My experience in the active trade policy world has made me realize that trade policy negotiations have moved away from the text-book comparative advantage intra-industry tariff negotiations to competitive advantage inter-industry non-tariff measures (NTM) negotiations. And that is where the problem of a researcher starts. 1. inter-industry research requires a more detailed level understanding of the input-output matrix of a product which needs to be beyond the 2-digit input-output matrix (for the US we have 4-digit now). Product-level input-output matrix helps in not only overcoming the difficulties in empirically studying the same (for instance, I am always faced with the question of aggregation bias whenever I land up using IO tables) but also help in understanding the input importance to an output industry which helps in understanding the implications of a non-tariff measure imposed along the product-level-input-output matrix. 2. Although researchers have been extremely active in developing NTM database (see, also see), the problem arises in the universe of NTMs which are observable at times (researchable) but in most cases, take shelter in a reciprocity-domestic “welfare” argument. This makes it difficult in identifying its existence, goes unobserved most of the times (unless of course you meet the industry and they rub the fact of it on your face) and hence the difficulty in researching the same.
Despite these problems, the issue at hand cannot and should not be put away. While I continue to make project proposals which looks at the implications of trade policy into an inter-industry context and have long discussions with my fellow researchers on the same, the question that unconsciously comes to an academic economist is: what is the background literature to these proposals?
I am glad that this topic has started to gain traction in the academic world, and synopsis of a new paper on this can be read at http://voxeu.org/article/costs-cascading-trade-protection#.V_zkc6bC59I.twitter.
I would be happy to discuss more on developing both product-level-input-output matrix and the unobservable NTMs, and of course on my proposals of this topic. You can reach me at email@example.com