Greetings. Yesterday was Budget Day in the UK — perhaps more interesting politically than economically, as it was the last one before the next general election. For all the coverage you need, see the FT’s Budget page.

Last week, we dissected the US’s strong productivity growth since the pandemic, and one reader asked for the same treatment of Europe. So here goes.

Parsing Europe’s productivity stagnation

As is well known, the EU is nowhere near matching the US’s post-pandemic productivity bounce. It has had a decent recovery — the bloc’s gross value added rose 3.9 per cent in the four years to the third quarter of 2023 — but this is left in the dust by the US’s stellar rebound.

And whatever output growth the EU did have was largely a factor of more people working, unlike in the US. In those four years, EU productivity — measured as value added per worker — grew a measly 0.6 per cent (in total, not per year). To be fair, it is also the case that each person worked fewer hours on average compared with before the pandemic. So if you measure EU productivity by value added per hour worked rather than per person employed, the four-year rise is a more respectable 2 per cent.

But contrast this with the US where, in the non-farm business sector, output per worker rose about 5 per cent and output per hour worked increased more than 6 per cent in the four years from the eve of the pandemic. The last Free Lunch attempted to understand this US productivity mini-miracle by breaking it down into how individual industry groups have evolved. With no similar overall productivity growth in the EU, there may be no story for a similar decomposition to illuminate. But it’s worth doing anyway since overall stagnation could still hide interesting differences between sectors.

And one thing that is immediately clear is that those sectoral variations — chart below — are very different from the US.

In both countries, the sector with the strongest productivity growth was the information industry, which includes telecoms and data processing. That’s no surprise given the leap in digitisation and remote working technology that the pandemic necessitated. But look at European manufacturing, which leapt ahead, producing 5.8 per cent more per worker towards the end of 2023 than four years earlier, despite the traumatic energy crisis of 2022. In the US, remember, manufacturing was a drag on overall productivity growth. As I have written before, don’t be too quick to cry for European industry.

Another contrast is that professional and business services recorded much lower productivity growth in the EU than in the US, although it did pull up the aggregate number in Europe too.

The most talked-about sectors during the lockdowns and reopenings — wholesale and retail trade, hospitality, and transport and warehousing, all of which EU statistics lump together in one big industry group — behaved rather similarly on both sides of the Atlantic, with poor productivity growth even as they lost employment share to other sectors.

Efficiency changes within a sector only move the needle for overall productivity, however, if the sector itself is sizeable. In the chart below, I account for the relative sizes of the EU economy’s industrial groups to show their absolute contribution to the evolution of the bloc’s productivity over the four years to the third quarter of last year (measured as value added per worker, to make it comparable with last week’s exercise for the US).

As suspected, there is not much of a story — not, in particular, a story of any one sector’s big productivity contributions being erased by devastating productivity losses in other sectors. All the contributions are small; remove any sector, and it wouldn’t change the big picture. Without the drag of the construction sector, productivity growth would double, but still only to 1.2 per cent over four years. Take away professional services, and value added per worker would have risen 0.2 percentage points less than it did; a tenth of the contribution the sector made in the US.

The within-sector rates of productivity growth, in fact, add up to so little that they together barely match the very small contribution to productivity growth from cross-sector reallocation. That contribution is about the same as in the US, just a quarter of a percentage point over the four years in question. One idiosyncratic feature of the European evolution, however, is agriculture (such a small employer in the US that we didn’t even include it last week). It is a very unproductive sector, so the shrinking of its employment share by 0.3 percentage points since before the pandemic made a positive difference to overall productivity. This is worth keeping in mind as we contemplate the farmers’ protests around the continent.

The other idiosyncratic European story here is that of manufacturing. It has shed labour (measured both in terms of workers and even more in terms of hours worked), which is bad for overall productivity because manufacturing is more productive than the economy on average. Yet it has expanded its within-industry value added more than its US counterpart, and this contributed 0.9 percentage points to overall output per worker (that is, more than the total). If I was asked to interpret those numbers, I would say they seem to reveal a sector that — behind the jeremiads about energy costs, Chinese competition and Washington’s Inflation Reduction Act — has made use of those headwinds to ruthlessly cut waste and rationalise its operations. It is not all bad news.

• This week the EU’s new rules on big “gatekeeper” internet companies come into force. Our reporters have detected scepticism about whether they will live up to their promise. At the Centre for European Reform, Zach Meyers has a useful guide to the factors that will determine the Digital Market Act’s success or failure.

• China’s new economic programme targets 5 per cent growth, opens up for modest stimulus and boosts defence and research spending. It’s a far cry from what the Chinese economy needs.

• More brainstorming is coming forth around the idea of lending Ukraine money against a claim on future reparations payment from Russia — with a view to seizing Russia’s blocked assets if (when) those payments fail to materialise. I first outlined such an idea last year. Recently, a related proposal for a “limited recourse” syndicated loan has been made by Lee Buchheit, Hugo Dixon and Daleep Singh. These ideas seem to have caught the interest of UK foreign secretary Lord David Cameron, who mentioned them positively in the House of Lords this week.

• Mining companies are asking the London Metal Exchange to separate contracts for nickel produced through low- and high-carbon production methods, which could lead to a “green nickel premium”. It’s another sign of rising pressure to prevent the undercutting of industrial carbon transitions, exemplified by the EU’s carbon border adjustment mechanism.

• Donald Trump won big on the Super Tuesday of US state primaries (FT results page here) — but as Peter Spiegel points out, far less big than what the polls had predicted. This strengthens my worry that polling data is no longer terribly informative in US elections.

  

• Bearing that caveat in mind, here are some other interesting poll findings. The New York Times reports how more people say Trump’s policies have benefited them economically than Joe Biden’s. And the polls showing Americans are worried about Biden’s age are not all that they seem, writes Patti Waldmeir.

• The European Central Bank meets today to decide monetary policy.