Recent years have seen interest grow once again in industrial policy, specifically micro-interventionist industrial policy. A search of the Nikkei Keizai Shimbun database produces a greater number of hits for the keywords "industrial policy" and "policy & system" for the period beginning 2016, showing a particularly high increase since 2018 (see graph).
In international academic circles of economics also, there has been a notable increase in interest in industrial policy. For example, the European scholarly publication Journal of Industry, Competition and Trade, compiled a special issue on industrial policy that was guest edited by Professor Karl Aiginger of the Vienna University of Economics and Business and Professor Dani Rodrik of the Harvard University. Both pointed out in their editorial notes "Rebirth of Industrial Policy and an Agenda for the Twenty-First Century" that "After a period of decline in interest and premature predictions of demise, industrial policy is back on the scene."
This academic interest in industrial policy has also extended to the United States. The American Economic Association's Summer 2019 issue of the Journal of Economic Perspectives included an article by Nicholas Bloom, Professor of Economics at Stanford University, and two other authors entitled "A Toolkit of Policies to Promote Innovation." Predicated on a survey of relevant empirical studies, they evaluated evidence for the effectiveness, cost-benefit differences, as well as other outcomes related to a total of nine policy tools that included R&D grants, R&D tax credits, and patent boxes (preferential tax systems on revenues generated from patents).
This same trend has also manifested in the topics of articles published in the so-called "top five journals," which are regarded as the most authoritative journals in the field of economics. An "industrial policy" keyword search found only one article in the American Economic Review (AER) for the period between 2000 and 2009, but generated six AER hits and a total of eight articles overall for the period between 2010 and 2019.
♦ ♦ ♦
The article by Professor Aiginger and his co-author points out that this increased interest in industrial policy is due to a growing demand in the developing world for changes in industrial structure, long-term labor market malaise and financial crisis in advanced economies, as well as major technological changes. Moreover they cite the presence of China looming behind all these trends.
China has not sought to merely expand its economy, but, as implied in its strategic plan "Made in China 2025" issued in 2015, the nation has employed industrial policy targeting advanced technology sectors in its aim to be the world leader even in terms of state-of-the-art technological development. That is why, in recent years, the Trump administration has been so concerned with China's industrial policy. The above graph also shows the number of hits when the keyword "China" has been added. Such a correlation may imply that the recent increase in articles on industrial policy is related to China.
However, the reality of the current situation is not the only factor behind the high interest in industrial policy in economics. The development of methods for conducting empirical research founded upon econometrics has been a key in enabling scholars to rigorously and quantitatively identify and evaluate the effectiveness of industrial policies. It is this point that I would like to explain in more detail using examples from specific articles.
In a 2014 Quarterly Journal of Economics (QJE) article by Patrick M. Kline and Enrico Moretti, both of the Department of Economics at the University of California Berkeley, the authors evaluated development policies for the Tennessee River Valley in the United States over the period of the 1930s to 1950s.
The Tennessee Valley Authority (TVA) was a core part of President Franklin Delano Roosevelt's New Deal. It is also a typical example of a place-based industrial policy similar to Japan's new industrial city construction and regional revitalization policies. Nevertheless, the TVA's effectiveness had never been rigorously evaluated before the research by Kline and Moretti. The professors' paper utilized an empirical research method known as a "natural experiment" to evaluate the policy's effectiveness.
The original reason why the Tennessee River Valley was chosen as the object of development policy was its severe underdevelopment and other unique attributes, which made it particularly suitable as a target of development policy. Due to its unique characteristics there was difficulty in isolating policy effects in terms of economic changes in the region before and after policy implementation compared to other regions not targeted under development policies. The authors therefore performed an econometric analysis for comparison using six other regions for which development policies similar to the TVA were proposed but not implemented due to political reasons as controls.
Their findings showed that the growth rate of manufacturing employment was relatively larger in the Tennessee River Valley not only during the period when public investment was provided, but also after the investment ended; whereas the growth rate of agricultural employment, despite being high while public investment was the greatest, fell below other regions after public funding was terminated. These effects show that regional development policies where public investment is intensively furnished to a specific region are able to produce sustainable positive effects in industrial growth through the effect of agglomeration.
In another AER article published in 2017, Sabrina T. Howell, Assistant Professor at the NYU Stern School of Business, employed a method known as a regression discontinuity design (RDD) to evaluate the effects of R&D grants. Her focus was a U.S. Department of Energy (DOE) research and development grant program. During the selection process for companies to whom grants would be provided, the DOE ranked companies based upon their applications.
Although there is continuity in the rankings, discontinuity occurs where companies above the threshold for a particular award receive grants and those below do not. A regression discontinuity design examines whether there is any discontinuity in the number of future patents or other innovation outcomes before and after the award is granted, both above and below the discontinuity. The results of the analysis found that the award significantly increased subsequent cite-weighted patents as well as the chances of receiving venture capital funding for the recipient companies.
♦ ♦ ♦
The natural experiment and regression discontinuity design which were employed in these two articles have been widely utilized in recent years in the field of economics as empirical analysis methods to identify causal relationships. These methods have provided a solid empirical basis for evaluating industrial policy effectiveness.
In addition, economic disparity among regions, which was the major factor determining the creation of the TVA, is also a major issue in Japan today. Needless to say, accelerating innovation is a crucial challenge facing the Japanese economy, which has grown at a very slow pace for over 30 years. Industrial policies may be a powerful tool in resolving this issue.
The important thing is that recent research has not only provided insight into the effectiveness of industrial policy, but also effective methods for implementing such policies. Assistant Professor Howell's article demonstrated that R&D grants are more effective when allocated to younger companies and to companies in new industries.
In addition, Harvard University Professor Philippe Aghion and his co-authors used a dataset of enterprises in China to verify the type of cases where industrial policy was most effective in raising productivity. Their results showed that industrial policies can foster productivity growth to a larger extent in cases where they are designed to promote competition among companies, such as cases where benefits are allocated to many companies within one sector or to younger and more productive enterprises.
When implementing industrial policy, it is important not only to conduct interim and retrospective evaluations of effectiveness, but also to design effective policy implementation schemes or frameworks in advance that are based on relevant insights.