(I am an occasional contributor to roundtables of economists in the magazine The International Economy. This month’s topic was: “What are the policy implications if productivity growth is being under-measured in the official data?” My answer is below.)
How many hamburgers equal one haircut?
In itself, the question doesn’t make sense. They’re just different things. What we can compare, is how much they cost. This is true across the board: The only way we can convert all the endlessly varied objects and activities that make up “the economy” into a single number, is through their market prices. Markets are what let us express all the various products of human labor as a single quantity we call output.
This means that productivity is only meaningful in the context of market prices. There are lots of things that people do that are useful, important, even essential to economic life, from raising children to following the law, that can’t be expressed as output per hour.
So it doesn’t really make sense to ask if the nonmarket effects of technological change mean we are undermeasuring productivity. A new technology may transform our lives in all sorts of ways, but we can’t talk about its effect on productivity except insofar as its products are sold. There’s no other basis on which productivity can even be defined – we have to go by market prices. And what market prices are telling us is that productivity growth is slower than it used to be.
This slowdown is not really surprising. Manufacturing – where the transformation of work by technology has gone farthest, and where productivity growth almost always fastest – is steadily shrinking as a share of the economy.
It is true that we often think of economic growth as something broader than market prices. It’s supposed to describe a more general rise in living standards. So a more meaningful way to ask the question might be: Does measured productivity growth accurately reflect the material improvements in people’s lives?
The answer here is indeed no. But unfortunately, in the rich countries at least, the mismeasurement probably goes the opposite way as the question suggests.
Measures like life expectancy used to be closely linked with economic growth. In poor countries, this is still the case – higher GDP is associated with longer lifespans, lower child mortality, and similar improvements in health and wellbeing. If anything, today’s GDP growth may be associated with even faster improvement than we would expect based on the historical record. But in richer countries the opposite is true – higher GDP no longer translates reliably into better health outcomes. In some places – like the UK, and much of the US – life expectancy is actually falling, even as income per capita continues to rise.
Leisure time is another measure of wellbeing — presumably if people were having an easier time meeting their material needs, they would choose to take more time off work. (Adam Smith once suggested that the amount of leisure people enjoyed was the only meaningful standard of economic value across countries.1) On this measure too, living standards seem to be falling short of GDP growth rather than running ahead of them. Between the end of World War II and the early 1980s, the average weekly hours of an employed American fell by about 15 percent. But since then, average hours per worker have been essentially flat. This makes the postwar growth performance look even better, and the more recent performance worse, than the headline numbers suggest.
It seems likely that measured productivity overstates, rather than understates, our real improvement in living standards, at least in the US. If so, the policy implications seem clear. Policymakers should worry less about growth, and more about concrete interventions that we know improve people’s lives – things like universal access to childcare and health care, high-quality education, and paid time off for all.
- “Equal quantities of labour, at all times and places, may be said to be of equal value to the labourer. In his ordinary state of health, strength and spirits; in the ordinary degree of his skill and dexterity, he must always lay down the same portion of his ease, his liberty, and his happiness: The price which he pays must always be the same, whatever may be the quantity of goods which he receives in return for it.” Wealth of Nations, 1.5.7
I was going to write an ultra cranky comment, but here is a shorter version:
The slowdown in productivity and the fall in the quality of life IMHO are unrelated.
The fall in the quality of life is a result of the fall in worker’s bargaining power, that shows both in the fall of the wage share, worse shift, generally increased competition among workers, more anxiety etc.
The slowdown in productivity IMHO is mostly a price effect: before of the rise of the east some products were only produced in the west and therefore were overpriced relative to a notional equilibrium price in a world where everyone can produce them.
Now we are entering that world where everyone can produce industrial stuff, so there is an adverse price effect as this stuff goes towards the notional equilibrium price. This shows as a slowdown in productivity.
Incidentally saying that we are understating productivity is the same of saying that we are overstating inflation, since inflation adjusted dollars represent a basket of commodities.
The opposite is also true, however I’ve never seen an inflation truther or a productivity truther make this point.
“Policymakers should worry less about growth, and more about concrete interventions that we know improve people’s lives – things like universal access to childcare….”
Shouldn’t childcare be the exclusive preserve of families? Why should any child except an orphan be handed over to the care of paid professionals in institutions?
You write that “raising children” is “useful, important, even essential” but “can’t be expressed as output per hour.”
Yet the whole point of institutionalized “universal access to childcare” is to raise output per hour. Women’s mothering of their own children is economically unproductive by definition. So they are bundled off to the labor market where their productivity as, say, hotel chambermaids or prison guards or advertising copywriters can be monitored, optimized and monetized. The money they earn in turn pays (directly or indirectly through taxes) for strangers to provide monetized care to their children, with startling productivity gains: twenty children raised by a single childcare professional instead of ten mothers.
You’re paying lip service to free, spontaneous, authentic life outside the scope of economic rationalization; the epitome of that is caring for one’s own children. But your policy platform envisions the formal economy subsuming and monetizing childcare—capitalism with a vengeance, except under the control of bureaucrats instead of entrepreneurs. Under that vision it’s inevitable that crass productivity metrics will dominate child-rearing, because that’s how childcare institutions work; they’re economic institutions, not families.
Families are economic institutions too, though.
By the way the “natural” free, spontaneous, authentic life is more likely to be the extended family or the clan, very likely a matrilineal clan: the atomic family is also a recent creation.
A late comment :
In an idealised market, all transactions would offset each other, there would be 0 debt and trade would be barter under another name.
In the real world, debt exists, therefore transactions don’t completely offset each other, and this makes possible, among other things, long term trade unbalances.
This is relevant because we can see trade unbalances as showing over valuations and undervaluations: if say Germany and Italy are net exporters and UK and USA are net importers, then it means the euro is undervalued and the dollar and the pound are overvalued.
This doesn’t mean that over or under valuations are the cause of the unbalances, but it means that whatever the causes the relative value of the currency didn’t change enough to offset it.
So if the dollar is over valuated, it means that actual USA productivity is even lower, even if we don’t know how much lower.