No, Protectionism Does Not Lead To Economic Development

See Kicking Away the Ladder: Development Strategy in Historical Perspective by Douglas Irwin. He reviewed Ha-Joon Chang's book Kicking Away the Ladder: Development Strategy in Historical Perspective. Excerpt:

"Chang’s book is provocative and interesting, but falls short of persuading. Perhaps the biggest disappointment is Chang’s extremely superficial treatment of the historical experience of the now developed countries. He has simply chosen not to engage the work of economic historians on the questions he is raising. For example, chapter one — “How Did the Rich Countries Really Become Rich?” — does not contend with the work that economics historians have done on the topic. Given the broad question posed in this chapter, one might have expected Chang to confront such landmark works as Douglass North and Robert Thomas’s The Rise of the Western World (1973) or Nathan Rosenberg’s and L.E. Birdzell’s How the West Grew Rich: The Economic Transformation of the Industrial World (1986). These works stress the importance of political systems that provide security to economic transactions and economic systems that allow for competition, broadly construed. But Chang does not explain why the lessons from these works are not relevant to developing countries today.

Rather, in chapter 2, Chang elaborates on his contention that “infant industry promotion (but not just tariff protection, I hasten to add) has been the key to the development of most nations … Preventing the developing countries from adopting these policies constitutes a serious constraint on their capacity to generate economic development.” In my view, this statement is erroneous on two counts — that infant industries were the key to economic development, and that developing countries are prevented from adopting such policies today.

Just because certain trade and industrial policies were pursued and the economic outcome turned out to be good does not mean that the outcome can be attributed to those specific policies. Yet Chang does not advance our understanding beyond this “correlation therefore attribution” approach. Perhaps the success of developed countries came despite the distortions and inefficiencies created by their earlier policies because the broader institutional context was conducive to growth.

For example, the United States started out as a very wealth country with a high literacy rate, widely distributed land ownership, stable government and competitive political institutions that largely guaranteed the security of private property, a large internal market with free trade in goods and free labor mobility across regions, etc. Given these overwhelmingly favorable conditions, even very inefficient trade policies could not have prevented economic advances from taking place. (As Adam Smith once commented, the effort of individuals to improve their condition “is frequently powerful enough to maintain the natural progress of things towards improvement, in spite … of the greatest errors of administration.”)

And yet, in Chang’s story, these other things get no credit for America’s economic success; rather, it all comes down to infant industry promotion. Chang writes: “Although some commentators doubt whether the overall national welfare effect of protectionism was positive, the U.S. growth record during the protectionist period makes this scepticism look overly cautious, if not downright biased.”

But, once again, correlation is not causation. Chang produces no evidence that protectionism was responsible for the growth. He does not investigate the various channels and mechanisms by which trade policy affects growth and compare them to other factors leading to economic expansion. He does not undertake a counterfactual analysis to determine the magnitude of benefits and costs of infant industry policies. In the reasoning style of Paul Bairoch, if tariffs were high and growth was strong, then there must be a causal relationship between the two. There is no need to examine alternative explanations, such as whether any effects of tariff policy were swamped by the advantages of other aspects of the American economy. Instead, Chang makes sweeping statements like “It is also clear that the U.S. economy would not have got where it is today without strong tariff protection at least in some key infant industries.”

The implication is that protecting manufacturing industries accounts for the success of rich countries. But Stephen Broadberry (1998) has shown that the United States overtook the United Kingdom in terms of per capita income in the late nineteenth century largely by increasing labor productivity in the service sector, not by raising productivity in the manufacturing sector. Broadberry’s research is not obscure, yet Chang makes no note of it.

Attributing the economic success of various other countries to their trade and industrial policies alone grossly inflates their role. In Europe, Broadberry and others have showed that growth was related to the shifting of resources out of agriculture and into industry and services. Yet trade policies may have slowed this transition for some countries. Britain industrialized with the textile industry in the late eighteenth and early nineteenth century, but the Corn Laws during this period kept more labor and capital resources in agriculture, not industry. Similarly, to the extent that Germany’s tariff code protected agricultural goods (where it was a net importer), it actually slowed that transition and may have retarded growth in the late nineteenth century.

A broader problem afflicts Chang’s approach — sample selection bias. Chang only looks at countries that developed during the nineteenth century and a small number of the policies they pursued. He did not examine countries that failed to develop in the nineteenth century and see if they pursued the same heterodox policies only more intensively. This is a poor scientific and historical method. Suppose a doctor studied people with long lives and found that some smoked tobacco, but did not study people with shorter lives to see if smoking was even more prevalent. Any conclusions drawn only from the observed relationship would be quite misleading. Chang also overstates the degree to which developing countries today are prevented from pursuing interventionist trade and industrial policies. Trade agreements such as the General Agreement on Tariffs and Trade (GATT) pose few barriers to countries that wish to pursue activist trade policies, and indeed many countries did so during the years when import substitution was the rage among developing countries in the 1950s and 1960s. Article XVIII of the GATT allows governments to undertake trade measure to promote development, including the promotion of selected industries. Many countries are choosing not to do so because their past experience with such policies has not been successful.

No economic historian will deny the importance of lessons from history in guiding policy today. The question is “which” economic history is relevant. (This point was raised in some insightful comments on Chang’s work by Ken Sokoloff at last year’s EHA meeting.) Which historical experience is most relevant for developing countries in Asia, Latin America, and Africa today — the perceived failure of state-led development and import substitution in those countries in recent decades, or the experience of Britain and the United States in the nineteenth century? Certainly China and India have answered by saying that their past policies of inward-looking socialism have failed them. Both countries have done better over the past decade or two by shedding heavy-handed government involvement in regulating the economy and allowing a greater role for market forces, even though they have not embraced every aspect of the “Washington Consensus.” In particular, China and India have decided to become much more open to world trade and investment and have reaped benefits by exposing long protected “infant industries” to global competition."