Productivity Surge: Mirage Or Miracle?
The US productivity surge might be for real - and Europe might be left behind again.
Much of the excitement about generative AI is driven by the hope that it will boost productivity, accelerating economic growth and lifting living standards. I have argued why I think for now these expectations are overblown: the intrinsic fallibility of the current wave of generative AI systems undermines their potential in a vast range of economic applications.
But productivity growth, meanwhile, is generating its own share of excitement, rekindling the US-Europe rivalry. Euro area labor productivity plunged 1.2% year-on-year in the last quarter of 2023 (and across the EU it fell by 0.7%), while US productivity surged by 2.6%. An uncomfortable comparison for the old continent, and Q4 was just the tip o the iceberg: European productivity has been declining for the past five quarters, whereas US productivity has been accelerating for the past three:
Source: OECD
Productivity here is measured as output per hour worked, so this is not because Europeans enjoy longer lunches and more time with the family—it’s because when they’re at work, they produce less than Americans. And US productivity has not been boosted simply by robots and automation—unemployment remains at record lows. What is going on?
The government did it!
The Financial Times puts the emphasis on government policy, and specifically on public investment: it notes that “The recent jump in US productivity comes after a massive fiscal stimulus centred on green industry,” while “By contrast, the eurozone has received less fiscal support from governments”.
There is no doubt that the recent unrestrained fiscal largess of the US puts to shame even the most fervent Keynesian policymakers in Europe. But before we rush — yet again — to argue that more government spending is the key to long-term prosperity, let’s take a look at productivity trends over a longer period:
Source: OECD
Productivity growth figures are highly volatile, and I’ll come back to that. But the US systematically outperforms Europe. We can see this even more clearly by taking averages. Below I have broken down the sample in two periods: 1996-2005, when productivity growth was faster across the board, and 2006-2023:
Source: OECD
Over both periods, US productivity rose at more than double the pace of the euro area, and nearly twice as fast as the EU. The cumulative impact of such sustained divergence is dramatic: if we set the level of productivity at end-1995 equal to 100 for both the US and Europe, by the end of 2023 US productivity levels are 3 1/2 times higher than the euro area and 2 1/2 times higher than the EU. The chart below says it all:
Such massive and consistent outperformance over a thirty year period cannot plausibly be ascribed to the recent burst of US green investment. And no-one in their right minds would argue that for the last thirty years Europe has “received less fiscal support from governments” than the US. Insurer Axa’s chief economist Gilles Moëc correctly suggests “we need to contemplate the possibility that something structural is happening.” Quite. And after an appropriate period of contemplation, maybe think about whether anything can be done about it.
Those pesky trade-offs again
There is really no mystery here. The US economy has consistently been less tightly regulated, with a more flexible labor market, a competitive environment that rewards innovation and a financing ecosystem that sustains it, and a legal framework that enables the orderly resolution of failures through bankruptcy procedures that do not leave a permanent stigma. Like everything in life, these features are all part of trade-offs. A more flexible labor market leads to wider swings in unemployment than in Europe, but with a lower average unemployment rate. Europe has every right to keep making different choices — but lower productivity should not come as a surprise. Not anymore.
There has been some reshuffling of relative performance within Europe. During 1996-2005 the UK enjoyed the fastest pace of productivity growth, followed by France, with Spain at the bottom; then Spain surged ahead, the UK fell behind and France collapsed. Italy…maybe in Italy productivity carries a social stigma, like ordering a cappuccino after 10:30am…. If the US seems too foreign, maybe European countries could at least compare notes to see what’s working a bit better and what’s obviously not working at all.
Source: OECD
Missing the train — again?
As European policymakers contemplate the situation, they should also consider the possibility that the stakes might be getting higher.
When we looked at the historical averages, I isolated the 1996-2005 period because that is when most advanced economies experienced a remarkable acceleration in productivity growth (in the US it doubled). That acceleration was driven largely by the Information and Communication Technology revolution, as computers were deployed broadly across the economy.
At that time, the US enjoyed a much stronger productivity boom than Europe; and many economic studies indicate that the US economy’s greater degree of flexibility is what made the difference: greater flexibility allowed US companies to change operating procedures, management practices and incentives, to redeploy workers and resources. Some studies have found that even Europe-based subsidiaries of US corporations enjoyed greater efficiency gains than their European peers, suggesting that corporate culture played an important role beyond that of the legal and institutional context.
It is too early to tell whether the most recent acceleration in US productivity growth marks the beginning of a new golden age — given how volatile the data are, the last three quarters might prove to be a fluke. But over the past decade, companies have been carrying out a massive effort in adopting new digital technologies. And while there has been plenty of hype, a lot of important innovation has taken place, and it would be surprising if none of it ever resulted in some boost to productivity.
We might have finally reached the point where digital innovation starts to yield some tangible economic benefits. Europe squandered the previous digital productivity boost. If it misses this one as well, the competitiveness gap with the US will widen even further. And given the pressing need for greater economic resources for national security and stressed social safety nets, it would be a crying shame to miss out on a potential boost to growth.
Governments do have a role to play. Setting regulations and taxes at levels that foster stability without sacrificing dynamism should be a key priority. Equally important is building high quality digital infrastructure and encouraging the rise of stronger diversified entrepreneurial ecosystems. Education remains paramount. Keynesian economists should not worry — there are worthy causes for governments to spend money on. But creating pervasive regulation, setting taxes and social benefits so high they undermine incentives, and trying to favor selected industries through regulations and subsides will prove, as always, self-defeating. In a highly dynamic and uncertain environment, the private market economy is much more likely to get it right than any central planner.
This piece makes you think! That labor productivity growth difference is indeed spectacular. But why does half of the US population believe that they got completely screwed in the labor market over the last 30 years? One (nagging) thought in my head is that US employment and US payroll figures have diverged over the time period considered (payroll figures were always much higher than employment figures; just look at recent months). Thus, what could be happening is that effective labor input in the US is much higher than indicated by the (household survey) employment data. But I have no idea whether this effect could be large or small given how the EU does its labor statistics.