It's astonishing to read the criteria by which this paper judges success or failure. 1/

Industry profitability is treated as a benefit, the lack thereof a cost. The framework is surplus — welfare is consumer surplus + profit — but a broader view of social welfare might (very much should!) treat peristent profit as distortative rent. 2/

Fragmentation and idle ("excess") capacity are inefficiencies, welfare costs, by definition. Neither benefits of competition nor the social value of price elastic response to demand shifts can play any role. 3/

They are admirably candid here: "our model does not feature any market failure; as a result, industrial policy is necessarily welfare-reducing and our results essentially speak to its costs." 4/

The exercise is basically (my words, not a quote from the piece) "from the perspective of a worldview in which policy can only be harmful by definition, here are details of just how this policy is costly." 5/

If that's your worldview, there's no point in learning the details. But if that's not your worldview, you need some understanding about what it is that you mean for industrial policy to accomplish and what you view as benefits and costs for the exercise to be meaningful. 6/

For people (like me!) who you a political-economic support of competition and a flexible capacity to produce as benefits, their analysis is 180 degrees topsy-turvy. 7/

It is perhaps interesting as an exercise to understand, under a partial equilibrium surplus-tallying framework, presuming otherwise perfect markets, where surplus was lost by China's intervention. But it's a curiousity, entirely unsuitable to any kind of normative or policy guidance. /fin