going going gone routine jobs in global value chains

Going going gone? Routine jobs in Global Value Chains

By Luca Marcolin.

Blog37.0Innovation and new technologies – especially Information and Communication Technologies (ICTs) – and globalisation are key features of 21st century economies. They contribute to shape firm and industry performance and dynamics, and may trigger radical changes in employment patterns and the skills profile of the workforce.

Many factors determine the way firms organise production across borders and, ultimately, countries’ specialisation patterns. Among them is the routine content of occupations, i.e. the extent to which the tasks undertaken follow a set of well-defined rules or sequences. If the relationship between technology and employment has been already explored on this blog, this post adds an extra dimension to the problem, one related to Global Value Chains (GVCs), and proposes a new way to identify routine intensive occupations.

Recent OECD work (Marcolin, Miroudot and Squicciarini, 2016a) develops new indicators of the routine intensity of occupations based on information from the OECD Programme for the International Assessment of Adult Competencies (PIAAC). These indicators (available on request to the authors) cover 20 countries and capture individual workers’ degree of independence in planning and organising their activities and time, and their freedom in deciding what to do on the job, and distinguish between what workers do on their jobs and their skill endowment.

Using the routine intensity indicator (RII), occupations are subdivided into four groups:

  1. Non-routine (NR) occupations, such as legislators and managers
  2. Low routine-intensive (LR) occupations, like secondary education teachers and hairdressers
  3. Medium routine-intensive (MR) occupations, such as machinery mechanics and shop salespersons
  4. High routine-intensive (HR) occupations, like assembly line workers and food preparation assistants.

The resulting indicator is used in Marcolin, Miroudot and Squicciarini (2016b) to calculate the average share of employment accounted for by the different groups of routine-intensive occupations over the period 2000-2011. Cross-country differences emerge: at the economy level, the number of non-routine and low-routine-intensive workers ranged between about 22% (Italy) and 56% (Luxembourg) and the average share of workers in high routine-intensive occupations ranged from 21% (Greece) to 37% (Poland) (see Figure 1). Services generally enjoy higher shares of employment in non-routine, low-routine and medium-routine jobs. Conversely, manufacturing accounts for higher shares of workers in high-routine occupations: 41% on average, as compared to an average 28% in services.


These differences mirror differences in the composition and structure of industries, the skills endowment of the workforce and countries’ stage of development. For instance, skill intensity (in terms of either proficiency or skills use) is found to be only weakly correlated with routine intensity: while it may be true that high-skill workers tend to specialise in non-routine tasks, a proportion of the highly educated may still be at risk of relocation or automation.

Offshoring does not always mean fewer jobs.

What drives routine rather than non-routine employment? Marcolin, Miroudot and Squicciarini (2016b) show that offshoring does not necessarily lower the levels of employment of routine-intensive workers. Especially in manufacturing, offshoring the inputs of production relates positively to employment in routine-intensive occupations. Such a relationship is consistent with the specialisation of manufacturing firms in specific stages of GVCs: as they import more inputs that need to be further processed, they also rely relatively more on routine-intensive jobs. Conversely, offshoring of final assembly is associated with a reduction of employment in non-routine jobs: in that case it is not only production and core assembly that are lost, but also some supervision and support activities.

Of course, this changes according to the specialisation of countries: European catching-up and transition economies have been gaining employment in MR and HR occupations, while more mature “service-based” economies have been experiencing more labour demand in NR occupations. While more open trade regimes might have facilitated such specialisation, these trends appear to be explained by other determinants including the skills of the workforce, technology endowments, innovation capabilities and industry structure.

Technological innovation helps employment but ICT does not always do so…

Technological innovation (using patenting as a proxy) matters positively for both routine and non-routine employment. The stronger competitiveness that technological innovation may confer, especially in manufacturing, seems to always translate into-higher employment levels. Conversely, ICT-related capabilities generally correlate positively with employment levels in all groups but for the high-routine one. Also, while comparatively higher skills are generally associated with higher employment especially in non-routine and low routine-intensive occupations, differences emerge between manufacturing and services industries. All of the above is found to be true independently of the industry structure of the country.

Complex dynamics need coordinated policies

These complex relationships point to the need of tailoring policies depending on whether manufacturing or services industries are considered. This underscores the need for well-functioning labour markets and appropriate labour market policies, able to strike the right balance between employment flexibility and aggregate welfare and to smooth the reallocation of the labour force according to the patterns of production and of trade in value added. Also, labour market policies need to be coupled with trade, industry and innovation and competition policies, aiming at creating the right business environment for firms which operate in open, globalised economies.