Should we retaliate against China's new export controls?
The case for free trade is as watertight as ever
China has not only implemented a new export licencing regime on all rare-earths used for advanced chip making, but also on lithium batteries and artificial graphite anodes. Given China’s dominance in the supply chain for these materials, and their centrality as a key input in the production of electronics, this could yield major ramifications for the global economy, should China block exports. Many perceive this as a grave act of escalation, and are calling for retaliatory measures. Is that the optimal response?
There are a plethora of reasons why we should not respond tit-for-tat in my view:
Given the American ban on contracting to Huawei, and their export restrictions to China, arguably this is itself a response to anti-China protectionist policy in Washington. If the US is perceived by China as the aggressor, and hence a removal of it's export controls would foster a cooperative move by China, then de-escalation seems far superior.
Just because China has enacted the capability to block exports of these materials, does not mean that China will actually use this power. In fact, as the Chinese economy is struggling, they will want all the export revenue they can get. Choking off supplies to the rest of the world would not only choke off its revenue from these materials, but also exacerbate decoupling efforts. China of course knows this, hence I see this move as posturing - a deterrent against further US aggression on trade. In other words, with no further protectionist measures, the status-quo will hold, yet if America escalates, then China will respond tit-for-tat. Moreover, given the statist nature of the Chinese economic model, they already had the ability to do this long ago, yet this did not significantly change the behaviour of firms and investors. So we should interpret this move as signalling a willingness to retaliate to further protectionism, as opposed to a concrete change in fundamentals.
The timing seems awfully suspect. Later this month, there is an expected meeting between Trump and Xi. As such, the CCP is clearly trying to gain leverage in any upcoming negotiations. America should use this opportunity to de-escalate, and negotiate for a removal of all export restrictions (and ideally tariffs too, but Trump will not do this).
Most importantly, assume the worst case scenario actually does happen, and China prohibits exports of all rare earths to the West. What is the elasticity of substitution for these inputs? Is the production function Leontief? If not, then the global economy will adapt, albeit with a mild downturn. Indeed, the predictions of Ben Moll et al regarding the energy shock facing Germany turned out to be as accurate as we could hope for. I recall many catastrophic predictions of a major German recession (2008 rates of negative GDP growth) at that time, yet this scenario did not materialise. Just as we were right to ignore the doomsayers then, perhaps we are right to ignore them today. Hence, caution is warranted before retaliation.
In general, Tyler Cowen has implied that although protectionism is bad, it is not uniquely bad compared to other distortionary taxes or levers of industrial policy. If one accepts the desirability of tax and transfer redistribution, and industrial policy, then in principle, there is no reason why one should reject protectionism outright. Indeed, I often hear the argument that tariffs are good, because they are an alternative source of revenue that can be used to cut other taxes.
Leaving redistributive concerns aside [1], as Brian Albrecht highlights, a model is necessary to evaluate such sweeping normative claims, and indeed there is reason to believe that tariffs are a uniquely bad form of taxation. I would add that many key imports used as inputs yield low elasticities of substitution, so this would amplify further the deadweight loss on taxing those.
This yields implications for optimal tariff theory too, as taxing these less substitutable imports would dampen international competitiveness, thereby dampening any gains in terms of trade. Moreover, market share is endogenous, so if misallocation reduces productive efficiency which worsens competitiveness, then the nation in question will lose market share, reducing its influence over import prices. To invoke a Hayekian argument, social planners do not yield all the available information on elasticities, nor on how agents will react, and cannot predict all the general equilibrium effects, with their unintended consequences. This alone should make anyone sceptical of economic defences of protectionism, which are based on faulty reasoning.
Underlying this post is a stark degree of scepticism regarding any forms of protectionism. Contrary to those calling for decoupling, or for retaliatory measures, the example of rare earths merely strengthens the case for free trade. We cannot, and should not, decouple from imports of every input low in substitutability, as the elasticities themselves influence the cost of these measures. Any attempt to distort global trade flows will yield unintended consequences that amplify their costs, whilst the supposed benefits are limited.
I think with tariffs, we will see a small increase in unskilled manufacturing jobs and wages relative to the counterfactual, yet this is outweighed by the effects of automation. In equilibrium, the overall effect on low-income individuals is ambiguous. No one will notice, no one will be happy, and the pessimistic discourse will continue as before. In any case, the sheer number of revisions to the “China shock” thesis of Autor et al 2013 (of which this is just the latest example) suggest that any gains in manufacturing employment will be even smaller than we might expect.