31 Krugman/Obstfeld/Melitz • International Economics: Theory & Policy, Ninth Edition
the size of the industry but not necessarily on the size of the firm; or as (2) internal economies, whereby the
production cost per unit of output depends on the size of the individual firm but not necessarily on the size
of the industry. Internal economies of scale give rise to imperfectly competitive markets, unlike the perfectly
competitive market structures that were assumed to exist in earlier chapters. Industries characterized by
purely external economies of scale will typically consist of many small firms and be perfectly competitive.
The focus of this chapter is on external economies, while the next chapter looks at internal economies.
External economies of scale (EES) lead to a clustering of firms in one location for three main reasons:
1. Specialized Suppliers: By locating next to firms in the same industry, you are able to specialize in one
aspect of the production process and outsource other stages of production to neighboring firms.
2. Labor Market Pooling: Firms with specific skill needs will prefer to locate near a large pool of workers
with those skills to limit labor market shortages. At the same time, skilled workers prefer to locate close
to the firms that hire them to limit unemployment.
3. Knowledge Spillovers: Having similar firms located next to one another can lead to increased sharing
of ideas and partnerships.
Market equilibrium in an EES industry is determined by the intersection of market demand and supply
as in the constant returns case. The key difference here is that the market supply curve is forward falling,
reflecting the fact that average costs in the industry actually fall as industry production (i.e., size) rises. This
distinction is the key driver of trade in this model. When two countries trade, it makes sense to concentrate
production in one country, since this will lead to lower average costs than splitting production across two
countries. With trade, the country with the lower average cost will export the good. This will lead to
more production in the exporting country and less production in the importing country. As the industry is
characterized by EES, this will cause costs in the exporting country to fall and costs to rise in the importing
country. Eventually, all production will locate in the exporting country at a lower market price than would
have prevailed without trade.
The pattern of trade is only partially explained by comparative advantage. Rather, it may be a historical
accident that led to the formation of an industry in a particular location. The chapter gives the example of
why global button manufacturing is concentrated in one town in China, mostly because one firm in the 1980s
began producing buttons there. That the location of production is not entirely dependent on comparative
advantage presents situations in which trade can actually make a country worse off. For example, if button
production is already established in China, then Chinese button producers have an advantage over firms in
countries without an established button industry (due to external economies of scale in the button industry).
Without trade, the button industry in a low-wage country could develop to the point where it is actually
producing at a scale such that the price of buttons is lower than the world price established by the Chinese
button industry. This suggests that a country could actually make itself better off by closing off from trade
to let external economies of scale industries develop. However, these cases may be difficult to identify and
protectionism can lead to unintended consequences such as retaliatory tariffs.
External economies may also be the result of learning curves (dynamic increasing returns). In this scenario,
the unit cost of production falls as the cumulative output of an industry rises. Industries that have been around
for a long time are further out on their learning curves and will have an advantage over new industries that
still have to undergo the process of “learning by doing.” The presence of these learning curves may justify
infant industry tariff protection, as a new industry in a country could potentially be competitive, but needs
to be protected until it develops the acquired knowledge of established global competitors. However, these
cases are hard to identify and present several problems to be discussed in later chapters (notably rent-seeking
behavior by protected firms).
© 2012 Pearson Education, Inc. Publishing as Addison-Wesley