Beyond urban form: How Masahisa Fujita shapes us

• Published date: 01 March 2017
• Date: 01 March 2017
• Author: Tomoya Mori,Marcus Berliant
• DOI: http://doi.org/10.1111/ijet.12115

doi: 10.1111/ijet.12115

Beyond urban form: How Masahisa Fujita shapes us

Marcus Berliant∗and Tomoya Mori†,‡

Literature from several phases of the career of Masahisa Fujita is surveyed chronologically, with

a view toward future contributions in these areas. First we address the economic structure of

the interior of a city with mobile consumers, adding production. Next we provide a critical dis-

cussion of the new economic geography,in par ticular distinguishing between recentapproaches

employing two regions and more than two regions, both in theory and in application to data. Fi-

nally, wediscuss knowledge creation in groups and briefly touch on his current work in artificial

intelligence.

Key wor ds urban economics, new economic geography,knowledge creation, knowledge diver-

sity, robot economist

JEL classification D83, O31, R12, R13, R14

Accepted 2 May2016

1 Introduction

Over the years, much has been written about both the work of Masahisa Fujita, henceforth Masa, and

its impact. Here, we wish to take the time and space not just to thank him, whichis r ather superficial,

but to give an integrative and critical view of the past, present and future of some Masa-relevant

literatures from our rather special vantage points, those of a long time collaborator and a student.

We do not wish to put words into his mouth, as it were, nor to imply that our vantage points are

more important than those of others. Rather, we think that our view of intellectual progress over

the course of Masa’s career is different from that of the many others who have contributed to urban

economics, economic geography,and related disciplines. Naturally, a researcher’smilieu and context

have much to do with how a person’s research proceeds. Clearly Masa’s give and take with others

working in parallel or at cross-purposes with him, in other words interactions in person or in print,

or coauthor and mentor relationships, haveproduced the research path we have experienced and will

experience. There is a great deal of path dependence and path co-dependence in the development of

new ideas.

Naturally,many of Masa’sinnovations come from tensions in the literature. Theory, stylized facts,

and data all play a role. Tensions can arise either within or between any of these. But what is unique

∗Department of Economics, WashingtonUniversity, St. Louis, MO, USA. Email: berliant@wustl.edu

†Institute of Economic Research,Kyoto University, Yoshida-Honmachi, Sakyo-ku, Kyoto,Japan.

‡Research Institute of Economy,Trade and Industry, Annex, Ministry of Economy,Trade and Industry (METI), Kasumi-

gaseki, Chiyoda-ku, Tokyo, Japan.

We thank the editors,Kazuo Nishimura and Makoto Yano, and an anonymous refereefor helpful comments, but retain

responsibility for the content of this paper.

International Journal of Economic Theory 13 (2017) 5–28 © IAET 5

International Journal of Economic Theory

Beyond urban form Marcus Berliant and TomoyaMori

about Masa is his systematic, exhaustive organization and categorization of the literature prior to

addressing any of the tensions. This may involve working out versions of models and results in his

notebooks, possibly results never published, in order to make surethat all the log ical possibilities are

known and available to him. Or it may involve cataloging the previous results in a literature. This

can be seen, for example, in the book Urban Economic Theory. It makes the referee process easy in

many cases, as that process simply involvesgiv ing page numbersin the book.

We shall returnto this big point in the Epilogue.

Our survey of Masa’swork is neither comprehensive nor random. Weselect work, from each phase

of his career thus far, which provides a way forward. It is generally not useful to survey deceased or

zombie literatures with no evident future.

The remainder of this article is organized as follows. Section 2 discusses material related to the

early part of Masa’scareer, namely classical urban economics, up to around 1990. Section 3 discusses

material related to the new economic geography (NEG), from around 1990 until around 2003.

Section 4 discusses material from around 2003 until around 2011. It is about knowledgecreation and

transfer in groups. Section 5 provides a brief discussion of work from around 2011 into the future,

related to artificial intelligence.

2 Episode 1: Inside a city

The literature from the early part of Masa’s career, summarized in his encyclopedic work Urban

Economic Theory, is to a good degree settled.1So this section will take the form of a pictorial tribute,

presenting some new material and ideas informally. It may tax your intuition some, but will not be

technical. Your homework is to write a paper or two.

Tobegin, we detail Alonso’s (1964) famous model of a city, as further developed by Berliant and

Fujita (1992) and as described pictorially in Berliant and LaFountain (2006). It is the analog of an

Edgeworth box exchange economy in the urban context. We shall cover this older literature briefly.

The city is linear, so there is one unit of land available at each distance from the central business

district (CBD), where the latter is located at 0, so the supply of land is the interval [0,l), where lis

the exogenous extent of the city. The totalamount of composite consumption commodity available

in the economy is C>0. There are two consumers, called Aand B. The use of two consumers

is essential for the diagrams employed, much as they are in an Edgeworth box economy, but the

definitions and results for our model extend easily to an arbitrary but finite number of consumers.

When dealing with positive issues such as equilibrium, it is necessary to add individual endowments

and an absentee landlord who is endowed with all of the land and likes only composite commodity,

but we shall focus on normative issues here.

Each consumer will be allocated some composite good, cA,cB≥0, and an interval of land,

[xA,x

A+sA), [xB,x

B+sB), where xAand xBare the driveway locations of consumers Aand B,

namely the closest point in their respective parcels to the CBD, whereas sAand sBare the interval

lengths for the respective consumers, namely the sizes of their parcels. Marginalcommuting cost per

unit of distance is t>0, so the total commuting cost in terms of consumption good for consumers

Aand Bis t·xAand t·xB.

1One areaof ongoing dispute, which we shall suppress here due to constraints on time, space, and energy, are the differences

between models with a continuum of agents and those with a finite number of agents.

6International Journal of Economic Theory 13 (2017) 5–28 © IAET