In defense of industrial policy


This page in:

Like others, I have been skeptical about industrial policy in Africa, where the government selects certain industries for support in order to trigger a process of structural transformation. It’s been tried before—with disastrous results. 

The selected industries were captured by political elites who continued to receive subsidies without generating anything close to labor-intensive growth (the Morogoro shoe factory in Tanzania never exported a single pair of shoes). Furthermore, most of the constraints to industrial growth in Africa are man-made: policies or regulations that stand in the way of poor workers’ employment prospects. 

Africa’s high transport prices are not the result of poor-quality roads, but mainly due to monopoly prices charged by trucking companies, who benefit from regulations that prohibit entry into the trucking industry—and block any attempt at deregulation. By targeting certain industries for assistance, industrial policy doesn’t address these government failures and may even divert attention from them.

A recent study by my colleagues Hinh Dinh and Vincent Palmade on light manufacturing in Ethiopia confirms the point that the constraints to the apparel and leather goods industries are largely existing policies and regulations—trade policies that inflate input costs by creating local monopolies, and land and financial regulations that favor large firms. But they also show that, should Ethiopia remove these constraints, it could expand employment in these industries by two orders of magnitude. 

Whereas Ethiopia has 9,000 workers in apparel and 8,000 in leather goods, the comparable figures for Vietnam—a country the same size as Ethiopia—are 1 million and 600,000 respectively.

This is the best case for industrial policy that I have seen. By focusing on particular sectors and showing the employment benefits of addressing some of the government failures, there is a better chance that the government will undertake these reforms. 

When the same reforms are presented as “good things to do” because they are “removing distortions”, governments often resist—partly because they are not sure of the potential benefits of implementing politically-sensitive reforms. But when presented with hard evidence of concrete benefits in specific sectors, it is easier to make the case for the reforms.

Note that this case for industrial policy is almost the opposite of the traditional case, where people identify a “market failure” and then advocate government intervention to correct it. Most of the problems in Africa have to do with government failure. But these are much harder to correct, because politically powerful interests may benefit from keeping the distortion in place. The light manufacturing study provides a possible way of out of this difficulty.


Shanta Devarajan

Senior Director, DEC and Acting World Bank Group Chief Economist

Join the Conversation

January 17, 2012

At this time in history, there are a lot of instruments in place put, to harness the development potential of our african nations… Not only those tools are available to us when it comes international cooperation, but domestically there are a lot that has been done to put the right concertation mechanisms in place, ready to get along with an entrepreneurship model based on sound investment strategies…
If to date development has not taken place as it was intended to be, I would point the donor community for not being proactive or at best flexible in adapting those mechanisms on their sides to the prevalent socio-economic realities in Africa (ie illiteracy rates, gender disparity, digital divide, access to water and basic infrastructure services…the list goes on and on)...

To this very juncture, with regards to all the woes this world is dealing with in the food supply front, energy and world peace concerns, only a shift, a genuine shift in the development rationale can ease the whole burden we face…

Our farmers on the ground, our entrepreneurs, along with our civil societies in Africa, have always been ready, far ahead of local government bureaucratic mannerisms… but to those who make the local informal work, it is unthinkable to expect them to have to comply with rigorous accountability criteria dictating conventional market approaches… We need minimal accountability systems in place to allow such potential relevant to our economy to thrive for the benefit of the whole…

Any other way would turn such sparse entities into workers of the big companies grabbing the lands and resources in Africa in the guise of a mistaken new world order…

Wake up donors and partners in development…

Intl. Partnerships

January 17, 2012

Thanks for your comment, Demba. I agree that development policies should be tailored to the socio-economic realities of Africa, but I don't agree that the problem is making informal workers "comply with rigorous accountability criteria dictating conventional market approaches." The problem is that politicians are not accountable enough to citizens, especially poor citizens. That is why we see subsidies that go mainly to the rich, and infrastructure that benefits urban residents, whereas it is the rural poor who could benefit the most. This is the government failure we need to correct if the informal sector in Africa is to grow.

Andy Mold
January 20, 2012

There is a bit of a logical inconsistency in the arguments here, isn't there? The blog claims that "Whereas Ethiopia has 9,000 workers in apparel and 8,000 in leather goods, the comparable figures for Vietnam—a country the same size as Ethiopia—are 1 million and 600,000 respectively." and that by 'removing government restrictions', the Ethiopian government could 'double' the size of the employment in these industries. Which would take us to 18,000 in apparel, and 16,000 in leather goods in Ethiopia. It wouldn't seem like government intervention is the main part of the story, if this account of the differences is to be believed. The difference in employment, once the 'restrictions' are removed, is still 982,000 in apparel and 584,000 jobs in leather goods that needs explaining! Perhaps a more convincing explanation could be put forward?

T.V. Somanathan
January 18, 2012

This is a very interesting insight. A government may find it much easier to remove distortions if there is a clear domestic constituency which demands their removal and if the removal were part of an 'industrial policy' meant to promote growth in specific sectors with credible prospects of job and income creation. Often the political economy of 'distortion removal' is itself distorted by the fact that the losers from distortion removal can clearly identify themselves (a la Nigerian fuel subsidy) while the beneficiaries are unidentified and do not even know that they might benefit, leave alone how much.


January 20, 2012

As this discussion advances, as much as I agree with you all on the fact that lack of accountability and democratic governance setback development in Africa, I also like to shed some positive light into the need to focus investment in areas with the highest potential. Even though evenly spread development is critical both for long-term economic development and stability, by emphasising on social equity, a country's resources might simply be too thinly spread out that in the end can slow down progress in promising areas. These areas and sectors with the highest potential for growth can be the drivers of national growth and the lagging areas can be supported by complementary social measures.

Looking forward to reading more constructive discussion on this.

Thank you

January 25, 2012

It is the best example of mentioning the reality in Ethiopia. The monopoly of the land by the government the manufacturing industry in Ethiopia is almost zero percent. Also the corrupted high officials forced you to work with you in case if you have good idea and business plans. So I completely agree mostly the problem occur with the wrong policy and immdite forced measures.

Shanta Devarajan
January 27, 2012

Andy, thanks for checking my arithmetic, but I think the problem is that you interpreted "two orders of magnitude" as a doubling. In fact, I meant that employment would increase hundredfold, not two times. Regards, Shanta

Industrial Policy
February 28, 2012

One of the best reasons to work in IT, is arguably the money. Information technology is known as a career that pays fairly well, across many areas.