Not So Fast Professor Stiglitz: “Industrial Policy” Still A Rotten Idea

Joesph Stiglitz, 2001 Nobel laureate in economics, wants to revitalize industrial policy through greater government intervention in favoring certain technologies over others. Stiglitz correctly points out the importance of learning and technological development in economic growth, citing such luminaries as Joseph Schumpeter and fellow laureate Robert Solow in defending his position that industrial policy is a productive tool for states in both the developed and developing world.

Stiglitz’s emphasis on development through innovation and technology is noteworthy if for no other reason than it shows that mainstream economics is finally moving beyond archaic models of development, where the panacea for developing nations was to airdrop “shmoo” capital upon an economy and wait for the catch-up effect to lift them out of poverty.

All the same, Stiglitz’s reimagining of industrial policy has the same weaknesses as these earlier models in desperate need of retirement. Reframing the issue from picking winners to picking ones with positive externalities does nothing to resolve the fundamental critiques of centrally planned allocation. Simply put, governments have neither the knowledge nor the incentives to successfully promote industrial policy.

Invoking the importance of ideas, Stiglitz puts forth a classic public goods argument where new ideas have large positive externalities and will therefore be underprovided. The focus is not just on new ideas but on learning, and specifically learning by doing (another nod to the work of Solow). To quote Professor Stiglitz:

the point of industrial policy is not to pick winners at all. Rather, successful industrial policies identify sources of positive externalities — sectors where learning might generate benefits elsewhere in the economy.

There is a lot going on in this statement that needs to be unpacked. Most importantly, how are these positive externalities to be identified and valued? Many firms will credibly claim that they can do a better job if only they had more money and a bit more time. How would an observer determine which firms have the most to learn?

This observing entity would not only need to select firms with the most to offer but be able to offer substantial financial support as well. Accordingly, a central pillar of Stiglitz’s proposed industrial policy is privileged access to credit and subsidized loans. If a government cannot generate funds domestically, perhaps an entity like the World Bank (Stiglitz was their chief economist from 1997 to 2000) could provide such loans to worthwhile businesses.

This protectionism is justified by claiming that what one firm learns can be spread throughout other sectors of the economy. Yet it is not obvious how what one firm learns from building cars or heavy manufacturing, for example, is readily transferable to other, different sectors of the economy. For any learning that actually occurs, some may be transferable but parts will only be applicable to the unique circumstances of that particular firm.

Furthermore, there is no a priori theoretical justification to believe that some firms or industries will be better at learning and spreading than others. Indeed, some of the best learning goes on when firms learn what not to do. Just ask Reed Hastings, CEO of Netflix.

More to the point, the market already performs the service of spreading transferable knowledge. To the extent that learning leads to transferable knowledge, the market provides ample incentive for firms to pay to import useful skills and technology. The days of lifetime careers at the same company have come and gone and it is a common occurrence for firms to poach talent from their competitors. One only needs to look at the field of venture capital to see that a large part of their success derives from connecting those with ideas to those who have the experience to properly execute them.

The larger issue with a grand decider wielding influence over industrial policy is that more money will inevitably find its way into the pockets of lobbyists as opposed to R&D. The irony here is that revitalizing industrial policy would likely lead to fewer ideas being generated by the marketplace in favor of greater rent seeking by firms eager for government support.

Finally, we should be especially skeptical of the merits of industrial policy as a form of foreign development. We must remember that many countries suffer from poor institutions and corruption. Even if they could correctly identify positive externalities, the cold calculus of politics might divert those funds to the coffers of cronies and supporters. The poor of the world need free and open markets that draw in foreign investors, not greater control by kleptocrats doling out favors on the public dime. Giving these governments a veneer of theoretical respectability to their looting of public coffers is more likely to benefit Swiss bankers than aid in opportunities for the poor.

Stiglitz’s idea of industrial policy should go the way of Netflix’s mercifully brief experiment with Qwikster. The market already provides a powerful means of encouraging firms with something to offer. In that, even a Nobelist has much to learn.