More new growth wisdom: Dani Rodrik and Jason Hwang

A loyal MR reader, Daniel Kahn, directs my attention to Dani Rodrik's latest; here is the bottom line:

Note how different this finding is from the conditional convergence results that dominate the work on economic growth.  The message of the conditional convergence literature is as follows: “what you need to do to converge to the income levels of rich countries is to get your policies and institutions in order.”  [Jason] Hwang’s (forthcoming) results say: “what you need to do to converge is to get a foothold in the goods that rich countries produce.”  The absence of unconditional convergence at the aggregate level must be due, in turn, to structural features of low-income economies.  Poor countries remain poor because they are not producing the kind of goods that will carry them towards riches.

Here are Jason Hwang's papers; he is currently on the job market and will be a big deal in the profession.  This abstract summarizes what will be an important paper...

How does the introduction of new goods affect growth?  While recent evidence has highlighted the role of new goods in raising the diversity and sophistication of a country’s production structure, which in turn matter for growth, little evidence tells us why.  I propose a simple channel of impact relying on two building blocks.  One, there is a convergence force operating at the product level.  The further behind the frontier you are in a given product, the faster you raise quality. Two, new goods are introduced with a greater distance to the frontier than in existing goods.  I construct a Schumpeterian growth model with these features to show how entry into new goods influences aggregate outcomes by determining the range of products in which convergence occurs.  Detailed trade statistics provide strong support for both building blocks of the model.  Using unit values as a proxy for quality, I find that unit values exhibit strong convergence – at about 5% a year – for the great majority of products in the sample.  Also the gap in unit values relative to the world frontier is larger for new goods. Confirming a key prediction of the model, I further show that, holding constant levels of development, unit values are inversely related to measures of diversification and sophistication of a country’s exports. This last finding helps to explain a recently documented puzzle regarding the high sophistication and low unit values of Chinese exports. (To be posted in early November)

Perhaps I will not be convinced (what goods can be produced reflects an unobserved heterogeneity in underlying conditions), but Hwang is worth watching.