Skip to content

On labour shortages and automation

Last month, the ABC reported that Singaporean organisations had taken to employing robots in response to a shortage of workers previously supplied by migration. A more recent article had the horticultural industry developing robots to address a shortage of fruit pickers in Australia. In between the two, I was reading Robert Gordon’s The Rise and Fall of American Growth (2015), which here and there mentions the impact of labour under- and over-supply on organisations’ propensity to automate (amongst an enormous number of other influences). The basic logic is that a shortage of labour and high labour costs make automation more attractive, while an abundance of labour and low labour costs make automation less attractive.

I can’t recall this logic appearing in any public debate around wages and productivity—yet the whole point of the “productivity” that economists and politicians talk so much about is to produce more with less work, so that people can have more (wages or profit) while doing less. Through that lens, one might expect economists and politicians to be cheering on labour shortages, increased labour costs (wages), and whatever automation organisations do in response.

Reading the ABC’s description of the cutesy robot that the National Library of Singapore uses to scan shelves for missing books, I wondered why a library would wait for a critical labour shortage to adopt such a thing, reported to be able to scan nearly a third of the library’s collection each day. There are several obvious answers: the robot costs money; the notice posted on its front suggests that it might fail if a library patron is standing in its way; and maybe the library just doesn’t need to scan a third of its collection every day.

I suspect the “costs money” explanation accounts for a good deal of resistance to improving productivity. In the short term, deploying new technology costs money, disrupts production, and comes with a risk that it may not work out. And if some of the benefit of increased productivity goes to paying higher wages, that’s even more money. From the employee perspective, if becoming more efficient means increasing an organisation’s profit for no reward, or automating oneself out of a job altogether, why do it? (To be fair to economists, many do note that rising wages are linked to increases in productivity.)

In other cases the automated solution might not work as well—I wrote a bit in my old blog about technology enthusiasts mistaking near-replacements for human skills with total replacements—or the technology might be overkill for the problem at hand. As the system administrator for, I have plenty of semi-automatic procedures that don’t quite solve the problem by themselves, access to mountains of open source software that does all manner of things I don’t need, and limited time improve the procedures and sort through software features.

Still, necessity is evidently the mother of the invention in at least some circumstances. I’m not sure anyone could reasonably prescribe labour shortages as a means to increasing productivity, though some approaches to four-day working weeks (such as Andrew Barnes’) might be thought of as imposing a “shortage” of working hours that forces organisations and workers to use them more carefully. But, on the flip side, what worker (or even economist or politician) is going to sign up for throwing abundant low-cost labour at the economy?

1 thought on “On labour shortages and automation”

Leave a Reply