The productivity paradox

People are worried about robots taking jobs. Driverless cars are around the corner. Restaurants and shops increasingly carry the option to order by touchscreen. Google’s clever algorithms provide instant translations that are remarkably good.

But the economy does not feel like one undergoing a technology-driven productivity boom. In the late 1990s, tech optimism was everywhere. At the same time, wages and productivity were rocketing upward. The situation now is completely different. The most recent jobs reports in America and Britain tell the tale. Employment is growing, month after month after month. But wage growth is abysmal. So is productivity growth: not surprising in economies where there are lots of people on the job working for low pay.

The obvious conclusion, the one lots of people are drawing, is that the robot threat is totally overblown: the fantasy, perhaps, of a bubble-mad Silicon Valley — or an effort to distract from workers’ real problems, trade and excessive corporate power. Generally speaking, the problem is not that we’ve got too much amazing new technology but too little.

This is not a strawman of my own invention. Robert Gordon makes this case. You can see Matt Yglesias make it here. Duncan Weldon, for his part, writes:

We are debating a problem we don’t have, rather than facing a real crisis that is the polar opposite. Productivity growth has slowed to a crawl over the last 15 or so years, business investment has fallen and wage growth has been weak. If the robot revolution truly was under way, we would see surging capital expenditure and soaring productivity. Right now, that would be a nice “problem” to have. Instead we have the reality of weak growth and stagnant pay. The real and pressing concern when it comes to the jobs market and automation is that the robots aren’t taking our jobs fast enough.

And in a recent blog post Paul Krugman concluded:

I’d note, however, that it remains peculiar how we’re simultaneously worrying that robots will take all our jobs and bemoaning the stalling out of productivity growth. What is the story, really?

What is the story, indeed. Let me see if I can tell one. Last fall I published a book: “The Wealth of Humans”. In it I set out how rapid technological progress can coincide with lousy growth in pay and productivity. Start with this:

Low labour costs discourage investments in labour-saving technology, potentially reducing productivity growth.

My background is in economic history. Economic historians often explain divergences in patterns of industrialisation by pointing to differences in labour costs. British workers were expensive relative to workers on the continent and relative to British energy. British firms therefore had an incentive to develop and deploy new technologies that economised on labour and used a lot of energy: industrialisation! There’s more to the story than that, but that’s a pretty big component. Later something similar happened in America, where workers were even more expensive and resources even more abundant, and where the phenomenally productive “American system” of manufacturing therefore emerged.

We all sort of know this: that whether and how companies decide to develop and use technologies depends on the costs they face. But we often forget it when it comes to debates about robots and automation.

Today, labour costs are relatively low. In real terms, wage growth has been close to imperceptible for most of the workforce since 2000, and in some cases going back much earlier. The real value of the minimum wage in America is quite low relative to what it was a half century ago. Now: it is also true that the cost of computing power and data storage has fallen, a lot. Perhaps unsurprisingly, the “capital share” of income has fallen by more in recent decades than the “labour share”. One could argue that because the cost of technology has fallen by more than the cost of labour, we ought to have seen mass automation if in fact digital technology is all it’s cracked up to be.

But the scarce factor isn’t capital equipment. What is expensive is the intangible capital that’s needed to overhaul production in ways that use cheap computing power to eliminate lots of jobs. It is complicated to figure out how to get these systems working and operating in a way that generates profits. While labour is cheap, firms face little pressure to make those massive investments in intangible capital in order to automate key processes.

Returning to the industrial parallel, it was not the case that James Watt developed his steam engine and everyone said “great, this technology is clearly superior to everything else and we will therefore use it all across the economy”. Rather, it was used in a small number of contexts in which the economics (expensive labour, cheap energy) pushed business owners to experiment. Then, over time, engineers improved the technology and firms built up a wealth of knowledge about how to use it to make a buck. Then eventually the technology was so good that it began to be adopted in places where labour costs were not all that high.

Cheap labour is reducing the incentive to push new technologies along a similar path.

The digital revolution is partly responsible for low labour costs.

The digital revolution has created an enormous rise in the amount of effective labour available to firms. It has created an abundance of labour. If you’re a company and your workforce is demanding higher pay or being difficult, you have many ways to get the labour you need without adding to your wage bill.

You can move work abroad. Technology enabled the growth of global supply chains, which helped bring billions of low-wage workers into the global labour force. You can restructure your business in ways which allow fewer, more skilled workers to use technology to do tasks which previously required lots of less-skilled workers; or you can restructure your business in ways that reduce the bargaining power of your employees, or reduce your obligations to them. And, of course, you can automate.

How does automation contribute to this abundance of labour? Well, there’s a long-run story in which the march of progress means that technology is increasingly capable of substituting for human workers across a broad range of tasks. If firms are indifferent between using people and machines, and if machines (or code) is abundant, then the effect of progress is to create a single mass pool of “labour” (meaning people and people-like machines) that is super abundant.

But there’s a more straightforward and important way in which automation adds to abundance now. When a machine displaces a person, the person doesn’t immediately cease to be in the labour force. It is not the case that in period one the economy produces x using y workers and in period two it produces x using y-1 workers and therefore productivity goes up. No, that displaced worker probably has all sorts of bills to pay and must therefore find another job.

In some cases workers can transition easily from the job from which they’ve been displaced into another. But often that isn’t possible. Generally speaking, the workers displaced by technology will tend to be those without much in the way of exceptional skills or training. Such workers find themselves competing for work with many other people with modest skill levels, and with technology: adding to the abundance of labour.

This is a critical point. People ask: if robots are stealing all the jobs then why is employment at record highs? But imagine what would happen if someone unveiled a robot tomorrow which could do the work of 30% of the workforce. Employment wouldn’t fall 30%, because while some of the displaced workers might give up on work and drop out of the labour force, most couldn’t: they need the money. They would seek out other work, glutting HR offices and employment centres and placing downward pressure on the wage companies need to offer to fill a job: until wages fall to such a low level that people do give up on work entirely, drop out of the labour force, and live on whatever family resources they have available, or until it becomes economical to hire people to do very low productivity work — serving as the fifth landscape worker on the household staff of a very rich tech magnate, for example.

Given the structure of our social safety net, automation tends to increase poverty and inequality rather than unemployment.

What effect does all this have? There are several.

First, as the economy attempts to absorb lots of relatively undifferentiated labour, wages stagnate or fall. As wages fall it becomes economical to hire people for low productivity work. So employment in low productivity jobs expands, affecting aggregate productivity figures.

Second, the abundance of labour, and downward pressure on wages, reduces the incentive to invest in new labour-saving technologies. Until pressed, firms don’t overhaul and automate their warehouse or swap out some wait staff for touchscreens. So productivity within sectors grows more slowly than it otherwise might. There are dynamic effects as well: when you don’t deploy new technologies because labour is cheap, you don’t get all the tweaks and knock-on innovations and accumulation of intangible capital that contributes to still more productivity growth down the road.

Third, the abundance of labour destroys worker bargaining power. That hurts in lots of different ways. Workers find themselves receiving a declining share of income. Workers are unable to bargain for the kinds of changes that might reduce the pain of economic shifts: like training to help some workers fill new niches complementary to technology, or profit-sharing.

Fourth, low wages and a falling labour share lead to a misfiring macroeconomy, because more resources are concentrated in the hands of those with high propensity to save while the impetus to invest is also diminished. And chronically weak demand is worse for workers and for productivity growth.

But this last point begins to take us into later chapters of the book, which examine the stresses that the dynamics described in this post place on other parts of the economy and society.

So there you are: continued high levels of employment with weak growth in wages and productivity is not evidence of disappointing technological progress; it is what you’d expect to see if technological progress were occurring rapidly in a world where thin safety nets mean that dropping out of the labour force leads to a life of poverty.