Econ 137

Urban Economics
Lecture Notes I

Guillermo Ordonez, UCLA

Course Information

Instructor: Guillermo Ordonez ([email protected])

Lecture hours: MW 1:00
- 3:05 pm, Dodd 175.
Office hours: MW 3:30
- 4:30 pm, Bunche 2265.
By appointment.
Midterm: August 27th, 2007 (Monday of 4th week).
Final Exam: September 12th, 2007 (Last class).
Course Webpage:
http://www.sscnet.ucla.edu/071/econ137
- 1/

Econ 137 - Summer 2007 2

Grading

„ Midterm (1 hour) 25% (Replaceable grade!)

„ Individual Essay 15%
„ Group Essay 10%
„ Final (2 hours) 50%

Econ 137 - Summer 2007 3

 Readings
ƒ [O] O’Sullivan, Urban Economics, Sixth Edition, McGraw-Hill 2006
E-reserve: College Library Circulation Desk
http://www2.library.ucla.edu/service/student.cfm

ƒ [LN] Lecture Notes:

http://www.sscnet.ucla.edu/071/econ137-1/linkpage.php
Print free from the printing kiosks (Public Policy)
Print them right before class!

ƒ [AR] Additional readings

ƒ News articles and papers given in class.
ƒ [H&G] Hoover and Giarratani, Regional Economics, Online Edition
http://www.rri.wvu.edu/WebBook/Giarratani/contents.htm
Econ 137 - Summer 2007 4
Plan for the Course
Week 1 Introduction [O] Ch. 1, 2, 3
Cities and Clusters [H&G] Ch. 1, 2
Week 2 Cities over space (Size, Shape) [O] Ch. 4, 5
Cities over time (Growth) [H&G] Ch. 7,8
Week 3 Land use pattern [O] Ch. 6,7
[H&G] Ch. 6
Week 4 08/27 - Midterm
Urban Transportation, Housing [O] Ch. 10, 11, 13, 14
Week 5 09/03 - Labor Day (No Class)
Education, Crime. [O] Ch. 12
Week 6 Local Government Economics [O] Ch. 15, 16
09/10 - Essays are due
09/12 - Final Exam
Econ 137 - Summer 2007 5
What is this course about?
CITIES

Why do cities exist? We’ll explain it by studying

What determines the location, utility-maximizing households
size, shape and growth of cities? and profit-maximizing firms

Once we understand this, we can see inside a city !

Land use, industry clusters and population distribution

Urban Problems: Crime, Education, Transportation, Housing

- Cities are important because they facilitate growth

- Cities exist because benefits (innovation, production and
trade) exceed costs (cities are noisy, dirty and crowded)
Econ 137 - Summer 2007 6
What is this course about?
CITIES

Who cares?

Businessmen: Where to locate firms? How big the market will be?

Real estate investors: Patterns of land use and population distribution

Policy makers: Policies that combat urban problems will increase the vitality of
cities, helping them to grow.

and hopefully YOU!

Econ 137 - Summer 2007 7

Questions for Lecture Notes I

„ What is a City?
„ Why do cities exist?
„ Why do cities became so important in the last century?
„ Why do we observe industry clusters?
„ What determines city size?
„ Why are there different types of cities?

Econ 137 - Summer 2007 8

Tools to answer our questions
Economic Models
„ Mathematics as a tool
… Exogenous vs. Endogenous variables
… Parameters

„ Equilibrium
… Partial Equilibrium
… General Equilibrium Have you ever played SimCity?

For review:
… Nicholson, Microeconomic Theory, Chapter 2
… O’Sullivan, “Urban Economics”, Appendix.

Econ 137 - Summer 2007 9

Tools to answer our questions
Golden rules of optimization
„ Marginal Benefit = Marginal Cost

„ Average Benefit >= Average Cost

Econ 137 - Summer 2007 10

Tools to answer our questions
Average vs. Marginal Effects
g ( x) > 0
g ( x)
A= >0 (Average contribution of all units)
x
∂g ( x )
M= (Contribution of the last unit)
∂x
∂A ∂g ( x ) / ∂x g ( x ) 1
= − 2 = [ M − A] >=< 0
∂x x x x
„ If Average decreases, M < A
„ If Average is constant, M =A
„ If Average increases, M > A
Econ 137 - Summer 2007 11
Tools to answer our questions
Economies of scale in production

From Production functions

to cost properties
„ DRS = AC < MC
„ IRS = AC > MC
„ CRS = AC = MC
Econ 137 - Summer 2007 12
 What is a city? Census Bureau definitions
„ Census block: Smallest geographical unit. Area bounded by
streets, property lines, political boundaries, etc.

„ Urban and Rural areas:

… Urbanized Area (UA): Densely settled core of blocks and
surrounding blocks such that pop. density > 1000 people per
mile2 & N>50,000
… Urbanized Cluster (UC): Small version of urbanized area. Pop.
density > 500 people per mile2 & 2,500<N<50,000
… Rural area (RA): Not UAs or UCs

Econ 137 - Summer 2007 13

 What is a city? Census Bureau definitions
„ Metro(micro)politan areas (MSA).
… Metropolitan area (MeA): Includes at least one urbanized area
(UA) such that N>50,000
… Micropolitan area (MiA): Includes at least one urbanized
cluster (UC) such that 10,000>N>50,000

„ Principal City
… Principal city is the largest place (municipality) in each MSA.
… Possibly many principal cities in a given MSA.
… Minimum requirements
„ Population size: At least 250,000 people.
„ Employment size: At least 100,000 workers.

Econ 137 - Summer 2007 14

Examples
of MSAs

Econ 137 - Summer 2007 15

 Principal cities in Los Angeles--Long Beach--Santa Ana, CA UA

Alhambra city, CA Fullerton city, CA Orange city, CA

Anaheim city, CA Gardena city, CA Paramount city, CA
Arcadia city, CA Garden Grove city, CA Pasadena city, CA
Baldwin Park city, CA Glendale city, CA Pico Rivera city, CA
Bellflower city, CA Hacienda Heights CDP, CA Pomona city, CA
Buena Park city, CA Hawthorne city, CA Rancho Cucamonga city, CA
Burbank city, CA Huntington Beach city, CA Redondo Beach city, CA
Carson city, CA Huntington Park city, CA Rosemead city, CA
Cerritos city, CA Inglewood city, CA Santa Ana city, CA
Chino city, CA Irvine city, CA Santa Monica city, CA
Chino Hills city, CA La Habra city, CA South Gate city, CA
Compton city, CA Lakewood city, CA South Whittier CDP, CA
Costa Mesa city, CA Long Beach city, CA Torrance city, CA
Diamond Bar city, CA Los Angeles city, CA Tustin city, CA
Downey city, CA Lynwood city, CA Upland city, CA
East Los Angeles CDP, CA Montebello city, CA West Covina city, CA
El Monte city, CA Monterey Park city, CA Westminster city, CA
Florence-Graham CDP, CA Newport Beach city, CA Whittier city, CA
Fontana city, CA Norwalk city, CA Yorba Linda city, CA
Fountain Valley city, CA Ontario city, CA
Econ 137 - Summer 2007 16
What is a city?
„ Economic City: Area with a relatively high population
density that contains a set of closely related and
economically integrated activities (flow of workers,
money, etc...). Using Census Bureau definitions this
roughly refers to UA, MeA or City.

„ Political City: Area over which a municipal corporation

exercises political authority, providing local government
services and collecting taxes. Using CB definitions
central city or municipality where a municipal corporation
exercises political authority and provides local services.
Econ 137 - Summer 2007 17
Census Bureau: More definitions

Geographic entities
„ Administrative or Legal:
State, County, Consolidated City, School District, Voting
District, ZIP codes, etc…

„ Statistical:
MSA, UA, UC, Census Tract, Blocks, etc..

Econ 137 - Summer 2007 18

Econ 137 - Summer 2007 19
Econ 137 - Summer 2007 20
Urban Population Growth

Econ 137 - Summer 2007 21

Urban Population Distribution

Econ 137 - Summer 2007 22

Why do cities exist?
Three facts of life (real imperfections) that make cities
desirable

ƒ Imperfect factor mobility (impeding the elimination of

natural resources advantages).
ƒ Imperfect input divisibility (generating scale and
agglomeration economies).
ƒ Imperfect mobility of good and services (it’s better to be
close to producers)

Econ 137 - Summer 2007 23

Why do cities exist?
Three basic conditions
(that have been getting more important over time)

ƒ Agricultural surplus (to feed cities)

ƒ Urban production (to buy food)
ƒ Transportation (to exchange food and urban production)

Econ 137 - Summer 2007 24

Why do cities exist?
First, let’s think the conditions under which cities won’t
arise (backyard production).

A region WILL NOT HAVE cities when:

1) Equal productivity across areas.
2) No scale economies in exchange.
3) No scale economies in production.
4) Equal preferences across areas.

Econ 137 - Summer 2007 25

Why do cities exist?
„ We will focus on two types of cities and the conditions
under which they arise.
… Trading cities: Cities specialized in trade services.
… Factory cities: Cities specialized in good or services
production.

Econ 137 - Summer 2007 26

Trading Cities

Relax 1): Comparative advantages in production

+
Time and transport cost not too big
_____________________________
Specialization and trade

Is that enough?

NO!!!, TRADE does NOT mean a TRADING CITY

Econ 137 - Summer 2007 27

Example: Absolute vs. comparative advantages
2 goods (Shirts and Bread)

Comparative Advantage
Output per hour Opportunity Cost
North South North South
Bread 1 2 1 shirt 3 shirts
Shirts 1 6 1 loaf 1/3 loaf

South region has absolute advantage in both goods, but

comparative advantage in _____
Shirts

Econ 137 - Summer 2007 28

Example: Autarky situation
Endowment: 24 hours

AUTARKY
Production = Consumption

Prices of bread in autarky

- 1 shirt in the North
- 3 shirts in the South

Econ 137 - Summer 2007 29

Example: Trading costs and Specialization
„ Assume Price is P= 2 shirts per loaf:
Specialization pays +1 loaf of bread to South and +1
shirt to North!!!

„ Is this enough to have trade?

NO
It depends on the transport and trading cost
t = 1 hour: NO specialization or trade
t = ½ hour: Specialization and trade

„ Why?
Econ 137 - Summer 2007 30
Example: Trading costs and Specialization
„ Assume P = 2 shirts per loaf and t= ½ hour per trade
„ South
… If trade: +6s- 2b=+6s- 4s=2s gain = 1 bread gain
… If transport cost: - 1/2 hour = 1 bread lose Transform hours
„ North into the
… If trade: +1b- 1s=+2s- 1s= 1 shirt gain good that you’d
produce if no trade
… If transport cost: - 1/2 hour = 1/2 shirt lose (the one you resign)

South is indifferent and North gains from trade

TRADE

„ Is this enough to have a Trading city? NO

Econ 137 - Summer 2007 31
Example: Gains from Trade

South specializes in producing shirts

North specializes in producing bread

This graph assumes TC=0.

How it changes with TC=1/2 hour

Econ 137 - Summer 2007 32

 Trading Cities
„ Relax 2): T’(q) > 0- T’’(q) < 0

„ Specialization in Trade:
… Middlemen (trading firms)
… Trading cities along rivers, railroads, ports, etc.

Comparative Advantages (Relax 1)

+ savings in production cost > transport cost = TRADE

Econ 137 - Summer 2007 33

 Trading Cities
„ Relax 3): T’(q) > 0- T’’(q) < 0

„ Specialization in Trade:
… Middlemen (trading firms)
… Trading cities along rivers, railroads, ports, etc.

Comparative Advantages (Relax 1)

+ savings in production cost > transport cost = TRADE
+ Economies of scale in Transport (Relax 2)= TRADING CITY

AND economies of scale should be enough to compensate for

higher cost of living faced by the workers (commuting, house rent)
Econ 137 - Summer 2007 34
Factory Cities
Recall….
A region WILL NOT HAVE cities when:
1) Equal productivity across areas.
2) No scale economies in production.
3) No scale economies in transportation.
4) Equal preferences across areas.

Econ 137 - Summer 2007 35

Factory Cities
Relax 3): Economies of scale in production of one good
+
Time and transport cost not too big
_____________________________
Specialization and trade

Is that enough?

Economies of scale should be enough to compensate for higher

cost of living faced by the workers (commuting, house rent)

Econ 137 - Summer 2007 36

Factory cities: Market Area of a Shirt Factory
Assumptions:
„ Home production: 1 shirt or 1 loaf of bread per hour.
„ Factory production
… A worker is 8 times more productive producing shirts in a
factory than at home (8 shirts/hour).
… Cost of indivisible input (e.g. capital) = ½ loaf per hour
„ Workers extra costs
… Cost of urban living (e.g. higher land price) = ½ loaf per hour
„ Consumers extra costs
… Linear travel cost [walking, 8 miles per hour].
„ How big is the market area?
Econ 137 - Summer 2007 37
Factory cities: Market Area of a Shirt Factory
„ Wages = 3/2 loaves/hour (otherwise they will work in rural
area)
… Cost from working at the factory and not at home = 1 loaf/hour
… + Cost of urban living = ½ loaf/hour
„ Hourly Cost of producing shirts = 2 loaves/hour.
… Cost of workers (wages) = 3/2 loaves/hour
… Cost of indivisible input (e.g. capital) = ½ loaf/hour
„ Cost per shirt = 1/4 loaf (since the factory produces 8 shirts
per hour and the hourly cost is 2 loaves).
„ In equilibrium the Price per shirt is the cost per shirt.

Econ 137 - Summer 2007 38

Factory cities: Market Area of a Shirt Factory
Cost of factory cloth Cost of factory cloth
1.00 Cost of homemade cloth
Net cost of cloth (hours)

0.75

0.50 Travel cost Slope=0.25=2*(1/8)

0.25
Factory cost
Miles from Factory
4 3 2 1 0 1 2 3 4

Factory

Econ 137 - Summer 2007 39

Factory cities:
How we obtained the market area of 3 miles?
„ Indifferent consumer lives x miles from factory. Recall
the worker losses 1/8 hours walking 1 mile, hence t=1/8

cost of shirt at home = total cost from factory

1 = 0.25 + 2tx
1
0.75 = 2 x
8
x=3
The limit to economies of scale in production is given by
the market area (and hence by transport costs !)
Econ 137 - Summer 2007 40
 Factory cities: Size Limits
Workers will live close to the factory to economize commuting
costs
Effects of More Workers
„ Price decreases (widening market area)
… More efficiency and lower AC.
„ Price Increases (shrinking market area)
… Workers live farther away, commuting costs increase, higher
salaries needed and increases in hourly costs of production.
… Higher density in cities, higher urban costs, higher salaries
needed and increases in hourly costs of production.
Econ 137 - Summer 2007 41
Factory cities: Size Limits
City size is limited by transport costs because
„ Transport costs limit market area (limiting production and the
possibilities of economies of scale)
„ Commuting costs increase costs of production, compensating
the effects of economies of scale.

Econ 137 - Summer 2007 42

 Factory Cities (with a single factory)

„ No specialization
… shirts are produced in factories,
… bread at home [surplus sold to workers, no need to
set up bread factory]

Factory Factory Factory

Bread producers,
(buy shirts)

Econ 137 - Summer 2007 43

Factory cities: Comparative Statics
Can this model account for the steady growth in urban
population along time?

„ Innovations in manufacturing
… Productivity shock that shift average labor costs. (e.g sewing machines,
interchangeable parts for manufacturing, etc).

„ Innovations in transportation
… Turnpikes, railroads, steamships, etc.

„ Innovations in agriculture.
… Steel plow, agricultural science, fertilizers, etc.
Econ 137 - Summer 2007 44
 Market Area: Innovation in manufacturing
Cost of factory cloth Cost of factory cloth
1.00 Cost of homemade cloth
Net cost of cloth (bread)
0.75

0.50

0.25
0.10
Miles from Factory
4 3 2 1 0 1 2 3 4
Worker
productivity
increases
from 8 3.6miles
shirts/hour
to 20
shirts/hour

Econ 137 - Summer 2007 45

 Market Area: Innovation in transportation
Cost of factory cloth Cost of factory cloth
1.00 Cost of homemade cloth
Net cost of cloth (bread)
0.75

0.50

0.25

Miles from Factory

4 3 2 1 0 1 2 3 4

4.5miles

Transportation
costs reduce from
8 miles/hour to 12
miles/hour
Econ 137 - Summer 2007 46
 Market Area: Innovation in agriculture
1.25 Cost of factory cloth Cost of factory cloth
Net cost of cloth (bread) 1.00 Cost of homemade cloth

0.75

0.50

0.25

Miles from Factory

4 3 2 1 0 1 2 3 4

4 miles

Shock in
agriculture from 1
bread/hour to 1.25
bread/hour
Econ 137 - Summer 2007 47
Factory cities: Material oriented firms
The same logic applies in cases in which the final output
can be easily transported but the raw materials have a
high transportation cost.
In these cases firms will locate around raw materials plants
(instead of locating around consumers).
Workers of firms then will locate in that area to save on
commuting costs, generating factory cities.

Econ 137 - Summer 2007 48

Cities: Historic Perspective
„ First Cities:
… Defensive city (ES in storage of surplus)
… Religious city (ES in worship)
„ Greek and Roman Cities (trading cities- >
Parasitopolis)
„ Feudal Cities
„ Mercantile Cities
„ Factory Cities

Econ 137 - Summer 2007 49

Cities: Industrial Revolution
Innovations:
… Agricultural (K/L)
… Manufacturing (SE and specialization)
… Transportation (steamship and railroad)
… Construction

… The location pattern started to depend also on energy

sources.

Econ 137 - Summer 2007 50

Summary Ch. 2 O’Sullivan
„ Employment opportunities are concentrated in cities because some
activities are subject to economies of scale.
„ A trading city will develop if a) the differences in productivity
underlying comparative advantage are large relative to transportation
costs and b) there are scale economies in transportation.
„ A factory city will develop if scale economies in production are large
relative to transporting goods (freight costs) and the cost of
transporting workers within the city (commuting costs)
„ The first cities developed for defensive and religious reasons.
„ The centralization of power during the mercantile period caused the
development of administrative cities and reduced trade barriers.
Combined with efficient ocean travel, this resulted in the development
of trading cities.
„ The rapid urbanization in the nineteenth and twentieth centuries was
caused by the industrial revolution and the associated innovations in
agriculture, transportation, and manufacturing
Econ 137 - Summer 2007 51
BUT….
Scale economies in production and transport help to
explain trading or factory cities where just one activity
exists (more if we consider more goods, but just one
producer of that good at a city).

Why do we observe big cities with a diverse collection of

economic activities?
Why similar industries locate close together instead of
having their own market area?
Answer: AGGLOMERATION ECONOMIES
Econ 137 - Summer 2007 52
Some industrial clusters in US

Econ 137 - Summer 2007 53

Agglomeration Economies

External Economies of Scale

„ (LE) Localization Economies [industry level]

„ (UE) Urban Economies [metropolitan level]

Econ 137 - Summer 2007 54

Localization Economies (LE)
„ Sharing Input Suppliers
… Minimum demand level to exploit Economies of Scale in
the production of intermediate input
… Reduce Transport costs

„ Sharing a Labor Pool

… Varying Demand [job search + switching costs]
… Matching [variety]

„ Knowledge Spillover
Econ 137 - Summer 2007 55
LE – Sharing Input Suppliers
A cluster of several firms:
„ Generate sufficient demand to allow intermediate input
producers to exploit scale economies.
„ Reduce face2face time and transport costs.
„ Allow intermediate suppliers to produce a wide variety.

In fact, you can think in a cluster as a market

area for intermediate inputs factories.

Econ 137 - Summer 2007 56

LE - Labor Pooling [Switching Costs]
„ Shocks on product demand (50% probability of good
times or high demand)
„ Firm’s labor demand in good (high) and bad (low) times:
1 1
w = 60 − L
h
w = 40 − L
l

4 4
„ Fixed labor supply (120 in isolated site (the only firm)
and 240 in clustered site (with another firm)).
„ Assume shock correlation =- 1when clustered. Total
Demand? Lh + Ll = LCluster = 400 − 8w
400 − L Total demand
w= and wages fixed
8
Econ 137 - Summer 2007 57
 LE - Labor Pooling [Switching Costs]
Isolated Place Clustered Site Corr=- 1

W, $ Wage, $
60 60
TD
40 40

30

20
10

DL DH DL DH

120 Labor units 80 120 160 240 L

Econ 137 - Summer 2007 58
 LE - Labor Pooling [Switching Costs]
1 1
„ Relative Benefits of Clustered Site = (1400) + (−1000) = 200
2 2
Gains in good times (lower salary)
Wage, $ =120x$10+40x$10/2=$1400

TD

30
20
Loss in bad times (higher salary)
=80x$10+40x$10/2=$1000
10

DH DL DH

80 120 160 Labor units

Econ 137 - Summer 2007 59

 LE - Labor Pooling [Switching Costs]
What if correlation is not -1
Probability Matrix
Firm I
----------- L H wHH = (480 − L) / 8
Firm II wLL = (320 − L) / 8
L 0.15 0.35 wLH = wHL = (400 − L) / 8
H 0.35 0.15 E [ w] = ?
Pr ( H ) = 0.5
Pr ( H / L ) = Pr ( L / H ) = 0.7

E [ w ] = 0.15 ⋅ w H H + 0.15 ⋅ w LL + 0.35 ⋅ w LH + 0.35 ⋅ w H L

= 50 − 0.125 L

Econ 137 - Summer 2007 60

 LE - Labor Pooling [Switching Costs]
Clustered Site [-1< corr <1]
Probability Matrix Wage, $
Firm I
DHL=DLH
----------- L H
Firm II

L 0.15 0.35 30
E[w]=$20
H 0.35 0.15
10
wHH = (480 − L) / 8
wLL = (320 − L) / 8 DLL
E[D] DHH

wLH = wHL = (400 − L) / 8 240 Labor units

E [ w] = 50 − 0.125 L
Econ 137 - Summer 2007 61
 LE - Labor Pooling [Switching Costs]
Isolated Place Each firm in Clustered Site [firm demand]
W, $
W, $

HH
30 30
20 HL
LH
10
10
LL

DL DH DL DH

120 Labor units 80 120 160 Labor units

Econ 137 - Summer 2007 62
LE - Labor Pooling [Switching Costs]
„ Expected gains in clustered site [- 1< corr < 1]
GHH = 0
GLL = 0
GHL = (30 − 20) *120 + (30 − 20) * 40 / 2 = $1400
GLH = −(20 − 10) *80 − (20 − 10) * 40 / 2 = −$1000

EG = 0.15GHH + 0.35GHL + 0.35GLH + 0.15GLL = 0.35* 400 = $140

What if correlation=1?

Econ 137 - Summer 2007 63

LE - Labor Pooling [Matching]

Econ 137 - Summer 2007 64

 LE - Labor Pooling [Matching]
„ Assume optimal size of the firm is 2 workers, net profits per
worker are $12 and unit training costs are $24
„ Four workers
… Profits per worker= 12-24/8=9 (wages under competition)
„ Six workers
… Profits per worker= 12-24/12=10 (wages under competition)
„ What happens as the number of workers goes to infinity?

„ MUTUAL ATTRACTION BETWEEN FIRMS AND WORKERS

Econ 137 - Summer 2007 65

 Localization Economies - Examples
„ Sharing Input Suppliers
… Minimum demand level to exploit SE in production of intermediate
input
„ Headquarters demand Legal Services, Advertising, Economic advising, etc..
… Transport costs
„ Face to face contact with input suppliers

„ Sharing a Labor Pool

… Varying Demand [job search + switching costs]
„ TV industry in LA and NY
„ Hi Tech industries
… Matching [variety]

„ Knowledge Spillover
„ Innovation is correlated with industry clusters [Dumais, Ellison and Glaeser
(2001)]. Econ 137 - Summer 2007 66
Localization Economies – Empirical tests
„ Worker Productivity:
… Elasticity of Output per worker with respect to Industry Output:
∂ ( qI / LI ) QI
ε qI / LI ,QI = Office sector=0.27
∂QI qI / LI
„ Plant Births:
… Location choice by new firms:
∂BI QI Electronics=0.43
ε B ,Q =
I I
∂QI BI
„ Employment Growth: [Rosenthal and Strange (2000)
Add 1000 Software Jobs in Jobs in New Establishments
in Zip-Code Area Increase by
1-mile radius around zip code area 11.7 jobs
1-5 miles from zip-code area 0.8
6-10 miles from zip-code area 0.8

Econ 137 - Summer 2007 67

Urbanization Economies
„ Intra
- n
i dustry share of inputs.
„ Intermediate inputs: banking, insurance, hotels, public
services, etc.
„ Labor Pooling [negative correlation makes more sense
now]
„ Spillovers of information [product design and methods]
„ Joint labor supply, learning and social opportunities

Econ 137 - Summer 2007 68

Urbanization Economies - Evidence
„ In the last decades shift in specialization of cities. Large
ones are specialized in managerial activities while small
cities specialize in production (laboratory cities).
„ A doubling in population increases output per worker by
between 3 and 8 per cent.
„ Diversity promotes employment and firm growth

Econ 137 - Summer 2007 69

 Typical cases of industrial clustering
Silicon Valley and Route 128
„ Network approach to external economies. Saxenian (1996)
„ Silicon Valley
… Area of Northern California (30 mile by 10 mile strip of land in
Santa Clara county, between San Francisco and San Jose)
„ Route 128
… 65 mile highway surrounding Boston and Cambridge
„ Both of them are technological concentrations that have been
devoted to the creation of new information technology and
electronics.
„ Silicon Valley has been more successful than Route 128
Econ 137 - Summer 2007 70
Typical cases of industrial clustering
Silicon Valley and Route 128
Number of companies
40 39

35 33

30
26
25 25
25
22

20
15
14 14
15
10
10
7

5 4

0
1985 1986 1987 1988 1989 1990
Silicon Valley Route 128
Econ 137 - Summer 2007 71
 Typical cases of industrial clustering
Silicon Valley and Route 128
„ Silicon Valley
… Regional advantages
„ World class academic institutions (Stanford and Berkeley)
„ Defense agencies in California (Lockheed in 1956)
„ Nice weather !!!
… Agglomeration economies
„ Professional and social networks
„ High turnover (35%). Average job tenure (2 years)
„ Good jobs search and matching
„ Flat hierarchies (entrepreneurship and experimentation)
„ Collective learning (encourages innovation)
Econ 137 - Summer 2007 72
 Typical cases of industrial clustering
Silicon Valley and Route 128
„ Route128
… Regional advantages
„ World class academic institutions (MIT)
„ Federal government. Department of Defense has accounted for
over 60% of federal R&D in the state.
… Agglomeration economies (compared to Silicon Valley)
„ More government contracts and secrecy (less entrepreneurship)
„ Less social networks
„ Less job turnover and more average tenure.
„ Vertical hierarchies

This may explain why Silicon Valley has been more successful
Econ 137 - Summer 2007 73
Summary Ch. 3 O’Sullivan
„ Location Economies (lower production costs as industry wide output
increases) occur because firms in an industry cluster benefit from
sharing (1) the suppliers of intermediate inputs, (2) a labor pool, and
(3) information

„ Urbanization economies occur if the production cost of a particular

firm decreases as the total output of the urban area increases

„ Agglomeration economies cause self-reinforcing changes in

location: the movement of one firm to a city increases the incentives
for other firms to move to the city.

„ Networks may be very important external economies.

Econ 137 - Summer 2007 74

Questions for Lecture Notes I

„ What is a City?
„ Why do cities exist?
„ Why do cities become so important in the last century?
„ Why do we observe industry clusters?
„ What determines city size?
„ Why are there different types of cities?

Econ 137 - Summer 2007 75

Practice Exercises - Lecture Notes I

„ All exercises in O’Sullivan - Chapters 2 and 3

Econ 137 - Summer 2007 76