OECD Compendium of Productivity Indicators 2023

1. Productivity and the COVID-19 pandemic

Productivity developments in the aftermath of Covid

Physical distancing, lockdowns and restrictive measures put in place worldwide to contain the COVID-19 pandemic triggered the most severe and abrupt global recession since the end of the Second World War. However, the depth of the economic downturn was heterogeneous across OECD countries. This reflects the uneven impact of the pandemic throughout the economy, resulting from differences in economic structure, variations in the nature and timing of restrictions, and differences in fiscal and monetary policy across countries.

Russia’s war of aggression against Ukraine and the energy crisis have derailed the post-pandemic economic recovery and, as of early 2023, a significant growth slowdown is expected for the world economy in 2023. Growth has lost momentum, inflation has hovered at elevated levels, confidence has weakened, and uncertainty is high (OECD, 2022a). As productivity growth in many countries had been declining before the pandemic, the downturn could potentially drive long-term productivity growth into zero or negative territory, lowering living standards (di Mauro and Syverson, 2020).

The economic literature on the impact of the COVID-19 crisis on productivity is inconclusive as several factors are at play (OECD/APO, 2022). On the one hand, recessions are likely to hit primarily less productive firms and result in a reallocation of resources towards more productive firms. On the other hand, permanent losses of capital and jobs can hamper long-term productivity developments. The specificities of the COVID-19 crisis make the assessment even more challenging. It affected both demand and supply, curtailing large areas of activity intermittently over months. The scale of impact was global and combined with strong uncertainties for an extended period, thus holding back corporate investment. Nevertheless, the policy reaction to limit the spread of the virus and cushion the downturn was unprecedented. The acceleration of digitalisation and take up of teleworking also helped to mitigate the depth of the recession.

In the OECD as whole, labour productivity, measured as GDP per hour worked, soared in 2020, growing at a record high rate of almost 4%, much higher than the 1% annual average growth over the period 2010-2019. This reflected a much larger decrease in hours worked than in GDP, particularly during the first half of 2020. Nevertheless, OECD labour productivity growth slowed in the second half of 2020, and then stagnated in 2021.

These aggregate developments mask significant heterogeneity across countries, with for example large swings observed between 2019 and 2021 in Canada, Chile, Colombia, Costa Rica and Türkiye, and overall stability in Denmark, Finland and Germany.

 

Compare

Multifactor productivity (MFP) growth fell in many countries in 2020, as capital inputs did not decline as fast as hours worked, and bounced back in most economies in 2021. These movements led to a stagnation of MFP growth in most countries between 2019 and 2021, adding further concerns to the MFP slowdown observed in many economies prior to the COVID-19 crisis.

 

Compare

Beyond its aggregate impact, the recession affected economic sectors in different ways, both in terms of output and hours worked, reflecting the uneven impact of the restrictions put in place to contain the propagation of the virus. While under normal economic circumstances, within-industry labour productivity growth is the major driver of aggregate labour productivity growth, the lockdowns put in place during the pandemic led to an unprecedented reallocation of hours worked across sectors, with hours worked declined much more in contact-intensive and transport sectors than others. Overall, hours worked declined less in high-productivity than in low-productivity sectors, which boosted aggregate labour productivity growth in 2020. The available evidence for countries in 2021 shows that most of these effects were temporary and got reversed in 2021 (see Chapter 4 on Industry contributions to labour productivity growth).

 

 

In the longer term, the main risk to productivity depends on the impact of the crisis on the productive capacity of the economy. While recessions can accelerate the process of resource reallocation from low-productivity to high-productivity firms, the policy response to the COVID-19 pandemic that prioritised preservation (over reallocation) of resources via job retention schemes may have softened the traditional cleansing (Andrews, et al., 2021). In addition, financial measures put in place by governments to support firms and temporary regulations on insolvency contributed to avoid a wave of bankruptcies in 2020, but have increased private debt and firm vulnerability to shocks (OECD, 2022b), especially in the current environment of rising interest rates. Besides, despite government actions to mitigate the crisis, the pandemic may also influence human capital accumulation. Notwithstanding tight labour markets and shortage of workers across the OECD (OECD, 2022c), the “scarring” effects of the COVID-19 pandemic on workers, in particular women, youth, low-skilled and low-income workers are likely to affect labour force developments over the medium to long term (ILO, 2021a, OECD, 2022c).

On a more positive note, the COVID-19 crisis has accelerated the adoption of digital technologies by firms. Indeed, evidence from business surveys conducted worldwide in the course of 2020 indicated that most firms have intensified their use of digital technologies during the pandemic (OECD, 2023 forthcoming). Nevertheless, smaller firms consistently report that costs, and lack of skills and awareness of digital tools, remain important concerns for investing in digital technologies (OECD, 2021b). The diffusion of digital technologies across firms is key to avoid a further widening of the productivity gap within economies and to ensure a sustained rebound of aggregate productivity after the crisis (Andrews et al. 2016). In addition, the take up of teleworking also helped to mitigate the depth of the recession. Survey evidence points to a positive short-term impact on self-assessed productivity by managers or by employees in OECD economies (Barrero et al., 2021; Criscuolo et al., 2021). However, the evidence on the long-term effects of teleworking on productivity remains scarce.

Monitoring the dynamics and composition of investment, with a focus on assets such as research and development (R&D) and information and communication technologies (ICT), will be key to assess the potential for a rebound in productivity after the crisis. The volume of gross fixed capital formation (GFCF) declined by almost 4% on average across OECD countries in 2020, with large cross-country differences, and rebounded by almost 5% in 2021, recording the highest growth of GFCF in the OECD since 2005. Investment in transport equipment, and to a lower extent other machinery and equipment, was the main contributor to these large swings in aggregate investment in 2020 and 2021, which is both related to the restrictions put in place to contain the pandemic and to the fluctuations in economic activity. Early evidence shows that investment in R&D and ICT has been more resilient and has even increased in 2020 in a few countries. In this respect, the COVID-19 crisis has reinforced the need for more timely data to monitor short-term developments, including of investment and productivity.

Statistical measurement challenges and cross-country comparability issues during the crisis

Measuring non-market output

The COVID-19 pandemic had a significant impact on the provision of education and health services which, in most countries, are largely non-market services. Contrary to market output, non-market output is provided most often for free or at subsidised prices. In the absence of a market, its value is imputed in the national accounts and calculated as the sum of the underlying costs of production. These costs include intermediate consumption, compensation of employees, consumption of fixed capital, and taxes less subsidies on production.

Compensation of employees represents by far the largest share of non-market value added in current prices (around 80% on average across OECD countries). In countries where governments maintained the compensation of their employees in the first phase of the crisis even though they may have worked longer or shorter hours than usual, the value added of non-market activities at current prices (i.e. in nominal terms) remained largely unaffected in the national accounts.

When it comes to the measurement of the volume of non-market output (i.e. in real terms), national statistical agencies tend to use different methods in the absence of observable market prices to deflate output at current prices (Schreyer, 2012; Mitchell et al., 2022). In a nutshell, they choose between four different methods: (1) deflate nominal output using output price indices in comparable market sectors (indirect output method), (2) deflate nominal output using input price indices (indirect input method), (3) measure the volume of inputs using direct input indicators such as the number of employees or the volume of hours worked (direct input method), and (4) measure the volume of output using direct output indicators such as the number of medical treatments (direct output method).

While most OECD countries use direct output indicators to measure the volume of education, all four methods are used to measure the volume of health services. In both cases, different adjustments may also have been implemented to account for COVID-19-related reductions in education output resulting from remote learning, and for the change in the composition and costs of medical treatments during the pandemic.

Countries most affected by the COVID-19 pandemic, as measured by their excess mortality rate, tended to have a larger decline in measured education and healthcare output during the first wave of the pandemic (second quarter of 2020). This decline was more significant in countries using direct output indicators and applying ad-hoc adjustments to measure non-market output during the pandemic (e.g. the United Kingdom), and mitigated in countries measuring the volume of non-market output by deflation or based on direct input indicators, especially if they relied on the number of employees rather than the volume of hours worked to measure inputs (e.g. Spain) (Mitchell et al., 2022).

Nevertheless, the issues related to the measurement of non-market output are probably more severe when relying on the first estimates of quarterly national accounts. Indeed, additional and better data feed into the successive vintages of annual national accounts and contribute to an improved measurement.

Moreover, even though non-market activities make up around 20% of GDP on average across OECD countries, differences in non-market output account for only a small part of cross-country variation in annual GDP growth in 2020 (OECD, 2021a). In other words, differences in measuring non-market output across countries played a much more limited role than other factors (e.g. economic specialisation, restrictions put in place to limit the propagation of the virus, magnitude of the monetary and fiscal impulse) in explaining cross-country differences in overall GDP growth during the COVID-19 crisis.

Measuring labour input

The COVID-19 pandemic occurred at a time when employment rates, i.e. ratios of employment to working age population, were at historic highs across OECD countries (OECD, 2019). According to the OECD Labour Market Statistics, the OECD employment rate declined by 2.7 percentage points, from 68.8% in 2019 to 66.1% in 2020. Nonetheless, the latest estimates for the second quarter of 2022 point to an employment rate in the OECD close to 69.4% - its record high since the start of the series in 2005.

While the total number of persons employed declined by more than 1% in most OECD countries in 2020, the situation varied substantially across countries, with Canada and the United States recording a drop in total employment of 4.2% and 4.3% respectively between 2019 and 2020, and the euro area recording a decline of 1.3%. These cross-country differences are partly related to differences in macroeconomic conditions and labour-market institutions, but they mostly reflect differences in the treatment of workers on temporary lay-off in official labour statistics (Arnaud, 2020).

According to the International Labour Organisation (ILO) guidelines, temporarily laid-off or furloughed workers are persons “employed” who, in their present job, are “not at work” due to economic reasons for a short duration but maintain a job attachment during their absence (ILO, 2013 and 2020). These guidelines treat furloughed workers as being employed when:

• the expected total duration of the absence is up to three months (or more if the return to employment in the same economic unit is guaranteed), or

• workers continue to receive a remuneration from their employer, including partial pay, even if they also receive support from other sources, including government schemes.

In turn, furloughed workers not satisfying the above criteria are classified as either unemployed, if they are actively looking for a job and available for work, or as outside of the labour force (i.e. neither employed nor unemployed).

In practice, departures from these guidelines in national practices do exist. In the United States for example, persons on temporary lay-off are classified as “unemployed” even if they expect to be recalled to their job within six months. “People who were effectively laid off due to pandemic-related closures [are] counted among the unemployed on temporary lay-off” without further testing for their return to their previous job (BLS, 2020). In Canada, persons on temporary lay-off are also classified as “unemployed” even if they have a date of return or an indication that they will be recalled by their employer. On the contrary, in Europe, only those furloughed workers that do not have an assurance to return to work within three months and receive less than half of their salary are treated as unemployed, if they are available to start working in the next two weeks.

In normal times, these differences only have a marginal impact on the international comparability of employment and unemployment measures, as furloughed workers typically account for a relatively insignificant share of the population. However, the confinement measures put in place to contain the spread of the virus significantly increased their number. Prior to the pandemic, furloughed workers represented 0.5% of the labour force in the United States and less than 0.2% in the European Union, but in the second quarter of 2020, these shares jumped to 9.3% and 9.6% respectively (Arnaud, 2020).

It is therefore important to use total hours worked as a measure of labour input, as opposed to job- and head-counts labour input measures, in particular when analysing the short-term impact of the COVID-19 outbreak on productivity. International differences in the treatment of temporarily laid-off workers are likely to be less relevant when labour input is measured as the total number of hours worked, as furloughed workers work by definition zero hours while being in temporary lay-off. National accounts data for 2020 reveal a decline in total hours worked of 6.8% and 8.1% in the European Union and the euro area, respectively, against a drop of 1.4% and 1.5% in total employment.

The effective quantity of labour input depends not only on the total number of hours actually worked but also on the skills and characteristics of the workforce. Indeed, workers with different education levels and experience are not perfect substitutes and can show different productivity performance. Accurate estimates of labour input for productivity analysis requires accounting for both the number of hours worked and the composition of the workforce, that is a measure of composition-adjusted or quality-adjusted labour input. During a recession, the compositional effect of labour input generally points to an increase in the average skill level of those in employment (i.e. an increase in labour quality), as firms are likely to shed labour and/or reduce hours worked among lower-skilled workers, and to hoard higher-skilled individuals.

Measuring investment and capital utilisation

The fluctuations in investment that were recorded in 2020 and 2021 have a direct impact on the measurement of capital stocks and capital services during the pandemic and beyond. Nevertheless, current measures of capital services do not capture the reduced rates of capital utilisation during the crisis. Differences in containment measures and in teleworking capacities across countries resulted in different patterns of capital usage across economic activities. For instance, many offices and commercial buildings were closed, in full or in part, during lockdowns, with much of the capital equipment being unused. This has certainly caused a fall in capital utilisation rates, which has been more or less pronounced across assets and industries. In productivity analysis, it is typically assumed that capital services are a time-invariant proportion of the productive capital stock, an assumption that is difficult to maintain during the COVID-19 crisis, when strict and widespread lockdowns were implemented. While estimating capital utilisation rates is far from trivial (OECD, 2001), future research could be oriented to adjust measured capital services for the reduced rates of capital utilisation to provide a more accurate estimate of capital services and MFP growth during the pandemic.

Data sources

OECD National Accounts Statistics (database), https://doi.org/10.1787/na-data-en.

OECD Labour Market Statistics (database), https://doi.org/10.1787/data-00046-en.

OECD Productivity Statistics (database), http://dx.doi.org/10.1787/pdtvy-data-en.

References and further reading

Andrews D., Criscuolo C. and Gal P.N. (2016), “The Best versus the Rest: The Global productivity Slowdown, Divergence across Firms and the Role of Public Policy”, OECD Productivity Working Papers, 2016-05, OECD Publishing, Paris, https://doi.org/10.1787/63629cc9-en.

Barerro J.M., N. Bloom and J. Davis (2021), “Why Working from Home will Stick”, NBER Working Paper N. 28731, April. http://www.nber.org/papers/w28731.

BLS, Bureau of Labor Statistics (2020), “The impact of the coronavirus (COVID-19) Pandemic on the Employment Situation for March 2020”, U.S. Bureau of Labor Statistics, March 2020, Washington DC. https://www.bls.gov/cps/employment-situation-covid19-faq-march-2020.pdf.

Criscuolo, C., et al. (2021), "The Role of Telework for Productivity during and post-COVID-19: Results from an OECD Survey among Managers and Workers", OECD Productivity Working Papers, No. 31, OECD Publishing, Paris, https://doi.org/10.1787/7fe47de2-en.

Di Mauro F. and C. Syverson (2020), “The COVID Crisis and Productivity Growth”, https://bfi.uchicago.edu/wp-content/uploads/BFI_White-Paper_Syverson_6.2020.pdf.

ILO (2013), “Resolution Concerning Statistics of Work, Employment and Labour Underutilization”, 19th International Conference of Labour Statisticians (ICLS), Geneva. https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/normativeinstrument/wcms_230304.pdf.

ILO (2020), “COVID-19: Guidance for Labour Statistics Data Collection”, International Labor Organisation (ILO), Geneva. https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/publication/wcms_741145.pdf.

ILO (2021a), “World Employment and Social Outlook: Trends 2021”, Geneva, 2021, https://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/---publ/documents/publication/wcms_795453.pdf.

Mitchell, J., et al. (2022), "International comparisons of the measurement of non-market output during the COVID-19 pandemic", OECD Statistics Working Papers, No. 2022/03, OECD Publishing, Paris, https://doi.org/10.1787/301f1306-en.

OECD (2001), “Measuring Productivity - OECD Manual: Measurement of Aggregate and Industry-level Productivity Growth”, OECD Publishing, Paris, https://doi.org/10.1787/9789264194519-en.

OECD (2019), “OECD Compendium of Productivity Indicators 2019”, OECD Publishing, Paris, https://doi.org/10.1787/b2774f97-en.

OECD (2021a), “OECD Economic Outlook”, Volume 2021, Issue 1 (June), OECD Publishing, Paris, https://doi.org/10.1787/edfbca02-en.

OECD (2021b), “OECD SME and Entrepreneurship Outlook”, OECD Publishing, Paris, https://doi.org/10.1787/97a5bbfe-en.

OECD (2022a), OECD Economic Outlook, Volume 2022 Issue 2: Preliminary version, OECD Publishing, Paris, https://doi.org/10.1787/f6da2159-en.

OECD (2022b), "Deteriorating conditions of global financial markets amid high debt", OECD Business and Finance Policy Papers, No. 15, OECD Publishing, Paris, https://doi.org/10.1787/89757fae-en.

OECD (2022c), OECD Employment Outlook 2022: Building Back More Inclusive Labour Markets, OECD Publishing, Paris, https://doi.org/10.1787/1bb305a6-en.

OECD (2023), “OECD SME and Entrepreneurship Outlook”, OECD Publishing, Paris, forthcoming.

OECD/APO (2022), Identifying the Main Drivers of Productivity Growth: A Literature Review, OECD Publishing, Paris, https://doi.org/10.1787/00435b80-en. Arnaud B. (2020), “Has COVID-19 Distorted International Comparability of Unemployment Rates?”, in The Statistics Newsletter, Issue No. 73, December 2020, https://www.oecd.org/sdd/theoecdstatisticsnewsletter-allissues.htm.

ONS, Office for National Statistics (2021), “International Comparisons of GDP during the Coronavirus (COVID-19) Pandemic”, article published on the ONS website on February 1, 2021, https://www.ons.gov.uk/economy/grossdomesticproductgdp/articles/internationalcomparisonsofgdpduringthecoronaviruscovid19pandemic/2021-02-01.

Schreyer, P. (2012), “Output, Outcome, and Quality Adjustment in Measuring Health and Education Services”, Review of Income and Wealth, Series 58, Number 2, June 2012, https://doi.org/10.1111/j.1475-4991.2012.00504.x.