School of Creative Arts and Humanities, Charles Darwin University, Darwin, Australia
Received Date: December 01, 2016 Accepted Date: March 16, 2017 Published Date:March 23, 2017
Citation: Doran C (2017) The Price of Prejudice: Women’s Work and Labour Force Discrimination in Australian History. Anthropol 5: 177. doi:10.4172/2332- 0915.1000177
Copyright: ©2017 Doran C. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Anthropology
This article examines changes in women’s participation in the workforce and female wage rates in late nineteenth and early twentieth century Australia, focusing on the period of boom, depression and recovery from 1890 to 1910. Instead of explaining the trends in terms of family income or labour market segmentation, it is argued that the changes observed can be explained on the basis of sex-based discrimination in the labour market. Evidence from Victorian manufacturing during this period supports the discrimination hypothesis. Statistics for manufacturing industry in Brisbane during this period also give support to the discrimination hypothesis. As the economy slipped into depression and the general level of incomes fell, people were less willing to pay the price of prejudice against women workers. Both employment discrimination and wage discrimination against women were temporarily alleviated. Consequently there was the concurrence of rising employment of women relative to men and rising women’s wages relative to men’s wages.
Women; Labour; Female discrimination; Prejudice; Wages
The idea that women’s income merely supplements that of a male breadwinner at the head of every household, the belittling of women’s earnings as “pin money” or “pocket money,” is of long standing, and has not yet been entirely laid to rest, as some analyses of the current unemployment problem show. Before the introduction of equal pay in Australia in 1972-75, this concept was a cornerstone of male economic domination. The efforts of both official statisticians and academic writers served to prop it up; governments, unions and employers, by institutionalising female dependence, turned it into a self-fulfilling prophecy. Recent historical research has cast doubt on the validity of this notion of the merely supplementary female wage for describing even nineteenth century Australian society , and has thus brought into question historical explanations founded upon it.
In many recent studies of women’s labour force participation, and of the continuing disparity between male and female wages, there has been a heavy emphasis on labour market segmentation as an explanatory factor [2-4]. Labour market segmentation refers to situations where the labour market is divided into clear sub-sections, with reduced mobility between the sub-groups. Applied to women’s work, it is the idea that women’s employment is largely confined to certain subsections of the labour market, often associated with women’s traditional nurturing roles. In Australia market segmentation has often been used to explain historical and continuing wage differentials between women and men, almost to the exclusion of any other explanatory factor [5- 8]. This article takes a different approach, inspired by the groundbreaking theories of Gary Becker with regard to the salient influence of sex-based discrimination on labour market outcomes. It focuses on a critical period of Australian history around the turn of the twentieth century. During the 1890s the Australian economy experienced a severe depression. The effects of the depression, and of the subsequent recovery, on women’s participation in the workforce and women’s wages are the subject of this study .
In two articles in the Economic Record and Historical Studies, W.A. Sinclair provided useful estimates of female wages and labour force participation rates in Melbourne and Adelaide over the century from 1871 to 1971 [10,11]. Analysing short term changes in participation between 1881 and 1911, Sinclair emphasised changing conditions of supply of female labour. Sinclair suggested that supplementation of family income by unmarried daughters is the key to the relationship between general economic conditions and female participation in the labour force.
This article suggests an alternative interpretation of the data. To explain short term changes in women’s participation and wage rates, it emphasises changing conditions of demand for female labour. The key to the relationship between general economic conditions and female participation and wages is found in the ingrained prejudices of a male dominated society. According to this alternative hypothesis, women workers are portrayed as the victims of discrimination rather than as economic subsidiaries of men.
Women’s participation in the labour force has to be understood within the context of the complex social system in which it is embedded. Economic discrimination is one factor that shapes women’s involvement in the labour force and should not be seen as the complete explanation. Nevertheless, this factor has been relatively neglected in the available studies of the history of women’s employment in Australia, in favour of the dominant theories of market segmentation and family income. It is the aim of this article to argue that discrimination needs to be given more weight in academic analyses.
The paper first outlines briefly Sinclair’s “family income” hypothesis; then presents the alternative “discrimination” hypothesis and empirical evidence which supports it; and finally considers some broader implications of the analysis.
W.A. Sinclair provides data on female participation in the work force in Melbourne and Adelaide from 1871 to 1971. The use of decennial census statistics poses some problems: as P.G. Macarthy has pointed out, the census years of 1891, 1901 and 1911 were rather uncharacteristic of the intervening decades, with relatively low rates of unemployment being reported . For this reason annual data are used to support the argument presented in this paper. Annual data, together with literary evidence from the period , reinforce Sinclair’s conclusion that female participation in the workforce did increase during the depression decade of the 1890s. Annual real wage statistics suggest that this greater participation was associated with rising rather than falling real wages (Tables 1 and 2).
|Year||$||% Change||$||% Change|
($=Weekly Wages in Pounds, Shillings and Pence) (Based on Macarthy 1970a, 61).
Table 1: Money wages in Victorian manufacturing.
|Year||Male||Females||Ratio of Women’s to
Male Wages Index: 1911=100
Female Wages Index: 1911=100
Table 2: Real wages in Victorian manufacturing during 1891-1911.
Sinclair’s statistics show remarkable long term stability in female participation in the labour force until after the Second World War; only after then is there a general rising trend. However, in the short term there is significant instability in female participation rates, which shows an inverse relationship to the general economic climate. Thus, considering the picture in Melbourne, female participation falls from 1881 to 1891, recovers by 1901, then falls again in 1911, rising again in 1921. As Sinclair points out, Melbourne enjoyed boom conditions in the 1880s, suffered severely in the general economic depression of the 1890s, and took part in the recovery of the first decade of the twentieth century. During prosperous times women’s participation in the workforce declined; during recession it increased. Sinclair concentrates on the situation Victoria, as I will too.
According to Sinclair’s decennial figures, wage rates, in contrast to participation rates, show a direct relationship with general economic conditions. Thus both real and money female wages rose in the decade to 1891 and then fell by 1901. Real wages rose again to 1911, and again fell to 1921 (Table 2).
To explain these short term movements in participation and wage rates, Sinclair employs a proportional supply curve for female labour, which represents the proportion of women of working age prepared to accept employment at various wage rates. Shifts in this curve are used to explain the observed movements in wages and participation.
The relevant evidence is consistent with instability in the proportional supply curve, so defined. The lower rate of participation of unmarried women in Melbourne in 1891…can be seen…to have coincided with a substantial rise in women’s wage rates. Assuming that the wage rates in 1881 and 1891 approximated to equilibrium positions, the proportional supply curve must have shifted inwards in the intervening period. In the next ten years, the reverse movement appears to have occurred .
Thus Sinclair posits an inverse relationship between the availability of women for work and economic prosperity. The key to the relationship, it is suggested, is that women’s wages were regarded as a supplement to family income. When the economy was buoyant, the family was less in need of this supplement, since the male “breadwinner” was more likely to be in work and male wages were higher.
An alternative explanation of the data rests upon the economics of discrimination. Discrimination is said to occur when equally productive workers receive different treatment in the labour market. Economic analysis of discrimination was given impetus by Gary S. Becker , whose landmark theorising on the subject suggested that prejudice could be seen as a “taste for discrimination,” for which consumers of prejudice would willingly pay a price. Thus members of a particular group would earn less than their abilities or productivity warranted if other persons were willing to “pay” in order to avoid employing, working with, or dealing with them. For example, an employer who is motivated by prejudice will be unable to select the most productive workers; consequently s/he will pay for this prejudice in lost productivity. Discrimination in the market place would not exist if everyone maximised money incomes and if all markets were competitive, as conventional neoclassical economists assume. It could persist, however, if some people were willing to sacrifice part of their money income in order to indulge their taste for discrimination . Becker’s theory of economic discrimination continues to be influential in studies of the labour market, and indeed it has experienced something of a revival in recent years, as the work of Kerwin Kofi Charles and Jonathan Guryan has shown [15,16]. As Guryan and Charles acknowledge, Becker’s work included “deep insights” regarding the working of discrimination in markets . Becker’s body of work as an economist has recently been criticised as representing the epitome of the mentality of neoliberalism , but his contributions on economic discrimination offer useful approaches for women and other disadvantaged groups.
According to Becker, discrimination could take either of two forms: employment discrimination, where members of a particular group were simply refused employment when they would have been employed if they had not been objects of discrimination; and wage discrimination, where they were offered wages less than their productivity would warrant. Employment discrimination is relatively straightforward; attention will therefore be given to analysing the forms of and scope for wage discrimination. During the late nineteenth and early twentieth century, in Australia and elsewhere, there was always a sizeable gulf between the wages of men and women. As a very rough approximation, men’s wages were more than twice those of women. The gap can be analysed as follows. First there were probably differences in what economists call the “marginal productivity” of male and female workers. For instance, it is probable that the greater average strength of the male would often have given a productivity advantage at a time when work was more laborious, and mechanisation less advanced. It is on grounds like this that the differences between male and female wages have traditionally been defended. However, even within this “productivity” component of the gap between male and female wages there was scope for prejudice. If goods produced by women, or services provided by them, were valued less by the consuming public than similar goods or services produced by men, then women’s “productivity” was indeed lower. For example, a consumer may have been willing to pay more for a suit made up by a tailor than for an identical garment produced by a tailoress. Clearly this factor would have been more significant when there was direct contact between the buying public and producers. The “productivity” of women workers would also have been influenced by the prejudices of co-workers. If employing a woman provoked resentment among male colleagues, or disharmony in the work place which reduced output, then in the eyes of the employer, the woman’s productivity was reduced.
Thus prejudice on the part of consumers or co-workers could actually reduce the productivity of a female worker; this would then be reflected in a lower wage. These aspects of prejudice have usually been neglected by those economists who maintain that discrimination is impossible under competitive market conditions. On top of these sources of prejudice, there is Becker’s notion of the “taste for discrimination” on the part of employers, where employers are willing to sacrifice money income in order to practise discrimination. The fact that perfect competition is seldom found in the real world allows further scope for the exercise of prejudice against women workers. When competition is less than intense, or when markets are not contestable, pure profits can be earned, and can be taken out in a variety of forms; one of these is indulgence in prejudice. Hence there are several kinds of prejudice which may contribute to the difference between men’s and women’s wages.
The effect of changing economic conditions on each of these forms of prejudice helps to explain the evidence about movements in wages and participation of female labour around the turn of the twentieth century. In general, as income falls, the price which consumers of prejudice will be willing to pay in order to indulge their “taste for discrimination” will be lower than in more prosperous times. The analogy here is with a “normal good” and a leftward shift in the demand curve for “prejudice” as income decreases, so that the consumer of prejudice will demand less at any given price.
Therefore as economic conditions deteriorated and incomes fell, employers were less willing to pay the price of prejudice against women. Maximising money income became a relatively more important motive. They were more likely to give equally productive workers equal chances and conditions of employment. There could have been less employment discrimination, or less wage discrimination, or, as appears to have been the case, a combination of both. Male workers who had previously been favoured were faced with falling relative wages and unemployment.
Depressed economic conditions can also affect the prejudice components of productivity differences between men and women. As their incomes fall, consumers are less willing to pay for their prejudice against goods and services produced by women. Male co-workers, more anxious to retain their jobs, are forced to tolerate women on the job; moreover the fact that numbers of women workers are increasing makes it more difficult and costly to enforce their exclusion. Thus the “productivity” of women actually rises as prejudice declines, which is reflected in higher wages. Another factor was that the productivity difference may also have been diminished because, under economic pressure, manufacturers speeded up mechanisation and rationalisation of work processes. Women tended to be more amenable to these changes, which sometimes fell foul of entrenched male skill categories [13,18].
Another source of changes in wages is trade union activity. W. Nicol has convincingly shown that in New South Wales the attitude of male trade unionists to women’s unionisation depended on general economic conditions and events in the other Australian colonies followed a similar sequence. In the boom period from 1880 to 1889 women were often excluded from male trade unions; men made little effort to organise women workers and often were hostile to women’s attempts to organise themselves. These attitudes underwent a remarkable reversal with the onset of the depression. Suddenly male unionists gave high priority to the recruitment of female co-workers. According to Nicol, male efforts to organise women workers in this period “are probably best described as frantic” . Unorganised, low paid female workers were now being given preference in employment, taking jobs from men, so men’s unions were desperate to bring women workers under their aegis.
The price male unionists had been willing to pay for their prejudice against women was the lack of a “closed shop” and the existence of a large pool of non-union, low-wage female labour. When employers’ willingness to indulge prejudice changed because of harsher economic circumstances, unionists found the price even higher than they had anticipated. Unfortunately accurate statistics for union membership are not available for the period before 1910 [19,20]. However, unionisation of women, relatively slow as it was, was probably another factor working towards closing the gap between male and female wages.
It is often observed that during economic crises there is a crescendo of publicly expressed antagonism to women’s participation in paid employment, which apparently contradicts the suggestion of diminishing prejudice. However, a distinction should be made between the verbal protests of an articulate few and the economic behaviour of the many, which is guided by material forces largely outside their control.
To check whether the evidence supports the discrimination hypothesis, I will look more closely at manufacturing industry in Victoria. At the end of the nineteenth century, the vast majority of working women in Melbourne were employed either in domestic service or in manufacturing industry. In the period 1891 to 1901, the depression era, domestic service declined as a field of women’s work and manufacturing rose at a time when total female participation was increasing. It is appropriate to focus on the rising sector as more characteristic of the period; in addition, statistics for manufacturing are more comprehensive.
The discrimination hypothesis would suggest that as wage discrimination decreased with the onset of depression, the gap between men’s and women’s wages would decrease. In fact women’s wages in manufacturing rose significantly relative to men’s in the first six years after 1890 and they were still higher in 1901 than they had been in 1891 (Table 1). It is true that this increase in women’s relative wages was not monotonic year by year, (See column 3, Table 2), but the overall trend is clear.
Data in Table 2 under the heading “Ratio of women’s to men’s wages” is actually a measure of the gender pay gap, and shows its fluctuation over time. It shows that over the decade of the 1890s, from 1891 to 1899, as the grip of depression tightened, the gender pay gap was closing. The rise in the ratio of women’s to men’s wages to 108 indicates an 8% reduction in the gender pay gap. Over the subsequent decade, from 1900 to 1909 the wage gap widened again, to women’s disadvantage.
Women’s real wages rose from 1891 to 1895 and were higher at the end of the decade than at the beginning. It is not surprising, therefore, that the higher real wage offered to women after 1891 called forth a greater supply of female labour. It is not necessary to postulate any theory of women’s “supplementary” income or any shift in the supply curve to account for this. The economist’s distinction between a “movement along” the supply curve due to a change in an endogenous variable, in this case, price, and a “shift in” the supply curve due to a change in an exogenous variable, is crucial here.
The shift was more likely to have been in the demand curve for female labour so that, their discriminatory tendencies diminished, employers were willing to hire more women at any given wage rate. Consequently employment of women in Victorian manufacturing rose appreciably; indeed by the end of the 1890s the proportion of women in Victorian manufacturing had doubled (Table 3).
|Year||% Employment of Female Workers|
Table 3: Female workers as a percentage of employment in Victorian manufacturing during 1890-1909.
It therefore seems that declining discrimination in Victorian manufacturing was reflected both in decreasing wage discrimination and in decreasing employment discrimination. In the first half of the decade declining employment discrimination seems to have been predominant, whereas in the latter half decreases in wage discrimination became more important. It is worth noting the high rate of participation by women in the manufacturing labour force by the end of the nineteenth century (29%), which gives the lie to the notion that women did not enter the work force in significant numbers until the latter part of the twentieth century.
Investigation of manufacturing industry in Brisbane, Queensland during this period lends further support to the discrimination hypothesis. The main problem is lack of consistent statistics. In Queensland legislation for shops and factories was passed only in 1896 with the Factories and Shops Act; not until then were statistics regularly kept. Manufacturing industry was less developed in Queensland than in the southern colonies, and was concentrated in the capital of Queensland, Brisbane.
In those years for which statistics are available, the Brisbane figures show a higher proportion of females in manufacturing and a pattern similar to that of Victoria. Note again the significant role of women in the manufacturing workforce (33% in 1901) (Table 4).
|Year||% Employment of Female Workers|
Table 4: Female workers as a percentage of employment in Brisbane factories.
The only figure available for comparison with earlier years is from the 1891 Queensland Census, which shows females making up 18% of those in occupations in Brisbane classed as “industrial” . However, this census category did not coincide exactly with “factories” under the Factories and Shops Act, but also included heavy construction industry in which women were probably sparsely represented. 18% is already higher than the figure for female representation in Victorian manufacturing in 1891 (16%), and because of the difference in categories it probably underestimates female involvement in manufacturing in Brisbane. Perhaps the fact that the depression began in Queensland well before 1890 is relevant here.
From 1891 to 1901 the labour force participation rate for women in Brisbane rose slightly from 38.2% to 39.1% (ibid, 1146, 1238; Census of 1901, 1902, 1008, 1192). The participation rate in Queensland as a whole fell from 34.4% to 30.1% . These statistics are consistent with the fact that the depression was markedly more acute in Brisbane than in the rest of Queensland. The difference between female participation rates in Victoria and Queensland may reflect the earlier onset of depression in Queensland, and earlier recovery.
Wage statistics, however, do provide clear support for the discrimination hypothesis. Comprehensive statistics for wages were kept from 1898; the only means of comparison with earlier dates is through the analysis of wage rates by the 1891 Royal Commission on shops and factories. Comparison indicates that women’s wages had indeed increased relative to men’s from 1890 to 1900. To give a few specific examples of weekly wages, females in the confectionery trade received from 5 shillings (s.) to 15s. in 1890, and men from 75 to 80 shillings; in 1900 females received an average of 15s and males 46 shillings. In this case the gender pay gap was more than halved. In cordial manufacturing females were paid from 11s. to 15s. in 1890, while men received 40 to 60; in 1900 females received 11 shillings and males 31. This represents a reduction in the gender pay gap of 46%. In the clothing industry, a tailoress received from 3s. to 3.5s. in 1890 for making trousers, while a tailor earned from 8s. to 10s.; by 1900 the women’s relative pay had increased significantly so that tailors earned an average weekly wage of 42.5s. for trouser making and tailoresses 18 . In each of these instances the ratio of men’s to women’s wages fell during the period. At the end of the century, the wages of women in Queensland relative to men were similar to those in Victoria . Compared to Victoria, declining wage discrimination seems to have been more important in Queensland manufacturing, instead of a reduction in employment discrimination.
The main objective of this paper is not to dispute Sinclair’s statistical evidence on female participation in the workforce at the turn of the century. Nor is it primarily to present new statistical data on female employment or wages. Instead, the purpose is to suggest a new interpretation of the available data, one which explains Sinclair’s employment data, but which is also consistent with evidence, derived from P.G. Macarthy’s work, on wage rates of the period. A lesser aim is to make available some new data on female employment in Brisbane. Admittedly these new figures, at least for the period before 1896, are somewhat limited; however they are the best available. Sinclair’s article gave information on female employment in Melbourne and Adelaide; the new Brisbane data helps to fill out a wider picture, as well as providing additional support for the new hypothesis being advanced.
W.A. Sinclair’s family income hypothesis not only fails to account for increases in women’s real wages during the 1890s depression; nor does it explain the rise in women’s wages relative to those of men during this period. The discrimination hypothesis is capable of explaining these wage movements as well as changes in female participation in the labour force. The discrimination hypothesis is supported by employment and wage data for manufacturing in both Melbourne and Brisbane.
As the economy slipped into depression and the general level of incomes fell, people were less willing to pay the price of prejudice. Both employment discrimination and wage discrimination against women were temporarily alleviated. Hence there was the remarkable concurrence of rising women’s wages relative to men’s, and rising employment of women relative to men.
This interpretation of the data places emphasis on the demand for rather than the supply of female labour. This emphasis is consistent with the conclusions presented in a significant study of Australian labour force participation, which suggested that empirical data on women’s wages and employment should be seen as points on the demand curve for female labour rather than the supply curve .
No claim is made here that this analysis provides a complete explanation of women’s employment experience during the 1890s depression. A considerable amount of fruitful research is currently underway on the changing structures of the economy, the labour force and the family in this period; all these would have to be taken into account in any comprehensive history, though a great deal more work needs to be done before the complex processes at work are fully understood, a task made more difficult in some cases by the paucity of reliable statistical material. The claim made here is that the discrimination hypothesis has sufficient theoretical cogency and empirical support to warrant a place within that comprehensive history, whenever it should be written.
Until recently, the idea that women represented a “reserve army” of labour, which could be drawn into the workforce in periods of economic growth and expelled from it during recession, has held considerable sway in historical analysis of female employment [25,26]. The evidence analysed in this paper contradicts this thesis, as does evidence from the 1930s depression [27,28]. Women’s participation in the labour force increased during depressions, rather than the reverse; this paper suggests an explanation of the phenomenon in terms of the economics of discrimination. According to this interpretation, the “reserve army” concept remains valid, but in a sense rather different from the original Marxist-inspired one; women still wait in the wings to be called upon by employers, but the timing of their appearances and the reasons for their entry are different to what was previously assumed. In periods of economic crisis, as in times of national emergency such as the Second World War, the social cost of indulging in discrimination becomes less acceptable. That society pays a price when it discriminates against women is the underlying point to be stressed. More than 2,000 years ago Plato pointed out in The Laws the absurdity of the state reducing its strength by half by not having women develop the same pursuits as men with all their strength and mind. Although much has changed for the better over two millenia, there is still a notable price to be paid when society indulges the “taste for discrimination.”