By Vidit Doshi
The practice of industrial policy has evolved in significant ways over the past few decades, yet our conversations around it continue to follow outdated frameworks. But what really is the recent evidence around industrial policies, what do such policies look like, and how have they evolved? And based on all of this, how should we think about their impact and success? A range of researchers have been thinking about these questions in different contexts.
Our panel Taking Stock: Recent Evidence on Industrial Policy, featured Nathan Lane, Associate Professor in Economics at the University of Oxford; Myrto Kalouptsidi, Professor in the Department of Economics at Harvard University; and Rohit Lamba, Assistant Professor of Economics in the Department of Economics at Pennsylvania State University. The panel was moderated by Dani Rodrik, Ford Foundation Professor of International Political Economy at the Harvard Kennedy School and Reimagining the Economy Faculty Co-Director. Here are some of the main takeaways from the discussion.
- The conversation on industrial policy needs more nuance and less binary analysis
Lane, drawing on his range of recent research on industrial policy, lamented that discussions on industrial policy often fall into the trap of being overly simplistic and binary: is it good or bad? Is it possible or not? This black-and-white analysis doesn't capture the richness of practice being undertaken around the world. There is currently a huge proliferation in industrial policy, especially among the OECD countries. While the empirical evidence points towards a mixture of success and failures, there are always lessons to be drawn. Additionally, there is a tendency to think of industrial policy in terms of big policy reforms. This however is a narrative that is imposed in retrospect. Modern thinking around industrial policy views it as a set of smaller, incremental policies. The empirical study of industrial policy has historically almost always given negative results. Most of these studies were based on correlations between level of intervention and economic performance but were unable to make claims about causality. More recently, since the credibility revolution in economics, economists have used more rigorous methods and found more positive results.
- Success depends on the objectives
Continuing with the broader theme of viewing industrial policy with nuance, the panel reinforced the need to view industrial policies in terms of their specific goals and objectives. Kalouptsidi discussed her work investigating the expansion of shipbuilding in China - which went from obscurity to dominance in less than 20 years. She noted several shifts in global dominance in ship building had occurred in the 20th century - from the U.S. during wartime to Japan, and subsequently to South Korea and China. In the early 2000s, China’s global market share in the shipbuilding industry was less than 10 percent. After shipbuilding was included in China’s National Five-Year Plans between 2006-2010 and then again 2011-2015, there was an extraordinary level of national support provided to expand the sector. The scale of the support was unprecedented (around $11bn annually) with the majority of subsidies going towards encouraging entry, followed by production and investment. Within two years, China overtook South Korea and Japan as the world’s largest producer. However, the ultimate impact on economic welfare remains debatable. The return on investment looks very poor, for every $1 spent, returns to welfare were close to only 20 cents.
However, the low returns to welfare don’t mean the policies were failures. Kalouptsidi highlighted two strategic rationales that might have been motivating Chinese policymakers: the reduction in global trade costs due to increased ship supply and the strategic advantage gained through enhanced military production capabilities. Against those objectives, the policy could be viewed more favorably - so analysis should be careful and cognizant of the true objectives undergirding the policy.
- Traditional critiques of industrial policy have new responses
Lamba set out responses to three typical critiques of industrial policy. First, the Hayekian critique is that the government is poorly placed to “pick winners” because information is dispersed across the system and markets should be left alone to allocate capital and resources. However, Lamba cited recent work from Dani Rodrik, Nathan Lane, and Reka Juhasz, which shows that successful industrial policy is more about “letting losers go” rather than “picking winners.” Second, industrial policy risks political capture from industries that would specialize in extracting rents from the state rather than creating productive activity within the economy. Lamba countered with examples of highly competent bureaucrats in South Korea and local competition in China - both of which mitigated capture. Third, industrial policies may not account for their full effects. For example, he cited India’s policy to restrict the import of foreign laptops; the goal was to build up domestic manufacturing, but it ended up delaying the growth of the services sector. The lesson practitioners have learned is that industrial policies should be time-bound so they can be constantly reviewed and updated as governments gain new information about the impacts that policies are having.
- Industrial policy needs the right conditions to flourish
Lamba compared China and India’s experiments with industrial policy in the 20th century as a pathway to development. China is widely regarded as successful, but India isn’t. He laid out three differences which could explain the disparate outcomes.
First, China had a moderately skilled labor force in large numbers while India’s primary and secondary education was unable to produce a similar level of human capital. India’s prioritization of higher education means it has produced many of today’s tech CEOs while much of the general population continues to suffer with malnutrition and live with a GDP per capita of $2,500. Therefore, India’s workforce could never realistically compete despite similar starting points in the 1960s. Second, China could decentralize governance and allow local experimentation whereas India’s post-partition governance was highly centralized. Indian elites were suspicious of local governance and therefore held much of the bureaucratic power in state capitals. Third, China was able to act decisively and thoroughly: suppressing market wages and building infrastructure at speed in a way India never could because of its internal divisions and slow-moving bureaucracy.
