Substitution or complementarity III: Context and strategy

In this series of posts: 

Carl B. Frey & Michael A. Osborne. The Future of Employment: How susceptible are jobs to computerization? Technological forecasting and social change. 2017. Published version here. Open access working paper here.

Melania Arntz, Terry Gregory & Ulrich Zierahn. The risk of automation for jobs in OECD countries. OECD working paper available here.

David H. Autor. Why are there still so many jobs? The history and future of workplace automation. Journal of economic perspectives. 2015. Published, open access version available here.

Susan F. Lu, Huaxia Rui & Abraham Seidman. Does technology substitute for nurses? Staffing decisions in nursing homes. Management Science. Forthcoming. Published open access version available here.


So we’ve covered the (anxiety-prompting) prospects for job destruction by technology and the (anxiety-dampening) history of job creation. Now, we’ll turn to something close to my heart, namely the role of context and firm-level strategies in deciding the actual effects of technology’s potentials.

Susan Lu and her colleagues look at when technology (specifically a system called CPOE) leads nursing homes to employ fewer (technology substitutes) or more (technology complements) nurses. It’s a wonderful case, I think, because traditionally we might think that nursing work certainly can’t be automated. It’s all about human care that robots can’t do and even if robots could do it, lots of people would probably prefer humans. But if you observe it, is it also clear that parts of their work is not about human care, but rather administrative and repetitive in nature. CPOE seeks to automate some of those parts. In principle, that should reduce the demand for nurses.

As it turns out, technology does substitute for nurses and lower employment, but only in some nursing homes. In others, it increases employment. When and how so?

In high-end nursing homes (defined as a home with a high nurse-to-resident ratio), adopting CPOE lets nursing homes achieve the same high quality with fewer nurses. Because the marginal benefit of more available nurses on quality is low (because the service is already good), the resources that technology frees up is turned into lay-offs. In low-end nursing homes (with low nurse-to-resident ratios), something else happens. When nurses are freed up in a low-end nursing home, quality increases (the marginal benefit of having more nurses available is high, because service is not so good). This allows the nursing home to charge more for its service, making it profitable for the low-end nursing home to expand and in doing so hire more nurses.

It’s a lovely and very elegant explanation and a fine illustration of how substitution and complemenarity is (in part at least) determined by the two factors I mentioned in the beginning: context and strategy.

Regarding context, it would not immediately be obvious that technology would substitute for nursing. But, in the particular context of a private market for nursing homes where homes choose their own levels of staffing and compete for residents of varying ‘profitability’, a technology substituting for nurses’ administrative work could clearly matter. In a Scandinavian context of publicly owned and operated nursing homes with centrally defined funding, the effects would be different – without political intervention, perhaps quality would improve (or work could become less stressful), with political intervention lay-offs might be forced through. And of course the nature of the technology matters greatly too. CPOE has one set of effects, a different technology has others.

I particularly like the strategic side of the story, because it shows how the same technology allows firms to take different strategic directions and positions (in this case, relating to the quality of service they provide) and interacts with previously-made choices. It made me think of the decisions to substitute humans for machines in places like bank service call centers, so that you chat with a machine instead of talking to an actual person (which, to my experience, currently works so poorly that I’ve found it to be deplorable service). Or of how MOOCs and online educations and courseware are substituting for traditional higher education – sometimes providing comparable or better services, sometimes more dubious ones, but definitely different. A university like Georgia Tech seems to use technology to fill an entirely new niche with their very affordable online master’s degree in computer science, with different roles and demands for university instructors (who some thought were hard to automate).

For sure, there are problems with the paper as with any empirical paper (as the comparison between Frey & Osborne’s analysis and the one from the OECD) snd there is such much more to explore around this issue – as soon as you say context matters, you’re opening up a whole range of questions about moderators, mediators, boundary conditions and contingencies (which, for a question as complex as how technology influences work, organizations and employment, are really healthy to try to answer or at least know that you haven’t answered yet). And of course, strategic choices about quality are just one of many  that technology can potentially influence. I look forward to following this line of research into the future.
Upcoming posts: 

  • The richness of work: The tensions between the OECD report and Frey & Osborne made it clear that a deep understanding of what work actually is matters, especially when talking about how that worl will be changed by technology. This made me think of Julian Orr’s brilliant ethnography of service work and some of the ideas that lead up to it, like the nature of knowledge and de-skilling.
  • Adjustment: Autor mentions James Bessen and his book on ‘Learning by doing’ is a great historical treatment of what it takes for the ‘bounty’ of new technology to be spread amongst the many. A review is coming up. LINK.
  • The essential fact of capitalism: I’m currently re-reading Schumpeter’s 1942 ‘Capitalism, Socialism and Democracy’ for a class I’m teaching on why organizations change. There’s a great, great analysis in that book on the nature of capitalism (where Schumpeter coined the term Creative Destruction) and whether capitalism can survive. I’ll post about that in the not-too-distant future. LINK. 



3 thoughts on “Substitution or complementarity III: Context and strategy

  1. Pingback: The end of scarcity (maybe) | The Gale

  2. Pingback: Substitution or complementarity II: (Quiet) optimism | The Gale

  3. Pingback: Teaching after the break, part 1 | The Gale

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s