On archives, macroeconomics and labour markets

October 24, 2017 | Blog
Home > On archives, macroeconomics and labour markets

Everything (well,… most things) you know about wages 1650 -1800 is wrong. That’s a great opportunity for historians

by Judy Stephenson (University of Oxford)

 

My forthcoming paper in the Economic History Review makes some big claims about the level of nominal and real wages in urban England before industrialization. There is an early working paper version available.

Specifically, I argue that the data used for the years between 1650 and 1800 are completely wrong because the people who compiled them (who go back in some cases to the 1930s and late nineteenth century) took figures from bills for construction services rather than actual wage books. As an actual wage book from the contractor who built the South West Tower of St Paul’s shows, men were not paid these charge out rates, they were paid considerably less.

This has some big ramifications for some influential economists and historians who have relied on long established data sets of ‘builders wages’, such as those of Phelps Brown and Hopkins (1955, 1956) to create macroeconomic models of the past to calculate real wages and infer GDP; to argue that Britain had ‘high wages’; or a comparative advantage in traded goods; or a narrower ‘skill premium’ and better institutions.

In truth, that these wages were ‘wrong’ is in no way surprising to anyone who has ever done work on early modern earning. Any historian of the eighteenth century sensed that these ‘average wages’ were unreasonably high and that their implied welfare ratios gave a falsely rosy picture. (As someone face palmed; ‘A labourer in London able to afford a respectable basket of goods for a family in the mid eighteenth century?? Have you read Dorothy George?’). Those who have ever worked with labour records and account books know that the homogenous figures found by Elizabeth Gilboy were questionable, and indeed in 2011, John Hatcher had successfully called into question the golden age of the fifteenth century. ‘Real’ day wages and wage accounts are always fascinatingly messy, unpredictable, and varied, yet econometricians clung to the old data sets because they believed it was too difficult to find anything else.

Untitled
Chart showing comparative real wages, in grams of silver, of European cities 1650 -1800, based on Allen, (2001), where the nominal wage has been reduced by 25%.

My findings make the idea that Britain was a ‘high wage’ economy in the long eighteenth century hard to sustain. If paid wages were 25% lower than we thought, the real wage for labourers through this period in London was not the highest by far. Rather, it seems, they were at the lower end of NW European advanced economies.

This is exciting for economists who think that explaining why the industrial revolution happened in Britain is the ‘Holy Grail’ (it’s back up for grabs). But, the debunking of these inaccurate wage series also makes it a really exciting time for people who want to understand the role of labour in the economy, and who think that the period before collective bargaining and factories has some strong parallels with our own. Lots has been written about the decline of ‘history’ in economic history, but the new opportunity is as wide and bright for historians as it is for economists and econometricians. This breakthrough in this long-run view on wages came not from new statistical techniques, but from the margins of dusty parchment, little iron pins, raggy old papers, smudged watered down ink, and the tentative ‘x’s’ and proud flourishes of the archives.

It’s time to stop recycling tired old data sets and expecting new technology to tell us something different about them. There is a wealth of sources and data in London archives, which have never been used before because they didn’t look comparable to Elizabeth Gilboy’s ‘day rates’, but which offer historians and economists the potential to look at earning, bargaining and the capital labour relationship in new ways. There is exciting work in progress from established and new scholars in the field. No one data set will ever be able to replace the supposed reliability of Phelps Brown and Hopkins, but even they were very tentative about their sources.

To contact the author: judy.stephenson@wadham.ox.ac.uk

References list available here

SHAPE
Menu