Global economic growth remains solid and broad-based (OECD, 2017b), even though the pace has eased in recent periods. But while the upturn is set to persist into 2018, it has been modest, partly reflecting continued relatively weak labour productivity growth in most countries (Figure 1.1). In the OECD as a whole labour productivity in the post crisis period has grown at about half the rate of the pre-crisis period.
OECD Compendium of Productivity Indicators 2018
OECD Compendium of Productivity Indicators
Chapter 1. Recent trends in productivity and employment
Economic growth is picking up but labour productivity growth remains weak
Figure 1.1. Labour productivity growth before and after the crisis
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, February 2018.
Productivity growth slowed markedly after the crisis in all major sectors, but particularly in manufacturing where productivity growth rates remain well below pre-crisis levels in most countries (Figure 1.2), such as the Czech Republic, Finland, Hungary, Korea, Latvia, Sweden and the United States. Contributions from the business sector services also declined in the post-crisis period, especially in Estonia, Hungary, Greece and the United Kingdom (Figure 1.3).1
Figure 1.2. Labour productivity growth in manufacturing and business services
Note: For Japan, Korea and the United States, labour productivity is measured as gross value added per person employed, as national accounts estimates of total hours worked by main economic activity are not currently available.
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, February 2018.
Figure 1.3. Contributions of manufacturing and business services to labour productivity growth in the business economy
Note: For Japan, Korea and the United States, labour productivity is measured as gross value added per person employed, as national accounts estimates of total hours worked by main economic activity are not currently available.
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, February 2018.
Capital deepening has also been weak
Although there are some signs that investment may be beginning to pick up (OECD 2017b), the recovery remains modest, with capital deepening, i.e. increases in capital per hour worked, from both ICT and non-ICT capital, stalling in many countries in the post-crisis period (up to 2016), compounding the longer term slowdown in productivity growth seen before the crisis in many countries. Indeed, the contribution of capital deepening was negative in Italy and Switzerland between 2014 and 2016, and in Portugal and Spain between 2014 and 2015 (Figure 1.4).
Slower capital deepening rates in part reflect higher employment, but are also in line with lower investment rates, especially in tangible assets, i.e. dwellings, non-residential construction, machinery and equipment and cultivated assets, which in most countries showed only a marginal improvement on the crisis lows and remain below pre-crisis rates (Figure 1.5). However, investment in intellectual property products, where the benefits of the investment may not accrue immediately and are often lagged, has remained more robust, marginally increasing in most countries, which may act as a catalyst for stronger economic growth going forward.
Multifactor productivity growth has also remained weak in many countries
Weak labour productivity growth also reflects continued slowing in multifactor productivity growth (MFP) (Figure 1.4; see Chapter 7. for methodology), notably in Austria, Belgium, the United Kingdom and the United States, lending some weight to the arguments that:
technological spillovers and diffusions from ICT and other new technologies may be lower than from earlier technology breakthroughs (Cowen, 2011; Gordon, 2012, 2016); and
the pace of technological progress requires parallel innovation in organisational structures and business models and, as such, the deployment of new technologies is only getting under way (Brynjolfsson and McAfee, 2011; Baily, Manyika and Gupta, 2013).
Other factors may also play a role. Declining MFP growth rates may at least in part reveal the financial crisis after-effects (OECD, 2015) and inefficiencies in the combined utilisation of labour and capital inputs, notably skills mismatches (Adalet McGowan and Andrews, 2015) but also capital misallocation (Adalet McGowan et al., 2017) or a slower pace at which innovations spread throughout the economy from frontier firms to other firms (Andrews et al., 2015).
Figure 1.4. Contributions of capital deepening and MFP to labour productivity growth
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, February 2018.
Figure 1.5. Investment rates, total economy
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, February 2018.
Employment growth has been a key driver of economic growth in many countries
Among advanced economies, in particular, the United Kingdom and the United States, and more recently in Mexico, Spain and Italy, the recovery in GDP growth has been largely sustained by increasing employment gains, making up for limited labour productivity growth (Figure 1.6).
Figure 1.6. Growth in GDP, labour productivity, average hours worked and employment
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, February 2018.
Employment gains have been mainly in low productivity activities
A concern that has emerged in recent years is that many jobs are being created in lower labour productivity activities, dragging down overall labour productivity (Figure 1.7), partly reflecting the on-going decline in manufacturing’s share of overall employment (Table 1.1). This also reflects, in part, the outsourcing of some service activities from manufacturing firms to (more specialised) service firms and the fact that more low-skilled jobs were destroyed during the crisis.
Figure 1.7. Contributions to employment growth by main economic activity, 2010-2016
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, February 2018.
Table 1.1. Employment in manufacturing, persons, percentage of total
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, April 2018.
Across major economies, on average, between 2010 and 2016, increases in employment in activities with below average labour productivity were about two to four times higher than in those with above average labour productivity (Figure 1.8). In the United States, for example, 9.7 million more jobs in activities with below average labour productivity levels existed in 2016 compared to 2010, over four times the additional 2.4 million jobs in above average labour productivity activities. Comparable figures in other major economies were: 0.5 million and 0.2 million in Canada, 0.4 million and 0.2 million in France, 1.5 million and 0.6 million in Germany; minus 0.02 million and minus 0.2 million in Italy; and 1.9 and 0.6 million in the United Kingdom.
In nearly all major OECD economies the top three sectors generating the largest net employment gains over the period 2010 to 2016 had below average labour productivity, with restaurants, health and residential care activities featuring highly in most economies (Table 1.2). Only France saw a sector, namely legal, accountancy, management consultancy, with above average labour productivity in the top three sectors. On the other hand, in sectors that had lost most (net) jobs over the same period, most major economies had at least one above average labour productivity sector in the top three; all three in the case of the United States (Table 1.3).
Insights on the contribution of these reallocation and structural effects on overall labour productivity growth can be made using a standard shift-share analysis that decomposes productivity growth into three components: (i) a component that accounts for the within-industry effect: i.e. the sum of industry productivity growth rates, weighted by the initial employment shares, in other words a measure of productivity growth unaffected by actual reallocation effects; (ii) a static structural component that captures the overall effect of changes in employment from low to high labour productivity sectors (or vice versa); and (iii) a dynamic component that captures the overall effect of changes in the shares of labour shifting to higher or lower productivity growth sectors (Box 1.1).
While the impact of the crisis and the different response mechanisms adopted by countries to mitigate its effects need some care in interpretation, the evidence points to relatively minimal boosts to productivity growth from reallocation effects in most countries2 (Figure 1.9) with declining within-sector productivity growth mirroring declines at the whole economy level. However, in the United Kingdom reallocation effects, which were strong drivers of overall productivity growth in the pre-crisis period, and helped mask a precipitous decline in within-industry productivity growth in the five years leading up to the crisis, petered out in the post-crisis period.
Figure 1.8. Labour productivity in 2010 and employment growth over 2010-2016 (or latest available year)
Note: Quadrant I (II) corresponds to sectors with below (above) average – whole economy – labour productivity levels in 2010 and above average employment growth between 2010 and 2016. Quadrant III (IV) corresponds to sectors with above (below) average – whole economy – labour productivity levels in 2010 and below average employment growth between 2010 and 2016. The size of the bubbles increases with the absolute change in employment on a person’s count basis in the different sectors. Bubbles in dark blue denote employment creation; bubbles in grey denote employment destruction. Measurement of value added, and hence labour productivity, in the primary sector, real estate services, public administration, health, education and other non-market activities is subject to a number of conceptual and measurement issues (Cho et al., 2017; OECD, 2017a; Schwellnus et al., 2017).
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, April 2018.
Table 1.2. Net employment creation between 2010 and 2016 (or latest available year)
Three sectors with largest net employment creation, selected countries, thousands of persons
Note: Average labour productivity is measured as gross value added per person employed in the total economy. Data for Canada refer to the period 2010-2013 and shows thousands of jobs created over that period. Data for France, Germany and Italy refer to 2010-2015.
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, April 2018.
Table 1.3. Net employment destruction between 2010 and 2016 (or latest available year)
Three sectors with largest net employment destruction, selected countries, thousands of persons
Note: Average labour productivity is measured as gross value added per person employed in the total economy. Data for Canada refer to the period 2010-2013 and shows thousands of jobs created over that period. Data for France, Germany and Italy refer to 2010-2015.
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, April 2018.
Figure 1.9. Shift-share analysis of labour productivity growth
Note: The shift-share analysis is based on information of all one-digit sectors in line with ISIC Rev 4 industrial classification. Labour productivity is measured as gross value added per hour worked in France, Germany, Italy and the United Kingdom, and gross value added per person employed in Japan and the United States.
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, February 2018.
Box 1.1. Decomposing aggregate labour productivity growth into intra-branch effects and structural changes: the shift-share analysis
Changes in an economy’s labour productivity can be due to either changes in labour productivity of the individual industries or structural shifts in resources between contracting and expanding industries. The shift-share analysis is the most commonly used method to analyse these effects, by decomposing aggregate labour productivity growth into the contributions of within-sector productivity gains and structural changes caused by resource reallocation among sectors. In other words, the shift-share analysis assess to which extent changes in aggregate labour productivity growth rates emanate from higher or lower labour productivity growth in the specific industries and shifts in employment to industries with lower or higher labour productivity levels and growth rates.
Under this approach, labour productivity in the total economy is expressed as the sum of labour productivity levels of each sector weighted by the sectoral employment shares, as follows:
where , and represent respectively labour productivity, output and employment in the total economy in period t, , and represent respectively labour productivity, output and employment in sector ( in period t, and represents the employment share of sector in the total economy in period t.
In a discrete time perspective, the difference in aggregate labour productivity levels at time 0 and time T can be written as follows:
Dividing both sides by , it follows that aggregate labour productivity growth can be decomposed into intra-branch labour productivity growth, represented by the first term in the right-hand side of the equation, and the effects of structural change in the economy, which consist of a static effect, represented by the second term, and the dynamic effect, represented by the third term.
Within-sector productivity growth effect: also known as intra-branch productivity effect, it captures the effect of productivity growth within the different industries in the absence of structural change, this is, on the assumption that there are no changes in employment shares in specific industries (i.e. the structure of the economy is assumed to remain fixed).
Static shift effect: also known as static shift effect or static structural change, it measures the contribution to aggregate labour productivity growth of a shift of employment resources towards sectors or branches with lower or higher labour productivity levels at the beginning of the period.
Dynamic shift effect: often referred as the interaction effect or the dynamic component of the structural change, it measures the interaction of changes in labour productivity and employment across sectors, this is, to measure the extent to which positive/negative efficiency gains interact with the expansion/contraction of different industries. This term is positive if sectors with higher (lower) productivity growth increase (reduce) their share in total employment, and negative when expanding sectors have below average labour productivity growth or if sectors with higher productivity growth have declining shares in total employment.
The sum of the static and dynamic shift effects is often used a measure of the overall resource reallocation process in the economy.
And, in turn, most growth has been in lower paid jobs
Labour compensation levels correlate highly with labour productivity levels, and, so, more jobs in lower labour productivity activities has also meant more jobs with below average wages in most economies, working to weigh down on average salaries in the economy as a whole. Between 2010 and 2016, for example, close to 90% of all new jobs in France were created in activities with below average wages; close to two-thirds in Germany and the United Kingdom and over three-quarters in the United States (Table 1.4).
Table 1.4. Change in employment over the period 2010-2016, or latest available year
Thousands of persons
Note: Data for Canada refer to 2010-2013. Data for France, Germany and Italy refer to 2010-2015. At the time of writing this publication, data for the period 2010-2016 in Italy are available only at the whole economy level and show net employment creation equal to 56 thousand of persons.
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, and OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, February 2018.
Growth in real wages, adjusted for inflation using the consumer price index, has also been lagging labour productivity growth in many countries (Figure 1.10) (OECD, 2017a; Schwellnus et al., 2017). Indeed real labour compensation per hour worked, adjusted for the CPI (which provides for a better measure of real purchasing power from a household perspective, compared to the GDP deflator), declined between 2010 and 2016 in Portugal, Spain and the United Kingdom. However, in some countries, such as Germany and the United States, real labour compensation has begun to rise in line with (albeit slow) labour productivity growth in recent years, helping to reverse pre-crisis decoupling (Figure 5.5).
Figure 1.10. Labour productivity and average labour compensation per hour, total economy, 2010-2016
Source: OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en, and OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, February 2018.
The decoupling of wages from labour productivity growth at whole economy level however can mask divergences within sectors (Figure 1.11). In France, for example, out of 63 sectors, according to the ISIC Rev. 4 classification, 41 saw growth in average real labour compensation lagging labour productivity growth in the post-crisis period; with the largest decoupling occurring in water transport services and telecommunication services. Similarly in the United Kingdom and Italy, just over half of all sectors saw real average labour compensation grow at a slower pace than labour productivity, with the largest gaps in the fishing and aquaculture and education sectors in the United Kingdom and basic metals and non-metallic mineral products in Italy. In the United States and Germany however, mirroring the improvement seen at the whole economy level, most sectors saw real wage growth outpacing productivity growth (40 out of 58 in the United States and 37 out of 63 in Germany). Across all major economies, sectors that saw the highest net increase in employment gains also saw wages outpacing, or keeping pace with, labour productivity growth.
Figure 1.11. Growth in labour productivity and average labour compensation since 2010
Note: The three sectors with highest employment change over the period 2010-2015 are denoted in black. ISIC Rev 4 code descriptions can be consulted here: https://unstats.un.org/unsd/publication/SeriesM/seriesm_4rev4e.pdf.
Source: OECD National Accounts Statistics (database), http://dx.doi.org/10.1787/na-data-en, April 2018.
Granularity matters
For ease of exposition past editions of the OECD Compendium of Productivity Indicators presented information at a relatively aggregated sectoral level, which can however mask developments at a more detailed level. The analysis presented above shows that more granularity helps to better understand the underlying drivers underpinning whole economy productivity growth, in particular with regards to structural changes and their impact on material well-being and inclusive growth. Indeed, employment growth has been an important driver of overall economic growth in most OECD economies in the post crisis period, but a more granular analysis reveals that in nearly all major economies most net job creation has occurred in activities with below average labour productivity and below average wages.
Future editions of the Compendium will work to explore granularity in more detail, capitalising on the OECD Structural and Demographic Business Statistics database that provides sectoral data at a 4-digit level of detail.
References
Adalet McGowan, M., D. Andrews and V. Millot (2017), "The Walking Dead?: Zombie Firms and Productivity Performance in OECD Countries", OECD Economics Department Working Papers, No. 1372, OECD Publishing, Paris, http://dx.doi.org/10.1787/180d80ad-en
Adalet McGowan, M. and D. Andrews (2015), “Labour Market Mismatch and Labour Productivity: Evidence from PIAAC Data”, OECD Economics Department Working Papers, OECD Publishing, Paris, No. 1209, http://dx.doi.org/10.1787/5js1pzx1r2kb-en
Andrews, D., C. Criscuolo and P.N. Gal (2015), “Frontier Firms, Technology Diffusion and Public Policy: Micro Evidence from OECD Countries”, OECD Productivity Working Papers, No. 2, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jrql2q2jj7b-en
Brynjolfsson, E. and A. McAfee (2011), Race Against the Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy, Digital Frontier Press.
Baily, M. N., J. Manyika and S. Gupta (2013), “U.S. Productivity Growth: An Optimistic Perspective”, International Productivity Monitor, No. 25, Spring.
Cho, T., S. Hwang and P. Schreyer (2017), "Has the Labour Share Declined?: It Depends", OECD Statistics Working Papers, No. 2017/01, OECD Publishing, Paris, http://dx.doi.org/10.1787/2dcfc715-en
Cowen, T. (2011), The Great Stagnation: How America Ate All The Low-Hanging Fruit of Modern History, Got Sick, and Will (Eventually) Feel Better, Dutton, June 2011.
Gordon, R. (2012), “Is US Economic Growth Over? Faltering Innovation Confronts the Six Headwinds”, NBER Working Papers, No. 18315.
Gordon, R. (2016), The Rise and Fall of American Growth: The U.S. Standard of Living since the Civil War, Princeton University Press.
OECD (2015), The Future of Productivity, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264248533-en
OECD (2017a), OECD Compendium of Productivity Indicators 2017, OECD Publishing, Paris, http://dx.doi.org/10.1787/pdtvy-2017-en
OECD (2017b), OECD Economic Outlook, Volume 2017 Issue 2, OECD Publishing, Paris, http://dx.doi.org/10.1787/eco_outlook-v2017-2-en
Schwellnus, C., A. Kappeler and P. Pionnier (2017), "Decoupling of wages from productivity: Macro-level facts", OECD Economics Department Working Papers, No. 1373, OECD Publishing, Paris, http://dx.doi.org/10.1787/d4764493-en
Notes
← 1. In this chapter, the statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities or third parties. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.
← 2. However it should be noted that the industry level decomposition used here (at the one-digit level) may mask reallocation effects within one-digit sectors.