It’s not about picking winners Since the change of guard at 10 Downing Street there has been much excited talk about industrial policy in the UK. Prime minister Theresa May has, in the words of an FT leader column, “signalled an enthusiasm for industrial policy going well beyond that seen in the Cameron and Osborne years”. Her main adviser is well known for his openness to state intervention in the pursuit of economic growth and employment. She has refashioned ministerial portfolios to create a department for business, energy & industrial strategy. It seems to fit the zeitgeist – especially in Britain. Laisser faire economics everywhere took a bad knock in the global financial crisis, and the UK referendum on EU membership revealed how many Britons felt decades of market liberalisation had done little for them (or worse). And there is growing realisation that past demonisation of government’s role was unjustified. As Mariana Mazzucato, a prominent advocate of active state involvement in industrial innovation, writes: “Almost every technology that makes the iPhone smart was funded by government. This was not just about basic research but rather engagement across the whole innovation chain, including demand-side procurement policies.” Far from the caricature of bureaucrats telling business people (badly) what to do, there is a perfectly reasonable theoretical case for industrial policy within standard neoclassical economics: namely that of correcting market failures, when incentives or information are not so aligned that individual choices coalesce into efficient economic production. In this regard, industrial policy is rather similar to economic development strategy, even though the specific substance may be different. In both, there are plenty of examples of grand policy visions or concrete projects gone awry. But this does not mean industrial or other development is always best left to the market to sort out on its own. One of the most promising lines of thought about how development policy can be done is the “growth diagnostics” approach developed by Ricardo Hausmann, Danik Rodrik and Andrés Velasco. Simply explained, it says economists can act as diagnosticians, identifying which factors in any one situation are the most relevant constraints on growth or industrial development, and formulate policies aimed at ameliorating those particular constraints. “Diagnostics requires pragmatism and eclecticism,” Rodrik writes, “in the use of both theory and evidence. It has no room for dogmatism, imported blueprints, or empirical purism”. In other words, it’s hard but not impossible work, which requires in particular a strong awareness of what we do not know. That means an explicitly stated industrial strategy can be better designed, and politically more robust, than industrial promotion carried out on the sly. Such transparency allows policymakers openly to recognise and address such challenging facts as the Hayekian insight that knowledge is diffused through society and the incentives any policy creates for the private sector to “game” it. These thoughts inform Rodrik’s highly readable analysis of green industrial policy, which includes this summary: “Good industrial policy does not rely on government’s omniscience or ability to pick winners. Mistakes are an inevitable and necessary part of a well-designed industrial policy programme; in fact, too few mistakes are a sign of underperformance. What is needed, instead, is a set of mechanisms that recognises errors and revises policies accordingly. This is a much less demanding requirement than that of picking winners.” We should also acknowledge that industrial strategy is about much more than funnelling money to favoured industries or technologies. A proper diagnostic exercise might find the best policies to involve removing subsidies that make it harder for new technologies to compete in the market (such as still-high subsidies for fossil fuels that tilt the market against green energy), or non-financial policies such as regulatory and urban planning. For example, a new study by MIT researchers shows that the vast majority of current US car use – 87 per cent of vehicle-days – could already be satisfied by existing electric car technology. Carbon emissions from transport could be therefore reduced significantly if greater availability of car-sharing for the remaining 13 per cent of vehicle-days encouraged more drivers to own electric cars. That suggests a big gain for climate change policy, as well as green vehicle manufacturing, can be achieved through a regulatory and planning environment more favourable to car-sharing solutions for travel that requires combustion engines today. The overall point is that there is a lot of scope for economies to improve, and much that smart policy can do to nudge them on to a better path. My colleague Janan Ganesh’s column on how the UK’s success at the Rio Olympics owes a lot to policy decisions made more than 20 years ago is as fine a parable as any of what industrial policy could do if done well.