2.2. Productivity
In 2016, labour productivity in information industries was higher than in other industries in the non-agricultural business sector in almost all OECD countries – and as much as one-and-a-half times higher in Israel and Turkey. This reflects the relatively higher investment by information industries in machinery and equipment, as well as knowledge-based capital such as software and Research and Development.
The labour productivity of information industries varies among countries for reasons including the roles different countries play in global value chains (many products of information industries are highly tradable) and variations in the weight of the different components of information industries (e.g. ICT manufacturing and services). Large differences in productivity relative to the rest of the economy contribute to very high ratios of labour productivity in information industries compared to other industries in India (nearly 5 to 1), Costa Rica and Israel. In contrast, the high ratio in the United States (over 2 to 1) reflects a focus on relatively higher value-added activities, while in Korea (2 to 1) it is indicative of the strength of ICT manufacturing. Conversely, the relatively low ratios observed in countries such as Switzerland and Norway are linked to high average levels of productivity in other industries.
In Finland, the Netherlands and Austria, labour productivity levels in ICT manufacturing are markedly higher than those in information and communication services, and about twice those of the rest of the economy. This suggests concentration in high value added ICT manufacturing such as advanced components. In contrast, the comparatively low productivity levels of ICT manufacturing in Poland and Estonia suggest that these countries are hubs for more simple ICT products.
Understanding the drivers of productivity growth requires an awareness of the contribution made by each industry (OECD, 2017a). Between 2006 and 2016 – a decade during which productivity growth slowed in most countries due to the Great Recession – the contribution of information industries remained generally positive. However, this contribution varied among OECD economies, with the highest relative contributions (more than half of the total) occurring alongside robust overall productivity increases in Sweden, the United States and Germany. In France, Finland, Italy and Norway, meanwhile, productivity growth in information industries compensated for weak or negative growth in the rest of the economy.
In Sweden, the United States and Ireland, content and media industries contributed relatively strongly to productivity growth. In many countries, telecommunication and information services also provided notable contributions. In most cases, this corresponded to a fast expansion of ICT services, accompanied by employment growth, while in telecommunications it resulted chiefly from a strong reduction in employment.
On average, in the OECD the level of labour productivity in information industries is about 65% higher than that of other industries in the business sector.
Definitions
Labour productivity is the amount of output (value added) produced per unit of labour input (number of persons employed or, when data allow, the number of hours worked). Industry values are computed relative to the whole economy (i.e. to GDP per person employed in each country), adjusting the indicator for differences in productivity levels across countries.
Information industries includes ICT manufacturing and information services i.e. ISIC Rev.4 Divisions 26 and 58 to 63. See page 2.1 for more detail.
The non-agricultural business sector excluding information industries refers to ISIC Rev.4 Divisions 05 to 25, 27 to 56, 64 to 66 and 69 to 82.
Measurability
Value added is measured in National Accounts based on structural business surveys and other sources. Persons employed is typically measured using national Labour Force Surveys. The use of persons employed rather than hours worked leaves differences in average working hours across sectors unaccounted for.
Measuring real value added can be challenging. For example, most countries assume no change in labour productivity for public administration, defence, education, human health and social work activities; these are therefore excluded. Real estate services are also excluded as their output includes a large imputation made for “services” from dwellings to owner-occupiers. In addition, industries such as construction, accommodation, and food and beverage services are characterised by high degrees of part-time work and self-employment, which can affect estimates of actual hours worked. See OECD (2017b) for more discussion of productivity measurement issues. Finally, using hedonic deflators to account for the improving quality of ICT products can significantly improve real value added measures and therefore sector productivity measures. However, such techniques are not applied in all countries and, more importantly, often applied to ICT manufacturing only, despite similar quality changes occurring in services (notably broadband).
Source: OECD, STAN Database (http://oe.cd/stan), National Accounts Statistics and national sources, September 2018. See 1. StatLink contains more data.
← 1. Real estate activities (68) are excluded from calculations as the value added of this sector includes an imputation made for dwelling services provided and consumed by home-owners.
Labour productivity is calculated as current price value added per person employed and per hour worked. Ratios are presented relative to the productivity of non-agriculture business sector excluding information industries
It is preferable to measure labour productivity based on hours worked rather than number of persons engaged. However, detailed hours worked per activity are sometimes not available. In such cases, a substitute based on employment is used to maximise country coverage.
The difference between the two measures reflects the average hours worked per person engaged. Higher relative value added per person reflects higher hours worked per person in the information industries.
For Brazil, Canada and Mexico, labour productivity estimates are based on jobs instead of persons engaged.
Source: OECD, STAN Database (http://oe.cd/stan), National Accounts Statistics and national sources, September 2018. See 1.
← 1. Real estate activities (68) are excluded from calculations as the value added of this sector includes an imputation made for dwelling services provided and consumed by home-owners.
Labour productivity is calculated as current price value added per hour worked. Ratios are presented relative to the productivity of non-agriculture business sector excluding information industries.
It is preferable to measure labour productivity based on hours worked rather than number of persons engaged. However, detailed hours worked per activity are sometimes not available. In such cases, a substitute based on employment is used to maximise country coverage.
The difference between the two measures reflects the average hours worked per person engaged. Higher relative value added per person reflects higher hours worked per person in the information industries.
Source: OECD, STAN Database (http://oe.cd/stan), National Accounts Statistics and national sources, September 2018. See 1. StatLink contains more data.
← 1. Real estate activities (68) are excluded from calculations as the value added of this sector includes an imputation made for dwelling services provided and consumed by home-owners.
Labour productivity is measured per hour worked. The contributions are calculated relative to the productivity growth of non-agriculture business sectors.
For Canada, data refer to 2006-14.