From Clio Infrastructure
General background
Why are data on occupational structure and population so important?
Data on occupational structure consist of counts of the numbers of people in particular occupational categories within a given area while population data consist simply of counts of the total population (though sex and age-breakdowns are sometimes available). No other form of historical economic data provides a comparable level of quantitative detail on the structure of economies over time nor such a comprehensive spatial and sectoral picture. Estimates of historical national (and per capita) income levels (the subject of another hub in the Clio-Infra consortium) are of self-evident importance but are critically dependent on the existence of occupational data. Moreover, occupational data are often much more robust than estimates of historical national income and (unlike historical income data) can easily be disaggregated within nations and aggregated or compared across national (currency) boundaries.
Such datasets can be used for many important research purposes in economic and social history. One example will have to suffice. Modern economic growth began in Europe. Yet the origins of modern economic growth remain poorly understood both during the industrial transition and in the centuries leading up to it. Long run data (over several centuries) on the occupational structure of different parts of Europe would allow economic historians to chart the economic development of Europe over time in a way which could fundamentally reshape our understanding of the process of economic development in Europe and hence the emergence of the modern world.
What can be done?
It is important to make a distinction between the modern ‘census’ period and the ‘pre-census’ period. In the second half of the nineteenth century almost all European states began to take periodic population and occupational censuses and to publish the data in tabular form both at the level of the nation-state and usually at more disaggregated spatial levels as well. There is a vast body of such census data available in print running down to the present day. However, most of it is not machine-readable and therefore cannot be analysed. What is machine-readable is in scattered locations. Furthermore occupational data are too complex to analyse without a degree of ‘coding’ into some limited number of occupational sectors. A long-run aim of the longer-term project (10-15 years) would be to make all the available European occupational and census data fully machine-readable at both nation state level and at a spatial level approximating to the British county. Such a project might require around 10-15m euros and can probably only be effected by multiple research teams based in different countries using, for the most part, national sources of funding. However, unless such work is co-ordinated at the European level and coded to a common occupational coding scheme the datasets from different countries will not be well suited to comparative analysis. It is therefore important to create an integrating infrastructure to ensure that common standards and coding schemes are in place. In the pre-census period (the exact dates differ from one country to another) it is much more difficult to acquire appropriate data and radically more expensive. Such research requires very extensive research work in numerous archives.
What has already been done?
Far and away the most important and sustained work on both population and occupational structure for the pre-census period to date has been undertaken by the Cambridge Group for the History of Population and Social Structure (a research group founded by Tony Wrigley and Peter Laslett in 1964). The occupational structure of Britain 1379-1911 project is run by Leigh Shaw-Taylor and Tony Wrigley and has received £1.3m worth of funding since 2003. Further funding applications (for over £1m) are in train to extend the datasets back to the fourteenth century and forward to 2001.This project has already begun to transform our understanding of the nature and timing of the first industrial revolution. To give two examples: the project has revealed that most of the growth in secondary sector (craft, construction and manufacturing) employment took place before the classic industrial revolution period began in 1750 and that over the next 100 years the most rapid growth of employment was not in industry but in the service sector.
Unfortunately no remotely comparable datasets currently exist for any other economy in the world. However, virtually all European countries have census occupational data available from the late nineteenth century. Moreover, some, perhaps many, other European countries have much better source availability for the pre-census period than Britain.
In Belgium, Professor Erik Buyst is applying for funding to emulate the Cambridge project, initially for the period 1846 to 1947 using the same methodology. There are several research groups in both Sweden and the Netherlands. Other projects are under consideration for Bulgaria and possibly other former Ottoman territories (Dr Martin Ivanov, Sofia), Germany (Dr Paul Warde, University of East Anglia), Spain (Dr Natalia Mora-Sitja, Cambridge) an Sweden (Professor Maria Agen, Upsalla).
Across Europe much census data is already machine-readable but it is currently in scattered locations, not commonly coded and not catalogued and is therefore currently of limited utility.
