TFP = ‘Mix & Methods’, not ‘Magic Fairy Dust’

Total factor productivity (TFP) has become that part of productivity growth that can’t be measured by changes in the other inputs. In a way, that makes it the stardust factor. It includes innovation, entrepreneurialism, je ne sais quoi, whatever. Everyone knows it is important, but no one knows precisely what it is.

Despite that, in 1957 Robert Solow, another Nobel prize winner, reckoned that about 80% of the growth in labour’s productivity came from this residual factor. And Baily has been able to tease out which parts of the US economy contributed most to its growth in total factor productivity (TFP) for the period 1987 to 2019. Overall, he suggests that America’s total factor growth was 0.75% a year and that almost all of it, about 85% came from just three industrial sectors – manufacturing, which accounted for nearly half the growth, retailing and wholesaling.

Digging down further, Baily finds that almost all of manufacturing’s increase in TFP came from the sub-sector for computers and electronics. While manufacturing TFP grew by 1.38% a year, the computers sub-sector grew by 1.35% – “a remarkable finding”, he says.

He cites the super-productivity of the Japanese auto industry which, in the 1990s, left its US equivalent, and even Germany’s, far behind. In particular, there is the familiar example of Toyota, whose productivity was enhanced by its work culture, the best-known elements of which were its practice of continuous improvement and its ‘keiretsu’. Continuous improvement did what it said by making incremental changes to designs and production processes, often based on suggestions from the shop floor. The keiretsu was the close relationship between Toyota and its web of suppliers, all of which were supposed to be working towards the goals of cutting costs and raising product quality.

The contrast between Japan’s best practices and the methods of the US auto industry was stark. To generalise, US carmakers maintained distance between themselves and their suppliers while usually having more than one company supply the same component. That put carmakers in a strong position to cut buying-in costs, but the squeeze it exerted on suppliers’ profits deterred investment and, in time, drove suppliers either out of business or into low-wage parts of the world.

Meanwhile, Japan’s productivity decline has been startling. Arguably, this is better illustrated by Table 1 than the chart. Its marvellous growth rates of the 1950s and 1960s were built on the combination of an economy running on newly built post-war infrastructure and a wonderful demographic dividend. Given that, productivity growth was always going to fade. However, in the 2010s growth was almost non-existent. In seven of the 14 years 2006-19 Japan’s productivity actually fell year on year. For a nation that still has so much going for it, one wonders, how did Japan actually manage that?

The other pertinent question is whether the UK is going in the same direction; granted, the two economies have important structural differences, in particular that too much of Japanese industry is sheltered from competition. But if so, Table 2 makes for miserable reading. The table juxtaposes Japan’s average annual productivity growth per decade against average changes in the Nikkei 225 index of Japanese equities. The volatile – but mostly falling – 1990s and 2000s are a reminder of what happens when investors fall deeply out of love with over-hyped markets. Even the bounce-back in the 2010s is partial – at its current 27,600, the Nikkei index is still 30 per cent below the level at which it peaked 33 years ago.

For long-term investors wanting to sidestep such a prospect, the direction should be plain – go where the productivity growth looks brightest.