The right way to do industrial policy

It’s a question nearly as old as government itself: when, if ever, should the state put its thumb on the scale to favor a particular industry?

Governments have nearly always been hot on the idea. To raise revenue for the empire, China’s Han Dynasty rulers set up salt and iron monopolies. To build the “wealth, safety, and strength” of medieval England, the government established Navigation Laws decreeing that all imports and exports from and to the colonies must travel on English-built ships. To protect the “national independence,” of the newly formed United States, Alexander Hamilton stacked the deck in favor of U.S. manufacturing.

Policy experts and scholars have been split, nearly throughout. Roughly 30 years ago, however, the World Bank took a position that helped stigmatize the idea in the Information Age: outside of a few exceptions in northeast Asia, we said, industrial policy is usually a costly failure. “The prerequisites for success,” our 1993 report on the East Asian Miracle noted, are “so rigorous that policymakers seeking to follow similar paths in other developing economies have often met with failure.” One set of preconditions was especially elusive: “macroeconomic stability and low inflation.”

That advice has not aged well—it has the practical value of a floppy disk today. Today, with the publication of Industrial Policy for Development: Approaches in the 21^st Century, we are updating the recommendations for a global economic landscape that would be unrecognizable to 1990s-era policymakers. The world’s average income per capita today is nearly double what it was in 1993, meaning the easiest avenues to prosperity have already been exploited. Educational attainment levels are substantially higher, average inflation is lower, and—with a few exceptions—the quality of national macroeconomic management is better.

Under these conditions, our analysis finds, industrial policy is far more replicable than previously thought—and it should be considered in the national policy toolkit of all countries. On the surface, that recommendation might seem moot: the use of industrial policy across the world is at record levels. A review of the national economic-growth strategies of 183 countries conducted for our report found that all target at least one industry, with developing countries engaging in the practice more indiscriminately than advanced economies do.

Yet governments in developing economies regularly ask us: are we getting it right? Last year, 80 percent of World Bank country economists reported that client governments sought their advice on how to use industrial policy more effectively. On the basis of the evidence amassed in recent decades, our view is that governments in developing countries are botching the job far too often—but not because industrial policy itself is the wrong choice. It’s because governments usually resort to blunt instruments, opting for the bludgeon of sweeping tariffs and subsidies over the scalpel of industrial parks and skills-development programs. It’s because they see industrial policy as a kind of magic bullet: nothing else is needed.

Consider the evidence. The 25 poorest countries—with per capita incomes of less than $1,200 a year—are the heaviest users of tariffs, averaging a 12 percent tariff rate, more than other developing economies and twice the rate of high-income economies. The 54 upper-middle-income countries—with incomes ranging from about $5,000 a year to about $14,000—are the biggest users of business subsidies, which now stand at a record 4.2 percent of their gross domestic product (GDP). The blast radius of these instruments is inevitably economy-wide, far beyond the targeted industry. They raise costs for nearly all citizens and businesses, they slow job creation, they invite retaliation from abroad—and they are notoriously difficult to unwind. Larger, wealthier countries can ride out the consequences. Smaller and poorer countries cannot.

All countries would be better off with a more pragmatic and precise approach. By itself, industrial policy has seldom been a game-changer: the record shows that, even under ideal conditions, it results in an average gain of just 1 percent of GDP. Still, in an era of persistently sluggish growth, every option to boost growth ought to be considered. Industrial Policy for Development offers the first comprehensive 21^st-century framework for industrial policy. Call it a feasibility framework: it argues for starting with a light touch and dialing up as state capacity and resources grow.

Three national characteristics define what type of industrial policy a government will be able to pull off: first, the size of the domestic market; second, the government’s capacity to use the chosen instrument of industrial policy; and third, its budgetary room for error. Those traits determine which of 15 possible industrial-policy tools will work best—ranging from industrial parks and job-skills training all the way to import tariffs, production subsidies and competitive exchange-rate devaluations. These tools are a mix of sharp and blunt instruments, and their effectiveness depends on progressively higher levels of market size, government capacity, and fiscal space.

Resorting to industrial policy, however, does not relieve governments of basic housekeeping duties—the obligation, as our 1993 report put it, to “get the policy fundamentals right.” That includes a healthy and educated workforce, a strong infrastructure for transportation and energy, and a sound macroeconomic framework. It’s not always necessary for countries to have these virtues in place before they initiate industrial policy. But long-term success will depend on them, because industrial policy cannot be a permanent solution. To the extent that governments use it to buy time, they should commit to getting the fundamentals right while the industrial policy is in effect—ideally within 10 years.

It is foolhardy to think that a question involving statecraft nearly as old as the state itself has an answer that is at once correct and concise. But let me hazard an attempt. 

• First, a government that is contemplating government action to influence a business activity that it deems pivotal for economic development should never convince itself that industrial policy can be a permanent substitute for the institutional arrangements needed to ensure macroeconomic stability and a balanced business climate.
  
  
• Second, it must self-critically assess whether the public services that are necessary inputs for the desired business activities are sufficiently provided and—if they are not—to provide them.
  
  
• Third, it might contemplate using taxpayer funds to provide incentives to businesses to boost production or accelerate innovation—provided it has the fiscal and administrative wherewithal to ensure that these monies will not be misappropriated.
  
  
• Finally, governments in small economies with limited administrative expertise and fiscal space should resist the temptation to intervene in foreign exchange markets or provide general tax exemptions to help businesses compete on world markets—because such macroeconomic interventions are invariably self-defeating. 

Industrial policy, in short, requires a judicious blend of public institutions, inputs, incentives, and interventions. This blend is now within the reach of more than just a handful of East Asian high performers. Yet no government—poor, middle-income or rich—can afford to ignore the lessons of history: when it comes to industrial policy, the government’s reach can easily exceed its grasp.