Chapter 8

Paul Krugman, New Trade Theory and New Economic Geography

New trade theory emerged as an explanation of trade in the late 1970s onwards with the
pioneering work of Krugman (1979, 1980, 1981, 1986, 1990), Helpman and Krugman (1985,
1989), Brander (1981), Brander and Krugman (1983) among others. The traditional explanation
of trade was dominated by comparative advantage which is measured by relative labour costs
in Ricardo and by relative differences in factor endowments in Hechscher-Ohlin. In both cases it
is the difference in countries which gives rise to trade. While in one case relative labour costs
(or labour productivities) differ, in the other a country produces and exports a product which is
intensive in a factor with which it is abundantly supplied. Adam Smith had couched his
explanation of trade in terms of absolute costs (based on the division of labour and increasing
returns) but his explanation got overshadowed by that of Ricardo. As noted by Krugman (1990,
p. 4), since increasing returns models were considered to be mathematically messy, the theory
of international trade followed the path of least mathematical resistance.

The comparative advantage theory assumes perfect competition where price equals marginal
cost, and this makes these models mathematically tractable. But once one introduces
increasing returns, implying economies of scale and imperfect competition, price is no longer
equated with marginal cost as marginal cost pricing results in universal losses. The only
exception is the case where increasing returns depend on the size of the industry instead of the
size of the firm. In this case, increasing returns are external to the firm but internal to the
industry as a whole. During the 1970s, however, with the developments in industrial
organization, there was an explosion of models of imperfect competition. When these new
theoretical developments were applied to trade theory, ‘new trade theory’ emerged. These
new mathematical models made it possible not only to construct trade models based on
internal economies of scale but also models with external economies where increasing returns
arise in the production of intermediate goods.1

The new trade theory recognizes that comparative advantage is one reason for trade. But it
focuses on the role of increasing returns as a source of trade. For example, trade between rich
and poor countries or between manufactures and food follows comparative advantage. But
trade between advanced countries themselves is based on economies of scale (or increasing
returns) where manufactures exchange for manufactures. 2 Here each country specializes in a
limited range of goods to take advantage of economies of scale. But since consumers have
1
In Krugman’s (1990, p. 3) opinion: “Since economics as practiced in the English-speaking world is strongly
oriented toward mathematical models, any economic argument that has not been expressed in that form tends to
remain invisible.”
2
It should be noted that the new trade theory equates increasing returns with economies of scale at the firm or
industry level.

1
preference for variety, international trade leads to an increase in the variety of goods available.
The pattern of trade here is based on specialization and increasing returns. While the pattern of
trade between food and manufactures is interindustry trade, trade between manufactures and
manufactures is intraindustry trade.3 Intraindustry trade within standard industrial classification
is a substantial part of world trade and accounts for about one-fourth of world trade (Krugman
and Obstfeld 2003, p. 139).4 However, there is an element of arbitrariness to this story. To take
an example of airplanes manufactured in Seattle, it is difficult to find an explanation in some
special attribute that Seattle’s location possesses. It may just be a historical accident that the
industry got concentrated there. Once established, increasing returns keep the industry there.
The same is equally true for other locations such as Detroit or Silicon Valley.

Before the new trade theory came on the scene, the proposition that increasing returns could
be a cause of trade in similar countries was not well understood. As Krugman (1979) notes, it
was never covered in most textbooks or courses whether graduate or undergraduate. While
some authors (e.g. Ohlin 1933, Balassa 1967, Kravis 1971) recognized the importance of scale
economies in international trade, they received little attention in formal trade theory. “The
main reason for this neglect seems to be that it has appeared difficult to deal with the
implications of increasing returns for market structure” (Krugman 1979, p. 469).

Krugman (2002) credited Ohlin’s (1933) Intraregional and International Trade with many of the
insights (but not all) of the ‘new trade theory’ and ‘new economic geography’. For example,
Ohlin was aware that increasing returns were a cause of trade but he downplayed their
practical importance. Ohlin (as quoted in Krugman 2002) stated: “[I]t is certainly the differences
in factor supplies that determine the course of interregional trade – unless regions are small –
whereas the advantages of large-scale production are more in the nature of a subsidiary cause,
carrying the division of labour and trade a little further than it would otherwise go, but not
changing their characteristics.” According to Krugman, Ohlin was also a bit inconsistent; after
disparaging the practical importance of increasing returns in Part I of his book, in Part II he
seemed to suggest that they would play a growing role over time. Samuleson was aware that
Ohlin had stressed increasing returns as a subsidiary cause, but chose to stress the primary
argument in his formalization of Ohlin. Others who followed Samuelson simply took their cue
from him without bothering to read the original Ohlin (1933). As Krugman notes, they may have
skimmed through Ohlin, but their mind was already conditioned by Samuelson’s interpretation.
That is how Ohlin’s insights about the role of increasing returns to scale got lost over time in the
neoclassical discourse on trade starting with Samuelson.
3
The standard formula for calculating intraindustry trade (I) within an industry is: I = 1 – (|exports –
imports|)/(exports + imports). See Krugman and Obstfeld (2003, p. 139, f.n. 5).
4
Krugman et al. (2018, p. 212) point out that the proportion of intraindustry trade in world trade has grown over
the last half century. Depending on the coarseness of the industrial classification (whether into hundreds or
thousands), intraindustry trade ranges from one fourth to one half of world trade.

2
According to Krugman, even if Ohlin had stressed increasing returns more forcefully than he
actually did, his insights would still have met the same fate. Although Ohlin was aware of the
Marshallian distinction between internal and external economies, he was not always careful to
separate the cases. Moreover, he was not entirely clear what he meant by increasing returns. If
he meant internal economies, the implication is imperfect competition. If he meant external
economies, we are in the world of market failure. It was only with the development of tractable
monopolistic completion model that economies of scale and external-economy models became
tractable. Moreover, the habit of thinking in two-goods framework gives rise to a convex
production possibility curve, unless there are accidental and arbitrary increasing returns strong
enough to give rise to a concave curvature. Minus this concave curvature, the only way left to
think about increasing returns is their effect in modifying comparative advantage and giving
advantage to large countries in scale-intensive goods.

Krugman notes that if (1) there is a strong tendency of concentration of factors – either due to
external economies or due to linkages – and (2) some factors are more mobile than others,
then factor mobility accentuates regional differences instead of reducing them. That is how
concentrations of beautiful people in Hollywood, ambitious people in London and nerds in the
vicinity of Stanford (Silicon Valley) emerged. Ohlin appears to touch on all these elements.
Krugman also notes that Ohlin’s description of ‘agglomerating’ and ‘deglomerating’ tendencies
is strikingly similar to one of new economic geography’s central themes of tension between
centripetal and centrifugal forces. Similarly Ohlin’s insight that “Inventions may by chance lead
to manufacture in one country or place when others would do just as well” is akin to new
economic geography’s stress on historical accidents in determining industrial location.
However, Ohlin was less surefooted about these insights as compared to his insight about
increasing returns as a cause of trade.5

This chapter will first discuss the role of economies of scale and external economies in new
trade theory. It will also discuss new economic geography or the so-called geographical turn in
economics. Next, the chapter takes up the role of strategic trade policy in the context of
sophisticated arguments for governmental intervention and whether free trade is passé.

5
Krugman gives three reasons for this. First, even though Ohlin recognized that there were ‘arbitrary elements in
location’ and that the ‘historical influence can lead to an uneconomical location of industry’ his main emphasis is
that location decisions are likely to be mistaken. The main point, however, is that an industry may be in the wrong
place even if nobody is making a mistake. Second, Ohlin downplays the endogeneity of factor location attributing
regional differences in natural resources as the cause rather than the effect of location of industry. This may be
true of Europe but not in America where the manufacturing belts owed their existence to early starts rather than
to resource advantages. The location of activity in America in the 1920s was driven by the logic of self-reinforcing
concentration. Third, unlike the integration of increasing returns with comparative advantage which reinforced his
basic vision, the integration of increasing returns with factor mobility went counter to this vision. While in the
former case the integration leads to convergence of factor prices, in the latter case it leads to their divergence.

3
Krugman’s contribution to trade theory is also taken up. Finally, we make a critical assessment
of new trade policy and new economic geography.

The Role of Economies of Scale

Krugman and Obstfeld (2003, Chapter 6) discuss the role of economies of scale and external
economies in trade theory. Comparative advantage models assume constant returns to scale
and perfect competition. In practice, many industries are characterized by increasing returns or
economies of scale where doubling of industry inputs more than doubles the industry’s output.
When increasing returns operate, large firms usually have an advantage over smaller ones, and
the market may be dominated by a single firm (monopoly), a few firms (oligopoly) or a larger
number of differentiated firms (monopolistic competition). Pure monopoly is rare in practice as
a firm making high profits attracts competitors. The usual market structure is characterized by
oligopoly with several firms where each firm is large enough to affect prices but none is an
uncontested monopoly. Besides, pricing policies of firms under oligopoly are interdependent
where the firm will consider not only the expected responses of consumers but also of
competitors while setting prices. These responses in turn depend on what the competitors’
perception of how the firm in question will behave. It is therefore a complex game in which
each firm is trying to second-guess the others’ strategies. A special case of oligopoly is
monopolistic competition, a model which is easier to analyze, and has become popular with the
trade theorists in the 1980s.6 To bypass the problem of interdependence, two crucial
assumptions are made: (1) each firm produces a differentiated product (a product which is
somewhat different of its rivals, and (2) each firm, though facing competition from its rivals,
behaves like a monopolist by ignoring the impact of its own prices on other firms (i.e., each firm
assumes the prices charged by others as given). Automobile production in Europe is a good
example of monopolistic competition where a number of producers such as Ford, General
Motors, Volkswagen, Renault, Fiat, Volvo, Peugeot and Nissan offer different but competing
automobiles. However, the main appeal of the monopolistic competition model is not its
realism but its simplicity in explaining how economies of scale give rise to mutually beneficial
trade.

Under monopolistic competition, industry consists of a number of firms producing

differentiated products, i.e., products that are not the same but substitutes for one another.

6
Krugman and Obstfeld (2003, p. 131) note that there are few industries in practice which are well described by
monopolistic competition. The most common market structure is likely to be “small-group oligopoly” where only a
few firms actively compete. Thus the key assumption of monopolistic competition model that each firm behaves
like a true monopolist is likely to be violated. Interdependence of firms will come into play. There are two kinds of
behaviour in oligopoly which are excluded by assumption in the monopolistic competition model. First is collusive
behaviour to charge high prices or a tacit understanding to allow one firm to act as a price leader for the industry.
Second is strategic behaviour where, for example, a firm may build extra capacity to deter potential rivals from
entering the industry.

4
Each firm is a monopolist in the sense that it is the only firm producing that particular product
(or brand). But the demand for its good depends on the availability of similar products as well
as the prices on which they are offered. Taking cost and price on the vertical axis, and the
number of firms on the horizontal axis, market equilibrium under monopolistic competition will
be determined by the intersection of the upward sloping supply curve and downward sloping
demand curve. That is, at equilibrium both the number of firms as well the average price they
charge are determined. Equilibrium describes the maximum number of firms the industry can
support at zero profits. If the market size increases, the supply curve will shift outwards to the
right, and the equilibrium price will be lower and the number of firms larger. Consumers would
prefer to be in a larger market where they can benefit from lower price as well as larger variety
of products. Larger market can also result from international trade where each country
specializes in a restricted range of products to reap economies of scale and also benefit from
the larger variety of goods. Thus, there are gains from international trade both in terms of
larger variety and lower price.

To illustrate how economies of scale interact with comparative advantage to determine the
pattern of international trade, let us consider an example. Suppose there are two countries in
the world, Home and Foreign and two factors, labour and capital. Further suppose that Home
has higher capital-labour ratio (i.e., is a capital-abundant country) than Foreign. Let us also
imagine two industries – manufactures and food, with manufactures being more capital
intensive. The presence of economies of scale ensures that neither country can produce the full
range of manufactures by itself. Although each can produce some manufactures, they produce
different goods. If manufactures was not a differentiated product, Home will produce and
export manufactures and import food from Foreign. If, however, manufactures is
monopolistically competitive, Foreign will partly produce manufactures and partly food. Home
will specialize in manufactures only but these will be different from those produced in Foreign.

The world pattern of trade will be partly intraindustry (i.e., exchange of manufactures for
manufactures) and partly interindustry (i.e., exchange of manufactures for food). While
interindustry trade reflects comparative advantage, intraindustry trade reflects economies of
scale. A large part of advanced nations’ (or industrial nations’) trade in manufactured goods
consists of intraindustry trade based on economies of scale. Over time, industrial countries
have become quite similar in terms of technology, capital abundance or skilled labour. Though
economies of scale can be an independent cause of world trade, comparative advantage still
retains its importance. The pattern of intraindustry trade itself is unpredictable and has more to
do with accidents and history than anything else. Krugman and Obstfeld (2003, p. 138)
summarize the pattern of trade when comparative advantage and economies of scale interact:

5
 1. Interindustry (manufactures for food) trade reflects comparative advantage. The
pattern of interindustry trade is that Home, the capital-abundant country, is a net
exporter of capital-intensive manufactures and a net importer of labor-intensive food.
So comparative advantage continues to be a major part of the trade story.

2. Intraindustry trade (manufactures for manufactures) does not reflect comparative

advantage. Even if the countries had the same overall capital-labour ratio, their firms
would continue to produce differentiated products and demand of consumers for
products made abroad would continue to generate intraindustry trade. It is economies
of scale that keep each country from producing the full range of products for itself; thus
economies of scale can be an independent source of international trade.

3. The pattern of intraindustry trade itself is unpredictable. We have not said anything
about which country produces which goods within the manufactures sector because
there is nothing in the model to tell us. All we know is that countries will produce
different products. Since history and accident determine the details of the trade pattern,
an unpredictable pattern of trade is an inevitable feature of a world where economies of
scale are important. Notice, however, that unpredictability is not total. While the
precise pattern of intraindustry trade within the manufactures sector is arbitrary, the
pattern of interindustry trade between manufactures and food is determined by the
underlying differences between countries.

4. The relative importance of intraindustry and interindustry trade depends on how

similar countries are. If Home and Foreign are similar in their capital-labor ratios, then
there will be little interindustry trade, and intraindustry trade, based ultimately on
economies of scale, will be dominant. On the other hand, if capital-labor ratios are very
different, so that, for example, Foreign specializes completely in food production, there
will be no intraindustry trade based on economies of scale. All trade will be based on
comparative advantage.

Intraindustry trade produces additional gains from international trade over and above those
from comparative advantage. This is because intraindustry trade allows countries to gain from
larger markets. As noted earlier, intraindustry trade allows countries to restrict the number of
differentiated products and this makes available greater variety of goods. By producing fewer
varieties, countries can gain from larger scale, higher productivity and lower costs. Consumers
also benefit from greater varieties on offer. In Hechscher-Ohlin theory, owners of abundant
factors gain and owners of scarce factors lose from trade. Because of wide differences in
resources, technology and barriers to trade among countries, complete factor price
equalization never takes place. So trade based on comparative advantage has strong income
distributional effects. In intraindustry trade based on economies of scale, these effects are

6
smaller and everyone tends to gain from trade. Intraindustry trade will predominate between
countries of similar level of economic development. This is characteristic of advanced
manufactured goods rather than traditional products like textiles and footwear. 7

With the establishment of the European Economic Community in 1957 (Britain joined in 1973
and exited in 2020), trade within the EEC grew twice as fast as world trade during the 1960s.
Growth of trade was almost entirely intraindustry, and the expected drastic dislocations did not
occur. Workers of Germany’s electrical machinery sector did not gain at the expense of France’s
electrical machinery industry; in fact workers in both sectors gained through increased
efficiency of European integration. As a result, fewer political and social problems resulted from
increased trade in Europe. Thus, according to Krugman and Obstfeld, intraindustry trade in
manufactures has few serious income distribution effects as the European example shows.

Monopolistic competition model helps us to understand how increasing returns promote

international trade. Although this model recognizes that imperfect competition is a necessary
consequence of economies of scale, it does not take into account the consequences of
imperfect competition for international trade. One important consequence is that firms do not
charge the same price in the foreign market as in the domestic market. There may be price
discrimination which in international trade commonly takes the form of dumping. This is a
practice where the exported good is sold at a lower price than the same good sold
domestically.8 For dumping to occur, two conditions must be satisfied: (1) the industry must be
imperfectly competitive, and (2) markets can be segmented so that domestic consumers are
not able to easily purchase goods meant for exports.9

The possibility of reciprocal dumping was first noted by Brander (1981). Krugman and Obstfeld
note that reciprocal dumping between two firms each producing the same good can actually
give rise to international trade. Assuming transportation costs, if firms charge the same price at
home and abroad, there will be no trade. Each firm will limit the quantity it sells in its home

7
In the US in 2009, intraindustry trade indices were high for goods such as metalwork machinery (0.97), inorganic
chemicals (0.97), power generating machines (0.86) and medical and pharmaceutical products (0.85), and low for
furniture (0.30), clothing and apparel (0.11), and footwear (0.10). See Krugman et al. (2018, p. 213).
8
Price discrimination can also be of reverse type where the domestic consumer pays a lower price, but price
discrimination in favour of exports is more common. Since international markets are imperfectly integrated with
the domestic market because of transportation costs and protectionist barriers, domestic firms are more likely to
have a larger share of the home market. This implies that foreign sales are likely to be more price sensitive than
domestic sales.
9
In many countries including the US, dumping is regarded as unfair trade practice, and firms which are injured by
foreign firms can appeal through a quasi-judicial procedure. If the appeal is successful, an anti-dumping duty is
imposed. In the US, the appeal is made to the Commerce Department (which accepts most complaints about unfair
pricing), but the actual injury is determined by the International Trade Commission (which rejects about half the
cases). Economists regard dumping as a form of price discrimination and have reservations on singling out
dumping as a prohibited practice. They regard charging lower price or selling at a loss in the foreign market as a
normal business practice of gaining experience or breaking into a new market.

7
market because if it sells more it drives down the price. If the firm sells a bit more in the foreign
market, it can add to its profits even if the price charged is lower abroad. By raiding the other
market, each firm sells at a lower price abroad (net of transportation costs) but that price is still
above marginal cost. Though international trade in exactly identical goods in both directions
may appear socially wasteful, reciprocal dumping eliminates pure monopolies, leading to some
competition. This may represent a benefit that may cover up for wasteful transportation. The
net effect of such trade on a nation’s welfare is therefore uncertain.10

The Role of External Economies

When economies of scale operate at the level of an industry rather than a firm, we call them
external economies. When such economies operate it is possible to localize industry, i.e.,
concentrate production in a few locations to reduce costs even if the size of firms remains
small. Alfred Marshall was struck by the phenomenon of industrial districts such as hosiery
firms in Northampton or cutlery cluster in Sheffield. Marshall (1890) noted that there could be
three reasons for industrial clusters: (1) specialized suppliers, (2) labour market pooling, and (3)
knowledge spillovers.

Many industries use specialized equipment or support services. While an individual firm may
not provide a large enough market for these services, a localized industrial cluster consisting of
many small firms can collectively provide a market large enough to support a wide range of
specialized suppliers. Modern examples of such external economies include the semiconductor
industry of Silicon Valley, investment banking in New York and entertainment industry in
Hollywood. In Silicon Valley high technology firms are clustered because key inputs are easily
available, and many firms compete to provide them. The firms then specialize in what they do
best and contract out other aspects of their businesses to others. A high technology firm
entering some other location would face immediate disadvantage of not being able to access
Silicon Valley suppliers, and would have to produce these inputs itself. Or alternatively it will
have to deal with the Silicon Valley suppliers from a long distance.

A cluster of firms can also create a pooled market of workers with highly specialized skills. This
benefits producers who are saved possible labour shortages. It is also beneficial for skilled
workers as they are less likely to be unemployed. If firms are located in different cities, a low
demand in a particular location is likely to result in unemployment. If an industry is
concentrated in a single location, high demand for labour from one firm may be offset by low
demand from others. As a result, workers have a lower risk of unemployment. Besides, it
becomes easier to switch employers in a place like Silicon Valley, so losing a job with one

10
Brander and Krugman (1983) present a model of reciprocal dumping where lower price abroad is not related to a
higher price elasticity of demand (and therefore to accidental differences in country demands) but to systematic
reasons associated with oligopolistic behavour.

8
employer may not be a big catastrophe as one is very likely to get another job with a different
employer.

In innovative industries, being a few weeks or months behind the cutting edge technology or
design can seriously disadvantage a company. Companies can acquire technology either
through their own R&D efforts or through reverse engineering the products of their
competitors. A third source of technical knowledge can be informal exchange of information
among workers when an industry is concentrated in a single place. The workers from different
companies then mix socially and talk freely about technical matters. Even Marshall wrote that
in an industrial district with many firms mysteries of trade remain no mysteries. A new idea
soon becomes known to others which when combined with their own additions becomes the
source of further new ideas. In Silicon Valley, it is easier for firms to stay near the technological
frontier than elsewhere. Some firms even locate their R&D centres there to keep up with the
latest technology.

External economies arising out of specialized suppliers, pooled labour market and knowledge
spillovers are possible in a country with a large industrial sector. A country with large industry is
more likely to reap these external economies than a country with small industry, other things
remaining the same. External economies also lead to a downward sloping industrial supply
curve, i.e., the larger the industry’s output, the lower will be the price at which firms sell their
output. This is reminiscent of Marshall’s downward-sloping industry supply curve where firms
are small but there are increasing returns at the industry level, i.e., costs and prices are lower as
the industry expands its output.

External economies play an important role in international trade. Countries can sometimes get
locked into undesirable patterns of specialization, and there can even be losses from
international trade. Countries that start as large producers in certain industries tend to remain
large even if some other country could potentially produce that good at a lower cost. In other
words, these countries enjoy first mover advantage even in the presence of cheaper producers.
For example, Swiss watch industry consisting of a large number of small producers was, for
historical reasons, established first. Although Thai watch industry which came up later could
produce watches at a cheaper rate as Thai wages are lower, but it did not lead to Thailand
taking over the world market. This is because at zero level of production, Thai costs are higher
than the Swiss costs at equilibrium level of production, although in both cases long-run average
cost curve is downward sloping and Thai curve is lower than Swiss curve. So although Thai
industry can potentially make watches at a lower cost than Switzerland, Switzerland’s head
start enables it to establish its hold. This implies that historical accident plays an important role
in the presence of external economies, and established patterns of trade persist even if they go
counter to comparative advantage.

9
If Thailand were to block all trade in watches, it would be able to supply its domestic market at
a lower price based on its own demand. In the presence of international trade its domestic
consumers pay the world price established by the Swiss industry. This price is low enough to
block entry of Thai producers who must initially produce at higher cost (when Thai output is
zero). This leaves Thailand worse off in the presence of international trade than it would have
been if international trade was absent. Thailand has an incentive to protect its watch industry
from foreign competition, but in practice it is difficult to identify cases of external economies.
This constitutes one of the main arguments against activist government policies in trade.

Knowledge spillovers in an industry may give rise to a situation where production costs depend
on cumulative output rather than on current output. This is because the production costs of
individual firms fall as the overall industry accumulates experience. This is shown by a
downward sloping learning curve where unit costs decline with cumulative output. When costs
decline as a result of cumulative output (rather than current output), this is known as dynamic
increasing returns. Like ordinary external economies, dynamic external economies can also lock
a country into a head start or initial advantage. The new country will be prevented from
entering the industry because of initial higher costs. A new country could potentially consider
temporarily subsidizing or protecting its industry experiencing dynamic external economies to
enable it to gain experience. This is known as infant industry argument for protection so as to
enable an industry to gain experience. However, in practice it is just as hard to identify cases of
dynamic external economies as those involving non-dynamic increasing returns.

Increasing Returns and Economic Geography

The foregoing analysis of industrial districts brings us to the concept of economic geography or
the center-periphery pattern of industrial development. Apart from new trade theory, Krugman
(1991a, 1991b) can also be credited with pioneering work in economic geography – the location
of factors across space. Krugman (1991b) notes that economic geography occupies a relatively
small part of standard economic analysis. This is surprising in view of the striking real world
facts about economic geography. “For example, one of the most remarkable things about the
United States is that in a generally sparsely populated country, much of whose land is fertile,
the bulk of the populations resides in a few clusters of metropolitan areas; a quarter of the
inhabitants are crowded into a not especially inviting section of the East Coast. It has often
been noted that nighttime satellite photos of Europe reveal little of political boundaries but
clearly suggest a center-periphery pattern whose hub is somewhere in or near Belgium”
(Krugman 1991b, pp. 483-84).

Krugman (ibid., p. 484) asks the question: “Why and when does manufacturing become
concentrated in a few regions, leaving others relatively undeveloped?” To answer this question,
Krugman develops a simple model of ‘industrialized’ core and an agricultural ‘periphery’.

10
Geographical concentration is based on the interaction between economies of scale and
transportation costs. Although Marshall offered an account of localization of particular
industries, Krugman’s concern is to answer why manufacturing in general gets concentrated in
one or a few regions of a country (manufacturing core) and the remaining regions playing the
role of agricultural suppliers (agricultural periphery). Accordingly, his explanation focuses on
generalized external economies rather than on those external economies specific to a particular
industry. It is also assumed that the external economies which lead to the emergence of core-
periphery pattern are pecuniary external economies associated with demand or supply linkages
rather than pure technological spillovers.

Because of economies of scale, manufacturing will get concentrated in a few locations,

preferably near large markets to minimize transportation costs. Other locations can then be
served from these centrally located sites. Though some demand for manufactures will come
from the agriculture sector, a major source would be the manufacturing sector itself. This
creates the possibility of ‘circular cumulative causation’ (Myrdal 1957) or ‘positive feedback’
(Arthur 1990). That is, manufacturing gets concentrated where the market is large and the
market is large where manufacturing is concentrated. To this analysis one can add Hirschman’s
(1958) forward and backward linkages as these linkages are likely to be the strongest where
manufacturing concentrations already exist. As noted by Krugman, though this story of circular
processes is not new and has been familiar to economic geographers since the emergence of
the US manufacturing belt in the second half on the nineteenth century, no attempts were
made to formalize it.

How far will geographical concentration proceed? It depends on the parameters of the
economy – transportation costs, economies of scale and the share of expenditure on
nonagricultural goods. Thus, small changes in these parameters may have huge qualitative
effects. “[W] hen some index that takes into account transportation costs, economies of scale
and the share of nonagricultural goods in expenditure crosses a critical threshold, population
will start to concentrate and regions to diverge; once started, this process will feed on itself”
(Krugman 1991b, p. 487). And where will manufacturing concentration actually end up? Initial
conditions are important. If one region has slightly more population than the other,
transportation costs decline below threshold, and the region in question gains and others lose.
Had the distribution of population at the critical moment been slightly different, the roles of
different regions may have been reversed.

Krugman (1991c), in his paper ‘History versus expectations’, points out that in models of
external economies there could be two or more equilibria in the long run. Which equilibrium is
chosen depends on the role of history versus expectations. Much literature emphasizes the role
of history in setting the initial conditions, and it is this which determines the outcome. In the

11
traditional literature (e.g., Myrdal 1957, Kaldor 1972, David 1985) the role of history is only
implicit. In the formal literature (e.g., Arthur 1986; Krugman 1981, 1987a) the dynamics are
formulated in such a way that history is decisive. For example, in Krugman (1981) the dynamics
of capital accumulation are such that the region which starts with a slightly higher capital stock
eventually occupies the dominant industrial position. An alternative view stresses the role of
expectations or self-fulfilling prophesy in determining the equilibrium outcome. Rosenstein-
Rodan’s (1943) ‘big push’ and its formalization by Murphy e al. (1989) are good examples of this
alternative view. Krugman (1991c) presents a simple model with external economies and
adjustment costs where the eventual outcome depends on the parameters of the economy.
The relative importance of history and expectations depends on the underlying structure of the
economy, particularly on adjustment costs. Marshall had assumed that resources move slowly
in response to differences in current earnings. This implies that it is costly for them to do so. If
there are zero adjustment costs history becomes irrelevant. The slower the rate at which the
economy adjusts, the more likely it is that history matters. If adjustment is slow enough, history
becomes decisive.

Krugman (1999) presents a table showing centripetal and centrifugal forces of geographical
concentration. The list is by no means comprehensive but is of practical value. Krugman points
out that there is a tug of war between centripetal and centrifugal forces. These forces are
shown as follows:

Table 1

Centripetal forces Centrifugal forces

Market size effects (linkages) Immobile factors
Thick labour markets Land rents
Pure external economies Pure external diseconomies
Source: Krugman (1999), p. 143.

The centripetal forces listed in the first column are three classical Marshallian sources of
external economies discussed earlier. The centrifugal forces listed in the second column consist
of immobile factors (land and natural resources), high land rents (due to concentration of
economic activity) and pure external diseconomies (such as congestion, pollution etc.). For
modeling economic geography, one has to move away from the complexities of the real world
and consider more limited set of forces. Thus, most modeling since the start of the field
(Krugman 1991a, 1991b) takes the first item in each column, i.e., the interaction between
linkages and immobile factors. Two strategic modeling considerations dictate this choice. First,
we want to be able to avoid an approach which asserts that agglomeration takes place because
of agglomeration. That is, some distance has to be created between assumptions and
conclusions. Second, if location is at the heart of the problem, distance as measured by

12
transportation costs needs to enter the picture. Since linkages and access to immobile factors
are both tied to distance, modeling in a spatial setting becomes easier.

Krugman is however cautious in drawing any strong policy conclusions from new economic
geography despite the temptations.11 First, it is difficult to jump from small suggestive models
to empirical models that can be used for evaluating specific policies. There has been a long
debate on how to make strategic trade policy operational (discussed below). In the case of new
economic geography, effects are general equilibrium rather than partial equilibrium making the
task even harder. Perhaps more research is required before any strong policy conclusions are
drawn.

Sophisticated Arguments for Governmental Intervention and Strategic Trade Policy

In the advanced countries during the 1980s, new sophisticated arguments emerged for
governmental intervention in trade. These arguments focused on high technology industries
such silicon chips (or semiconductors). While some of these arguments were related to market
failure, a new controversial theory of strategic trade policy, based on different ideas, emerged.

The main problem with the market-failure arguments is the problem of identifying it. In
advanced countries two such failures have been identified: (1) inability of high technology firms
to completely capture knowledge spillovers to other firms, and (2) the presence of monopoly
profits in highly concentrated oligopolistic industries. In the high technology industries where
there are substantial externalities, there is a good case for providing a subsidy to the industry.
This is akin to infant industry protection of less developed countries, but in advanced countries
the argument acquires a special edge, as in high technology industries which are central to the
generation of new knowledge through R&D expenditures. If other firms are able to imitate the
products of innovating firms through reverse engineering, and because patent laws provide
weak protection for innovators, the incentive for innovation is weakened.

The government can subsidize industries where knowledge generation is believed to take place,
but this involves using the instrument bluntly. Alternatively, the government can subsidise R&D
wherever it happens. But here the problem is of definition, a lose definition leading to abuse
and a strict definition would favour large bureaucratic organizations (where the allocation of
funds can be strictly documented) over smaller informal organizations (where original thinking
is widely believed to happen). How much should be the level of subsidy? No one has a good
idea of the optimal level of subsidy – whether it should be 10, 20 or 100 percent. Knowledge
spillovers are in the nature of externalities, do not carry a market price, and are therefore hard
11
These temptations for governmental intervention from new economic geography literature may include: (1) the
market may not get it right, (2) small policy interventions have large effects, and (3) cumulative processes of
concentration produce winners and losers, so government authorities have an incentive to ensure that their
country emerges as a winner.

13
to measure. Moreover, the benefits may spillover to firms in other countries, so appropriability
at the level of a nation may be a problem even in a large country such as the US. Despite these
limitations, technological spillovers argument is probably the best case one can make for an
active industrial policy.

In a Journal of International Economics article Brander and Spencer (1985) put forward a new
argument for governmental intervention which generated considerable controversy. 12 They
point out that some industries are dominated by small number of firms enjoying excess returns
over and above what equally risky investments can earn elsewhere. This will lead to
international competition over who gets these profits. In principle, the government by
subsidizing domestic firms, can shift these profits from foreign to domestic firms. Subsidy to
domestic firms will deter foreign firms from investing in the domestic excess-returns industries,
domestic firms are able to earn profits exceeding the amount of the subsidy. Ignoring the effect
on the consumers, subsidy raises a country’s national income at the expense of other countries.

To take a simple example to illustrate the point, suppose there are two countries – the US and
Europe, and two firms which can produce a new design aircraft, namely, Boeing and Airbus. As
the example is set up, either firm alone can make profits; if both produce the new design both
make losses. The question is, who gets the profits? Suppose Boeing has a small head start,
Airbus will have no incentive to enter. Suppose Europe is able to subsidise Airbus, this creates
an advantage in favour of Airbus comparable to the strategic first mover advantage Boeing had.
Critics have pointed out that the use of strategic trade policy is likely to require more
information than is likely to be available. Secondly, such policies which increase domestic
welfare at the cost of some other country are beggar-thy-neighbour policies, and therefore risk
retaliation. The resulting trade war leaves everyone worse off. Third, the domestic politics may
itself prevent the use of such subtle policy tools. 13 Special interest groups may be able to buy
influence. For example, in the US contributions by business and labour had an impact (in terms
of the number of votes polled) on the 1993 vote on NAFTA or 1994 vote on GATT (See Baldwin
and Magee, 1998).

Krugman (1987b) asks the question: is Free Trade Passé? He observes that free trade based on
comparative advantage was an almost sacred tenet in economics since the publication of
Ricardo’s Principles of Political Economy. Yet, in the current context doubts are being cast on

12
Other early contributions to strategic trade policy include Brander and Spencer (1981, 1983), Krugman (1984),
Dixit (1984) and Eaton and Grossman (1986). See also Spencer and Brander (2008) entry on strategic trade policy in
The New Palgrave Dictionary of Economics.
13
Krugman and Obstfeld (2003, p. 221) point out that in practice there is a political argument for free trade though
there may be better policies in principle. Trade policy in practice is often dominated by special-interest politics
rather than national costs and benefits. Though selective tariffs and subsidies can sometimes be shown to increase
national welfare, but in practice a sophisticated programme of trade intervention is subject to capture by interest
groups and there may be redistribution of income in favour politically influential sectors.

14
the case for free trade. This has happened not because of political pressures but due to radical
changes in international trade theory which emphasize increasing returns and imperfect
competition. These new models not only doubt the extent of trade which can be explained by
comparative advantage but also make out the case that import restrictions and subsidies under
some circumstances may increase national welfare.

Krugman points out that the positive conclusions of the new trade theory that much of trade
reflects increasing returns and that many international markets are imperfectly competitive
have been readily accepted by the profession. However, the normative conclusion that this
justifies greater degree of governmental intervention in trade has been sharply criticized
including by those who created the new theory. The critique partly reflects the politics of trade
policy which was mentioned earlier. “There are, however, three economic criticisms. First,
critics suggest that it is impossible to formulate useful interventionist policies given the
empirical difficulties involved in modeling imperfect markets. Second, they argue that any gains
from intervention will be dissipated by the entry of rent-seeking firms. Third, it is argued that
general equilibrium considerations radically increase the empirical difficulty of formulating
interventionist trade policies and make it even more unlikely that these policies will do more
good than harm” (ibid., p. 139).

Thus political economy concerns as well as the economic criticisms combine for the
reemergence of free trade as a rule of thumb. Also, when the expected gains from intervention
become sufficiently diluted, the political economy argument becomes overriding. Krugman
points out that the reemergence of free trade is a “sadder but wiser” argument especially “in a
world whose politics is as imperfect as its markets”. It is therefore possible to believe that while
comparative advantage as a model is incomplete, free trade nevertheless is a good practical
policy. The conclusion is:

[F]ree trade is not passé, but it is an idea that has irretrievably lost its innocence. Its
status has shifted from optimum to reasonable rule of thumb. There is still a case for
free trade as a good policy, and as a useful target in the practical world of politics, but it
can never again be asserted as a policy that economic theory tells us is always right
(ibid., p. 132).

Krugman’s Contribution

Krugman (2008a) in his Nobel Prize lecture places his contribution to new trade theory and new
economic geography in perspective. He describes ‘new trade theory’ as an unfortunate phrase.
For one thing, what was once new has already become old. For another, the increasing returns
revolution in trade and geography may have already peaked, and the world may be more
classical today than when the revolution in the trade theory began. For example, the index of

15
regional manufacturing specialization based on Krugman (1991a) reached its peak during the
interwar period and has since declined dramatically. After WWII, the manufacturing belt began
to dissolve, and after the 1950s industry spread to the sunbelt. The role of increasing returns, as
opposed to comparative advantage, in trade was applicable after 1950 when similar-similar
trade between advanced countries increased. However, during the last few decades similar-
similar trade is giving way to trade between rich and poor low-wage countries like China.
Krugman (2008b) proposes an indicator of this shift – the average hourly compensation of
workers in top US trading partners as a percentage of US compensation. The indicator which
stood at 76 in 1975 slightly rose to 81 in 1990, and by 2005 had fallen to 65 largely reflecting
the rapid increase of trade with China and Mexico. In 2006, for the first time, the US trade with
developing countries exceeded that with advanced countries. Krugman (2008a, pp. 346-7)
concludes:

And nobody doubts that trade between the United States and Mexico, where wages are
only 13 percent of the US level, or China, where they are about 4 percent, reflects
comparative advantage rather than arbitrary scale-based specialization. The old trade
theory has regained relevance… Both new geography and new trade, then, may describe
forces that are waning rather than gathering strength. Yet they are hardly irrelevant.
And even the fact that they may be losing force is itself an important insight… Whether
the influence of increasing returns on trade and geography is rising or falling, one thing
is clear: much was learnt from the intellectual revolution that brought increasing returns
to the heart of how we think about the world economy.

In his first year as assistant professor when Krugman began working on international trade
theory, he was dissuaded by his colleagues as they thought that the theory was complete and
there was nothing interesting left to do. However, Krugman was aware that there was an
undercurrent of dissatisfaction with the conventional trade theory as exemplified by Balassa
(1966), Gruber and Lloyd (1975) among others. Besides, there was huge similar-similar trade
between similar countries as exemplified by two-way trade in automotive products between
the US and Canada. Still in 1980, it was a new phenomenon and trade was dominated by
comparative advantage between dissimilar countries. Balassa (1966) was quite clear that
countries specialized in a narrow range of products, importing the rest from others, because
this helped produce machinery and intermediate products and reap economies of scale. Yet
such ideas of similar-similar trade were incomprehensible because economies of scale at the
firm level implied imperfect competition, and there were no readily usable general equilibrium
models of imperfect competition. This led Harry Johnson (1967) to remark that the theory of
monopolistic competition had virtually no impact on the theory of international trade. But
when new models of monopolistic competition were put forward by Spence (1976), Dixit and
Stiglitz (1977), and Lancaster (1979), it became obvious that these new ‘gadgets’ could be used

16
to explore the role of increasing returns in a number of ways including intra-industry trade. In
the works of Norman (1976), Krugman (1979) and Lancaster (1980), it became obvious that one
could use monopolistic competition models to bypass comparative advantage. In this setting,
countries with similar resources and technology specialize in different products and obtain the
rest through trade as consumers prefer variety. An extension of this approach by Dixit and
Norman (1980) and Helpman (1981) sought to bring back comparative advantage by assuming
that all differentiated products were produced using the same factor proportions. One could
then explain inter-industry trade in terms of Hechscher-Ohlin, with the overlay of intra-industry
specialization due to increasing returns.

Balassa (1966) pointed out that trade liberalization among industrial countries had proved to be
largely non-disruptive as Europe’s industrial landscape had not changed much after the
formation of the common market and nor were there any serious income distribution effects.
This was because trade had taken the form of intra-industry specialization rather than inter-
industry specialization. Krugman (1981) tried to capture this phenomenon by showing that
Stolper-Samuelson theorem where trade hurts a country’s scarce factor is strictly valid if
comparative advantage is strong and/or economies of scale weak. In case of intra-industry
trade between industrial countries, trade was win-win for all parties. There was another
important insight from the application of Dixit and Stiglitz (1977) type models to trade.
Krugman (1980) shows that increasing returns provide an incentive to concentrate production
in one location to minimize transportation costs. Production is concentrated near the largest
market and this location then is used to supply other locations.

Helpman and Krugman (1985) tried to recreate Samuelson’s insight that factor-price
equalization theorem could only work if specialization and trade took place as if in an
integrated world economy. They tried an approach to trade which combined both increasing
returns and comparative advantage. The key insight, in an integrated economy framework, was
that all goods subject to economies of scale be located in a single country. This approach also
made it clear that increasing returns reinforce, rather than call into question, the gains from
trade. Although it is possible to conjure up examples where trade under increasing returns
leads to losses, this approach points to gains from trade even under increasing returns, an
outcome which is better than what was previously thought.

Krugman points out that monopolistically competitive models gave an impetus to new trade
theory, but it did not remain confined to those models. By the mid-1980s, approaches involving
external economies, oligopoly and even contestable markets were applied to new trade theory.
What made this possible was a shift in the attitude of trade theorists. First, there was a change
in focus from proving general results (based on some upfront assumptions) to special cases.

17
Second, attention shifted from detailed prediction – which country produces what – to the
aggregate description of the pattern of trade.

Krugman observes that since the publication of Ohlin’s (1933) Intraregional and International
Trade, the motives for shipment of goods within countries are similar to those between
countries. One might have expected that international trade theory and economic geography
theory to have developed in tandem. However, the truth is that trade theorists paid little
attention to trade within national boundaries or location of production across space. Despite
Marshall, in his Principles of Economics, devoting a whole chapter on ‘The concentration of
specialized industries in particular localities’, the subject stayed away from the standard
economics curriculum. The reason, as Krugman notes, was the centrality of increasing returns in
geographical concentrations. “As long as trade theorist shied away from increasing returns in
general, economic geography wasn’t an inviting field” (Krugman 2008a, p. 342). Further, work
in urban economics led economists to gibes like ‘agglomeration takes place because of
agglomeration economies’. Particularly, thinking about spatial reach of agglomeration
economies posed a problem. Fujita (1988) used monopolistic competition models to derive
endogenous externalities, but his assumption that monopolistically produced goods are
completely nontradable once again meant that the problem of spatial limits of spillovers
remained unresolved.

The character of US geography, Krugman points out, underwent a phase change in mid-
nineteenth century and the nation became differentiated between farm and manufacturing
belts. The features of this change were the rise of large-scale production (economies of scale),
railroads (lower transportation costs) and declining share of agricultural production (more
mobile production). Krugman’s (1991b) ‘core-periphery’ model was the starting point of new
economic geography. It was essentially a model of agglomeration, and it became obvious that
one also needed models to deal with other aspects of geography such as regional specialization
of different industries and the system of cities. By assuming a few things such as increasing
returns in intermediate as well as final production, it became possible to address many of these
issues. The willingness to focus on tractable special cases, as was the case in trade theory,
became important. For example, in Fujita, Krugman and Venables (1999), increasing returns
models of Dixit-Stiglitz type are combined with ‘iceberg’ transportation costs that are
proportional to fob prices; use simple adaptive dynamics to arrive at equilibria, and numerical
simulations to deal with complex calculations. The method has been described as ‘Dixit-Stiglitz,
icebergs, evolution, and the computer’.

There was renewed interest in Marshall’s trinity of reasons for local specialization. The original
new trade models only modeled specialized suppliers, but with the emergence of new
economic geography the other two reasons, namely knowledge spillovers and labour-market

18
pooling, also received attention. Also, there was an explosion of empirical work after 1990.
Economists became interested in cross-city comparisons to shed light on such subjects as
externalities, innovation and growth.

Critical Appraisal

Comparative advantage theory assumes constant returns to scale and perfect competition.
Given these assumptions, trade can arise as a result of differences in tastes, technology or
factor endowments. The Ricardian theory emphasizes differences in technology, while
Hechscher-Ohlin-Samuelson model focuses on differences in factor endowments. As a result of
these differences, trade is shown to be mutually beneficial to the participating countries.
Comparative advantage theory makes free trade as an almost sacred idea in economics.
Thoughtful economists have long known that comparative advantage is an incomplete story in
trade. But the development of theory which could incorporate increasing returns was long held
up because of the difficulties in modeling imperfect market structure. The developments in
industrial organization theory solved this problem when these models were applied to trade
theory. The new trade theory, which emerged in the late seventies and early eighties,
emphasizes increasing returns and imperfect competition. Here trade arises not because of
dissimilarities between countries but due to economies of scale. Intraindustry trade takes place
between industrial countries at a similar level of development. A case is made out for
sophisticated arguments for protectionism and also for the use of strategic trade policy under
certain circumstances.

We noted several shortcomings of the new trade theory. First, the pattern of intraindustry
trade is itself unpredictable. History and accident determine the details of intraindustry trade.
Second, evidence suggests that similar-similar trade between advanced countries may be giving
way to trade between the rich (high wage) countries and poor (low wage) countries. 14 Also, new
trade theory and new economic geography may be waning forces. Although these forces have
hardly become irrelevant, the world itself may have turned more classical. Third, although
sophisticated arguments for intervention argue that intervention may be welfare enhancing
under some conditions, these welfare effects may be exaggerated. Once the net benefits of
interventionism are sufficiently whittled down, political economy arguments take over and call
for free trade as a practical policy that avoids succumbing to special interest groups and
lobbies. But the idea of free trade has lost its earlier innocence – free trade may not always be
right, contrary to what the economic theory tells us.
14
Krugman et al. (2018, pp. 47-8) point out that till the 1970s, developing countries mainly exported primary
products, but since then are moving rapidly into manufactured exports. During 1960-2001 there has been a
complete reversal of relative importance of agricultural goods and manufactures in developing countries exports.
While agricultural exports as a percentage of developing countries’ exports declined from about 60 percent to
about 10 percent, manufactured exports increased from about 10 percent to over 65 percent. China’s
manufactured exports now constitute more than 90 percent of its total exports.

19
The main shortcoming of the new trade theory perhaps lies in its supply side view of the
matter. It looks at increasing returns as an important source of trade, but the original authors of
increasing returns such as Smith (1776) thought of increasing returns as the effect of trade. By
increasing the size of the market, international trade carries the division of labour in a country
to utmost perfection. In terms of cumulative causation framework, as propounded by Young
(1928), increasing returns are as much the effect of trade as its cause. Moreover, these authors
had a generalized (or macroeconomic) notion of increasing returns rather than a notion
confined to particular firms or industries. Young had also talked about an imperfect market
structure as closer to reality, but for him imperfections arose because of the imperfect working
of competitive process in practice. Thus, imperfect competition for him was a part of the real-
world competition where each producer tried to be a bit different from others and where
selling costs were not to be thought as social wastes but competitive investments. Threat of
entry put a limit on the prices charged by producers and hence on their profits.

Deraniyagala and Fine (2001, p. 812) see new trade theory as basically neoclassical economics
but incorporating four innovations. These innovations are market imperfections, strategic
behaviour, new growth theory and political economy arguments. Though the policy of
intervention is allowed for, the main thrust is on free trade mainly arising from political
economy (or rent seeking) arguments. However, these political economy arguments are not
very convincing:

If there are underlying economic and political interests in favour of trade policy, why
should they allow trade liberalization to proceed? And, if they have no choice, might
they not engage in even more costly forms of pursuing their advantage? This is exactly
what is perceived to have happened, if not anticipated, in the wake of the Uruguay
Round, with trade policy pursued by other means and, most notably, through
antidumping measures which have become the new form of (privatized) trade policy in
the WTO era (ibid., p. 816).

Derniyagala and Fine point out that the neoclassical consensus in favour of trade liberalization
has done a double disservice both by undermining interventionist trade policy and its
integration with other policy areas. After surveying the empirical literature on trade
liberalization, they conclude: “First, there is little to suggest that trade policy is itself an
important determinant of industrial performance… Secondly, at least as important as trade
policy have been the other elements of industrial policy, such as R&D, and the impact of
technology transfer and the scale and growth of domestic markets… Thirdly, trade policy
involves a very wide variety of complex instruments with an equally varied set of outcomes
depending upon how trade policies interact with other policies… Consequently, trade policy

20
should not be seen in isolation from other policies and as a bias in one direction or the other”
(ibid. p. 820).

Ron Martin (1999), while welcoming the new found interest by economists in geography, points
out that thus far geographers have not been particularly impressed with this ‘geographical turn’
in economics. The new economic geography has very little resonance with the theoretical and
empirical concerns of economic geography proper. It represents reworking of regional science
and urban economics models – approaches which were discarded by geographers long ago.
Their preoccupation with mathematical modeling leads to neglect of real geography. The
results of these mathematical models are not particularly novel and the empirical applications
trivial. Thus, the conclusion reached is: “the ‘new geographical economics’ represents a case of
mistaken identity: it is not that new, and it most certainly is not geography” (ibid., p. 67).

References

Arthur, W. Brian (1986), ‘Industry location patterns and the importance of history’, Center for
Economic Policy Research, Paper No. 84, University of Stanford.

Arthur, W. Brian (1990), ‘Positive feedbacks in the economy’, Scientific American,

262(February), pp. 22-99.

Balassa, B. (1966), ‘Tariff reductions and trade in manufactures among industrial countries’,
American economic Review, 56(3), pp. 466-73.

Balassa, B. (1967), Trade Liberalization Among Industrial Countries: Objectives and Alternatives,
New York: McGraw-Hill.

Baldwin, Robert E. and Christopher S. Magee (1988), ‘Is trade policy for sale? Congressional
voting on recent trade bills’, National Bureau of Economic Research, Working Paper No. 6376.

Brander, James A. (1981), ‘Intra-industry trade in identical commodities’, Journal of

International Economics, 11, pp. 1-14.

Brander, James A. and Barbara J. Spencer (1981), ‘Tariffs and the extraction of foreign
monopoly rents under potential entry’, Canadian Journal of Economics, 14, pp. 371-89.

Brander, James A. and Barbara J. Spencer (1983), ‘International R&D Rivalry and Industrial
Strategy’, Review of Economic Studies, 50, pp. 707-22.

Brander, James A. and Barbara J. Spencer (1985), ‘Export subsidies and international market
share rivalry’, Journal of International Economics, 18, pp. 83-100.

21
Brander, James A. and Paul Krugman (1983), ‘A reciprocal dumping model of international
trade’, Journal of International Economics, 15, pp. 313-21.

David, Paul (1985), ‘Clio and the economics of QWERTY’, American Economic Review, LXXV, pp.
332-7.

Deriyagala, Sonali and Ben Fine (2001), ‘New trade theory versus old trade policy: a continuing
enigma’, Cambridge Journal of Economics, 25, pp. 809-25.

Dixit, A. K. (1984), ‘International trade policy for oligopolistic industries’, Economic Journal, 94
(Supplement), pp. 1-16.

Dixit, A. K. and V. Norman (1980), Theory of International Trade: A Dual, General Equilibrium
Approach, London: Cambridge University Press.

Dixit, A. K. and J. E. Stiglitz (1977), ‘Monopolistic competition and optimum product diversity’,
American Economic Review, 67(3), pp. 297-308.

Eaton, J. and G. M. Grossman (1986), ‘Optimal trade and industrial policy under oligopoly’,
Quarterly Journal of Economics, 101 (May), pp. 383-406.

Fujita, M. (1988), ‘A monopolistic competition model of special agglomeration: a differentiated

product approach’, Regional Science and Urban Economics, 18, 87-124.

Fujita, M., Paul Krugman and A. Venables (1999), The Spatial Economy, Cambridge: MIT Press.

Gruber, H. G. and P. J. Lloyd (1975), Intra-Industry Trade: The Theory and measurement of
International Trade in Differentiated Products, London: Macmillan.

Helpman, Elhanan (1981), ‘International trade in the presence of product differentiation,

economies of scale and monopolistic competition’, Journal of International Economics, 11, pp.
305-40.

Helpman, Elhanan and Paul Krugman (1985), Market Structure and Foreign Trade, Cambridge:
MIT Press.

Helpman, Elhanan and Paul Krugman (1989), Trade Policy and Market Structure, Cambridge:
MIT Press.

Hirschman, Albert O. (1958), The Strategy of Economic Development, New Haven: Yale
University Press.

22
Johnson, Harry (1967), ‘International trade theory and monopolistic competition theory’, in R.E.
Kuenne (ed.), Monpolistic Competition Theory – Studies in Impact: Essays in Honour of Edward
H. Chamberlin, New York: John Wiley, pp. 203-18.

Kaldor, Nicholas (1972), ‘The irrelevance of equilibrium economics’, Economic Journal, 82, pp.
1237-55.

Kravis, I. (1971), ‘The current case for import limitations’, in United States Economic Policy in an
Interdependent World, Commission on International Trade and Investment Policy, Washington:
Government Printing Office.

Krugman, Paul (1979), ‘Increasing returns, monopolistic competition, and international trade’,
Journal of International Economics, 9 (November), pp. 469-79.

Krugman, Paul (1980), ‘Scale economies, product differentiation, and the pattern of trade’,
American Economic Review, 70, pp. 950-9.

Krugman, Paul (1981), ‘Intraindustry specialization and gains from trade’, Journal of Political
Economy, 89, pp. 959-73.

Krugman, Paul (1984), ‘Import protection as export promotion: international competition in the
presence of oligopolies and economies of scale’, in H. Kierzkowski (ed.), Monopolistic
Competition and International Trade, Oxford: Oxford University Press.

Krugman, Paul (1986) (ed.), Strategic Trade Policy and New International Economics,
Cambridge: MIT Press.

Krugman, Paul (1987a), ‘The narrow moving band, the Dutch disease, and the competitive
consequences of Mrs. Thatcher: notes on trade in the presence of dynamic economies of scale’,
Journal of Development Economics, XXVII, pp. 41-55.

Krugman, Paul (1987b), ‘Is free trade passé?’, Journal of Economic Perspectives, 1(2), pp. 131-
44.

Krugman, Paul (1990), Rethinking International Trade, Cambridge: MIT Press.

Krugman, Paul (1991a), Geography and Trade, Cambridge: MIT Press.

Krugman, Paul (1991b), ‘Increasing returns and economic geography’, Journal of Political
Economy, 99, pp. 483-99.

Krugman, Paul (1991c), ‘History versus expectations’, Quarterly Journal of Economics, 106(2),
pp. 651-67.

23
Krugman, Paul (1999), ‘The role of geography in development’, International Regional Science
Review, 22(2), pp. 142-61.

Krugman, Paul (2002), ‘Was it all in Ohlin?’, in R. Findlay, L. Jonung and M. Lundahl (eds.), Bertil
Ohlin – A Centennial Celebration (1899-1999), Part IV, Cambridge: MIT Press.

Krugman, Paul (2008a), ‘The increasing returns revolution in trade and geography’, Prize
Lecture (8 December), [Link], pp. 335-48.

Krugman, Paul (2008b), ‘Trade and wages reconsidered’, Brooking Papers in Economic Activity,
Spring.

Krugman, Paul and Maurice Obstfeld (2003), International Economics, Sixth Edition, Boston:
Addison Welsley.

Krugman, Paul, Maurice Obstfeld and Marc J. Melitz (2018), International Economics, Eleventh
Edition, Harlow: Pearson.

Lancaster, K. J. (1979), Variety, Equity and Efficiency, New York: Columbia University Press.

Lancaster, K. J. (1980), ‘Intra-industry trade under perfect monopolistic competition’, Journal of

International Economics, 10, pp. 151-75.

Marshall, Alfred (1890), Principles of Economics, Eighth Edition, London: Macmillan (1920).

Martin, Ron (1999), ‘The new ‘geographical turn’ in economics: some critical reflections’,
Cambridge Journal of Economics, 23, pp. 65-91.

Murphy, Kevin M., A. Schliefer and R. Vishny (1989), ‘Industrialisation and the big push’,
Journal of Political Economy, 97(5), pp. 1003-26.

Myrdal, Gunnar (1957), Economic Theory and Underdeveloped Regions, London: Gerald
Duckworth.

Norman, V. (1976), ‘Product differentiation and trade’, UK Economic Theory Study Group,
University of Warwick.

Ohlin, Bertil (1933), Intraregional and International Trade, Cambridge: Harvard University Press.

Rosenstein-Rodan, Paul (1943), ‘Problems of industrialization in eastern and south-eastern

Europe’, Economic Journal, 53(June-September), 202-11.

Smith, Adam (1776), An Inquiry Into the Nature and Causes of the Wealth of Nations, Edwin
Cannan (ed.), Chicago: University of Chicago Press (1976).

24
Spence, A. M. (1976), ‘Product selection, fixed costs and monopolistic competition’, Review of
Economic Studies, 43, pp. 217-36.

Spencer, Barbara J. and James A. Brander (2008), ‘Strategic trade policy’, in Steven N. Durlauf
and Lawrence E. Blume (eds.), The New Palgrave Dictionary of Economics, Second Edition,
Basingstoke and New York: Palgrave Macmillan.

Young, Allyn (1928), ‘Increasing returns and economic progress’, Economic Journal, 38, pp. 527-
42.

25