4 minute read

By Pierre-Alain Pionnier (Pierre-Alain.PIONNIER@oecd.org), Belén Zinni (Belen.ZINNI@oecd.org) and Nhung Luu (Nathalie.Nhung.LUU@oecd.org), Statistics and Data Directorate (OECD)

Originally written for VOX EU

When the COVID-19 crisis hit the world economy in 2020, concerns arose that it would further dampen productivity growth, compounding the poor productivity performance experienced by many OECD economies since the mid-2000s. However, the 2023 edition of the OECD Compendium of Productivity Indicators shows that most OECD countries experienced a sizeable increase in aggregate labour productivity growth at the onset of the COVID-19 crisis. These results are scarcely a cause for celebration, with temporary factors playing a significant role in 2020, many of which were already receding by 2021.

Aggregate labour productivity growth masks significant heterogeneity across industries, as lockdowns and travel restrictions put in place to contain the pandemic disproportionately affected some industries. Insights on the industry contributions can be obtained with a shift-share analysis that decomposes productivity growth into three components: a within-industry effect, accounting for labour productivity growth within each industry, and static and dynamic reallocation effects, capturing the impact of labour resources shifting between industries with different labour productivity levels and growth rates, respectively. The combination of the second and the third component forms the overall reallocation effect.

Outside of recessions, industry reallocations play a limited role

Figure 1 compares the contribution of within-industry and overall reallocation effects to aggregate labour productivity growth across OECD countries in 2020 and the period 2010-2019. While the overall reallocation effect tends to play a limited role in normal economic circumstances, its contribution to aggregate labour productivity growth was unusually large in 2020.

Figure 1: Decomposition of labour productivity growth, average annual percentage change

Figure 2 presents a breakdown of the overall reallocation effect into industry contributions for the United States. The 2020 boost in aggregate productivity largely reflected the impact of lockdowns and travel restrictions and the fall in hours worked in a few selected industries, notably Transport, accommodation and personal services, whose productivity is below-average.

Other industries experienced a smaller decline in hours worked in 2020, which had a heterogenous impact on aggregate labour productivity growth. On the one hand, the resilience of economic activity and hours worked in high-productivity industries such as information and communication, or finance and insurance, likely related to the rapid deployment of digitalisation and teleworking, helped to boost aggregate labour productivity growth. On the other hand, the limited decline in hours worked in low-productivity industries, such as agriculture and construction, damped aggregate labour productivity growth.

Early evidence from the few countries for which detailed industry data are available for 2021 suggests that the reallocation effect observed in 2020 was largely related to temporary disruptions brought by the pandemic. In most cases, the reallocation across sectors started to revert to its pre-pandemic level in 2021.

The rapid increase in job quit rates that followed the first lockdowns, a phenomenon often referred to as the “Great Resignation”, was initially thought to reflect a durable change in worker preferences for certain types of jobs. However, recent empirical evidence for the United States (Hobijn, 2022) shows that only a small share of job quitters were actually changing industry of employment or occupation. Additional data for 2021 and beyond will cast some light on the longer-term effects of reallocations induced by the pandemic, although it will not be easy to disantangle those effects from those of the twin transition – the green and digital transformations which are likely to alter durably business models and economic behaviours.

Figure 2: Industry contributions to the overall reallocation effect, United States

Within-industry developments explain most of aggregate labour productivity growth

In recent decades, within-industry developments have accounted for most of aggregate labour productivity growth (Figure 3). Information and Communication Technology (ICT) drove the surge in aggregate labour productivity growth at the turn of the 21^st century, most notably in the United States (Jorgenson et al., 2008; Cette et al., 2016). This included productivity gains within ICT-producing sectors, as well as the increased use of ICT by some services sectors. Similarly, the slower aggregate labour productivity growth following the ICT-related productivity boom mainly reflected within-industry productivity developments.

Figure 3: Industry contributions to the within-industry effect, United States

In 2020, the within-industry contribution to aggregate labour productivity growth declined further in most OECD countries. In general, Transport, accommodation and personal services contributed negatively, while trade, ICT services, finance and insurance, and other business services contributed positively. Nevertheless, earlier evidence in a number of countries shows that the within-industry contribution to labour productivity growth usually had opposite sign in 2021. This, as a consequence, will bring back the average contribution over 2019-2021 closer to the value recorded for 2010-2019. The United States, where this contribution was both higher in 2020 than over 2010-2019, and even higher in 2021 than in 2020, stands out as an exception among OECD countries. Macro-economic data, such as the one used in the 2023 Compendium of Productivity Indicators, may not fully capture the unusual evolutions of within-industry contributions to labour productivity growth during the COVID-19 pandemic. This data is also subject to revisions, especially in the years 2020 and 2021 when collecting information is challenging due to the crisis.  Firm-level analysis can complement macro-economic analysis, unveiling heterogenous impact of the COVID-19 on firm performance. This, however, goes beyond the scope of this work.

References and further reading