I return to this theme in the present contribution. This is stimulated by my participation last month in a day-long conference held at the Peterson Institute for International Economics in Washington DC, on the theme "Making sense of the productivity slowdown" 2. The conference brought together leading researchers on productivity from the US but also from other parts of the rich world: Germany, Japan and the UK, as well as comparative analysis across the Organisation for Economic Co-operation and Development (OECD). In addition, two important keynote addresses were delivered by Lawrence Summers of Harvard University and Robert Gordon of Northwest University, each of whom have shaped the US (and global) discourse on this issue. All materials presented at the conference are helpfully made available on the Peterson website.
Two issues dominated the US discussion: The relationship between the financial crisis and the longer-term prospects for productivity growth; and the magnitude of possible mis-measurement of productivity given the shift toward the digital economy. As earlier, I remind readers that the professional literature moves freely and without warning between two linked but separate concepts. These are labour productivity (typically defined as growth in real output per hour in either the total economy, or, more narrowly, the business sector); and total factor productivity or TFP (the decomposition of this growth in labour productivity into improved labour quality; capital deepening; and a residual calculated by inserting the first two into an equation called the production function).
More From This Section
The stylised facts for the US (as presented by John G Fernald) made it clear that there have been two "golden ages" for growth in labour productivity since the Second World War. Real output per hour (in the business sector) grew at somewhat more than three per cent per year in this era, the "high noon" of American manufacturing dominance and steady growth in living standards. This rate halved for the next twenty years (1973-95; paradoxically overlapping with the supply side reforms of Ronald Reagan and George H W Bush) only to rebound under Bill Clinton. Since 2010, this three per cent has dwindled to 0.5 per cent.
Mr Fernald examines the contribution of slower growth in TFP to these labour productivity trends by sector and time. He concludes that the slowdown in TFP growth preceded the financial crisis and has been widespread across the market economy, not just associated with Information Technology or the so-called "bubble sectors" of construction, real estate, financing and mining. His examination of "mis-measurement" concludes that, if anything, better measurement "would make this broad-based TFP slowdown in market goods and services even worse." He concludes that prospects for a recovery of labour productivity are highly uncertain.
This story-line was given further depth and colour by Mr Gordon and other speakers. Mr Gordon examines US TFP through what he sees as three US industrial revolutions since the late 19th century: Roughly 1890-1920; 1920-1970; and 1970 till the present. His celebrated thesis is that the impact of inventions in the "second" industrial revolution of the beginning of the 20th century (electricity; running water; electronics; chemicals; communications; pharmaceuticals) was much more profound and long-lasting than the digital revolution that has followed.
The result has been a pronounced slowing of TFP growth since 1970. According to Mr Gordon's estimates, US annual TFP growth since 1970 (whole economy) at 0.65 per cent is barely a third of what the US enjoyed in the fifty years between 1920 and 1970. Strikingly, Canadian and European measures of labour productivity presented by Mr Gordon show deterioration compared to the US since the mid-1990s, indicating that the U.S. experience is not particularly idiosyncratic.
Space does not permit recounting of the other rich insights gleaned from the conference, but given their potential importance for India's future let me list two. Both work done on the US (by Peter R. Orszag and Jason Furman) and by the OECD for a cohort of its wealthy member countries, demonstrate that in recent years there has been increasing differentiation within sectors as certain "superstar" companies break away from the rest of the pack. In the language of the OECD, there has been a widening of the gap between "frontier companies" (emphasis mine) and the rest of the cohort, leading to the question of whether and why the process of diffusion has broken down. In a similar vein, John van Reenen summarised recent literature that assigns differences in average labour productivity across countries to the efficiency in allocation of labour and capital among firms; this, in turn, is related among other things to the diffusion of superior management practices across the economy.
I come away from this deep dive into advanced country productivity analysis with several reflections for India. First, we should wish for a rebound in advanced-country labour productivity growth. Without the rise in real incomes that this would permit, there will be a political backlash against globalisation, signs of which are already evident in Western democracies. This will harm India. Second, there is sound logic in the government's policies to encourage foreign direct investment by leading multinationals, in both manufacturing and services. But, to obtain leading edge technology, including management practices, a transparent and predictable policy, taxation and competition regime are essential. Fortunately, these are firmly on the government's agenda. Third, analyses of the kind done by Mr Gordon are only possible on the basis of reliable, high-quality factor input data, particularly on labour input. This should be put on the agenda of the Statistical Commission.
1 "What will drive India's growth": Business Standard, July 2015
2 http://www.iie.com/events/event_detail.cfm?EventID=416