Friday 7 October 2011

The Globalization Paradox

Have just finished Dani Rodrik's book on The Globalization Paradox. He argues that the world faces a trilemma in terms of nation states, democracy and (deep) globalization. In short, you can only pick two of the three. As Rodrik rules out global government and believes in democracy, he argues that deep globalization will have to go.

I'm not going to provide a review of the book, but wanted to pick up on two things that have interesting parrels to debates on urban economic policy.

First, Rodrik argues that future trade negotiations should focus on legitimising globalisation rather than seeking to deepen it. Can we come up with mechanisms that seek to maintain similar levels of globalisation but allow, e.g., the US to impose constraints where it doesn't like say the environmental or labour standards regimes of trading partners? I think this is an interesting argument even if there is a long way to go on how you might implement in practice (which is presumably why Rodrik calls for more effort to go in to thinking about it). It is premised on the idea that the removal of further tariff and other barriers to trade is unlikely to deliver very large gains (from trade) but will require lots of redistribution. I suspect many of my international trade colleagues would argue that Rodrik overplays his hand on this point - because he is mostly looking at the static gains compared to the redistribution. I have some sympathy with that counter argument.

The point on redistribution raises an interesting parallel with the rebalancing debate that is taking place in the UK. If we achieved spatial rebalancing by improving the economic performance of the North of the UK and having population shift, how big would be the economic gains relative to the amount of redistribution involved? We don't have a good answer for that for the UK - although in the US case Klaus Desmet and Esteban Rossi Hansberg are doing some fascinating research which suggests that the welfare effects from eliminating all efficiency and amenity differences across US cities would be tiny relative to the huge population movements that would result. Efficiency effects of between 0.2 to 1% (from eliminating all efficiency or amenity differences) require 40% of the population to move! Constrained systems (like the one in China) can see welfare gains that are order of magnitudes larger. But these come from movement towards the bigger places, not away from them. I have no idea where the UK would like on this spectrum, but found the parallel with the trade debate interesting.

The second (spatial) parallel that I would draw with Rodrik's argument is what government might do with it's newly negotiated powers that increase policy discretion. Rodrik makes the case for carefully designed industrial policy. Here I confess to being much less sympathetic. It's fine to point to examples of industrial policy where we might argue that the policy turned out well. But that doesn't pass my test for thinking about the average effects of introducing an industrial policy. To do that, you need to take into account all the industrial policies that fail and average out their effects with the successful policies. I am not at all convinced that the balance favours industrial policy. Rodrik might call for better design - but it is terribly hard to know what that might mean in practice. The obvious parallel in spatial policy is what, if anything, policy might do about clusters. As I argued recently, this is an incredibly hard question to answer.