Industrial Policy in the Middle East and North Africa

Rethinking the Role of the State

Ahmed Galal

• Publisher: The American University in Cairo Press
• Year: 2008
• ISBN:
  9781617975240
• Topic: Betriebswirtschaft
• Subtopic: Buchhaltung & Budgets

Most governments in the Middle East and North Africa (MENA) region use trade policy to protect certain industries, provide tax incentives to promote a particular type of investment, and make subsidized credit available to firms of a certain size. Such government intervention, known as industrial policy, is the topic of this book. The aim is to assess whether state intervention leads to net benefits to society, why policymakers intervene, and how to bring about a healthier balance between states and markets. Answers to these questions are given in six chapters based on research papers that were presented at a conference held in Cairo in November 2005, and include case studies on Egypt, Morocco, Turkey, and Jordan. Contributors: Hasan Ersel, Ahmed Galal, Najib Harabi, Nihal El Megharbel, Mustapha Nabli, and Marcus Noland. An Egyptian Center for Economic Studies / World Bank Publication

CHAPTER 1

Comparative Assessment of Industrial Policy in Selected MENA Countries: An Overview

Ahmed Galal

No debate in the development literature has persisted as long and with such intensity as that related to government intervention in economic activity. In the last half century alone, views and actual policies have changed considerably. In the 1950s and 1960s, it was believed that markets failed widely and government intervention was necessary to speed up the process of economic transformation and the rate of economic growth. Most developing countries, including those in the Middle East and North Africa, adopted import substitution strategies in conjunction with high levels of protection, central planning, public ownership, and nonuniform policies across sectors and activities. In the 1970s and 1980s it became increasingly evident that governments fail too, and in many cases they fail even more than markets. So, the pendulum swung in the opposite direction. Promarket reforms were adopted, especially in the 1980s, frequently with the support of the World Bank and the IMF. The Washington Consensus emphasized macroeconomic stability, trade and price liberalization, privatization, and competition as key ingredients for rapid economic growth.

The experience of the last couple of decades has given grounds for rethinking the balance between governments and markets. On one hand, the colossal failure of the socialist system in Eastern Europe and the Soviet Union supported the argument that markets were the best mechanism for allocating resources and motivating economic agents. On the other hand, the failure of market reforms in Latin America in achieving high and shared economic growth demonstrated that government intervention could do some good (De Ferranti et al. 2002). The pendulum has thus swung to the middle. Extreme views aside, there is a broad consensus that markets and governments can play positive and complementary roles in achieving more rapid and shared economic progress.

The evolution in thinking about the role of the state in economic development is very similar to the evolution in thinking about industrial policy, defined by Pack and Saggi (2003) as “any type of selective intervention or government policy that attempts to alter the sectoral structure of production toward sectors that are expected to offer better prospects for economic growth than would occur in the absence of such intervention, i.e. in the market equilibrium.” Currently, supporters of industrial policy argue, for various economic reasons that will be discussed below, that selective intervention is essential for economic diversification and long-term gains in productivity. They cite the success of industrial policy in East Asia as evidence in support of their view (World Bank 1993; Rodrik 2006). Opponents doubt the success of industrial policy anywhere and argue that the contribution of selective intervention in economic progress in East Asia is very modest (Krueger 1980; Noland and Pack 2003). The broad conclusion of the empirical evidence is mixed.

In the MENA region, systematic empirical evidence regarding the consequences of industrial policy is rare if not nonexistent. Nevertheless, many countries are actively engaged in attempts to alter the structure of their economies in an effort to gain new competitive advantages. Surely, the level of intervention has certainly subsided in the last couple of decades, but the legacy of selective interventions lingers on. Differentiated protection from foreign competition continues, and preferred sectors continue to receive tax exemptions, implicit subsidies of inputs (such as energy), or explicit tax exemptions. Moreover, investment promotion agencies are busy advocating certain projects or zones over others. These policies may or may not be good for economic development, no one really knows.

The purpose of this volume is to contribute to the debate on industrial policy in general and in the MENA region in particular. The following five chapters attempt to address these broad questions:

1. Has industrial policy worked in the MENA region?
2. What lessons can the region derive from the experience of East Asia?
3. Finally, what are the political economy factors that produced the current industrial policy in the region?

The first question is addressed by carefully assessing the experiences of Egypt, Morocco, and Turkey in Chapters 2 through 4. The second question is addressed by reviewing the relevance of the East Asain experience to the MENA region in Chapter 5. Finally, the political economy question is addressed in Chapter 6.

The remainder of this overview chapter does two things: First, it offers a brief account of the arguments for and against industrial policy; Second, it offers succinct answers to the three questions posed above—without pretending to be a substitute for the more careful and detailed analysis found in subsequent chapters.

I. The Industrial Policy Debate

The literature on industrial policy, both theoretical and empirical, is extensive and inconclusive. We have no intention of reviewing that literature here.¹ Rather, we are interested in analyzing the following case studies within the context of the current debate. For this reason, we offer only a brief summary of the rationale for and arguments against industrial policy, followed by our analysis of the case studies of Egypt, Morocco, and Turkey.

The Rationale for Industrial Policy

Traditionally, the rationale for industrial policy was linked to the infant industry argument. This argument is based on the notion that new industries will not be able to compete against their rivals, especially foreign competitors, because they initially incur high production costs. Protection and other forms of direct and indirect subsidies (such as tariffs or cheap credit) enable these firms to grow, increase productivity, and reduce the cost of production over time. Without support, Baldwin (1969) argued, entrepreneurs would not have the motivation to invest in knowledge acquisition because of knowledge spillover, train their workers because of labor mobility, produce new products with static positive externalities because they could not internalize the benefits, or undertake new projects if the initial cost of assessing these projects is high. From the perspective of society, extending support to such activities is justified as long as the discounted stream of benefits generated from learning-by-doing outweigh the discounted stream of subsidies (Pack and Saggi 2006).

Hausmann and Rodrik (2003) and Rodrik (2004) reformulated the arguments for industrial policy, emphasizing two types of market failures that weaken the motivation of entrepreneurs to diversify in low-income economies: information externalities and coordination externalities. With respect to information externalities, they point out that the diversification of the productive structure requires entrepreneurs to ‘discover’ the cost structure of new activities through random experimentation with new products and the adaptation of foreign technologies to local conditions. By providing support to this process, countries would be able to move beyond specializing in products in which they currently have a comparative advantage to products in which they could acquire one. Once the discovery is made by one entrepreneur, it is followed by imitative entry by others.

In support of the randomness of the process, they point out several examples of countries that have very similar factor endowments, but end up specializing in different types of products. Among these examples, they cite Bangladesh and Pakistan, two countries that appear to have similar initial resources. Yet Bangladesh exports a substantial number of hats while Pakistan exports virtually none. At a higher level of income, they indicate that Korea is a major exporter of microwave ovens but exports few bicycles; this pattern is reversed in Taiwan. A similar point is made regarding the successful cases of garments in Bangladesh, cut flowers in Colombia, and the IT industry in India. Finally, they use the Chilean experiment in promoting the salmon industry to point out that a state entity can successfully act as entrepreneur.

Beyond the examples cited above, perhaps the most compelling argument in favor of industrial policy is the observation that few developed or newly developed economies have made the transition without some kind of industrial policy that ignited a process of diversification. And it was because of diversification, spillover, and the sharing of knowledge that these economies were able to move to a higher and more sustainable level of economic growth and prosperity.

With respect to coordination externalities, the point is simple but compelling. Many projects require simultaneous, large-scale investment in order to be profitable. Investment in one project is not profitable without investments in other related projects. If the fixed cost of these other projects is high and no one is playing the role of coordinator, none of the investments in that industry will take place. The coordination failure is particularly acute where new industries exhibit scale economies in the presence of nontradable inputs (Rodrik 1996). It is also a common characteristic of most low-income countries.

The need to coordinate investment and production decisions, especially in the early stages of development, is not new. The idea finds its origin in the ‘big push’ strategies of development, and more recently in the concept of clusters in particular sectors, such as tourism and pharmaceuticals. The practical problem is that most industries tend to operate as clusters, although many of them can operate without clusters as pointed out by Rodriguez-Clare (2004). This observation led Rodrik (2004) to argue against extending support to specific sectors. Instead, he argues in favor of supporting the adoption of new technologies, the development of new products, and training that equips workers with the skills associated with new techniques. Supporting existing establishments or traditional products does not necessarily help the process of diversification and economic growth, and may result in lost resources to society.

The Case against Industrial Policy

Notwithstanding the strong appeal of arguments for industrial policy, the counterarguments seem equally powerful. To begin with, the success of industrial policy hinges on the assumption that the government is better informed than the private sector about potential winners, their geographical location, and the appropriate technologies. It is also based on the assumption that governments can identify instances of coordination failures and design support schemes that generate more benefits than costs. However, both assumptions may not necessarily hold in practice, and the private sector may in fact be better informed on these counts. Imperfect information on the part of the government is further exacerbated by the lack of penalties for bureaucrats who make bad decisions. Bureaucrats rarely pay for their mistakes and politicians are not typically penalized through the ballot box in less than democratic countries. Thus, societies may be better off if their governments refrain from adopting an active industrial policy, and focus instead on only what government can and should do. The latter includes protecting property rights, enforcing contracts, and providing sound policies.

The second counterargument is that governments may not always do what is best for advancing the development process. Motivated by the desire to stay in power, governments are likely to use industrial policy to favor their political supporters at the expense of their opponents. In addition, because industrial policy favors some business ventures and not others, it could lead to corruption and rent-seeking behavior on the part of some bureaucrats and private entrepreneurs (Nogues 1990).

The above arguments are not merely fear-based but are in fact supported by empirical evidence, which broadly suggests that industrial policy has either been ineffective or has been abused, with minimal returns to society (Krueger 1980; Pack 2000; Noland and Pack 2002). Without restating the findings of this literature, suffice it to note that it is widely believed that industrial policy in Latin America resulted in the inefficient allocation of resources, discrimination against exports, and even a deterioration of income distribution (Edwards 1994; Noland and Pack 2003). Even in East Asia, where industrial policy is believed to have worked, there is evidence that industries that received support did not experience higher productivity growth in comparison with those that received none (Pack 2000). There is also evidence that industrial policy promoted capital-intensive sectors at the expense of employment creation, or low export performers. More recently, the Asian crisis of the late 1990s has been partly blamed on earlier government direction of credit (Noland and Pack 2002).

The Bottom Line

It is clear from the above discussion that there are strong arguments for and against industrial policy. The empirical evidence is equally divided, offering support for the claims of both advocates and opponents. The dilemma is that industrial policy is needed to enable developing countries to escape the trap of specialization in a few traditional commodities. But there are no guarantees that this policy works. Even if the right intervention is made, politics, rent-seeking behavior, corruption, and weak institutions could stand in the way of the benefits of industrial policy.

From the perspective of learning from experience—which is the subject matter of this volume—one way to resolve the above dilemma is to modify the question being posed. Instead of merely asking whether industrial policy is needed or not, it is probably more useful to ask whether or not it has worked in a specific context and to identify the conditions that have produced observed outcomes. These are essentially the questions addressed in the case studies of this volume.

II. Summary of Findings

No summary can do justice to the rich details and rigorous expositions presented in the chapters that follow. Nevertheless, it is useful to capture the essence of the answers they provide to the three questions raised at the outset of this introduction. These questions and answers are briefly elaborated below.

Has Industrial Policy Worked in the MENA Region?

The short answer for Egypt, Morocco, and Turkey is ‘no.’ Active industrial policy may have increased diversification in the early stage of import substitution, given the initial narrow industrial base. However, the evidence does not support the view that industrial policy in recent years is effective in achieving further diversification, let alone improving total factor productivity over time.

In Chapter 2, Galal and El-Megharbel assess the merits of selective intervention in the Egyptian manufacturing sector, disaggregated into sixteen industries, over the period spanning 1980–2000. They measure the extent of diversificatio...