An evaluation of South Africa’s industrial policy

“The theoretical case for industrial policy is a strong one. The market failures that industrial policies target − in markets for credit, labour, products, and knowledge − have long been at the core of what development economists study. The conventional case against industrial policy rests on practical difficulties with its implementation. Even though the issues could in principle be settled by empirical evidence, the evidence to date remains uninformative. Moreover, the conceptual difficulties involved in statistical inference in this area are so great that it is hard to see how statistical evidence could ever yield a convincing verdict.” Dani Rodrik: “Normalizing Industrial Policy”.  Commission for Growth and Development. World Bank. 2008.

by Jac Laubscher, group economist

Industrial policy is arguably one of the most controversial (and perhaps misunderstood) areas of economic policy. This is partly due to the fact that it involves the direct intrusion of the state into what is generally regarded as the domain of the private sector, viz. it carries explicit ideological overtones.

But it is mostly due to the fact that the empirical evidence on the success or otherwise of industrial policy is inconclusive. Advocates of industrial policy will selectively quote evidence in favour of such policies (mostly from East Asia), while its detractors will likewise refer to instances where it has not been a success (mostly from Latin America and Africa, but also India). Indeed, industrial policy, like beauty, can be in the eye of the beholder.

Perhaps, as pointed out above by Dani Rodrik of Harvard University, the problem is located in issues of statistical methodology stemming from economics not being a laboratory science in which controlled experiments can be conducted. The arguments for and against industrial policy tend to rely on stylised facts, which unfortunately do not reflect cause and effect or the absence thereof.

Even a development economist of the standing of Rodrik (a protagonist of industrial policy) has fallen victim to this fallacy, arguing that there is a high probability that countries that experienced sustained high growth followed an active industrial policy. This caused the equally eminent William Easterly to point out that Rodrik is approaching the issue from the wrong end, viz. that he should have asked what the probability is that if a country adopted industrial policy, it would experience sustained high growth. Easterly then concludes that the accumulated evidence points to this probability being pretty low.
 

It is furthermore not self-evident that the relative success that some East Asian countries have achieved with industrial policy can be replicated elsewhere, with the role of other factors such as high investment rates, high levels of education, and gains in total factor productivity because of technology imported from abroad, dominating. In fact, it is doubtful whether industrial policy was the most important contributor to higher growth – it at best added marginally to already high growth rates. As pointed out by Marcus Noland and Howard Pack, it is a misconception that Japan’s erstwhile success was attributable to its industrial policy, with the sectors not receiving government support outperforming those that did.

It has also been argued that the financial crises in Japan and in Korea in the 1990s had their roots in the manipulation of the banking sector to achieve the selective objectives of industrial policy.

Not all the research and data on industrial policy in East Asia are relevant to South Africa today – the early period of post-war industrial policy in Japan, for example, concentrated on the reconstruction of sectors that were already well established before the war. The experience of Korea and Taiwan was likewise heavily influenced by the legacy of the Japanese occupations, while both were also blessed with exceptionally high levels of education at the start of their industrialisation drives, which greatly enhanced the capacity of the work force to absorb new technologies. The most relevant experiences are therefore arguably those of Korea and Taiwan since the mid-seventies.

Perhaps the correct answer is that industrial policy can be successful, depending on the realism of its objectives and how well it is implemented.  As Ann Harrison and Andrés Rodrígues-Clare put it: “While many kinds of market failures could justify government intervention in theory, the critical questions remain: Does industrial policy work in practice?”

In this regard one should bear in mind that the success of industrial policy is not dependent on a once-off successful intervention by government, but on doing it consistently over a long period of time. The main purpose of industrial policy is after all to speed up the process of structural change towards higher productivity activities. Industrial policy is also not only about introducing all sorts of measures, but also about knowing when to adapt or terminate a previous intervention as circumstances change, or to rectify outright mistakes. The latter especially can be politically difficult because of the development of vested interests.

I have a distinct impression that the reasoning behind South Africa’s decision to adopt a more active industrial policy, as embodied inter alia in the recently released “2010/11 – 2012/12 Industrial Policy Action Plan” (IPAP), very much follows Rodrik’s line of thought. If one believes William Easterly, the chances are therefore high that the results will turn out to be disappointing. After all, South Africa has a lot more in common with Latin America, Africa and India than with East Asia − structurally, institutionally and culturally. The DTI therefore has its work cut out.

 

A brief summary of the Industrial Policy Action Plan

As pointed out by Rodrik, industrial policy was put firmly on South Africa’s policy agenda by ASGISA, with much of the work of the so-called “Harvard Group” of economists that advised the South African government at the time focusing on industrial policy issues. It would therefore be wrong to dismiss the policy that subsequently emerged as an ineffectual part of the purported “shift to the left” in economic policy.

 

Objectives

South Africa’s industrial policy has the following objectives:

• The diversification beyond traditional commodities and non-tradable services, requiring the promotion of increased value addition.

• The long-term intensification of South Africa’s industrialisation process and movement towards a knowledge economy.

• The promotion of a more labour-absorbing industrialisation path.

• The promotion of a broader-based industrialisation path characterised by the increased participation of historically disadvantaged people and marginalised regions in the mainstream of the industrial economy.

• Contributing to industrial development on the African continent.

 

 

Industrial policy response

Seven sets of policies are regarded as crucial to scaling up industrial policy:

• Stronger articulation between macro- and micro-economic policies, in particular in striving for a competitive and stable exchange rate regime and a competitive real interest rate regime.

• Industrial financing channelled to real economy sectors, with both investment incentives and concessionary (although conditional) finance from development finance institutions such as the IDC being utilised as instruments in allocating capital to productive investment.

• Leveraging public and private procurement, focusing on a limited number of large and strategic procurement “fleets”.

• Developmental trade policies, lowering tariffs on intermediate inputs and increasing tariff and non-tariff protection for final goods.

• Competition and regulatory policies that lower costs for productive investments by reducing the market power of dominant firms.

• Skills and innovation policies that are aligned to sectoral priorities.

• Deploying these policies in general and in relation to more ambitious sector strategies.
  

 

Targeted sectors

The IPAP targets thirteen widely different sectors of the economy, viz:

 

Manufacturing

• Metal fabrication, capital equipment and transport equipment

• 'Green' and energy-saving industries

• Automotives, components and medium and heavy commercial vehicles

• Plastics, pharmaceuticals and chemicals

• Clothing, textiles, leather and footwear

• Cultural industries: crafts and film

• Advanced manufacturing.

 

 

Primary production/ manufacturing nexus

• Agro-processing

• Downstream minerals beneficiation

• Bio-fuels

• Forestry, paper and pulp, and furniture.

 

 

Services

• Tourism

• Business Process Services.

 

 

However, no explanation of the rationale behind the choice of each favoured sector is provided, e.g. what the nature of the deemed market failure in each case is and how it will contribute to achieving the policy goals described above. Nor is there a set of quantifiable objectives explaining the expected welfare enhancement of each intervention. The diversity of the chosen sectors also shuns the widely-accepted principle of approaching industrial policy in a cluster context.

One gets the impression that some of the targeted sectors made the list purely because they are fashionable (e.g. ‘green’ and energy-saving industries and bio-fuels) or ideologically attractive (e.g. downstream minerals beneficiation), while others respond to strong lobbying from vested interests (e.g. clothing, textiles, leather and footwear). It is therefore not surprising that the “Harvard Group” opposed their inclusion

 

Assessment

The IPAP is obviously an attempt at contributing to South Africa’s quest for a new growth path, the need for which there is agreement across a broad spectrum of people, ranging from government to labour and business. One can also not object to some experimentation in economic policy in view of the fact that past policies have not fully produced the desired results, especially with regard to job creation.

The question is how successful its implementation is likely to be. In this regard the following comments apply:

• I basically disagree with the point of departure of the IPAP that only mining, agriculture and manufacturing can be regarded as “productive” sectors, implying that the rest is “unproductive”. In my opinion all the sectors of the economy are productive and they complement each other. It is not only the manufacturing sector that has the capacity to create large numbers of low-skilled jobs, but also the services sector (e.g. tourism, which is fortunately one of the targeted sectors).
• Industrial policy should follow the revealed competitive advantage of the different sectors instead of trying to force them in a bureaucratically desired direction. To put it differently, industrial policy should aim at discovering the competitive advantages in the economy in close collaboration with the private sector instead of prescribing what they should be.
• A major objective of South Africa’s industrial policy is to promote a more labour-intensive growth path. Yet the IPAP refuses to get to grips with the dependence on sustained low real wages this objective would imply. The East Asian example furthermore points to the importance of labour flexibility, especially high mobility between different sectors of the economy, for workers to move into higher productivity jobs to achieve an improvement in real wages.
• A standard criticism of industrial policy is that governments cannot pick winners; only businesses responding to the profit motif can do so. (Or, perhaps, as Rodrik pointed out, the real test is whether governments can let losers go). The counterargument is of course that markets can fail to incorporate all the relevant information, e.g. the fiscal and social cost of unemployment, into the set of prices on which businesses will base their calculations of prospective profits and thus arrive at a sub-optimal conclusion.
Industrial policy boils down to a country acting similarly to a corporate entity, except for adding the missing information into the mix. Taiwan is an interesting example, with the government first hiring the Stanford Research Institute to identify promising industries, followed later by consultants Arthur D Little at the next step. In other words, Taiwan acted as Taiwan Inc. as if it was one diversified business, and it followed a dynamic development path, adapting as conditions changed. I find it difficult to envision a dynamic, vibrant South Africa Inc. based on the history of local industrial policy.
• A second standard criticism of industrial policy is that because its implementation relies on the creation of rents, whether in the form of direct benefits such as subsidies or indirect benefits such as concessionary finance, it encourages rent-seeking behaviour and corruption. The way to counter these tendencies is to put in place definite performance requirements for the recipients of government assistance, with penalties if they are not complied with. Once again I find it difficult to envision the South African government pushing through such an approach, especially since a secondary aim of the IPAP will be the promotion of previously excluded groups.
• One way of interpreting the IPAP is to regard it as an attempt to address South Africa’s poor performance in the entrepreneurial stakes, but this objective is bound to be frustrated by other government policies that discourage entrepreneurship, e.g. black economic empowerment policies in their present form.
• Although the IPAP is to target some services-related activities, its main focus is on growing the manufacturing sector. The possibility of utilising the services sector as the lead sector in economic development was not considered seriously. In the meantime, experiences in South Asia, in particular in India, have shown services-led growth can be sustainable – services are the largest sector in the world, accounting for more than 70% of global output, while globalisation of services has merely started.
• In those instances where industrial policy has been relatively successful, a focused approach was followed. The IPAP targets too many diverse sectors, resulting in a lack of strategic thrust. For all its emphasis on reducing unemployment, which as is generally known is concentrated among unskilled and low-skilled workers, the skills requirements of most of the sectors targeted in the IPAP appear to be rather high. As one commentator put it, “The IPAP is aimed at the labour force South Africa would like to have, not the one that it actually has”. In addition, the choice of sectors shows that the IPAP does not reflect a clear priority for promoting small business or big business as the agent of industrialisation, which would require vastly different approaches.
• One of the key elements of successful industrial policies has been their focus on the transfer, absorption, diffusion and development of technology. In the current WTO-dominated environment the protection of intellectual property rights of course complicates the transfer of technology from the advanced to emerging and developing economies, requiring a special policy focus on enabling such transfers. The best way available remains via the involvement of multinational companies in the economies of the latter group of countries. However, the capacity of an economy to absorb technological advance is heavily dependent on the quality of its human capital, which depends very much on education and skills development, and South Africa is unfortunately not well positioned in this regard.
• The IPAP puts a lot of emphasis on import replacement, which is a rather outdated approach to industrialisation. A focus on exports, as was the case in East Asia, would ensure the need to remain competitive will install discipline in the economy, forcing through gains in productivity. Import replacement by definition implies an introverted approach to development, making one more reliant on one’s own resources instead of tapping into the global pool. It means the potential growth rate of the economy will be limited to the growth in domestic demand, and that the option of leveraging off more strongly-growing global demand will not be available. It is striking that the import-substituting countries, e.g. in Latin America, did not achieve the increase in saving and investment rates and in education levels as the countries in East Asia. Evidence also points to technological advance being retarded in import-substitution regimes. It is currently fashionable to state that the global financial crisis and the accompanying collapse in world trade imply the demise of the export-driven growth and development model. However, this is taking it too far – while it does imply that for an economy the size of China it is not sustainable to forever run huge current account surpluses, a niche country such as South Africa will still be able to follow this approach. To argue that the global recession has proved that it is better to rely on domestic demand than global demand denies the high correlation in national business cycles that has been the consequence of global economic integration, while elevating a cyclical phenomenon to a structural one. The focus should rather be on improving the ability of South African industry to be included in global supply chains. This would inter alia require steps to overcome the disadvantage of the long distances between South Africa and major markets through a super-efficient transport and logistics system.

Industrial policy and the financial sector

The IPAP by and large displays a negative attitude towards the financial sector. Bashing the financial sector has of course become politically expedient in the wake of the financial crisis, but that does not detract from the essential role of the financial sector in enabling growth and development. Furthermore, some of the criticism of the financial system is based on erroneous arguments; for example, equating the cost of capital to the real repo rate is far off the mark. Viewing short-term bank credit as the only source of investment capital, ignoring the capital market and its role in mobilising and allocating capital, is likewise fundamentally wrong.

 

The fact that the contribution of the “productive” sectors (especially manufacturing) to GDP has declined while that of the “unproductive” sectors (especially finance) has increased as a result of the variation in sectoral growth rates, should not be construed as an indictment of the financial sector. (This argument has its origin in the general objection to the “financialisation” of the economy). It is mostly the result of changes in the relative competitiveness of different sectors – one only needs to take a glance at the various global competitiveness reports to see that the South African financial sector is rated way ahead of sectors such as manufacturing.

 

 It is also in line with what is generally regarded as the normal development path for economies from domination by the primary sector to the secondary sector and eventually the tertiary sector. Consciously turning back the development clock would not make sense to me. The South African financial services industry is well positioned to provide services to Africa; in fact, the interest expressed by global financial institutions in using South Africa as a gateway into Africa proves the point. The same comment applies to other “unproductive” sectors, e.g. the retail sector.

 

Conclusion

The odds are therefore heavily stacked against the IPAP reaching its lofty objectives. However, that does not imply that it should be abandoned without first asking in what ways it could possibly be improved. In this regard I would suggest the following:

Firstly, the DTI should determine what capacity it really has to pursue an active industrial policy and avoid trying to do more than it can realistically deliver.

Secondly, the focus of industrial policy should be narrowed down to what is manageable, viz. it should concentrate on fewer sectors that form a coherent whole.

Thirdly, there should be more interaction with the private sector to identify constraints on profitability and to discover which sectors hold the most promise for the least effort.

Fourthly, industrial policy should be aligned with factor markets, viz. the available skills, FDI, technology transfer, infrastructure and support institutions.

Fifthly, “soft” policies aimed at increasing productivity by increasing the supply of skilled workers, encouraging technological adoption, and improving regulation and infrastructure, stand a better chance of enhancing welfare.

A last thought: There has been a worldwide revival of interest in industrial policy in the past two years, with South Africa not the only country that wants to grow the share of the manufacturing sector in its economy – even advanced countries such as the UK are looking anew at industrial policy to enhance economic growth. South African policymakers are therefore faced not only with the challenge of implementing industrial policy successfully, but also of doing it better than the competition.

Fifthly, industry protection must be off-set by increased competitive pressures to introduce discipline and control rent seeking behaviour.

“The development landscape is littered with white elephants, products of industrial promotion efforts that resulted in low-productivity, uncompetitive enterprises that never operated at full capacity. At the same time, there is no shortage of cases that suggest industrial policies may have worked in some instances.” But “it cannot be ruled out that many of the industries just mentioned would have come out even better in the absence of government support. “ Dani Rodrik: “Normalizing Industrial Policy”.  Commission for Growth and Development. World Bank. 2008.

 

References

1. “Is Dani Rodrik moody?” Available at http://psdblog.worldbank.org/psdblog/2009/05/is-dani-rodrik-moody.html.
   

2. Sanjaya Lall: “Reinventing industrial strategy: The role of government policy in building industrial competitiveness.” University of Oxford. September 2003.
   

3. Irfan ul Haque: “Rethinking industrial policy.” UNCTAD Discussion Paper no. 183. April 2007.
   

4. “2010/11 – 2012/12 Industrial Policy Action Plan” Economic Sectors and Employment Cluster. February 2010.
   

5. Dani Rodrik: “Industrial Policy for the Twenty-first Century”. Harvard University. September 2004.
   

6. Dani Rodrik: “Normalizing Industrial Policy.” Commission for Growth and Development. World Bank. Working Paper no. 3. 2008.
   

7. Marcus Noland and Howard Pack: “Industrial Policy in an Era of Globalization. Lessons from Asia.” Institute for International Economics. March 2003.
   

8. Mauricio Cárdenas: “Rethinking Latin America’s Development Strategy.” Brookings Institution. May 2010.
   

9. Ejaz Ghani and Homi Kharas: “The Service Revolution.” World Bank. Economic Premise no. 14. May 2010.
   

10. Ricardo Hausmann, Danny Rodrik, and Charles F Sabel: “Reconfiguring industrial policy. A framework with an application to South Africa.” CID Working Paper no. 168. May 2008.
    

11. José Guilherme Reis and Thomas Farole: “Trade and the Competitiveness Agenda.” Economic Premise no. 18. World Bank. June 2010.
    

12. Howard Pack and Kamal Saggi: “The Case for Industrial Policy: A Critical Survey.” Policy Research Working Paper no. 3839. World Bank. February 2006.
    

13. Andrés Rodríguez-Clare: “Coordination Failures, Clusters and Microeconomic Interventions.” Working Paper no. 544. Inter-American Development Bank. December 2005.
    

14. Ann Harrison and Andrés Rodríguez-Clare: “Trade, Foreign Investment, and Industrial Policy for Developing Countries.” Working Paper no. 15261. National Bureau for Economic Research. August 2009.
    

15. Robert Wade: “Governing the Market. Economic Theory and the Role of Government in East Asian Industrialization.” Princeton University Press. 2004.
    

16. Masahiko Aoki, Hyung-Ki Kim and Masahiro Okuno-Fujiwara: “The Role of Government in East Asian Economic Development.” Clarendon Press. 2005.
    

17. Otaviano Canuto and José Manuel Salazar: “Challenges in the Coming Phase of Globalisation: A Sense of Déjà Vu.” Available at www.voxeu.org 28 June 2010.
    

18. Ann Harrison and Andrés Rodríguez-Clare: “From Hard to Soft Industrial Policies in Developing Countries.” Available at www.voxeu.org 27 June 2010.