Leigh Shaw-Taylor

What proportion of the male labour force do you think worked in agriculture at the beginning of the 18th century? Was it around 80 percent? Or 60 percent? Or 40 percent? The Occupational Structure of Britain c.1379-1911 project has shown that for England and Wales the correct figure is around 47 percent. This makes the English and Welsh economy much less agricultural (and much more industrial) than historians have previously believed.  

William Henry Midwood, The Reapers. Kirklees Museums and Galleries.

The importance of occupations and occupational structure was introduced in last week’s blog. If you haven’t read that, you might want to do so before reading on. 

In last week’s blog we looked at occupational data deriving from parish registers. From 1813 it was a legal requirement to record fathers’ occupations at the baptism of  a child, and for the period 1813-20 we have occupational data from virtually all 11,400 Anglican baptism registers, giving us complete national coverage. However, between 1600 and 1800, while we have hundreds of observations from parish registers, they are neither sufficiently numerous nor sufficiently well-spread geographically to create a reliable national estimate of occupational structure except around 1710. 

Probate documents

For the earlier period, data are available on a much larger scale and with much more comprehensive geographical coverage in the preambles to probate inventories and wills between 1600 and 1800. A probate inventory is a list and valuation of the goods belonging to a recently deceased individual. Both wills and inventories are known as probate documents. The picture below shows the preamble to a probate inventory. From each probate inventory or will, we extracted three key pieces of information: location, occupation, and date. 

Figure 1. The preamble to a probate inventory. Berkshire Record Office, Wiltshire and Swindon History Centre P23/331.

We have over a million observations between 1600 and 1800. However, the raw probate data cannot be used to estimate occupational structure, because there is a systematic bias in the probate record. Basically, the wealthier the individual was, the more likely s/he was to be probated.  

Sebastian Keibek devised an ingenious way round this problem as part of his PhD thesis. If we look at parishes where we have both data from baptism registers (which do not suffer from a wealth bias but cover all social groups), and probate data, we can calculate, for each occupation, the relative propensity to appear in a will or a probate inventory. This analysis shows that farmers, for instance, were 20 times more likely to be probated than labourers. If we used the raw data, we would massively overstate the share of farmers in the labour force, and severely underestimate the number of labourers.   

Having calculated the relative propensity to leave a probate document for the many hundreds of parishes where we have both parish register evidence and probate evidence, we can correct the bias in the whole probate dataset. This generates reliable estimates of occupational structure not just for the whole country but also for much smaller units, allowing us to map the geography of occupational change. Figure 2 (below) shows the estimated male occupational structure for 1381 to 2011. 

Figure 2. The male occupational structure of England and Wales 1381-2011.

Prior to 1600 we have estimates for just one year: 1381. These estimates were made by Richard Smith, and derive from documentation produced to administer the infamous poll tax of that year, which led to the 1381 poll tax revolt. Some villages and towns in the poll tax returns recorded male occupations consistently, and there are enough of these to allow us to estimate the male occupational structure for 1381.  The dashed line between 1381 and 1600 is purely conjectural, but we hope to replace this with a series of estimates from the Court of Common Pleas in the near future. 

The Peasants’ Revolt, miniature from a manuscript of Jean Froissart’s Chronicles, 15th century. British Library, London.

The evolution of the primary sector 

The rest of this blog considers the evolution of the primary sector over time. Future blogs will look at the secondary sector and the tertiary sector. If you need a reminder of what these terms mean, please see last week’s blog.

In Figure 2 (above), we have divided the primary sector into two components: agriculture shown in green, and mining shown in black. We have done this because the trends are very different. Mining, which was overwhelmingly coal mining (but also included lead, tin and copper mining), grew explosively and continuously from the middle of the 16th century up to 1921, before beginning its downward trajectory across the rest of the 20th century. This shows that the economy of England and Wales was using increasing amounts of coal rather than wood across the early modern period.

By the 1650s the amount of energy generated by coal exceeded that from wood, something which happened nowhere else in the world at this date.  This underpinned the transition from an organic economy to a mineral-based energy economy. The growth of mining was so rapid that by 1851 there were 57 times as many miners as in 1600. By comparison, total agricultural employment was only 1.7 times higher in 1851 than in 1600. 

Turning now to agriculture, in 1381 we estimate that 78 percent of adult males worked in agriculture. By 1601 this had fallen to 64 percent. We suspect but cannot yet prove that much of this change took place between 1550 and 1601. As we do more work on occupational data from the Court of Common Pleas, we will be able to test this hypothesis.

By 1701 the share of agricultural employment had fallen sharply to 47 percent, and by 1761 this had decreased slightly more, to 43 percent. The curve then plateaus for half a century before resuming its decline in the second quarter of the 19th century, reaching 28 percent in 1851. By 1911 it was down to 11 percent. 

W. Williams, ‘Plough Boy‘ (1795). The Royal Agricultural University Collection.

These changes in the share of the male labour force have major implications for productivity in agriculture, and in particular for labour productivity (how much one worker produces). In 1597-8, when the share of the labour force in agriculture was around 64 percent, England and Wales experienced its last major famine and one to two percent of the population is estimated to have died in famine-related mortality.

In the 1690s atrocious weather struck Scotland, England and Wales, and France, and led to reduced harvests in all three countries. Whereas in Scotland and France around 10 percent of the population died in famines in the 1690s, no-one seems to have died from famine in England. So, with a greatly reduced labour force in agriculture (47 percent in 1701), England and Wales could comfortably feed the whole population.    

In the early 18th century, as the share of the labour force decreased further, England and Wales became a net exporter of grain. There is a major historiographical controversy about whether an agricultural revolution took place between 1550 and 1700 or in the 18th century. The new data suggests revolutionary change across both periods. 

Comparisons with other countries

Thanks to another Campop project, the International Network for the Comparative History of Occupational Structure (INCHOS), directed by Leigh Shaw-Taylor in Cambridge and by Osamu Saito in Tokyo, we can put these developments in a comparative perspective. Figure 3 (below) shows the labour force shares in the primary sector (excluding mining, so this is overwhelmingly agriculture) for a number of European countries together with the US, India and China.  

Figure 3. International comparisons of primary sector labour force shares.

It is clear from figure 3 that the decrease in the primary sector share in England and Wales was extraordinarily precocious compared with that anywhere else except the Netherlands. Germany did not reach the level of England and Wales in 1650 until 1850, and France not until around 1870. Spain and Italy did not reach this level until the 20th century. In other words, the shift out of agriculture in England began 200-300 years earlier than elsewhere in Europe. 

Our data also allow us to look at the geography of agricultural employment. Figure 4 (below) shows the share of adult males in the primary sector excepting mining (so, very largely agriculture) in 1813-20. As is immediately apparent, this was very varied, with highs above 70 percent and lows below 5 percent. At this date the national average was 42 percent. Yet there were large swathes of the country where the figure was as high as 50 percent, 60 percent or even 70 percent. The low figures, 20 percent and below, were on the coal-fields where industry was located, or in London.  

Figure 4. The share of the male labour force in the primary sector (excluding manufacturing).

To summarise

In 1381, 78 percent of adult males worked in agriculture. By 1601 this had already fallen to 64 percent. By 1701 the figure had decreased to 43 percent, and by 1911 it was only 11 percent.

In an international perspective these changes were extremely precocious. France and Germany did not reach the levels England and Wales had in 1650 until the second half of the 19th century, and Spain and Italy did not reach these levels before the early 20th century.

Further reading 

• Shaw-Taylor, L., and Wrigley, E. A., ‘Occupational Structure and Population Change’, in R. Floud, J. Humphries, and P. Johnson (eds.), The Cambridge Economic History of Modern Britain, Volume 1: Industrialisation, 1700–1870 (Cambridge, 2014), pp. 53–88. Link to chapter  
• Keibek, S., The male occupational structure of England and Wales, 1600-1850, Cambridge University Ph.D. thesis (2017). https://doi.org/10.17863/CAM.8960 
• Shaw-Taylor, L., and Saito, O., ‘The Occupational Structure of Britain c.1379–1911 and the International Comparative History of Occupational Structure: An overview of findings and where to find them’. Available here.