Download PDF

Successful industrial policy must account for political economy

Once relegated to the junk heap of economic ideas by mainstream policymakers, industrial policies, or state actions to change the composition of economic activity, are making a comeback. In the United States alone, major new industrial policies grapple with the green energy transition, geopolitical competition, and supply-chain resilience. And the US is not alone; it’s part of a global renaissance of industrial policy.

Industrial policy is back in advanced economies, but so are questions about its merits, drawbacks, and practicality. Yet these debates don’t address the wide variation in global practice, why policies succeed or fail, or which policies are feasible in the real world. Although new literature has started to update our empirical understanding of these policies, we argue that this “new economics of industrial policy” (Juhász, Lane, and Rodrik, forthcoming) calls for serious consideration of the political forces behind policymaking.

Politics imbues industrial policymaking, sometimes more so than in other areas of economic policy. Since industrial policies have concentrated benefits and diffused costs, their allocation is often politically fraught. They may also lead to rents that are subject to lobbying. Transformative policies can be politically controversial and may threaten incumbents who rely on the economic status quo.

Although market failures and economic constraints may shape policy choices, so do policymakers’ political incentives. Thus, economics alone cannot explain the vast differences in experiences with industrial policy. The industrial policies that go into effect are those that correspond with our political world, yet modern political economic analysis of this area is sparse.

Optimists and skeptics

New empirical work in economics, which, along with Dani Rodrik, we call the new economics of industrial policy, shows that some episodes of industrial policy have delivered large, positive, and, at times, transformational effects.

The Heavy and Chemical Industry push under President Park Chung-hee in the 1960s set out to transform South Korea into a heavy-industry powerhouse—a proposition so fantastic that no external funder, including the World Bank, was willing to finance it. This initiative drove increased output and export development in targeted sectors, shifted comparative advantage toward these same sectors, and made the economy better off—just as policymakers had envisaged (Lane 2022; Choi and Levchenko 2023).

Postwar Italy pursued a decades-long massive industrial policy aimed at jump-starting development in lagging southern regions. The policy launched durable clusters of economic development both in targeted high-skill manufacturing jobs and the knowledge-intensive service jobs that emerged to support them (Incoronato and Lattanzio 2023). It is estimated that the policy raised national industrial production, which suggests that it did more than merely shift production from untargeted to targeted areas (Cerrato 2024).

Yet not all industrial policy has the scale and scope of these efforts. Case studies from Latin America show that much-smaller-scale industrial policies contributed to export market success (Sabel and others 2012). Examples include cargo flights by state-owned airlines to transport flowers from Colombia to the US export market, as well as collaboration between private growers and research and extension services in agriculture to bring soy cultivation to the northern savannas of Brazil. Similar proof-of-concept demonstration projects show signs of success in Africa (Bienen and Ciuriak 2015). One example is a multipronged policy to promote cut flower exports in Ethiopia.

The new economics of industrial policy has revealed its potential, but skeptics rightly point to the many flops that litter development economics. The debacles of industrial policy in postindependence African countries and the disappointing performance of Southeast Asia’s “Look East” policies have informed thinking about government failures: interventions that introduce more inefficiencies and distortions than they resolve. Where market failures justified industrial policy intervention, government failures repudiated them. Concern surrounding government failures and industrial policy coalesced at the crest of big development in the 1970s. After decades of enthusiasm, developing economies found themselves, in the words of Anne O. Krueger, “mired down in economic policies that were manifestly unworkable.”

The risks and failures around industrial policy are real. However, economists’ skepticism has translated into strong and deterministic claims. If government failure is an endemic feature of industrial policy, there is little reason to pursue it. Nor are there reasons to question success. At the extreme, economic pessimism has culminated in whole cloth, impossibility theorem–style rejections of industrial strategy. Some scholars have argued that industrial policy is impossible—best exemplified by economist Gary Becker’s 1985 assertion that “the best industrial policy is none at all.”

For decades, much mainstream economics embodied varying degrees of this view. The debate over whether to pursue industrial policy left little room for understanding of the conditions for success, much less the means. Political economy may lie at the heart of mainstream economics’ traditional theory of failure, but its deterministic nature rules out a consistent account of success.

Mind the constraints

Yet taking politics seriously helps us understand the wide variations in industrial policy’s successes and failures. On one hand, successes were supported by the domestic political environment. Sound policies were politically feasible, were favored by those with power, and worked within the parameters of the state’s administrative capacity. On the other hand, policies that were incongruent with the political world at the time failed. Restated in the language of modern political economy, government failure is most likely when industrial policy choices violate political economy constraints.

In practice, industrial policy is constrained by both politics and state capacity. Policy decisions are shaped by political institutions and those with political power and by their underlying incentives. Societal conflict means that policies implemented are often far from ideal in economists’ eyes. Even growth-enhancing policies may not be politically viable. For example, new export-promotion policies may threaten both incumbent industries reliant on import protection and their political benefactors. The most efficacious policies may be those facing the steepest political resistance.

State capacity is manifest in state administrators’ ability to implement policies in the real world. Even if fortuitous politics promote the adoption of good industrial policy, a government must still be able to implement it. For example, the successful export policies of one environment may entail a platoon of skilled administrators, detailed data on export performance, and more. Implementing such policies in a low-capacity environment may be entirely impractical without more investment in administrative capacity.

Through the lens of these governance constraints, the East Asian miracle was as much about the political environment as it was about the policy mix. South Korea’s political environment in the 1960s and 1970s supported outward-oriented industrial policies and the controversial reforms they required. The country’s all-out export-based industrial policy was forged under external duress; enduring military threats from North Korea aligned political and industrial elites, and recurring balance of payments problems made early import-substitution industrial policy unsustainable.

Such conditions fed political demands for new policies to spur industrialization and a sustainable source of export revenue. South Korea’s political climate allowed it to pursue a total export push that entailed devaluation and investments in competitive bureaucracies.

Thus, the East Asian miracle is miraculous not because of the export-oriented industrial strategy per se, but rather because of the fortuitous political environment that allowed it to emerge. The political economy constraints on large-scale, sweeping policies—such as those behind the East Asian growth miracle—can be binding in many cases. Duplicating policies without understanding their compatibility with local political constraints often leads to failure. The incomplete adoption of East Asian industrial policies in Malaysia and Thailand is a case in point.

Industrial policies, thus, should be considered within the constraints imposed by the political world. Best practices can be gleaned from others, but understanding the politics at play is critical.

Lessons for policymakers

Given this context, we highlight three key lessons for success with industrial policy.

First, policymakers should carefully evaluate how industrial policy aligns with the domestic political environment. They must consider who will benefit, who stands to lose, and how political incentives support good policies. It is important to consider how policies implemented today will change the future political environment and which are most likely to hold up across political cycles.

Second, countries’ ability to implement various industrial policies differs vastly, in ways beyond direct political constraints. Such policies must be tailored to administrative and fiscal capacity. For example, developing economies may not have the fiscal capacity today to implement green industrial policy with the tools currently deployed in advanced economies. Since there is no one-size-fits-all approach, we should expect industrial policies to differ across countries.

Third, industrial policy almost always requires investment in administrative capacity. This was true in South Korea in the 1960s and 1970s, when the government centralized bureaucracies and invested in administrative capacity, and it is true today. For example, the US Department of Energy Loan Programs Office has expanded to disburse 10 times more funding since enactment of the Inflation Reduction Act. (The office also claims that its administrative processes and institutional safeguards are much improved since the ill-fated decision to fund Solyndra over a decade ago.) 

Overall, we agree with the critics of industrial policy who say that ignoring politics is perilous. A pessimistic reading of this criticism is that the political stars must align—and such circumstances are rare. An exceptionally conducive political environment may be necessary to pursue the scale and scope of policies used across postwar East Asia, but that is not the point. Carbon copies of these policies are not necessary; smaller-scale successes across the globe attest to that.

To us, this suggests that the risks of government failure can and have been overcome in the past. When industrial policy is chosen to work within the local political and governance constraints, and when the state aggressively invests in building the required administrative capacity to deploy and monitor industrial policy, the odds are stacked for success.

There is still much to learn about the “how” of industrial policy. The literature is only beginning to explore empirically the diverse and rich experience of industrial policy; more measurement and evaluation are essential. Economists and policymakers must focus not only on the market failures and the policy mix, but on the politics as well.

Opinions expressed in articles and other materials are those of the authors; they do not necessarily reflect IMF policy.