Gunther Maier and Michaela Trippl
Introduction
Economic policy intervention in general and regional economic policy intervention in particular can be justified by two sets of arguments: an efficiency argument and an equity argument (Richardson 1978; FĂźrst et al. 1976; Armstrong and Taylor 1993). The efficiency argument is usually rooted in the notion of market failure and the need for policy to correct for it. In the case of substantial externalities, transaction costs or other major barriers to free trade and free resource allocation, it is the aim of policy to remove these barriers to allow market forces to allocate resources more efficiently. Rooted in the belief that market forces (when they can operate uninhibitedly) lead to an efficient allocation of resources, such a policy is expected to lead to higher efficiency of resource allocation and, consequently, higher welfare, since formerly underutilized resources are now used in an optimal way.
The equity argument, on the other hand, views it as morally unacceptable and politically risky when certain groups of society, or certain regions or sectors, do not get their adequate share of economic welfare. Excessive disparities in income, for example, may be seen as a threat to the existence of a state or an economic and political union like the European Union, and should therefore be counteracted by respective policy measures. The cohesion-oriented part of EU regional policy is solidly rooted in this set of arguments. Combatting regional disparities through various policy instruments like provision of information, public investment, public support for private investment, and provision of infrastructure are traditional tasks of regional policy makers.
In either case, the policy maker intervenes in the economic system with the aim of moving it away from one undesirable, or less desirable, state to another more desirable one. In the case of the efficiency argument, the policy will allow the economic system to better utilize the available resources, thus generating growth and a higher level of welfare. In the case of the equity argument, the policy will redistribute resources between regions and thus lead to a new distribution of economic activities, which policy makers suspect is politically more desirable for the population. In either case, policy intends to make a lasting impression on the economic system rather than just generating short-term effects. In this sense, a successful form of regional economic policy is one that steers the economic system toward a new, more desirable development path, which the economic system will then follow automatically without continuous policy intervention.
These arguments show that the policy maker needs an excellent understanding of how the economic system works. In order to design policies in such a way that they lead to the desired outcome, the regional policy maker needs to know which way the economic system will react to his/her policy incentives. This knowledge may come from theoretical knowledge, past experience, or the experience of comparable regions. This issue is of particular relevance to regional policy makers in a globalized economy, where the regional economic system is not only integrated into the national one, but, through numerous linkages, also connected to the economic systems of other countries and their regions.
In the recent two decades, our understanding of how the economic system works has changed considerably. At the micro level, substantial doubt has been raised about the validity of the basic assumptions of the standard neoclassical model used to analyze the behavior of economic actors (Tversky and Kahneman 1974, 1981; Smith 2000; Camerer and Loewenstein 2003). Contributions at the macro level have led to new perspectives on the role of trade (Krugman 1979, 1984, 1986; Ethier 1982; Eaton and Grossman 1986), on growth processes (Grossman and Helpman 1991; Aghion and Howitt 1998), and on the significance and relevance of spatial agglomerations of economic activities (Krugman 1991; Fujita and Thisse 2002; Fujita et al. 1999). Many of these arguments have led to policy suggestions and have been picked up by policy makers and policy consultants. In a regional context, cluster policies (e.g. Porter 1998), which attempt to form and support spatial agglomerations of related economic activities, have received widespread attention and have been developed and implemented by many regional authorities around the world.
Although such new views of how the economic system works are obviously adopted willingly by regional policy makers, it seems to us that this adoption is selective and quite incomplete. Policy makers tend to accept certain arguments and conclusions from new theoretical developments, while ignoring or pushing aside others. This may lead to inconsistent policies, the effectiveness of which can be seriously questioned.
In this chapter we will investigate various views of how the economic system works and what their consequences are for regional policy. As far as the views of the operation of the economic system are concerned, we take on a rather abstract perspective and distinguish between the âneoclassical viewâ on the one hand and the âagglomeration viewâ on the other. The second section of the chapter is devoted to a comparative discussion of these two views. We argue that these are actually the only two comprehensive views of the operation of the economic system currently available, and that all the above mentioned theoretical advances can be subsumed under the âagglomeration view.â
While the âneoclassical viewâ provides only few arguments for regional policy, the âagglomeration viewâ actually requires (regional) policy intervention, as we will discuss in the third section of the chapter. However, the same characteristics ascribed to the economic system by the âagglomeration viewâ that call for regional policy, also challenge the adequate implementation of regional policy. The non-linear relations between key elements of the economic system that the âagglomeration viewâ postulates imply that the reaction of the economic system to a policy intervention crucially depends on when, where, and in what intensity the policy measure is implemented. The chapter closes with a concluding section.
The theoretical background of regional economic policy
When we want to analyze the opportunities that regional policy has for influencing the long-term economic prosperity of a region, we need some understanding of how a regional economy works. As we have already mentioned in the introduction, at a general level we can distinguish between two main views:
1 the âneoclassical view,â and
2 the âagglomeration view.â
Both views that we will discuss in more detail below take a comprehensive look at the economy and its various elements and interactions. In a general equilibrium perspective, they take into account commodity, labor, capital, and various other markets as well as the transfers between these markets and between the spatial units considered.
The two views differ by their treatment of agglomeration forces and transportation costs. The neoclassical view excludes both agglomeration economies1 and transportation costs by assumption. This simplifies the formal structure of the model and allows us to derive some very clear and specific results. The agglomeration view, on the other hand, allows for positive agglomeration forces that are counteracted by transportation costs. Depending on the relationship between these forces, the models of the agglomeration view lead to different results. Since the other assumptions of the neoclassical view usually remain unchanged in agglomeration view models, one can view the neoclassical view as a special case of the agglomeration view. The former results from the latter when we assume that neither agglomeration economies nor transport costs exist in the economy.
Neoclassical view
In regional economics the neoclassical view results from the application of the principles and assumptions of neoclassical economics to a regional economy. The neoclassical view has been formalized by Arrow and Debreu (1954) and is taught to students today as mainstream economics. Neoclassical economics is based on a set of assumptions and its theoretical conclusions are deduced therefrom. The most important tool of neoclassical economists is mathematics, in particular calculus. The application of these assumptions, and of the economic toolbox, has led to a consistent set of arguments and to principles that are widely shared by economists. In the context of our chapter, a brief discussion of the neoclassical view is justified by the influence it still tends to have on policy considerations. Most of the market principles that guide the design and implementation of some of our policies are derived from the â fairly limited â perspective of the neoclassical view.
The most important assumptions of the neoclassical view are utility/profit maximization, perfect information, perfect mobility, and perfect competition. The neoclassical assumptions guarantee that economic actors always pursue their self-interest and in doing so are not restricted by limited information, transaction, and mobility costs or power of rival actors. Market prices always reflect the scarcity in the respective market and signal opportunities to increase their utility or profit. Because of the assumptions, actors will always pursue these opportunities. In the context of our chapter, the assumption of âperfect competitionâ and the assumption of âperfect mobilityâ are of particular relevance.
The assumption of âperfect competitionâ implies that production takes place according to a linear homogeneous production function. This implies that changing all the inputs to a production process by a certain factor changes the output by the same factor. So, irrespective of the scale of production, the costs per unit remain the same, larger or smaller units do not have an economic advantage over others, and agglomeration economies are excluded by assumption. In summarizing, it can be said that in the neoclassical view there is no economic incentive toward the concentration of production. If such an incentive existed in one sector, for example, it would work toward the concentration of all the global production in this sector in one location. Because of the assumption of perfect mobility, there is no force counteracting this concentration tendency. This would generate one monopolist in this sector, contradicting the requirements of perfect competition.
The assumption of perfect mobility eliminates one of the most important elements of spatial economics: transport costs. The assumption implies that all goods, services, and individuals can be transferred across space without any costs in terms of money, time, and effort. This assumption guarantees that there is âoneâ market with many suppliers and customers, instead of various spatial markets, which conflict with perfect competition because of their limited numbers of actors.
Because of the assumption of perfect mobility it is difficult for neoclassical economics to deal with regions and space. When we relax this assumption and allow for physical distance and transportation costs, we either conflict with other neoclassical assumptions, like perfect competition, or end up in the perverse structure of âbackyard capitalismâ (Mills 1972; Starrett 1978). With transportation costs, it is profitable for producers to break up production facilities and to locate them as closely as possible to the location of the customers. Because of the assumption of perfect competition, which implies the absence of agglomeration economies, this can be pursued until all products are produced in the backyard of their respective customer. In this âbackyard capitalismâ there are only subsistence industrialists and no large-scale production facilities that require external labor. No products are shipped from one location to another, no workers commute from home to work. Although transportation costs are the ultimate reason for this structure, in equilibrium, no resources are used for transport.
This line of reasoning shows clearly how the assumption of perfect competition and the assumption of perfect mobility are related. If we want to avoid outcomes of complete concentration and of âbackyard capitalism,â then whenever we relax one of the assumptions, we also have to adjust the other. They relate to counteracting forces in a spatial economy. Agglomeration economies pull production together while transportation costs push them apart.
An immediate consequence of its assumptions is that the neoclassical model is a linear model. This has a number of important implications for its policy recommendations:
⢠Marginal changes lead to marginal reactions: In all markets, a small increase in demand yields a small increase in the respective price, which in turn stimulates a small increase in supply, which brings supply and demand back to equilibrium. There are no threshold levels that need to be crossed before markets react. Consequently, there are no beaten paths that the economy will follow, nor can it be locked into a specific equilibrium by the structure it generates.
⢠There exists only one equilibrium outcome: This equilibrium is Pareto-efficient and socially optimal. The assumptions ensure continuously increasing supply and continuously decreasing demand curves, which intersect at one point and thereby define the market equilibrium. Since actors will trade or produce only when they can increase their utility or profit, everyone reaches the highest level of the objective function that can be achieved under the existing constraints. The resulting equilibrium is therefore optimal for every individual actor as well as for society as a whole.
⢠Disturbances are washed away over time: They lead to deviations from equilibrium, which triggers market forces that bring the economy back toward equilibrium. Therefore, disturbances can only have a temporary effect that disappears over time. The long run outcome of the economy cannot be changed by external disturbances.
These implications lead to the famous policy recommendation of neoclassical economics, that when all the assumptions hold, policy should adopt a âhands offâ approach and not intervene in the economy. Policy is only needed to correct for deviations (âmarket failureâ) from the neoclassical set of assumptions. This task corre...