2. Empirical Knowledge of Agglomeration Economies

For more on the spatial impossibility theorem, refer to Fujita and Thisse (2002).

According to Ricardo’s theory, division of labor between regions is explained through the difference in comparative production cost using labor as a measurement, and in the end, is determined by the unique climate of the region, or the regional differences in production technologies. Furthermore, according to the Heckscher-Ohlin theorem, even in cases where there are no differences in production technology between regions, the regional differences in early endowment of factors of production are used to explain the comparative advantage.

In international trade theory, the “increasing returns” approach is understood as the ability for nations to conduct trade due to the advantage provided by specialization, even when endowment of production factors is similar. The reason this approach was not adopted for a long time is because it was difficult to determine whether the increasing returns were something exogenous or endogenous to firms (Krugman 1991a). That is, in the case of the former, the perfect competition model could be applied, but if it were the latter, an imperfect competition model would be necessary. Therefore, to adopt the latter approach, it was necessary to wait until industrial organizational theories advanced in the 1970s and the theory of imperfect competition model was formulated.

Kaldor (1970) attempted to formalize the model of cumulative causation, which was later refined by Dixon and Thirlwall (1975). According to Kaldor (1970), regions that experience faster growth than other regions will start to specialize in exportation due to its competitive advantage. As a result, the growth process of that area will take on a cumulative character. Dixon and Thirlwall (1975) considered the feedback effect of a region’s growth against the increased competitiveness of the export departments to formulate the process of cumulative causation. “Verdoorn’s law” plays a central role in determining the region’s productivity growth rate. This law states that the rising output rate y improves the labor productivity rate q. This law is expressed in the following formula,where, a is the constant term representing autogenous growth of productivity, and λ is the Verdoorn coefficient.

In the field of business administration, this type of knowledge is called “tacit knowledge” and it is discussed in contrast to “codified knowledge.”

Detail can be found in Rosenthal and Strange (2004).

The size of the urban area is nothing but a proxy variable, assuming there is a high correlation between the range of economic activity and size of the urban area (Parr 2002).

According to the micro-foundation by Ciccone and Hall (1996), with the density effect, in addition to the influence of agglomeration and congestion that final goods manufacturing firms would receive, intermediate goods supplying firms’ markup ratio in relation to marginal cost is included.

However, there may also be an aspect in which productivity growth may invite more workers to flood in, and as a result employment density may increase. The method of instrumental variables was employed in the estimation to consider this endogenous problem.

Particularly, Henderson (2003) conducted a panel analysis using the plant data to increase the number of observations. He analyzed the effect of industrial agglomeration at the plant level by considering the observable characteristics of the plant (i.e., the size and scale of the plant, its age, and organizational structure) in addition to regional characteristics.

Originally, the supply and demand linkage effect for goods included not only the improvement of productivity but also an increase in consumers’ real income. According to spatial economics, firms try to produce differentiated goods at a single location to satisfy both the consumer’s desire for variety and economies of scale, but as they consider transportation cost to the market, they will try to build their plant near a large market to save on these costs. For this reason, firms will build in an area that has easy access to a large-scale market. At this point, there will be more specialized consumer goods suppliers moving into the area that start to supply a variety of consumer goods and the utility level for the consumers who prefer variety will increase. This increase in real income brings an influx of consumers from other areas, and because of expanding regional demand, the location of the firms in that area progresses. However, to the best of my knowledge, there have been few empirical studies that have attempted to find the influence of industrial agglomeration effects on a consumer market.

Such an effect where production activity in a particular industry influences the production activity of other industries through the market is called the “market linkage effect”. (Fujita et al. 1999; Fujita and Thisse 2002). Details will be covered in Chap. 5.