assignment

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

6

I . I N T R O D U C T I O N

A s the global economy has become more integrated and urbanized,

fueled in large part by technology, major cities and metropolitan

areas have become key engines of economic growth. The 123 largest

metro areas in the world generate nearly one third of global output

with only 13 percent of the world’s population.

In this urban-centered world, the classic notion of a

global city has been upended. This report introduces

a redefined map of global cities, drawing on a new

typology that demonstrates how metro areas vary in

the ways they attract and amass economic drivers

and contribute to global economic growth in distinct

ways. New concerns about economic stagnation—in

both developing and developed economies—add

urgency to mapping the role of the world’s cities and

the extent to which they are well-positioned to deliver

the next round of global growth.1

Instead of a ranking or indexed score, which many

prior cities indices and reports have capably deliv-

ered,2 this analysis differentiates the assets and

challenges faced by seven types of global cities.

This perspective reveals that all major cities are

indeed global; they participate as critical nodes in

an integrated marketplace and are shaped by global

currents. But cities also operate from much differ-

ent starting points and experience diverse economic

trajectories. Concerns about global growth, productiv-

ity, and wages are not monolithic, and so this typology

can inform the variety of paths cities take to address

these challenges. For metro leaders, this typology

can also ensure better application of peer com-

parisons, enable the identification of more relevant

global innovations to local challenges, and reinforce a

city-region’s relative role and performance to inform

economic strategies that ensure ongoing prosperity.

This report proceeds in four parts. In the following

section, Part II, we explore the three global forces of

urbanization, globalization, and technological change,

and how together they are demanding that city-

regions focus on five core factors—traded clusters,

innovation, talent, infrastructure connectivity, and

governance—to bolster their economic competitive-

ness. Building on these factors, Part III outlines the

data and methods deployed to create the metropoli-

tan typology. Part IV explores the collective economic

clout of the metro areas in our sample and introduces

the new typology of global cities. Finally, Part V

explores the future investments, policies, and strate-

gies required for each grouping of metro areas. Within

the typology framework, we explore the priorities for

action going forward, including the implications for

governance.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

7

U R B A N I Z AT I O N

The world is becoming more urban, placing cities at

the center of global economic development. The share

of global population in metropolitan areas has grown

from 29 percent in 1950 to well over half today, and it

is predicted to reach 66 percent by mid-century.4

History indicates that urbanization both accompanies

and facilitates economic transition from agricul-

ture to manufacturing and services, activities that

tend to demand clusters of labor and capital as well

as the proximity to other firms that cities provide.

Urbanization and industrialization, therefore, tend

to occur in concert. These twin forces, which revolu-

tionized Europe and North America in the late 19th

century and early 20th century, have now touched

Asia and Latin America. However, this process is not

preordained. Africa’s urbanization, for instance, has

not been accompanied by widespread industrializa-

tion.5 Notwithstanding Africa’s challenges, millions

of rural residents each week flock to urban regions

in the Global South in search of the living standards

that new production and service jobs provide. Since

2010 annual urban populations have grown fastest in

Africa (3.55 percent) and Asia (2.50 percent), greatly

exceeding the pace of urban growth in North America

(1.04 percent) and Europe (0.33 percent).6

The pressures and opportunities accompanying

urbanization will be felt most intensely and directly

in the Global South, but the knock-on effects will be

worldwide. Urbanization in developing economies has

resulted in a much greater number of urban areas

in which firms and workers can thrive. In techni-

cal terms, agglomeration externalities—the benefits

that accrue to firms, workers, and local economies

from clustering—now exist in many more parts of the

world.7 As a result, along with their growing human

footprint, metro areas are flexing even greater

economic muscle on the world stage. Overall, the 50

percent of the world’s population that lives in urban

I I . G L O B A L M E G A T R E N D S A N D C I T I E S

T hree significant forces—urbanization, global integration, and techno-

logical change—are reshaping the international economy.3 We focus on

these three forces because they are distinctly positioning cities as the

world’s competitive economic units while simultaneously redefining

what it takes for them to excel in today’s economy.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

8

areas produces roughly 80 percent of the world’s

total output.8

Urbanization, however, comes with risks if it is unman-

aged. Rapid population influxes in the megacities of

Africa, Latin America, and Southeast Asia are strain-

ing the ability of local governments to provide basic

housing, transportation, energy, water, and sewage

infrastructure.9 The world will need to invest $57

trillion in new infrastructure by 2030 to keep pace

with expected growth, the bulk of which will occur in

the developing world.10 If the negative externalities of

congestion, insecurity, and health risks overwhelm the

positive agglomeration externalities that cities provide,

countries run the risk of urbanizing without growth.11

The rise of developing metro areas creates both chal-

lenges and opportunities for developed world cities.

There is now more direct competition for firms and

talent, but metro areas in developed markets can also

look to developing metros with expanding populations

and wealth for new sources of demand. Brookings’

Homi Kharas and Geoffrey Gertz project that China

and India, which account for only 5 percent of global

middle-class consumption today, could together

account for nearly half of that consumption by 2050,

with most of it occurring in their cities.12

G L O B A L I Z AT I O N

Global integration, a defining trend of the postwar

era, is intensifying.13 The volume of goods, services,

and investments between countries increased from

$5 trillion in 1990 to $30 trillion in 2014, or from 24

percent to 39 percent of global gross domestic prod-

uct (GDP).14 Moreover, the nature of global exchange

seems to be shifting. While goods trade has stagnated

in recent years, cross-border flows of data and infor-

mation have grown robustly.15

Broadly measured, these connections matter.

Countries that are more internationally connected

can expect to increase GDP growth by up to 40

percent more than less-connected countries.16 These

findings affirm a wide array of economic literature

citing the benefits of participating in global flows of

trade, investment, and talent. Much of these benefits

stem from the presence of globally-engaged firms.

Local companies that embed themselves in global

value chains gain access to high-quality imports,

lowering their overall costs and allowing them to

become more globally competitive. This process

tends to boost productivity and wages.17 Firms selling

internationally inject new wealth from abroad that,

when spent locally, creates a multiplier effect in the

regional economy, spurring new jobs, growth, and fur-

ther tax revenue to be reinvested locally.18 Households

living in metro areas open to trade are able to access

a greater diversity of goods made elsewhere.19

Furthermore, global exchange is how regions with

fewer industrial capabilities often obtain the knowl-

edge required to move up the economic ladder, create

new jobs, and boost productivity.20

But cities also bear the brunt of the dislocations

caused by global integration. For instance, China’s

insertion into the global trading system resulted

in significant job losses in U.S. labor markets that

specialize in manufacturing.21 In the developing world,

there is an argument to be made that the globaliza-

tion of labor, trade, and capital markets, along with

bringing new knowledge and technologies, has con-

tributed to economic instability and rising inequities

within nations.22

Indeed, even those cities that have thrived in a more

globally integrated world are experiencing challenges

of unevenly shared prosperity. As Saskia Sassen has

argued, the rise of the globally integrated city has

coincided with the rise of the unequal city, across

both developed and developing countries.23 Indeed,

the Organization for Economic Cooperation and

Development (OECD) has found that inequality tends

to be higher and rising more quickly in large cities

than in their surrounding nations due to skills’ distri-

bution and the rise of high earners.24 Inequality may

limit upward mobility and overall economic growth if

it hinders investments in education and skills among

earners at the bottom of the income distribution.25

Recognizing these costs is an important and urgent

matter for public policy. But barring adoption of

severe isolationist policies, global integration will con-

tinue apace, and all cities must respond accordingly.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

9

T E C H N O L O G I C A L C H A N G E

The information technology revolution, digitization,

and labor-saving automation are altering modes of

communication, the processes firms use to create and

deliver products and services, and the very nature of

work itself.26

The scale of these technological changes is signifi-

cant and the pace of change has been relentless. The

McKinsey Global Institute predicts that 12 emerg-

ing technologies will generate an annual economic

impact of up to $33 trillion by 2025.27 A recent

Brookings study found that many of these technolo-

gies will be developed and deployed within a set of 50

“advanced” industries, characterized by a reliance on

high levels of research and development (R&D) and

significant numbers of science, technology, engineer-

ing, and mathematics (STEM) workers.28

Advanced industries matter because they drive pro-

ductivity growth in an environment in which overall

productivity growth has been lackluster.29 The aver-

age worker in advanced industries is twice as produc-

tive as the average worker outside the sector, due

to these firms’ unique abilities to productively utilize

new technologies and platforms. This productivity dif-

ferential matters because it allows workers within the

sector to earn wages double those of workers outside

of it.30 Cities that can foster environments in which

highly productive firms and workers can thrive enjoy

the associated wage benefits.

Risks accompany these high-tech breakthroughs,

however. In the United States, a useful proxy for other

advanced economies, already demonstrated technolo-

gies have the potential to automate 45 percent of

work activities in the United States.31 Indicative of the

deployment by advanced industries of labor-saving

technology, employment in advanced industries

in U.S. cities has been flat since 1980, even while

the sector’s value-added growth has soared. And

technology-induced labor market changes are not

a challenge just for the developed world. Increased

automation in manufacturing is one reason why

developing countries are deindustrializing at much

lower levels of income. This trend suggests that

manufacturing may not provide the same on-ramp

for lower-income countries going forward, and the

economic and political consequences of this shift may

be significant.32

Especially as populations age and workforces retire,

productivity growth, rather than labor force growth,

will have to do the heavy lifting to maintain overall

economic growth, especially in developed metro

areas. In a study of 20 large national economies, the

McKinsey Global Institute estimates that, to achieve

global growth rates comparable to those experienced

over the last 50 years, productivity growth will need

to be 80 percent faster to compensate for slowing

employment growth.33 Since technology appears to

be such a critical input to worker, firm, and industry-

level productivity, cities must understand and adapt

to its impact.

These three trends underscore a new economic real-

ity for cities. For starters, urbanization has placed

developing metro areas alongside their more devel-

oped peers as the main sites for economic growth and

development. This shift means that understanding

global market currents requires an understanding

of the economic dynamics playing out in the world’s

cities. The opportunities and pressures of global

integration mean that, to deliver prosperity for their

residents, cities must proactively adapt and position

workers, industries, and communities for the upsides

of global engagement by investing in a competitive

traded sector, maintaining infrastructure connec-

tivity, and being open to global flows of capital and

talent. To manage technological change and reap the

productivity gains that will improve living standards,

cities must cultivate innovation systems, skilled

workforces, and digital infrastructure. All of these

competitiveness assets must be stewarded by good

governance and a stable business environment.34

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

1 0

A framework for regional competitiveness

Infrastructure

Enablers

Governance

Trade

Innovation TalentInnovationInnovation TalentTalentTalent

Prosperity

Source: Brookings Institution, RW Ventures, and McKinsey and Company.

I I I . D A T A A N D M E T H O D S

D E F I N I N G A N D M E A S U R I N G C O M P E T I T I V E N E S S FA C T O R S

Given this global environment, this report focuses on

the assets that matter for a metro economy’s com-

petitiveness. We draw on the Harvard Business School

definition of a competitive market as one in which

firms can compete successfully in the global economy

while supporting high and rising living standards for

local households.35 Competitive regions are, by this

definition, supportive environments for both compa-

nies and people.

This report draws on a five-factor competitiveness

framework—tradable clusters, innovation, talent,

infrastructure, and governance. Globally competitive

traded sectors, innovation ecosystems, and skilled

labor are the key drivers of overall productivity,

employment creation, and income growth. “Enablers”

support these drivers: well-connected infrastructure

and reliable governance, public services, and the

business environment (see box).36 Focusing on these

fundamentals positions metropolitan economies

to compete based on the distinct long-term value

their industries and people can provide, and avoids

economic strategies that attract firms through “race-

to-the-bottom” techniques that compete via one-time

tax breaks or low wages.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

1 1

Measuring competitiveness factors

Tradable clusters: Tradable industries are a critical driver of prosperity and competitiveness. These

industries are typically anchored by globally engaged firms, which have valuable spillovers for local

economies. The traded sector can be measured in several ways. We measure tradable industries using

data on greenfield foreign direct investment (i.e., investments that bring new plants or offices), which

is inextricably bound up with traded industry clusters, and the productivity differential (measured as

output per worker) between a metro area’s traded sector and that traded sector nationwide.37 Due to data

limitations at the metropolitan scale, we are unable to standardize and measure domestic investments

across industries or include data on global trade flows.

Innovation: A region’s innovative capacity and levels of entrepreneurship both have implications

for its ability to develop and deploy commercial applications, start new businesses, and maintain

industrial competitiveness in the face of disruptive technological change.38 We measure innovation through

patenting, venture capital flows, and the scientific impact of research universities.39

Talent: Human capital—the stock of knowledge, skills, expertise, and capacities embedded in

the labor force—is of critical importance to enhancing productivity, raising incomes, and driving

economic growth. We measure talent through the share of population with tertiary education.40

Infrastructure connectivity: Infrastructure connectivity matters for regional competitiveness

because firms rely upon global access, both physically and digitally, to participate in the efficiencies

of global value chains. We measure infrastructure connectivity through aviation passenger flows and

internet download speeds.41 Due to data limitations we are unable to utilize standardized indicators on

other important infrastructure metrics such as the quality of freight and logistics systems, roads, and

public transit.

Governance: Governance matters for competitiveness because proactive government, public,

and civic groups can marshal investment from a variety of domestic and international sources to

enable new growth strategies. Similarly, the efficiency with which government can deliver services and

investments matters; highly fragmented metro areas tend to be less productive than their more cohesive

counterparts. Central, provincial, and municipal governments also have unique and complementary roles

to play in enabling firms and their wider regions to succeed in global markets.42 However, data limitations

limit our ability to quantitatively measure governance in this report.

S E L E C T I O N A N D D E F I N I T I O N O F M E T R O P O L I TA N A R E A S

We deploy new, standardized metropolitan-level data

to measure these factors for 123 large metro areas.

This sample constitutes the largest metropolitan

economies in the world in 2015 at purchasing power

parity (PPP) rates for which data on these factors

were available.43 With a few exceptions, these metro

areas all tend to have economies larger than $100

billion in nominal terms. The sample’s average popula-

tion is 7.6 million. As previous studies have shown,

including Brookings’ own Global MetroMonitor and

those by the McKinsey Global Institute and the World

Bank, global growth is not solely powered by these

large metro economies; in fact, small and mid-sized

cities matter greatly.44 Data limitations, however, pre-

vent us from analyzing a larger sample of economies

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

1 2

on all these factors. Given these limitations, we focus

on the largest city-regions because they uniquely

concentrate the assets that undergird global growth.

They are the main infrastructure connection points to

second- and third-tier cities. They cluster universities,

skilled workers, and other innovation assets that yield

the positive externalities and knowledge spillovers

that generate endogenous growth.45

This study uses the general definition of a metro-

politan area as an economic region comprising one

or more cities and their surrounding areas, all linked

by economic and commuting ties (see Appendix A).

These definitions are the same as those used in previ-

ous versions of Brookings’ Global MetroMonitor. We

use the terms city, city-region, metro, metro area, and

metro economy interchangeably to describe eco-

nomic regions.

M E T R O P O L I TA N T Y P O L O G Y

A significant body of research has sought to classify

global cities and measure their economic competitive-

ness. This literature began with the seminal work of

scholars like Peter Hall, John Friedmann and, most

famously, Saskia Sassen, each of whom documented

the unique role of a select handful of cities as the

command and control centers of global finance.46

That work has since been extended. Perhaps the most

commonly known classification of global cities comes

from the research group Globalization and World

Cities (GaWC), which has provided a rich theoretical

and analytical understanding of how cities engage in

the global economy through their unique concentra-

tions of advanced services firms.47 In their capacity

as analysts and investors, multilateral institutions

like OECD and the World Bank offer valuable, rigor-

ous assessments of growth and competitiveness in

global metro areas. Greg Clark and Tim Moonen have

found more than 200 indexes that have a global cities

focus.48

In a summary of global city rankings, the Chicago

Council on Global Affairs notes “how methodologies,

definitions, data use, and conclusions vary wildly

from ranking to ranking.” It also notes “biases and

challenges common to many indexes, including the

author’s perspective, lack of reliable and interna-

tionally comparable data, and the routine presence

of lagging indicators.”49 That report concludes that

city officials and policymakers seek out assessments

based on standardized data, look beyond topline rank-

ings, and uncover comparative strengths and weak-

nesses using relevant peers as a baseline comparison.

Against the backdrop of these previous efforts, we

develop a metropolitan typology based on regional

economic characteristics and competitiveness factors.

Classifying and identifying peers allows policymakers

and stakeholders to better understand the position of

their economies in a globalized context as well as to

conduct constructive benchmarking. To select peers

we utilized a combination of principal components

analysis (PCA), k-means clustering, and agglomera-

tive hierarchical clustering.50 These commonly used

data science techniques allowed us to group metro

areas with their closest peers given a set of economic

and competitiveness indicators. We used 35 variables

in the PCA analysis (see Table 1). We do not include

change-over-time metrics in the clustering algorithm,

but analyze change variables within and across

metropolitan groupings to summarize key trends. For

more details, see Appendix A.

This report creates metropolitan groupings based

on these factors, summarizes the distinguishing

characteristics of each group, and then examines

trends within each using a range of indicators. It is

important to clarify the two ways in which we use

these data. First, we use point-in-time data to create

the metropolitan typology. Those indicators and their

vintage are outlined in Table 1. Second, we examine

change-over-time trends for these same indicators

within the analysis. The variables used to measure

competitiveness factors come from a variety of

sources, including public and private datasets, and

as a result the periods for which we can measure

key characteristics vary considerably. The analysis of

economic and industrial characteristics looks at data

between 2000 and 2015; for flows of greenfield FDI

we use data corresponding to 2009-2015; for venture

capital flows we use data for 2006-2015; for patents

we look at stock of patents between 2008 and 2012;

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

1 3

to measure impact of university research we use the

2010-2013 period; the analysis of population with

tertiary education corresponds to 2014 or latest year

available; aviation passengers uses data for 2004

and 2014; and internet average download speed

corresponds to the 2008- 2015 period. For a more

detailed description of the data sources please see

Appendix A.

Table 1. Indicators used in the clustering algorithm, 2015 or most recent year available

Dimension Indicator Source

Economic and Industrial

Characteristics

Population, 2015 Oxford Economics, U.S. Census

Bureau

Gross domestic product, 2015 Oxford Economics, Moody's Analytics

Gross domestic product per capita, 2015 Oxford Economics, Moody's Analytics,

U.S. Census Bureau

Output per worker, 2015 Oxford Economics, Moody's Analytics

Industry share of overall output, 2015 Oxford Economics, Moody's Analytics

Industry output per worker, 2015 Oxford Economics, Moody's Analytics

Traded Clusters Greenfield foreign direct investment, 2009-2015 fDi Intelligence data

Greenfield foreign direct investment per capita,

2009-2015

Greenfield foreign direct investment jobs

created, 2009-2015

Innovation Share of total publications in top 10 percent

cited papers, 2010-2013

Centre for Science and Technology

Studies (CWTS) and Leiden University

dataShare of total publications done with industry,

2010-2013

Total patents, 2008-2012 REGPAT

Total patents per capita, 2008-2012

Venture capital investments, millions of dollars

per 1,000 inhabitants, 2006-2015

Pitchbook

Venture capital investments, millions of dollars,

2006-2015

Talent Share of population 15+ with tertiary education,

2014 or latest year available

Oxford Economics, U.S. Census

Bureau

Infrastructure Connectivity Total aviation passengers, 2014 SABRE

Total aviation passengers per capita, 2014

Average internet download speed, 2015 Net Index

Governance Data not available across all metro areas N/A

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

1 4

These metros concentrate economic activity because

they house the competitiveness assets required to

drive global growth. They have attracted more than

$5.4 trillion in greenfield FDI since 2009, more than

one-quarter of the global total; six of the top 10 larg-

est inflows were destined for the Asian metros of

Singapore, Shanghai, Hong Kong, Beijing, Suzhou, and

Chongqing. When controlling for population size, FDI

concentrations are still greatest in many of these Asian

metros, but smaller metro economies in North America

(Austin and Vancouver), Europe (Birmingham and

Barcelona), and Australia (Sydney) also join the top 10.

The top 123 metro economies are critical generators

of new scientific research and innovation. Together,

they account for 44 percent of the world’s most

scientifically impactful research universities, gener-

ate 65 percent of all patents, and attract 82 percent

of all venture capital. The largest patent-producing

metros are among the largest economies in the

world, including Tokyo, Seoul-Incheon, Shenzhen,

Osaka, and San Jose. However, in terms of patents

per capita a smaller set of highly innovative cities

rises to the top: San Jose, San Diego, San Francisco,

Boston, and Stuttgart. Many of these metro areas

I V. M A P P I N G T H E E C O N O M I C A S S E T S O F G L O B A L C I T I E S

T he world’s large metropolitan areas are notable in their economic

primacy. With about 13 percent of the world’s people, 123 large metro

economies generate nearly one-third of global economic output. Nearly

all of the 123 largest metro economies studied in our analysis generate

more than $100 billion in annual economic output (in nominal terms), led by Tokyo

($1.6 trillion) and New York ($1.5 trillion).51

Figure 1. Global share of competitiveness factors, 123 largest metros, 2015 or most recent

year available

Airports in

Top 50 by

Passenger Traffic

Venture

Capital

Stock

PatentsResearch

Universities

Global

Output

FDI flowGlobal

Population

86% 82%

65%

44%

32% 27%

13%

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, and Pitchbook.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

1 5

are also among the most educated in the world.

San Jose, San Francisco, and Boston join Singapore,

London, Washington, and Madrid as the metros with

the highest shares of their populations with tertiary

education.

These metros also concentrate much of the world’s

critical infrastructure. In 2014, airports in these metro

areas transported more than 4.9 billion air passen-

gers. The largest metro economies in the world, which

house multiple large airports, move the most avia-

tion passengers. New York, London, Shanghai, Los

Angeles, Tokyo, Beijing, Chicago, and Atlanta had the

highest passenger volumes in 2014. The 123-metro

sample contains 86 percent of the world’s 50 busiest

international airports.

T H E S E V E N T Y P E S O F G L O B A L C I T I E S

This collective economic clout, however, masks the

significant variation in which competiveness factors

are distributed across these cities. While each met-

ropolitan economy in our sample possesses a unique

trade, innovation, talent, and infrastructure connectiv-

ity profile, the distribution of these assets reveals a

clear typology of places. We used advanced statistical

techniques to cluster metro economies based on their

size, industrial structure, and competitiveness fac-

tors. In some cases, these groupings align to specific

regions, like in China or the United States. But just as

often the groupings unite metro economies from dif-

ferent parts of the world, showcasing that they share

more in common with far-flung counterparts than

with their regional neighbors. And while we include

only point-in-time measures in the clustering algo-

rithm, the resulting groupings perform quite similarly

on growth metrics.

Map 1. Seven Types of Global Cities, 2015

123 Largest Metropolitan Areas Groups

Factory China

Knowledge Capitals

Emerging Gateways

Asian Anchors

Global Giants

American Middleweights

International Middleweights

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

1 6

When grouped into seven metropolitan categories,

the distinct competitive positions of the world’s larg-

est metro economies become sharper, and the result

is a resource that peer metropolitan areas can utilize

for common solutions and investments to enhance

economic growth:

ä Global Giants: six large, wealthy hubs with concen-

trations of corporate headquarters; they serve as

the command and control centers for the world’s

largest advanced economies.

ä Asian Anchors: five large, business and financial

nodes anchoring inward investment into the Asia-

Pacific and Russia.

ä Emerging Gateways: 28 large business and

transportation entry points for major national and

regional emerging markets in Africa, Asia, Eastern

Europe, and Latin America.

ä Factory China: 22 second- and third-tier Chinese

cities distinctly reliant on export-intensive manu-

facturing to power economic growth and global

engagement.

ä Knowledge Capitals: 19 mid-sized, highly produc-

tive knowledge creation centers in the United

States and Europe with talented workforces and

elite research universities.

ä American Middleweights: 16 mid-sized U.S. metro

areas striving for a post-recession niche in the

global economy.

ä International Middleweights: 26 mid-sized cities

in Australia, Canada, and Europe globally con-

nected by people and investment flows, but where

growth has lagged after the financial crisis.

Table 2. Seven types of global cities, 2015

Group name Metro areas

Number of

observations

Global Giants London, Los Angeles, New York, Osaka-Kobe, Paris, and Tokyo 6

Asian Anchors Beijing, Hong Kong, Moscow, Seoul-Incheon, Shanghai, and Singapore 6

Emerging

Gateways

Ankara, Brasilia, Busan-Ulsan, Cape Town, Chongqing, Delhi, East Rand,

Guangzhou, Hangzhou, Istanbul, Jinan, Johannesburg, Katowice-Ostrava,

Mexico City, Monterrey, Mumbai, Nanjing, Ningbo, Pretoria, Rio de Janeiro, Saint

Petersburg, Santiago, Sao Paulo, Shenzhen, Tianjin, Warsaw, Wuhan, and Xi'an.

28

Factory China Changchun, Changsha, Changzhou, Chengdu, Dalian, Dongguan, Foshan, Fuzhou,

Haerbin, Hefei, Nantong, Qingdao, Shenyang, Shijiazhuang, Suzhou, Tangshan,

Wenzhou, Wuxi, Xuzhou, Yantai, Zhengzhou, and Zibo

22

Knowledge

Capitals

Atlanta, Austin, Baltimore, Boston, Chicago, Dallas, Denver, Hartford, Houston,

Minneapolis, Philadelphia, Portland, San Diego, San Francisco, San Jose, Seattle,

Stockholm, Washington DC, and Zurich

19

American

Middleweights

Charlotte, Cincinnati, Cleveland, Columbus, Detroit, Indianapolis, Kansas City,

Miami, Orlando, Phoenix, Pittsburgh, Riverside, Sacramento, San Antonio, St.

Louis, and Tampa

16

International

Middleweights

Brussels, Copenhagen-Malmö, Frankfurt, Hamburg, Karlsruhe, Köln-Düsseldorf,

Milan, Munich, Nagoya, Rome, Rotterdam-Amsterdam, Stuttgart, Vienna-

Bratislava, Athens, Barcelona, Berlin, Birmingham, (UK), Kitakyushu-Fukuoka,

Madrid, Melbourne, Montreal, Perth, Sydney, Tel Aviv, Toronto, and Vancouver

26

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

1 7

G L O B A L G I A N T S

G lobal Giants serve as the command and con-

trol centers of the world’s largest advanced

nations. This group includes the largest cities

in the United States (New York and Los Angeles),

Japan (Tokyo and Osaka-Kobe), France (Paris), and

the United Kingdom (London). These metro areas

not only serve as the main entry points for their

extremely powerful nations, but as the world’s most

significant concentrations of wealth, corporate deci-

sion making, and international exchange.

The first characteristic that binds these metro areas

together is their size. On average, Global Giants house

19.4 million residents and generate over $1 trillion in

real output, three times more than the next largest

set of economies, the Asian Anchors. If they were a

single country, they would be the world’s third largest

economy. Beyond their overall economic clout, these

metro economies are highly productive and generate

enormous wealth. They have the second highest aver-

age nominal GDP per person ($58,000) and GDP per

worker ($116,000) among the metro groups, behind

only the Knowledge Capitals.

These wealth levels stem from the concentration of

financial and business services, which generate 41

percent of gross value added (GVA), on average, in

this group. About 20 percent of the Forbes Global

Map 2. Global Giants, 2015

Figure 2. Global Giant indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

1 8

Table 3. Global Giants economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per capita

2015 ($)

Tokyo 37,004 1,623,904 43,884

New York 20,182 1,492,242 73,938

Los Angeles 13,340 927,562 69,532

London 14,855 831,100 55,947

Paris 12,524 818,522 65,354

Osaka-Kobe 18,640 680,997 36,535

Global Giants Average 19,424 1,062,388 57,532

Source: Oxford Economics, U.S. Census Bureau, and Moody’s Analytics.

Figure 3. Average metropolitan gross domestic product, 2015

American

Middleweights

Factory

China

International

Middleweights

Emerging

Gateways

Knowledge

Capitals

Asian

Anchors

Global Giants

$148,797 $205,657$234,238

$264,926$282,801

$668,056

$1,062,388

Source: Oxford Economics and Moody’s Analytics.

Figure 4. Gross value added by type of service, 2015

Other Sectors

Business,

Financial,

Professional

Services

Factory

China

Emerging

Gateways

American

Middleweights

International

Middleweights

Knowledge

Capitals

Asian

Anchors

Global

Giants

88%

74% 69%68%

64% 60%59%

12% 26%

31%32%36% 40%41%

Source: Oxford Economics, U.S. Census Bureau, and Moody’s Analytics.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

1 9

2000 and 18 percent of global firms with more than

$1 billion in revenue, plus five of the world’s seven

largest stock exchanges by market capitalization, are

headquartered in these six markets. Dense clusters of

advanced-producer-services firms in law, accounting,

management consulting, and advertising have formed

to support the complex decision making occurring in

the financial markets and board rooms of multina-

tional firms.53

These are also the world’s major nodes for flows of

people, capital, and knowledge. In 2014, over 800

million aviation passengers traveled through these

markets, by far the highest total of any grouping.

Global travelers often stay to live and work; a little

under one in six residents of a Global Giant is foreign

born.54 Capital flows seamlessly through Global Giants.

Foreign investors parked an average of $25 billion in

these markets between 2009 and 2015, the second

highest after the Asian Anchors. Finally, knowledge

creation is increasingly a major function of these

metro economies. Among the seven types of metro

areas, Global Giants have the highest education levels,

the second highest patenting rates, and the second

highest share of high-impact scientific publications

in their universities. Every metro area except Osaka

is among the top 15 globally in terms of digital data

flows.55 And venture capital investment data reveal

that they are also sites for budding entrepreneurship,

especially London and New York.56

By nearly every measure these cities are globally inte-

grated and fluent. Saskia Sassen mainstreamed the

phrase “global city” in her 1991 book about London,

New York, and Tokyo. The world’s mobile talent and

capital seek them out, and they have benefited from

multiple cycles of high demand.57 Paris is regularly

cited in this class of global city, but Los Angeles and

Osaka may be more surprising additions given that

they are not generally considered among the world’s

leading financial hubs. However, they loom large

on the global stage by dint of their shear economic

weight—Los Angeles and Osaka are the fifth and sixth

largest metro economies in the world, respectively.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

2 0

A S I A N A N C H O R S

A sian Anchors include five Pacific-facing metro

areas—Beijing, Hong Kong, Seoul-Incheon,

Shanghai, and Singapore— as well as Moscow,

which, while more aligned with Europe, falls in this

group due to its similarity in size, wealth, and reliance

on business and financial services with many of these

Asian metro economies.58 Asian Anchors have many

of the same characteristics as their established coun-

terparts in Europe, Japan, and the United States, but

are not yet as wealthy and globally connected.

The rise of the metros in this group has everything to

do with the rise of Asia. The ascent of the Asian Tiger

economies followed by the gradual liberalization of

China and Russia positioned these cities as the gate-

ways between the global investment community and

their fast-growing nations. Those foreign investment

streams brought new industries and capabilities to

many of these cities, which have since been bolstered

by local investments in infrastructure and skills.

Asian Anchors are now among the cities with the larg-

est concentrations of people and market activity in the

Map 3. Asian Anchors, 2015

Figure 5. Asian Anchors indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

2 1

world. These metros have an average population

of 16.1 million residents and an average GDP of

$668 billion, the second largest figures among the

seven groups. GDP per capita in these regions has

grown by a robust 4.2 percent per year since 2000.

On average residents of the Asian Anchors are now

firmly rooted in the global middle class. Interestingly,

this average masks significant differences in nominal

GDP per capita among the wealthiest metros in this

group, Singapore ($84,000) and Hong Kong ($57,000),

and the lowest-income metros, Shanghai ($33,000)

and Beijing ($30,000). In line with convergence theory,

the lower-income city-regions in this group have seen

the fastest income growth since 2000.

Despite their disparities in wealth, several character-

istics bind this group, especially the five Asian metro

areas. First, the generous inflows of FDI distinguish

these regions from the rest of the world. On average,

$46 billion in greenfield FDI entered each of these

markets between 2009 and 2015, nearly double the

average of the next highest grouping. No metro areas

in the world attracted more FDI than Hong Kong

and Singapore during this period, and Beijing and

Shanghai were not far behind. These cities provide a

distinct value proposition for foreign investment: they

afford access to a rapidly growing Asian consumer

market; they provide strong infrastructure connec-

tivity—Asian Anchors rank second in total aviation

Table 4. Asian Anchors economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per capita

2015 ($)

Seoul-Incheon 25,095 903,466 36,002

Shanghai 24,768 809,507 32,684

Moscow 12,194 749,686 61,482

Beijing 21,876 663,590 30,335

Singapore 5,546 468,087 84,399

Hong Kong 7,295 413,999 56,751

Asian Anchors Average 16,129 668,056 50,276

Source: Oxford Economics.

Figure 6. Greenfield foreign direct investment in metropolitan groups (millions of $US), 2009-2015

American

Middleweights

Knowledge

Capitals

Factory

China

International

Middleweights

Emerging

Gateways

Global

Giants

Asian

Anchors

$2,414 $4,671$5,894

$8,681 $10,823

$25,417

$45,966

Source: Brookings analysis of fDi Intelligence and Oxford Economics data.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

2 2

passengers, behind Global Giants, and first in average

internet download speed and relatively well-educated

workforces; and they offer a more conducive regula-

tory and political environment than many peers in the

region.59 It is notable that Moscow has not kept pace

with the other Asian metros in this category in regard

to FDI attraction.

These metro areas, along with Tokyo and Osaka-

Kobe, are where Asia’s business gets done. About 32

percent of gross value added in these six metros is

generated by financial and business services, 10 per-

cent of Global 2000 firms are headquartered in these

markets, and major stock exchanges are located in

Shanghai, Hong Kong, and Seoul. Singapore is a sig-

nificant financial trading hub in its own right. And 41

percent of Moscow’s GVA is in financial and business

services.

Yet, labor productivity in this sector is only about one-

third as high as in Global Giants, revealing that much

work needs to be done to move further up the value-

added chain. These metro areas are not yet on par

with their Western counterparts in terms of patenting

intensity or the scientific impact of their universities,

although they can be considered the innovation hubs

of their respective countries. Beijing and Shanghai

together generate 23 percent of China’s patents,

Moscow generates 55 percent of Russia’s, and Seoul-

Incheon generates 67 percent of South Korea’s.

Patents per capita increased by 78 percent across

Asian Anchors between 2007 and 2012. And the share

of scientific publications generated in these markets

that can be considered high-impact increased by 18

percent between 2009 and 2013, the second fastest

increase among the seven groupings.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

2 3

E M E R G I N G G AT E WAY S

E merging Gateways are 28 large metropolitan

areas from developing economies that serve

as the business, transportation, and oftentimes

political centers of their countries and regions. Nearly

one-third of the cities in this group are the official

capital of their respective countries (e.g., Ankara,

Brasilia, Cape Town, Mexico City, Pretoria, Santiago,

and Warsaw). In fact, eight of the metropolitan

areas in this group serve as the financial centers of

their countries and house the largest national stock

exchange. Many of these cities served as the focal

point of their national economies as the countries

liberalized their markets for flows of trade, invest-

ment, and people at the end of the 20th century.60

Additionally some of these cities also serve as

gateways for entire regions, as is the case for São

Paulo in financial and business services within South

America61; Istanbul connecting the Middle East and

Europe; Johannesburg as the business hub of sub-

Saharan Africa; and Shenzhen as a major complemen-

tary business hub in China to Beijing, Hong Kong, and

Shanghai.62

Map 4. Emerging Gateways, 2015

Figure 7. Emerging Gateways indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

24

Metropolitan areas in this group house on aver age

10 million inhabitants and have an average GDP

of $265 billion, with some megacities boasting

economies of more than $400 billion (São Paulo,

Guangzhou, Shenzhen, Mexico City, Tianjin, Istanbul,

and Chongqing). The average inhabitant of these

metro areas entered the global middle class over

the past 15 years. Real GDP per capita in Emerging

Gateways has grown 5.5 percent annually since 2000

(second fastest after Factory China metros). Nominal

GDP per capita now stands at around $28,000. Asian

metro areas in this group experienced greater GDP

per capita gains (8.1 percent annually) between 2000

and 2015 than did their Latin American (3.2 percent)

and African counterparts (3.6 percent).

Table 5. Emerging Gateways economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per

capita 2015 ($)

Sao Paulo 21,175 579,473 27,366

Guangzhou 13,155 523,554 39,800

Shenzhen 10,816 490,761 45,374

Mexico City 21,099 485,621 23,017

Tianjin 15,646 477,808 30,538

Istanbul 14,627 449,388 30,723

Chongqing 30,159 425,472 14,108

Delhi 23,513 396,449 16,861

Wuhan 10,261 323,517 31,529

Busan-Ulsan 7,812 305,931 39,160

Hangzhou 8,922 274,969 30,820

Nanjing 8,245 271,934 32,983

Rio de Janeiro 12,172 233,238 19,162

Ningbo 7,724 233,000 30,166

Mumbai 21,799 221,192 10,147

Santiago 7,300 213,908 29,303

Jinan 7,066 174,317 24,671

Warsaw 2,901 164,068 56,564

Xi’an 8,606 160,578 18,658

Brasilia 4,076 159,587 39,150

Saint Petersburg 5,190 158,084 30,459

Monterrey 4,404 140,512 31,906

Katowice-Ostrava 5,008 136,218 27,200

Ankara 5,226 133,934 25,630

Johannesburg 4,725 94,096 19,913

Cape Town 3,976 66,599 16,750

East Rand 3,306 62,492 18,904

Pretoria 3,200 61,240 19,141

Emerging Gateways Average 10,432 264,926 27,857

Source: Oxford Economics.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

2 5

These regions disproportionately concentrate their

nation’s competitiveness assets. All the cities in

this group have a higher share of their working-age

population with tertiary education compared to their

national economies. Many are home to their nation’s

only globally relevant research universities. Cities like

Istanbul, Santiago, São Paulo, and Shenzhen account

for more than 40 percent of all the patents produced

in their countries. Business, professional, and techni-

cal services accounted for 25 percent of total output

in these metro areas. However, the productivity of the

average worker in this sector is one fifth that of their

peer metros in the Knowledge Capitals, Global Giants,

and American Middleweight group.

Emerging Gateways are the entry points for global

flows of people and capital. They typically house the

best-connected international airports of their nations.

In 2014 all the airports in these metropolitan areas

transported 800 million passengers, up from the 273

million in 2004. In fact, the average metro, which in

2014 transported 28 million passengers per year, up

from 9 million passengers in 2004, registered the

second fastest annual passenger growth rate—3.5

percent—among all groups, behind only Factory China.

Metropolitan areas in this group received FDI flows

of $58 billion between 2009 and 2015, but on a per

capita basis these investment flows trail most of the

other metro groups. They are not yet on par with the

Global Giants in terms of international business or

with Knowledge Capitals in terms of global innovation,

although their prominence is growing quickly. FDI

flows doubled between 2011 and 2015, and the stock

of venture capital investment grew by 300 percent,

from $4.3 billion in 2010 to $14.1 billion in 2015.

Figure 8. Output per worker in business, financial, and professional services in metropolitan groups,

(thousands of real USD), 201563

Emerging Gateways

Asian Anchors

International Middleweights

American Middleweights

Global Giants

Knowledge Capitals

53

66

158

229

248

254

Source: Brookings analysis of Oxford Economics data.

Figure 9. Aviation passengers compound annual growth in metropolitan groups, 2004-2014

American Middleweights

Knowledge Capitals

Global Giants

International Middleweights

Asian Anchors

Emerging Gateways

Factory China

0.2%

0.3%

0.4%

1.0%

2.3%

3.5%

7.1%

Source: Brookings analysis of SABRE data.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

2 6

FA C T O R Y C H I N A

F actory China comprises Chinese manufacturing

hubs, and the 22 cities are a good represen-

tation of the geographic diversity of China’s

industrial revolution. Factory China includes metros

on China’s east coast (Hefei and Nantong), inland

regions (Chengdu and Zibo), and the Pearl River Delta

(Foshan and Dongguan).64

The metro areas in Factory China are second- and

third-tier population centers that are growing quickly.

The typical city in this group has an average popula-

tion of 8 million and a nominal GDP of $205 billion.

Output and employment have grown in these met-

ros by an outstanding 12.6 and 4.7 percent annually

between 2000 and 2015, the fastest pace among

our seven groups. Real GDP per capita has expanded

fivefold since 2000, from $2,500 to $12,000, rooting

these metros firmly in the global middle class.

Map 5. Factory China, 2015

Figure 10. Factory China indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

2 7

The most salient feature of this group is the extreme

reliance on manufacturing, which accounts for nearly

40 percent of total output in the typical Factory China

city, the highest among all groups. In fact, Factory

China cities were more manufacturing-intensive in

2015 than they were in 2000, when manufacturing

accounted only for 30 percent of their GDP. With only

25 percent of national population, Factory China met-

ros generate one-third ($800 billion) of China’s total

manufacturing value added.

Factory China metro areas plug into the global econ-

omy as nodes in international manufacturing supply

chains, typically providing goods to wealthier con-

sumer markets in advanced economies. Multinational

corporations like Unilever (operating in Hefei),

Goodyear (Dalian), Samsung (Dongguan), DuPont

(Dongguan and Changshu), Intel (Dalian), Pfizer

(Dalian and Hangzhou), and Dell (Chengdu) anchor

manufacturing operations in Factory China.65 This

specialization has proved effective in building wealth

and moving millions of Chinese households into

the global middle class. But growth has come with

significant environmental costs. The heavy industrial

Table 6. Factory China economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per

capita 2015 ($)

Suzhou 10,658 440,255 41,306

Chengdu 14,407 306,458 21,272

Wuxi 6,526 269,957 41,368

Qingdao 9,054 265,789 29,357

Changsha 7,308 245,571 33,604

Dalian 6,942 245,161 35,317

Foshan 7,424 234,737 31,620

Shenyang 8,257 230,103 27,869

Zhengzhou 9,203 209,690 22,784

Tangshan 7,803 190,743 24,446

Dongguan 8,466 186,042 21,976

Yantai 7,057 183,501 26,003

Nantong 7,357 169,781 23,079

Changchun 7,601 162,933 21,435

Fuzhou 7,444 159,572 21,437

Haerbin 10,669 159,238 14,926

Hefei 6,043 156,989 25,979

Shijiazhuang 10,644 156,264 14,681

Xuzhou 8,660 149,682 17,284

Changzhou 4,727 147,281 31,155

Wenzhou 9,275 131,441 14,172

Zibo 4,633 123,273 26,608

Factory China Average 8,189 205,657 25,804

Source: Oxford Economics.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

2 8

activity has resulted in pollutant levels that are 40

times above what the World Health Organization

recommends, and 40 percent of China’s rivers are

polluted.66

Currently, business, financial, and professional ser-

vices—economic activities typically associated with

urban agglomeration—account for only 12 percent of

total output in this group, well below the average of

32 percent for the other groups. The lack of economic

diversification partly explains why cities in this cluster

rank last in flows of FDI, venture capital attraction,

and international passengers. Additionally, only 13 of

the cities in this group house a top-ranked research

university. Factory China metros file only 0.03 patents

per 10,000 employees, and less than 10 percent of the

population 15 years or older has tertiary education.

Figure 11. Manufacturing share of real gross value added in metropolitan groups, 2015

Global Giants

American Middleweights

Knowledge Capitals

International Middleweights

Asian Anchors

Emerging Gateways

Factory China

10.1%

10.7%

12.4%

14.5%

17.4%

23.7%

39.5%

Source: Brookings analysis of Oxford Economics and Moody’s Analytics data.

“Factory China metro areas plug into the global econ omy as nodes in international manufacturing

supply chains, typically providing goods to wealthier con sumer markets in advanced economies.”

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

2 9

K N O W L E D G E C A P I TA L S

K nowledge Capitals tend to be mid-sized

population centers that are among the

wealthiest and most productive in the world.

This group of 19 metropolitan economies has an

average population of 4.2 million, the second smallest

group by population. But because they are so

productive, these metro areas have the third highest

average economic output ($283 billion) and the

highest nominal GDP per capita ($69,000) and GDP

per worker ($136,000) of any group.

Map 6. Knowledge Capitals, 2015

Figure 12. Knowledge Capitals indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

3 0

Knowledge Capitals are the world’s leading knowl-

edge creation centers. They compete in the highest

value-added segments of the economy, relying on

their significant stock of human capital, innovative

universities and entrepreneurs, and relatively sound

infrastructure connectivity.

These places are supremely well educated: 41 per-

cent of their 15-and-over population has obtained a

college degree. Many of these are graduates from the

elite research universities that anchor these metro

economies’ distinct positions in science and technol-

ogy. Universities in this group boast the largest share

of highly cited scientific publications. Of the 100 most

scientifically impactful universities in the world, 20

are located in these cities.

Scientific research tends to translate to new inven-

tions in these regions, which have the highest average

rates of patenting in the world. With only about 1

percent of the world’s population, Knowledge Capitals

generated 16 percent of global patents between 2008

and 2012; shares were even higher in information

technology (22 percent) and life sciences (19 per-

cent). Led by San Jose, San Francisco, and Boston,

Knowledge Capitals also have, by far, the highest ven-

ture capital investment rates per capita in the world.

More than half of all global venture capital funding

flowed to these 19 markets over the past decade.

Finally, controlling for their population size, these

metro economies have the greatest volume of avia-

tion passengers in the world, signifying the substan-

tial flows of business and leisure travelers flocking

Table 7. Knowledge Capitals economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per

capita 2015 ($)

Chicago 9,551 582,496 60,988

Houston 6,657 505,218 75,893

Dallas 7,103 458,043 64,488

Washington 6,098 454,088 74,469

San Francisco 4,656 375,055 80,551

Boston 4,774 370,731 77,651

Philadelphia 6,070 363,644 59,910

Atlanta 5,711 310,822 54,427

Seattle 3,734 285,634 76,504

Minneapolis 3,525 227,417 64,523

San Diego 3,300 217,562 65,938

San Jose 1,977 180,757 91,437

Denver 2,814 179,882 63,916

Baltimore 2,797 178,121 63,673

Stockholm 2,615 167,911 64,223

Portland, Ore. 2,389 159,219 66,640

Zurich 1,972 135,596 68,761

Austin 2,001 119,234 59,591

Hartford 1,211 101,787 84,029

Knowledge Capitals 4,155 282,801 69,348

Source: Oxford Economics, U.S. Census Bureau, and Moody’s Analytics.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

3 1

to these places. However, foreign direct investment

inflows are not as substantial as in other groupings,

revealing that, for all their assets, many of these mid-

sized metros must proactively assert their visibility in

the global marketplace.

Knowledge Capitals overwhelmingly are located in

the United States. All but two (Stockholm and Zurich)

are U.S. cities, including well-known coastal innova-

tion hubs like Boston, San Francisco, San Jose, and

Seattle. But they also include metro economies in the

Midwest (Chicago, Minneapolis-St. Paul) and the South

(Atlanta, Austin, Dallas, Houston), which now tend to

compete in technology-intensive advanced industries

across both manufacturing and services.67 Stockholm

and Zurich represent two of Europe’s wealthiest and

most productive economies, specializing in profes-

sional, scientific, and technical services; finance; and

information technology. Overall, output per worker in

these metro areas is 9 percent higher than in the next

most productive metro grouping.

Not only are Knowledge Capitals more productive

than the rest of their advanced economy peers, but

the gap is widening. Between 2000 and 2015, growth

in annual GDP per capita and GDP per worker aver-

aged 0.9 and 1.4 percent, respectively, in Knowledge

Capitals. This is by no means a blistering pace, but

these growth rates are 37 percent and 69 percent

faster, respectively, then average growth rates across

the other three developed-economy groupings.

Figure 13. Global Share of innovation assets in Knowledge Capital metros, 2015 or most recent

year available

Venture CapitalPatentsTop 750 Research

Universities

Population

50%

16%

6%

1%

Source: Brookings analysis of Oxford Economics, U.S. Census Bureau, Centre for Science and Technology Studies (CWTS)

and Leiden University, REGPAT, and Pitchbook.

“Knowledge Capitals are the world’s leading knowl edge creation centers. They compete in

the highest value-added segments of the economy, relying on their significant stock of human capital,

innovative universities and entrepreneurs, and relatively sound infrastructure connectivity.”

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

3 2

A M E R I C A N M I D D L E W E I G H T S

S ixteen cities form the American Middleweights.

Metropolitan areas in this group are almost

evenly divided between mid-sized production cen-

ters in America’s North and East (Cincinnati, Cleveland,

Pittsburgh, Indianapolis, Detroit) and Southern

cities that have experienced significant population

growth (Miami, Phoenix, Orlando, St. Louis, Tampa,

Sacramento). The average metropolitan area has 3

million inhabitants, generates $149 billion in nominal

output, and has a nominal GDP per capita of $52,000.

Growth in overall output (1.6 percent), GDP per capita

(0.4 percent), and employment (0.7 percent) has

lagged most other metro groupings between 2000

and 2015, perhaps due partly to the high concen-

tration of non-traded clusters in their economies.

American Middleweights have the highest concen-

tration of local services (health care, real estate,

education, and public services), accounting for 28

percent of output and 42 percent of employment.

Moreover, their tradable industries tend to be less

productive than national averages. While many of

the cities in this group are still finding their global

niche, they all maintain at least one globally relevant

Map 7. American Middleweights, 2015

Figure 14. American Middleweights indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

3 3

export sector. For instance, Charlotte, Detroit, and

Phoenix are among the leading metro exporters of

engine and power equipment, motor vehicles, and

semiconductors, respectively. As a group, American

Middleweights increased their exports by 1.9 percent

per year between 2008 and 2014, slightly below the

national average of 2.4 percent in the same period.68

The prevalence of local services accentuated the

impact of the 2008 economic and financial crisis,

particularly in Sunbelt cities that relied heavily on

construction and real estate development to power

economic growth.69 Between 2008 and 2010 the

construction sector shrank 11 percent per year, the

highest drop among all the groups, while the average

home lost 29 percent of its value between 2008 and

2012.70 Cities like Detroit, Miami, Orlando, and Phoenix

saw home price declines of more than 30 percent.

Table 8. American Middleweights economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per

capita 2015 ($)

Miami 6,012 282,514 46,989

Detroit 4,302 218,080 50,692

Phoenix 4,575 214,809 46,958

Riverside 4,489 167,864 37,393

St. Louis 2,812 146,024 51,937

Pittsburgh 2,353 141,339 60,066

Tampa 2,975 140,263 47,144

Charlotte 2,426 131,636 54,253

Sacramento 2,274 126,103 55,449

Orlando 2,387 125,898 52,740

Cleveland 2,061 117,493 57,013

Cincinnati 2,158 115,552 53,553

Indianapolis 1,989 114,936 57,791

San Antonio 2,384 113,910 47,779

Columbus 2,022 113,875 56,328

Kansas City 2,087 110,456 52,914

American Middleweights 2,957 148,797 51,812

Source: U.S. Census Bureau and Moody’s Analytics.

“American Middleweights have a base of educated

workers, research universities and hospitals,

and trad able clusters. Aligning these assets

to improve export competitiveness through

coordinated economic strat egies will be critical

if these metros are to compete in global

markets.”

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

3 4

Figure 15a. Share of output in traded sectors in metropolitan groups, 2015

American Middleweights

International Middleweights

Knowledge Capitals

Global Giants

Asian Anchors

Emerging Gateways

Factory China

51%

54%

57%

57%

60%

61%

69%

Figure 15b. Share of output in local services in metropolitan groups, 2015

Factory China

Asian Anchors

Emerging Gateways

Global Giants

Knowledge Capitals

International Middleweights

American Middleweights

10%

13%

17%

22%

24%

25%

28%

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

At the same time, the manufacturing sector—once the

engine of export-led growth in places like Cleveland,

Detroit and St. Louis—has seen its share of output and

employment decline relative to other sectors of the

economy.71 Due to automation and strong competition

from abroad, manufacturing employment declined 2.1

percent annually since 2000. Today, manufacturing

accounts only for 7 percent of total employment in

this group.

American Middleweights have assets, however. They

house well-regarded research universities. Cities in

this group ranked third among all other groups in the

share of scientific publications in the top 10 percent of

most-cited academic journals. Additionally, one-third of

the working-age population in these markets boasts a

tertiary degree, ranking it fourth among all groups. The

combination of a highly skilled labor force and world-

class research universities is also strengthened by ven-

ture capital per capita, an indicator on which American

Middleweights ranked third among all their peers.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

3 5

I N T E R N AT I O N A L M I D D L E W E I G H T S

I nternational Middleweights include a diverse group

of wealthy cities in Canada (Toronto, Vancouver),

Europe (Brussels, Berlin, Munich, Rome, Milan,

Munich), Asia (Kitakyushu-Fukuoka, Nagoya, Tel

Aviv), and Australia (Sydney, Melbourne). These 26

metros have an average population of 4.8 million,

output of $234 billion, and nominal GDP per capita

of $49,000, fifth among our groups.

International Middleweights are the most varied

group of metro economies. Cities like Toronto, Sydney,

Frankfurt, Madrid, and Copenhagen play a fundamental

role in the provision of business and financial services

in their national and regional economies. In parallel,

industrial centers such as Kitakyushu-Fukuoka, Nagoya,

Stuttgart, Karlsruhe, Milan, and Barcelona gener-

ate significant levels of manufacturing value added

in Japan, Germany, and Southern Europe, respec-

tively. Most have diversified tradable sectors that

tend to specialize in knowledge services, advanced

manufacturing, or some combination of both.

Map 8. International Middleweights, 2015

Figure 16. International Middleweights indicators, 2015 or most recent year available

Download speed

Aviation passengers

Population with tertiary

education

Venture capital

investment

Patents per capita

Scientific research impact

FDI per capita

FDI stockTraded sector productivity differential

GDP per worker

GDP per capita

GDPGDP per worker

GDP per capita

GDP

Economic characteristics Economic growth Traded clusters Innovation Talent Infrastructure connectivity

(+)

(-) Metro Average (n=123)(-) Metro Average (n=123)(-) Metro Average (n=123)

Source: Oxford Economics, U.S. Census Bureau, Moody’s Analytics, fDi Intelligence data, Centre for Science and Technology

Studies (CWTS) and Leiden University data, REGPAT, Pitchbook, and SABRE.

B R O O K I N G S

M E T R O P O L I TA N

P O L I CY

P R O G RA M

3 6

Several shared characteristics bind International

Middleweights. First, they are globally connected by

migration and capital flows. About 22 percent of the

population in these cities is foreign born, the high-

est share among any cluster. Similarly, these met-

ros boast the second highest level of foreign direct

investment per capita, with almost $2,000 dollars

of FDI stock per inhabitant. These metros are well-

educated (33 percent of the working-age population

has tertiary education), house elite universities (the

highest number of research universities of any group

on both an absolute and per capita basis), and gener-

ate new knowledge (third highest rate of patenting

intensity).

Table 9. International Middleweights economic indicators, 2015

Cities

Population 2015

(thousands)

Nominal GDP 2015

($ millions)

Nominal GDP per

capita 2015 ($)

Köln- Düsseldorf 11,488 548,379 47,735

Rotterdam-Amsterdam 7,146 397,399 55,610

Milan 7,722 380,609 49,286

Nagoya 9,049 377,075 41,672

Madrid 6,586 315,507 47,905

Toronto 6,124 292,432 47,750

Brussels 5,540 290,522 52,445

Frankfurt 4,483 270,396 60,321

Munich 3,981 265,693 66,739

Sydney 4,916 251,254 51,115

Rome 4,468 207,502 46,444

Vienna-Bratislava 3,822 200,062 52,341

Barcelona 4,711 197,889 42,010

Melbourne 4,527 197,774 43,690

Kitakyushu-Fukuoka 5,563 194,550 34,970

Stuttgart 3,166 193,143 61,013

Hamburg 3,188 186,506 58,499

Berlin 4,314 185,910 43,100

Karlsruhe 3,056 159,066 52,050

Montreal 4,058 157,734 38,872

Copenhagen-Malmö 3,045 151,041 49,610

Tel Aviv 3,699 144,875 39,162

Perth 2,080 139,282 66,959

Athens 3,844 138,715 36,082

Birmingham (UK) 3,869 132,439 34,233

Vancouver 2,502 114,447 45,738

International Middleweights Average 4,883 234,238 48,667

Source: Oxford Economics.

REDEFINING

GLOBAL CITIES

THE SEVEN TYPES

OF GLOBAL METRO

ECONOMIES

3 7

For International Middleweights, unfortunately,

another characterization they share is sluggish

economic growth. Between 2000 and 2015, output,

GDP per capita, and employment grew 1.6, 0.7, and

1.0 percent annually, each the slowest of any group.

The solid economic growth of metropolitan areas

in Australia (Perth, Sidney, and Melbourne), Canada

(Toronto and Vancouver), and Israel (Tel Aviv), whose

metro economies posted real output growth rates of

3 percent on average, contrasts starkly with the 1.1

percent experienced by their metropolitan peers in

Europe. Further, the international financial crisis of

2008-2009 divides the economic trajectory of this

group of cities. Output, GDP per capita, and employ-

ment all grew faster in the 2000-2007 period than in

the following years. As a result, 12 cities in this group

have yet to return to their pre-crisis GDP per capita

levels and five cities have yet to regain their pre-crisis

employment base. Further, in half of these markets,

employment was lower in 2015 than in 2005, reflect-

ing both a demographic transition as well as lower

participation in the labor market.

Figure 17: Total number of world ranked research universities in metropolitan groups, 2010-2013

American Middleweights

Factory China

Global Giants

Asian Anchors

Knowledge Capitals

Emerging Gateways

International Middleweights

20

21

42

46

51

63

85

Source: Centre for Science and Technology Studies (CWTS) and Leiden University.

“International Middleweights are the most varied group of metro economies. Cities like Toronto, Sydney, Frankfurt, Madrid, and Copenhagen play a funda mental role in the provision of business and financial services

in their national and regional economies.”