Breadwinner Models and Historical Models PDF

Breadwinner Models and Historical Models. Transitions in Labour Relations and Labour Markets in Belgium, 19th - 20th Centuries Eric Vanhaute...

Breadwinner Models and Historical Models. Transitions in Labour Relations and Labour Markets in Belgium, 19th - 20th Centuries Eric Vanhaute Universiteit Gent Kostwinnersmodellen en historische modellen. Transities in arbeidsverhoudingen en arbeidsmarkten in België, 19de en 20ste eeuw Het ‘kostwinnersmodel’ is een geïntegreerd concept dat vraagt om een geïntegreerde benadering op de drie maatschappelijke niveau’s: micro (de kostwinnersfamilie), meso (de kostwinning op de werkvloer) en macro (de kostwinnersideologie). Een gecombineerde kijk op de interactie tussen individuele en gezinsstrategieën en tussen arbeidsmarkten en overheidspolitiek is noodzakelijk. Ons onderzoek beoogt de constructie en deconstructie van een kostwinnersmodel in de context van de Belgische en Vlaamse samenleving. Ik concentreer mij op de ‘timing’ van de historische tijd, zoals gedefinieerd door Tamara Hareven: ‘de bredere processen aan dewelke de tijdsspecifieke familiepatronen zijn gekoppeld’. De geschiedenis van de arbeid tussen 1750 en 2000 wordt gemarkeerd door drie periodes van transitie, die de contouren van de arbeidsorganisatie vastleggen voor de volgende ‘eeuw’. Vanaf de tweede helft van de 18de eeuw, parallel met de economische veranderingen getypeerd als de ‘eerste industriële revolutie’, verslechteren de arbeids- en inkomensperspectieven in zowel de stedelijke als de rurale samenlevingen. Dit beperkt de marges van de gezinsvorming. Ondanks de intensifiëring van de arbeidsparticipatie van vrouwen en kinderen is het evenwicht in het gezinsbudget fragieler dan ooit tevoren. De geleidelijke vervanging van een familiegebonden productiemodel door een open en onzekere arbeidsmarkt blokkeert voor een groeiend aantal volwassenen de perspectieven op huwelijk en gezinsvorming. In stedelijke samenlevingen worden de beschermende functies van de familiebanden meer en meer vervangen door bredere overlevingsnetwerken. Het ‘extensieve’ productiemodel van de eerste industriële revolutie, steunend op lage lonen en een snelle vervanging van arbeid, put de arbeidskracht uit en tast de reproductieve kracht van de gezinnen aan. Vanaf het einde van de 19de eeuw wordt deze flessenhals opgeheven door de overgang naar meer ‘intensieve’ modellen gebaseerd op hogere mannelijke lonen, een beschermende wetgeving (die vrouwen en kinderen uit de sterke en zichtbare industriële sectoren weert), een betere scholing, een betere behuizing en hygiëne. De huisvrouw is gedwongen zich terug te trekken in de onbeschermde/informele niches van de arbeidsmarkt, zoals huishoudelijke diensten, onderaanneming en landbouw. De uitbouw van een publiek transfersysteem van kinderbijslagen en belastingvoordelen is de finale stap in de realisatie van de mythe van het kostwinnersloon. De huwelijksmarkt breekt open, gevolgd door een beperking van de vruchtbaarheidsratio. Vanaf de jaren 1960 verschijnen (gehuwde) vrouwen terug op de formele arbeidsmarkt, in de eerste plaats in de nieuwe industrieën en de diensten. Deze herintrede gaat echter gepaard met een deregulering van de arbeidsmarkt door een toename van onregelmatig en tijdelijke jobs. De huwelijksvruchtbaarheid gaat nog verder achteruit, samen met het terugdringen van de ‘traditionele’ gezinspatronen. Breadwinner Models and Historical Models Problems, Programmes and Models 59  De 19de eeuw wordt gekenmerkd door een steeds meer ingrijpende deregulering, binnen familieverbanden, op de werkvloer, in het overheidsbeleid. Het vrijmaken van de arbeidsmarkt wordt onderbouwd door een offensieve ideologie van arbeidsdisciplinering. Het einde van de eeuw markeert een verandering in de richting van een (20ste-eeuws) model van regulering: beschermende maatregelen op de werkvloer en een nieuwe ideologie van het kostwinner/huisvrouw-model gaan samen met een hechter gereguleerde arbeidsorganisaztie (Taylorisme-Fordisme). Dit komt opnieuw op losse schroeven te staan vanaf het laatste kwart van de 20ste eeuw, wanneer nieuwe vormen van flexibele productie ondersteund worden door een minder stringente arbeidsregulering, een fragmentatie van de collectieve samenlevingsvormen, een individualisering van consumptie en een commercialisering van de vrije tijd. Het ‘kostwinnersmodel’ in België is verbonden met een specifieke wijziging in de ‘historisch tijd’, een verandering die een hoge publieke welvaart koppelt aan een stringente arbeidsregulering en massieve fincanciele overheidstranfers. Eric Vanhaute was born in 1959 and obtained his Ph.D in 1989. He teaches econom- ic and social history at Gent University (Belgium). His original research area was the rural and agricultural history of the 18th till the 20th centuries, focussing on labour and income strategies of peasant families in Flanders. His major fields of interest are now changing labour relations in and between families, rural crises (esp. the mid-19th cen- tury potato crisis), life course analysis, the social stratification debate, structural history and world-systems analysis. He is currently working on a book on the transformation of rural society in Flanders in the 18th and 19th centuries. MODELS AND STRUCTURES In his description of the English working class of the mid-19th century, Friedrich Engels expressed serious concern about the preservation of the family: “Thus the social order makes family life almost impossible for the worker. […] Neglect of all domestic duties, neglect of the children, especially, is only too common among the English working-people, and only too vigorously fostered by the existing institutions of society” 1. The fast processes of change characterising the ‘First Industrial Revolution’ caused many contemporary commentators to write down their impressions. From Malthus to Marx, they were all convinced that the societal shifts they witnessed were revolutionary in nature. However, long-term historical research tends to minimise ‘revolutionary’ shifts. Indeed, recent research on the interaction between family formation and the organisation of pro- duction and work stresses both continuity and change during periods of strong economic, social and cultural changes. Katrina Honeyman pointed out that British industrialisation made use of the existing gender hierarchy in the work place, at the same time reconstruct- ing the different gender roles and identities at the interface of home and work 2. In this chapter I will look for structural changes in the relationship between work, family and gender in 19th and 20th century Belgian society. This long-term approach only makes sense when it looks for structures, processes and models, to reconstruct ‘the larger social processes’ which form ‘historical time’ 3. Working with models, though, still differs from a 60 Eric Vanhaute model-based approach. In the first case we attempt to understand social order and its evo- lution by comparing them to other forms of social behaviour. The aim of this exercise is to arrive at an empirical interpretation of a social system. Much attention is devoted to recur- rence versus singularity; ‘historical time’ holds centre stage here. In the second case the analysis of human behaviour starts from a model designed in advance. There, time is not historical, but a background variable devoid of content. I will attempt to ‘historicise’ a widely-accepted social science model: the breadwinner model. In historical and sociological literature notions like ‘the breadwinner ideology’, ‘the breadwinner model’, ‘the breadwinner system’, ‘the breadwinner norm’, ‘the breadwinner ideal’, ‘the breadwinner family system’, are used rather indiscriminately and smoothly, without ever establishing when something becomes an ideology, a standard, an ideal or a model. My ambition is not to give a universally applicable definition of these notions, which would not make sense, given the attempt to place these concepts in their historical context. I want to make clear, however, how a historical model, in this case the ‘bread- winner model’, can be placed within the context of Belgian history. I start by discussing a few definitions of production and reproduction, labour and labour relations, especially the relations within the breadwinner model. The way in which these notions are defined determines the outlook of social history. Afterwards I will briefly exam- ine the issue of sources and the way in which the latter yield information. We will identi- fy a ‘Belgian’ breadwinner model by interpreting two statistical parameters. Obviously this is a selective angle. Finally I want to place the breadwinner model within the context of the social dynamics of the last two centuries. I see these dynamics as the development/tran- sition/confrontation of different historical society patterns or models. The search for a time-scale showing the advent, the development and the decay of a bread- winner model ‘has to be sensitive to variations by region, trade and industry’ 4. A long term structural analysis is a complementary approach. Moreover, comparative analysis seems indeed to be the only way to approach the mentioned variations. PRODUCTION, REPRODUCTION AND LABOUR Following the logic of academic disciplines, notions like family and household formation are still thought to belong to the department of ‘demography’. In the course of many years, demographic history has acquired an impressive knowledge about behaviour as determined by the individual, the family or the group. A more integrated approach, more sensitive to interaction with social and economic variables in family formation as well as to the rela- tion with macrostructures, has proved much more complex. The rigid boundaries between social research disciplines have always proved bothersome to demography. Thus we remain with the consensus, explicit or implicit, that ‘demographic behaviour’ is about reproduc- tion, whereas (wage) labour is always associated with production. This distinction has become untenable 5. Where labour generates the survival goods required and, consequent- ly, is necessary for ‘reproduction’, demographic choices and family life guarantee the ‘pro- duction’ of this labour, both literally (births) and in a more figurative sense (raising chil- dren, feeding them, providing rest, training, affection). It does not seem very sensible to establish a hierarchy between various forms of production, they are inextricably linked. Breadwinner Models and Historical Models Problems, Programmes and Models 61  Hence, demographic analysis is not only a story of ‘production relations’ (man, woman, children), but also of input (income), output (labour) and surplus value. The choices made by men, women and families about living together, having children and raising them are part and parcel of their general strategies about labour and income. The notion of labour holds centre stage in the concepts of production and reproduction. By labour I understand each ‘productive’ human activity. Productive activity adds user value to goods or services, however this may happen. Work or labour modes are not caused by ‘the call of the market’. Historically most labour has taken place outside what we call the (official) labour markets. Labour markets are therefore not natural, universal phenom- ena, but the historically contingent products of the struggle to gain control over labour conditions. In other words, the organisation of the labour market can only be understood in the social context of time and place, rather than in the timeless logic of individual pref- erences, technological stimulants, market forces or ideology 6. As capitalist labour markets increasingly monopolise labour, the notion of labour narrows 7. Unpaid labour, household labour, non-formal labour are excluded from the modern defini- tion of labour. In other words, the notion of labour moves from an inclusive one to an exclu- sive one, responding to the new, strongly segregated labour markets. The use of labour as an exclusive notion almost always carries a connotation linked to class, group and gender. Some productive activities receive increased value while others are reduced in value. In this context, various forms of labour are defined antagonistically: tra- ditional vs. modern, skilled vs. unskilled, paid vs. unpaid, individual vs. collective, male vs. female, respectable vs. not respectable, independent vs. dependent. Such distinctions are either based on class or group (unskilled, traditional, etc.) or based on gender (household work, women’s labour, etc.). The exclusive connotation of labour is an ideological weapon used to create a hierarchy among social groups and among men and women, regulating new forms of social organisation. The way in which this happens may vary according to time and place (nature of the capitalist organisation, core versus periphery, etc.). WHAT IS A BREADWINNER MODEL? Social actors (men and women, families, social groups, businesses, government and so forth) continually make choices within their particular societal context. Such choices constitute the basis of strategies and behaviour. When strategies and patterns of behaviour become standardised in particular locations and at given times, we may refer to a model. Although a standard may leave some room for other patterns, it does set the agenda as regards demo- graphic, social and economic issues, even right up to politics and ideology. However this only happens when the pattern has achieved sufficient specific weight to set the agenda. The standard model then turns into a role model, a societal standard, a point of reference. Historians tend to investigate, not so much the standard as such, but deviations, resistance and hidden or open deviance in relation to the standard. Obviously, this is a legitimate enterprise, but, to my mind, it is sensible only if the standard is mapped out too. Each social model results from the interaction between (micro)individual choices and (macro)societal order. Thus the model – like the breadwinner model – will be a multilay- 62 Eric Vanhaute ered integrated notion linking individual and family strategies to shifts in the work place and changes in the public order. The ‘breadwinner model’ has to be understood as an inte- grated concept which can only be studied in an approach integrating the three societal lev- els: micro (‘the breadwinner family system’), meso (‘breadwinning’ in the workplace) and macro (‘the breadwinner ideology’). A combined view on the interaction between indi- vidual and family strategies, labour markets and public policies is necessary to reconstruct ‘the ways larger social structural forces constrain the repertoire of available adaptations’ 8. A few recent contributions have revalued the breadwinner’s concept in historical research 9. They have identified the busy scientific crossroads where research into fami- ly relations and household labour strategies is conducted. ‘The rise of the male bread- winner family’ confronts us with economic (the logic of capital), social (the conflict between labour and capital, between groups of workers, between man and woman) and political (social and tax policy) statements and especially with the need to integrate such motivations. The debate is a strongly ideological one, often taking the shape of a capi- talist, middle-class or male offensive. Moreover the research model includes micro- processes (household strategies) and macrostructures (labour markets). Nevertheless (or, perhaps, therefore) we have not seen, up to now, the emergence of an integrated vision on the success of the breadwinner model. What are the main components of such a breadwinner model? – a strongly separated allocation of labour: the external, paid labour is performed by to the male head of the family; the internal, unpaid labour is left to the housewife; – rigid family relations, with the husband filling a paternalistic father role and acting as the family representative on the public forum; the wife remains scarcely visible, inside the fam- ily, the children become more dependent, among other reasons, because of compulsory school attendance; – at the workplace we have a strongly segregated and regulated work organisation (tay- lorism-fordism); – the model is fostered by an active and regulating government; – an ideological offensive regarding the social position of men, women and children legit- imates the new relations between the sexes. HOW TO MAP OUT A BREADWINNER MODEL? It is no simple matter to map out family patterns and household strategies. In the course of many years, historical demography has accumulated vast and varied expert knowledge on the analysis and explanation of population patterns in families and small communities. Integrating such knowledge into economic, social and political-ideological analyses has proved to be much more problematic. Unavoidably researchers have been using the serial processing of average coefficients or the analysis of some cross-sections. Both methods are rough tools. Coefficients hide the dynamics of life. Averages wipe out the variations in the family cycle. Trends emphasise a statistically created stability and uniformity. Moreover the main sources, i.e. the population registers, civil status registers, population censuses, are Breadwinner Models and Historical Models Problems, Programmes and Models 63  strongly subject to standardisation and only give a partial view of demographic behaviour and decisions. Deviant behaviour, unconventional modes of living together, extended fam- ily bonds… in their search for averages researchers run the risk of being blind to the great variation of patterns of ‘living together’ among people. A classical case in point is the obsession with the average number of family members. Extensive research demonstrates that the average ‘density’ of the European family scarcely changed from the Middle Ages to the early 20th Century, with the typical family numbering 4 to 5 persons. This average hides more than it reveals. Indeed, it does not take into account the family cycle, whereby family composition may vary considerably, nor does it look at whether or not tempo- rary/permanent inhabitants are registered, and, last but not least, it overlooks the types of partnership outside the conventional family pattern. We know, for instance, that the 19th century industrialisation did not imply the confirmation of the core family. On the con- trary, rapid economic and social change reinforced the bonds of the extended family. Occupational statistics are another example, based on the ‘mother’ of all statistics, the pop- ulation census 10. If we use a broad definition of labour input, we may assume that an aver- age of two-thirds of the population was ‘potentially’ labour-active. Occupational figures hardly ever yield an official activity level of 65%. In Belgium in the middle of the 19th cen- tury, this figure stood at 57%, but a hundred years later it is reduced to about 30%. Is this a development in the ‘real world’, did half the potential active population really drop out of the labour force? Or does this reduction rather reflect changed and narrowed counting criteria? Statistics are the product of a certain reality, but simultaneously they also shape that real- ity. They shape and are shaped and in this sense they do not differ from other sources. Paid female labour is a case in point: it becomes increasingly invisible in the statistics between 1850 and 1950. This partially results from counting agent choice (that is, choices made by those responsible for the counting), but it also reflects the actual process of social margin- alisation of the labour-active woman. Sources only yield their secrets if put into context: in time, in place and in interaction between social actors 11. A BREADWINNER MODEL IN BELGIUM? The first graph (Graph 1) presents a general overview of the rates of labour participation on – what can be called – the ‘formal labour market’. Most striking in these figures is the growing gap, between 1850 and 1950, between the group of potentially active women and the female participation rate in the ‘external’ labour market. In the middle of the 19th century, one out of every three women between 15 and 64 years old is part of the ‘housewife labour reservoir’. One century later, a majority of women belong to this cate- gory. From the 1960s onwards, this long-term trend turns. Nowadays, only 29% of the potentially active women are professionally not active, against 56% in 1950 and (maxi- mum) 43% in 1900. The expulsion of female labour from the formal markets affects married women first. Around 1900 only 18% of the married women formally have a profession. In centralised industries only 39% of the female labour force is over 21 years old, and barely 17% is mar- 64 Eric Vanhaute Graph 1. Registered labour participation in Belgium, 1850-1990, based on professional censuses. Percent of population 15-64 years old 12. ried. In 1950 over 40% of the women in the age group of 15 to 24 years is active in the labour market (the figure for men is 75%), a ratio that drops to below 30% in the age group of 25 to 34 years (men: +90%). This means that less than three out of ten women in that age group are professionally active. Four decades later this participation ratio has risen to more than eighty percent. This re-entry of women in the labour market is first of all the re- entry of married women. In 1950 only 15% of all married women are active on the labour market, while in 1990 the figure rises to 60%. Nowadays the majority of women who re- enter the labour market are older than 35 years old. Due to the expansion of part-time work, the increase of the total female labour input is smaller. Almost half of the active women do not have a full-time job (44%), whereas only a small fraction of their male part- ners (16%) have a part-time job 13. Graph 2 shows the women’s share of the official labour market, as a total and then subdi- vided according to the three classical sectors. In general, the women’s share has decreased from 42% of the registered active people in 1850 (with a very equal distribution over the three sectors) to 28% a century later. Today this share has risen again to almost 40%. The drop in the official labour participation of women occurred first in the industrial sector, with the marginalisation of the production in cottage industries, but also, in the late 19th century, with the expulsion of women from the factories. However, the same period saw the feminisation of the expanding trade and services sector (especially with the strong growth Breadwinner Models and Historical Models Problems, Programmes and Models 65  Graph 2. Women’s share on the registered labour market. Belgium, 1850-1990. in the numbers of female servants). After 1900 women also lost ground in agriculture and the tertiary sector. The last column shows that the re-entry of women into the official labour market after 1950 was almost exclusively in the service sector. Again, as in 1900, women constitute the majority of workers there. 66 Eric Vanhaute The ratios of labour participation point at an exclusion of the ‘housewife’ from the formal labour market between 1850 and 1950, and, perhaps, to an advance of the breadwinner family. Can this be confirmed by budget inquiries? Until the first half of the 20th century, the wage income is a shared responsibility of all family members. Throughout the 19th century, the weight of the female wage is decreas- ing. This is compensated, first, by an increase of the earnings of the children, and from the end of the century onwards, by higher male wages. What is striking for the first half of the 20th century is not only the marginalisation of wage income from wife and children, but, even more so, the sharp increase of non-wage income. Public allowances and returns of savings and investments diminish the share of the male pay-packet in the family budget. After 1950, the re-entry of the housewife on the external labour market is reflected in a higher share of the female wage. These figures too point at the predominance, in the first half of the 20th century, of the breadwinner family. But the data can be read in a different way. In fact, the male wage never represented more than three quarters, and in most cases, not more than two thirds of the total family income. In ‘regular’ working-class families, the male wage-packet was never large enough to make family ends meet (graph 4). Before the interwar period, extra income via the earnings of wife and children was necessary. The retreat of young and female workers from the formal labour markets was only possible after a massive public subsidising of the family budget (or the male wage) through a range of allowances (family allowances, unemployment benefits, health insurance, pension allowances) and tax benefits. The relatively fast rise of the male wage from the end of the century onwards encouraged the withdrawal of women and young Breadwinner Models and Historical Models Problems, Programmes and Models 67  children from the external labour market, however, without ever reaching the level of the ‘family wage’ 15. The re-entry of the married partner in the labour market has had a strong effect on the living standard. In roughly the last two decades, the households with two wage earners have set a new standard of consumption 16. The removal of women from the registered labour market from the second half of 19th cen- tury mainly boils down to the expulsion of married women. In 1900 only one out of every five married women was officially occupationally active, undoubtedly a substantial reduc- tion compared to half a century earlier (although we are unable to supply general figures here). This movement goes hand in hand with a more explicit ‘protective’ attitude of leg- islators toward the married housewife. After a long preparatory phase during the last decades of the 19th century, 1919 saw the voting of the first comprehensive law restricting (paid) female and child labour. The family had to be protected from excessive absence by the mother. During the recession of the thirties the formal labour participation of women dropped to an all-time low. In 1934-1935 there was an unusually intensive legislative activ- ity to limit the presence of women on the labour market: per company, a quota of female labour was established, a recruitment ban was decreed for women in government depart- ments and agencies, jobless women and jobless partners of working women no longer received benefits, or if they did, lower ones, and tax policy summed the income of married couples. This deterrence policy was continued after World War II, but now especially through ‘rewarding’ the housewife. Mothers without independent income received a ‘fam- ily benefit’ and higher child benefits. By 1950 the level of female participation had dropped to a mere 15%. On the other hand, the re-entry of women on the labour market during the last three decades also mainly con- cerned married women. Today half of them are again occupationally active. This change matches a u-turn in government policy. From the second half of the 1950s the Belgian economy went into a fast growth, causing an enormous demand for additional labour. Now wage labour by women had to be stimulated. First the benefits for housewives were abol- ished and then the system of discriminatory unemployment benefits. The law of 1967 on child and female labour no longer sought to discourage working women, but rather to ‘pro- tect’ them. This led, among other things, to compulsory maternity leave and the restric- tion of employers’ right to fire the employees concerned. The year 1975 is a milestone in the painfully slow process to achieve equal treatment of men and women in the work place. The Belgian government put into practice a European directive setting a general ban on wage discrimination in any shape or form based on gender. Three years later this discrimi- nation ban was extended to the whole field of labour (labour conditions, vocational train- ing, career opportunities, etc.). During the 1980s this principle of equality was further developed, with the introduction of parents’ leave, career break schemes, positive actions to recruit women. The 1988 tax reform law separated the incomes of partners in two-earn- er families, thereby removing the tax ‘disincentive’ on the second wage earner (the woman in most cases). The above analysis leads to the following conclusions. First we observe that the accelerat- ed entry of women in the formal labour market during the last two to three decades is not unprecedented, but, in a sense, a return to an earlier, more general pattern. Moreover, even in the age of the breadwinner model and government disincentives – during the second 68 Eric Vanhaute and third quarter of the 20th century – a substantial part of the women remained formally active: one out of three potentially active women, one out of seven married women. Here we necessarily have to leave out much labour done in the uncharted informal and tax-hid- den labour circuits. Thirdly the breadwinner model could only be generalised through massive government subsidies. Part of the expenses of households was ‘collectivised’ through a system of direct and indirect public financing. FROM AN EXTENSIVE TO AN INTENSIVE LABOUR MODEL Any analysis of changing labour relations has to be linked to the successive phases in the process of capital accumulation, that is, contextualised in changing historical time. The history of capitalist accumulation is marked by successive periods of transformation and stabilisation. In Belgium’s economic and social history we can trace three such periods in the last two centuries. The transition to 19th century industrial society goes hand in hand with a deregulation on all societal levels. The ‘extensive’ production model of the first industrial revolution is based on low individual wages and a high labour turnover. The female participation in the ‘external’ labour market has never been higher, except for the last decades, than in the first half of the 19th century 17. Labour input per worker is more intensive than in the periods before and after 18. In the Ghent cotton industry in 1816 only 30% of the labour force were men older than 16. In 1846 one quarter of the workers was younger than 16, a percentage which only slowly decreased after 1870 19. Women in contrast strengthened their presence in the textile factories, from 34% in 1846 to 50% in 1910. Adult male workers were never the majority in 19th century Ghent textile industry 20. Female and child labour was no pre- rogative of textile towns. In Antwerp, the expansion of a commercial economy was accom- panied by the withdrawal of working women in the unstable labour markets of ‘sweating’ industries, personal services and other casual labour 21. Female (and child) labour was mas- sive, cheap, flexible and reliable and was the main ‘buffer’ against the uncertainties of the rapidly growing industrial capitalist system, both within the household and as ‘a reserve army’ on the labour market 22. The deterioration of labour and income prospects in both rural and urban societies narrows the margins of family formation. In spite of an intensification of female and child labour participation, certainly in the first decades of the industrial revolution, the balance in fam- ily budgets was more fragile than before. The gradual substitution of a home-based production model by open and unstable labour markets blocked the perspectives of mar- riage and ‘regular’ family formation for a growing number of adults. In the expanding city the protective functions of the household became more and more the responsibility of broader networks 23. Public authorities, especially in Belgium, followed the liberal adage of non-intervention. Until the beginning of the 20th century a central social policy was almost non-existent 24. The last quarter of the 19th century marked a new turning point in the history of western society. Together with a ‘second industrial revolution’, a new societal model of increasing Breadwinner Models and Historical Models Problems, Programmes and Models 69  regulation took shape. The bottleneck of the ‘19th century model’, the exhaustion of the labour market and the family reproductive capacities, was avoided thanks to the transition to an ‘intensive’ model based on higher male wages, protective legislation (banning women and children from heavy industrial sectors), better schooling, a higher quality of housing and living. The housewife was forced to withdraw to unprotected/informal niches of the labour market (domestic service, ‘sweating’, industries, agriculture) and, during the inter- war period, to household labour. The development of a public transfer system of family allowances and tax benefits was the final step in the realisation of the myth of the breadwinner wage. The costs of capitalist growth were more and more ‘externalised’, borne by the public governments. The marriage market opened up, followed by a restriction of the fertility rate. In short, this period was marked by: a) The withdrawal of the (married) woman from the formal labour market. The participa- tion rate of women between 15 and 65 years old dropped from 67% in 1850 to 39% a cen- tury later. In 1900 already 43% of the ‘potentially active’ women were registered without a official occupation, against only 3% of their male counterparts. b) A sector-based segregation in the labour market. The decline of the female labour par- ticipation was sharpest in the strong and in the new industries 25. On the other hand, women exercising market activities were over-represented in more invisible sectors such as cottage industries (‘sweating’ industries such as lace-making and clothing), personal ser- vices and family farm labour. The ‘gendering of jobs’ was accompanied by segregation with- in the female labour force. About 1900 two thirds of the unmarried female industrial work- ers were factory labourers, one third was engaged in cottage industries. For married women the ratios were reversed. In 1896 61% of all workers in cottage industries were female, in 1910 this was already 76%. Until World War I, an ever increasing part of the young female labour reservoir was drawn into the sector of the personal services. Between 1850 and 1910 the number of (rural) maids serving in urban households quadrupled. More than 80% was not married. In the 20th century, the group of domestic servants systematically shrank, strongly accelerating the diminution of the labour possibilities for women. c) The institutionalisation of the relations of dependence via a dual process of discourag- ing or restricting female access to an (external) job, and by denying women equal treat- ment in public insurance 26. d) The ideologisation of breadwinner and household labour. This had mainly a ‘megaphone’- function, through the male dominated labour movements 27, through justifying initiatives for protective legislation 28, through the propaganda of hygienist movements, the professionali- sation of ‘housewifery’ in schools, the rhetoric of nationalist movements, and so forth. All are different sides of the same picture. The segregation of the labour market along gen- der lines goes hand in hand with the ousting of women from the formal labour market, in the first place married women, in the interwar period also unmarried girls. In 1950 the par- ticipation rate of girls between 15 and 24 years had dropped to 40% (men: 75%). In the age group 25-34 years the ratio was below 30%, while more than 90% of the men of the same age had a registered occupation. 70 Eric Vanhaute The public policy of subsidising the breadwinner family was promoted by an ideologically based construction of a gendered labour concept, along the lines of external and internal, or paid and unpaid labour. In the words of Honeyman and Goodman 29: “marital status took on a new significance as protective legislation, the cult of the family wage, and the ideol- ogy of domesticity interacted to emphasise gender inequality in the labour market”. A NEW TRANSITION During the last quarter of the 20th century the single-earner model quickly lost its attrac- tion. It is significant to see the re-emergence of ‘looser’ relations in three societal fields: the weakening of the strict social relations linked to the regulated consociational economy, the deregulation in the work place and freer choices regarding modes of living together. The end of the 20th century’s bipolar labour model goes hand in hand with a greater diversity in labour, income and cohabitation forms. Where the traditional ‘breadwinner model’ rest- ed on the foundation of a ‘passive welfare state’ (with low labour participation and high social protection), the last two decades have seen an unmistakable shift to an ‘active com- bination model’ 30. This new project seeks to combine high (male and female) labour par- ticipation with high social protection. This happens simultaneously with fundamental shifts in the work place, in government actions and within households. More flexible mod- els replace the rigid Taylor-Ford type labour organisation. Recently the government has been resolutely reducing participation-inhibiting measures (in tax and social security poli- cies, etc.) and embarking on an ‘activating’ policy. The ‘traditional’ family has been losing ground increasingly, with various types of cohabitation being combined again in the course of people’s lives. In 1991 one household out of three no longer matched the ‘traditional’ two-partner model. In 1950 unconventional cohabitation only occurred in one out of every six cases. Moreover the trend toward a more equal internal and external division of labour is moving ahead. Children will again become active partners in the upbringing and cohabitation process. Since the 1960s, (married) women have re-appeared on the labour market, applying for jobs in new industries and the services sector. The ‘housewife labour reservoir’ (the num- ber of potentially active women [16-64 years old] who have no official occupation, are not officially unemployed, and do not have the status of student or pensioner) shrank from almost 50% in 1950 to barely 10% at the end of the 1990’s. This re-entry seems to have been a ‘lever’ for the entrepreneurs to deregulate the labour market (irregular and part-time work). After a century-long trend of modulating household income patterns towards the breadwinner/housewife model, the two last decades of the 20th century show a new vari- ety in household labour and income patterns. These seem to be closely linked to the income level. Households in the lowest income quarter depend mostly upon social securi- ty subventions (65%), much more than on income from labour or savings (each 17%) 31. For households from the highest income quarter, labour wages amount to 75% of the household income, against 14% from social security and 11% from savings. Families with at least two wages depend for more than 80% on their labour income. Contrary to the for- mer century, this means a higher and more secure income status. Again, a new labour and income standard is set by the upper middle classes. Breadwinner Models and Historical Models Problems, Programmes and Models 71  HISTORICAL MODELS AND HISTORICAL TIME The breadwinner model is a symbol of the 20th century social order, not only in Belgium, but also in the rest of Western Europe. This model, which has its roots in the closing decades of the 19th century will probably last into the first quarter of the 21st century. The 20th century family model replaced the deregulated pattern established in the wake of the first industrial capitalist acceleration. The breadwinner model is not only a model because it exhibits the characteristics of an empirical-logical system, but also because it has an ‘exemplary’ value. Therefore any model is something of a myth, in view of its ‘recruiting’ and ‘hiding’ character. It is a concealing myth, because, in narrowing the definition of the labour market, it hides many female tasks from formal, statistical government registration. As a myth, it is also ‘recruiting’, because the model has been used as an ideological tool to impose a particular gender model on society and households. Here I refer not only to the old social and political elites, but also to the new emancipating ‘middle groups’: the lower middle classes, the lower clergy and teachers, the worker elites leading the rising social movements. The central link in the analysis of the rise and decline of the breadwinner model is the household, or, to put it more accurately, the totality of choices made within the household. There lies the junction where individual aspirations, gender relations, entrepreneurial strategies as well as ideological standards meet. An integrating research model can only be built up upon a systematic and comparative reconstruction of ‘the breadwinner model’ in a series of regions and countries according the centre-periphery division 32. This geograph- ical analysis should be linked to the different phases (acceleration and deceleration) in the growth of capitalist production. That is how the concept of family strategies can be made operational within the framework of a changing ‘historical time’. After 1750 western labour history is marked by three periods of transition, which outline labour organisation for the next ‘long century’ 33. These periods are marked by accelerated changes in capital accumulation, labour organisation, household formation, gender rela- tions and normative regulation. The first period is characterised by the deregulation of labour markets and household structures and a privatisation (de-collectivisation) of social Graph 5. The timing of historical models in Belgium, 19th - 20th centuries. 72 Eric Vanhaute relations. This goes hand in hand with an ideological offensive of economic liberalisation and social disciplining. From the last decades of the 19th century, measures to promote protective legislation and a new ideology of the breadwinner-and-housewife-model go hand in hand with a more rigorous labour organisation. The trend is towards a more regu- lated societal model, on a macro-level (a fordist government policy), on a meso-level (a taylorist labour organisation) and on a micro-level (a patriarchal household organisation). This regulated and collectivist model is unsettled again in the last two decades of the 20th century, when a new acceleration in the capitalist accumulation model generates a new de- regulation and de-collectivisation. New forms of flexible production are backed by a less stringent labour legislation, the fragmentation of household patterns, the individualisation of consumption and formerly collective life transitions, and the commercialisation of leisure. As Saskia Sassen argues, this new phase in the globalisation of capitalism generates a ‘movement to a more informal modus operandi’ of economic production, which stimu- lates a rising demand for flexible, non-regulated and therefore more easily exploitative forms of family, female and immigrant labour 34. Comparative research demonstrates that there is no such thing as a simple set of circum- stances which generated the male breadwinner family 35. However comparative research should be sensitive to the wider outline of historical time as well as to the periods of shifts and transitions shaping such times. Any research project aiming at a comparative analysis over time and space has to bridge the periods of transition 36. Moreover it has to integrate the changes in the social concepts that order and transform societies: household, group, class, gender, labour, social movement etc. 37. Without doubt, this will be the hardest nut to crack. NOTES 1 F. Engels, The condition of the working class in England, Chicago 1984, p. 159. 2 K. Honeyman, Women, gender and industrialisation in England, 1700-1870, Basingstoke-London 2000. 3 T.K. Hareven, The history of the family and the complexity of social change, “American Historical Review”, 1991, p. 124. 4 S. Horrel, J. Humphries, The origins and expansion of the male breadwinner family. The case of nineteenth century Britain, “International Review of Social History”, Supplement 5, 1997, pp. 26-27. 5 W. Seccombe, Weathering the storm. Working-class families from the industrial revolution to the fertility decline, London - New York 1993. 6 C. Tilly, C. Tilly, Work under Capitalism, Boulder, 1998. 7 See the texts of the workshop Changing Boundaries and Definitions of Work over Time and Space in the 19th International Congress of Historical Sciences, Oslo 2000 (Proceedings 19th International Congress of Historical Sciences, Oslo 2000, pp. 309-327). 8 Ph. Moen, E. Wethington, The concept of family adaptive strategies, “Annual Reviews of Sociology”, 1992, p. 234. 9 C. Creighton, The rise of the male breadwinner family: a reappraisal, “Comparative Studies in Society and History”, 3, 1996, pp. 310-337; A. Janssens (ed.), The rise and decline of the male breadwinner family?, “International Review of Social History”, Supplement 5, 1997; Horrel, Humphries, op. cit.; E. Vanhaute, Between patterns and processes. Measuring labor markets and family strategies in Flanders, 1750-1990, “The History of the Family. An International Breadwinner Models and Historical Models Problems, Programmes and Models 73  Quarterly”, 4, 1997, pp. 527-545. 10 See e.g. G. De Brabander, Regional specialization, employment and economic growth in Belgium between 1846 and 1970, New York 1981; H. Pott-Buter, Facts and fairy tales about female labor, family and fertility. A seven-country comparison, Amsterdam 1993. 11 About the counting of female labour in English censuses see e.g. E. Higgs, Women, occupations and work in the nine- teenth century censuses, “History Workshop Journal”, 23, 1987, pp. 59-80; Honeyman, op. cit. 12 Professionally active: having a profession according the professional censuses; Professionally non-active: students, officially unemployed, officially retired; Non-professional: not belonging to the first two categories (household labour reservoir). 13 Vanhaute, op. cit. 14 Cf. Vanhaute, op. cit., p. 537. Information about urban working-class families. 15 P. Scholliers, Family income, needs and mother’s wages. A critical survey of working-class budget inquiries in Belgium, 1853-1929, in T. Pierenkemper (ed.), Zur Ökonomik des privaten Haushalts. Haushaltsrechnungen als Quelle historischer Wirtschafts- und Sozialforschung, Frankfurt - New York 1991, p. 169. 16 In 1976 63% of all households with two partners were one-income families. In 1995, this ratio had dropped to 40%. 17 M. Berg, Women’s work and the industrial revolution, in A. Digby, C. Feinstein, D. Jenkins (eds.), New directions in eco- nomic and social history, vol. II, Basingstoke - London, 1992, p. 34; T. McBride, Women’s work and industrialization, in L.R. Berlanstein (ed.), The industrial revolution and work in nineteenth-century Europe, London - New York 1992, pp. 63-80; Honeyman, op. cit. 18 From an average of 2500 hours per year in 1750 to 3000 in 1850; Seccombe, op. cit. 19 P. Scholliers, Wages, manufacturers and workers in the nineteenth-century factory. The Voortman cotton mill in Ghent, Oxford-Washington 1996, pp. 93-94. 20 The importance of child labour can be illustrated by the recruitment policy of one of the Ghent cotton mills, Voortman. Until the beginning of the 20th century, one out of every three starting girls was younger than 15 (Scholliers, op. cit., p. 102). 21 C. Lis, Social change and the labouring poor. Antwerp, 1770-1860, New Haven - London 1986. 22 Ellen Jordan points at the high degrees of female unemployment in the 19th century British industry (E. Jordan, Female unemployment in England and Wales, 1851-1911. An examination of the census figures for 15-19 years old, “Social History”, 13, 1988, pp. 175-190). 23 Lis, op. cit., pp. 158 ff.; Seccombe, op. cit., pp. 54-60; Berg, op. cit. 24 P. Hilden, Women, work and politics. Belgium, 1830-1914, Oxford 1993, pp. 4-5. 25 Creighton, op. cit., p. 327; Honeyman, op. cit., pp. 144-146. At the end of the 19th century female factory workers had almost disappeared in the metallurgical city Liège: wives as workers in families without children: 8%, mothers with one or two children: 6%. In the cotton city Ghent the ratios were completely different: 86% of the wives in families without children, 32% of the mothers of one or two children were registered with a profession (G. Alter, Work and income in the family economy. Begium, 1853-1891, “Journal of Interdisciplinary History”, 1984, p. 271; Scholliers, op. cit., 1996, p. 179). 26 S. Pedersen, Family, dependence, and the origins of the welfare state. Britain and France, 1914-1945, Cambridge 1994. 27 The strategies of English male trade unionists have been the subject of much debate. Their exclusionist strategies towards female labour in certain professions seem to have been responses to a complex mixture of class conflict and gender antagonism. See J. Humphries, Class struggle and the persistence of the working class family, “Cambridge Journal of Economics”, 1, 1977, pp. 241-258; S. Rose, Gender antagonism and class conflict: exclusionary strategies of male trade unionists in nineteenth century Britain, “Social History”, 13, 1988, pp. 191-208; Honeyman, op. cit., pp. 115-137. 28 Until as late as 1889 Belgium had no limitation on child labour. The bill on compulsory school attendance was only voted in 1914. In the second half of the 20th century however, Belgium was an exemplary student in the class of public regulation and taylorist-fordist labour organisation. 29 K. Honeyman, J. Goodman, Women’s work, gender conflict, and labour markets in Europe, 1500-1900, “Economic History Review”, 4, 1991, p. 165. 74 Eric Vanhaute 30 W. Van Dongen, M. Beck, E. Vanhaute (eds.), Gezinsleven en beroepsleven. Het combinatiemodel als motor voor een actieve welvaartsstaat?, Leuven-Apeldoorn 2001. 31 Based on the budget inquiry 1996-1997 of the National Institute for Statistics. 32 See e.g. Seccombe, op. cit.; A. Janssens (ed.), The rise and decline of the male breadwinner family?, “International Review of Social History”, 1997, Supplement 5. A first comparison can be based on the calculation of crude par- ticipation rates of female labour, cfr. Pott-Buter H., Facts and fairy tales about female labor, family and fertility. A seven- country comparison, Amsterdam 1993, p. 27, p. 200. 33 I lend the concept of ‘long centuries’ and structural crises from G. Arrighi, The Long Twentieth Century. Money, power, and the origins of our times, New York 1994. 34 S. Sassen, The informal economy: between new developments and old regulations, in Id., Globalization and its Discontents, New York 1998. 35 S. Horrel, J. Humphries, The origins and expansion of the male breadwinner family. The case of nineteenth century Britain, “International Review of Social History”, Supplement 5, 1997, p. 26. 36 This is not restricted to the ‘modern’ period. Honeyman, Goodman, op. cit., stress the changes in the urban labour markets in the period of economic acceleration of the 16th century. 37 R. Milkman, E. Townsley, Gender and the economy, in N.J. Smelser, R. Swedberg (eds.), The handbook of economic soci- ology, Princeton - New York 1994, pp. 600-619. SELECTED BIBLIOGRAPHY Alter G., Work and income in the family economy. Begium, 1853-1891, “Journal of Interdisciplinary History”, 1984, pp. 255-276. Arrighi G., The Long Twentieth Century. Money, power, and the origins of our times, New York 1994. Berg, M., Women’s work and the industrial revolution, in Digby A., Feinstein C., Jenkins D. (eds.), New directions in eco- nomic and social history, vol. II, Basingstoke - London, 1992, pp. 23-36. Creighton C., The rise of the male breadwinner family: a reappraisal, “Comparative Studies in Society and History”, 3, 1996, pp. 310-337. De Brabander G., Regional specialization, employment and economic growth in Belgium between 1846 and 1970, New York 1981. Engels F., The condition of the working class in England, Chicago 1984. Hareven T.K., The history of the family and the complexity of social change, “American Historical Review”, 1991, pp. 95-124. Hilden P., Women, work and politics. Belgium, 1830-1914, Oxford 1993. Higgs E., Women, occupations and work in the nineteenth century censuses, “History Workshop Journal”, 23, 1987, pp. 59- 80. Honeyman K., Women, gender and industrialisation in England, 1700-1870, Basingstoke-London 2000. Honeyman K., Goodman J., Women’s work, gender conflict, and labour markets in Europe, 1500-1900, “Economic History Review”, 4, 1991, pp. 608-628. Humphries J., Class struggle and the persistence of the working class family, “Cambridge Journal of Economics”, 1, 1977, pp. 241-258. Horrel S., Humphries J., The origins and expansion of the male breadwinner family. The case of nineteenth century Britain, “International Review of Social History”, Supplement 5, 1997, pp. 25-64. Breadwinner Models and Historical Models Problems, Programmes and Models 75  Janssens A. (ed.), The rise and decline of the male breadwinner family?, “International Review of Social History”, Supplement 5, 1997. Jordan E., Female unemployment in England and Wales, 1851-1911. An examination of the census figures for 15-19 years old, “Social History”, 13, 1988, pp. 175-190. Lis C., Social change and the labouring poor. Antwerp, 1770-1860, New Haven - London 1986. McBride T., Women’s work and industrialization., in Berlanstein L.R. (ed.), The industrial revolution and work in nineteenth- century Europe , London - New York 1992, pp. 63-80. Milkman R., Townsley E., Gender and the economy, in Smelser N.J., Swedberg R. (eds.), The handbook of economic soci- ology, Princeton - New York 1994, pp. 600-619. Moen Ph., Wethington E., The concept of family adaptive strategies, “Annual Reviews of Sociology”, 1992, pp. 233-251. Pedersen S., Family, dependence, and the origins of the welfare state. Britain and France, 1914-1945, Cambridge 1994. Pott-Buter H., Facts and fairy tales about female labor, family and fertility. A seven-country comparison, Amsterdam 1993. Rose S., Gender antagonism and class conflict: exclusionary strategies of male trade unionists in nineteenth century Britain, “Social History”, 13, 1988, pp. 191-208. Sassen S., The informal economy: between new developments and old regulations, in Id., Globalization and its Discontents, New York 1998. Scholliers P., Family income, needs and mother’s wages. A critical survey of working-class budget inquiries in Belgium, 1853- 1929, in Pierenkemper T. (ed.), Zur Ökonomik des privaten Haushalts. Haushaltsrechnungen als Quelle historischer Wirtschafts- und Sozialforschung, Frankfurt - New York 1991, pp. 145-181. Scholliers P., Wages, manufacturers and workers in the nineteenth-century factory. The Voortman cotton mill in Ghent, Oxford-Washington 1996. Seccombe W., Weathering the storm. Working-class families from the industrial revolution to the fertility decline, London - New York 1993. Tilly C., Tilly C., Work under Capitalism, Boulder 1998. Van Dongen W., Beck M., Vanhaute E. (eds.), Gezinsleven en beroepsleven. Het combinatiemodel als motor voor een actieve welvaartsstaat?, Leuven-Apeldoorn 2001. Vanhaute E., Between patterns and processes. Measuring labor markets and family strategies in Flanders, 1750-1990, “The History of the Family. An International Quarterly”, 4, 1997, pp. 527-545. Vanhaute E., Het kostwinnersmodel als historische fictie. Arbeid en inkomen van gezinnen in langetermijnperspectief, in Van Dongen W., Vanhaute E., Pauwels K. (eds.), Het kostwinnersmodel voorbij? Naar een nieuw basismodel voor de arbei- dsverdeling binnen de gezinnen, Leuven-Apeldoorn 1998, pp. 55-70. 76 Eric Vanhaute Breadwinner Models and Historical Models Problems, Programmes and Models 77