A What Works Centre post I thought would be good here. Written with Henry Overman.
Much has already been written on the UK Government’s Industrial Strategy Green Paper. This post isn’t intended to provide an overall assessment or spell out our individual views on the approach being set out (they differ, depending on which of us you ask). But there are areas where the proposed strategy will shape the work that we’ll do at the Centre and where we also hope that our work will influence the implementation of the eventual strategy.
[Full disclosure – the Centre is cited in the document as one of the institutions the Government hopes will help improve local economic growth.]
The first area relates to what we do and don’t know about policy effectiveness, particularly when it comes to some of the Strategy’s 10 pillars – which are a mix of tech (science, research and innovation) cross-cutting (skills, infrastructure, supporting business growth, procurement, trade and inward investment) and sector (new sector deals, clean energy). Academics would call this a ‘matrix’ approach.
Take, for example, policy to support business to start and grow. We know that there are market failures here – entrepreneurs often make avoidable mistakes, which better information could help fix; many young firms need better access to early stage finance (the Green Paper talks about ‘patient capital’).
The crucial question is: what’s the right policy mix to help address these challenges? Our evidence reviews on business support and on access to finance suggest that around half of schemes have measurable impact against policy objectives but around half don’t. Our reviews and associated toolkits start to identify the elements that might go in to the design of a more effective set of interventions. And we’ll soon be publishing more toolkits on incubators, accelerators and science parks. All this material provides guidance on how we might improve support to businesses but major challenges remain – both in terms of gaps in our understanding and embedding the evidence in policy development.
We are in a similar position when it comes to policy to develop skills. We know quite a lot – see, for example our evidence reviews on employment support and apprenticeships and our toolkit on training (soon to be supplemented by a toolkit on apprenticeships). Changes to policy design can improve effectiveness but, once again, there are gaps in our knowledge and challenges in implementation.
Things are more complicated when it comes to investing in science and promoting innovation. We can say something about the specific policy tools – e.g. from our evidence reviews we know that both R&D grants and tax credits drive up innovative activity. But it’s not so clear whether increased innovation at the firm level feeds in to improved local economic performance and there are lots of unanswered questions about the appropriate policy mix. That ambiguity is one of the reasons why people advocate such different approaches to strategy.
In the interests of openness – we should note that one of the things our review did find was that grants and loans programmes that target particular production sectors appear to do slightly worse in terms of increasing R&D expenditure and innovation, compared to those that are ‘sector neutral’. So, while it makes sense for government to recognise that different sectors might need different policy responses (e.g. in terms of the institutional structure that supports those sectors) this might increase the challenge of effective policy implementation in some of the other policy areas.
Questions of infrastructure are similarly challenging. The evidence that we do have on the link from transport to local economic growth raises some questions about the effectiveness of these policies for turning around areas that are struggling. But at the same time, we know that such investments can help drive growth in areas where travel times and congestion are a big issue (and not all of those areas are in London and the South East). Getting the right balance will be crucial.
As with innovation expenditure, people are willing to advocate for very different approaches – particularly when it comes to the overall pattern of expenditure. We’ll continue to make the case that focussing on the overall pattern of expenditure isn’t helpful when it comes to shaping effective policies. What we need is a better understanding of the economic impact of different schemes and improved ways of feeding this information back in to decisions about scheme prioritisation. This will be where our work will focus in the coming years.
We could make similar points about the other pillars, but in the interest of space, let’s turn instead to a final cross cutting issue – whatever happens we think that to be successful, industrial policy will need to be inherently experimental. How we deliver and develop the policy will matter a lot.
Industrial strategy is always going to involve unknowns. Most fundamentally, because it involves funding basic science (or commercialising new ideas) – not all of which are going to work out, so wouldn’t be delivered by the market. In other cases, investments will trigger spillovers between parts of the economy that are hard to see upfront.
Finally, unknowns crop up because – for a lot of the things Governments want to do as part of industrial strategy – we still have a long way to go in understanding what is an effective policy mix. In addition to the policy areas covered above, at least three of the Pillars – strategic procurement, innovative place strategies, and institutions – are subject to big knowledge gaps in terms of what works. As a result, how we implement future industrial strategy will be crucial.
As you might expect, we will be arguing for an experimental approach. We need to test lots of different ideas, figure out what works, scale up the things that do and drop those that don’t. Many of those calling for a more interventionist policy – such as Harvard’s Dani Rodrik – have consistently emphasised this point. Many people have argued that the Green Paper’s approach isn’t such a fundamental break with the past. But a greater focus on flexibility, on experimentation, and on testing and improving, would help differentiate this from the past and increase the chances of success where so many other strategies have failed.
Originally posted here on 27 January.