Bob Hancké of the LSE applauds a rare empirical study of the effects of automation on inequality

An interesting blog post here, with a link to a paper by David Hope (Kings College London) and Angelo Martelli (LSE). They examine the inequality effects of the ‘knowledge economy’. There are two important reasons why this is a very relvant piece of research.

Very few pieces on the knowledge economy actually have falsifiable arguments, and/or are based on systematic (rather than somewhat anecdotal) evidence. The trailblazing works by Brynjolfsson and McAfee at MIT may be forgiven for that; theirs are agenda-setting pieces, which explore the contours of the new economy rather than test its effects. But much of this literature is methodologically quite weak: depending on your accounting as much as 60% of jobs are under threat from automation – or as little as 15%. That’s a pretty wide confidence interval.

The debate on the social effects of automation today and in the future also usually progresses along the lines of ‘there’s this big steam roller, and it will ultimately have the same effects everywhere’. But we know from previous debates on technological change and its organisational effects that these effects are always mediated through institutions and struggles over the technology. This time may be different, but without a good argument for why today political and institutional mediation of the effects is negligible this is theology, not an argument.

The paper and blog give a good sense of how key institutions in the labour market – wage setting and employment protection – modify, strengthen or counter what the rise of new sectors in the ‘knowledge economy’ may mean for wage inequality. Very useful, not least since this also opens a wider research agenda that can provide a more grounded counterpoint to overly optimistic or pessimistic views. Highly recommended.