Industrial strategy and inclusive structural change

17 December 2018

Maria Savona - Professor of Innovation and Evolutionary Economics at the University of Sussex

Industrial strategy must pay attention to left-behind places as well as mature sectors and vulnerable categories of workers.

Last year’s industrial strategy white paper includes a plan for developing concerted local industrial strategies that tap into ‘local comparative advantages’. While it is of course vital that innovation does not leave certain people and places behind, it has been argued that the current pleas for local industrial strategy are still very much centred around large city-regions, and risk exacerbating the striking regional inequality in the UK (Rodriguez-Pose, 2017; see also Lee, 2018 and Tomaney and Pike in Berry ed. 2018). Similarly, the preoccupation around the productivity slow down and puzzle have led to a focus on key sectors, rather than privileging more pervasive interventions to ensure technological upgrading of mature sectors. Here I argue that the unquestionably positive support for research and development, and innovation at large, should be part of a more articulated industrial policy that ensures inclusive structural change of the economy, that pays attention not only to left-behind places, but, relatedly, to mature sectors and vulnerable categories of workers.

The industrial strategy in the EU context of industrial renaissance

The industrial strategy has been devised in light of the recently revamped interest – in Europe and worldwide (Stiglitz et al., 2013; Warwick, 2013) – in the role of industrial policy. The need to ‘reindustrialise’ countries has been advocated by several international programs, such as the Juncker Plan in Europe, the Made in China 2025 programme, and the Indian National Manufacturing Policy. The Juncker Plan, launched at the end of 2014, has been welcomed as a substantial fiscal policy intervention to get Europe out of the recession, despite criticism from some that it is ‘too little, too late’, in terms of a counter-cyclical boost to the economy (Dosi et al., 2017). The plan has also been understood as a concerted vision of public and public-private funding in areas that are considered strategic to ensure a European industrial renaissance.

When briefly commenting on the Juncker Plan as a platform for industrial policy (Savona, 2018), I previously raised a few questions that could be similarly posed to the industrial strategy white paper. First, in the UK context, widely considered as a service-based economy, what type of (further) de-industrialisation would represent a threat? Second, how would a concerted platform of innovation and industrial policy interventions steer a transformation of the sectoral structure that considers the large differences in sectoral specialisation and technological opportunities across UK regions?

The industrial strategy, through the sector deals, focuses on key, strategic sectors such as automotive and artificial intelligence, and plans to invest in large infrastructures, digital and logistics. As relevant as it is, this logic of an industrial renaissance might not be enough, when aiming for a more pervasive and spatially balanced industrial upgrading.

One such missing piece of the puzzle that the sector deals aim to tackle is how to prepare the ground, facilitate and support the maintenance and growth of backward linked services (engineering, technical consultancy, etc.) to the more traditional (mature) manufacturing industries. In addition, attention to forward linked sectors would also require a vision that identifies new technological opportunities – not only those offered by frontier technologies such as Artificial Intelligence and Robotization, but those that would help to ‘rejuvenate’ traditional manufacturing and non-tradable services. The importance of supporting a certain type of sectoral structural change (labour demand) is similar to planning adequate education and training policies for the labour force (labour supply) and directly affects what the future of employment in the UK could be. The next section presents some evidence in support of this view.

R&D and changes in the employment composition: a policy conundrum?

The industrial strategy includes plans to increase total R&D investments to 2.4 per cent of GDP by 2027 and to 3 per cent over a longer time span; increase the rate of R&D tax credit to 12 per cent; and invest £725 million in the new Industrial Strategy Challenge Fund programmes to leverage on the value of innovation (HM Government, 2017).

While the positive link between R&D spending and innovation performance has been relatively uncontroversial for many decades (Griliches, 1979), the heterogeneity of this effect across sectors and places, as well as its unintended consequences, are much less so.

Firm R&D spending is a strategic choice that might represent a trade-off with other investments, and certainly requires changes in the organisation, and demand for new types of workers and skills. R&D might have spillover learning effects within the firm and a multiplicative effect in terms of labour demand within the local labour markets in which the firm is located. Much depends on the initial industrial structure and employment composition of local labour markets. Creating incentives (i.e. tax credits) for firms to spend in R&D might have very different effects on the sectoral composition of employment and on the type of employment created depending on the local characteristics.

Figure 1: Share of routine employment across travel-to-work-areas in Britain, 2001 Reproduced from Ciarli et al., 2018a; each TTWA reports the share of category NS-SEC7 (Routine occupations)

Across the UK’s local labour markets, Ciarli et al. (2018a) find that firm spending in R&D has, on average, a low multiplicative effect on employment rates, while it has quite a remarkable effect on the changes in the employment composition, depending mainly on the initial (in 2001) industrial structure of the local labour market. In areas highly dense in non-routinised occupations, which represent about 85 per cent of the UK population (see figure 1), R&D spending does increase the share of the highly educated, working in manufacturing paid employment sectors (see note 1).

In contrast, when R&D spending occurs in areas with a higher share of routinised jobs, the (positive) boost to employment is concentrated in lower educated workers employed in non-tradable services. A notable, and relatively overlooked phenomenon, is that the creation of employment in highly routinised areas is substantially driven by an increase in self-employment, a long-term trend in the UK, and to a lesser extent in the main European countries (see figure 2).

Figure 2: Self-employment as a % of employed labour force (1995-2016) Reproduced from: Ciarli et al., 2018a

It seems that innovation, in the form of R&D spending, is a mixed blessing, depending on the initial industrial structure. While rightly supporting the generation of new ideas, the maintenance and improvement of scientific excellence and the competitiveness of firms, innovation might also have side effects of increasing regional polarization in terms of industrial structure, share of routinised employment and self-employment.

There seems therefore to be a two-speed, innovation-led, structural change in different areas: an increases in manufacturing jobs of the areas that are already highly dense in non-routinised jobs, while areas that are already highly dense in routinised jobs – the ‘left-behind’ areas – experience a loss of manufacturing and an increase of non-tradeable (hotel and restaurants, trade), lower skilled jobs. It is here that the increase in self-employment is also concentrated: The increase in self-employment in these areas might have little to do with entrepreneurial jobs (Thurik et al., 2008), and instead represent ‘refugee’ self-employment, that is, turning to self-employment as a coping strategy, due to skill mismatches (Vona and Consoli, 2015), extreme skills complementarities (Mazzolari and Ragusa, 2013; Eeckhout et al., 2014) or simply ‘hidden unemployment’ (Blundell et al., 2014).

Our findings pose a challenging policy conundrum that the industrial strategy should address more substantially. The potential polarizing effects of innovation on structural changes in the composition of employment and the labour market should be prevented, and actions taken to ‘prepare the ground’ for local investments in R&D. In this context, ensuring a direction to industrial policy that supports innovation and higher quality jobs in non-tradable services is a necessary complementary measure to harness the beneficial effects of R&D.

Innovation policies for inclusive structural change

The industrial strategy should aim at a more substantial integration between innovation and industrial policy to achieve inclusive structural change (Ciarli et al. 2018b). Inclusive structural change should undoubtedly be starting from the recent proposals around inclusive innovation that aim to broaden the access of individuals, industries and places to innovation processes. However, it should also go beyond these, based on a comprehensive, evidence-based understanding of how innovation affects industrial and employment structure, including how it shapes inter-sectoral linkages between the frontier sectors and more mature, linked ones. This knowledge is important when aiming for an inclusive innovation policy that does not leave behind places and vulnerable categories of workers.

The sector deals proposed in the strategy focus on a few strategic sectors. The deals seem to follow a mission-oriented logic of investing into specific, potentially frontier areas, but are relatively blind as to what the more pervasive impact on related – or unrelated – sectors will be, depending on the initial industrial structure of specific places.

It has been argued that industrial policy should be characterised by ‘ambidexterity’ (Frenken, 2017; Frenken et al., 2007), that is, the ability to intervene in both radically unrelated areas – for instance through public procurement, and a clear identification of strategic areas – and in related areas, building upon accumulated technological capabilities. Mission-oriented policies should be complemented with more targeted instruments that ensure structural transformation of economies that are based on extant specializations and related technological capabilities. The industrial strategy seems to be pursuing much of the former, but less of the latter.


1.) Travel-to-work-areas in the UK are sub-district units and are approximated to local labour markets.

This post is based on the authors’ contribution to What We Really Mean When We Talk About Industrial Strategy, edited by Craig Berry, and published by Future Economies at Manchester Metropolitan University with the support of the British Academy. The ebook is available for free at: 

Related posts

Towards a globalist feminist trade politics

Erin Hannah, Adrienne Roberts & Silke Trommer - 12 November 2019

A truly progressive agenda has to recognise the centrality of gendered and other social hierarchies to the deeper workings of the global political economy. This is part 5 of the series 'Reglobalisation in action'.

Read more

The European periphery and the Eurozone Crisis

Niall Dooley - 9 November 2019

European elites have argued that ‘peripheral’ Eurozone states such as Greece and Ireland must ‘follow the rules’. But ‘following the rules’ of European integration – particularly in the domain of financial services – drove divergence and led to their original crises. This is part 6 in the series 'Industrial development in a post-crash world'.

Read more