EHS 2010, Durham
J. F. Field
Service, gender and wages in England, c. 1700-1860
Jacob F. Field, Cambridge Group for the History of Population and Social Structure, University of Cambridge, [*]
Introduction
Domestic service was a vital occupational sector in England in the eighteenth and nineteenth centuries – particularly for women. By 1851, according to the Census, at least 24.5 per cent of women over twenty in reported employment were engaged in domestic service – compared to 1.5 per cent of men.[1] This was by no means a nineteenth century phenomenon. Even in the eighteenth century, England had high numbers of servants, fuelled by rising urbanization and growth of a middle class.[2] Domestic service was essential to the English economy. As Eddy Higgs points out, by performing domestic labour, women maintained the efficiency of workforce and so contributed to the economic system.[3] In addition, service played an important role in the lifecycle of many individuals before marriage. Entry ranged from early to late teens.[4] It was frequently undertaken as a means to save money and gain skills before setting up their own households and can be defined as ‘lifecycle’ servants. However, there were also many who were ‘career’ servants. They frequently served as senior staff in multi-servant households.[5] They tended to be older, better trained and skilled, and hold some form of supervisory authority, in comparison to ‘lifecycle’ servants, who tended to be relatively unskilled and directed.[6]
In spite of its importance, the wages paid to servants on a national level are not well understood. The only national series of wage levels for servants was produced over fifty years ago by J. J. Hecht, using information from eighteenth-century newspaper advertisements and literary sources.[7] Although this is very useful, this dataset has the drawback that this does not record the cash wages that were actually paid, or agreed to be paid, in a household. Also, it is not subdivided by region, and is probably very heavily skewed towards the upper end of the social scale. There are also some useful local wage series collected for domestic servants – for example, Pamela Sharpe’s for Essex from 1736 to 1835, and Tim Meldrum’s for London from 1660 to 1750.[8]
Sources and methodology
The dataset used for this paper is based on records of wages paid to domestic servants in household and estate account books from across England. So far, around 1,500 wage observations from 56 separate households or establishments have been collected from c. 1700 to 1860. These observations were collected as part of a major nationwide survey and sampling exercise of primary sources (funded by the Leverhulme Trust as part of the Occupational Structure of Britain Project at the Cambridge Group for the History of Population and Social Structure, University of Cambridge) that might be of use in quantifying women’s occupations from c. 1600 to 1850.[9] This survey aimed to find primary sources that specified when, where and what were women were paid for work in the pre-Census period (at the moment only paid remuneration has been considered). Payments made to domestic servants were one of the few types of sources that consistently stated the exact names and occupations of the payee. However, it appears that they only begin to exist in large numbers from around 1700. As such, once the survey was completed, we were left with a large amount of detailed observations for service wages, and information about the precise duties of domestic servants, which formed the dataset for this paper. As many of the sources have only been sampled for one or two years, it should be possible to at least double the size of the dataset by going back to the archives and taking complete transcriptions of the most useful and detailed sources describing service wage payments. The full list of primary sources used in this paper is set out in the appendix. I would welcome any suggestions on possible additions to this list.
Figure 1: The dataset, subdivided by Census Registration District
Source: See Appendix.
Figure 1 shows the geographical spread of the dataset. It is fairly comprehensive, although more data from the South West, the East Midlands, East Anglia and the North East would improve its coverage. There are three establishments from which large numbers (more than 100) observations were taken. Two are large aristocratic households: the Earls of Lonsdale in Lowther, Westmoreland, and the Earls of Lichfield in Shugborough, Staffordshire. The third is not from a large aristocratic household – rather it from Sir William Oglander’s establishment in Parnham, Dorset, and there are so many observations because it wage book maintains a very high level of detail for over fifty years.[10]
In order to fully appreciate the wage differentials between different types of servant only wage information with exact occupation considered – those only described as ‘maid’ or ‘servant’ not considered. Some of the accounts used were specific ‘servants’ wage books’, which recorded all of the wages paid to the establishment. Others were records of the servants in a household for a particular year, and some were extracted from general household or estate accounts. This approach has its limitations. The level of detail varied considerably even in the same document. Frequently exact occupations or full names were not recorded, and many wage books only recorded amounts paid to new servants, meaning long-term members of the household may have been ignored. A further major drawback is that this approach relied strongly on wages paid in large gentry or aristocratic households for which suitable accounts survive. This also meant that the dataset was strongly rural, with very few observations from urban areas.
In almost all cases, the wage recorded in contemporary documents was an annual amount. Most domestic servants were retained for a yearly term, and paid wages quarterly.[11] This was not always the case in practice. Many servants were not paid regularly, but given a lump sum at the end of their period of service. I have found one rather extreme example of a housemaid in Yorkshire who was not paid for over twenty years from 1749 to 1769, and was eventually owed £205 when her master died. The executor refused to pay this amount to her, as when the estate was settled she only claimed a house bill for £15 as a debt outstanding to her. Subsequently she was refused payment of her backdated wages.[12] Also, most of the wage observations collected were for at the start of the time in service, so they do not take into account any raises in wages over time. However, after examining the accounts that do, for the most part these only tended to be a pound or a guinea a year for no more than two or three years.
A further problem when using annual salaries for domestic servants is the matter of non-wage compensation. Compared to other sectors, the nominal wage of a servant seemed low because it did not include the value of the food and lodging they received. In all probability, English servants probably only earned half of their compensation in terms of monetary wages.[13] Eddy Higgs valued this non-wage compensation as being worth five shillings per week in the first half of the nineteenth century, making domestic service wage rates comparable to the general industrial average.[14] This ‘flat rate’ approach would be most useful for adding the value of food and lodging, as the only variation would be by gender for the differential in the amount of food consumed. In addition, some senior domestic servants may have often ‘lived out’, and so would have had to pay their own rent.
As well as food and board, there were the expected gratuities, gifts and rights due to servants. These varied by position. For example, cooks had the right to sell cooking fats and ashes.[15] Clothing was frequently purchased for some servants. In aristocratic establishments, it was commonplace for male servants to have their livery purchased for them, whereas female servants would have had to purchase their own clothing to work in. The type of clothing that would be purchased varied. For example, in the 1760s, in the Earl of Lonsdale’s Cumberland household, male servants who worked outside were given from 2s 2d to 13s 6.5d per year ‘boot money’, but in 1800 the footman of the Jervis Family in Staffordshire had to purchase his own boots.[16] ‘Tea money’ was also frequently added to wages, but this did vary. In the Collier household of Hertfordshire in 1783, only the housekeeper was given tea and in Sir William Oglander’s household in Dorset in the 1810s, all servants had to find their own tea.[17] Tips would have added considerably to a domestic servant’s wages. For example in the household of the Baronets of Foremark in Derbyshire in the 1740s, housemaids were generally given an extra guinea per year in addition to their £5 wages ‘for encouragement’.[18] The practice of giving vails, tips given to the household by visitors, would have been an important extra for most servants, and could double the wages of a maidservant.[19] There was pressure to phase out this practice in the 1760s, but it was never formally abolished.[20] The essential point when considering extra-salaried income is that it varied over time, position and household. For this paper, I have not attempted to add anything to my service wage levels for non-wage earnings, but I would be grateful for suggestions for how to go about this, and do appreciate it is necessary if this dataset is to be comparable to other sectors.
Although the data collected assigned each individual an exact position, there is the problem that for the most part the label applied to servants probably did not matter as far as day-to-day duties were concerned.[21] Carolyn Steedman notes, ‘authority over the servant’s time was perpetual: the employer had obtained the day-in, day-out attentions of a servant’.[22] Particularly in smaller households servants had to be generalists – performing whatever work was at hand.[23] Some servants occupied multiple positions – the most common was the label ‘housekeeper and cook’, but I have also found positions such as ‘baker and footman’, ‘gardener and bailiff’ and ‘nurse and laundry maid’.[24] Occasionally accounts of servants’ wages reveal more than just a position. The servants’ wage book of the Anderton Family in Lancashire for the early nineteenth century recorded that ‘Margaret Burchall entered my service as dairy maid, & to assist in the Kitchen’ at £8 per year and ‘Ellen Walker entered my service as Laundry Maid & to take care of Master Anderton’s Clothes when at home to work at her needle & assist occasionally in the house work’ at £8 per year.[25] The labour of a servant was undertaken both indoors and outdoors, and distinctions between the duties of different types of servants were usually blurred, and domestic work not confined to four walls of the house.[26] For example, the descriptions of duties given for the servants of the Churchill Family in mid nineteenth century Gloucestershire recorded that the laundry maid was also meant to ‘keep clean the yard’ and the hall boy was required to work in the stables ‘if wanted’. Conversely, the dairy maid was required ‘when that work diminishes… to employ her leisure time in needlework & make herself generally useful’.[27]
In an ideal world, it would be helpful to have some form of time use sheet for domestic service to gain an idea of how exactly their time was spent. This may have broader implications for the sectoral division of occupations. Classically, service has been considered an example of the tertiary sector par excellence. However, particularly for females, employment as a servant encompassed roles and tasks associated with the primary and secondary sectors. Maidservants would have often performed ‘primary’ tasks such as milking, husbandry and harvesting. In all probability, they would have also worked alongside family members in secondary sector productions such as brewing, baking, spinning, knitting and plain sewing.[28] This would have varied by household – in larger and wealthier ones there would have been greater specialisation amongst domestic servants. For example the Earls of Lichfield employed Mary Dakin solely as a ‘burnisher’ (silver polisher) at £31 10s per year in 1817,[29] whilst in smaller households, domestic servants would have been required to engage in a wide variety of tasks.
Analysis of the wage observations
The nominal wages were adjusted for changes in prices, using Bob Allen’s consumer price index for London,[30] and then rescaling to 100 to the earliest, lowest, wage observation: £4 per annum paid to a groom, John Duport, in Glemham, Suffolk.[31] The dataset was divided into three periods: 1700-75, 1776-1825, and 1826-60, and the male and female servants were subdivided into fairly broad occupational groups (see table 1, below). These broad occupational groups were then divided between ‘career’ and ‘lifecycle’ type servants. Geographically, the data was divided into three zones: North, Southeast, and Southwest.[32] I have used such broad divisions in order to maintain numerically large subsamples of data. Adding to the dataset would enable me to break down the data into smaller subdivisions, by time, region and position. Ideally, I plan to add London later, where wages were generally higher – perhaps by ten to fifteen per cent - than the rest of the nation.[33] However, I have had difficulty finding sufficient numbers of detailed account books for London, and would appreciate any suggestions on potentially useful sources.
Table 1: Domestic service annual wage levels in England, by position in household, 1700-1860, (adjusted for consumer price index, 100=Groom in 1704)
Position / Number / Mean / Median / Coefficient of variationMale
Bailiff / 33 / 730 / 568 / 0.727
Butler / 55 / 494 / 522 / 0.412
Cook / 33 / 640 / 516 / 0.645
Gardener / 98 / 364 / 392 / 0.407
Coachman / 69 / 323 / 281 / 0.676
Under Butler / 24 / 251 / 268 / 0.313
Gamekeeper / 73 / 313 / 249 / 0.662
Footman / 106 / 221 / 208 / 0.352
Groom / 175 / 241 / 186 / 0.666
Boy / 92 / 125 / 119 / 0.368
Female
Housekeeper / 65 / 375 / 368 / 0.536
Lady’s Maid / 32 / 256 / 186 / 1.213
Cook / 75 / 197 / 174 / 0.571
Nurse / 32 / 177 / 151 / 0.576
Maids (all) / 545 / 116 / 105 / 0.419
-House / 149 / 123 / 119 / 0.387
-Laundry / 132 / 122 / 115 / 0.385
-Nursery / 31 / 117 / 114 / 0.340
-Kitchen / 86 / 117 / 106 / 0.419
-Scullery / 22 / 94 / 102 / 0.288
-Still Room / 23 / 104 / 99 / 0.394
-Dairy / 68 / 106 / 86 / 0.583
-Chamber / 27 / 88 / 84 / 0.331
Source: See Appendix.
Note: Each position must have at least twenty observations to be included in this table. Thus, it excludes governesses, poultry maids, laundresses and female gardeners.
Table 1 shows the range of household positions. As Sharpe pointed out, service wages were on a sharply graded scale.[34] There are no real surprises in its results – the wages broadly match Hecht’s proposed hierarchy of servants.[35] The groups were divided into ‘lifecycle’ and ‘career’. The division made between ‘lifecycle’ and ‘career’ servants is based on the wages paid to them as well as recording of their duties, expertise and responsibilities. Thisis a very broad division, and there were exceptions to it. For women, I have deemed all maids (except lady’s maids) to be ‘lifecycle’, and the rest ‘career’. For men (and boys) the division is more difficult. For this paper I have deemed boys, grooms and footmen to be ‘lifecycle’ servants, and the rest ‘career’. However, other divisions may be more helpful – perhaps by creating a ‘professional’ group (bailiff, butler, and cook). Ultimately I would like to collect more data than just wage information, such as age and years served, before making a final division into ‘lifecycle’ and ‘career’ for male servants.
For male servants, cooks, butlers and senior estate administrators received the highest wages, whilst the various ‘boys’ employed around the household received the lowest. The greatest variation was amongst bailiffs, and the lowest amongst ‘boys’ and footmen. In general, it appears that there was the greatest variation amongst those positions where the skill level and responsibility would tend to vary most. For female servants, occupations with increased responsibility and some kind of managerial capacity received the highest wages, such as housekeepers and cooks. The large mean and variation for lady’s maids is partly due to the £66 14s 5d annual wage paid to Lady Cowper’s maid in Hertfordshire in 1743 (which was the highest female wage in the entire dataset).[36] The multiplicity of different types of maid received the lowest wages, and there was the least variation amongst their wages. However, when the maids were broken down into more precise categories, some more trends emerge. House, laundry and nursery maids were paid at the highest levels, with dairy and chamber maids paid the lowest. Even when the maids were subdivided by type, it is striking how little variation there was in the wages paid to them.
It is clear that male servants generally received higher wages than female servants, but direct comparison of the level of this difference is complicated by the fact that their positions are not strictly comparable given the differences in duties and work. However, some brief comparisons can be made. Firstly, the wages paid to housekeepers and butlers (both positions of similar managerial authority) were compared between the first and last fifty years of the sample, and secondly the position of cook, where female cooks were regarded as inferior in talent and knowledge to males.[37] From 1701-50 to 1811-60, the difference between the wages paid to housekeepers and butlers grew from two per cent to 23 per cent. The difference in wages paid to male and female cooks was even more dramatic: growing from 265 per cent in the earlier period to 350 per cent in the later one. Based on this observation, it is possible that the wage differential between men and women was increasing from the early eighteenth century to the second half of the nineteenth. However, these differences could also be compositional, and this trend needs further examination.