Seite - (000052) - in Disrupted Development and the Future of Inequality in the Age of Automation

44 L. SCHLOGL AND A. SUMNER they eliminate (“technological unemployment”) or create jobs in sectors which are potentially less desirable and productive (“premature deindus- trialization”). Politically, the recommendations of the pessimist camp range from a “robot tax” to redistributive responses such as a universal basic income (with the latter potentially funded by the former) and ques- tions of public versus private ownership of production and technology. It is fair to say that the second, more pessimistic, camp has been increasingly visible in recent years. Yet, unemployment is generally not considered to be the main issue. With a view to the United States, eco- nomic experts from the IGM Panel (2014) agree that automation has not (yet) markedly reduced employment but has rather led to a stag- nation of median wages, a decoupling of real-wage growth from pro- ductivity growth, and a labor market polarization or “hollowing out” of middle-skill employment. Technology can depress or enhance wage growth depending on whether it substitutes or complements tasks (see for discussion, Acemoglu & Autor, 2011; Autor, Katz, & Kearney, 2004; Firpo, Fortin, & Lemieux, 2011; Goos & Manning, 2007). Further, it should not be taken as given that lower skilled work will necessarily be automated, but it can contribute to a “missing middle” whereby most jobs are low or high skilled, and those in-between are relatively more susceptible to automation, or whereby employment expansion in those middle-skill jobs is weaker than that of low- and high- skilled jobs (see Autor, Levy, & Murnane, 2003). The problem thus may not be so much that jobs are lost, rather than that other types of jobs expand in number. People are being driven into the jobs below their skill level, with either lower or slower growing wages than the middle-skill jobs that previously existed. A key question is what happens to productivity growth in any given country. In short, who “captures” the productivity growth in terms of capital or labor and the functional distribution of income. And how what is captured is then distributed within the capital share (which may be distributed between reinvestment, dividend payments, reserves building, or other activity e.g. rents), or within the labor share which may be distributed between employment growth, real-wage growth, or social security entitlements (see discussion of Atkinson, 2009; Francese & Mulas-Granados, 2015). This matters from an individual income inequality perspective, as reductions in the labor share of income are correlated with rising income inequality between individuals (see for detailed discussion, Chapter 3 of IMF, 2017).