Industry policy just might work â shock!

Yesterday I posted an introductory post on industry policy summarising some of the very good reasons to be suspicious of ‘picking winners’. But that’s only one side of the story. Here’s another side. As Fred Astair says in some movie “That idea’s so crazy it just might woik”.

It’s often difficult to untangle the political and the economic objections to industry policy.

The free marketeers attempts to demonstrate that economic development is strongly correlated with free markets and free trade are frequently extremely tendentious as for instance World Bank’s (and following them our own PC’s) 1990s analyses of what drove successful Asian development. That was until the more recent and much more modest effort by the World Bank written up by Dani Rodrik and discussed by James Farrell on Troppo. As Rodrik points out it is certainly true that egregious departures from basic market and macro-economic disciplines lead to disaster, but that does not demonstrate the same case of correlation with much more moderate departures.)

Industry policy â including the idea of picking winners and generating ‘good jobs’ could quite easily be an important part of the success of Japan, Korea, Taiwan, Malaysia and Singapore. (It is obvious that heavy government involvement to promote emerging industrieswas instrumental in all these countries’ economic strategies (and in some cases like Japan’s aluminium industry actually deliberately shrink unpromising industries). The stronger â also pretty plausible claim is that they would have done worse â perhaps substantially worse â without such policies).

Of all the really successful developing countries, I can only think of a couple in which industry policy hasn’t had quite a bit to do with the discovery and expansion of competitive industries. Hong Kong is one. And India is probably the other exception. It’s got industry policy alright but of the worst kind. Its successful industries â like software and back office processing â have not in my understanding been the product of industry policy but have tended to emerge despite rather than because of protection to other industries.

It is also remarkable that aggressive industry policy isn’t just associated with most of the biggest development successes. It’s also associated with the fastest growing countries in the OECD. Ireland and Luxembourg have each targeted specific industries â in Luxembourg’s case mainly finance which occupies an astonishing 30% of its economy â compared with less than 10% for almost everyone else. In Australia and the US at least â and I’d assume it’s similar elsewhere, finance workers are paid nearly twice others’ wages. And Ireland and Luxembourg are the two highest growing countries I know of in the OECD â with their success dating from their success in attracting high value adding industries.

One of the pieces of economic commonsense that is frequently cited in industry policy studies is the idea that industries with high value added per worker and industries paying high wages (economic theory says these should be the same industries) are ‘good’ ones and industries paying low wages are ‘bad’ ones and that industry policy should be about getting as much of the good ones as you can. It’s certainly a crude idea. What if the high value jobs that are added just involve poaching high skilled workers from adjacent industries or firms? I certainly know of a case where the Victorian Government crowed about Toyota establishing a design centre in Melbourne. Local companies were not impressed as the new facility set about hiring its workforce â poaching local designers away from their existing employers. It’s not clear there are a lot of gains there!

Still, in Australia people in finance get around 185% of the wages of those not in finance, if you control for skill (or at least educational qualifications) they still get around 20% more â suggesting there’s rent to be had. And that rent’s been growing strongly over the last two decades and seems to be continuing to grow. Perhaps that’s one of the things that has been making money for the Irish and for Luxembourg and one of the reasons why financial centres seem to get so rich! Markets are fairly efficient, but it never hurts to get near money!

Then there is this NBER paper which has just been published. It goes to the heart of the ‘good jobs v bad jobs’ debate. As the authors explain

The accounting approach used to dispel the ¹ view consists of mechanically computing the fraction of a wage change that can be directly attributed to the loss or gain of employment in high versus low wage paying industries. The result from such exercises almost always indicates that the wage change directly accounted for by changes in sectoral composition of employment is small. . . .

The accounting approach hinges critically on the assumption that a change in employment opportunities in one sector does not affect the wages paid in other sectors, i.e., that there are no general equilibrium effects from shifts in industrial composition. . . . There are many ways to justify the no-GE effects assumption, which is part of the appeal of this approach. The easiest defence is to note that if wages are simply a function of productivity and returns to labour are close to constant, one just needs to assume that changes in industrial composition do not change productivity within sectors to arrive at the conclusion that there are no GE effects.

Beaudry et al, the authors of the study “use US Census data over the years 1970 to 2000 to quantify the relationship between changes in industry-specific city-level wages and changes in industrial composition”.

Our finding is that the spill-over (i.e., general equilibrium) effects associated with changes in the fraction of jobs in high paying sectors are very substantial and persistent. Our point estimates indicate that the total effect on average wages of a change in industrial composition that favors high paying sectors is about 3.5 times greater than that obtained from a commonly used composition-adjustment approach which neglects general equilibrium effects.

One might expect this to happen if higher wages in one industry drive up wages elsewhere through labour market competition. But the study finds that the effects are not transitory. Their persistence could only be explained in a competitive economy if they were associated with productivity improvements. This has long been a justification for various policies which share with Keynesian macro-economic policy that galling quality that the medicine that makes you better is also easy to swallow. Thus, so such arguments run, minimum wages, collective bargaining and industry policy don’t raise unemployment (or, more reasonably, don’t do it much) because the various actors â workers and bosses â behave in such a way that higher wages come to be affordable (workers work harder and train more, bosses invest more and act smarter and perhaps both workers and bosses bargain better because they’re under more cost pressure).

One commentator suggests these things can come about through firms

seeking more productive ways of going about their business â for example, through improved workplace arrangements; restructuring opportunities including strategic alliances, mergers and acquisitions; process innovations and commonisation; and closer integration of production with wholesaling and retailing activities.

Does all this sound too good to be true? Well it does a bit â but then the answer of those who support such policies is only to interfere in the market a little. Clearly you can’t reach nirvana by doubling the minimum wage. (And for my money our minimum wage being much higher than most other OECD countries is too high). But it’s all interesting stuff and I’ll be very interested to see considered reactions to the study.

Oh â and by the way, the PC has endorsed this basic logic. In the early naughties it had a couple of inquiries into whether we should reduce tariffs below their current pretty low level. It said ‘yes’ but it had a problem. The modelling showed that it wasn’t worth it because the little we’d gain from our economy’s improved access to imports would be outweighed by the lower prices we’d get when we expanded some exports. I covered the issue to some consternation of a couple of commenters here. So it started talking about the ‘cold shower’ effect of tariff cuts. Put people under cost pressure and they improve their productivity. When they built this assumption into their model it turned out their preferred policy â zero tariffs â was right all along. Of course as John Quiggin pointed out that meant that the best policy would have been an even colder shower with import subsidies.

In any event, the words I’ve just quoted above on the productivity improvements that come from putting firms under competitive pressure are quoted from the Commission’s report on the car industry in 2002 (p. 144). Don’t expect any conversions any time soon.

1. good jobs[↩]