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 [good jobs] 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.

This entry was posted in Economics and public policy, Politics - national. Bookmark the permalink.

16 Responses to Industry policy just might work â shock!

  1. derrida derider says:

    It’s not that industry policy can’t work, or hasn’t worked in other times and places (though of course it has had plenty of spectacular failures too). It’s the specifics of what “industry policy” is taken to mean in Oz. In Oz it means transferring resources to industries in which we have no hope of generating an eventual comparative advantage. Pouring money into, eg, the car industry is just pissing it up against the wall.

  2. observa says:

    Or is it the problem of diminishing returns. Picking winners in advanced developed economies simply has such poor prospects, it’s not worth it and there is probably better returns from lessening red tape, general infrastructure investment, reducing taxation disincentives and the like. The dramatic returns in IP in LDC’s may well be due not so much to picking winners, but the fact that this obseved process is a symptom of pro-business, facilitative govt overall, in contrast to the past. When the miracle occurs like Japan, inevitably the old recipe no longer works, due to diminishing returns.

  3. Patrick says:

    Hey, I think we can all hold hands and dance in a circle now – we do everything Observa says and call it:
    Industry Policy for the Finance Industry!!


  4. observa says:

    Perhaps Patrick, the major flaw in IP boils down to, if you believe the dopey bastards are no good currently at picking winners, then why would you make a bunch of them bureaucrats with lousy incentives to improve the result?

  5. Paul Frijters says:

    its a mistake to hold up Ireland and Luxembourg as examples of great industry policy. Both effectively free-ride off their close neighbours in terms of either bank secrecy or capital taxation. ‘Beggar thy neighbour’ may work for small countries with direct access to big markets. It doesnt work for big countries.
    Australia is a small country though. Maybe in the future Australia can choose to become a tax-haven for the South Asia region? The ‘Switserland of the South’? We could probably pull it off.

  6. Patrick says:

    Er, I think Singapore has tried and liked the idea, not to mention Hong Kong! So sure, but there is competition in that market.

    But your point is quite valid – as is the more general point that N Gruen’s industry policies that work aren’t ‘industry policies’ as anyone in the world understands the term. Hence our suprisingly similar tongue-in-cheek contributions.

    Which he actually makes a fairly good case for.

  7. Fred Argy says:

    Nicholas, I think there is some validity in the argument that government support of high-value, high-wage industries would, through labour market competition, push up the general wage level and that this in turn would put firms

  8. Yes Paul,

    There’s certainly some successful tax competition in Ireland and Luxembourg’s success.

  9. James Farrell says:

    The notion that ‘high-value’ industries should be a policy priority is never easy to sustain.

    If a good has high value-added per worker in a given country, this must be because either (1) its production uses an exclusive technology or resource that earns monopoly rents, or (2)the labour is itself expensive. In the case of a resource or an established technology, the country will have a cost advantage anyway. In the case of a technology that is not yet realised, an advantage might be achieved by means of astute investment in R&D. If the benefits can be captured by an individual firm, the profit motive will do its job and there shouldn’t be any need for subsisdies. On the other hand, if somehow the technology can be kept within the country but not limited to a particular firm, then in principle R&D subsidies could contribute to the creation of a high value added industry, until such time as the technology is copied elsewhere and the price falls to match the cost of production.

    That leaves the case where labour is itself expensive. To the extent that this is due to a high skill component requiring years of education or training, such an industry has no obvious advantage over others. In some cases, however, the work involves native talent as well as skill, so the workers in question earn monopoly rents. A policy to promote such industries creates more for such people to enjoy their power, but has no broader social benefit – it is actually a form of middle class welfare.

  10. Patrick says:

    Um, what about plain adding value? Ie capital markets advisors might be highly paid because they add a lot of value (positive sum, ie to all parties) to their clients?

    The only ‘rent’ is that of knowing people and a fairly accessible skills set, which hardly seems monopolistic.

  11. Patrick says:

    Sorry – the relevance of which is that anyone can have capital markets industries if you attract them with favourable conditions, and they create genuine broader social advantages from higher incomes sustaining greater local industry such as restauration, retail, etc, which are great ‘first-rung’ jobs (although sometimes only-rung, which is sometimes a worry), and gardening etc (which is quite lucrative).

    Then their kids go to school which pushes more money into schools, and improves their facilities etc and gives them money to offer scholarships, ditto for Unis, etc.

    Finance and capital markets even creates lots of lefty-bonuses like funding for operas and ballets :)

  12. James Farrell says:

    …lots of lefty-bonuses like funding for operas and ballets.

    Where do I sign?

  13. James,

    Like you I’m skeptical, and one of the (many) grounds is – as you point out – it’s middle class welfare. On the other hand I wonder what you make of the NBER study? And, I’m quite happy with the idea that there are rents in the labour market for finance. My intuittion (and the empirical work) suggests a reasonable kind of case for this. It goes along these lines.

    There are rents in finance because of assymetric information, economies of scale and various other things. Being close to money is a good place to be if you want some. Some of the rents earned by finance companies get shared with the workforce. (Ask people in Macquarie Bank). So it might be worth it to compete for some of it – but it will certainly benefit the (upper) middle class. My response to this is to say that that’s yet another reason why I don’t want to reduce the progressivity of our tax system.

    In fact I think it’s not that important to hand subsidies over to finance – but it is important to ensure that we regulate with sensitivity to the ability of our finance industry to export. The financial entrepots all do this. By contrast we have a finance sector that looks very competitive in terms of cost and capability but it exports less than 3% of its output! I think that’s because we have run tax and financial regulation systems in a way that’s pretty oblivious to export issues. We’re just starting to turn that around.

    Declaration of interest – Lateral Economics is currently consulting on these issues.

  14. Patrick says:

    I think the progressivity of our tax system is partly, albeit not per se, responsible for that.

    In England foreigners are not taxed on their overseas income – we should try that. Also the CGT exemption is good, and was overdue, but we are still fairly high-taxing in company tax.

    But we have significantly improved our international tax – the necessary trade-off should be, since the improvements have effectively broadened the base in areas like fx, to reduce the load.

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.