Book Reviews 491

uneven nature of aircraft purchasing makes learning-by-doing of little significance. Rather, the
rising concentration in this industry is best explained via Sutton's R&D escalation mechanism.
As industry R&D/sales ratios rose, profitability fell and fewer and fewer firms were able to
survive.
It is difficult to survey adequately the richness of Sutton's analysis. He has established a
benchmark for research in the impact of technological change on market structure. His modes
of analysis should have wide application as the pace of technological changes continues.

John Vahaly
University of Louisville

The Spatial Economy: Cities, Regions, and International

Economic Geography has received substantial attention in recent years and the authors of this
book have been at the forefront of this literature. In essence, this book is a revision and en-
hancement of journal articles written by subsets of these authors. In the time since the original
works were published, the authors have figured out ways to sharpen their results and to add
some new ones. With the additional space afforded by a book as opposed to a journal article,
they have room to consider a wider set of cases. There are also more references to earlier work
in regional science and geography. It is clear that if one wants to cover this material in a graduate
course, one should use this book rather than the original journal articles.
Since the book aims to be a complete theoretical analysis, much of the book involves
complicated derivations (many chapters have more than 30 equations). In no way is this book
aimed at a popular audience or for any kind of typical undergraduate course. Nevertheless, those
who have come to expect excellent writing in any manuscript with Krugman's name on it will
not be disappointed. Serving the equations up along side nicely written prose makes the equa-
tions easier to swallow. Also, the authors do a good job of highlighting the key elements of the
models in introductory chapters; by the time you start trudging through the equations you have
a pretty good idea of what is in them.
The book has three main sections, each corresponding to a journal article antecedent. The
first section looks at "regional economics" and is based on the famous Krugman (1991) Journal
of Political Economy article. In the authors' view, a regional model is one where factors such
as labor are free to flow across regions but other factors such as land are fixed. The point of
the analysis is to show how with scale economies, product differentiation, and transportation
costs, a core-periphery structure can emerge endogenously. The core will attract workers because
of the appeal of the large variety of differentiated products found in the core. The core will
attract firms because of the appeal of large markets for their goods and a large pool of potential
workers (forward and backward linkages in the parlance of regional economics). Thus a grav-
itational pull of workers and firms to the core is endogenously created. The second section
looks at "urban economics," and part of this section draws from Fujita and Mori (1997). This
section succeeds in getting arent gradient (the price of land falls as we move away from the
city) as well as a system of cities that looks something like Christaller's central place theory.
The third section looks at "international economics" and is based on Krugman and Venables
492 Book Reviews

(1995). Here international economics is just like regional economics except that labor is no
longer a mobile factor.
Ideas such as central place theory and core-periphery theory are as old as the hills. The
authors make no claim that they are the first to propose these ideas. They recognize that others
have put forth these ideas either with informal discussions or with formal models of social
planner problems. As the authors see it, their contribution is to show that these outcomes can
arise in a formal equilibrium model where agents are maximizing. What makes this a technical
challenge is that scale economies are crucial in these stories. From the outset, we have to rule
out our workhorse model of perfect competition. They are able to overcome the technical
challenge by trotting out the Dixit-Stiglitz model of monopolistic competition, where the use
of constant elasticity of substitution (CES) production functions results in substantial simplifi-
cations. Their use of this model in the book is so pervasive that the authors joke they could
have renamed the book, Games You Can Play with CES Functions.
Economists do not usually pay attention to an argument until someone shows the argument
holds together in a formal equilibrium model. Core-periphery theory and central place theory
are important ideas so the authors have done a service to the profession by working them out
in formal equilibrium models. However, the authors seem to suffer from the conceit that they
have laid a groundwork upon which armies of graduate students can now build their theses,
with perhaps minor alterations of the authors' structure. I am less optimistic. The Dixit-Stiglitz
structure has a number of features that simplify things for theoretical exercises but make things
difficult for empirical work. Consider the issue of firm size, for example. All firms are the same
size in the Dixit-Stiglitz model. If we want to be serious about taking the theory to the data,
it would be helpful to have a model that allows for firms to have different sizes. It is easy to
come up with alterations of the Dixit-Stiglitz structure that will allow for differences in firm
size in equilibrium (see, for example, Holmes 1999). The point is that these changes are a big
departure from their setup. The various little tricks the authors use won't be too helpful here.
The researchers going this direction will be pretty much starting from scratch in figuring out
how to solve their models.
This book is at its best when it analyzes static models. Its efforts at dynamics are less
convincing. First, agents are not forward-looking. Second, there are no state variables. That is,
the location choice an agent makes in one period has no bearing on his or her subsequent
location choice; there are no sunk costs. I recognize that it is very difficult to put true dynamics
into these models; these models are difficult to work with even as static models. The annoying
thing is that the authors spend a lot of time pontificating about the importance of "doing things
right." It seems obvious that doing dynamics right means having forward-looking behavior and
links between past decisions and future choice sets. The authors should stick to the analysis of
static models because they are doing the static analysis right
The material on urban economics is the newest material in the book and I close with a
few comments about this section. I am not impressed with their monocentric model of a city.
In this model, manufacturing workers and firms collapse onto a city that is a single point on a
line segment. Outside the city, the only industry is agriculture. The price of agricultural land
falls as we move away from the city as a compensating differential since farmers further from
the city have to pay higher transportation costs to get manufactured goods. This is not so much
a monocentric model of a city but rather a monocentric model of a region. This theory has little
empirical relevance for regions (I doubt the price of agricultural land 50 miles from a city is
much different than the price 100 miles from a city) and this theory gives us no help in
 Book Reviews 493

understanding the organization of cities. I do like their central place theory. They succeed in
getting a hierarchy of cities to emerge where lower order cities produce a limited range of
products and higher order cities produce a wide range of products. It is the only equilibrium
model of central place theory that I have seen. However, the analysis is clunky; essentially an
analysis of an example on a computer. It may be that further progress in this area will be
obtained by considering alternative new structures rather than pushing this particular structure
further on the computer.

References
Fujita, Masahisa and T. Mon. 1997. Structural stability and evolution of urban systems. Regional Science and Urban
Economics 27:399-442.
Holmes, Thomas J. 1999. Scale of local production and city size. American Economic Review Papers and Proceedings
89:317-20.
Krugman, Paul R. 1991. Increasing returns and economic geography. Journal of Political Economy 99:483-99.
Krugman, Paul R. and Anthony J. Venables. 1995. Globalization and the inequality of nations. Quarterly Journal of
Economics 110(4):457-880.

Thomas J, Holmes
University of Minnesota