Automation, employment, and capital investment

At his blog STUMBLING AND MUMBLING, CHRIS DILLOW discusses recent reporting on rapid automation fears in the United Kingdom:

“‘More than six million workers are worried their jobs could be replaced by machines over the next decade’ says the Guardian. This raises a longstanding paradox – that, especially in the UK, the robot economy is much more discussed than observed.

What I mean is that the last few years have seen pretty much the exact opposite of this. Employment has grown nicely whilst capital spending has been sluggish. The ONS says that ‘annual growth in gross fixed capital formation has been slowing consistently since 2014.’ And the OECD reports that the UK has one of the lowest usages of industrial robots in the western world.

My chart, taken from the Bank of England and ONS, puts this into historic context. It shows that the gap between the growth of the non-dwellings capital stock and employment growth has been lower in recent years than at any time since 1945. The time to worry about machines taking people’s jobs was the 60s and 70s, not today.… If we looked only at the macro data, we’d fear that people are taking robots’ jobs – not vice versa.”

Link to the post. ht Sidhya

Dillow offers some possible theories for declining capital stock. Among the various possible causes are the findings from a 2005 paper by WILLIAM NORDHAUS, which shows that innovation doesn’t pay well except for a small fraction of firms. Link to that paper.

  • Robert Gordon in 2012 on the links between capital investment, innovation, and economic growth: “The paper is about “how much further could the frontier growth rate decline?” The analysis links periods of slow and rapid growth to the timing of the three industrial revolutions (IR’s), that is, IR #1 (steam, railroads) from 1750 to 1830; IR #2 (electricity, internal combustion engine, running water, indoor toilets, communications, entertainment, chemicals, petroleum) from 1870 to 1900; and IR #3 (computers, the web, mobile phones) from 1960 to present. It provides evidence that IR #2 was more important than the others and was largely responsible for 80 years of relatively rapid productivity growth between 1890 and 1972. Once the spin-off inventions from IR #2 (airplanes, air conditioning, interstate highways) had run their course, productivity growth during 1972-96 was much slower than before. In contrast, IR #3 created only a short-lived growth revival between 1996 and 2004.” Link.
  • A paper on the decline in research productivity from Nicholas Bloom, Charles I. Jones, John Van Reenen, and Michael Webb, whose work on investment and innovation we’ve linked to previously. Link.
  • Mark Paul’s ambitious Roosevelt Institute report on labor and automation, “Don’t Fear the Robots,” which we shared here on its publication in June, examines the situation and offers broad policy proposals for the management of technology’s effect on labor—rapid or otherwise. Link.


Contrasting two study methodologies

A paper by DAMON JONES, DAVID MOLITOR, AND JULIAN REIF, covered in a widely-shared New York Times article, shows the enduring importance of randomized controlled trials, and emphasizes that observational analyses are prone to selection bias and so do not provide an accurate assessment of impact, here on wellness programs’ effectiveness. From the Times:

“The results [from the randomized controlled trial] were disappointing. There seemed to be no causal effects. Here’s the nerdy fun part, though. In addition to this analysis, the researchers also took the time to analyze the data as if it were an observational trial.”

These methodological issues have specific consequences here, as they likely do elsewhere:

“The prior literature has overlooked important questions regarding selection into wellness programs. Wellness incentives may shift costs onto unhealthy or lower-income employees if these groups are less likely to participate in wellness programs. Furthermore, wellness programs may act as a screening device by encouraging employees who benefit most from these programs to join or remain at the firm—perhaps by earning rewards for behaviors they already enjoy.”

Link to the full paper.

  • On the observational vs. RCT debate: a systematic overview of over one thousand education-related RCTs conducted between 1980 and 2016, addressing major criticisms of the use of RCTs in education contexts. Link.
  • In 2014, Cochrane published a systematic review of RCTs vs. observational studies in the healthcare context. “Our results underscore that it is important for review authors to consider not only study design, but the level of heterogeneity in meta-analyses of RCTs or observational studies…In order to understand why RCTs and observational studies addressing the same question sometimes have conflicting results, methodological researchers must look for explanations other than the study design per se.” Link.

+ + +

  • The Electronic Frontier Foundation exposes the Sacramento Country Department of Human Assistance’s startling misuse of automated license plate readers to track welfare recipients in Sacramento County. Link.
  • A new study on solar geoengineering examines expected impacts on crops. Link. ht Bobby and Jack
  • An AMA from a participant in the (now cancelled) Ontario basic income pilot: “A person deserves the chance to realise the potential they have irrespective of the barriers they face. What that potential is will vary from person to person, but it is in society’s best interest to extract it, and the best interest of the individual to bring it forward. I find it baffling that some cannot look past the dollar figures. This is about equality of opportunity, not equality of results, yet the detractors of progressive policy claim it to be exactly the opposite.” Link. ht Lauren
  • “Branchless banking”: Examining the effects of India’s recent financial inclusion drive, which increased rural access to financial institutions through mobile technologies. Link.
  • A 2013 paper from Suresh Naidu and Noam Yuchtman examines law and the labor market in industrial Britain: “British Master and Servant law made employee contract breach a criminal offense until 1875… Positive shocks in the textile, iron, and coal industries increased prosecutions. Following the abolition of criminal sanctions, wages differentially rose in counties that had experienced more prosecutions, and wages responded more to labor demand shocks.” Link.
  • A new paper published by Equitable Growth and co-authored by Mark Paul (already referenced above), Khaing Zaw, Darrick Hamilton, and William Darity Jr. (whose work we’ve shared several times in the past) on intersectionality and the wage gap. “There is no single ‘gender’ or ‘race’ wage penalty… holding multiple identities cannot readily be disaggregated in an additive fashion. Instead, the penalties associated with the combination of two or more socially marginalized identities interact in multiplicative or quantitatively nuanced ways.” Link.
  • From Growth Econ, a review of Mariana Mazzucato’s new book The Value of Everything. Link.
  • Rachel Thomas at Fast.AI has written a clarion response to Alex Miller’s HBR piece “Want less biased decisions? Use algorithms.” Thomas writes: “Miller’s HBR article points out (correctly) that humans are very biased, and then compares our current not-so-great approaches to see which is less terrible. The article does not ask the question, how can we develop less biased ways to make decisions (perhaps using some combination of humans and algorithms)? which is a far more interesting and important question.” Link.
  • Religious prophesying glitch from Google Translate. Link. ht Margarita
  • In an unprecedented data visualization project, Mapping Inequality offers access to the HOLC’s “security maps” and explores the New Deal’s lasting effects on intergenerational poverty and spatial segregation. Link.
  • New research by Katherine Eriksson and Zach Ward examines trends in spatial segregation and the rate of immigration assimilation in the late 19th and early-20th century. “New methods and newly digitized data suggest that segregation in the US between 1850 and 1940 was both higher and more widespread than previously thought. However, despite slow rates of spatial assimilation, immigrants tend to assimilate culturally at a fast rate.” Link.

Each week we highlight research from a graduate student, postdoc, or early-career professor. Send us recommendations:

Subscribe to Phenomenal World Sources, a weekly digest of recommended readings across the social sciences. See the full Sources archive.