Censuses and the work women really did: case studies 1720-1920

by Amy Erickson (University of Cambridge)

This research will be presented during the EHS Annual Conference in Belfast, April 5th – 7th 2019. Conference registration can be found on the EHS website.


Women workers Woolwich Arsenal 1917 London. Available at Wikimedia Commons.

Museums and popular histories typically repeat the idea that women only entered the labour market in large numbers in the twentieth century. In fact, in the first British census that can be analysed for paid employment in 1851, more than 40% of all women reported regular paid employment. They contributed nearly one third of all hours in the paid economy. Participation rates were probably even higher before mechanisation.

The occupational structure of men in the past has recently been explored as a key to understanding economic development. But national data on women’s work are not available until the advent of censuses in the nineteenth century.

A set of studies to be presented at the Economic History Society’s 2019 annual conference makes uses of censuses and alternative sources – focusing on eighteenth century Europe, nineteenth century textiles, the 1881 British census and twentieth century Canada – to demonstrate the formative role of women’s labour contribution to economic development.

Professor Carmen Sarasua (Autonomous University of Barcelona) points out that the most important feature of the European economy in this period was the rapid spread of manufacturing, organised on a domestic basis in the eighteenth century and in mechanised factories in the nineteenth century. Domestic industry did not always evolve into industrialisation in the same places, but long before the advent of factories, manufacturing relied on the labour of women and children.

Most analyses of occupational change as an indicator of economic development consider only male occupations, which show heavily agricultural societies, but when women’s occupations from tax registers are included, manufacturing is roughly equivalent to agriculture by the later eighteenth century. This means that manufacturing was important long before we currently think it was, and long before the application of new technology.

For centuries, textiles were the largest manufacturing industry in most European countries, and women dominated that labour force. Professor Manuela Martini (University of Lyon) focuses on Lyon, which was the most important silk-producing city in Europe in the nineteenth century.

She compares population censuses, which often omitted details of women’s occupations, with trade union and administrative sources on silk workers’ wages, to understand the occupational distinctions at the level of tasks performed by men and women, establishing a vocabulary with which to compare textile trades internationally.

Dr Xuesheng You (Cambridge University) presents the first detailed evidence from British censuses showing there were wide geographical variations in female labour force participation rates and in the sectoral distribution of female employment between agriculture, manufacturing and services. Factors such as age, marital status and number of children were relatively insignificant compared with the demands of the local economy. In other words, if work was available, women took it, regardless of their household situation.

Dr Keith Sugden (Cambridge University) and Professor Roger Sugden (University of British Columbia) find that in a fast-growing agricultural area of Western Canada, early twentieth century censuses recorded almost no married women’s employment, while for men and single women, they provided the occupation, wage, and number of weeks worked in the previous year.

But in an area dominated by small family farms many if not all married women would have been employed in market-oriented production on the farm even if they did not receive a wage. Sugden and Sugden propose ways to quantify the value of married women’s ‘hidden’ economic contribution.

So censuses may or may not record women’s work consistently. But other sources show high levels of labour force participation and demand-led employment, placing women’s labour at the centre of the growth of manufacturing and services that characterises economic development.

People, Places and Business Cultures: Essays in Honour of Francesca Carnevali

review by Jim Tomlinson (University of Glasgow)

book edited by Paolo Di Martino, Andrew Popp and Peter Scott.

People, Places and Business Cultures: Essays in Honour of Francesca Carnevali’ is published by Boydell and Brewer. SAVE  25% when you order direct from the publisher – offer ends on the 2nd April 2019. See below for details.




Festschriften are usually produced at or around retirement, and to celebrate long academic careers. This collection, tragically, marks the end of a foreshortened career, that of Francesca Carnevali, who died in 2013 at the age of 48.

The chapters of the book have all been written by historian colleagues and friends of Francesca. The authors come from a diverse set of academic backgrounds, including the prominent medievalist Chris Wickham and the social and cultural historian Matthew Hilton. But most of the contributors come, as did Francesca, from the broadly-defined subject of business history.

Francesca’s own work provided a broad and variegated set of concerns and approaches that enables the contributors to link her work to their own diverse areas of expertise. Thus, for example, Leslie Hannah (who supervised Francesca’s PhD, and co-authored an article on banking with her), provides a new approach to the old question of the comparative performance of British banking before 1914. He stresses the paradox (at least for those who think competition is always the key to efficiency), that by any standards Britain at that time had a highly competitive banking system, yet suffered a growth ‘climacteric’. More broadly, Hannah, like Francesca herself, adheres to a broadly declinist view of British economic history, whilst clearly identifying the unsatisfactory nature of many declinist stories.

Francesca’s own work on banking contrasted Italy and Britain, and the financing of Italian small business is the concern of Alberto Rinaldi and Anna Spadavecchia’s chapter. The conclusion of this analysis emphasizes the embeddedness of financial institutions in legal, social and political conditions as well as economic circumstances, a conclusion that links to Francesca’s broadening concerns after her early work on banking. Key to this broadening was an examination of social capital and trust, as key, if problematic, concepts for understanding business behaviour.

This behaviour is examined in a variety of contexts in this book, ranging from Andrew Popp’s study of Liverpool cotton brokers and their ‘public staging of business life’ to Lucy Newton’s study(jointly authored with Francesca) of making and selling pianos in Victorian and Edwardian England. This concern with consumer goods is linked by Peter Scott and James Walker to an innovative study of how mass consumption and mass marketing, to some degree at least, blurred class demarcations on interwar Britain.

These empirical studies are complemented by more conceptually focussed chapter, by Chris Wickham on the genealogy of ‘micro-history’, by Kenneth Lipartito on the concept of social capital and its limits, and by Andrea Colli on the problems of doing comparative European history.  Last, but very far from least, there is a characteristically wide-ranging and insightful chapter by Mathew Hilton on the problems of writing the economic and social history of twentieth-century Britain in the light of the recent ‘turns’ in how that history is being written.

The diversity of this book’s contents is a strength not a weakness. Business historians of almost any bent will find something interesting and important to engage with. The breadth of analytical and empirical concerns, allied with the close attention to important conceptual puzzles, makes this book a fitting reflection of, and tribute to, Francesca’s productive and well-lived life.


SAVE 25% when you order direct from the publisher using the offer code BB500 online hereOffer ends 2nd April 2019. Discount applies to print and eBook editions. Alternatively call Boydell’s distributor, Wiley, on 01243 843 291, and quote the same code. Any queries please email marketing@boydell.co.uk


To contact Jim Tomlinson: jim.tomlinson@glasgow.ac.uk


Note: this review was originally published on-line in Business History, 2019.  It is reproduced by kind permission of  Lee-Ann Anderson (Permissions and Licensing, Taylor and Francis).

‘Money talks – give yours an Empire accent’: the economic failure of Britain’s Empire Marketing Board

by David M. Higgins (Newcastle University) and Brian Varian (Swansea University)

This research will be presented during the EHS Annual Conference in Belfast, April 5th – 7th 2019. Conference registration can be found on the EHS website.


The formation of the Empire Marketing Board (EMB) in 1926 was a unique experiment in interwar Britain: it was the first, publicly funded marketing board in the UK that sought to encourage domestic consumption of empire foodstuffs and raw materials. Using a diverse range of marketing methods – including films, cinema broadcasts, newspaper advertisements and especially posters – the Board aimed to increase public awareness of the strong economic interdependence between Britain and its Empire.

This relationship was longstanding. Britain was unquestionably the ‘vent for surplus’ for many empire products such as New Zealand butter, cheese and lamb, Indian tea and Australian frozen beef. Yet by the 1920s, the Dominions and other primary-producing countries were increasingly competing in the British market, necessitating a reassessment of Britain’s relationship with the Dominions. Importantly, it was recognised that the purchase of empire produce provided the means for the Dominions to increase their consumption of British manufactures.

Unlike most countries, Britain pursued an essentially free-trade policy in the 1920s. There was simply no scope for Britain to favour empire produce through the extension of preferential tariff rates, a policy known as ‘imperial preference’. Yet the Dominions had applied preferential tariff rates to their imports of manufactures from Britain.

Consequently, Britain attempted to correct this imbalance through the creation of the EMB, which would favour Dominion produce not through tariffs, but rather through publicity. This approach was a stop-gap that was superseded by the Ottawa agreements of 1932, when Britain’s abandonment of free trade finally allowed the country to implement preferential tariff rates.

To date, much of the research on the EMB has claimed that it served an important function by fostering imperial ideology within the empire. But while the cultural impact of the EMB has been studied, its economic impact has not. Our study, to be presented at the Economic History Society’s 2019 annual conference, is the first to evaluate the economic impact of the EMB.

We begin by describing the scale of the problem confronting the Board: by 1924, for example, almost 80% of Britain’s beef imports originated from ‘foreign’ countries, principally Latin America while over 55% of butter imports originated from outside the Empire. In fact, cheese was the only major food product in which the empire dominated Britain’s imports.

In our study, we focus on a sample of the iconic posters that were issued by the EMB and displayed in cities throughout Britain. Using a series of regression analyses, we test whether these posters – the EMB’s most prominent form of advertising – raised the empire’s share of Britain’s imports of key foodstuffs, including butter, cheese, wheat, tea, rice, sugar and beef. Our econometric results indicate that the EMB did not exert a statistically significant effect on the empire’s share of these imports. In economic terms, Britain’s short-lived EMB was a failure.


Higgins and Varian
Figure 1: Stylised grocer’s window exhibiting domestic and empire produce caption


From one perspective, our results suggest that appeals to patriotism and imperial ‘self-help’ were fundamentally misguided. Empire suppliers had to compete with well-established foreign producers whose products were held in high-esteem by UK consumers: chilled Argentine beef was a far superior product to the frozen product from the Antipodes; while Danish butter predominated in much of northern England.

There were other key problems that the EMB ignored. Possibly the greatest shortcoming of the EMB was its failure to differentiate sufficiently the products of individual dominions from those of the empire.

In this regard, the Merchandise Marks Act 1926 was unhelpful: vendors were at liberty to sell butter and beef either with a definite indication of origin or the term ‘empire’;  for other produce, such as cheese, retailers were not required to indicate origin. While the EMB was advertising ‘Empire’ to little effect, the marketing campaigns of Dominion control boards, such as the New Zealand Dairy Produce Control Board, were winning the British consumer through more differentiated advertising.

Moreover, we argue that the EMB was underfunded. From 1928 to 1931, the EMB’s publicity expenditure averaged 0.07% of the value of Britain’s imports from the empire. In contrast, the New Zealand Dairy and Produce Control Board’s expenditure on its well-defined marketing campaign was 0.13% of the value of Britain’s imports of dairy products from New Zealand.

Our study of the EMB has contemporary relevance to debates on ‘trade blocs’. The EMB represented an attempt to forge a trade bloc through the non-conventional approach of advertising – extending preference within the constraint of free trade. While this innovative experiment of the interwar era failed (economically), it is nevertheless indicative of Britain’s desire to reorient its trade toward the Empire. In this respect, the EMB was a precursor to the ultimately ‘successful’ formation of trade blocs and, indeed, disintegration of the world economy in the 1930s.


The age of mass migration in Latin America

by Blanca Sánchez-Alonso (Universidad San Pablo-CEU, Madrid)

This article is published by The Economic History Review, and it is available on the EHS website.


Blanca Blog
General Carneiro station which belonged to Minas and Rio railway. Minas Gerais province, Brazil, c.1884. Available at Wikimedia Commons.

Latin America was considered a ‘land of opportunity’ between 1870 and 1930.  During that period 13 million Europeans migrated to this region.  However, the experiences of Latin American countries are not fully incorporated into current debates concerning the age of mass migration.

The main objective of my article, ‘The age of mass migration in Latin America’,  is to rethink the role of European migration to the region in the light of new research. It addresses several major questions suggested by the economic literature on migration: whether immigrants were positively selected from their sending countries, how immigrants assimilated into host economies, the role of immigration policies, and the long-run effects of European immigration on Latin America.

Immigrants overwhelmingly originated from the economically backward areas of southern Europe. Traditional interpretations have tended to extrapolate the economic backwardness of Italy, Spain, and Portugal (measured in terms of per capita GDP and relative to advanced European countries) to emigration flows. Yet, judging by literacy levels, migrants to Latin America from southern European countries were positively selected. Immigrants to Latin America from Spain, Italy and Portugal were drawn from the northern regions which had higher levels of literacy.  There were very few immigrants to Latin America  from the southern regions of these countries. .  When immigrant literacy is compared with that of potential emigrants from regions of high emigration, positive selection appears quite clear.

One proxy often used to signal positive self-selection is upward mobility within and across generations. Recent empirical research shows that it was the possibility of rapid social upgrading that made Argentina attractive to immigrants. First-generation immigrants experienced faster occupational upgrading than natives; upward occupational mobility occurred for a large proportion of those who declared unskilled occupations on arrival. Immigrants to Argentina experienced a very fast growth in occupational earnings (6 per cent faster than natives) between 1869 and 1895. For the city of Buenos Aires in 1895, new evidence shows that Italian and Spanish males received, on average, 80 per cent of average native-born earnings. In some categories, such as crafts and services, immigrants obtained higher wages than natives. These findings provide an economic rationale why some Europeans chose Argentina over the US, despite a smaller wage differential between originating country  and destination.

Immigrants appear to have adjusted successfully to Latin American labour markets.  This is evidenced by access to property and in the large ownership of businesses.  Almost all European communities experienced strong and fast upward social mobility in the destination countries. Whether this was because of positive selection at home or because of the relatively low skill levels in the host societies is still an open question.

European immigrants to Latin America had higher levels of literacy than the native population. Despite non-selective immigration policies, Latin American countries received immigrants with higher levels of human capital compared to natives. Linking immigrants’ human capital to long run economic and educational outcomes has been the focus of recent research for Brazil and Argentina. The impact of immigration in those areas with higher shares of Europeans appears to be important since immigrants demanded and created schools (public or private). New research presents evidence of path dependency linking past immigrants’ human capital with present outcomes in economic development in the region.

Immigration policies in Latin America raised few barriers to European immigration. However, the political economy of immigration policy of Argentina shows a more complicated story than the classic representation of landowners constantly supporting an open-door policy.

Brazil developed a long-lasting programme of subsidized immigration. The expected income of immigrants to São Paulo was augmented by prospective savings, a guaranteed job on arrival, and the subsidized transportation cost. Going to Brazil was perceived as a good investment in southern Europe. Transport subsidies and the peculiarities of the colono contract in the coffee areas seem more important explanations than real wage differentials for understanding how Brazil competed for workers in the international labour market. The Lewis model merits further investigation for two main reasons. First, labour supply increased faster than the number of workers needed for the coffee expansion because of subsidies and, second, labour markets in São Paulo were segmented. European immigrants supplied only a fraction (though a substantial one) of the total labour force needed for the coffee plantations. The internal supply of workers became increasingly important and must be included in the total labour supply.

Recent literature shows that researchers are either identifying new quantitative evidence or exploiting existing data in new ways. Consequently,  new research is providing answers and posing questions to show that Latin America has much to add to debates on the economic and social impact of historical immigration.


To contact Blanca Sánchez-Alonso: blanca@ceu.es

The gender division of labour in early modern England: why study women’s work?

by Jane Whittle (University of Exeter) and Mark Hailwood (University of Bristol)

This article is published by The Economic History Review, and it is available on the EHS website.


Interior with an Old Woman at the Spinning Wheel. Available at Wikimedia Commons.

Here are ten reasons to know more about women’s work and read our article on ‘The gender division of labour in early modern England’. We have collected evidence about work tasks in order to quantify the differences between women’s and men’s work in the period from 1500-1700. This research allows us to dispel some common misconceptions.


  1. Men did most of the work didn’t they? This is unlikely, when both paid and unpaid work are counted, modern time-use studies show that women do the majority of work – 55% of rural areas of developing countries and 51% in modern industrial countries (UN Human Development report 1995). There is no reason why the pattern would have been markedly different in preindustrial England.
  2. But we know about occupational structure in the past don’t we? Documents from the medieval period onwards describe men by their occupations, but women by their marital status. As a result we know quite a lot about male occupations but very little about women’s.
  3. But women worked in households headed by their father, husband or employer. Surely, if we know what these men did, then we know what women were doing too? Recent research undertaken by Amy Erickson, Alex Shepard and Jane Whittle shows that married women often had different occupations from their husbands. If we do not know what women did, we are missing an important part of the economy.
  4. But we have evidence of women working for wages. It shows that around 20% of agricultural workers were women, surely this demonstrates that women’s work wasn’t as important as men’s in the wider economy? This evidence only relates to labourers paid by the day, and before 1700 most agricultural labour was not carried out by day labourers, so this isn’t a very good measure. Our article shows that women carried out a third of agricultural work tasks, not 20%.
  5. But women mostly did domestic stuff – cooking, housework and childcare – didn’t they, and that type of work doesn’t change much across history? Women did do most cooking, housework and childcare, but our research suggests it did not take up the majority of their working time. These forms of work did change markedly over time. A third of early modern housework took place outside, and our data suggests the majority was done for other households, not as unpaid work for one’s own family.
  6. But women only worked in a narrow range occupations, didn’t they? Our research shows that women worked in all the major sectors of the economy, but often doing slightly different tasks from men. They undertook a third of work tasks in agriculture, around half of the work in everyday commerce and almost two thirds of work tasks in textile production. But women also did forms of work we might not expect, such as shearing sheep, dealing in second-hand iron, and droving cattle.
  7. Women’s work was all low skilled wasn’t it? Women very rarely benefitted from formal apprenticeship in the way that men did, but that does not mean the tasks they undertook were unskilled. Women undertook many tasks, such as making lace and providing medical care, which required a great deal of skill.
  8. But this was all in the past, what relevance does it have now? Many gendered patterns of work are remarkably persistent over time. Analysis by the Office of National Statistics states that one third of the gender pay gap in modern Britain can be explained by men and women working in different occupations, and by the lower rates of pay for part-time work, which is more commonly undertaken by women than men.
  9. So nothing ever changes …? Well, not necessarily. In fact looking carefully at patterns of women’s work in the past shows some noticeably shifts over time. For instance, women worked as tailors and weavers in the medieval period and in the eighteenth century, but not in the sixteenth century.
  10. But we know why women work differently from men, particularly in preindustrial societies – isn’t it because they are less physically strong and all the child-bearing stuff? Physical strength does not explain why women did some physically taxing forms of work and not others (why they walked for miles carrying heavy loads on their heads rather than driving carts). And not all women were married or had children. Neither physical strength nor child-bearing can explain why women were excluded from tailoring between 1500 and 1650, but worked successfully and skilfully in this and other closely related crafts in other periods.

We now have data which allows us to look more carefully at these issues, but there is still much more to uncover.


To contact Jane Whittle: j.c.whittle@ex.ac.uk, Twitter: @jcwhittle1

To contact Mark Hailwood: m.hailwood@bristol.ac.uk, Twitter: @mark_hailwood

‘Stop-go’ policy and the restriction of post-war British house-building

by Peter Scott (Henley Business School, University of Reading) and James T. Walker (Henley Business School, University of Reading)

This article is published by The Economic History Review, and it is available on the EHS website.


British house-building
A member of the Pioneer Corps assists a civilian building labourer in tiling a roof. Available at Wikimedia Commons.

Britain’s unusually high house price to income ratio plays an important role in reducing living standards and increasing “housing poverty”. This article shows that Britain’s housing shortage partly stems from deliberate long-term government policies aimed at restricting both public and private sector house-building. From the 1950s to the early 1980s, successive governments reduced housing starts as part of `stop-go’ macroeconomic policy, with major cumulative impacts.

This policy had its roots in the Second World War, when an influential coalition of Bank of England and Treasury officials pressed for a post-war policy of savage deflation, to restore sterling’s credibility and re-establish London as a major financial centre. John Maynard Keynes warned that prioritising international ‘obligations’ over the war-time commitment to build a fairer society would be repeating the 1920s gold standard error – though his direct influence ended with his untimely death. Deflationary policy proved politically impracticable in the short-term, as evidenced by Labour’s 1945 landslide election victory, though its supporters bided their time and were able to implement much of their agenda in the changed political climate of the 1950s.

The Conservatives’ 1951 election victory was based on a pledge to build 300,000 new homes per year. This was achieved in 1953 and building peaked at 340,000 completions in 1954. However, officials took advantage of the 1955-57 credit squeeze to press for severe cuts in housing investment. Municipal house-building was cut, while private house-building was depressed largely through restricting the growth of building society funds (by pressurising the building societies’ cartel to keep interest rates at such low levels that they were starved of mortgage funds). While the severity of policy varied over time, these restrictions were maintained almost continually until the early 1980s.

These restrictions were never formally announced and were hidden from Cabinet for much of this period. Meanwhile, given the political importance of housing, the Conservative government simultaneously proposed ever-larger housing targets (culminating in a 1964 election pledge to build 400,000 per annum). This created a perverse situation, whereby the government was spending substantial sums on highly publicised policies to increase demand for private housing (such as the 1959 House Purchase and Housing Act and the 1963 abolition of Schedule A income tax), while covertly reducing housing supply through restricting mortgage funding, limiting building firms’ access to credit, and reducing municipal housing investment. The following Labour government found itself drawn in to a similarly restrictive housing policy, as part of its ill-fated commitment to avoid sterling devaluation (arguably based on misleading Treasury advice), while housing restrictions were also used as an instrument of macroeconomic stabilisation in the 1970s.

A 1974 Bank of England analysis found that this policy had created both an exaggerated housing cycle and a structural deficit (with house-building being held below market-clearing levels at all points in the cycle). This had in turn reduced the capacity of the housing market to respond to rising demand, by reducing builders’ land banks, building materials capacity, and building labour, which raised house-prices while lowering productivity and technical progress. There is also evidence of “learning effects” by house-builders, who avoided expanding their activities during cyclical upturns, as they correctly perceived that tighter government restrictions might be imposed before their houses were ready to sell. These pressures fuelled house price inflation, both directly, and because housing became increasingly regarded as as a hedge against inflation.


Figure 1: Capital formation in dwellings, as percentage of total capital formation, and housing completions per thousand families, private houses and all houses, 1924-38 and 1954-79

Housing Graph


British house-building during this era compared unfavourably to inter-war levels, as shown in Figure 1. Moreover, private house-building was even more depressed that total housing – as the Treasury found it easier to covertly restrict private housing than to reduce municipal building starts, where policy was more open to Cabinet and public scrutiny. British gross domestic fixed capital investment in housing was also very low relative to other European nations. Our time-series econometric analysis for 1955-1979 corroborates the `success’ of the restrictions and also shows the predicted asymmetric impact in `stop’ and `go’ phases of policy. This is an important finding – as stop-go policy is often examined in terms of the volatility of the variable under examination – based on the unrealistic assumption that industry would fail to realise that demand upturns might be rapidly terminated by the re-imposition of controls.

Housing restriction policy has persisting consequences. Additions to the housing stock were depressed for several decades, while the inflationary-hedge benefits for house-purchase became a self-fulfilling prophecy. Meanwhile restrictive planning policy (which was substantially intensified in the 1950s, as a further measure of housing restriction) has proved difficult to reverse. Average house-prices to income ratios have thus continued the upward trend established in this era, currently excluding a substantial and growing proportion of the population from owner-occupation.


To contact Peter Scott: p.m.scott@henley.ac.uk

To contact James T. Walker: j.t.walker@henley.ac.uk



Unions and American Income Inequality at Mid-Century

by William J. Collins (Vanderbilt University) and Gregory T. Niemesh (Miami University)

This article is published by The Economic History Review, and it is available on the EHS website.


EHS Great Depression
Crowd of depositors gather in the rain outside the Bank of United States after its failure. Available at Wikimedia Commons

Rising income inequality in the United States has attracted scholars’ attention for decades, resulting in an extensive and detailed literature on the trend’s causes and consequences.  An equally large but much less studied decline in income inequality occurred in the US during the 1940s.  This led to an era of relatively compressed income inequality that lasted into the 1970s. Goldin and Margo (1992) called this ‘The Great Compression.’

Our recent research has explored the role of changing labour market institutions in contributing to the Great Compression, with a focus on the role of labour unions.  In the US, labour unions rose to prominence starting in the late 1930s, following the Wagner Act of 1935 and a Supreme Court decision in 1937 upholding the Act.  This recast the legal framework under which unions formed and collectively bargained by creating the National Labor Relations Board to oversee representation elections and enforce the Act’s provisions, including prohibitions of various ‘unfair practices’ which employers had used to discourage unions.  Unions continued to grow through the 1940s, especially during the Second World War, and they peaked as a share of employment in the early 1950s.

Time series graphs of union density and income inequality over the full twentieth century in the US are nearly mirror images of each other (Figure 1).  But it is difficult to evaluate the role of unions in influencing this period’s inequality due to limitations of standard data sources.  The US census, for instance, has never inquired about union membership, which makes it impossible to link individual-level wages to individual-level union status in nationally representative samples for this period (see Callaway and Collins 2018 and Farber et al. 2017 for efforts to develop data from other sources).  Research on US unions later in twentieth century, when data are more plentiful, highlight their wage compressing character, as does some of the historical literature on wage setting during the Second World War, but there is much left to learn.


Figure 1: Unions and income inequality trends in the 20th-century United States

Great Depression Table

Sources: See Collins and Niemesh (forthcoming).


In a paper titled ‘Unions and the Great Compression of wage inequality in the United States at mid-century: evidence from labour markets,’ we provide a novel perspective on changes in inequality at the local level during the 1940s (Collins and Niemesh, forthcoming).  The building blocks for the empirical work are as follows: the “complete count” census microdata for 1940 provide information on wages and industry of employment (Ruggles et al. 2015); Troy’s (1957) work on mid-century unionization provides information on changes in unionization at the industry level over the 1940s; and subsequent censuses provide sufficient information to form comparable local-level measures of wage inequality.  We use a combination of local employment data circa 1940 and changes in unionization by industry after 1939 to create a variable for local ‘exposure’ to changes in unionization.

We ask whether places with more exposure to unionization due to their pre-existing industrial structure experienced more compression of wages during the 1940s and beyond, conditional on many other features of the local economy including wartime production contracts and allowing for differences in regional trends. The answer is yes: a one standard-deviation increase in the exposure to unionization variable is associated with a 0.072 log point decline in inequality between the 90th and 10th wage percentile in the 1940s (equivalent to 32 percent of the mean decline).  The association between local union exposure and wage compression is concentrated in the lower part of wage distribution.  That is, the change in inequality between the 50th and 10th percentile is more strongly associated with exposure to unionization than the change between 90th and 50th percentile.  As far as we can tell, this mid-century pattern was not driven by the re-sorting of workers (e.g., high skilled workers sorting out of unionizing locations) or by firms exiting places that were highly exposed to unionization.

We also explore whether the impression unions likely made on local wage structures persisted, even as private sector unions declined through the last decades of the twentieth century. In fact, the pattern fades a bit with time, but it remains visible to the end of the twentieth century. We leave for future research important questions about the mechanisms of persistence in local wage structures, non-wage aspects of unionization (e.g., implications for benefits or safety), implications for firm behaviour in the long run, and international comparisons.


To contact William J. Collins: william.collins@Vanderbilt.Edu

To contact Gregory T. Niemesh: niemesgt@miamioh.edu



Callaway, B. and W.J. Collins. ‘Unions, workers, and wages at the peak of the American labor movement.’ Explorations in Economic History 68 (2018), pp. 95-118.

Collins, W.J. and G.T. Niemesh. ‘Unions and the Great Compression of wage inequality in the US at mid-century: evidence from local labour markets.’ Economic History Review (forthcoming). https://doi.org/10.1111/ehr.12744

Farber, H.S., Herbst D., Kuziemko I., and Naidu, S. ‘Unions and inequality over the twentieth century: new evidence from survey data.” NBER Working Paper 24587 (Cambridge MA, 2018).

Goldin, C. and R.A. Margo, ‘The Great Compression: the wage structure in the United States at midcentury.’ Quarterly Journal of Economics 107 (1992), pp. 1-34.

Ruggles, S., K. Genadek, R. Goeken, J. Grover, and M. Sobek. Integrated public use microdataseries: version 6.0 [Machine-readable database]. (Minneapolis: University of Minnesota, 2015).

Troy, L., The distribution of union membership among the states, 1939 and 1953. (New York: National Bureau of Economic Research, 1957).

from Microeconomic Insights: ‘When Britain turned inward: lessons from the impact of 1930s protectionism’

by Alan de Bromhead (Queens University Belfast), Alan Fernihough (Queens University Belfast), Markus Lampe (Vienna University of Economics and Business) and Kevin Hjortshøj O’Rourke (All Souls College, Oxford)

The full post and online access to the full article is available on the Microeconomic Insights website


With a protectionist president in the White House, the future of the multilateral, rules-based international trading system seems much less certain. So it is not surprising that politicians and commentators are turning to the 1930s for examples of what protectionism can imply for international trade flows.

World trade not only collapsed during the early 1930s: it also became much less multilateral. Countries like Britain and France, which already had empires, traded more with those empires. And countries like Germany and Japan, which were looking to acquire empires of their own, similarly traded more intensively with their respective spheres of influence.

This research focuses on the experience of Britain, which in 1931 broke decisively with a longstanding tradition of free trade. From November that year, substantial tariffs could be imposed on manufactured goods from outside the Empire. Similar duties on non-Empire fruit, flowers, and vegetables were possible soon after. And following the Ottawa conference of 1932, Britain’s trade policy explicitly served the interests of ‘the home producer first, Empire producers second, and foreign producers last’.

What was the impact of this dramatic policy shift? This study analyzes detailed data on British imports of 258 consistently defined commodities from 42 countries over the period 1924-38, as well as information on tariffs, quotas, voluntary export restraints, and other variables potentially influencing trade flows. To quantify the impact of the switch to protection, the authors compare actual trade flows from 1931 with counterfactual flows that would have taken place had tariffs and quotas remained unchanged.

The shift towards protection reduced the value of British imports by 9-10% on average, with the biggest impact being felt in 1933. Protection accounted for about a quarter of the total decline in British imports, which is consistent with results for the United States.

But in contrast with the findings of previous studies (which analyze aggregate data on trade and trade policies), the new research finds that the shift towards protection had a big effect on the geographical composition of British imports. For example, the Empire’s share of British imports rose from 27% to 39.2% between 1930 and 1935, while in the absence of protection it would only have increased to 31.4%.

Overall, the research shows that using disaggregated data does not significantly change the estimated impact of protection on the total value of trade. But it matters a great deal for the estimated impact of protection on the geographical composition of trade. Studies using aggregate data find that imperial trade blocs did not have a big influence on trade patterns during the 1930s. In contrast, this research finds that trade policy was crucial in increasing the share of the British Empire in British imports.

The clear ‘Balkanization’ of world trade shown in these results had wider effects, as several contemporary observers recognized. It reflected and probably also exacerbated the international tensions of the times, the later outcomes of which are well known.

Shoplifting in Eighteenth-Century England

by Shelley Tickell (University of Hertfordshire)

Shoplifting in Eighteenth Century England is published by Boydell and Brewer Press. SAVE  25% when you order direct from the publisher – offer ends on the 5th March 2019. See below for details.



What would you choose to buy from a store if money was no object? This was a decision eighteenth-century shoplifters made in practice on a daily basis. We might assume them to be attracted to the novel range of silk and cotton textiles, foodstuffs, ornaments and silver toys that swelled the consumer market in this period. Demand for these home-manufactured and imported goods was instrumental in a trebling of the number of English shops in the first half of the century, escalating the scale of the crime. However, as my book Shoplifting in Eighteenth-Century England shows, this was not the case. Consumer desire was by no means shoplifters’ major imperative.


Shoplifting occurred nationwide, but it was disproportionately a problem in the capital. A study of a sample of the many thousand prosecutions at the Old Bailey reveals that linen drapers, shoemakers, hosiers and haberdashers were the retailers most at risk. Over 70% of goods stolen, particularly by women, were fabrics, clothing and trimmings. Though thefts were highly gendered, men also stole these items far more frequently than the food, jewellery and household goods which were largely their preserve. Yet items stolen were not predominantly the most fashionable. Traditional linens, wool stockings and leather shoes were stolen as often as silk handkerchiefs and cotton prints. A prolific shoplifter who confessed to her crime found it profitable over the course of a year to steal printed linen at four times the quantity of the more stylish cotton, lawns, muslins and silk handkerchiefs she also took.

The shoplifters prosecuted were overwhelmingly from plebeian backgrounds. Professional gangs did exist but for most the crime was a source of occasional subsistence. Shop thieves came from the most economically vulnerable sections of society, seeking to weather an urban economy of low-paid and insecure work; many were older women or children. As the stolen goods needed to be convertible to income they were very commonly sold. So thieves sought the items which were most negotiable, those in greatest demand and least conspicuous in the working neighbourhoods in which they lived. A parcel of handkerchiefs stolen unopened was found to be ‘too fine’ for a market seller to whom it was offered. While there was undoubtedly an eagerness for popular fashion, the call for neat and appropriate daily dress in working communities was as insistent. We find the frequency with which shoplifters stole different types of clothing is consistent with a market demand governed in great part by the customary turnover of clothing items in labouring families. Handkerchiefs, shoes and stockings which were replaced regularly, were stolen frequently, jackets and stays more rarely.

There were also some practical reasons why shoplifters avoided the high-fashion goods that elite shops sold. To enter the emporiums in which the rich shopped added a heightened degree of risk. Testimony confirms shopkeepers’ deep reluctance to suspect any customer who appeared genteel, but in elite areas such as London’s West End retailers had an established clientele and a new face was likely to draw attention. A few shoplifters did try their luck by making an effort to dress the part and their polite fashioning and acting skill, witnesses recall, was often masterly. But an accidental slip into plebeian manners was easily done. Three customers dressed in silk drew the suspicion of a Covent Garden shopwoman as, she explained, ‘they called me my dear in a very sociable way’.

In general, shoplifters restricted themselves to plundering smaller local shops that were convenient to reconnoitre and with fewer staff to mount surveillance. A mapping of incidents in London shows this bias towards poorer and less fashionable districts, particularly to the north and east of the capital. The research found that within these working neighbourhoods shoplifted goods played an instrumental role in the intricate social and economic relations that underpinned community survival. Local associates earned money selling or pawning goods for the thief, their reputation serving to give the transaction an added credibility. Neighbours were informally sold stolen items on favourable terms, often including an element of exchange and credit, which acted to secure their complicity and future loyalty. We also come across shoplifted goods that were pawned to fund the shoplifter’s ongoing business or even recommodified as stock for their small retail concerns. Need rather than consumption fever motivated these shoplifters. Shoplifting was a capital crime throughout the century but this seems to have been of very little moment when the dictate was economic survival. As a shoplifter bluntly testified of her friend in 1747, ‘The prisoner came to me to go with her to the prosecutor’s shop, she wanted money, and she should go to the gallows’.


SAVE 25% when you order direct from the publisher using the offer code BB500 online at https://boydellandbrewer.com/shoplifting-in-eighteenth-century-england-pb.htmlOffer ends 5th March 2019. Discount applies to print and eBook editions. Alternatively call Boydell’s distributor, Wiley, on 01243 843 291, and quote the same code. Any queries please email marketing@boydell.co.uk


To contact Shelly Tickell: s.g.tickell@herts.ac.uk

The returns to invention during the British industrial revolution

by Sean Bottomley (Max Planck Institute for European Legal History)

This article is published by The Economic History Review, and it is available on the EHS website


british industry
Ancoats, Manchester. McConnel & Company’s mills, about 1820. Available at Wikimedia Commons

Since the Victorian period, it has been commonly assumed that inventors were rarely remunerated for their inventions. To contemporaries they were ‘the miserable victim of [their] own powerful genius’, ‘Martyrs of Science’ who worked ‘alone, unfriended, solitary’, while ‘the recorded instances of the[ir] martyrdom would be a task of enormous magnitude’. Prominent examples of important inventors from the industrial revolution period, but who had the misfortune to die in penury (the steam engineer Richard Trevithick, for example), has meant that this view has passed into the modern literature almost without scrutiny.

This assumption, though, is significant, as it directly informs how we might explain probably ‘the’ big problem in economic history: what were the origins of the industrial revolution, and concomitantly, of modern economic growth. In particular, if inventors did usually fail to obtain financial rewards, this precludes potential explanations of the industrial revolution that invoke incentives to explain the actions of those who invented and commercialised the new technology industrialisation required. It also precludes the applicability of endogenous growth theory to the industrial revolution (theory which has earnt two of its progenitors 2018 Nobel prizes) as it assumes that profit incentives determine the amount of inventive activity that occurs.

In an attempt to determine the wealth of inventors, I have collected probate data for over 700 inventors born in Britain between 1660 and 1830, from a list first compiled by Ralf Meisenzahl and Joel Mokyr. This probate data indicates that inventors were in fact extremely wealthy. For instance, in one exercise, I compared the probated wealth of 422 inventors who died between 1800 and 1870, with that of the overall adult male population.


Table 1.           Probated wealth of inventors, 1800-1870

Probated wealth Adult male population (1839-1841) Adult male population (1858) Inventors
<£200 or no will 73302 (88.14%) 87043 (87.70%) 124 (29.4%)
<£1,000 5570   (6.70%) 6690   (6.74%) 39   (9.2%)
94.84% 94.44% 163 (38.6%)
<£10,000 4296 (5.16%) 4554 (4.59%) 104 (24.6%)
<£50,000 812 (0.82%) 95 (22.5%)
   £50,000+ 154 (0.16%) 60 (14.2%)
5.16% 5.56% 259 (61.4%)

Notes: For details on how the distribution of male probated wealth was estimated for 1839-41, and 1858, please refer to the appendix in the original article published in the Economic History Review.


The table above shows us that approximately 5 to 6 percent of adult males who died in 1839-41 and 1858 (years for when these figures can be collated), left behind wealth probated in excess of £1,000. The equivalent figure for inventors was over 60 percent. The disparity only increases as we move up through the wealth categories. Whereas only 0.16 percent of adult males left behind wealth probated in excess of £50,000 in 1858 (one in 650), for inventors it was 14.2 percent (one in 7).

It does not, however, automatically follow that the wealth of inventors was actually derived from their inventions. These were presumably talented individuals and their income may have been accrued over the course of a ‘normal’ business career and/or inherited. Unfortunately, this is a prohibitively difficult subject to approach directly: accounts rarely survive for these inventors and in any case, it is doubtful whether income from an invention could be neatly distinguished from ‘normal’ business income. As an indirect approach, I have also collected probate information for the brothers of inventors. Brothers are an especially apposite group for comparison: they would have enjoyed a very similar inheritance to their brothers (although inheriting financial capital appears to have mattered less than inheriting social capital) and they tended to enter similar occupations to their (inventive) brothers. Indeed, 24 of the inventors in the entire dataset were related as brothers – the talents and opportunities required to become an inventor were clearly not evenly distributed among the adult male population.

For 143 of the 422 inventors discussed in table 1, it was possible to confirm the existence of at least one adult brother who reached at least the age of 25 and who died in Britain between 1800 and 1870 (253 brothers in total). In the table below, the top row divides these 143 inventors into the same wealth categories as those used in the table above, with the number in parentheses denoting how many of the 143 inventors are in each category. The columns beneath this then show the distribution of the wealth of their brothers. So, there are 25 inventors in this exercise whose estate was worth less than £200. Of their 45 brothers, 31 were also left behind less than £200. Three had probated wealth between £200 and £1,000, nine between £1,000 and £10,000 and two between £10,000 and £50,000. None left behind more than £50,000.


Table 2.           Brother’s Probates, 1800-1870

< £200 (25) < £1,000 (11) < £10,000 (35) < £50,000 (44) £50,000+ (28)
     < £200 31 12 26 35 23
  < £1,000 3 3 7 7 2
< £10,000 9 2 14 31 13
< £50,000 2 2 3 9 8
    £50,000+ 3 2 6

Notes: as Table 1


Overall, if inventors were wealthier than their brothers, then the latter should be concentrated at the top and to the right of the table, and away from the bottom left corner. Clearly, they are – overwhelmingly so when one considers how important simple happenstance can be in influencing an individual’s financial success over the course of their career.

Previous work has relied on impressionistic evidence to suggest that inventors in this period rarely obtained financial rewards commensurate with their technical achievements. Probate information, though, shows that inventors were extremely wealthy relative to the adult male population. Inventors were also significantly wealthier than another group who would have received a similar inheritance (in terms of both financial and social capital) and entered similar occupations: their brothers. Their additional wealth was derived from inventive activities: invention paid.


To contact Sean Bottomley: bottomley@rg.mpg.de