of paid and unpaid work. Here, we gain some insight into the subtleties of 'feminised' work andemployment patterns.

6 Peter - Class, Gender and Work Identities in Flux

We have argued throughout the paper that analytical categories (principally class and gender) need to be explored in the context of place and theoretical constructions of dominant (and masculine) economic discourse. Our interviews with 15 displaced workers in Cornwall (8 men and 7 women who had worked in secondary manufacturing plants that were shut down between1989 and 1992) helped us contextualise points of contact between the social, spatial and economic processes shaping restructuring events in positioning gender in contemporary restructuring debates. This section examine's aspects of Peter's5 work and life experiences which reflect on various aspects of the restructuring process and interpretations of feminised employment patterns.

At the time of the interview (June 1995), Peter was 39 years old and had lived inCornwall all his life 4 Interviews (June to September 1995) with various agencies in Cornwall and unpublished statistical records 19891995.

s All the interviews were anonymous and fictitious names given to each interview participant.

25

. One of seven children, his parents and his partner's parents are also lifetime residents of Cornwall. Like many of his generation in Cornwall, Peter went to secondary school until the mandatory age of 16 when he 'dropped out' and started working atodd jobs until he secured a place in Courtauld's textile fibre mill. At that time, a completed secondary school diploma was not a necessary condition of being hired on any shop floor inCornwall (this changed for most companies in the early 1990s) as, according to Peter, 'jobs were there for the taking'. Concerning the plant shut down in 1992, Peter began hearing strong rumours about closure in the late 1980s when Canada entered into Free Trade (FTA) talks with the United States. However, he did not take these rumours seriously as "the company was always giving us pep talks - you know about everything going good and profits and all. So, we bought a house and my wife was pregnant with our third child" (Peter, June 1995). Peter's reaction to the closure announcement (not six months after he had moved into his new house)was therefore one of 'extreme shock'. At the time of the closure, Courtauld's set up a Labour adjustment Board which was necessary by law, to help workers in their 'transition time' after loss of employment. Typical of other interview participant's views of this process, Peter claimed, those things were shit. I had no use for them at all.Adjustment to what, anyhow? All the plants were shutting down in Cornwall. There were no damn jobs anywhere no matter who you were (Peter, June 1995).

With no job prospects in manufacturing, no high school education a new mortgage and a young family, Peter faced profound conditions of change on a social, economic, personal and cultural level. Having lost his job at Courtauld's after working 17 years on an all-male shift,Peter increased his unpaid workload at home while his wife took on more hours at a city operated child care facility to compensate for lower family income. Being the parent who stayed at home was a 'new' role for Peter as before the mill shut down, Peter stated that he did little domestic labour and often worked at Courtaulds on weekends.

In the first year after losing his factory job, Peter got up at 4 o'clock in the morning to deliver newspapers and then worked as a night clerk after balancing child care and sleep requirements during the day. Most of the time, sleep and any leisure time were compromised.When his wife was working out of the home, Peter performed unpaid domestic labour, looked after his three children and occasionally the neighbor's children. He was usually responsiblefor starting supper by the time his wife got home (around 4:30 in the afternoon), and then would go to the local gym to work out, return in time for dinner, then study in the evenings to upgrade his secondary school qualifications so that he could attend community college. After several months of looking for jobs in Cornwall, Peter realized that an upgraded education was essential.In the second year of his 'unemployment' (1993) he commuted daily to community college in ottawa (120 km away) to attend chef courses. Training for work in the service sector was a conscious choice for Peter, as he was convinced that Cornwall's days as a manufacturing centre were over with the implementation of Free Trade with the United States (implemented in 1989).While attending college courses during the day, evenings were spent assisting with domestic work and studying. Peter reflects on this time as the most stressful of his life. The routine demands of post-secondary education were difficult as was the knowledge that he was the oldestin his class and, as a male, not in the majority for his most of his classes. He was also balancing family care tasks, trying to keep his marriage together and make large mortgage payments. In general, Peter's life became much more fluid and there were fewer boundaries surrounding different elements of his life and how he defined 'work'. Everything became 'work' - family,marriage, education and struggling to find time for himself. Peter was redefining himself given different and constantly fluctuating identities as a stay at home parent, dependent husband,student and unemployed person.

When I was at Courtaulds, I used to have the mentality that things didn't change. But now, with all I've been through -my thinking is that anything can happen! (Peter, June1995).

Having gained his chef qualifications in 1994 (the third year after Courtaulds shut down),Peter juggled three part time cooking jobs for almost a year before he found more steady work at a senior's residence (30 hours per week) in Cornwall. His work place is now all-female (Peter being the only exception) which he found 'a bit weird at first'. And, given his part-time work status, Peter remains the primary after school care giver for his children and often, his neighbour's children. He continues to do most of his family's food preparation, performs housecleaning tasks and continues to work out at the local gym several times per week. Even though he is earning an income at present, Peter stated that cutting back on his portion of the housework was never an issue. At the time of the interview, Peter was also on the negotiating committee for the Union for workers at the senior's residence.

In terms of class identity, Peter's employment history as a mill hand and present job as a cook places him in the working class (proletariat). As a union representative, his political interests are also aligned with 'the working classes'. His class position in Cornwall however, isnot that clear cut. Peter perceives his cooking job as preferable to his factory job given the different quality of his work environment and overall life style. His hours are more regular, his weekends are free to spend with his family and he feels much more relaxed. Despite the fact that he earns about $20,000 (Cnd) per year less as a chef than when he worked in the mill, Peter's efforts to improve his education and skills greatly increased his self-confidence and status among his family and friends. Gaining a post-secondary education combined with changing employment sectors from manufacturing (which was 'dirty and hot') to services (which is a'clean' environment) has, in Peter's estimation, also has a positive impact on his employment image. In fact, Peter suggests that he is now 'better off' in terms of his overall quality of life and contends that a return to millwork (if an opportunity arose) would be a low priority.

According to Peter and our other interview participants, paid employment not involving mill work (on the shop floor) in a 'working town' like Cornwall is often perceived as 'higher class' and contributes to the complex processes of local identity (gender and class) formation.Indeed, according to the interview group, there was universal agreement over the fact that almost any type of job outside of millwork was better than working on a shop floor - even though factory work paid relatively well. The majority of our interview group (12 out of 15 participants)indicated that even part-time work in the service sector (25 hours per week or more) was preferable to part- or full-time work in a mill. Thus, the 'location' of work activity has as much or more to do with class perceptions in Cornwall than 'individual location' in capitalist production systems.

Restructuring labour market options through deindustrialisation in Cornwall may be contributing to work force 'feminisation' in terms of mirroring national trends towards employment tertiarisation, however, feminisation does not necessarily equate with lower status and perceived class polarisation in Cornwall. For instance, in terms of the construction of his work identity, Peter's experience with deindusttialisation and moving from well-paid 'masculine' millwork to lower paid 'feminine' service sector work has had a complex effect on his employment status and class image as defined in Cornwall . The feminisation of work therefore,is not necessarily seen as a 'negative' experience by those people who have been 'restructured' out of goods production into service sector employment or retraining schemes for service sector jobs. Herein we see the importance of locality in understanding the dynamics of regional labour markets, interpretations of 'feminisation' and how the social processes of production shape gender and class formation. Regarding the definition of 'class' in particular, Chouinard's (1996)observations concerning the subjective and objective nature of group and individual identity formation are particularly useful.

Regarding Peter's unpaid labour at home, he is involved in a non-capitalist (feudal) class process of surplus labour production and appropriation. However, contrary to the 'usual 'arrangement of the husband appropriating labour from his wife, Peter's gendered identity is further fragmented and difficult to capture in the context of 'traditional' class identities given among other things, his primary involvement in the 'feminine' activities of cooking, cleaning and child care. In fact, six out of the eight males interviewed had 'non-traditional' and highly fragmented class and gender positions given their experiences with 'feminised' employment patterns (job loss in manufacturing and job gain in the service sector). It may be the case that men performing 'feminine' unpaid labour at home do not experience the type of (patriarchal)gendered power relations with their (female) partner which often shape the socialized value of women's unpaid domestic labour. However, 5 of the 6 female interview participants who were married indicated that their male partners had always done some domestic and child care work particularly in the context of shift working when both people had been employed in factory jobs.In terms of further research concerning fluctuating and fragmented identities, more questions need to be asked about how we view male gender in the context of different types of employment conditions, occupations and adaptation to restructuring labour markets.

In terms of making theoretical links between industrial restructuring, changing employment patterns 'in place' and gender, our research experience underscores how social processes are rooted in peoples' specific ways of engaging in and reacting to lived socioeconomic relations. Our example points to the fact that Peter's gender and class identity are 'in flux' as he negotiates the changing material conditions, power relations and divisions oflabour which structure his life.

7 Discussion and Conclusions

What our research has shown us is that the particular ways in which people (andcommunities) transform and adapt to uneven economic development does not necessarily correspond with dominant discursive expectations or explanations. The particular strength of ourresearch design (integrating statistical and interview data) was that it allowed us to examine dominant trends structuring national and regional labour market change in Canada in combination with how that structuring was experienced by individuals. In terms of new approaches to regional geographies of change, this allowed us to: (a) focus on gender as a key analytical construct in critically exploring theoretical constructions of feminisation and the importance of locality and diversity; (b) build an understanding of restructuring events that linked micro-level specialisation and diversity with macro-level processes of change - both of which are in flux and mutually recursive and; (c) devise an approach to restructuring studies that breaks down the restrictive boundary differentiating 'qualitative' from 'quantitative 'research.

Developing pluralism approaches to geographic research allows investigation to reach beyond the socio-statistical level, to the level of exploring 'processes' and 'context' in geographic theorising and research. Particularly in the study of gender - a socially, economically, culturally and politically constructed and contested category - flexibility of research techniques and an understanding of how individuals experience changes in flux, is vital. In turn, this helps us move towards a greater concern with how we frame questions, interpret analytical categories and form explanations.

In seeking to understand the socio-spatial, economic and political implications of restructuring economies and regional labour markets, economic geographers might start by renegotiating fundamental categories of analysis and, by considering the central role gender plays in constituting class and non-class processes. Where spatial divisions of labour and arguments concerning post-Fordist working relations have placed gender firmly on the restructuring agenda, we suggest that more work needs to be done on: (1) de constructing hegemonic and masculine notions of 'the economic' in geographies of employment and regional change; (2) exploringdiversity of socioeconomic forms of production and exchange in restructuring communities, and;(3) examining the complex and contradictory experiences of multiple gender and class identities in the context of regional functional specialisation and locally experienced change. These issues are particularly relevant given our concern with 'placing' restructuring experiences in localitieswhere the social processes of production and class formation do not necessarily follow dominant fordist and post-Fordist theoretical constructions.

Another set of issues bearing on the use of pluralist methods (and relevant to making theoretical links between macro and micro processes and experiences) concerns the type and quality of data available in the exploration of gender and feminised employment patterns. A significant challenge we encountered involved dissecting macro employment trends in terms of their complex constituent parts, and speculating on the processes and mechanisms that link them to local experiences of deindustrialisation and individual employment change. Until recently (for example in the forthcoming 1996 Canadian Census), there has been a paucity of data sources outlining macro-level categories describing gendered (particularly female) experiences of paid work and life (which are by no means homogeneous, but are gendered). Among other things,reasons for this may be linked to the way researchers and policy makers define and value contributions to social well being - which includes economy and polity. That this has impeded progress into research for women (aimed at equity) as opposed to research on women (counting them) is a problem of gender-neutrality being addressed by feminist researchers. We suggest that female and male experiences relating to work and life need to be better documented as a precursor to reexamining diversity and explaining social process.

Drawing on this last point, and relating it to concepts of feminised employment patterns and social constructions of work, is the paucity of information on how 'male gender' is changing in restructuring labour markets. This is particularly acute in communities like Cornwall where gendered divisions of labour in the work place (dominated by secondary manufacturing) are suddenly dissolved in light of rapid deindustrialisation~ on. In a place with relatively little economic diversity and prevailing rates of high unemployment, men and women are having to adjust to new employment opportunities, sectoral participation and types of paid jobs. Recalling Peter's fragmented gender, class and work identity, more research might be directed towards ways men are coping with changed employment patterns; how their gendered identities are reproduced in the context of new paid and unpaid work relations and how these changes may be affecting female work opportunities and gendered identity formation.

While the nature (or at this point, existence) of a 'new' and emergent economy in cornwall is not clear, our research has demonstrated the importance of understanding how individuals, groups and communities experience contradictory and fluctuating conditions of change. Too often, it is assumed that 'old order' (Porter, 1991) manufacturing communities like cornwall either succumb to the forces of globalised~ on and fade away as economically (and by extension) socially irrelevant places, or, restructure themselves out of crisis (Massey, 1984) and emerge as global.competitors in an increasingly integrated world economy. And, in terms of employment restructuring, the complexity of gendered divisions of labour and feminisation are often not adequately explored in the context of dominant discursive portrayals of Fordist-post fordist labour market transition. What we have argued throughout this paper is the need for geographers to critically reevaluate the theoretical understandings of conditions of change and diversity in building de-centred and placed geographies of restructuring. It may be the case that communities like Cornwall will not reemerge as leaner and more competitive manufacturing centres in the global economy, but rather as our research seems to indicate, community members and organisations may adopt a variety of strategies based on trading and exchanging goods and services in order to meet broader social and individual needs. Whether these strategies are viewed as successful points of resistance to global forces of restructuring, means of survival orsome combination of coping and resisting is an area requiring more investigation. In the meantime, the challenge is to identify economic and political options aimed at mitigating the most acute effects of socio-economic adjustment with trying to promote and support options for community redevelopment.