Knowledge intensive sectors

The OECD identifies high and medium tech manufacturing; high value added “knowledge intensive” market service industries such as finance and insurance and telecommunications; and business services. The current OECD definition also includes education and health. The Work Foundation has extended this definition in the recent Ideopolis report to capture a higher share of employment in the cultural and creative industries.

The OECD wide definition of knowledge based industries indicates that Ireland was the most knowledge based economy in the OECD, with these industries accounting for 48 per cent of GDP followed by the US, Germany, and Sweden with around 43 per cent. The knowledge based industries accounted for around 40 per cent of GDP in the UK and France. Estimates for Japan are only available on the more restricted market based industry definition, excluding health and education. However, on this basis Japan has a less knowledge based industrial structure than Germany, the US, the UK or France. (based on 2005data)

Knowledge jobs and knowledge workers

There are (at least) three ways we can work towards a definition of knowledge workers:

  1. All those who work in the top three standard occupational classifications (managers, professionals, associate professionals)
  2. All those with high levels skills, indicated by degree or equivalent qualifications
  3. All those who perform tasks that require expert thinking and complex communication skills with the assistance of computers.

Knowledge workers account for about 42 per cent of all employment in the UK in the first quarter of 2006, using the occupational definition of the top three occupational groups. This compares with 31 per cent of total employment in 1984, according to the latest projections prepared for the Sector Skills Development Agency (SSDA). The SSDA is projecting the share will grow to just over 45 per cent by 2014.

The underlying story is one of fairly stable constant structural change in the labour market decade on decade. The share of knowledge economy jobs has increased by between 4 and 5 percentage points in each decade, while the share of unskilled jobs has fallen by about 2 and 3 percentage points in each decade.

International comparisons are difficult because of differences in occupational classifications and how these are interpreted in national surveys. The best comparable data we have found to date suggests that in 2004 or latest year available between 40 and 45 per cent workers in the smaller North European economies, North America, and Australia are knowledge workers. The UK lies alongside Germany and Canada, but behind the Nordic economies, Switzerland and the Netherlands.

Knowledge workers in the UK economy 1984-2014






Knowledge workers





Personal services; sales; admin/clerical





Skilled/semi- skilled; manual





Unskilled jobs





Note:2014 projected. Knowledge economy jobs are managerial, professional, associate professional standard occupational classifications. Personal services include care, recreational, and some hospitality jobs. Employees and self-employed.Source: Working Futures 2004-2014, table 4.1

Human skill definitions

An interesting alternative approach to using broad occupational categories is set out in a recent paper by Autor, Levy and Murname5 that divides human skills into five categories:

  1. Expert thinking: solving problems for which rule based solutions do not exist. Computers cannot substitute for human beings but can assist by making information more readily available
  2. Complex communication: interacting with other people to acquire or convey information and persuading others of the implications – examples might include some managers, teachers, sales people
  3. Routine cognitive: mental tasks closely described by rules such as routine processing application forms and claims – these jobs are often vulnerable to computerisation
  4. Routine manual: physical tasks closely described by rules, such as assembly line work and packaging. These repetitive tasks can in some circumstances also be undertaken by programmed machines
  5. Non-routine manual tasks: physical tasks hard to define by rules because they require optical and fine muscle control, including truck-driving and cleaning. Such jobs are unlikely to be assisted or replaced by computers.

The authors applied these categories to the US workforce between 1969 and 1998 and found that jobs requiring complex communication increased by nearly 14 per cent, and jobs requiring expert thinking increased by just over 8 per cent. All other jobs saw a declining share of employment over this period.

A recent large scale survey carried out by the Economist Intelligence Unit of top company executives and managers used a similar approach in identifying which skill sets would be most valuable in terms of competitive advantage in the year 2020. This adopted a five-category definition:

  • Complex knowledge based roles that are primarily outward facing and require developed communication and judgement skills
  • Complex knowledge based roles that are primarily inward-looking and require developed communication and judgement skills
  • Simple knowledge based roles that are rules-based, outward facing and do not require developed communication and judgement skills
  • Simple knowledge based roles that are rules based, inward facing and do not require developed communication and judgement skills
  • Production roles directly related to manufacturing or production processes.

Perhaps not surprisingly, 62 per cent of respondent’s said outward facing complex knowledge based roles would be most important for the organisation’s future competitive advantage, followed by 28 per cent saying inward facing knowledge based roles would be most important. The rest were cited by only 2 to 4 per cent of respondents as being important for future competitive advantage.

 This approach undoubtedly gets closerto defining knowledge jobs in terms of both cognitive complexity and the relationship to computers – in other words, what people actually do – and might be regarded as superior to simply classifying jobs by occupational title or educational qualification of the job holder. The disadvantage is that it requires either an extensive re-working of the statistics or original survey work and may not easily lend itself to direct comparisons with previous work or international comparisons.