The project blog has been quiet for a while – I’ve been working away on a series of articles from the project, Mark has taken up a lectureship in History at Bristol University, and the project has been joined by Dr Charmian Mansell in January 2018. However, to celebrate International Women’s Day, we’re starting a new series of blogs looking at the project’s main findings. Please read these alongside our piece on methodology, also available on the project website.
So what work did women and men do in early modern England? We collected 4300 instances of specific work tasks carried out by specific people from church court depositions, quarter sessions examinations and coroners’ reports from Devon, Somerset, Wiltshire, Hampshire and Cornwall, dating from between 1500 and 1700 (although mostly after 1550). We sorted these work tasks into ten broad categories of types of work, shown in table 1 below. These show that none of these broad areas of work were gender exclusive: all involved both women and men.
Table 1: Work tasks by category: raw data
|Agriculture and land||1077||864||213||19.8|
|Crafts and construction||443||335||108||24.4|
|Mining and quarrying||28||25||3||10.7|
However, there are two problems with the data in this form. (1) Women’s work in under-represented and (2) types of court cases affect the reporting of some types of work. 29.4% of work tasks we recorded were carried out by women (and 70.6% by men). Because we included all types of work, including housework and care work, this could only be a true reflection of the distribution of work tasks between women and men if women had more leisure time than men. This is extremely unlikely to have been the case. Time-use studies from almost all societies show that women work longer hours than men. Early modern commentators such as Fitzherbert and Tusser suggest that women’s work was never-ending and more time-pressured than men’s. In addition, we can show that the under-reporting of women’s work is largely a consequence of the courts’ preference for male witnesses. Only 26.5% of the witnesses who reported the work tasks we collected were female. Both men and women were more likely to report the work activities carried out by people of the same gender (91% of male work tasks were witnessed by men).
A simple solution to this issue is to make the (quite conservative) assumption that at least 50% of total work tasks were undertaken by women. To turn 29.4% into 50% requires us to multiple the number of tasks recorded as done by women by 2.41. If this is then applied to the totals of women’s work tasks in the different categories the findings look like this:
Table 2: Female work tasks by category with 50/50 multiplier applied
|Category||Female tasks recorded||Female tasks multiplied by 2.41||50/50 multiplier % of tasks by women|
|Agriculture and land||213||513||37.3|
|Crafts & construction||108||260||43.7|
|Mining and quarrying||3||7||22.4|
The final column shows the percentage of tasks within that category that were carried out by women rather than men with the multiplier applied. These findings look plausible: women dominate housework and care work, but are well represented in other categories – making up half of these engaged in commerce and over half of those carrying out management tasks (financial and work arrangements). Women make up a significant proportion of those carrying out manufacturing, construction and food processing, but slightly less than men. Even in the male dominated activities of agriculture and transport, over a third of work tasks were carried out by women.
While these figures offer a reasonably robust picture of the gender division of labour within each category, they are more problematic when used to infer a picture of women’s or men’s overall work patterns by making comparisons between categories. Table 3 shows the percentage of work tasks undertaken in each category by either women or men (that is, the percentages are calculated down the columns, not across the rows). Here the influence of types of court cases becomes problematic in some categories.
Table 3: The distribution of work tasks across categories by gender
Incidental and related data only (%)
All data (%)
Incidental and related data only (%)
|Agriculture and land||16.9||21.0||28.4||37.7|
|Crafts & construction||8.6||8.9||11.0||13.4|
|Mining and quarrying||0.2||0.3||0.8||1.1|
To guard against this we labelled tasks in the database according to whether they were ‘integral’, ‘related’ or ‘incidental’ to the court case (see the methodology piece for more details). Including work tasks that were integral (or very closely related) to cases inflated the totals of commerce and to a lesser extent, food processing (tasks commonly reported in theft cases in the quarter sessions). If we take these cases out of the analysis, and use only the incidental and related evidence, which was not heavily influenced by the court cases, we get a distribution that reflects the likelihood of women or men carrying out these activities, rather than of the courts recording them.
This is as close as we can get to time-use data. It has some similarities with the modern technique of ‘random spot check observation’, where data on time use is collected by contacting people at certain times of the day, and their activity at that time is recorded. It suggests the relative importance of different types of work to women and men. The findings show that agricultural work tasks (21.0%) had a similar importance to women as housework (21.3%), and that women’s work tasks in commerce (17.9%) were more frequently reported than care work (10.8%). Together women’s housework and care work took up about a third of their work time (32.1%): while this was much higher than the proportion of work tasks carried out by men involving housework and care work (5.2%), it suggests very strongly that the majority of women’s work time was spent doing other things – such as agriculture, commerce, craft production and food processing (53.4% in total).