Critics of industrial policy claim that state interventions into the market lead to inefficient subsidies, protectionism, or the “picking of winners.” Kartik Akileswaran and Rohan Sandhu argue that when industrial policy is instead viewed as an iterative, dynamic process between the private and public sectors, and undergirded by the state’s capacity to play this role, it leads to a more sophisticated understanding of industrial policy’s manifold historical manifestations and benefits.
Recent commentary has belabored the claim that industrial policy is back. The truth, however, is that governments around the world have always provided subsidies to spur private investment. More recent is the recognition that the state should play an active role in shaping markets and that the way it approaches economic development requires recalibration.
Even as practitioners and researchers attempt to find the most effective mechanisms to design and implement such policies, critics in the commentariat have already declared industrial policy to be doomed. Critiques like these gloss over the evolution in thinking about industrial policy, continuing to view it narrowly through the lens of tax breaks and financial incentives aimed at “picking winners” among large firms, or favoring protectionism over free trade. Even industrial policy boosters sometimes take a similarly narrow aperture.
This framing mischaracterizes industrial policy as a one-time static intervention rather than a dynamic process based on learning and iteration. This idea that industrial policy is a process and not a specific intervention, is itself hardly new. The title of Dani Rodrik’s 2009 paper “Industrial Policy: Don’t Ask Why, Ask How,” is self-explanatory. Tim Bartik, Mariana Mazzucato, and Charles Sabel have similarly reinforced the need to think of industrial policy in terms of providing a range of public inputs through governance structures that enable iterative public-private partnerships over time.
The disconnect between the commentariat and actual practice is perhaps a function of grammar. The term “industrial policy” doesn’t quite have the action orientation that the actual practice of industrial policy entails. Chris Hughes and Peter Spiegler’s recent rebranding of industrial policy as “marketcrafting” is more fitting, describing the process in the grammar of action. It also explicitly calls attention to the understanding that the state must play a more holistic role in shaping the economy, particularly to generate broad-based gains and investments in critical sectors. Industrial policy is a process of the state enabling firms to discover what they can competitively produce, rather than the outdated notion of “picking winners.”
The implication is that governments must inculcate a range of capabilities to play this role effectively—from program design and delivery to monitoring and re-design. At the core of the state’s role in shaping markets are mechanisms for public-private coordination (Table 1). The logic is that firms have more information about what they need to enhance their productivity than do governments—such as cost structures and barriers to capital—and some of these things may require public inputs. To make informed choices about whether and how to intervene, governments need to create processes and platforms that enable them to take the “firm’s-eye view” and facilitate public-private learning.
One example of this is Peru’s Mesas Ejecutivas, the literal translation of which is “Execution Tables,” but a more apt English name might be “Working Group.” Mesas Ejecutivas are sector-specific convenings of public and private stakeholders that have resulted not only in dialogue and learning, but also real-time problem solving. As Piero Ghezzi, the architect of this process, highlights, the term “ejecutivas” was bestowed by the private sector, which saw the government actively execute on business and economic concerns. For instance, as Peru tried to grow its aquaculture sector, the mesas revealed that the absence of a regulatory authority and specific laws around fishing and fish farming were key constraints. Similarly, existing entities that governed labor and environmental standards didn’t have aquaculture-specific rules. Such learning enabled the government to take concrete steps, like the creation of SANIPES, the sanitary authority for aquaculture.
Through the mesas, the government not only broadened participation of the private sector beyond the handful of big firms already represented in chambers of commerce, it also, at the outset, took cash subsidies off the table. Colombia’s Productive Transformation Program is another example where this strategic choice served a public-private collaboration mechanism well. Removing subsidies from consideration directly reduced opportunities for blatant rent-seeking by private actors. More importantly, it enabled the government to focus attention on the tangible barriers that firms faced in increasing productivity and creating jobs.
The ability to play this barrier-bashing executive role is predicated on governments addressing a range of internal silos and coordination problems. Notably, there is no singular “public” in the public-private equation described above. Government, far from being a monolithic entity, is a network of organizations with disparate (though sometimes overlapping) missions and functions, often suffering from deep coordination failures. As industrial policy increasingly is about a range of public inputs—workforce development, access to finance, small business assistance, R&D, and entrepreneurship training—we need improved horizontal coordination across sector-specific government departments. In South Korea, a monthly “Expanded Meeting for Export Promotion,” chaired by the president, served as a platform for inter-ministerial dialogue and coordination in the late 20th century. A range of other examples underscore the importance of creating linkages between technical and vocational education and training (TVET) and labor demand by industry. Singapore’s National Manpower Council, the Vietnam–Singapore Technical Training Center, and the Penang Skills Development Centre in Malaysia are all cases of such intersectoral collaboration involving different departments.
Aligning national, state, and local departments and agencies—vertical public-public coordination—is just as important. This is evident in the context of place-based policies like the CHIPS & Science Act and the Hydrogen Hubs Programs in the United States, where a bedrock of local infrastructure and state government support will be critical in bringing these federally-defined policies to life. For instance, semiconductor firms will have to work closely with local community colleges, economic development organizations, and unions, to support workforce development efforts. Similarly, the CHIPS Act’s Recompete and Tech Hubs programs—targeting, respectively, distressed regions and potential regional technology centers—call for regional consortia to work together to design strategies to spur local economic development.
This approach of federal policy relying on local and contextual problem-solving is well-encapsulated by Charles Sabel and David G. Victor’s call for a local and experimentalist approach to address the climate crisis. Their model of experimentalism involves an organizational apparatus that balances top-down decree with bottom-up innovation: This framework enables mutual correction between “ground-level problem-solvers and those orchestrating their efforts.” Matt Andrews and Peter Harrington bring these concepts to life in a case study about a time-bound and locally centered discovery process, designed to understand and alleviate the constraints in Albania’s struggling textiles sector. While Albania had a Delivery Unit situated within the prime minister’s office to focus on broad policy implementation and performance, this approach complemented that effort with the recognition that top-down solutions like “steps taken to improve the externally defined, globally generic Doing Business indicator scores would not address the peculiar, local problems ailing Albania’s manufacturers.” Instead, they learned that the problem would require several small interventions: funding new technology, expanding training opportunities, simplifying hiring processes, et cetera—as opposed to one grand solution.
In practice then, industrial policy—or marketcrafting—involves a series of incremental steps. It is fundamentally iterative in nature, rooted in a dynamic and ongoing relationship between firms and various levels and departments of government. The questionable uses of public funds that directly subsidize firms and are subject to abuse should undoubtedly be scrutinized. But that doesn’t mean we give up on industrial policy and relegate the state to a minor—and ultimately inadequate—role.
We need to finally move beyond the decades-old, fruitless debate around “picking winners” through subsidies and tax breaks, which has undermined state capacity to engage in industrial policy, properly defined. Instead, we should ask—and answer—how governments can effectively play the role of detective, convenor, coordinator, and financier at various points over time to facilitate the emergence of new thriving economic activities, as well as jobs and productivity for workers.
Table 1: Types of Industrial Policy
Note: Effective industrial policy requires a range of coordination—within government and between government and the private sector. The table summarizes these.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
