«Learning and Industrial Policy: Implications for Africa1 Bruce Greenwald2 and Joseph E. Stiglitz3 Over the past thirty years, Africa has suffered ...»
Moreover, in almost all countries, governments play a central role in education, health, infrastructure and technology, and policies and expenditures in each of these areas --- and the balance of spending among these areas --- also shapes the economy. In short, all governments really do have an industrial policy. The only difference is between those who construct their industrial policy consciously, and those who let it be shaped by others, typically by special interests, who vie with each other for hidden and open subsidies, for rules and regulations that favor them, usually at the expense of others. Even the agenda of financial market liberalization was an industrial policy--one pushed by the banks and the financial sector, the effect of which was to lead in many countries to a bloated financial sector, rife with explicit and implicit subsidies (reaching record levels in the crisis of 2008-2009), diverting resources from other uses that arguably would have led to high sustained growth. It was an industrial policy that led to more macro-economic instability, which, as we explain below, itself was adverse to learning.
I. The Washington Consensus and learning
The Washington Consensus policies referred to earlier in this paper focused on static efficiency.
They didn’t consider even, what were the consequences for innovation and learning. If there was learning and technological progress, it was assumed to be exogenous, outside the purview of policy, and certainly outside the purview of the economic policies on which they focused.
That this was so was striking, given the observation, made earlier, that development was so much about learning and economic transformation.
Standard theory has long recognized that there could be a trade-off between learning, or dynamic efficiency, and static efficiency. The patent system creates a temporary monopoly and imposes restrictions on the usage of knowledge, but these significant static inefficiencies are justified on the basis of the increased innovation that results.
The success of the most successful countries in development—those in East Asia—is largely attributable to their recognition of the importance of learning. Korea, for instance, paid little attention to its static comparative advantage. Its static comparative advantage would have led that country to focus on rice farming. But it knew that even if it became the most productive rice farming country in the world, its prospects would be limited. Only by focusing on sectors from which it could learn, and on the basis of which it could close the knowledge gap with more advanced countries. It developed complementary education and technology policies, and it succeeded, increasing its per capita income more than eight-fold in a span of less than four decades.
Had it followed the dictates of the Washington Consensus policies13 it would have eschewed industrial policies, and it would have focused investments in education at the primary level— and it would have, at best, been a middle income rice growing country. Unfortunately, many countries in Africa have followed the dictates of the Washington Consensus policies, and through the structural adjustment programs they have made a step backwards, as we have noted, becoming increasingly resource dependent economies.
The Washington Consensus policies were predicated on the assumption that markets, by themselves, are efficient; and that therefore the major source of inefficiency or malperformance of the economy arises from government intervention. Hence, the first item in the reform agenda is to eliminate these interventions with the market. The only role of the government was to ensure price stability.
Even before these doctrines became fashionable, their intellectual underpinnings had been taken away. Greenwald and Stiglitz (1986) showed for instance that whenever information was imperfect (asymmetric) and risk markets incomplete (which is always the case, and especially so in developing countries) markets are not constrained Pareto efficient (that is, taking into Broadly understood—not in the more restricted sense that the term was used by Williamson (1989).
account the costs of obtaining and disseminating information or creating and maintaining markets.) But the financial crisis of 2008 reinforced the conclusion that markets, on their own, may be massively inefficient, and unstable. It showed also that maintaining price stability did not necessarily lead either to growth, stability, or efficiency.14 In short, the crisis has re-emphasized the importance of market failures. It is thus natural that there should not only be a rethinking of macro-economic theory and policy, but also of microeconomic theory and policy, including the most important subject for Africa today, that of industrial policy, of how governments can help change the structure of the economy to promote learning—thereby increasing long-term sustainable growth.
Learning and one-size-fits all policies One critique of the Washington Consensus is that it has attempted to impose one-size-fits all policies. Such policies may be particularly inappropriate when it comes to creating a learning society.
A critical aspect of “learning” is that it takes place locally and must adapt to local differences in culture and economic practice. Thus “learning” prescriptions that apply in some environments will not apply in others. For example, in some economies what has been called (by outsiders) “crony capitalism” has a long and successful record. In others it does not.15 Learning how to relate to government has value in most economies, but in some, the skills required may concern those related to bidding processes, in others to interpersonal connections. American firms have had to learn to adapt to the Foreign Corrupt Practices Act.16 Labor norms differ too among countries, and personnel policies have to accommodate such differences. Differences in consumer preferences and norms as well as in distributional channels necessitate different For a discussion of the implications of the crisis for economic theory and policy, see Stiglitz (2011).
It is, perhaps, worth noting that what is viewed as corruption in one society may not be so viewed in that way by others. Many outsiders, look at the American system of large campaign contributions and revolving doors, which seems to “buy” favorable legislation as a form of corruption, even if there issn’t money stuffed into brown paper envelopes for the politicians themselves.
Dixit (2012) has argued that firms from developing countries may have a knowledge advantage in dealing with governments of other developing countries.
“learning” about marketing. Most importantly, and perhaps obviously, relative factor prices may differ, so that the returns to learning on how to save on the utilization of one factor versus another may differ.
These cross country differences have numerous implications. It helps explain why learning in a firm may spill over more easily to other firms in the same country than to firms in other countries. The learning in one country may simply be less relevant to production in the other country.
They help explain too why it is that in some economies public enterprises function well. In others they do not.17 They also help explain the limitations of globalization: local firms have a competitive advantage in having more knowledge about local circumstances.18 Much financial information is chiefly available locally, and even when information is available, outsiders may have less of an understanding of the nuances of the country’s distinctive institutional structure – as foreign investors have learned to their cost about US mortgages. Thus, effective capital deployment will often require local financial institutions.
Unfortunately, Washington Consensus policies which pushed capital and financial market liberalization did not take into account this local knowledge. Foreign banks succeeded in attracting depositors away from local banks, because they were perceived as safer (and in some cases, may have been, because they had the implicit guarantee of governments with deeper pockets.) But foreign banks were at an information disadvantage relative to local banks about small and medium sized local firms, and it was thus natural that lending be diverted away towards loans to government, consumers, and large domestic firms (including local monopolies Herbert Simon emphasized that if there are differences in the performance of public and private enterprises, the differences could not be explained just be differences in incentives, since in both typically most individuals work for others, and have to be incentivized. See for example Simon (1991, 1995).
“This examination of authority and organizational identification should help explain how organizations can be highly productive even though the relation between their goals and the material rewards received by employees, if it exists at all, is extremely indirect and tenuous. In particular, it helps explain why careful comparative studies have generally found it hard to identify systematic differences in productivity and efficiency between profitmaking, nonprofit, and publicly controlled organizations.” (Simon 1995: 288).
See Greenwald and Kahn (2005).
and oligopolies.) But in doing so, local learning and entrepreneurship may have been undermined, and growth weakened. Rashid's paper in this conference (2013) provides data strongly supporting this conclusion.19 By the same token, WTO restrictions on industrial policies and domestic sourcing (and possibly other restrictions on financial markets) may impede the ability of developing countries to foster learning, to garner for themselves the full learning benefits of foreign direct investment, or, as we shall see shortly, force them to employ second-best methods for promoting learning within their economies.
II. Macro-conditions for creating a learning society
Most of this paper is concerned with micro-economic policies, but in our earlier paper, we argued that one of the objectives of industrial policies is to create an economic environment that is conducive to learning. For this, the macro-economic environment is central. Economic stability appears to play an important role in creating a successful “learning” environment.
Evidence for this comes from the experience of developed economies during recessions.
Productivity growth is normally low during contractions and there is no off-setting gain during subsequent expansions.20 The productivity loss during the dislocation associated with the recession appears to be permanent.21 There are several reasons why stability is important for learning. The first is that much information is embodied within existing institutions, in complex webs of interactions. Key institutions—firms—often die in the face of high levels of instability.
Greenwald and Stiglitz (2003) present the general theory.
There are exceptions, including the increase in productivity in the current US recession. While there are several explanations of this distinctive aspect of the downturn, one is the increasingly short sighted behavior of firms ignores the long run costs of firing or laying off trained workers. In that case, it will still be the case that there will be long run adverse effects of the downturn on productivity. In the Great Depression productivity growth also appears to have been quite high in part due to important investments made by government (including in transportation) (Field 2011).
This is, of course, consistent, with standard results on unit roots. See Dickey and Fuller (1981) and Phillips and Perron (1986).
Moreover, managerial attention is limited. When firms are focusing on survival, they have less attention to devote to “learning,” except learning how to survive.
Thirdly, high levels of macro-instability lead firms to act in a more risk-averse manner.
When firms go into recessions, among the first things to be cut are investments in R & D, and this is even true among firms that are relatively dependent on innovation. Part of the reason is that learning is future-oriented. One has to make sacrifices today and undertake risks today, for future benefits. But in the presence of instability, there is a risk that there will be no future—and hence less reason to make the requisite investments today. Instability weakens future oriented incentives.
And fourthly, learning requires resources, including access to capital. Instability may make capital less accessible and more costly.22 In downturns, capital is likely to be rationed, and investments in R&D are often sacrificed.23 This has important implications for policy: policies which expose countries to a high level of instability, or which increase the economy’s instability (e.g. by weakening automatic stabilizers) have an adverse effect on learning. Examples include financial and capital market liberalization and deregulation (Rashid 2012; Stiglitz et al 2006; Stiglitz 2008), and tarrification (Dasgupta and Stiglitz 1977).
By the same token, policies that focus on price stability, at the expense of real stability, may actually be counterproductive. (See Stiglitz et al, 2006.) Inflation targeting, with its focus on price stability attained by interest rate adjustments, may be “doubly” bad: Responding to inflation by increasing interest rates--even when the cause of the inflation is an exogenous supply shock--is an example of a pro-cyclical policy. And the increases in interest rates have a disproportionate effect on certain sectors, those that are most interest sensitive and which rely most on bank financing. Small businesses, in particular, bear the burden. Small firms that may be killed when interest rates are raised dramatically don’t come back to life when they are This can be put slightly differently: With capital (debt and equity rationing) the shadow price of capital often increases dramatically. (See Greenwald, Stiglitz, and Weiss, 1984; Greenwald Stiglitz, 2003).
Greenwald, Salinger, and Stiglitz (1990); Stiglitz (1994).
subsequently lowered: there are important hysteresis effects. This is especially important in developing countries where there may be a dearth of entrepreneurship. If, as some claim, much of the learning and innovation in society occurs within small and young enterprises, then these policies increase the burden on these key “learning” sectors. But whether that is the case or not, these policies exacerbate the already adverse effects arising from the cyclical volatility in the “shadow” cost of capital.