Most professional economists would agree that open world trade increases economic growth and raises living standards. Trade barriers such as tariffs and non-tariff barriers – which include rules-of-origin clauses or sanitary and phytosanitary conditions – reduce countries’ ability to specialise in those goods and services which they are good at, and force them to produce things that they are not good at.
But despite this important insight that dates back to Adam Smith and David Ricardo, economists also know that free trade is not always good for everyone. Industries that are uncompetitive but employ many thousands of people can suffer when trade barriers protecting those industries fall. Many countries protect certain key industries, arguing that they are industries of national security. The classic examples here are military spending or food security. (This can have ridiculous consequences: Gilette argued that its razors deserve tariff protection during the Second World War, ostensibly because soldiers could not go unshaven.) Other industries are protected because they are young and, it is argued, will become more efficient once they obtain certain economies of scale. This is known as infant industry protection. The problem is: many infants never grow up. The South African clothing and textile industry has received government support since the 1930s, and we still pay exorbitant import tariffs on clothes.
But sometimes it does work. As Concrete Economics, a new book by Stephen Cohen and Brad de Long, explains, the United States became the manufacturing hub of the late nineteenth and early twentieth centuries because it was protecting its local industries from cheap British imports. In Economics jargon, their comparative advantage (the thing they were relatively better at making) switched from agricultural goods to manufacturing goods. And with manufacturing came higher paying jobs and more dynamic technological innovation.
It is this same model that the East Asian Tigers followed, copying the basic and later advanced technological products of the West, building a domestic industry behind high tariffs, and once they’ve built up the necessary technological know-how, exported their way to prosperity. Now they are at the technological frontier designing and building new phones (Samsung, Korean) and computers (Lenovo, Chinese) and robots (Honda, Sony, Fujitsu, Hitachi and Toyota, all Japanese firms, have built human robots).
But this strategy did not work everywhere. The evidence for Latin America is mixed: attempts at import-substitute industrialising failed to propel Argentina, Brazil and many other smaller South American countries to prosperity in the same way it did East Asian countries. And in postcolonial Africa it only managed to impose a heavy burden on poor consumers without stimulating any large-scale industrial activity. Many African countries remain incredibly protected – just ask any importer to Nigeria, for example – and this has contributed little to the rise of African industry.
So are open borders good or bad? A new paper by Pable Fajgelbaum and Amit Khandelwal in the Quarterly Journal of Economics gives the standard economist response: it depends. Some consumers buy more tradable goods and are therefore more affected by relative price changes caused by international trade. They find, using a novel methodology, that those consumers who gain most are often the poor, who buy more tradable goods and services. Open borders, they claim, is a very good thing if you are a poor person.
So policy makers are stuck between a rock and a hard place: close borders in the hope that some industries grow beyond infants, at the cost of cheaper goods and services for poor people. Or open the borders and allow the poorest access to cheap goods and services, but with the caveat that some uncompetitive industries suffer injury.
Take South Africa’s dispute over chicken imports. Chicken is the largest protein for poor South Africans. By denying them access to cheap food we not only hurt them but also their children’s ability to consume nutritious protein so critical for early childhood development. We thus perpetuate the cycle of poverty. And by protecting chicken imports, do we really stimulate local economic development in dynamic industries with agglomeration and spill-over externalities? Probably not.
In contrast, we protect the local automotive industry because it not only creates direct jobs but because vehicles support an entire value-chain, from raw material to assembly. Unlike chicken producers, building a car requires vast numbers of engineers and other skilled artisans that may have large (and unexpected and unplanned) spill-overs in related industries.
What made Japan, the first Asian Tiger, so successful was a capable bureaucratic administration that could, with little political influence, judge which industries required support and which did not. Some that received support failed to deliver, and support was quickly removed. We only recognise the successful ones: Panasonic, Kawasaki, Canon.
Wherever protection has failed, it has done so because supported firms gain political influence to protect their support. Bureaucrats are people too – often poorly paid – and find it difficult to challenge entrenched interests of the firms they initially supported. The government bureaucrats that steered Japan’s miracle were not only well remunerated (making them less corruptible) but were also the top graduates from Japan’s best universities. They had the foresight to invest in industries of the future.
In general, then, open borders are likely to be more beneficial than closed ones, especially to those people who are worst off in society. But this is not to say that there is no role for industrial policy. If political influence can be thwarted – and that’s a big ‘if’, especially given the recent revelations of state capture in South Africa – support for strategic industries that have large spill-overs can play an important role in building a thriving economy. The hard questions remain, though: Who picks the winners? And what happens when they fail?
*An edited version of this first appeared in Finweek magazine of 21 April.