SciELO - Scientific Electronic Library Online

vol.1 special edition author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Revista Brasileira de Ciências Sociais

Print version ISSN 0102-6909

Rev. bras. Ci. Soc. vol.1 São Paulo  2005


Towards a sociology of unemployment*


Por uma sociologia do desemprego


Pour une sociologie du chômage



Nadya Araujo Guimarães

Translated by Knight Dundonald Campbell
Translation from Revista Brasileira de Ciências Sociais, São Paulo, v.17, n.50, p.104-121, Oct. 2002.




The paper aims to reflect on the links between the phenomena of employment and unemployment in a context governed by the intense flexibilization of work and the institutional and normative reconstruction of the patterns of workers' protection. The argument unfolds in three directions. Firstly, it accompanies the movement to redefine the notion of "unemployment," pursuing, with the help of the recent sociological literature, the constitution and social legitimacy of the new phenomenon of the "long-term unemployed." Secondly, it compares theoretical developments that have attempted to understand this phenomenon with the efforts made by the Brazilian labor sociology to interpret the problematics of unemployment. Finally, it argues in favor of the heuristic value of contextualized comparisons to test, for the Brazilian case, the hypotheses developed by the sociology of unemployment in Europe, particularly France.

Keywords: Unemployment; occupational trajectories; Brazil.


O artigo busca refletir sobre os elos entre os fenômenos do emprego e do desemprego em contextos de intensa flexibilização do trabalho e de reconstrução institucional e normativa dos padrões de proteção ao trabalhador. Para tanto, a argumentação se desenvolve em três direções. Em primeiro lugar, acompanha o movimento de re-significação da noção de "desemprego", perseguindo, com a literatura sociológica recente, a constituição, e legitimação social, da nova figura do "desempregado de longa duração". Em segundo lugar, compara desenvolvimentos teóricos voltados a entender este fenômeno com os esforços empreendidos pela sociologia brasileira do trabalho no sentido de interpretar a problemática do desemprego. Em terceiro lugar, argumenta em favor do valor heurístico de comparações contextualizadas para testar, para o caso brasileiro, hipóteses desenvolvidas pela sociologia do desemprego na Europa e, sobretudo, na França.

Palavras-chave: Desemprego; trajetórias ocupacionais; Brasil.


L'article propose une réflexion sur les liens entre les phénomènes de l'emploi et du chômage dans des contextes d'intense flexibilisation du travail et de la reconstruction institutionnelle et normative des valeurs de protection du travailleur. Pour cela, l'argumentation se développe en trois directions. En premier lieu, elle accompagne le mouvement de re-signification de la notion de ìchômage", tout en poursuivant, avec la littérature sociologique récente, la constitution et la légitimation sociale du nouvel aspect du "chômage de longue durée". Ensuite, il compare les développements théoriques qui cherchent à comprendre ce phénomène grâce aux efforts entrepris par la sociologie brésilienne du travail dans le sens d'interpréter la problématique du chômage. Finalement, il soutient la valeur heuristique de comparaisons contextualisées en vue de tester, pour le cas brésilien, les hypothèses développées par la sociologie du chômage en Europe et, surtout en France.

Mots-clés: Chômage; trajectoires occupationnelles; Brésil.



The following reflections have arisen from a desire to confront an interpretive challenge, that of understanding the links between the phenomena of employment and unemployment in the context of the intense flexibilization of work and the institutional and normative reconstruction of worker protection standards. The argument is developed in three directions.

Firstly, I shall attempt to trace the movement to redefine the notion of "unemployed", pursuing, with the help of the recent sociological literature, the constitution (and social legitimization) of the new concept of the "long-term unemployed".

Secondly, I shall compare the theoretical developments aimed at understanding this phenomenon with the efforts of Brazilian labor sociology to interpret the problematics of unemployment, by analyzing the links between this phenomenon and the nature of social organization and the patterns of inequality prevailing in our country.

Thirdly, I shall argue in favor of the heuristic value of resorting to contextualized comparisons – in the light of recent findings regarding the fluidity between occupational borders and the transit between situations of employment/unemployment/inactivity in the labor market in other countries – in order to discuss the validity of using the hypotheses developed by the sociology of unemployment in Europe (and France in particular) to interpret the Brazilian situation.


From the sociology of work to the sociology of employment? Structural changes and the redefinition of concepts.

The intense restructuring of production that took root in the most industrialized nations as of the 1980s, has been subjected to a good deal of study, particularly from the point of view of its effects on the technological and organizational reconfiguration of productive environments, most notably in the manufacturing industry. Careful analyses of the organization and management of both production and work have identified changes that point to new paradigms of production.

From early on, the literature on the sociology of work has drawn our attention to substantial alterations in: (i) working practices, with an impact on the size of the workforce (managerial and operational), leading to new types of work contract that have segmented internal and external labor markets; (ii) the content of job positions and the division of tasks in the direct operation of productive processes; (iii) the division of work within companies, and the resulting occupational structure; (iv) hierarchical relations, i.e. the new ways in which power is divided and circulates within organizations, with a particular emphasis on the mechanisms for generating consent in contexts involving intense change and restrictions on rights.

Since these studies were conducted from the privileged position of the workplace, their results were especially concerned with the impact of the technical and organizational changes on those workers who might be termed as "surviving" the restructuring process. Thus concern was directed towards the forms of hiring and use of labor which arose in environments undergoing this restructuring. How, then, to use these reflections to formulate questions about the occupational horizons of those who are not included in, or were recently excluded from, these environments?

In fact, the analysts have followed two paths. I shall examine both, focusing not only on the empirical results but (particularly) on the nature of the analyses themselves.

The first was aimed at producing elaborate and meticulous descriptions of the "survivors", either by characterizing the new employment practices, or by examining the chances of reinsertion (or "employability conditions", as the jargon has it) of those outside good jobs in the labor market. The assumption (not always explicit) underlying this first analytical tendency was that, in conditions of ample job supply, and where the employer had absolute authority in the selection process, it was reasonable to assume that the characteristics of the employees primarily expressed the requirements of the employer's recruitment and selection policies, which (precisely because of this almost unlimited supply) could become disassociated from any constraints imposed by the profile of those individuals in search of work.

The second attempted to analyze the characteristics of the unemployed themselves. The magnitude and types of unemployment, as well as the profiles of the unemployed workers, would provide useful indicators for localizing and describing those social groups with the least prospects of inclusion. But why do this if, thanks to the first analytical path, we already possessed a good approximation of corporate requirements? Because only such an approach was capable of delineating the hiatus between the characteristics desired by the employers and the "assets" of those seeking work. In addition, without this knowledge it would be impossible to draw up policies – governmental, union and corporate – aimed at ensuring employability.

What is the link between these two analytical tendencies? The recognition that under productive restructuring and intensified globalization of corporate business strategies, job opportunities are affected by two trends. One is towards convergence (and thus homogenization) and the other is towards hybridization (and thus a combination of diverse local forms). To clarify: on the one hand (and under the banner of productive and financial globalization), there is a growing convergence between normative and institutional, national and international structures shaping the supply of goods and services.1 There is also a convergence regarding the normative models of the work culture (the so-called "new paradigms of production" and their accompanying "best practices"). However, it is equally possible to recognize a movement in the opposite direction, a countertendency, towards hybridization. This is because the norms and institutions that are spreading in a globalized environment interact with the specific characteristics (national and sectorial) of the social contexts in which this diffusion is taking place, and where it is embedded. How are these tendencies expressed in the current patterns of employment and unemployment?

Looked at from the convergence point of view, the hypothesis that there is an irresistible diffusion of a new pattern of hiring and using labor gains ground. This is the so-called "flexible specialization model". How, in this context, is the link between employment and unemployment expressed? The classical Fordist framework – founded on collective bargaining, the power of the unions and a system of social protection – is inadequate when faced with contemporary corporate strategies, fueled by intense rationalization (of both production and work) in a climate of harsh competition and unprecedented exposure to globalized production and consumption patterns. "Focalization" and "deverticalization" are tied to the growth of "subcontracting" and the "externalization" of work. Thus production is the result of corporate networks, in which the major firms ("lean" and "focalized") are tied to a select number of qualified suppliers.2

What are the effects on employment? The distribution of positions between sectors and companies is altered, according to their size. An inter-sector shift leads to an increase in service jobs.3 Another type of recomposition increases the relative weight of jobs in small and mid-sized firms. In addition to this redistribution, however, there has been an important alteration in labor relations, since the shift of employment from big corporations to small and medium-sized businesses is accompanied by an increase in the so-called "atypical" types of employment4 to the detriment of full-time workers protected by the results of collective bargaining and public welfare. The numbers of self-employed are also increasing. This growth of precarious labor relations has only been possible, in the central nations, due to the flexibilization of both the welfare system itself and the regulatory framework governing hiring procedures. The latter movement has been aided by the fact that the shrinkage of the internal labor markets has simultaneously attacked wage workers (governed by stable and protected employment) and their unions, reducing their ability to resist the changes and, consequently fueling them. This signifies a third order of effects, those that have affected the industrial relations system itself, determining the future expansion prospects for the flexibilization of labor relations.

This conjunction of observations can perhaps be reduced to two major (and dramatic) tendencies. In quantitative terms, we have entered an era in which production has grown without a corresponding expansion of jobs, leading to an increase in that portion of open unemployment that can best be understood as structural in nature.5 In qualitative terms, jobs have become polarized into "good" and "bad". One question that immediately springs to mind is: which workers have benefited from the so-called "good" jobs? In other words, what is it about these workers that has enabled them to survive the restructuring and allow them to maintain the higher-quality positions?

From the point of view of convergence, one response is particularly appealing: the new productive contexts have led to a massive differentiation among workers in terms of human capital – i.e. the greater and broader one's skills, the greater one's chances of survival. Thus a first means of resolving the question would be to suppose that the higher the worker's level of education, the greater his or her chances of professional rerouting. Such workers therefore have a higher "trainability" and thus possess a greater degree of defensive professional mobility (intra- or inter-corporate) when faced with the restructuring process. They are therefore more "employable" (Alves and Soares, 1997).

However, as certain authors who have adopted this line of approach have pointed out (such as Amadeo et al., 1993), flexibilization was associated with both general and specific skills. This is because, on the one hand, workers with more accumulated general capital were in a comparatively advantageous position in times of adjustment and restructuring. They were better equipped to perceive and interpret the economic changes; had a higher capacity for learning new, lower-cost techniques; were more productive; and were given preference in retraining. We can therefore conclude that those who had accumulated the highest general human capital had also accumulated the highest specific human capital. On the other hand, however, the picture was a complex one, given that these same workers were also the ones who lost their accumulated specific human capital most rapidly in times of productive restructuring. Thus the loss of firm-specific skills was more rapid in restructured contexts. Consequently, in order to understand the chances of survival, the key variable appeared to be the amount of specific human capital in relation to general human capital accumulated by each individual, with this ratio in turn determining the cost of restructuring for each type of worker.

However, if we look at the question from the point of view of those who did not survive in these restructured contexts, there has been a clear increase in unemployment, so much so that it has become a structural problem, even in the most advanced economies. More than this, the very forms of unemployment and the profile of the unemployed themselves have altered. However, these changes have not been unidirectional; on the contrary, from the ways in which they have become manifest in different countries, it would appear that the previously-mentioned tendencies towards convergence, far from being univocal, were in fact constrained by important societal limits (Bénoit-Guilbot and Gallie, 1992).

Observing these same tendencies, but this time not through the lens of the convergence hypothesis, these limits began to be deeply examined. Among others, they include: the distinct types of employment practices adopted and the greater or lesser role of the State in employment policies; the diversity of size and format (public or private) of the different countries' welfare systems; the varied nature of the industrial relations system and its consequences for wage bargaining and professional relations; and certain broader cultural determinants, what D'Iribarne (1990) called "the codes of legitimacy" and their implicit effects on individual rights.6 All of these factors help determine corporate competition strategies, their human-resource policies and the varying ways in which the different countries are economically inserted into the new global competitive scenario.

Bearing this in mind, at least three typical situations can be discussed in order to offer explanations for the inter-societal differences and the specific forms of global and national hybridization (Demazière, 1995). Considering these situations allows us to shed light on the ways in which national societies, due to their specific characteristics, react to the challenge that productive restructuring poses for their respective employment systems. Let us look first at Japan. There, unemployment rates have remained relatively low over the long term7, thanks to two main escape valves. Firstly, the growing underutilization of labor was not apparent in the open unemployment figures, given the rapid movement of many unemployed people into the ranks of the economically inactive; this is particularly true in the case of women, showing the weight of specific cultural factors on Japanese gender relations. Secondly, the "survivors" had remained in the "good" jobs by submitting to intense wage flexibilization.8 This flexibility was in turn part of the socio-cultural characteristics of the so-called "Japanese life-time employment system" in which workers were allocated to specific positions in the production chain and not necessarily within the firm itself; as a result, a type of predatory mobility towards the most distant links in the chain ensured that occupational levels were maintained, flexibilizing and degrading the hiring and working conditions for a significant portion of the workforce (Hirata, 1992; Hirata and Zarifian, 1994).

The United States perhaps exemplifies a second social situation vis-à-vis the effects of restructuring on unemployment. There, an exceptionally flexible labor market allowed the number of available jobs to expand, helping rein in the rise in the unemployment rate (which, even so, exceeded 6% at the beginning of the 1990s).9 Is this a solution that can be generalized or is it once again specific to a particular society? Analyses of comparative figures from the OECD countries (Demazière, 1995, and Dedecca, 1996) suggested that deregulation, founded on the creation of precarious and low-paid jobs, would not be capable of confronting growing unemployment. The experiments in flexibilization undertaken in certain advanced economies, notably Canada10 and the UK, ended up raising the rates of forced economic inactivity among men at the peak of their activity, i.e. between the ages of 35 and 44.11

Finally, a third situation may be said to characterize such countries as France, Germany and Italy, economies suffering from reduced job-creation capacity and rapid unemployment growth. In these cases, given the still extremely strong public welfare system,12 not only is the underutilization of labor growing, but it has assumed a new pattern, that of the so-called "long-term unemployment". This was an entirely new phenomenon and the very means of categorizing it presented us with a paradox: unemployment ceased to be codified as an involuntary and occasional deprivation of work (and, as such, legally recognized and statistically measured) and began to acquire the extraordinary character of permanence.

And why did it appear so extraordinary? Because, in the way it had been previously conceptualized, unemployment showed two strong components of the code of legitimacy prevalent in our societies. Firstly, by being temporary, work deprivation freed those subjected to it from being branded as "lazy" ("idle", "weak", "not doing enough" to find a job). Secondly, by being involuntary, work deprivation was something "suffered", so individuals in this situation could be differentiated from "indisciplined", "unstable" and "irresponsible" workers who were regarded as being the agents of their own exclusion in that they lacked (once again) the values attached to the normative culture of wage work.13 Thus social recognition of the legitimacy of unemployment – defined as a transitory and involuntary phenomenon – did not threaten any of the normative components central to the work ethic. (Demazière, 1995a).14

However, the invention of a new social category – that of the "long-term unemployed" – together with the force that the phenomenon assumed in precisely those capitalist societies with the greatest tradition of labor regulation and protection, reveals the unprecedented rupture of the nexus between employment and unemployment. Once this nexus is ruptured, the underutilization of labor ceased to assume the classical and unique form of open unemployment and begins, as we have seen in the situations cited above, to express itself in multiple forms, such as individuals at the height of their active life joining the ranks of the economically inactive, precarious and/or atypical forms of so-called "low-quality jobs" and long-term unemployment itself.

What appears to be at stake when one considers these multiple forms of underutilization of labor? The fact is that unemployment is not only increasing quantitatively and assuming a diversity of forms, but is affecting individuals according to their gender, age, socio-professional category and educational level, variables so dear to the sociological and socio-demographic analysis of inequality. Or rather, unemployment, as well as being involuntary, as our classical normative model of work would like, is also highly selective, since the chances of employment are unequally distributed among the different social groups.

This selectivity became particularly clear in "long-term unemployment", as the latter was expressed in the most advanced capitalist countries as of the 1980s. In these nations, the handicaps which had become morally legitimate and which justified difficulties in securing a job, were no longer, as in the 19th century, purely physical, but also social. In the case of France, for example, data collected between 1983 and 1989,15 a crucial period in the expansion of long-term unemployment, allow one to identify a conjunction of social risk factors that altered individuals' employability.16 As Demazière (1995, p. 55) puts it, "the most favored group consists of young male graduates with little time out of work (a 68% chance of being re-employed), and the least favored one comprises older, non-graduate women not actively seeking work (a 7% chance of being re-employed after two years of unemployment)".

The contemporary labor market is also characterized by another trait. Individuals are no longer leaving the ranks of the unemployed by obtaining a stable occupation through a high-quality job. On the contrary, the empirical evidence, gathered from the tendencies in the most advanced capitalist countries, suggests that "there is a simultaneous trend taking place – more people are becoming unemployed and less people are becoming re-employed" (Dedecca, 1996, p. 14). Moreover, even those who are re-employed are likely to have more tenuous employment linkages so that they become more likely candidates for future unemployment. Thus definitively leaving the ranks of the unemployed (or at least for some considerable time) by acquiring a new job is by no means certain.

We were faced with strong evidence that the link between employment and unemployment, hitherto regarded as "natural", had broken down. Consequently, the duration of a given individual’s unemployment became a sensitive measure of his or her future re-employment difficulties. "It seems as if unemployment has helped redistribute jobs" (Demazière, 1995, p.52), in that the degradation of hiring conditions had become a common characteristic for a substantial portion of the unemployed re-entering work.17

If this line of argument proved sound in its capacity to interpret realities where social protection of the unemployed was effective, what, then, to say about the scope of this unemployment tendency in those realities where the welfare systems were fragile? Or, borrowing from Esping-Andersen (1990 and 1999) and de Gallie (2001), what happened in contexts where the "public regulation systems geared towards the protection of individuals and the maintenance of social cohesion through intervention (by legal and distributive measures) in the economic, domestic and community spheres" (Gallie, 2001: p. 2) were largely ineffective. I shall return to this institutional aspect later; but for now, I would like to dwell a little longer on the question of selectivity and its consequences for our research.

If we agree with what has been proposed so far, it is clear that the traditional idea of employability has to be enhanced.18 The probability of obtaining a job has become dependent not only on the classical sociological measurements of position, which document the characteristics of unemployed individuals seeking work (gender, age, educational level, etc.). It now requires a longitudinal analysis, i.e. knowledge of the individuals' occupational trajectories, given that their chances of professional (re-)employment will largely depend on their previous experience of employment and unemployment. In addition, the various social networks, constructed during times of work or periods of unemployment, play an important role in determining re-employment success. This is particularly true in the case of the family group.

Thus employability, more than mere individual ability, should be understood as a social construct (Demazière, 1995 and 1995a; Outin, 1990), since the action of successfully acquiring a job is governed by factors that go beyond individual will and conduct. This is because, in the new productive contexts worker's occupational trajectories depend on: the way in which their individual attributes are appreciated in different times and spaces; their "social skills", i.e. the accumulation of sufficient social capital through the effective networks to locate and acquire a job; and, finally, structural factors absolutely beyond their control, such as corporate locational and investment strategies (DiPrete and Krecker, 1991). Thus employability becomes the result of the interaction between these strategies, both individual and collective, as much among those seeking work as those offering it (Outin, 1990).19

It is not without reason that the international literature based on comparative studies (see, for example: Gallie and Paugam, 2000; Gallie, 2001) calls attention to the fact that the profile of those groups most vulnerable to unemployment varies substantially from country to country.

Towards an agenda of new themes for a sociology of unemployment: the normative-institutional and personal-subjective dimensions20

Although there is a certain convergence among the analysts regarding the heuristic value of comparisons in the process of revealing unemployment's specifics and/or social determinants, it is a fact that sociological analyses of this issue based on international comparisons are rare; in most cases, such studies limit themselves to constructing statistical series, based on calculations consistent with international measurement standards (mostly based on ILO operational definitions). Whatever their degree of sophistication, such analyses are implicitly founded on the idea (extremely uncomfortable for we sociologists) that unemployment means the same thing in the different countries where the statistics have been collected.21

In fact, however, a sociological analysis of unemployment cannot possibly ignore the fundamental point that the meaning of the term varies from country to country. After all, this point of departure is at the very heart of our disciplinary approach!

We know that the mechanisms of access to a professional activity or the obtainment of material resources are not uniform; on the contrary, they vary in line with national and personal contexts (gender, age and ethnicity). Similarly, the various unemployment insurance systems and the dispositions for distributing resources among those seeking work operate in distinct ways, and are based on equally diverse and socially-determined eligibility criteria. In the same way, there is considerable variation in the types of institutional support that each society offers those in search of work (government-run employment agencies, policies to combat unemployment, small job announcements, local collectives, informal networks, etc.).

Even if we consider a single society at different moments in time, these variations also occur. Consider Brazil. In the recessive climate at the beginning of the 1980s, although the phenomenon of unemployment had acquired enormous social visibility, the ways in which the State agencies intervened, and the mechanisms of the public systems themselves, were much less active and developed than they were in the 90s. In addition, the unions played an important role in defending "their" unemployed. This suggests both that re-employment in the same sector remained feasible (as soon as the crisis abated), and that the unions had sufficient social legitimacy to raise funds (albeit temporary ones) to support "their" unemployed. On the political front, the unemployed movements were key agents in the demand for a more effective public regulatory system, an indication that being unemployed was not a stigma; on the contrary, it could be assumed as a diacritical sign, an identity, even if based on a transitory situation. Under these conditions at the beginning of the 80s, "unemployed" workers remained, for example, "metal workers"; i.e. their occupational origin was proof of their self-recognition and their recognition by others (notably their own union). This was possible because of a belief in the possibility of re-employment in one's "original sector"; and this resulted from the strategy of work management based on the turn-over of workers. Such a subjective construction, and its institutional counterpart, lost sense in the crisis of the 90s.

Thus, the variation in the meaning of unemployment – which I have sought to illustrate, albeit rapidly – was expressed as much in the normative plane, via the differing ways the institutions interacted with those designated as unemployed, as in the subjective one, the experiences of the unemployed themselves. With this, I am calling attention to the fact that the duty of a sociology of unemployment is to analyze and conciliate two dimensions that are pillars of our disciplinary tradition: on the one hand, the phenomenon’s institutional and normative framework; on the other, its subjective meaning, built up through experience in the labor market and resignified by the subjective interpretation of individual histories. In this sense – and seen in strict sociological terms – being unemployed means being institutionally recognized, being counted and considered as unemployed, and, at the same time, subjectively defining and considering oneself as such.

For this reason, comparative studies have become a useful method for describing and understanding the forms (and transformations) assumed by unemployment in distinct societies (D'Iribarne, 1990). Despite the existence of measurement conventions, the basis for the statistics that underpin our habitual international comparisons, a sociological view of the phenomenon cannot ignore the fact that the boundaries between unemployment, activity and inactivity are relative, in that they result from social constructions specific to each reality. For this reason, we must not only consider the institutional forms, the systems of rules, the normative dynamics which, differing from country to country, play a specific part in shaping employment and unemployment, but also the forms of activity, the ways that individuals categorize their situation, their personal histories, realities and professional values. In short, comparisons between societies (or between different time periods in a single society) allow us to elucidate the institutional logic in play and its role in shaping unemployment, employment and inactivity.

In this field, there are still many ways to investigate a possible agenda for a sociology of unemployment. The first consists of identifying the institutional actors that participate in regulating employment and analyzing their legitimacy, considering the different types of institutions intervening in the process: the State, clearly; non-State public institutions (e.g. unions and union federations) and their social partners (e.g. NGOs); the family (which plays a crucial part in role structuring, dividing the responsibilities for ensuring survival and devising job-market strategies for its members); and the conjunction of mutual-aid networks (founded on blood ties, such as the family networks, or other types of communal identification, either permanent, such as regional or ethnic origin, or transitory, such as neighborhood or church organizations).

However, given that unemployment is a subjective construction, the frequent variations in its statistics (when comparing different social contexts) results from the fact that, in order to be socially recognized (and computed) as being unemployed, the individuals concerned must recognize themselves as such and resort to those institutions responsible for recognizing the condition. In order to understand if, when and how this self-identification arises, it is essential to analyze personal experience. This includes studying the professional biographies of the unemployed as well as the subjective feelings attributed to the condition by those individuals experiencing it.

These different dimensions of the employment experience are, in turn, inscribed in culturally-diverse normative frameworks that must also be considered if one intends to construct a sociology of the different types of unemployment.


Thinking comparatively: Brazil in the light of the other realities

In countries like Brazil, the issue of unemployment is a critical one. In this context, the stable wage relation was never universalized as the dominant norm of unemployment, nor did mass production give rise to a movement of extended citizenship and social protection in the form of a public welfare system capable of protecting jobs. As Silva has wisely pointed out (1990), the specifics of our labor market and our industrial relations, allied to the authoritarianism of State-union relations, meant that the organization of rigid mass production came to dominance free from the social counterparts that gave it legitimacy, in contrast to what Boyer has described as "true Fordism".

In the United States, these counterparts welded the compromise between the social policy of the State, corporate interests and workers’ demands. The principle of "five dollars a day" became the best-known expression of this compromise, whereby greater access to consumption was only the external face of a citizenship achieved within the ambit of production. Here, on the other hand, the trade union movement was unable to set itself up as the legitimate intermediary for negotiating working conditions and wages, nor did workers achieve their full status as consumers, as they did under the American mass production system.

In addition, labor relations were characterized by excessive employment instability, wage scales with widely differing levels, greater rigidity in the definition of job positions and no tradition of team-work. In our society, the political culture has never favored stable wage negotiation rules, the inviolability of labor agreements or the legitimacy of workers' action to claim their rights. The high level of unemployment and the lack of any tradition of worker participation in negotiations on the introduction of new technologies meant that the model of the industrial relations system was entirely different from that which prevailed in the countries where the post-Fordist paradigms of organized labor originated (Ferreira et al., 1991).

If we add to all this the lack of any sense of universalism and republicanism in the social actions of the State, which resulted in the constitution of a pale imitation of a public welfare system (compared to France's, for example), without the corresponding and compensatory development of private welfare (see the Japanese system). Here, such a development became absolutely unnecessary, its place being taken by an authoritarian paternalism that arose at a time when consent was still dictated by the old market despotism.

Thus productive restructuring finds itself facing social problems that affect the "survivors" of this process as much as those "excluded" from it. Unlike in the major capitalist countries, where, as I have mentioned, the symbolic and material frontiers between unemployment and pauperism can be easily established, there being a diversity of policies and institutions to deal with both, in Brazil's case there is a distinct lack of employabilty and social welfare policies.22

However, even in the case of the small contingent of possible "survivors", one is immediately struck by a serious educational deficit, the fruit of the low educational level of the Brazilian population in general, and workers in particular, and the fact that the public school system has been all but destroyed.23

How did this "Brazilian Fordism" shape the characteristics of our vocational training system? Unlike almost everywhere else, here this system developed in complete autonomy from the educational system. But, unlike other realities regarded as virtuous (the German model, for example), it was also autonomous from the State and the workers' unions. Distant from the State, under the current understanding that the latter's source of funding (payroll charges) is essentially "private" in nature, the professional training system was also distant from workers' representative organizations, which, until very recently, did not understand that the training issue could be part of their political agenda (Fogaça and Salm, 1995).

Finally, companies themselves have not become heavily involved in vocational training. In fact, the growth of capitalism in Brazilian has been marked by exceptionally meager investments in human capital (Carvalho, 1992). The large number of institutional factors that regulate the labor market here (the CLT, or consolidated labor laws, the FGTS, unemployment insurance, pay rise rules) ended up giving rise to a system that did not encourage workers to invest in their jobs or firms to invest in the training of their workers. The data collected by Amadeo et al. (1993) suggest that either dismissal costs in Brazil have always been extremely low, or that the incentives of the FGTS were so strong that they offset those costs (the institution of the FGTS, a compulsory workers' savings fund levied as a payroll charge, gave employers the right to rescind work contracts no matter how long the period of employment provided they paid compensation). Herein, perhaps, lie the reasons for the lack of corporate investments in workforce training. Given the high probability of losing such investments, only specific and indispensable expenditures were made and, therefore, the less qualified the worker, the lower his or her opportunities within the firm and the lower his or her wage. From the company’s point of view, the best strategy was to get the most possible out of workers when they were still under contract with no commitment to their future. Hence the paltry investments in human capital (especially regarding the least qualified workers) and high workforce turnover.

Can we shed more light on the Brazilian case if we compare it to other experiences of the institutional construction of employment and unemployment?

It has become increasingly common to find authors who believe that comparative analysis is a fruitful strategy for describing the processes by which unemployment is configured. Production chains have been compared; within those, complexes, and within those, individual firms. Our own analyses (Cardoso, Comin, Guimarães, 2000; Cardoso, 2000; Guimarães, 2002) have already led us to recognize that, given the way in which the labor market and our welfare regime have become institutionalized, the restructuring of companies has led to enormous disorganization in the occupational trajectories of their ex-employees. At the same time as they have severed their links with their former workers, they have also severed the link with registered and (minimally) protected labor. And this, no matter where or when we look, has become the dominant tendency – challenging internal labor markets, professional careers, collective identities, public policies and individual strategies for skills acquisition.

Increased transit in the labor market, the loss of links with a prior professional destiny, has meant that studies of workers' trajectories according to their original sector/production chain have become virtually meaningless. Or at least it lost meaning for that portion of workers whose most visible and inescapable fate was to lose the links with their professional past, being launched back into the job market on paths increasingly characterized by recurring long-lasting unemployment. We had, therefore, to put the weight of long-term unemployment into perspective; this presupposes a system that anchors (and social networks that protect) the unemployed, keeping them in this situation and with this identity.

In conditions of recurrent unemployment, with workers constantly moving in and out of work, it is in the labor market that recent literature seeks an understanding of individual mobility possibilities, since workers' professional pasts have little bearing on their occupational future. The labor market institutions, notably those which regulate (institutionalizing) unemployment, then became prime foci for new interpretations.

In this sense, intra-national comparisons between production chains give way to comparisons between societies, in an attempt to identify the forms of normative institutionalization and their consequences for the individuals involved. What, then, can two other highly distinct welfare regimes (one, public; the other, private), built on entirely different normative frameworks, like those of France and Japan, tell us about the Brazilian case?24 And why compare Brazil with precisely these two countries? Let us proceed step by step.

At first sight, these three countries could hardly be more different, both in terms of their respective unemployment situations and their economies. Their unemployment rates (measured in line with ILO norms, the international standard) are also very different. When we formulated our initial ideas for a comparative study, the French rate was the highest (10.9% in 1998); Japan's the lowest, albeit steadily increasing (4.5% in the same year); and Brazil’s, curiously, in the middle (7.6%, also in 1998).

In addition, each country had experienced a distinct socio-economic trajectory: passing from a rural and agrarian economy to rapid industrialization and urbanization in the second half of the 20th century, in Brazil's case; the move towards an articulation between modern service society and restructured manufacturing, transforming a very old industrial economy, in France's case; and impressive economic growth from the 1950s on, in Japan's. Such huge differences led us to underline the similarities as well as the particularities.

Despite the massive social, cultural and economic differences, all three were part of a context characterized by global trade, the debilitating deregulation of public authorities and the standardization of production, management and employment norms. Even with their highly contrasting histories, they were all involved in the globalization process so typical of the contemporary period. More precisely, they were all experiencing a common phenomenon, albeit to a differing degree and in varying time periods, marked by a threat to the norms of employment, with the rapid development of "atypical" types of job.

The effects of this process, however, were different in each country. In Brazil, there was not only increased mobility between occupations with distinct forms (formal and informal, regular and irregular), but also intensive transit between different conditions of activity, i.e. unemployment and inactivity, or occupation and inactivity. In France, we witnessed the growth of particular forms of employment in terms of the duration and stability of the work contract and working hours. In Japan, the "life-time employment" model, adopted by the major corporations, began to give way to precarious or atypical forms of employment, including jobs not subject to any kind of formal recognition, frequently part-time and with no social protection whatsoever.

These changes in the forms of employment and (consequently) unemployment affected both the institutional regulation of labor relations and workers' professional trajectories. It therefore became strategic to cast a sociological eye on the process, focusing on an analysis of the interaction of the institutional and normative systems that configured the management of employment and unemployment, as well as the personal trajectories that arose from these contexts of change. Or rather, each country could be characterized by a convention, a particular social norm that would define activity and inactivity, employment and unemployment, creating and institutionalizing a frame of reference for the subjective experience.

In Brazil, the institutionalization of unemployment was both weak and recent. The frailty of social protection and the various public help schemes for the unemployed, as well as unemployment insurance itself, led to a significant underestimation of work deprivation. This gave rise to a substantial clouding of the boundaries between employment, unemployment and inactivity, a clouding whose public manifestation was in the controversy surrounding the unemployment figures.25 Employment conditions here are exceptionally heterogeneous, ranging from occupational spaces subject to strong legal norms and social protection to those in a highly developed and diversified informal underground economy. In a context where the number of officially registered and protected jobs has fallen sharply, most notably in the industrial sector in the 90s, unregistered occupations (such as informal jobs or self-employment) have ensured the predominance of work flexibility (accounting for more than half of jobs), at the same time fueling the growth of service activities and irregular work.

Thus, in a context where the weight of formally-registered wage workers is "limited" (Lautier, 1987), the destabilization of employment conditions has led to accelerating transitions towards unregistered and unprotected jobs. The resulting defensive strategies began to base themselves on a combination of these two types of situation, either on the part of a given individual (accumulating both types of job at once, or alternating between the two), or within the family group. Most economically active women work in precarious or informal occupations, which are also occupied in great part by younger people. It is clear that such social mechanisms for the distribution of forms of employment have structural effects on the subjective reactions to unemployment and, more generally, on professional trajectories and activities. The social construction of unemployment in Brazil is marked by a rupture of the equivalence between work deprivation and unemployment. In addition, when the boundary between employment and unemployment is so permeable, other categories of subjective and political identification gain ground ("homeless", "landless", for example). Not only are they more utilized, but they are also more effective (than "unemployment") when negotiating social protection. It is not without reason that the strong movements of the unemployed at the beginning of the 80s gave way to social movements where individual interests are cemented by other collective identities. And even if most of the "homeless" are also "workless", it is not their collectively-shared subjective identity as "unemployed" that determines their actions.

In France, on the other hand, unemployment is strongly institutionalized. The public replacement institutions cover the entire country with a fine net, which has been in place for some considerable time. This helps explain why the unemployment rate in the Paris region (around 11%) was higher than São Paulo's (around 8%) in the year used for our comparisons. And not without reason; encouraged by the institutionalization of the social and employment protection system, people deprived of work tend to declare themselves as unemployed, registering with the ANPE (Agence Nationale pour l'Emploie) as soon as they lose their job. Only by doing so will they gain the right to enter the welfare system.26

However, unemployment in France has also been rendered banal as the forms of experiencing it have multiplied. But, above all – and this is the most salient characteristic from the sociological point of view – unemployment is occurring less and less in the form of an abrupt break in the midst of a continuous professional career. And, in this way – and this is the other theoretically important finding – the forms (or institutional configurations) of unemployment depend on and express the forms (or institutional configurations) of employment. In France, the latter have diversified greatly in recent years, expanding to include the so-called "particular or atypical types of job", all of which were exceptions to the previous prevailing norm. This exceptionality is manifested in two ways. Firstly, in the sphere of work contract duration and stability, Fixed Duration Contracts (CDDs), temporary contracts and internships have become institutionalized. Secondly, regarding actual working hours, part-time work has been gaining increasing ground. In the mid-90s the French job market contained around 2.5 million contracts involving part-time workers (out of a total of 22 million). The young, entering the labor market for the first time, made up the biggest contingent subject to the precarious work contracts, while women were the prime candidates for part-time work (85% of such jobs in 1998).

Activity rates were unequal in terms of gender and age. Among women in general, they remained extremely high, but there were marked age differences across the genders (Gaullier, 1999). Not only were people starting their working life later, but they were more subject to unemployment or increasingly frequent transitions between employment and unemployment. At the same time, professional lives were being interrupted more prematurely (either through unemployment of the elderly, or through the effects of social policies encouraging early retirement and the abstinence of job searches by older people). Thus gender and age differences appeared to be particularly important, whether in the analysis of professional trajectories and subjective relations vis-à-vis the different types of social statute, or in developing an understanding of the institutional policies for managing unemployment and the workforce.

In Japan, unemployment is certainly less institutionalized, but for very different reasons from those in Brazil. It is true that unemployment rates were exceptionally low until very recently. In addition – and this is a huge difference – employment management was handled much more by the companies themselves (normally big corporations) than by the State. Only recently have these major corporations begun to dismiss workers on a large scale, reversing the previous policy of retaining excess man-power by allocating it to their networks of "affiliated" companies.

It is hardly surprising, therefore, that for workers culturally and subjectively shaped under the "lifetime employment" system – the overwhelming majority of whom are men – the experience of unemployment was (and still is) marked by social shame and feelings of personal humiliation. For this reason, registering with an employment agency is seen as dishonorable. Unemployment is not only structurally lower, but its "undercounting" (reducing the issue to its purely statistical aspect) would be incomprehensible outside a cultural framework which keeps individuals from registering with the new protection system which is taking shape. For sure, atypical jobs are also multiplying (22% of the wage-earning population in 1998) and under diverse forms: temporary hirings (arubaito), mostly concentrated among young people; short-term contract employees (shokutaku), concentrated among older workers; and finally part-timers (a misleading term, to say the least, since such workers often work comparable hours to full-timers, but without the benefits of social security). The latter category includes 95% of working women, most of whom have returned to work after raising their children.27 Moreover, for the categories of workers who do not have a registered job, the boundaries between employment and unemployment, and even activity and inactivity, are more fragmented and blurred, because in Japan (unlike in France), breaks in employment are not systematically categorized and recognized as unemployment, particularly when they do not involve the right to some kind of indemnification. One of the frequent reasons put forward to explain the low level of unemployment during periods of crisis is that women, having lost their precarious jobs, do not seek a new one (Freyssinet, 1984). The norms of activity and behavior in relation to unemployment and employment are highly differentiated in terms of gender, but also depend upon the moment in one's life cycle (Sugita, 2002).

These observations, although initial and provisional, are, however, sufficient to convince us that there is a case for building a comparison between Brazil, France and Japan based on the different (and/or specific) ways in which unemployment is manifested, defined and experienced in each country.

For sure, there will be those who are content to apply the habitual analytical parameters to the subject . For those policy makers (happy with the established forms of technocracy) it may appear sufficient not only to measure (and measurement for them is indisputable), but to retain the institutional definition for those two highly obvious clusters: on the one hand, stable employment, and on the other, institutionalized unemployment. And, in fact, for a long time our public policy directives have only concerned themselves with these two groups. It is true that this reflected a civilizing model whose core was the wage relation and which postulated – for strong intellectual, political, but also perhaps moral, reasons – that the universalization of this norm was an expression of modernity and, therefore, the basis for guiding political action, public intervention and academic analysis. This justified focusing the analysis on the two poles, each of which was typified in a unique form of expression.

But there is another way of dealing with the subject (a more sensitive one, I would suggest), but one which has triggered a dispute in the "good measurement" camp under the argument that, in the structure of labor markets in societies where industrialization came late, the "other" of stable employment cannot only be institutionalized unemployment. Because of this, some have proposed substituting simple open unemployment alone with a wide range of other types of job deprivation, since the former is only the "other" of the classical wage relation. The contrast with the so-called marginal economies alerts one, therefore, to the existence in these countries of distinct structuring patterns for the labor market and the employment relation and, consequently, of distinct types of unemployment. The challenge for analysts and for public policy is no longer the so-called (and paradoxical, as we have seen) "long-term unemployment", but the new phenomenon of "recurrent unemployment". This means of reacting to the problem calls attention to the need for theory and practice to concern themselves not only with two clusters of concrete situations – that of stable employment and that of institutionalized unemployment – but also to extend over a broad zone between these two extremes, quantifying it and making it a legitimate focus of public policy (if not the focus).

The aim of this paper is to shed light on this ample gray zone which challenges our ideal of modernity, universalism and inclusion, and which, therefore, needs to be adequately measured and theorized so that the particular nature of the phenomenon does not become lost in the quantification process, once again "throwing out the baby with the bath water". This zone is far from being just another manifestation of "Brazilianization" (to use a particularly fashionable adjective and one that which touches us most), a term which, by its provocation, alludes to the transposition of a somewhat spurious model into social systems with a strong regulatory, protectionist and republican tradition. On the contrary, the recurrence of unemployment can occur, as suggested above, in distinct societal contexts and for equally diverse reasons. Nor do I believe the question can be resolved by resorting to the terms of the old debate on marginality which causes so much furor among Latin Americans and Brazilians, in the sense of indicating how the insertion of capitalism into our societies ruptured the social universe and created two cycles of reproduction.

On the contrary; I believe that this zone (which can be more precisely quantified by introducing new categories and forms of unemployment and/or employment) has an origin, a basis, a normative representation, an institutional construction and a subjective introjection which should be sought out in each case, not being regarded exclusively as an epiphenomenon of a certain "globalization" or "internationalization". Despite their recurrence (which leads us to put the onus for explanation on the global plane), the phenomena, while well understood in terms of their normative, institutional and subjective dimensions, not only vary from country to country (and it would be trite to stop here), but only acquire meaning, can only be explained and effectively understood in their entirety, if subjected to comparative studies, i.e. taking the specifics of each societal context into consideration.

Am I challenging the interpretive statute of an economy of work? Not at all; I am only attempting to advance towards a sociology of unemployment, something which I believe it is vital to defend and which is in urgent need of a body of comparative, thematic, theoretical and methodological reflection.



1 Perceptible, for example, in trade regulations and property rights.

2 Here I have deliberately resorted to a typical set of categories derived from current jargon. My aim is to call attention to the fact that such a convergence also takes place in the interpretive systems, giving rise to its own lexicon and grammar (developed in order to deal with what were believed to be the new tendencies of restructured work in globalized contexts).

3 Often without a corresponding increase in productivity (which, in the current scheme of things, leads to a reduction in earnings).

4 Such as part-time work or employment for a specified (usually short) period.

5 Dedecca (1996) refers to the OECD warning, formulated in 1995, that "… even if there were continued economic expansion, the unemployment rate for the OECD area could remain high at around 7 per cent in the year 2000, above the level it reached before the latest recession." (p. 18; my italics).

6 National unemployment insurance systems are full of examples of these specific legitimacy codes, in that it is up to them to define the conditions under which workers will be eligible for benefits and for how long. Maruani and Reynaud (1993) draw attention to a particularly eloquent example: in various countries (e.g. Germany, the UK and the Netherlands), there is a type of legitimacy code that states that married women should receive less benefits than single women. Most probably, this has played a not inconsiderable role in the increasing move into inactivity among married women, exposed to prolonged periods without work, and which results in the prevailing collective representation model of gender-based social relations. I shall return to this issue later.

7 However, its steady increase resulted in a rate of 5% in 2001, an extraordinary level when one considers that, historically speaking, even in times of relatively high unemployment, the average was less than 3% of the working population.

8 Most notably by means of productivity awards.

9 Such flexibility is based on such characteristics as ease of dismissal, unattractive unemployment benefits and an exceptionally low minimum wage, among others, all indicative of a social context that allied working practices (traditionally deregulated) to the growing importance of private welfare (rooted in benefit policies under the direct responsibility of the companies themselves).

10 Canada's unemployment rate doubled in the 1970s and 80s, reaching 9% in 1990.

11 These reached 6%, equivalent to double the French rate, for example (Demazière, 1995).

12 Expressed in more rigid dismissal regulations, relatively high minimum wages and levels of unemployment benefit permitting minimal subsistence.

13 In fact, long-term unemployment was only morally acceptable when it concerned individuals who were physically incapable of work, either through illness (work-related or not) or accident (idem). Until the end of the so-called "golden years", only such individuals were entitled to social security; others out of work were computed as idle, not unemployed. It is hardly surprising that this conception was prevalent at the same time that the work relation was represented, in these same societies, by the stable and protected wage-earning standard.

14 For an analysis of the introduction to the category of "unemployment" in our system of collective representations, see Salais, Bavarez and Reynaud (1986) and Topalov (1994).

15 From employment surveys carried out by the INSEE (Institut National de la Statistique et des Études Économiques).

16 Considered here in the sense attributed by Ledrut (1966): the prospects of remaining in employment, or of obtaining re-employment if dismissed.

17 I would like to diverge somewhat from the main line of argument to record a parallel reflection, which I believe to be interesting. This change in meaning (structural and symbolic) of the unemployment phenomenon is intellectually perceived in a different manner by the different academic communities that have adopted and processed it. In France, for example, it is manifested in a polarization between a "sociology of work" and a "sociology of employment" (systematically reconstructing their roots, theoretical lineages and loyalties; see, for example, the recent reflections of Maruania and Demazière). In Brazilian labor sociology, there was a gradual convergence between studies based on the workplace (a dominant part of our tradition of analyzing work processes and social movements, which arose at the end of the 70s, took off in the 80s and was still strong in the early 90s), and sociological studies of labor markets (important in the 60s and 70s, exemplified by analyses of development, modernization and occupational marginality, which were dropped by mainstream sociology in the 80s when the mantle of these studies was passed to the labor economists). The compromise appears to be woven around a new thematic agenda, some of whose highlights include: analyses of corporate restructuring strategies and their connection with workers' occupational destinies; studies on corporate relocation (greenfields, globalized negotiating strategies) and the impact on regional job markets; and analyses of privatization, lay-off strategies and the mobility of dismissed workers.

18 At this point, I would suggest that we try (at least momentarily) to turn away from the short-term interpretive connotations and the political debate that this concept triggered in the 90s in Brazil, to reflect on the relevancy and heuristic value of the concept’s initial formulation (cf. Ledrut, 1966), as "the probability, more or less high, of an individual seeking work actually finding it". Although difficult, this is occasionally a fruitful exercise, particularly in a discipline such as the sociology of work whose reflective agenda owes so much to, and strongly emulates, that of the social agents in the so-called "world of work".

19 My intention here is to draw attention to our duty to fuel academic debate on a concept, which has a not entirely negligible trajectory in the field of work studies, but which has only recently entered the Brazilian debate, in turn dominated by the polarization of the arguments underlying public vocational training and employment policies as of 1995.

20 In this item, I owe a great deal to the form of problematization that I have been developing recently with other colleagues, first jointly formulated in Demazière, Guimarães, Hirata, Pignoni and Sugita, 2000.

21 Only very recently have the studies begun to question this supposed similitude, showing its limits when considering countries with a relatively similar capitalist development (cf. Gallie and Paugam, 2000).

22 Tracing the history of the concepts can be a fruitful means of perceiving how the problems were being framed and, from this, delineating their solutions. Thus it is not insignificant that the notion of "employability" only achieved a central position here in the discourse governing social policies very recently, although it had been on the foreign bureaucratic menu for some time, being an integral part of the discourse of the international agencies (see the EEC). With even more reason, it would appear to fit very well into government discourse in realities such as the Brazilian one, with its intense occupational transit. As for assistance policies, the accumulated social deficit is notorious and its analysis lies outside the scope of this article's aims.

23 For an analysis of the industrial employment figures, see Carvalho, 1992.

24 In the reflections making up this final part of the paper, I once again turn to the arguments and hypotheses of the project I am then engaged in: "Desemprego: abordagem institucional e biográfica – Uma comparação Brasil, França Japão" (Demazière, Guimarães, Hirata, Pignoni, Sugita, 2000).

25 Once again, turning to the history of the disputes surrounding the construction of these concepts and measurements is fruitful. The controversy surrounding employment measurement in Brazil arose in the midst of a massive growth-and-employment crisis at the beginning of the 1980s, but also from another crisis involving the political legitimacy of the military rulers, exemplified by the ample opposition victory in the 1982 elections. At this precise moment, the official index, produced by the IBGE (Brazilian Institute of Geography and Statistics) and based on ILO norms, was challenged by another type of measurement, tested by a survey of living conditions in São Paulo conducted by the Dieese (Inter Trade Union Department of Statistics and Socio-Economic Studies) . This in turn gave rise, from 1984 on, to a second major statistical survey, now in continuous use for 17 years: the PED (Employment and Unemployment Survey), conceived and carried out by the Dieese and a local government body (at that time the oppositionist government of Franco Montoro), the Fundação Seade, linked to the São Paulo State Planning Secretariat. The Seade unemployment measurement, based on a distinct methodology, introduced two forms of hidden unemployment in addition to the classical open one: that caused by the precariousness of work and that caused by despair. Just to exemplify, according to this alternative measurement, the unemployment rate in 1998 (considered as a reference year, since it was the last one for which we had information when we began our international comparisons in 1999), was not 7.6%, but 18.3%.

26 It is true (and this is supported by empirical research) that less individuals in more precarious social positions or at the lower end of the poverty scale tend to register. It is also true that prolonged periods of unemployment may lead to despair, implying that such individuals cease their active search for work (such a search being a central element in defining the unemployed) and, as a result, their benefits may be withdrawn and they may be disqualified from the institutional protection system. But these are not the main drivers governing the conduct of the "clientele" vis-à-vis the French unemployment institutionalization system. For this reason, it is estimated that more than 80% of the unemployed register with the ANPE.

27 In Japan, the "S" curves are still a good model for representing female employment rates. For a meticulous and richly-detailed study, see Sugita (2002). In France (and Europe in general), they tend towards male activity curves (in the French case, this is clear; see Maruani, 2000), and even in Brazil the strong recent growth in female employment rates (see Bruschini, 1998; Lavinas, 1998) is running parallel with a change in the profile of those women in the job market. It is no longer the young, single and childless that are fueling the employment rates, but older married women with children (Bruschini, 1998; Guimarães, 2001).



ALVES, Edgard Luiz Gutierrez & SOARES, Fábio Veras. (1997), "Ocupação e escolaridade: tendências recentes na Grande São Paulo". S. Paulo em Perspectiva, 11(1), Jan.-Mar.

AMADEO, Edward; BARROS, Ricardo Paes de; CAMARGO, José Márcio; MENDONÇA, Rosane; PERO, Valéria & URANI, André. (1993), "Human resources in the adjustment process". IPEA, Série "Textos para Discussão", 317, Oct.

BASTIN, N. (1987), "De la gestion déliberée du paradoxe". Grand Angle sur l'Emploi, 2, apud Demazière, 1995.

BÉNOIT-GUILBOT, Odile & GALLIE, Duncan (eds.). (1992), Les chômeurs de longue durée. Actes Sud, Arles, apud Demazière, 1995.

BRUSCHINI, Cristina. 1998. "Gênero e trabalho feminino no Brasil: novas conquistas ou persistência da discriminação? Brasil, 1985 a 1995". Presented at the seminar, "Trabalho e Gênero: Mudanças, Persistências e Desafios", ABEP/NEPO, Campinas, Apr. 14-15.

CARVALHO, Ruy de Quadros. (1992), "Projeto de primeiro mundo com conhecimento e trabalho do terceiro?". Comunicação apresentada ao GT "Processo de Trabalho e Reivindicações Sociais", Encontro Anual da Anpocs, Caxambu.

CARDOSO, Adalberto. (2000), Trabalhar, verbo transitivo: destinos profissionais dos deserdados da indústria automobilística. Rio de Janeiro, Editora da FGV.

CARDOSO, Adalberto; COMIN, Álvaro & GUIMARÃES, Nadya. (2000), "Os deserdados da indústria: reestruturação produtiva e trajetórias intersetoriais de trabalhadores demitidos da indústria brasileira". Presented at the III Congreso Latinoamericano de Sociología del Trabajo, Buenos Aires, May (subsequently published under the same title in the Revista Latinoamericana de Estudios del Trabajo, Buenos Aires, ALAST, year 7, 13: 17-52).

CASTRO, Nadya Araujo. (1995). "Modernização e trabalho no complexo automotivo brasileiro", in _________. A máquina e o equilibrista: inovações na indústria automobilística brasileira, São Paulo, Paz e Terra.

CASTRO, Nadya Araujo; CARDOSO, Adalberto Moreira & CARUSO, Luis Antonio. (1997), "Trajetórias ocupacionais, desemprego e empregabilidade: há algo de novo na agenda dos estudos sociais do trabalho no Brasil?". Contemporaneidade e Educação, 2.

DEDECCA, Claudio Salvadori. (1996), "Desregulação e desemprego no capitalismo avançado". S. Paulo em Perspectiva, 10 (1): 13-20, Jan.-Mar.

DEMAZIÈRE, Didier. (1995), La sociologie du chômage. Paris, Éditions La Découverte.

_________. (1995a), Le chômage de longue durée. Paris, Presses Universitaires de France.

DEMAZIÈRE, Didier; GUIMARÃES, Nadya A.; HIRATA, Helena; PIGNONI, Maria Teresa & SUGITA, Kurumi. (2000), Chômage: approches institutionel et biographique: une comparaison Brésil, France, Japon. Research project, Paris.

DIPRETE, Thomas A. & KRECKER, Margaret L. (1991), "Occupational linkages and job mobility within and across organizations". Research in Social Stratification and Mobility, 10: 91-131, JAI Press Inc.

D'IRIBARNE, P. (1990), Le chômage paradoxal. Paris, Presses Universitaires de France.

ESPING-ANDERSEN, G. (1990), The three worlds of welfare capitalism. Cambridge, Polity Press.

_________. (1999), "Les trois mondes revus", in _________. Les trois sociétés de l'État-providence. Paris: Presses Universitaires de France (Collection "Le Lien Social").

FERREIRA, Cândido; HIRATA, Helena; MARX, Roberto & SALERNO, Mario. (1991), "Alternativas sueca, italiana e japonesa ao paradigma fordista: elementos para uma discussão sobre o caso brasileiro", in Anais do SeminárioModelos de Organização Industrial, Política Industrial e Trabalho, São Paulo, ABET, Apr.

FOGAÇA, Azuete & SALM, Cláudio. (1993), "Estudo da competitividade da indústria brasileira: competitividade, educação e qualificação. Nota técnica temática do bloco condicionantes sociais da competitividade". Campinas, mimeo.

FREYSSINET, Jacques. (1984), Le chômage, Paris, La Découverte.

GADREY, J. ; JANY-CATRICE, F. & RIBAULT, T. (1999), France, Japon, Etats-Unis: l'emploi en détail. Essai de socio-économie comparative. Paris, PUF.

GALLIE, Duncan. (2001). "Unemployment, welfare regimes and social exclusion". Paper presented at the CNRS/ESRC Anglo-French Seminar on Social Exclusion, London, Dec. 11-12.

GALLIE, Duncan & PAUGAM, Serge. (2000), Welfare regimes and the experience of unemployment in Europe.Oxford, Oxford University Press.

GAULLIER, X. (1999), Les temps de la vie: emploi et retraite. Paris, Esprit.

GUIMARÃES, Nadya (2001), "Laboriosas, mas redundantes", Estudos Feministas, 9 (1): 82-103.

_________. (2002), Caminhos cruzados: estratégias de empresas e trajetórias de trabalhadores. Tenure thesis for the Department of Sociology, São Paulo, Universidade de São Paulo.

HIRATA, Helena (org.). (1992), Autour du modèle japonais. Paris, L'Harmattan.

HIRATA, Helena & ZARIFIAN, Philippe. (1994), "Le modèle français sous le régard du Japon: l'example de l'agro-alimentaire". Atas do Deuxième Rencontre International du GERPISA (Groupe "Hybridation"), Paris, Jun.

LAUTIER, Bruno. (1987), "Fixation restreinte dans le salariat, secteur informel et politique d'emploi en Amérique Latine". Revue Tiers Monde, 110.

LAVINAS, Lena. (1998), "Evolução do desemprego feminino nas áreas metropolitanas". Rio de Janeiro, s.p.i., mimeo.

LEDRUT, Raymond. (1966), La sociologie du chômage. Paris, Presses Universitaires de France.

MARUANI, Margaret. (2000), Travail et emploi des femmes. Paris, La Découverte.

MARUANI, Margaret & REYNAUD, Emmanuèle. (1993), Sociologie de l'emploi. Paris, Éditions La Decouverte (Collection Repéres, no. 132).

OUTIN, J. L. (1990), "Trajectoires professionelles et mobilité de la main-d'oeuvre: la construction sociale de l'employabilité". Sociologie du Travail, 4.

SALAIS Robert; BAVAREZ, N. & REYNAUD, B. (1986), L'invention du chômage. Paris, Presses Universitaires de France.

SILVA, Elisabeth Bertolaia da. (1990), Refazendo a fábrica: contrastes da indústria automobilística no Brasil e na Grã-Bretanha. São Paulo, Hucitec.

SUGITA, Kurumi. (2002), "Emploi atipique et chômage dans la société japonaise", Lyon, IAO/Univ. Lyon, mimeo.

TOPALOV, Christian. (1994), Naissance du chômeur, 1880-1910. Paris, Albin-Michel.



* Revised version of the paper presented to the XXV Annual Anpocs Meeting, Caxambu, October 16-20, 2001, as part of the Thematic Seminar "Workers, Unions and the New Social Order" (session entitled "Theories and Configurations of the Working Class Today").