I apologise for another post that’s indirectly about AI, but it’s hard to think about much else at the moment!
This week’s article focuses on the slightly less direct effects of new technologies like it. As they increase productivity, they tend to decrease costs while increasing wages (since individual workers are producing more). But what happens to other industries, ones that aren’t experiencing the same increases in productivity? Frighteningly, they also experience rising costs, because their workers could ultimately choose to go to the industries where productivity is increasing.
It’s scary to be in an industry that’s being directly disrupted by new technology. It’s not a fun time to be a writer or a programmer at the moment. But it can be scarier still to be in an industry that isn’t being directly disrupted: you’re impacted nonetheless, but have none of the productivity benefits to soften the blow.
This week’s article
Why you don’t just need to worry about your competition – other industries matter too
Why do the members of a string quartet earn more today, even adjusting for inflation, than they did in the 1700s? Their productivity hasn’t increased; it still takes the same amount of people the same amount of time to perform a particular piece now as it did three hundred years ago. In some sense, shouldn’t their wages have remained static with their productivity? What gives?
You can figure out the answer by imagining what would happen if you tried to pay modern musicians inflation-adjusted 1700s wages. Presumably after telling you in no uncertain terms where you could stick your offer, they’d then go to work somewhere else. They could, after all, wander into a manufacturing or technology job, in a sector where productivity has increased, and earn much higher wages than an 18th-century musician would have earned.
If you want to employ a string quartet these days, then, you’ve got to pay them more than you would have needed to in the past. That’s because there are other areas of the economy that have undergone improvements in productivity, whose wages have increased as a result, and in which your musicians could choose to work. This observation is known as Baumol’s cost disease, and it’s named for the economist William J. Baumol who first described it with his colleague William G. Bowen in the 1960s.
To put it into more abstract terms: some sectors of the economy experience productivity increases, but others don’t. Wages rise in the high-productivity-growth sectors, but they must also rise in the other sectors too; otherwise, there would be an exodus of employees from the low-productivity-growth sectors to the high-productivity-growth ones. This simple effect explains much of the broader economy, such as why services are increasing in price faster than goods and why so much more of the economy and employment is moving towards services like nursing, teaching, childcare, and food service. Because productivity doesn’t increase in some sectors, relatively more and more of the population ends up working in them; and because prices increase, we end up spending more and more of our money on them, which means overall productivity (and economic growth) starts to slow.
At the time that Baumol was writing, the greatest differences in productivity growth were between manufacturing industries and service industries. Looking back thirty years from when Baumol was writing, for example, it took far less time and labour to manufacture, say, a car or a radio in 1960 than it did in 1930. But in the same period, the time it took for a university professor to mark an essay, a surgeon to perform an operation, or a teacher to teach a class didn’t change at all.
Simplification of the products themselves, better tools, improvements in assembly line technology and processes, more efficient supply chains, better training of workers; all these things increased productivity in manufacturing. But there were no such increases in service industries, because so much of what they did was constrained by unchanging human factors.
As a business, cost disease is an unpleasant malady. Your productivity isn’t growing, but your costs are. Eventually, something has to give. You can raise prices, but that naturally affects demand for what you make or do. You can lower quality, but that also affects demand and has other tradeoffs. You can accept lower margins and fewer dividends, but that gives you less room to manoeuvre. None of these options are particularly appealing. The much better option is to find a way to increase productivity. But how?
In the 1960s – when Baumol was writing – the stark divide was between manufacturing businesses and service businesses. Today the largest difference is between those sectors of the economy that have been able to harness technology and those that haven’t. Those businesses that have been able to become digital, exploiting the zero-marginal-cost nature of the internet as a distribution channel, have seen an explosion in productivity. Those businesses that have been forced to remain physical have seen a less dramatic increase – even if technology has increased their productivity in some ways.
Ten years ago, Marc Andreessen wrote presciently that “software was eating the world”. And as Roy Amara said, “we overestimate the impact of technology in the short term and underestimate the effect in the long term”. Andreessen was rightly assumed to be hyping up enormous pure-play technology businesses like Facebook and Twitter, which were the focus of everyone’s attention in the late 2000s and early 2010s. But his point was that software would eventually eat the world – that it would touch every industry and change every business.
That’s because software has magical properties that nothing before it has been able to match. It has the potential to achieve zero marginal costs. It has the ability to increase productivity in sectors that have previously been untouchable. It’s probably coming for your industry, if it hasn’t already. If you’re suffering from cost disease, software might well be the cure – but will it be you that sees the benefit?
This week’s four interesting links
I knew that butter lasted for a while out of the fridge, but didn’t realise quite how long – to the point that many people advocate never refrigerating it in the first place:
“In 2015, Ms. Mertzel sent samples of four brands of butter to a lab for testing. The finding: No sign of spoilage after three weeks of storage at 68 to 77 degrees Fahrenheit [20 to 25 degrees celsius]. She commissioned a similar analysis this year and found no spoilage after 30 days.
“‘This is a quality issue, not a safety issue,’ said Gina Mode, a butter researcher at the University of Wisconsin’s Center for Dairy Research. Butter will eventually go rancid but that won’t make people sick, she said. Ms. Mode in an informal survey of her colleagues found that 24 of 31 keep butter out, a telling data point among experts.”
Adam Rutherford showcases five data visualisations that changed the world, for both good and ill:
John Snow’s dot map of Soho, that led to an understanding of the transmission of cholera
Florence Nightingale’s Coxcomb of military deaths in the Crimean War
W. E. B. Du Bois’s graphs of African-American advancement in the years following slavery
Henry Goddard’s Kallikak family tree, used to justify eugenics
Ed Hawkins’s global warming colour stripes
Richard Jones, whose Soft Machines blog is a great read on industrial strategy, sets out what the UK should do to build an effective semiconductor strategy – of huge importance given the emergence of compute-hungry AIs. The options aren’t amazing, mainly because we’ve been neglectful in the past:
“The UK’s limited options in this strategically important technology should make us reflect on the decisions – implicit and explicit – that led the UK to be in such a weak position.
“Korea & Taiwan – with less ideological aversion to industrial strategy than UK – rode the wave of the world’s fastest developing technology while the UK sat on the sidelines. Their economic performance has surpassed the UK.”
Andreas Wagner, in an extract from his new book, details the evolutions – in both biology and human culture – that lay dormant for years before suddenly encountering the conditions in which to become successful:
“These and many other new life forms remained dormant before succeeding explosively. They are the sleeping beauties of biological evolution. They cast doubt on many widely assumed beliefs about success and failure. And these doubts apply not just to the innovations of nature, but also to those of human culture.”