Skip to content
Back to search results

Occupational Structure Reconstruction

Introduced by
E. A. Wrigley and the Cambridge Group for the History of Population and Social Structure
Year
2010
Type
descriptive-reconstruction
Family
economic-structure-measurement
Font size

What does it do?

Occupational structure reconstruction uses the distribution of the workforce across economic activities to chart how an economy's sectoral composition evolved before modern statistics.

The Cambridge Group, led by E. A. Wrigley, developed the Primary-Secondary-Tertiary (PST) system, a classification purpose-built for historical occupational descriptors that sorts every recorded occupation into agriculture and extraction (primary), manufacturing and processing (secondary), or trade, transport, and services (tertiary), with finer subdivisions beneath. By coding occupations from censuses, parish registers, probate inventories, trade directories, and militia lists, the method estimates the share of workers in each sector and tracks its movement across benchmark years. Because the shift from primary to secondary and tertiary employment is the structural signature of economic development, these reconstructions illuminate the timing and geography of industrialization with a directness that aggregate output figures cannot match, complementing and constraining national-accounting estimates of growth.

The intuition behind the method

The clearest fingerprint of economic development is what people do for a living. In a poor agrarian economy almost everyone farms; as productivity rises, labour is released into making things and then into services. So if you can count, for a given place and decade, how many were farmers, how many were artisans and manufacturers, and how many were traders, carriers, and servants, you can read the stage and pace of development directly off the workforce. The PST system is the ruler for this measurement, designed so that the strange occupational vocabulary of old records, 'husbandman', 'fustian weaver', 'higgler', maps cleanly onto sectors. The hard part is that records under-count women, conceal the farmer who also wove in winter, and describe occupations inconsistently. Reconstruction therefore combines careful coding with corrections for these biases, producing sector shares that, tracked over benchmark years, reveal an economy quietly transforming long before any statistician thought to measure it.

History

The systematic study of historical occupational structure was transformed by E. A. Wrigley and the Cambridge Group, who from the 1990s built the PST system to exploit England's exceptional occupational records. The flagship 'Occupational Structure of Britain' project, with Leigh Shaw-Taylor and others, coded millions of occupational observations from parish registers, censuses, and probate to reconstruct the sectoral structure of the English and Welsh economy from the seventeenth century onward. Their findings, notably an early and substantial secondary sector well before the canonical Industrial Revolution, reshaped debates on the timing of industrialization. The PST framework has since been extended to other countries, and the project's data underpin much subsequent regional economic history.

When to use it

Use occupational structure reconstruction when you want to measure the sectoral composition of a pre-statistical economy and its change over time, particularly to study the timing, pace, and geography of industrialization or the release of labour from agriculture. It is ideal where occupational descriptors survive in abundance, as in early modern England, and where the question concerns economic structure rather than social class. Prefer it to output-based methods when labour allocation is the object of interest or when output sources are too thin. It is less suitable where occupational recording is sparse or where you need monetary aggregates rather than employment shares, which require national accounting instead.

Limitations and considerations

  • Severely hampered by under-recording of women's and secondary occupations
  • By-employment makes a single occupational label misleading
  • Source coverage varies over time, confounding real and measured change
  • Yields employment shares, not the monetary output aggregates economists often need

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Reads the stage and pace of development directly from labour allocation
  • PST scheme is purpose-built for archaic and compound historical occupations
  • Exploits abundant parish-register and probate occupational data
  • Provides an independent structural check on output-based growth estimates

Applications

  • Reconstructing the sectoral workforce of England and Wales since 1600
  • Dating the rise of the secondary sector before the Industrial Revolution
  • Mapping regional patterns of industrialization and de-industrialization
  • Cross-checking national-accounting growth estimates against labour shares
  • Comparing the occupational structures of different early modern economies

How it works

How it works

1

Reconstruction draws on every source that records occupations: early censuses, parish registers (which note fathers' occupations at baptism and marriage), probate inventories, trade directories, and militia or tax lists. Each source has distinct coverage and biases, censuses sample the whole population but late, parish registers run earlier but skew toward married men. The analyst combines sources to extend coverage backward and assess consistency, treating their overlaps as cross-checks on the emerging picture.

Formula
R = \{\text{census, parish, probate, directory, militia}\}
2

Every occupational descriptor is coded into the PST system, which assigns a sector (primary, secondary, tertiary), a subsector, and a specific activity. The scheme was designed expressly for historical English occupations and handles archaic and compound titles with explicit rules. Unlike HISCO, which targets cross-national occupational comparison, PST is optimised for resolving the sectoral location of work, making it the preferred tool when the analytical goal is economic structure rather than social class.

Formula
o_i \mapsto \text{PST code}\ (\text{sector}, \text{subsector}, \text{activity})
3

Raw counts are adjusted for the well-known biases of historical sources. Women's and children's work is systematically under-recorded; by-employments, the farmer who also weaves, hide secondary activity; and some sources over-represent particular groups. The analyst applies weights or correction factors, informed by sources that do capture female and secondary work, to move from recorded occupations toward a fuller picture of actual labour input across sectors.

Formula
w_i\ \text{(correct for under-recording, by-employment, gender)}
4

Weighted counts are converted into sector shares, the proportion of the workforce in primary, secondary, and tertiary activity. These shares are the headline output, summarising the economy's structure at a point in time. Care is taken to define the denominator consistently, whether all adults, the male labour force, or full-time equivalents, since the choice strongly affects comparability and is a frequent source of disagreement between competing reconstructions.

Formula
\pi_s = \frac{\sum_i w_i \mathbb{1}[s_i = s]}{\sum_i w_i},\ s\in\{P,S,T\}
5

Finally sector shares from successive benchmark years are linked into a time series showing structural change. This requires reconciling differences in source coverage and coding across periods so that movement reflects real change rather than shifting measurement. The resulting trajectory of declining primary and rising secondary and tertiary shares is then compared with, and used to discipline, output-based growth estimates from national accounting and back-projection.

Formula
\{\pi_{s,t}\}_{t}\ \text{linked across benchmark years}

Common pitfalls

  • Ignoring female and child labour and so overstating the primary sector's decline
  • Treating a single recorded occupation as full-time when by-employment was common
  • Mixing denominators (all adults vs. males) across benchmark years
  • Attributing shifts to real change when they reflect changing source coverage

Frequently asked questions

How does PST differ from HISCO?

PST is optimised for placing occupations into economic sectors, primary, secondary, tertiary, to study economic structure, whereas HISCO is designed for cross-national occupational comparison and feeds social-class schemes. PST handles archaic English occupational vocabulary in fine sectoral detail. The two answer different questions: PST asks which part of the economy a worker belonged to, HISCO asks what occupation it was in an internationally comparable index.

Why is under-recording of women such a problem?

Historical sources recorded men's occupations far more consistently than women's, and much female labour, in agriculture, textiles, and services, was casual, seasonal, or unpaid and so left little trace. Because women's work concentrated in particular sectors, ignoring it biases the estimated sector shares and can exaggerate apparent structural change. Reconstructions therefore apply explicit corrections drawn from the rarer sources that do capture female employment.

Can occupational structure measure economic growth?

Indirectly. Occupational structure measures how labour is allocated across sectors, not output, so it cannot by itself give GDP. But the movement of labour out of agriculture is a strong structural indicator of rising productivity and development, and sector shares are used to cross-check and constrain output-based growth estimates. The two approaches are complementary, with structure illuminating the composition behind aggregate growth.

Related methods

Sources

  1. Wrigley, E. A. (2010). The PST System of Classifying Occupations. Cambridge Group for the History of Population and Social Structure.

    Open publication reference
    Copied.
  2. Maddison, A. (2007). Contours of the World Economy 1-2030 AD: Essays in Macro-Economic History. Oxford University Press.

    ISBN: 9780199227204
    Copied.