This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies.Find out more here


To discuss how you and your organisation can get more involved with The Work Foundation, please contact us.

Call 020 7976 3575 or email


Professor Geraint Johnes
Professor of Economics at Lancaster University
T 020 7976 3516
Professor Geraint Johnes

The productivity puzzle - real or imagined?

Authors: Professor Geraint Johnes

28 May 2014

The economic history of recent years has been quite remarkable in many dimensions. One of the most notable features has been the sharp decline and subsequent stagnation of labour productivity. Between 2007 and 2009, output per worker fell by over 5%. Over the subsequent four years, it rose by just 1%. Since, owing to advances in technology, productivity usually grows year on year, the full effect of this is that productivity is now almost 20% below the long run trend.

These statistics make the recent recession very different from many other recessions, particularly those of the 1980s and 1990s, after which productivity recovered quite quickly. The initial decline in productivity is likely the result of a fall in demand; but one would normally expect firms to adjust their labour input in line with this fall, thereby raising productivity, quite quickly. The prolonged fall in productivity is thus something of a puzzle. It is not easy to imagine why we should have stopped being as good at doing things as we once were.

This productivity puzzle was the theme of a recent roundtable held at The Work Foundation. The context was set in presentations by Jonathan Haskel and Charles Levy, with subsequent discussion involving representatives from the Treasury, the Department for Business, Innovation and Skills, the Office of National Statistics, the Bank of England, and several other key agencies. While it is fair to say that (unsurprisingly enough) the roundtable did not solve the puzzle, it succeeded in identifying several issues that are likely to be key to a solution, and also identified a promising and important agenda for further research.

Several partial candidate explanations for the apparent slowdown in productivity were identified. Amongst these were:

  • Changes in the composition of industry, notably the decline of capital intensive, high productivity, extractive industries such as oil. This may be a partial explanation, but we know that productivity has declined within a wide range of sectors, not just in some industries.
  • A decline in successful innovation, particularly in industries such as pharmaceuticals. This may contribute to falling productivity in some industries, but it cannot explain the drop in productivity in industries that are less sensitive to innovation.
  • Change in the composition of investment, with much investment now taking the form of R&D, design, brand management, reputation, and software development. Over the last 20 years, such ‘intangibles’ have come to represent a far greater proportion than before of the economy’s investment – up from under 10% to some 35%. When firms invest in intangibles, they seem to be spending resource on workers’ pay without getting a return – the intangibles are not (by and large) treated as an investment in the national accounts, and so national output is mismeasured. This would suggest that GDP is, in reality, higher than official figures suggest, and that productivity has not really fallen (or at least not by as much as the published data would lead us to believe). However, investment represents a relatively small portion of GDP, and the increase in (unmeasured) investment would need to be far larger than is currently the case if it were to explain more than a small part of the apparent decline in productivity.
  • A change in the capital:labour ratio brought about by lower real wages. As real wages fell during the recession, firms may have substituted labour for capital, with the hiring of lower productivity workers becoming increasingly worthwhile. Certainly, investment in physical capital has fallen and has failed (so far) to recover – though many observers expect a marked bounce-back this year. To the extent that this has led to a deterioration of the toolkit with which workers can use to undertake their jobs, one would expect it to lead to declining productivity. But whether the fall in the relative price of labour has been sufficient of a trigger to bring this about is moot – especially given that a change in investment needs to be sustained over several years to have a marked effect on the total stock of capital.
  • The hollowing out of the labour market – a process whereby technology has led to the erosion of opportunities for workers with intermediate level skills. Employment gains in recent years have been concentrated in low skill sectors, including retail and hospitality, and this may explain in part the decline in overall productivity aggregates. Yet we know that this cannot be the full story because productivity appears to have declined across the piece, not just in some sectors.
  • Underemployment is a phenomenon that has drawn much recent attention, with many workers now being employed for fewer hours than they would like. If the decline in productivity manifested itself only as a fall in output per worker, this might provide a convincing explanation. However, we have also observed a significant decline in output per hour worked. So underemployment does not appear to be the story.
  • The increase in self-employment. We still know relatively little about the nature of the new cohorts of self-employed workers. If they are concentrated in low productivity occupations, then the rising incidence of this category of work might play a part in explaining the decline in output per unit of labour employed. The self-employed still constitute a relatively small part of the workforce, however, and they are likely therefore to provide only a small part of an explanation. Moreover, the increase in the numbers of self-employed workers has accelerated sharply over the last twelve months, and of course the fall in productivity predates that by several years.
  • The operation of the financial sector. In the years leading up to the recession, banks may have supported the development of firms that are characterised by relatively low productivity. Following the downturn, by the same token, they may have been unable to support firms whose productivity profile is better. In this respect, we may now have inherited a stock of firms that is suboptimal.

This last point, in particular, has generated considerable discussion – given the role played by the financial sector in supporting businesses in all industries, it is a candidate explanation that is likely to have affected firms right across the economy, and it does not rely on an heroic assessment of the impact on the capital stock of a fairly small change in the relative price of labour. We know, from work done at the Bank of England (see especially Chart 18) that most of the change in productivity over the last decade or so is due to change that has happened within firms – not between firms in different sectors. So anything that can offer an explanation for declining output per worker (or per hour) that is not sector-specific has particular appeal as a prospective explanation.

In light of the above, the roundtable agreed that an appropriate programme for further research in this area might profitably focus upon:

  • Establishing a better understanding of the nature of the new cohorts of self-employed workers
  • Evaluating the changing nature of investment in the modern economy, disaggregated into land, buildings, physical capital, intangibles etc.
  • Exploring the nexus of links between the banking system, capital markets, ability to borrow, and new firm formation; assessing the potential for new models of banking to deliver a stock of employers with optimal characteristics
  • Evaluating the role played by uncertainty in the low rate of investment (and hence low productivity).



Comments in Chronological Order (Total 1 Comments)

Marina Savkovic

28 Jul 2014 10:25AM

Dear professor,

Thank you for sharing this post. I have learned so much from it.

Looking forward to reading your articles on this and similar topics.

Best regards,

Marina Savkovic