«Learning and Industrial Policy: Implications for Africa1 Bruce Greenwald2 and Joseph E. Stiglitz3 Over the past thirty years, Africa has suffered ...»
Learning and Industrial Policy: Implications for Africa1
Bruce Greenwald2 and Joseph E. Stiglitz3
Over the past thirty years, Africa has suffered from deindustrialization. The quarter century
from the early 1980s was a period of declining per capita income and increasing poverty.
Structural adjustment policies advocated by the IMF and the World Bank were predicated on
the belief that by eliminating “distortions” in the economy, Africa would grow faster—by
constructing an economy based on principles of free and unfettered markets, with the government restrained to ensuring macro-stability (which typically just meant price stability), economic performance would be increased and all would benefit.
It was recognized, of course, that eliminating trade protection would result in the loss of jobs, some in agriculture, many others in industry. The strongly held belief, however, was that these workers would quickly find jobs in new industries, consistent with the country’s comparative advantage. Moving resources from inefficient protected sectors to more efficient competitive sectors would raise incomes. Little attention was paid to the distribution of income, perhaps because of an implicit belief in trickledown economics—somehow, if the economic pie grew, all would benefit.
Things didn’t turn out as the advocates of these policies had hoped. Rather than growth there was decline. Job creation didn’t always keep pace with job destruction, and so workers moved from low productivity protected sectors to even lower productivity unemployment, open or disguised. When there was growth, the benefits often went disproportionately to those at the top, and didn’t trickle down to the rest of the economy.
Paper presented to an International Economic Association roundtable conference on “New Thinking on Industrial Policy: Implications for Africa,” Pretoria, July 3-4, 2012, co-sponsored by the World Bank, UNIDO, and the South African Economic Development Department. Research support from Laurence Wilse-Samson and the helpful comments of the other participants in the seminar is gratefully acknowledged. This paper is a companion to B.
Greenwald and J. E. Stiglitz, Industrial Policies, the Creation of a Learning Society, and Economic Development, presented to the International Economic Association/World Bank Industrial Policy Roundtable in Washington, DC, May 22-23, 2012. Both papers are based on Greenwald and Stiglitz (2006, forthcoming) and Stiglitz (2010).
Professor at Columbia University, Graduate School of Business University Professor, Columbia University When, in the first decade of the twenty first century growth resumed, it was largely based on the boom in commodity prices. The share of global manufacturing value added in Africa in 2008 was 1.1 percent in 2008, from 1.2 percent in 2000 (UNCTAD 2011). Even countries that achieved macroeconomic stability and evidenced reasonably good governance seemed unable to attract much investment outside of the extractive sector.
It is imperative that this course of events be changed, particularly since the extractive sector typically does not give rise to many jobs, and certainly not enough jobs for the burgeoning labor force in many of the countries. (The African labor force is expected to grow --- working age Africans today comprise some 500m people; by 2040, that number will be 1.1bn4.
A Propitious Time for Africa
Fortunately, there are a set of events that may be propitious for the sub-continent. First, increasing wages and an appreciation of exchange rate in East Asia may enhance Africa’s comparative advantage in manufacturing. The high levels of productivity growth in manufacturing—exceeding the increases in demand—imply that global employment in manufacturing will be declining; but it may be possible for Africa to seize a larger share of these jobs.
Moreover, there are some spillovers from even imperfectly managed natural resources: higher incomes give rise to a demand for more consumption, and some of this will be locally produced and/or serviced. There is an increasingly large middle class. Indeed, by some estimates, only around a quarter to a third of the sub-continent’s recent growth is directly attributable to natural resources.5 Moreover, with the weaknesses in Europe and the United States that began with the Great Recession of 2008 looking likely to extend for at least a decade, those with funds are looking McKinsey, 2010, “What’s driving Africa’s Growth?” available at http://www.mckinseyquarterly.com/Whats_driving_Africas_growth_2601 (accessed September 5, 2012).
See for example, McKinsey, 2010, “What’s driving Africa’s Growth?” available at http://www.mckinseyquarterly.com/Whats_driving_Africas_growth_2601 (accessed September 5, 2012), Exhibit 1 indicates that 24% of the growth between 2000 and 2008 in sub-Saharan Africa is attributable to resources (but a further 8% is derived from resource-financed government expenditure).
elsewhere for places in which to invest their money. Africa is looking more attractive, with an increase in the share of global foreign direct investment projects to 5.5% in 2011.6 But many African countries still face serious disadvantages. Deficiencies in infrastructure both increase the cost of production and increase the costs of bringing goods to market and to obtaining necessary inputs. There are also important shortages of skilled personnel, even in an environment where there unskilled workers are in abundance.
This paper is predicated on the belief that these disadvantages can be overcome by appropriate government policies, but such policies necessitate moving further away from the structural adjustment /Washington Consensus (WC) policies, by embracing industrial policies—policies which were shunned under the WC programs. Industrial policies are what we call those policies that help shape the sectoral composition of an economy. The term is used more broadly than just those policies that encourage the industrial sector. A policy which encourages agrobusiness, or even agriculture, is referred to as an industrial policy.
Such government policies can enhance the ability of African economies to seize an even larger share of global foreign direct investment, to create new domestic enterprises and to expand existing enterprises. While many countries within Africa are benefitting from natural resources, most countries have not taken full advantage of those resources, to create new industries and to provide employment for more of their citizens.
Industrial policies and market failures
At the International Economic Association/World Bank meeting on industrial policy in Washington, in May, 20127, there was a broad consensus on why countries should have such policies: to correct market failures, situations where markets by themselves do not lead to efficient, or desirable, resource allocations, and in some cases, even to correct other government failures, where other, harder to alter, government policies “distort” resource allocations.
See Ernst & Young (2012).
Proceedings available as Lin and Stiglitz (2013) in accompanying volume.
Market failures arise whenever private rewards and social returns differ, and since the work of Greenwald and Stiglitz (1986) it has been recognized that such discrepancies are pervasive.
Industrial policies are designed to correct major sectoral or other misallocations.
Objectives of industrial policies
For Africa, there are at least three objectives of such policies: With many countries facing high unemployment, there is an imperative to create more jobs. The labor market is not working the way it does in neoclassical models, where there is full employment. That means that the market price of labor almost surely is markedly higher than the “shadow price,” the opportunity cost of labor. Government should encourage labor intensive sectors and technologies. To the extent possible, government should be sensitive to the kinds of labor that are being demanded, using both industrial and educational policies to bring the demand and supply of, say, school leavers and university graduates into better alignment.
Secondly, many African countries have been marked by large increases in inequality.8 Industrial policies can affect the extent of inequality, by increasing the demand for lower skilled workers, driving up their wages and lowering their level of unemployment. While policies focusing on distribution have traditionally been centered on tax and transfers, it has long been recognized that it may be better (more efficient) to have policies that change the before tax and transfer distribution of income. Such policies reduce the burden imposed by distortionary redistributive policies (Stiglitz 1998a).
Thirdly, it has increasingly been recognized that development requires the structural transformation of the economy (see Lin 2012; Stiglitz 1998c). Markets themselves are not very good at such structural transformations, partly because the sectors that are being displaced— resources that have to move from one sector to another—typically suffer large capital and income losses, and are thus not well placed to make the investments required for It is difficult to track inequality due to data limitations. The Africa Progress Report (Africa Progress Panel 2012) states that 24 countries in Africa have Gini coefficients in excess of 42, the level in China. It also points out that in a number of cases, recent growth has not been matched by falling poverty --- which they attribute to inequality (p.16), “In many countries, the pattern of economic growth is reinforcing these inequalities.” redeployment, and well-understood capital market imperfections (based on information asymmetries) limit access to outside resources.9 Fourthly, it has long been recognized that what separates developed from developing countries is not just a gap in resources, but a gap in knowledge (Stiglitz 1998b). More broadly, even in developed countries a large fraction of the increase in per capita income over the last two centuries is attributable to technological progress, to learning how to produce things more efficiently (see Solow 1957). And the fact that some countries and firms have “learned how to learn” helps explain why the last two centuries have seen such remarkable increases in standards of living, in comparison to the millennia that preceded it, which were marked by stagnation (see Maddison, 2001).
If this is so, then it means that development strategies should be centered on promoting learning, closing the knowledge gap between developing countries and less developed countries.
Market failures, learning, and industrial policies10 We suggested earlier that industrial policies are motivated (in part) by an attempt to correct market failures, by the failure of markets by themselves to yield socially desirable outcomes.
There can be too much inequality, too high unemployment, too little growth. This paper centers around the failure of markets in learning.
Knowledge is different from ordinary products. Knowledge is essentially a public good, that is, its consumption is non-rivalrous (Stiglitz 1987a, 1999). When one individual shares knowledge with someone else, it does not diminish the amount of knowledge that the first person has.
Markets by themselves are never efficient in the production and utilization of public goods.
The producer of the knowledge may restrict the usage of the knowledge (through secrecy or patents), in an attempt to appropriate returns, in which case there is underutilization. More generally, there will be underproduction, because—even with effectively enforced patents— See Delli Gatti et al., 2012a, 2012b.
See Greenwald and Stiglitz (2013), for a more extensive discussion of these market failures.
there are important spill-overs from learning. What one firm or industry learns enhances the productivity of others. When learning is a by-product of investment or of production, a corollary is that there will be underinvestment or under production (Arrow 1962; Stiglitz, 2012a).
There are other market failures associated with learning: because learning is a fixed, sunk cost, sectors in which learning is important are likely to be imperfectly competitive.11 Because investments in learning cannot be collateralized, imperfections of capital markets may restrain research expenditures, say, relative to real estate speculation. With learning by doing, optimal production may entail firms increasing production today, beyond the point where they are breaking even, in return for the benefit of lower production costs in the future, but with capital market imperfections, firms cannot finance the ensuing losses (Dasgupta and Stiglitz 1988a).
The fact that investments in learning are highly risky, and risk markets are absent (especially in developing countries), also discourages such investments.12 The general theory of learning and industrial policies is taken up in Greenwald and Stiglitz (2013). Here, we focus on several topics that illustrate the general themes discussed there and that are of particular relevance to Africa.
The inevitability of industrial policy
First, however, we want to reiterate an important point raised in our earlier paper:
governments are inevitably involved in industrial policy, in shaping the economy, both by what they do and what they do not do. If they don't manage well the macro-economy, then more cyclically sensitive industries will be discouraged. If they use interest rate adjustments to stabilize the economy, interest sensitive sectors will suffer. If they don't stabilize the exchange rate, then non-traded sectors are encouraged.
Some are wont to say, just let market forces shape the economy, but market forces don't exist in a vacuum. Every market is shaped by laws, rules, and regulations. A bankruptcy law that Moreover, potential competition is not an effective substitute for actual competition. See Dasgupta and Stiglitz (1988b); Stiglitz, 1987b.
These failures (imperfections in capital markets) can themselves be explained by imperfections of information.
gives priority to derivatives encourages these financial products. A bankruptcy law that says that student loans can't be discharged, even in bankruptcy, encourages banks to make more student loans. A tax law that provides for deductibility of mortgage interest leads to more mortgages. A tax law that taxes capital gains at lower rates than ordinary income encourages land and financial market speculation.