Gastbeitrag: Productivity and the crisis: Revisiting the fundamentals

by Dirk Elsner on 15. Juli 2009

by David BrackfieldJoaquim Oliveira Martins

Most narratives of the crisis start with problems in the financial sector that then spilled over into the real economy. This column looks at the real side first and shows that labour productivity growth declined significantly in the years prior to the crisis, particularly in the US construction sector. Financial markets may have failed in that they didn’t detect the deterioration of structural productivity trends in the early 2000s.

Discussions about the current crisis often present events in a sequence, such as that the US sub-prime crisis in August 2007 triggered a major inter-bank credit crisis, which transformed itself into a general credit crisis, the latter having an impact on the real economy and overall business and consumer confidence. However, it is interesting to investigate what was actually happening to the real sector, especially productivity trends, in the major OECD member economies before the crisis. An issue with productivity data, notably multi-factor productivity (MFP), which is a measure of technical progress, is that these data are available with some delay (typically a one- to two- year lag). Nevertheless the data we present below make it clear that labour productivity growth was already slowing down well before the crisis. In particular, the US construction sector, underpinned by loose credit arrangements underpinning, had displayed dismal and worsening productivity performance since 2002.

The vanishing productivity gap

Figure 1 shows annual labour productivity growth in the US, Europe, and Japan since 1997. The US has historically maintained a higher productivity growth rate than Europe, which has usually been attributed to labour conditions (hiring and firing), product market regulation, and the uptake of technology. A striking feature of the graph is the downward convergence to low growth by 2007, which could be dubbed “a race to the bottom”. . It is important to note that this convergence happened before the start of the crisis –the US experienced a labour productivity growth decline since 2004.

Figure 1. EU, US, and Japanese productivity growth

Source: OECD ULC Database

Which sectors are responsible for the slowdown of labour productivity in the US?

Figure 2 breaks aggregate labour productivity into three broad sectors: industry, construction, and market services. While the picture for the industry sector is somewhat volatile, growth rates since 2005 are certainly lower than in the past ten years. The market services sector has also seen a long-term fall in the trend of labour productivity growth since its 1998 peak. However, an important driver in the fall in labour productivity growth for the total economy is construction. The construction sector is around 5% of GDP, but its poor productivity performance has been a significant drag on aggregate productivity. The latter culminates in 2007, with an abysmal -12% labour productivity growth in construction accounting for -0.6 percentage points of aggregate productivity growth. Importantly, these trends started between two to four years before the crisis.

Figure 2. US productivity growth by sector

Source: OECD ULC Database

Price and financial signals were disconnected from productivity fundamentals

Figure 3 shows the disconnection between the productivity trends and the price and financing conditions in the construction and housing sectors. The construction of new dwellings in the US remained high through early 2006; afterwards it started to decline steadily. In parallel, the negative labour productivity growth in the construction sector, after a short improvement in 2005, started again to decelerate strongly during 2006 and 2007. This suggests a serious excess supply problem. In contrast, the housing price bubble was still inflating in 2005; it levelled off during 2006 but only started to decline in the beginning of 2007. It is striking that the boom-years of the mortgage and subprime business (2005-2006) coincides precisely with the periods where the productivity performance was deteriorating rapidly. As if a substantial boost in demand through extended credit conditions could have compensated for the supply-side problems. But the real side of the economy took finally its toll. By mid-2007, it was obvious that the construction sector was unsustainable.

Figure 3. Trends in construction and housing

Source: S&P, Bank of England, OECD MEI, and OECD ULC databases

Stable but low productivity performance continues in the euro area

In the euro area, the first noticeable point about labour productivity growth, with the exception of the industry sector, is its stability and low level. Aggregate labour productivity growth has remained in a one-point range for at least the last ten years. In contrast, the industry sector has fluctuated around the higher value of 3% annual growth. It is worth noting that labour productivity growth in the construction sector has been negative for virtually the whole of the last ten years. Market services had slowly improved their productivity performance since the early 2000s before declining again in 2007.

Figure 4. European productivity trends

Source: OECD ULC Database

Multi-factor productivity had deteriorated in major OECD countries

One of the major drivers of labour productivity is the pace of technical progress, usually measured by multi-factor productivity (see OECD, 2008). Figure 5 presents data from the OECD Productivity Database. MFP trends in the major OECD countries provide additional evidence of the deterioration of the real economy performance in the pre-crisis years for the US and Japan. The picture is less clear and fewer data is available for the Euro area, but in 2007 the low MFP performance is also apparent in France and Germany.

Figure 5. Multi-factor productivity growth

Source: OECD Productivity Database


This brief analysis suggests the financial crisis is not wholly responsible for the deterioration in the real economy. In some sense, financial markets did not detect the deterioration of structural productivity trends in the early 2000s early enough. The latter continued to fuel asset price inflation up to a point where the disconnection became apparent. Economic historians will revisit these numbers in the future, but at least one lesson should be learned – it is important to have timely and reliable information about productivity. Better and timelier statistics may not prevent crises, but they could help agents build more consistent expectations about the future.


OECD Compendium of Productivity Indicators (2008), Paris.

OECD System of Unit Labour Cost and Related Indicators Database. Available at:

OECD Productivity Database. Available at:

Beitrag veröffentlicht auf, übernommen für den Blick Log unter den Copyright-Bedingungen von

This article may be reproduced with appropriate attribution. See Copyright by

Comments on this entry are closed.

Previous post:

Next post: