Ideal versus reality? The domesticity ideal and household labour relations in Dutch industrializing regions, circa 1890

Abstract

For long, international comparisons of female labour force participation (FLFP) have been based on aggregate source material, most notably censuses. However, the lion’s share of today’s historians agree that censuses have systematically underreported women’s work activities. Consequently, scholars relying on this source have found a nineteenth-century Dutch male breadwinner society while others have found that the Dutch female labour force was quite extensive. This discrepancy in the historiography is in need of closer scrutiny. The current study shows that by the end of the nineteenth century, in industrial regions married women indeed withdrew from the registered labour market but instead engaged in other types of labour relations that could easily be combined with homemaking duties and that remained invisible in the census. Furthermore, this article argues that the fact that married women provided an income did not necessarily contradict the growing ideal of domesticity. The alternative types of work married women took up were rather a way of reconciling this ideal with keeping the household on a respectable level of existence.

Keywords:

• Labour relations
• domesticity
• women’s work
• Netherlands
• nineteenth century

1. Introduction

For a long time, most research on female labour force participation (FLFP) in Western Europe was based on aggregate source material, most notably censuses (Table 1). Nowadays, however, most scholars contend that censuses systematically underreported women’s labour owing to their irregular work patterns and their involvement in informal labour (Higgs, 1987; Hill, 1993; Horrell & Humphries, 1995; Humphries & Sarasúa, 2012; Janssens, 2014, p. 113; Muños Abeledo, 2012; van Nederveen Meerkerk, 2012; van Nederveen Meerkerk & Paping, 2014).¹ Moreover, women who were self-employed or assisted their husbands on a farm were usually not listed as working. Even women who were full-time, gainfully employed were often not listed with an occupation while part-time employed men were. Underreporting was in part caused by the specific characteristics of women’s work but even more so by the enumerators’ values and dominating stereotypes about gender labour division. Therefore, ‘they [censuses] appear objective and value free, but their meaning grows out of socially constructed concepts that are laden with cultural and political values’ (Folbre, 1991, p. 463). Thus, census data has skewed our understanding of FLFP within countries and may have additionally given us the wrong impression about national FLFP rates relative to each other.

Table 1. Female labour force participation in Western Europe 1850–1930 (percentage of women with a listed occupation).

Year	Belgium	France	Germany	Netherlands	UK
1850	38			24	30
1860	36	25		18	28
1870	36	24			28
1880	34	30	24		25
1890	29	33	25	15	27
1900	29	35		17	25
1910	25	39	30	18	26
1920	21	42	35	18	26
1930	24	37	34	19	27

Source: Pott-Buter (1993, p. 21).

Fortunately, the strand of literature using alternative, disaggregate sources for investigating FLFP in pre-industrial and industrial economies has been growing in the past years (see among others: Boter, in press; Humphries & Weisdorf, Humphries & Sarasúa, 2012; Humphries & Weisdorf, 2015; Janssens, 2014; Muños Abeledo, 2012; Schmidt & van Nederveen Meerkerk, 2012; van den Heuvel & van Nederveen Meerkerk, 2014; Whittle, 2014). Moreover, the broad definition of work by Tilly and Tilly – ‘Work includes any human effort adding use value to goods and services’ (Tilly & Tilly, 1998, p. 22) – is increasingly applied in empirical research. These studies have taken more types of labour relations into consideration. For instance, next to men’s wages, reciprocal labour – i.e. labour without the involvement of markets or monetized transactions – is now considered to have been crucial for determining households’ living standards throughout history (Humphries, 2013). Furthermore, more and more scholars acknowledge the importance of using a regional approach because differences within countries are crucial for understanding the mechanisms behind changing FLFP in different types of local labour markets (Janssens, 2014, pp. 45–49). In summary, the extent and importance of women’s ‘hidden tasks’ are increasingly unveiled by means of the creative use of disaggregate source material.

The Netherlands makes an especially interesting case for investigating women’s (hidden) labour because the census data (Table 1) suggests that FLFP in the Netherlands decreased earlier and more rapidly relative to surrounding countries. According to some scholars, this decline was caused by middle-class social norms regarding domesticity spreading through the working class during the second half of the nineteenth century (de Regt, 1984; de Swaan, 1989). Because the housewife was crucial for a domestic lifestyle, task division within the household became more pronounced with men providing an income and women specializing in homemaking tasks (Van Poppel, Van Dalen, & Walhout, 2009). Moreover, in the Netherlands, the domesticity ideal was presumably already prevalent during the early modern period (de Vries & van der Woude, 1997). Some scholars have concluded that, therefore, the Dutch ‘male breadwinner society’ was already established in the seventeenth century (Pott-Buter, 1993).

Recent research has shown, however, that Dutch women were actually much more active in the early modern (van den Heuvel, 2007; van Nederveen Meerkerk, 2007; Schmidt, 2005) and nineteenth-century labour markets (Schmidt & van Nederveen Meerkerk, 2012) than previously believed. Ariadne Schmidt and Elise van Nederveen Meerkerk have estimated that female participation rates in the Dutch 1899 census (1900 in Table 1) need to be adjusted from 17% to at least 24%. According to them, above all married women working in agriculture and wives of self-employed men in the industrial and retail sectors have remained invisible in the census (Schmidt & van Nederveen Meerkerk, 2012, pp. 88–89).

This discrepancy in the historiography between a male breadwinner society and a considerable female labour force is in need of closer scrutiny. The present research shows, based on qualitative evidence, that such contradictory findings can partially be explained by the ways in which Dutch women attempted to reconcile their desire for domesticity with the need for income. Furthermore, by specifically focusing on various types of women’s labour in industrializing regions, it argues that FLFP in the 1899 census was even higher than the proposed adjusted percentages by Schmidt and van Nederveen Meerkerk (2012) that were principally based on adding women working in agriculture and women assisting their self-employed husbands. As such, this research analyses two sides of the same coin. On the one hand, it examines the working and middle class’ perceptions of labour relations in light of their desire for domesticity (see also: Hofmeester & Moll-Murata, 2011). On the other hand, it looks at the actual work activities of working-class households. By analysing hundreds of interviews with industrial labourers from 1890, this article sheds new light on Dutch women’s hidden labour in a domestic world.

Two main conclusions will be presented. First, it shows that married women often combined different types of labour relations, i.e. wage labour, reciprocal labour, and reproductive labour, to both live up to the domesticity ideal and provide additional income. Including these types of work means that FLFP in the 1899 census was even higher than the previously proposed (lower-bound) estimate of 24%. Second, a household in which women provided an income did not necessarily contradict the ideal of domesticity. Letting the wife combine different types of labour relations was a way for households to reconcile their longing for domesticity with the need to provide enough income.

This article is structured as follows. Section 2 considers the theoretical framework of a dichotomy between a public and a private sphere which has often been applied in research on women’s labour history. Furthermore, it elaborates on the sources used for this research and how they can help us understand the significance of separate spheres in real life. Section 3 is concerned with the working class’ and middle class’ perceptions of household labour relations. Section 4 gives an impression of actual household labour relations and considers the implications of the results for working-class households’ income composition. Section 5 concludes.

2. Theory and sources

In the historiography, the discourse of domesticity is almost invariably linked with theories about a dichotomy between the public and the private sphere. The former is usually associated with outdoor work, dominated by men, while the latter is thought of as a domestic, female sphere. This conceptual framework has for long been the dominant way in which scholars approached the history of women (Vickery, 1993). An especially persistent idea is that industrialization stimulated women’s retreat from the labour market. The reasoning is that as a result of mechanization, production was increasingly carried out in factories instead of at home whereby women could no longer combine their homemaking duties with gainful labour. Consequently, women became increasingly restricted to the private sphere while their husbands became the household’s sole wage earners. To use the words of one of the most outspoken opponents of the use of this framework in women’s history: ‘According to customary wisdom, sometime between 1600 and 1800 a wholesome “family economy” wherein men, women and children shared tasks and status gave way to an exploitative wage economy which elevated the male breadwinner and marginalized his dependants’ (Vickery, 1993, p. 402).

Vickery’s critical account of the use of the separate-spheres theory in women’s history is already 20 years old. Nevertheless, her arguments are still relevant in light of more recent research because the separate-spheres theory is still often interpreted as a dichotomy between a male and a female sphere, and as one between labour and domesticity. For instance, recent research has argued that women increasingly withdrew from the labour market to live up to society’s increasing desire for domesticity and full-time housewifery (Minoletti, 2013; Van Poppel et al., 2009). This conclusion implies that the private sphere ideally was to be freed from any type of gainful work and that women needed to stay out of the public sphere, i.e. withdraw from the labour market, to create domesticity in the private sphere. Thanks to our increasing understanding of women’s hidden tasks within the home, we know that this was not true in reality. This research offers new insight into work in the private sphere and additionally shows that even ideals were not this stringent.

To further explore this issue, the Dutch labour surveys of 1887 and 1890 have been consulted. These surveys contain hundreds of interviews with a great variety of people such as (factory) labourers, vicars, mayors and producers. The surveys were executed to investigate the working and housing conditions of labourers and to test the application of respectively the Child Labour Act of 1874 and the Labour Law of 1889. The former law prohibited children younger than 12 from working in factories but did not apply to children working in the agricultural and service sectors (Schenkeveld, 2003). The latter law, among other things, limited women’s and children’s working days to 11h, prohibited their night-time and Sunday labour, and prohibited women from working within four weeks after giving birth (Brouwer & van Eijndhoven, 1981; Smit, 2014). Hence, the focus of the surveys was on the nature and extent of women’s and children’s labour, the relationship between labourers and employers, housing conditions and working hours.

This qualitative information provides a rare insight into the daily lives of labourers in an industrializing economy. Their accounts tell us more about their conception of domesticity and how they attempted to realize this. Furthermore, implicitly, their answers increase our understanding of the existence of separate spheres on a micro level: did the respondents actually think of the public and private spheres in terms of male versus female and work versus domesticity?

The surveys of Tilburg (1887) and Twente² (1890) have been selected as case studies. Both regions industrialized relatively early in the Dutch context, while in the Netherlands in general, industrialization took off much later relative to surrounding countries (de Jonge, 1976; Griffiths, 1979; Jansen, 1999; Mokyr, 2000a; van Zanden & van Riel, 2004). Already during the early modern period, both Twente and Tilburg had been important textile centres. Farmers’ households dedicated themselves to spinning and weaving during the winter months, slack periods for agricultural work (Slicher van Bath, 1960). In Twente, the textile industry expanded swiftly after 1830 because the Dutch cotton industry, previously located in the southern provinces, was transferred to Twente after Belgium gained independence in that year. Consequently, the demand for labour in this region boomed and men as well as women were mobilized to supply the sudden need for a large labour force (de Groot, 2001; pp. 186–196). In contrast, the wool industry of Tilburg expanded more gradually. Supposedly, these different processes of industrialization resulted in a stricter gender segregation in Tilburg than in Twente (de Groot, 2001, p. 186; Janssens, 2009, p. 94).

These regions were selected for two reasons. First, former research on women’s hidden work has successfully adjusted FLFP in the 1899 census (Schmidt & van Nederveen Meerkerk, 2012), but has principally focused on women working in agriculture and on wives of craftsmen and retailers working in their husband’s business. The surveys will show that there were more types of women’s work that have remained obscured in the census. Second, industrial regions make an interesting case to investigate the ideal of domesticity since (registered) labour was normally carried out in the public sphere. In line with the growing desire for domesticity, women needed to stay at home, at least most of the time, to fulfil their tasks as housewives, but at the same time the household had to provide enough income to make ends meet. How did industrial households cope with these possibly conflicting wishes? In other regions, such as the service oriented coastal cities, households might have responded very differently and they might have had different perceptions of women’s and children’s labour relations.

The respondents’ accounts were systematically searched for comments about household income, ideas about domesticity, women’s and children’s work activities, and perceptions of this work. It is imperative to distinguish between comments on unmarried and married women’s work and to, where possible, account for the number of children because the different life-cycle stages of the household greatly affected household labour allocation. Besides the real possibility that the father/husband died, left his family or was otherwise absent, almost every household experienced periods during which his earnings were not sufficient to feed and clothe the entire family (Schneider, 2013). This was especially the case in households with many small, dependent children (Humphries, 2013, p. 708).

We have to keep in mind that the interviews were conducted to investigate labourers’ working conditions and, in the case of the 1890 survey, to test the success of the Labour Law. Employers were probably more inclined to lie about their preferences and policies – especially when they were breaking the law, which was implemented by the same authorities who carried out the survey. Likewise, labourers themselves could have been more inclined to obscure their perceptions and work activities because they wished to make a good impression on the survey conductors. On the other hand, the survey conductors at times put words into the respondents’ mouths, pushing a certain issue when receiving unsatisfactory answers (Smit, 2007, p. 93). Moreover, the number of women questioned is negligible (19 out of 244 interviewees in Twente and none in Tilburg) and children were not questioned at all.

Additional source material was therefore needed. Despite the objections against the use of censuses, still the 1899 census has been consulted to give an idea of the size and composition of the registered labour force. Furthermore, several surveys on the home industry and married women’s factory labour from the first decade of the twentieth century have been examined. These surveys were conducted for different reasons and therefore place the 1890 surveys in a (temporal) perspective. Finally, municipal records provide additional information on local textile production.

3. Perceptions of working-class household labour relations and domesticity

This section explores nineteenth-century conceptions of domesticity and the survey respondents’ perceptions of the type(s) of labour relations necessary to achieve this. Furthermore, it will touch upon the question whether labourers indeed viewed the public sphere as male and industrious, and the private sphere as female and domestic. Before moving on to the results, a closer reading of the Dutch historiography on the development of the domesticity ideal is required.

As mentioned in the introduction, the early modern Dutch Republic is considered to have been a front runner in the development of middle-class social norms regarding domesticity (Schmidt, 2011; de Vries & van der Woude, 1997), although other studies give a more nuanced view (Everard, 2005; Schuurman, 1992). At the same time, ample literature has shown that the size of the early modern female labour force was considerable. According to Ariadne Schmidt, these two apparent contradictory findings could coexist because the ideal of industriousness concurrently intensified: perceptions of hard work changed positively for men as well as for women (Schmidt, 2011). In reality, therefore, domesticity and industriousness could coexist because many types of work, such as textile production, were still carried out at home.

The domesticity ideal further gained ground during the nineteenth century when the middle class started to actively propagate its way of life among the working class. Its main goal was to ‘rescue’ the working class from life in dire poverty by promoting the virtues of, among other things, self-restraint and frugality. It has been argued that the underlying motivation was to prevent the poor from revolting against the established order and against spreading contagious diseases (de Swaan, 1989; p. 132; Smit, 2007, p. 74). Indeed, besides domesticity, spreading knowledge about hygiene was an important aspect of this mission, just like in nineteenth-century England (Mokyr, 2000b).

This ‘civilizing mission’ had been institutionalized already in 1784, when the Maatschappij tot Nut van ‘t Algemeen (‘the society for public welfare’, henceforward: ‘t Nut) was founded. Its ideas about women’s roles in society were clear: ‘For women from all social classes there is only one possible career. All women have merely one destiny: that is, to become friends, wives, mothers and care givers’ (quoted in: Kruithof, 1983, p. 378). In this line of reasoning, the most crucial requisite for a proper domestic life was the housewife who was supposed to devote all her time to homemaking tasks and consequently had to withdraw from the labour market (van Zanden & van Riel, 2004, pp. 319–328). Other examples of ‘t Nut’s activities are the establishment of local libraries and savings banks, and increasing interference in education.

3.1. Women

Perceptions of women’s labour depended upon the type of work and their marital status. Unmarried women’s factory labour was, especially in middle-class circles, considered negative for their development. According to contemporary observers, in a factory young women would be corrupted by immoral gossip and intercourse with the opposite sex. Even worse, here they would never acquire sufficient knowledge about the ins and outs of good housewifery. An engineer from Tilburg mentioned:

Factory girls cannot become decent housewives, and when the male labourer is lucky enough to marry a decent girl who served in a proper household for five or six years, and therefore knows how to manage a household, in that case this man is very fortunate. However, when the labourer marries a factory girl who knows nothing about housekeeping and will find his house in chaos after coming home, he will go to the pub and destroy himself. The woman plays a large part, and this pleads against women working in the factories. (Giele, 1981, p. 90)

However, not all respondents were equally negative. A producer from Twente stated that: ‘I consider it crucial that girls do not go to the factory before they learned how to manage a household’ (Arbeidsinspectie, 1890a, p. 437). Another producer emphasized the importance of factory work for saving for the foundation of a new household upon marriage (Arbeidsinspectie, 1890a, p. 153). Some factories even offered a special savings arrangement for their employees (Arbeidsinspectie, 1890a, p. 109). Probably, producers were more inclined to respond positively because young women were cheap labourers. Likewise, working-class respondents had ample economic incentives to approve unmarried women’s factory labour. Women’s factory wages were often an important resource for either their parents’ or their (future) husbands’ households. Still, working-class respondents were not that outspoken about this issue. Most labourers casually mentioned their adult daughters and other young women working alongside them, although they did stress the importance of good education in domestic skills such as knitting and sewing.

Perceptions of married women’s work were more straightforward. Their factory work was seriously frowned upon by the majority of the respondents, although some did explicitly acknowledge the necessity of the wife’s income for her household. We have to keep in mind that, as mentioned in section 2, the interview conductors often asked suggestive questions and that, on the other hand, the respondents may have been inclined to give desirable answers. However, considering the remarkable consensus among the respondents on this issue, I believe that these sentiments were indeed prevailing.

The respondents’ arguments against married women’s factory labour can be roughly divided into two groups. First, women should stay at home to fulfil their domestic duties because working in a factory all day long inevitably resulted in chaotic homemaking. This type of reasoning corresponded to the goals of the middle class’ civilizing mission: domesticity needed to be pursued by all households by means of a devoted housewife. The working-class respondents’ answers therefore indicate that this mission had been successful. Cor Smit has similarly concluded for the city of Leiden that: ‘[t]he civilizing mission of the bourgeoisie aligned with the ideals of a considerable share of the working class’ (Smit, 2007, p. 110).

Second, several male labourers expressed their fear of female competition in the labour market. A labourer from Oldenzaal mentioned: ‘I wish that female labour would be abolished altogether because male wages would rise and the prosperity of the household would improve’ (Arbeidsinspectie, 1890a, p. 211). Cor Smit has found comparable answers in the surveys of Leiden (Smit, 2007, p. 100). Similar arguments have been advanced in the British historiography on FLFP (Burnette, 2008, p. 15; Horrell & Humphries, 1997, p. 50; Seccombe, 1993, p. 74). Joyce Burnette has even argued that: ‘[t]hese men used gender ideology to increase public support for the entry barriers [i.e. against women in the labour market] they erected, but their primary motivations were economic’ (Burnette, 2008, p. 15). Indeed, female competition in the labour market was considerable, causing men’s wages to go down and women to perform work that would otherwise have been done by men who were now unemployed.

Surprisingly, despite ample negative accounts, the majority of the respondents pleaded against stricter legislation abolishing married women’s factory labour altogether. For instance, one producer from Enschede argued that young men were not capable of saving enough money before getting married. During the first years of marriage, women’s incomes were indispensable for the establishment of a new household. A prohibition of married women’s factory labour would encourage young unmarried couples to postpone marriage, which would result in immoral relations between the sexes (Arbeidsinspectie, 1890a, p. 153). Even more remarkable is the quote by textile producer Albert Jan Blijdenstein who did not believe in depriving women of their freedom:

The household’s interests could make the wife’s wage labour necessary. Generally, she doesn’t desire to work. [...] The adult woman should have the freedom that we all wish to have. I believe that further interference of the law would cause much harm. (Arbeidsinspectie, 1890a, p. 44)

All in all, the prevailing idea was that married women should principally reside in their home, i.e. the private sphere, because she had important tasks to fulfil: (i) turn the house into a home, (ii) raise children, and (iii) manage the finances. The first task was clearly illustrated by a preacher from Hengelo who said that: ‘a good woman can make a bad house more habitable than the best house would be under the supervision of a careless and sloppy woman’ (Arbeidsinspectie, 1890a, p. 447). The wife’s dedication to the physical environment of her family was of crucial importance for the ‘production’ of hygiene and good nutrition. This too was something the working class needed to learn, at least according to this producer:

[...] I would like to point out the necessity of the abolishment of excise tax on soap, one of the primary necessities of life, and secondly the necessity of the abolishment of excise tax on salt. [...] People do not wash themselves as often as they should. [...] If one wants to do something for the working class, they should start by abolishing these taxes. [...] I pointed out soap and salt, as soap is essential for cleaning and salt for nutrition. (Arbeidsinspectie, 1890a, p. 223)

This type of sanitary propaganda is in line with Joel Mokyr’s argument that three nineteenth-century scientific revolutions (the growing knowledge of the connection between filth and diseases, germ theory of disease and nutritional sciences) greatly attributed to the origins of full-time housewifery. A clean home was crucial for the health of all household members (Mokyr, 2000b). Women were to be held accountable for the creation of such a healthy home. Mokyr then continues his argumentation by exploring how women were persuaded to play this role. According to him, ‘[m]iddle-class notions of a culture of respectability were a subtle means by which concepts of proper housekeeping were diffused through the working class’ (Mokyr, 2000b, p. 22). Hence, in Mokyr’s view, the spreading of middle-class social norms was merely the means with which knowledge about health was spread through the working class.

The housewife’s second responsibility, raising children, links up to the first in terms of both health and respectability. Germs were transmitted via polluted water and cows’ milk. Therefore, breastfeeding was actively promoted to combat child mortality (Dyhouse, 1978; Mokyr, 2000b). Recent research on the Dutch case has led to the albeit cautious conclusion that breastfeeding indeed reduced child mortality (Janssens & Pelzer, 2012). Whether mothers actually chose to breastfeed their children principally depended on local culture (Walhout, 2010). Evidently, the promotion of breastfeeding was one of the key points for the interview conductors to combat child mortality and this could only be realized when women would stay at home.³ The same applied to raising children and teaching them good morals. Hiring a babysitter was regarded irresponsible, although it was perceived as less harmful when the babysitter was the child’s grandmother. A preacher from Enschede said:

As long as the children are small, I consider married women’s factory labour to be negative for the children, because they fall into the hands of others and lack motherly care. It is a different story when in such families there is still a grandmother present who takes over the care of the children from the mother [...] (Arbeidsinspectie, 1890a, p. 189)

Finally, housewives were expected to be frugal in order to manage the household’s finances. How incomes were managed determined the well-being of a household more than the actual size of these incomes did. A producer from Oldenzaal marvelled at the fact that labourers with a low income still managed to live a comfortable life. He concluded that: ‘everything depends on the woman’ (Arbeidsinspectie, 1890a, p. 498). However, as a preacher from Borne noticed, by 1890, frugality had become a virtue of the housewives that once were. He stated that:

[t]hey all hoard their money, but the housewife of today is not as clever as she was before. She used to be sparing by tactically keeping everything in order with little means. Those housemothers have become a curiosity among labourers. (Arbeidsinspectie, 1890a, p. 387)

3.2. Children

Child labour was another important topic of the labour surveys. Most respondents considered 12 to be a proper age to start working and that children who were not taught a craft at a young age would suffer the consequences during their adult lives (Arbeidsinspectie, 1890a, p. 62). Factory work was in this way considered to be an investment in human capital. However, one producer said: ‘In general, it is not the producer’s desire to hire such young labourers, but it is a request from the parents themselves’ (Arbeidsinspectie, 1890a, p. 101). Indeed, from the surveys it is apparent that parents anticipated the moment their children were able to earn money. Thus, also perceptions of children’s labour relations were for an important part economically motivated.

Children’s individual wages were reason for concern about their morality because they sometimes paid their parents boarding money and kept the rest of their wage for themselves. This supposedly encouraged them to spend their money on alcohol or other extravagances. Moreover, in this way adolescents realized that they played a crucial role in the household economy which could make them undermine their parents’ authority. A priest from Almelo stated that: ‘I think it is an abominable thing when children pay board to their parents. They give them f2,50 or f3,- and they spend the rest of their wage themselves which is obviously bad for their morality’ (Arbeidsinspectie, 1890a, p. 248).

3.3. Perceptions of labour relations in relation to domesticity and spheres

These findings illustrate several important aspects in light of the debates about domesticity and the separate-spheres theory. First, they show that unmarried as well as married women’s industriousness was highly encouraged. Unmarried women were expected to save money for the foundation of a new household when they married, which is further reflected in the numerous saving banks, some even founded by ‘t Nut. Furthermore, married women’s incomes were by many considered indispensable for their households.

The way in which women were supposed to earn money was, however, highly debated. Unmarried women should preferably work as domestic servants, although producers and working-class respondents did approve of factory labour. Married women on the other hand should not be working in factories. These statements were by no means a plea for the abolishment of married women’s gainful work altogether, as long as they could fulfil their homemaking duties accordingly. Hence, according to contemporaries, industriousness in the form of gainful work could be expressed and executed in the private sphere. This conclusion links up to Schmidt’s argument that during the early modern period, domesticity and labour were not mutually exclusive. In other words, even in industrial economies the dichotomy between the public and the private spheres cannot simply be interpreted as an (ideological) dichotomy between gainful work and domestic work (Schmidt, 2011).

Second, women’s education and work experience were considered essential investments in the future. A producer from Twente even mentioned that, according to him, women’s education was more important than men’s since only a well-educated woman would be able to ‘guide her husband, to withhold him from extravagances like drinking, and to make sure he will be home in time’ (Arbeidsinspectie, 1890a, p. 206). Thus, the longing for domesticity was already reflected in how the respondents perceived unmarried women’s labour. Therefore, we can conclude that domesticity in all its facets was indeed pursued by most nineteenth-century households.

4. Combining labour relations

As extensively argued in the introduction, occupational censuses do not give a complete picture of FLFP. Still, to get an idea of the extent of women’s and children’s formal participation rates Figure 1 shows the percentage of married and unmarried women and children listed with an occupation in the 1899 Dutch census in Enschede and Tilburg. As expected, married women were poorly represented in the registered labour market, unlike children and unmarried women. The labour surveys offer a great opportunity to highlight some of women’s (and children’s) hidden tasks. In what follows, evidence of several labour relations will be presented.⁴

Figure 1. Percentage of women and children with a listed occupation in the 1899 census.

Source: Dutch occupational census (1899).

4.1. Factory wage labour

The vast majority of the registered labour force of Tilburg and Enschede displayed in Figure 1 was working for an industrial enterprise. Table 2 illustrates this based on the results of an industrial census from 1889 showing the share of the total population of Enschede working in an industrial enterprise classified according to age, gender and marital status.

Table 2. Men and women working in an industrial enterprise and the total population, Enschede, 1889.

	Men/boys			Women/girls
Age	Working	Total	% working	Working	Total	% working
12–13	65	152	43	50	189	27
13–14	93	190	49	74	163	45
14–16	209	312	67	194	282	69
16–18	188	297	63	187	301	62
>18 unmarried	2096	4176	50	692	1869^a	37
>18 married				228	2531	9
Total	2651	5127	52	1425	5053	28

Source: Industrial census (1889).

^a Including widows.

Table 2 shows that out of all the unmarried, adult women living in Enschede, 37% worked in a factory. Factory work was more attractive for unmarried women than, for instance, domestic service because factory wages were much higher (Janssens, 2014, p. 91). The share of married women’s factory labour was relatively small, but they were not entirely absent from the factories. The respondents’ answers from Twente confirm this. Out of the 56 labourers that were explicitly asked whether their wives performed factory labour, 23 confirmed that they did. According to a different survey on married women’s factory labour in the Netherlands conducted by the government in 1908, Twente remained the region with the highest number of married female factory workers (Departement van Landbouw, Nijverheid en Handel [DLNH], Departement van Landbouw, Nijverheid en Handel, 1911, pp. 22–25).

Thus, women were an important part of the industrial labour force (see also: Janssens, 2014, pp. 98–102). In 1889, circa 35% of all the industrial labourers included in Table 2, including children, were female. This conclusion is in sharp contrast with the theory that industrialization forced women out of the labour market because they were not willing, able or permitted to dwell in the public sphere. An opposing theory is that industrialization caused an increase of the demand for cheap labour due to the introduction of new machinery. New jobs for women and children were created because these machines required less physical strength (Burnette, 2008; Humphries & Weisdorf, 2015, p. 427; Pinchbeck, 1969). This was also the case in the Dutch textile industry (Smit, 2014, pp. 462–463). A producer observed this development and stated that:

Due to an increase of improved machines, the number of labourers has grown as well. Physical strength of men and women is no longer necessary. Instead, care and intelligence have become important. As a result, the number of women and children in the factories is much larger than it used to be, because even with little physical effort, they are still able to earn some money. (Arbeidsinspectie, 1890a, p. 43)

Indeed, the municipal records of Enschede show that the number of spindles and looms increased dramatically from the 1890s onwards (Table 3), meaning that the demand for labourers to operate them rose accordingly. When we zoom out and look at the textile industry from a macro perspective the same picture of expansion emerges. The total value added of Dutch textile production rapidly increased from the 1890s onwards (Smits, Horlings, & van Zanden, 2000). Recent research has shown that the textile industry had a significant, positive effect on (unmarried) women’s LFP. Especially from the 1880s onwards, FLFP in the major Dutch textile centres drastically increased again after a short period of decline (Boter, 2014). In short, several sources confirm that the expanding Dutch textile sector was to a significant degree realized by women and children.

Table 3. Increasing number of spindles and looms in use in Enschede 1895–1917 (1895 = 100).

	No. of spindles	Index	No. of looms	Index
1895	110.780	100	6.130	100
1900	198.690	179	7.722	126
1905	253.358	229	9.587	156
1907	288.086	260	10.784	176
1917	439.240	396	14.677	239

Source: Municipal records of Enschede (1895–1917).

Married women’s choice whether or not to work in a factory was heavily influenced by the number of children. It has been shown for several Western European countries that childless women or mothers with only one or two children were more likely to be employed outside their homes than women with five or six small children (Garrett, 1998, p. 288). Indeed, research on industrial England by Paul Atkinson has shown that fertility and FLFP were highly correlated. However, while many have reasoned that the extent of FLFP was determined by fertility rates, Atkinson turns causality around. He argues that in regions with ample working opportunities for women, fertility was low while in regions with a low demand for female labourers, women chose to marry earlier and have more children (Atkinson, 2012). Supporting evidence for the relationship between the number of children and married women’s factory labour was found in the surveys:

If labourers only have one or two children, both the husband and the wife work in the factory. However, the moment they have three children, the husband has to work alone and this means a difficult time until the first child can start working. When there are two or three children allowed to work, people can start saving money for their old age. (Arbeidsinspectie, 1890a, p. 410)

In contrast, recent research on Spain has shown that the number of children did not affect women’s employment rates at all. According to Muñoz Abeledo, Spanish women in the tobacco and textile industries brought their children to the factory or left them in the care of neighbours. She even found evidence of nurseries in the factories (Muños Abeledo, 2012, p. 138). The Dutch sources do show that mothers with older children often returned to the factory. A producer from Almelo said that: ‘as soon as a child is 12 years old, he or she starts working in the factory. Frequently, their mother then returns to the factory’ (Arbeidsinspectie, 1890a, p. 241).

Moreover, many households could not afford to go without the wife’s income. Keeping married women out of the factories, even though they would still be able to contribute to the household income in other ways, made households increasingly dependent on children’s earnings (de Vries, 2008, p. 200). A warehouse worker from Oldenzaal stated that a labourer ‘lives from the money his children earn with him’ (Arbeidsinspectie, 1890a, p. 211). A weaver from Enschede mentioned that:

one should not forget that there are households with four or five children. If the father earns nine guilders, then the household could use the support of a child that earns two or three guilders. Only then, better foodstuffs and clothing can be consumed. (Arbeidsinspectie, 1890a, p. 48)

Indeed, Table 2 shows that in 1889, almost half of all children in Enschede aged 12–14 worked in a factory and nearly 70% of all children aged 14–16.⁵

4.2. Commodified and reciprocal (subsistence) work at home

There were three common alternatives to factory work: the home industry, reciprocal labour at home, such as the cultivation of land, and running a (family) business. To start with the first, the home industry was common in industrial economies. Honeyman and Goodman have argued for the case of England that: ‘as women were squeezed out of employment in the public arena, they were forced either into purely domestic activities or into homeworking or sweatshop employment’ (Honeyman & Goodman, 1991, p. 622). In the Netherlands too, until at least the first decades of the twentieth century, home industrial work remained important and encompassed a great variety of work ranging from peeling shrimps to producing rosaries. In the textile industry, many tasks were outsourced to homeworkers as well.

The surveys show that there was a large supply of labour in the home industries of Twente and Tilburg as well, in particular from married women. A factory director from Twente mentioned that the sewing of sacks was performed by homeworking, married women earning some 50–60 cents a day. His factory employed 200 of those women and therefore, ‘about as many households’ (Arbeidsinspectie, 1890a, p. 378) which implies that other household members assisted the homeworker. The mayor of Tilburg stated that ‘at least 400 households’ (Giele, 1981, p. 5) were working in the home industry. The chair of the Chamber of Commerce of Tilburg estimated this number to be between 200 and 300 households (Giele, 1981, p. 18). Guillaume Pollet, producer of woollen cloth, stated: ‘I think that there are at least 1,000 home weavers,⁶ maybe even 1,500’ (Giele, 1981, p. 42). Five interviewed producers together reported circa 600 employed homeworkers. Gertjan de Groot found that as late as 1938, there were 850 women working at home in Tilburg, compared to 1360 women working in factories (de Groot, 2001, pp. 205–206).

Information on wages in the home industry provided by the surveys is scarce. One producer from Almelo commented that,

[s]pun yarn has to be bundled. I let married women do this, who gladly want to perform some work at home to earn a little extra. They earn f2,50-f4 per week, depending on household tasks: the larger the household, the less time they have for extra work. At the moment, I employ twelve: but were I to decide today I needed 24 or 36, I would have them by tomorrow. Seeing that one puts so much effort into earning extra money, it is apparent that wages are not high. (Arbeidsinspectie, 1890a, p. 241)

A special survey on the home industry from 1909 provided more meticulous information on wages. The catalogue of the objects displayed at an accompanying exhibition provides ample information on the producer(s) of each object that was registered, including gender, age, hours worked per week, and weekly wages. Earnings differed considerably, depending on the type of work and the amount of hours worked per week. The producer’s estimation quoted above seems to have been representative: a 66-year-old woman from Goirle – nearby Tilburg – spent 48 hours per week on sewing sacks earning f2.50 (Posthumus, 1909, pp. 46–47) and a woman from Mierlo, in the south of the Netherlands, spent 72 hours per week weaving velvet, earning f3.90 (Posthumus, 1909, pp. 48–49). These incomes must have been an important share of households’ incomes since the average male factory labourer in Twente earned circa f8–f10 per week.

Table 4 displays the average wages in the textile and apparel home industry in 1909.⁷ The catalogue included objects from 114 women who were working alone. They earned an average wage of f3.82 in a working week of almost 57 hours. These relatively short average work weeks made it possible to combine home industrial work with other labour relations such as homemaking. Men’s significantly higher wages can be explained by longer average workdays and different types of work with higher turnover, such as shoemaking. Married couples also worked together. Children rarely worked alone, but did occasionally assist their parents a couple of hours per day.

Table 4. Wages and working hours of homeworkers, 1909.

	N	Average wage per week (f)	Average wage per hour (f)	Average amount of working hours per week
Man alone	146	7.87	0.10	76.1
Woman alone	114	3.82	0.07	56.7
Man + woman	75	8.64	0.09	100.5

Source: Posthumus (1909).

The second alternative to factory work was the cultivation of land and keeping livestock. Subsistence agriculture is often overlooked in the historiography on women’s labour and households’ living standards in general. Indeed, in Twente, 56 workers mentioned whether or not they had a plot of land next to or nearby their home: 47 (72%) of them confirmed that they did. This was a legacy from the initial stages of industrialization, when factory workers were still part of the peasantry. As mechanization progressed, less time remained for activities outside the factory and eventually, agricultural labour was mostly abandoned. However, cultivating food crops for self-provisioning did not die out completely.

The cultivation of this land seems to have been a joint effort of the entire family. The housewife played an important part as she was, usually, not bound to the unrelenting jingling of the factory bells. A doctor from Haaksbergen noticed that: ‘[w]hen the wife does not work in the factory, more land is cultivated. The wife cultivates the land and the husband does so when he comes home from work’ (Arbeidsinspectie, 1890a, p. 353). Where both spouses worked in a factory or the wife could not manage on her own, households occasionally paid a day labourer to do the work. Moreover, other kin living nearby could have been able to assist.

According to a carpenter from Enschede, the habit of cultivating land was on its return by the end of the nineteenth century:

It used to be customary that every civilian in Enschede and Hengelo possessed his own land to cultivate his potatoes, cabbage, etc. Those people are still familiar with agriculture, although this is decreasing. The weekly markets are more frequently visited than in former times, because there are people who do not want to cultivate their own land anymore. However, he who is wise will cultivate himself because it will save him f3 per week. (Arbeidsinspectie, 1890a, p. 16)

Although according to this carpenter this practice was diminishing, cultivating food was still, for many households, a significant addition to the household income. Several labourers stated that without their land, they would not have been able to get by every week (Arbeidsinspectie, 1890a, p. 132). Moreover, if the above quoted comment is accurate, subsistence agriculture could save a household f3 a week: about one-third of the average man’s factory wage. It has to be noted that the self-provisioning of food was largely impossible in the larger cities in the Western provinces and is therefore typical for rural regions or smaller cities with a long tradition of agriculture.

Finally, some respondents mentioned that they ran their own business as well as their factory work. Although this was only mentioned in the surveys sporadically, it was yet another way for women to combine different types of labour relations. Indeed, one factory labourer owned a beer room which, according to him, was managed by his wife when he was at the factory. At night, he took over (Arbeidsinspectie, 1890a, p. 113). Just like work in the home industry and the cultivation of food, part-time running a business was not registered in official documents like censuses.

In sum, the surveys have demonstrated the ubiquity of women working in the home industry, in subsistence agriculture, and, to a lesser extent, in a (family) business. These are all types of gainful work that have remained invisible in the occupational census. The adjusted lower-bound share of working women in the 1899 census proposed by Schmidt and Van Nederveen Meerkerk therefore needs to be raised even more.

4.3. Combining labour relations in the private sphere: significance for the household income

The results presented in this section show that most children older than 12 and unmarried women were principally engaged in some sort of wage labour. In contrast, married women managed to combine several types of labour relations to reconcile their longing for domesticity with the need to provide sufficient income. They engaged in reciprocal labour relations that were both reproductive (cleaning, cooking, child rearing, etc.) and productive (subsistence agriculture). Furthermore, wage labour in the home industry could easily be combined with reciprocal labour because working hours were flexible and the work could be performed from home. In this way, married women’s work could mean a substantial contribution to the total household income. Therefore, they cannot be ignored by scholars investigating, among other things, the history of households’ living standards.

Figure 2 demonstrates the implications this conclusion has for working-class households’ income composition. Four prototypes of households were constructed, each representing one of the possible combinations of labour relations regularly mentioned in the surveys. The average factory wages of men, women and children noted down by the survey conductors have been used for determining the incomes from wage labour, the average female wage in the home industry (Table 4) for part-time wage labour, and the comment made by a carpenter from Enschede quoted in section 4.2 for subsistence farming. Households 3 and 4 account for the possibility that children were absent or too young to work (cannot be expected to work) and that the wife merely performed homemaking labour (reproductive labour). Table 5 further specifies the type of labour relation(s) performed by each household member for each prototype of household. Subsistence farming was often a joint effort of the whole household.

Figure 2. Income composition of four prototypes of households.

Source: Arbeidsinspectie (1890a, 1890b).

Table 5. Labour relations in four prototypes of households.

	Household 1	Household 2	Household 3	Household 4
Husband	Wage labour	Wage labour	Wage labour	Wage labour
Wife	Wage labour + reproductive labour	Part-time wage labour + reproductive labour	Part-time wage labour + reproductive labour	Reproductive labour
Children	Wage labour (1 child)	Wage labour (2 children)	Cannot (be expected) to work	Cannot (be expected) to work
Subsistence farming	Reciprocal labour for the household	Reciprocal labour for the household	Non-existent	Reciprocal labour for the household

5. Conclusion

This article has provided supporting qualitative evidence for the claim that what we know about national FLFP rates based on occupational censuses in Western Europe needs to be reconsidered. Previous research has estimated that the Dutch female participation rates in the 1899 census (Table 1) need to be adjusted from 17% to at least 24%. These adjustments were principally based on the inclusion of women’s hidden work in agriculture, and women assisting their self-employed husbands in the industrial and retail sectors. By exploring various household labour relations in highly industrialized regions, the present research has further increased our understanding of the types of work that have remained invisible in the 1899 census and has argued that participation rates in the 1899 census mostly have been even higher than 24%. Future, quantitative research will tell to what extent.

Furthermore, from the surveys it has become clear that domesticity was indeed pursued by the working class by the end of the nineteenth century. Multiple respondents answered that they did not want their wives to work in a factory because that would harm the household. Married women therefore needed to dwell in the private sphere to create a domestic environment for their families. However, this did not mean that they were discouraged from providing an income. The private sphere was not perceived as a place exclusively for leisure, but rather considered as a place where domesticity and industriousness could go hand in hand.

The coexistence of domesticity and work is reflected in the ways in which households allocated their labour. Most married women indeed stopped working in factories but instead, found other ways to provide an income to reconcile the desire for domesticity with the need for sufficient income. Cultivating food, keeping livestock, working in the home industry, and running a (family) business were common types of work that married women could perform in addition to their homemaking and child rearing duties. In other words, married women could combine several types of labour relations to generate sufficient income. This may explain why some scholars have found a nineteenth-century Dutch male breadwinner society while others have presented evidence that the Dutch female labour force was considerable.

These conclusions have implications for the ways in which the separate-spheres theory has been applied in the historiography. In the context of nineteenth-century, industrial regions, the public and the private spheres cannot be understood as dualities between male and female, or labour and domesticity: the home and its immediate surroundings turned out to be important workplaces. Working was not necessarily the same as leaving the house.

Disclosure statement

No potential conflict of interest was reported by the author.

Funding

This work was supported by the Nederlandse Organisatie voor Wetenschappelijk Onderzoek [grant number 276-53-007].

Acknowledgements

The author thanks Elise van Nederveen Meerkerk, Michiel de Haas, Kate Frederick, the anonymous reviewers, and the editors of History of the Family for their constructive comments on earlier versions of this article. Many thanks also to the organizers and participants of the workshop ‘Family, Demography and Labour Relations’, December 2014, International Institute for Social History, Amsterdam.

Notes

1. Opposing arguments have been made however: see Hatton and Bailey (2001) and Shaw-Taylor (2007).

2. Twente is a region in the east of the Netherlands. Throughout the text, the largest city of this region (Enschede) will in several instances be used as a case study within Twente.

3. The connection between child mortality and women’s industrial labour has been explored in great depth for the case of England and Wales (Garrett, 1998).

4. See for the underlying data: www.volkstellingen.nl

5. The prohibition of children’s factory work until the age of 12 positively affected the number of children receiving primary education. According to an investigation by the ministry of national affairs in 1892, circa 90% of all the Dutch children aged 6–12 were in school (Nationaal Archief, Archives of National Affairs, department education). Furthermore, in Enschede, factories were connected to a factory school providing education for employees between the age of 12 and 16. To be accepted in the factory and the factory school a primary school diploma was required. Since for many households the children’s incomes were indispensable, parents were therefore forced to send their children to school until they turned 12.

6. Including men.

7. The difficulty with this source is that sometimes, two different objects were produced by the same homeworker. However, the total weekly wage and amount of working hours were calculated per person, but noted down again for every object. I removed all the duplicates when calculating the averages in Table 4.