WTO 2.0: Governance of 21st century trade

Hausmann and Hidalgo think of economies as collections of “capabilities” and they make a series of assertions about these: i) capabilities determine a nation’s ability to make “sophisticated” products; ii) economies that export many types of products are “sophisticated”; iii) sophistication and wealth do not always go hand in hand, but economies that are relatively sophisticated but relatively poor, have a potential for quick growth. The flaw is that the authors implicitly assume that capabilities can be inferred from exports. This may have made sense in 1980s, but no longer. China, for example, makes very little of the iPad; Chinese workers merely assemble the sophisticated parts that are made elsewhere, so using Chinese exports to measure Chinese capabilities is incorrect.

Formally, the welfare impact of a tariff decomposes into a first-order small terms of trade gain (area R) and a second-order small distortion loss (areas a + b). For an infinitely small tariff, the second-order effects are zero, so the nation gains. The optimal tariff is where the change in R just equals the change in a + b. See Kaldor (1940).

As noted above, various offshoots of trade theory have allowed for this in the past (see the intermediates literature cited in this section). The statement here reflects the mainstream view that it is reasonable to view national exports as being mostly containing the nation’s own technology and factors.

Also see Gereffi (2001) for early examples and the website http://​www.​globalvaluechain​s.​org/​ for abundant recent case studies.

Note that terms of trade effects can also be expressed in terms of the ‘market access’ concepts prevalent among practitioners. A terms-of-trade loss is the ‘price symptom’ of restricting market access. Tariffs reduce market access and exporters naturally react by some combination of reduced sales and lower border prices–in effect, exporters find it optimal to absorb some of the tariff by lowering border prices.

This draws heavily on my unpublished paper, Baldwin (2010).

For example, in 1959 the EFTA nations really wanted free trade among themselves but only on industrial goods. In 1965, the US and Canada really wanted to liberalize bilateral trade but only in the auto sector. GATT rules prohibiting such less-than-substantially-all trade agreements weren’t applied because the concerned parties used the “escape hatch”; EFTA and the US blocked the negative Panel findings.

The list includes standard border measures like tariffs on industrial and agriculture goods, standard measures that could offset the lowering of border measures (subsidies, unfair competition, biased public procurement, onerous customs procedures), and a few behind the border barriers such as investment restrictions, trade-related intellectual property rights and technical barriers to trade (basically standards for industrial goods).

Note that the list of 52 provisions does not include the large number of extremely deep integration provisions in the EU’s Treaties since Horn, Mavroidis and Sapir (2009) considered EU RTAs with third nation.

This may explain why Washington Consensus policies have been so widely adopted, even by nations who have not received offshored manufacturing jobs.

Technically, strict disciplines is the dominate play for the government and given this, invest is the dominate strategy for the firm, so offshoring industrialization is the iterative dominate strategy in a one shot game.

For early thinking on the multilateralization of deep regionalism, see Baldwin, Evenett and Low (2009).

The matrix includes all the nations in the WIOD database (eliminated small nations).

It is important to note, however, that this information is severely incomplete when it comes to East Asia where supply-chain trade is rampant, RTAs are abundant, but participation in the WIOD project is scarce (only Japan, China, Korea, Indonesia and Taiwan). All the major ASEAN nations have RTAs with Korea, Japan and China as well as with themselves, so surely some of these would have blue borders if they were in the data set. The lack of information on Latin American and African nations is less problematic since they have many RTAs but engage in relatively little supply-chain trade.

See Baldwin, Evenett and Low (2009) for an analysis of why many supply-chain trade disciplines are not discriminatory per se.

The WTO classifies SDT measures into six categories: granting better market access to developing countries, safeguarding interests of developing countries, flexibilities and longer transitions, technical assistance; and flexibilities for Least Developed Countries (LDCs).

Technically this is just the effective rate of protection argument.