I have never been what The Kinks once lyrically described a ‘dedicated follower of fashion’, being an anorak in every sense. As an undergraduate in the 1970s, at a time when abstract theory and econometrics were becoming all the rage, my preference was for historical perspective and statistical data rather than ever more sophisticated statistical manipulation. The labour market being my favoured subject I was thus drawn to the work of the likes of Lydall, Phelps Brown and Routh. A B (Tony) Atkinson also crossed my radar at that time and this book demonstrates his continued mastery of the minutiae of longitudinal data sets.
I am usually wary of books based on lectures (in this case one delivered in 2006 at an Italian university). But Sir Tony’s lecture on the nature and causes of long-run trends in the dispersion of individual earnings forms only part of a near 500 page tome which combines general theoretical discussion with detailed individual chapters looking at data sources for, and the experience of, 20 OECD countries. And if the source material were not reason enough to read this book – which amongst other things usefully updates Lydall’s 1968 classic reference The Structure of Earnings for the countries covered – its fresh challenge to conventional wisdom on the matter of earnings dispersion should be. It raises a whole host of issues in political economy as well as in economic analysis.
Most contemporary economists deliver a factoid-rich narrative which portrays the past two decades as a period of widening individual pay dispersion following a prolonged era of stability in the pay structure. The driving force is said to be a combination of technology and trade effects which have favoured skilled (educated) individuals over the less skilled, with the observed impact on earnings dispersion being generic to all the developed economies. As Atkinson demonstrates, however, the actual historical record tells a more nuanced story.
Analysis of five countries for which Atkinson is able to assemble reasonable long-run data sets suggests that there was considerable change in pay dispersion prior to the 1980s. The UK, for example, experienced widening dispersion in the decade to the mid-1960s, a narrowing in the decade to the mid-1970s as workers at the bottom of the pay league caught up, and a subsequent decade long deterioration in the position of the latter. Since then there has been stability in the entire bottom half of the pay league but continued widening dispersion in the top half (what Atkinson calls the ‘fanning out’ of the distribution) with the very highest earners in particular racing away from the pack.
Looking specifically at the period since 1980, Atkinson finds that widening pay dispersion is indeed observed in most of the countries studied but that there is also considerable variation in the pattern and degree of this outcome. This variation seems due in part to differences in the policy response to skill-biased technological change (Atkinson mostly considers the possibility that some countries may be better than others in ensuring that the supply of skilled workers matches shifts in the demand for skills, though as he acknowledges his analysis doesn’t extend to the entire range of available policy responses). However, the book concludes that the simple trade/technology story is inadequate to fully account for the fanning out of pay in the upper reaches of the pay distribution in countries where this is observed. To explain this phenomenon Atkinson combines recent theories of ‘superstar’ pay with those which give a role to pay norms at the top of organisational hierarchies.
Most satisfyingly of all for those of us with an interest in social justice as well as economics, Atkinson ends the lecture section of his book by concluding that the distribution of earnings is not beyond our control: “All of these are matters over which citizens have some say; they are not purely spectators of an inevitable process”. Amen to that.
John Philpott
Chief Economist, Chartered Institute of Personnel and Development (CIPD) |