In my last post, I promised a quick overview of what analysts, from those at the World Bank to those at Harvard, are saying about contemporary industrial policy.  That promise was in response to a dismissive article in The Economist in August.  While I concurred with the magazine that we do not know a lot about how to prescribe good industrial policy, I dissented from their conclusion that this lack of knowledge indicated that we should abandon it altogether.  That left me in the position of having to say something constructive about how to use industrial policy in the modern era.  Moreover, since this site is dedicated to issues related to budgets and civil society, I also had to connect this new thinking on industrial policy back to those issues. I will try to do that today.

As it happens, there is no consensus about the new industrial policy. Opinion ranges from those whose views deviate a relatively short distance from traditional comparative advantage economics, to those who take a much more radical view of the need for state intervention.  All in all, however, there is a general recognition of the need to combine targeted state intervention with market discipline in an effective, if elusive, elixir.

Some of the elements of this elixir which are relatively less (but not non-) controversial include:

Imitation. There is a widespread understanding that poor countries have to imitate (or, as Erik Reinert prefers, “emulate”) wealthier countries with respect to diversification of the economy. There is some debate as to how (and how much) poor countries should imitate rich countries. For example, Justin Lin argues that poor countries should should target industries in countries which are just a bit wealthier than they are, while others argue that poor countries may need to target more mature industries in more advanced countries. While there is no agreement about exactly what constitutes an appropriate target for imitation, it seems reasonable to argue that citizens can expect their governments to explain how and what they are targeting, and how they plan to transform the existing economic structure into one that can support this type of imitation.

The advantage of imitation is that there is a record of how other countries developed their industries that can be referred to by civil society (as well as government officials) when monitoring the effectiveness of industrial policy.  Nevertheless, while imitation is essential, many analysts argue that innovation in areas where there is no well trodden path is also an important part (maybe more important) of the development process and this may be more unpredictable and difficult to monitor. At the same time, governments that profess to be subsidizing innovation still need to have clear plans for industrial restructuring that can be monitored.

Subsidize competition. There is nearly universal consensus that the goal of subsidy is to create industries which are competitive both at home and in international markets.  Reinert refers to this as ‘first emulation, then comparative advantage,’ because it relies on state protection in the first phase and competitive markets in the second. It follows that any government subsidy should be aimed at the creation of competitive industries which can ultimately survive without the subsidy. Thus governments should have a plan for the reduction and ultimate elimination of any subsidy, and that plan can be monitored.

The difficulty, of course, is knowing for how long the subsidy is really needed. But, even if this is unknown, regular plans to eliminate the subsidy require public debate about its extension.  This can contribute to ensuring that if a subsidy is continued, it is because there is a belief that the subsidy is ultimately leading to self-sufficiency.  Caution is necessary, however, because public debates of this kind can be dominated by beneficiaries of the subsidy seeking to perpetuate narrow benefits to themselves, rather than broader development.

Industries before firms.  The biggest risk of industrial policy is that it is captured by private interests and serves patronage interests rather than development goals.  There is no simple way to avoid this risk, given that successful industrial policy requires, as Peter Evans has described it, active “midwifery” or “husbandry” of private sector innovation.  This means that the state has to develop close ties to entrepreneurs in order to push them toward greater discovery and innovation.  One way to mitigate the risks of patronage is to focus industrial policy on removing constraints to industry development, rather than providing benefits to specific firms.  Industry-wide policies include subsidies for research, business incubators, industry marketing support, support for certification processes, and so on (see here for some examples).  These are subsidies available to market participants either on an equal basis, or on the basis of their market success.

Don’t ignore comparative advantage.  If strict fidelity to comparative advantage does not leave enough space for countries to diversify, most analysts recognize that comparative advantage needs to be consulted when decisions are made about what to subsidize and which mature industries to target.  Since economists do not agree about the degree to which policy ought to deviate from comparative advantage, there is no reason to expect anyone else to do so, but governments can still signal through their plans how these build on a country’s comparative advantage, and, if they try to “defy” it, how and why this is likely to be successful.  Public debate can only achieve so much here, but it can certainly act to curb some of the more fanciful attempts to defy comparative advantage.

In all likelihood, governments are going to continue to experiment with industrial policies, as they always have.  Indeed, they are only likely to do more of this experimentation in the near future, given that it is no longer considered, after a couple decades of ideological warfare, taboo.

We still have a lot to learn about how and when these policies can be effective, but one thing is certain: industrial policy implies the use of public money to reach development goals, and is therefore an important area for increased debate and monitoring by civil society.  That debate and monitoring requires, in turn, greater understanding of what industrial policies are, and how they work.  This is just the beginning.