mgi urban america full report[1]

7/31/2019 MGI Urban America Full Report[1]

1/54

McKinsey Global Institute

Urban America: US citiesin the global economy

April 2012


2/54

Copyright McKinsey & Company 2012

The McKinsey Global Institute

The McKinsey Global Institute (MGI), the business and economics researcharm o McKinsey & Company, was established in 1990 to develop a deeperunderstanding o the evolving global economy. Our goal is to provide leadersin the commercial, public, and socia l sectors with the acts and insights onwhich to base management and policy decisions.

MGI research combines the disciplines o economics and management,employing the analytica l tools o economics with the insights o businessleaders. Our micro-to-macro methodology examines microeconomicindustry trends to better understand the broad macroeconomic orcesa ecting business strategy and public policy. MGIs in-depth reportshave covered more than 20 countries and 30 industries. Current researchocuses on our themes: productivity and growth; the evolution o globalnancial markets; the economic impact o technology and innovation;

and urbanization. Recent research has assessed job creation, resourceproductivity, cities o the uture, and the economic impact o the Internet.

MGI is led by three McKinsey & Company directors: Richard Dobbs,James Manyika, and Charles Roxburgh. Susan Lund serves as director o research. Project teams are led by a group o senior ellows and includeconsultants rom McKinseys o ces around the world. These teams drawon McKinseys global network o partners and industry and managementexperts. In addition, leading economists, including Nobel laureates, act asresearch advisers.

The par tners o McKinsey & Company und MGIs research; it is notcommissioned by any business, government, or other institution.For urther in ormation about MGI and to download reports, please visitwww.mckinsey.com/mgi.


3/54

McKinsey Global Institute

James ManyikaJaana RemesRichard DobbsJavier OrellanaFabian Schaer

Urban America: US citiesin the global economy

April 2012


4/54

In todays challenging economic environment, the United States needs toovercome the hangover rom the Great Recession and build a plat orm or long-term growth. The overwhelming role that c ities play not only as home to the vastmajority o Americans but also as the dominant driver o the nations economicgrowth and a major continuing contributor to world GDP growth argues or a keenocus on their prospects.

In this report, MGI, the business and economics research arm o McKinsey &Company, puts US urban growth prospects into a global context as par t o ourcontinuing work on global urbanization and on US growth and renewal. Wecompare the economic weight o US large cities with the urban centers o otherdeveloped regions, describe GDP growth patterns across US urban centers overthe past ew decades, and highlight the key challenges and opportunities or USurban regions as they plot their course through to economic recovery. We plan topublish an update to our March 2011 report Urban world: Mapping the economic

power of cities in spring 2012.

Jaana Remes, MGI senior ellow based in San Francisco, led this project, wi thguidance rom James Manyika, a director o MGI based in San Francisco, and

Richard Dobbs, a director o MGI based in Seoul. The project team comprisedFabian Schaer and Javier Orellana. We are grate ul or the advice and input o many McKinsey colleagues, including Shannon Bouton, David Cis, John Horn,

Alex Maasry, Lenny Mendonca, Marcela Merino, Scott Nyquist, Asutosh Padhi,Michael Tavilla, Samantha Test, Jonathan Woetzel, and Marwa Joy Zohdy. Theteam also appreciates the contribution o Janet Bush, MGI senior editor, whoprovided editorial support; Rebeca Robboy, MGI external communicationsmanager; Julie Philpot, MGI editorial production manager; and Marisa Carder,graphics specialist.

We are grate ul or the invaluable guidance we received rom many experts in

academia, industry, and government. Our particular thanks go to Richard Cooper,Maurits C. Boas Pro essor o International Economics in the Department o Economics at Harvard; Michael Storper, pro essor o urban planning at theUniversity o Cali ornia, Los Angeles; Alan Berube and Emilia Istrate o theMetropolitan Policy Program at the Brookings Institution; Jed Kolko, chie economist at Trulia; and Doug Henton, chairman and CEO o CollaborativeEconomics.

Pre ace


5/54


6/54

US cities

of US GDP was generatedby 259 large cities in 2010,while the large cities ofWestern Europe contributedless than

Almost

85%

65%

of the US population

lives in large cities,vs. less than

in Europe80% 60%

of global GDP growthto 2025 will comefrom large US cities

More than

10%


7/54


8/54


9/54

Urban America: US cities in the g lobal economy McKinsey Global Institute

Executive summary 1

Mapping the role o US cities in an evolving urban world 7

1. Middleweight cities elevate the US urban economic weightabove those o other regions 9

2. There is no single path to rapid growth among US large cities 17

3. To regain growth momentum, US cities need to navigatechoppy waters 27

4. Implications or policy makers and companies 35

Bibliography 41

Contents


10/54


11/54


1

I the 21st century is the century o cities, as some observers characterizeit, urban America begins the millennium in a strong position. Large cities inthe United Statesand in particular the nations broad swath o dynamicmiddleweightsdominate the economy as in no other region o the world. 1 Theyalso loom large in the urban world. Almost one in seven o the City 600, thegroup o cities that is expected to contribute 60 percent o global GDP growthto 2025, is in the United States. Large US cities are expected to generate morethan 10 percent o global GDP growth in the next 15 years, a larger contributionthan all o the large cities o other developed countries combined. So although theburgeoning cities o Asia have seized the public imagination, US cities will remainan important part o the US and global growth story over coming decades.

But US cities ace turbulent times ahead as the economy strives to recover romthe Great Recession. In the next ew years, many cities are likely to grapple withthe dampening impact o deleveraging on economic activity as the public sectorand individuals attempt to pay o high debt levels, as well as persistently highpockets o unemployment. They also ace longer-term headwinds including theaging o the population, which will require even more emphasis on boostingproductivity, innovation, and skills. Policy makers and businesses need to ndways through these di culties in order to play their part in the growth andrenewal o the US economy. 2

In the past, the diverse pool o US cities has ound many di erent ways to expandand become more prosperous. There has been no single recipe or successnor is there likely to be one in the period ahead. In this repor t, MGI examinesthe importance o cities in the US economy and compares their role to the citieso other regions. The report describes urban GDP growth patterns over pastdecades and highlights some o the major trends acing urban America to arriveat a sense o how cities will navigate the challenges ahead.

1 Consistent with the MGI Cityscope database, we de ne large cities as metropolitan areaswith populations o 150,000 or more. The 259 large US metropolitan areas consist o the two

megacities o New York, New York and Los Angeles, Cali ornia, with populations o ten mil lionor more, and 257 middleweight cities wi th populations o between 150,000 and ten million.

2 For readers interested in MGIs work on growth and renewal and employment in theUnited States, see, or example, Growth and renewal in the United States: Retooling Ameri caseconomic engine , February 2011; and An economy that works: Job creat ion and Americasfuture , June 2011 (www.mckinsey.com/mgi).

Executive summary


12/54

2

US CITIESPARTICULARLY MIDDLEWEIGHTS WILL CONTINUE TO DRI VE GROWTH

Today, large US cities have more weight in the US economy than do large cities inany other major region. In 2010, 259 large US cities generated almost 85 percent

o US GDP. During the same per iod, large cities in Western Europe accounted orless than 65 percent o the regions GDP. Among emerging regions, metropolitanChina accounted or 78 percent o Chinas GDP and the large cities o Latin

America contributed 76 percent to regional GDP.

Large US cities have such relative economic weight or two reasons. First, theyare home to 80 percent o the population compared with less than 60 percent inWestern Europe. Second, they have a relatively high per capi ta GDP premium.

The average per capita GDP o large US cit ies is a lmost 35 percent higherthan in smaller cities and rural areas; in Western Europe, this premium is about30 percent (Exhibit E1).

The relative weight o di erent regions in the world economy changes when wehome in on the economic clout o their large cities. Even though Western EuropesGDP exceeded that o the United States by nearly 10 percent in 2010, thecombined GDP o large US cities exceeds that o large Western European citiesby more than 20 percent.

Exhibit E1

383

257

130134

100

ChinaLatin America

WesternEurope

UnitedStates

485558

80

ChinaLatin America

WesternEurope

UnitedStates

The relative importance of cities as centers of economicgravity varies across regions

SOURCE: McKinsey Global Institute Cityscope 1.5; McKinsey Global Institute analysis

Large city GDP as a share of region 1

Large city population as a share of region

Per capita GDP gap between large cities andthe rest of the economy 2

1 We define large cities as having 150,000 of more inhabitants in the United States, and Western Europe. In China and Latin America, we include only cities with 200,000 inhabitants plus in 2010. GDP is in PPP US dollars.

2 The rest of the economy comprises cities with fewer than 150,000 inhabitants as well as rural areas.

%, 2010

787664

84

ChinaLatin America

WesternEurope

UnitedStates

186259 289 710

Number of largecities in region

Developed

Developing


13/54

3Urban America: US cities in the g lobal economy McKinsey Global Institute

It is Americas cities that explain why the United States continues to enjoy higherper capita GDP than Europe. The higher share o US urbanitesand the act thatthey command a larger per capita GDP premium over US smaller towns and ruralareas than do their European counterpartsexplains three-quarters o the per

capita GDP gap between the two economies.

The nations largest and well-known megacit ies o New York, New York, andLos Angeles, Cali ornia, wi ll continue to prosper. 3 New York is on course to remainthe second-largest city by GDP in the world in 2025, and Los Angeles to riserom sixth place today to become the ourth-largest city. But the weight o these

megacities in the US economy is not decisive to the overall importance o cities inthe United States. London and Paris have a smaller share o the overall WesternEuropean population6 percent, compared with the combined population o the US megacities o 10 percent o the total US populationbut they enjoy asigni cantly higher per capita GDP premium than their US counterparts. Paris and

London contribute 9 percent to Western Europes overall GDP, compared with the13 percent contributed by New York and Los Angeles.

Instead, the true vigor o Americas urban economy comes rom a broad base o dynamic middleweights and the relatively high per capi ta GDP they achieve. Thereare just over 255 middleweight cities in the United States, compared with justover 180 in Europe. And they generate more than 70 percent o US GDP today,compared with just over 50 percent in Western Europe. In act, the top 28 USmiddleweights alone contribute more than 35 percent o US GDP. The dynamismo middleweights in the United States is a characteristic o todays global urbanexpansion, making them an interesting group to understand or both US andglobal growth prospects.

US CITIES H AVE VARIED WIDELY IN THEIR PERFORMANCETHERE IS NO SINGLE BLUEPRINT FOR FUTURE SUCCESS

While the overall per ormance o urban America has been a strong one, the ateo individual cities has varied widely. Among middleweights in the top 30 cities,considerable changes have occurred in their rankings by GDP over the past30 years. Cleveland, Ohio, or instance, dropped rom 17th place to 27th, whilePhoenix, Arizona, rose rom 28th place to 13th. Five cities have dropped out o the top 30 completely and been replaced by newcomers. 4

3 The metropolitan area o Los Angeles includes the Cali ornian cities o Long Beach and Santa Ana, and the metropolitan area o New York includes Newark, New Jersey.

4 The metro areas that have dropped rom the top 30 by GDP rom 1978 to 2010 are NewOrleans, Louisiana; Milwaukee, Wisconsin; Columbus, Ohio; Indianapolis, Indiana; and Bu alo,New York; and the new entrants to the top 30 are Riverside, Cali ornia; Portland, Oregon;

Tampa and Or lando in Florida; and Sacramento, Cal i ornia.


14/54

4

The GDP growth o a c ity consists o growth in its population and increases in itsper capita GDP. Looking at large cities on these two dimensions, it is the diversityo their per ormancenot the similaritythat is striking (Exhibit E2). Four eatureso US urban growth over the past three decades stand out.

Di erent population growth rates explain most o the di erences inGDP growth per ormance among US cities. Fast-growing cities on averageposted growth in their populations o two and a hal times the national averageeven while experiencing per capita GDP growth rates nearly identical to thenational average. 5

A avorable mix o sectors is a actor in the ast growth o top-per orminglarge US citiesbut is less important than observers o ten assume. Themix o sectors o the astest-growing cities can expla in 15 percent higher GDPgrowth than the average urban GDP growth rate. 6 But in act this group hasoutpaced the average by more than 45 percent, indicating that a avorable mixo sectors explains only one-third o their outper ormance.

Broad economic trends contribute to the diversity o experience acrossUS cities. Changes in the economic environment help explain why some citiesthrive and others dont. We have seen the rise and decline o manu acturingcities; the li t that Sun Belt cities in the South and West have received romtheir avorable climates; and the impact on Eastern and Western cities rom ashi t in global economic activity, away rom Europe and toward Asia, and romthe Atlantic to the Paci c.

The diversity o growth patterns among strongly per orming USmetropolitan areas suggests that there is no single path to economicsuccess. Cities that have outper ormed their peers in GDP growth includerapidly growing gazelles such as Austin, Texas, and Raleigh, NorthCarolina, which have outper ormed the US average in both per capitaGDP and population growth by build ing on their high-tech presence andstrong collaboration with local universities. Others such as Dallas, Texas;

Atlanta, Georgia; and Salt Lake City, Utahwhich we might call a ordablemetropoliseshave outper ormed the average national average GDP growthbecause their populations have expanded rapidly despite per capi ta GDPgrowth that was slower than average. Yet another set o large, establishedcities such as Boston, Massachusetts, and Washington, DCalphamiddleweightsoutper orm others with signi cantly above-average per capi taGDP and sustain moderate growth by leveraging the strength o their existingeconomic base.

5 We de ne astest-growing cities as those that have achieved GDP growth that is 25 percenthigher than the US average between 1978 and 2010.

6 The sector breakdown is based on a 20-sector split o the economy.


15/54


Exhibit E2

Compound annual growth rate, 19782010%

The growth patterns of individual cities have varied widely

Tallahassee

McAllen

Anchorage

Colorado Springs

Bakersfield

Oklahoma City

New York

2.0

Per capita GDP growth

3.0

1.2

1.1

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

2.9

2.8

2.7

2.6

2.5

0

-1.1

Population growth

5.04.83.83.63.43.23.02.82.62.42.21.81.61.41.21.00.80.60.40.20-0.2-0.4-0.6

Dover

Fort Collins

2.4

2.3

2.2

2.1

2.0

1.9

1.8

1.7

1.6

1.5

1.4

1.3

Detroit

Minneapolis

Phoenix

Miami

AtlantaSan Francisco

Boston

Philadelphia

Dallas

Houston

Chicago

Los Angeles

Raleigh

New Orleans

Buffalo

Austin

Las Vegas

Charlotte

Kansas City

Orlando

Pittsburgh

Portland (OR)

Riverside

San Jose (CA)

Size of bubble =

GDP, 2010US average

SOURCE: Moodys Analytics; McKinsey Global Institute analysisNOTE: For metropolitan regions, we use the first name of the region: e.g., New York for New York-Newark.

New England

Southwest

Great Lakes

Far West

Mid East

Alaska, Hawaii, Puerto Rico

Rocky Mountain

Plains

Southeast


16/54

6

POLICY MAKERS AND COMPANIES NEED TO NAVIGATEDIFFICULT TIMES AHEAD

In the years ahead, US cities will ace a number o strong headwinds againstgrowth. In the short term, many urban centers will need to grapple with high

unemployment and deleveraging. In the longer term, cities will need to dealwith declining population growth, demographic shi ts and aging, and less labormobility than in the past. These are likely to constrain purely population-drivengrowth strategies and urther increase the intense competition or talent.

Yet large US cities have time and again demonstrated that they collec tively havethe resilience and capacity to adjust to new situations. There is no reason this willnot be the case in coming decades, too. As in the past, the trump card or urban

America as it navigates its way toward growth and renewal will be the diversity o strategies and experiences o individual cities. While there is no single blueprintor all cities to ollow, there are workable approaches that have proved e ective

or policy makers and businesses.

Key approaches for policy makers

Know thysel and tailor strategy accordingly. Cities need to understandtheir strengths and weaknesses, as well as the impact o demographic andother trends on their prospects, and set their strategy accordingly. Largemetropolises need to compare themselves and benchmark against anincreasingly global urban landscape. Smaller cities need to prioritize to makethe most o their unique strengths.

Excel in execution. The way policies and strategic plans are carried out

is critical to success. Involving the private sector and a broad range o stakeholders has a proven track record.

Be connected. Rather than seeing other cities as competition, buildingconnections to other US and global ci ties can be a source o strength and newideas.

Key approaches for businesses

Be granular in the search o growth. In a diverse US urban landscape,companies need a detailed understanding o where the growth opportunitiesare and need to ensure that they are su ciently nimble to respond to changing

circumstances.

Engage actively in the search or talent. Skills, particularly technical ones,are going to be in increasingly shor t supply. Businesses need to understandwhich cities can o er the most attractive work orce and production assets.

Collaborate with cities to carve out a competitive environment.Companies should work with cities, many o which are keen to attractbusiness, to in orm local leaders o their needs and what policies would bemost conducive to a competitive urban environment.

* * *

Despite the severe challenges they ace, US cities start rom a more robustplat orm than their peers in other developed regions. The strength o urban

America lies in i ts diversity and the broad swath o middleweight citiesan urbanpro le that we are beginning to see play out across the urban world.


17/54


7

Amid the amassed dark clouds o the global economy, the massive wave o urbanization that is rolling across the emerging world i s a welcome shot o dynamism. 7 The most dramatic developments are un olding in Asia, wherethe scale and pace o urban expansion is unprecedented. Chinas economictrans ormation is happening on 100 times the scale o the rst country in theworld to urbanizethe United Kingdomand in just one-tenth o the time.By 2025, the number o urbanites in China will be triple the number in theUnited States.

With the unprecedented pace o growth in urban Asia grabbing the spotlight, it iseasy to orget that US urban centers will continue to be economic powerhouses.Large cities in the United States are the center o gravity o the US economy,today generating almost 85 percent o US GDP. But urban America also has avital role to play in the world economy, generating nearly 20 percent o globalGDP today. Consistent with our global Cityscope database, we de ne largecities as metropolitan areas with populations o 150,000 or more (see Box 1,MGIs Cityscope database). The 259 large US metropolitan areas consist o the two megacities o New York, New York and Los Angeles, Cali ornia, and 257middleweight cities.8

We expect the 259 large US cities to contribute more to global growth thanthe 355 large cities o all other developed countries combined. 9 Within thedynamic group o City 600the 600 cities that are expected to generate morethan 60 percent o global GDP growth to 2025almost one in seven is in theUnited States. The 79 US cities in this dynamic group are expected to contributemore to global GDP than the other 80 cities o the developed world in theCity 600.

In 2025, ranked by GDP, New York is projected to remain the second-largest cityin the world, behind Tokyo, and Los Angeles is projected to become the ourth-

largest city. And it is not just the very largest US cities that will continue to beglobally important. It is the large number and vigor o US middleweight cities thathave contributed to the nations urban clout. How wellor poorlythe cities o the United States per orm as the economy recovers is critical not only or theirhost nation but also or the global economy.

7 Urban world: Mapping the economic power of cities , McKinsey Global Institute, March 2011(www.mckinsey.com/mgi). MGI will publ ish an update to its 2011 urban world work in spring2012.

8 Cities in this report re er to Metropolitan Statistical Areas (MSA) as de ned by the US O ce o Management and Budget. However, in order to compare these cities with urban areas in otherregions, we only include 259 MSAs in this analysis with 150,000 or more inhabitants, andexclude 115 MSAs with populations below that threshold. So when we re er to cities, we meanthe broader metropolitan areas that include not just central city areas within the jurisdictionbut the surrounding MSA. The two megacities o Los Angele s and New York have populationso ten million or more, and 257 middleweight cities have between 150,000 and ten millioninhabitants.

9 Measured at expected real exchange rates.

Mapping the role o US cities inan evolving urban world


18/54

8

Box 1. MGIs Cityscope database

The MGI Cityscope is a database o more than 2,600 cities around the worldthat allows us to understand the evolving shape o global urban economies;extract many di erent city rankings and groupings by region, variable,and target market; and test the growth momentum that comes rom doingbusiness in particular geographies.

The database is, to our knowledge, the largest o its kind. I t can help answera range o questions relevant or the decisions that companies and policymakers need to make: Which cities wil l contribute the largest number o children to the world? Where will most o the new entrants to the work orceand most senior citizens be, and which cities will experience the astestexpansion among consuming middle-class income groups?

For each city, the database includes the ollowing data or 2010 and 2025.

Population by age group

Children (below the age o 15)

Working-age population (aged 15 to 64)

Older population (aged 65 and above)

GDP and per capita GDP

At market and purchasing power parity, or PPP, exchange rates

At predicted real exchange rate, or RER

Number o households by income segment in our income categoriesde ned by annual household income in PPP terms

Struggling (less than $7,500)

Aspiring ($7,500 to $20,000)

Consuming ($20,000 to $70,000)

Global (more than $70,000)

MGIs city-level datasets were developed as part o our previous research,rom existing public survey data, selected data rom external providers, and

MGIs country- and region-speci c models o city growth to 2025.

MGI will publish an update o its 2011 urban world report in spring 2012,with a broader set o variables included in the Cityscope database that willshed light on the diversity o urban market growth prospects across di erentindustries. The new Cityscope metrics include a view on markets such asbank deposits by city, estimates o residential and commercial foor space,and municipal water demand.


19/54


9

The economic weight o large c ities in the United States is the highest amongthe major regions o the world. While the very largest cities o the United Statesaccount or part o the economic clout that urban centers have in the USeconomy, it is the nations broad swath o middleweights that di erentiate the USurban landscape rom its peers in developed regions. 10

LARGE CITIES HAVE MORE ECONOMIC WEIGHT IN THE US

ECONOMY THAN IN AN Y OTHER DEVELOPED REGION The cities o the United States punch above the weight o their very largecombined populations. The 84 percent o US GDP that 259 large cities generatedin 2010 compares with a share o 64 percent rom large cities in Western Europe(Exhibit 1). This is roughly on a par with Japan, which has a much smallerpopulation and where Tokyo alone represents almost one-third o the nationseconomy. Among emerging regions, metropolitan areas in China accountedor 78 percent o Chinas GDP in 2010, and the large cities o Latin America

contributed 76 percent to regional GDP. 11

10 Middleweight cities re er to large metropolitan areas with population o 150,000 to ten million.

We make a urther distinction between, small, midsize, and large middleweights (up totwo million inhabitants, up to ve million inhabitants, and up to ten million inhabitants,respectively), recognizing that middleweights are a diverse set o cities.

11 In China and Latin America, MGIs Cityscope includes only cities with 200,000 or moreinhabitants, excluding cities with populations o 150,000 to 200,000 that would make thesegures ully comparable.

Exhibit 1

383

257

130134

100

ChinaLatin America

WesternEurope

UnitedStates

485558

80

ChinaLatin America

WesternEurope

UnitedStates

The relative importance of cities as centers of economicgravity varies across regions


Large city GDP as a share of region 1

Large city population as a share of region

Per capita GDP gap between large cities andthe rest of the economy 2

1 We define large cities as having 150,000 of more inhabitants in the United States, and Western Europe. In China and Latin America, we include only cities with 200,000 inhabitants plus in 2010. GDP is in PPP US dollars.

2 The rest of the economy comprises cities with fewer than 150,000 inhabitants as well as rural areas.

%, 2010

787664

84

ChinaLatin AmericaWesternEuropeUnitedStates

186259 289 710

Number of largecities in region

Developed

Developing

1. Middleweight cities elevatethe US urban economic weightabove those o other regions


20/54

10

There are two reasons that US urban regions have such economic weight. First,cities are home to a higher share o the US population80 percent o US citizenslives in large cities, compared with 58 percent in Europe. In the United States,the traditionally mobile population has gravitated into clusters o large cities,

particularly in coastal regions.12

In contrast, mobility has been lower within, andbetween, European countries. 13 National policies aimed at reducing regionaleconomic disparities have limited migration within individual countries, and acombination o language barriers and national borders within Europe has limitedcross-border migration. European Union structural and cohesion unds have alsotrans erred unds rom richer metropolitan regions in its member states to poorerrural ones. This helps to explain why Europe has a relatively higher share o population living in small cities and rural areas.

A second reason that US metropolitan regions dominate the US economy is theirrelatively high per capita GDP premium. The average per capita GDP o large

US cities is almost 35 percent higher than in smaller cities and rural areas in theUnited States, versus a premium o about 30 percent in Western Europe. 14 Thehigher US per capita GDP premium relative to Western Europe largely refectsdi erences in the mix o citiesa higher share o US population lives in very largemetropolitan areas that tend to have higher average per capita GDP. In addition,the share o large city populations in Europe is higher in Southern Europe, whereper capita GDP tends to be lower, than in Northern Europe, which is wealthieroverall.

So, not only does a big share o the population live in large cities in theUnited States, but the per capita GDP o those inhabitants is higher, urthercontributing to the large economic weight o large cities in the United S tatesrelative to Europe. Taking the population and per capita GDP o US citiestogether, we nd that they account or around three-quar ters o the overall US percapita GDP lead over Western Europe (Exhibit 2).

12 The share o population living in large US cities has risen rom 75 percent in 1980 to77 percent in 1990, 78 percent in 2000, and 80 percent in 2010. We expect migration to largecities to slow down, with the share living in large citie s climbing to only 81 percent in 2025.

13 See Raven Molloy, Christopher L. Smith, and Abigail K. Wozniak , Internal migration in theUnited States , NBER, Working Paper No. 17307, August 2011. The authors nd that USmobility is about twice as high as mobility in most European countries outside o NorthernEurope, based on a 2005 Eurobarometer survey that allows or comparisons with US data.Data rom the European Union Labor Force Survey generally con rm lower mobi lity rates inEurope than across in the United States. It is worth noting that cross-country comparisons o mobility rates are notoriously hard to make due to conceptual di culties in orming a commonde nition o internal mobility.

14 While large metropolitan areas have a dense urban core at their heart, they typically alsoencompass less dense and rural areas with commuting ties to the core. Characterizing largemetropolitan areas as purely urban there ore gives an incomplete picture. For an in-depthdiscussion o the interdependence o rural and urban components o US metropolitan areas,see B. Dabson, Rural-urban interdependence: Why metropolitan and rural Amer ica need eachother, Metropolitan Policy Program, Brookings Institution, November 2007.


21/54


A BROAD BASE OF MIDDLEWEIGHT CITIES EXPLAINS THEECONOMIC CLOUT OF US METROPOLITAN REGIONS

When we look within the pool o large cities across regions, it is not the largestmegacities but the broad base and relatively high per capi ta GDP o a large groupo middleweight cities that di erentiates the United States rom other regions.

The metropolitan area o New York is the second-largest city in the world byGDP, and Los Angeles is the sixth largest. 15 Yet their weight within the rest o the US economy contributes little to explaining the overall importance o citiesin the United States. While London and Paris have a smaller share o the overallEuropean population6 percent, compared with the combined population o the two US megacities o 10 percent o the total US populationthey have asigni cantly higher per capita GDP premium than their US counterparts. Thisnarrows the gap between the GDP o the megacities in the two regions. Parisand London contribute 9 percent to Europes overall GDP, compared with the13 percent share o New York and Los Angeles (Exhibits 3 and 4).

In contrast, Tokyo, the most populous city in the world, towers above theJapanese economy. The Japanese capital alone accounts or 34 percent o thenations GDP. Tokyo and Japans second-largest city o Osaka together generatemore than 49 percent o Japans GDP. 16

15 The metropolitan area o Los Angeles includes the Cali ornian cities o Long Beach and Santa Ana, and the metropolitan area o New York includes Newark, New Jersey.

16 Part o the explanation is the smaller size o Japanese economy ($5.5 trillion in 2010)compared with the United States ($14.7 trillion) and Western Europe ($16.0 trillion). Incomparison, London accounts or 33 percent o UK GDP ($2.2 trillion in 2010), and Paris30 percent o French GDP ($2.6 trillion).

Exhibit 2The larger share of US cities and their higher incomeaccounts for three-quarters of the per capita GDP gapbetween the United States and Europe

34.03.30.96.8

1.946.8


US average Pure share of urban popula-tion

Pure averageincome inlarge cities

Per capitaGDP in ruralareas

WesternEurope 1average

Per capita GDP Per capita GDPpremium in

Europes largecities in lower share of urbanpopulation

Per capita GDPdifference

between USand WesternEurope largecities


between USand WesternEurope ruralareas

Per capita GDP

Urban popula-tion + averageincome


between USand WesternEurope in lower share of urbanareas

1 Austria, Belgium, Switzerland, Germany, Denmark, Spain, Finland, France, United Kingdom, Greece, Ireland, Iceland, Italy,Luxembourg, Netherlands, Norway, Portugal, and Sweden.

53% 26%15% 6%

74% of gap comes from higher populationand per capita GDP of large cities

Per capita GDP, 2010$ thousand 2010, PPP

% of gap

100% of gap = 12.8


22/54

12

Exhibit 3A broad base of large cities explains the economicclout of US metropolitan regions

21

42

35

24

20

7

100% = 127.0

44

410.86

34

WesternEurope

Japan

313.410

UnitedStates

27

29

Concentration of population in top cities%; million

Concentration of GDP in top cities 1

%; $ trillion

2010

16

37

33

23

15

100% = 5.5

4937

16.09

UnitedStates

14.7

13

WesternEurope

Japan

31

29

7

SOURCE: McKinsey Global Institute Cityscope 1.51 GDP measured at real exchange rate (RER).

Small cities and rural areas

Other large cities

Top 30 by GDP

Top two by GDP

Exhibit 4US cities have the highest per capita GDP amongthe largest cities, yet Europes London and Parishave the highest per capita GDP premium

122

144

159

99

133

119

SOURCE: McKinsey Global Institute Cityscope 1.5

Los Angeles 732

London 752

Paris 764

Osaka 815

New York 1,180

Tokyo 1,868

Per capita GDP premiumrelative to region 1Index: 100 = regional per capita GDP at PPP

1 The US average was used for New York and Los Angeles; the Western Europe average was used for London and Paris; theJapan average was used for Tokyo and Osaka.

GDP, 2010$ billion at RERWorld rank

1

2

3

4

5

6

Per capita GDP$ thousand, PPP

United States

Western Europe

Japan

40.6

62.4

53.8

49.1

56.9

33.7


23/54


Instead, it is the strong base o middleweight cities that di erentiates Americasurban landscape. 17 The 28 middleweights that, together wi th New York andLos Angeles, orm the top 30 US cities are more populous than their WesternEuropean peers, home to one-third o the total US population, and responsible

or 37 percent o US GDP. Among these cities are Chicago, Illinois, which is hometo 3 percent o the US population; Dallas-Fort Worth, Texas, with 2 percent;Philadelphia, Pennsylvania, with 1.9 percent; and Boston, Massachusetts; Detroit,Michigan; and San Francisco-Oakland, Cali ornia, each o which accounts orabout 1.5 percent o the US population. These cities together command a higherper capita GDP than the top 30 cities in other regions, contributing to theireconomic weight (Exhibit 5).

Beyond the top 30, there are 227 other middleweight cities in the United States,compared with 183 cities o this size in Western Europe. Together, all the USmiddleweight cities are home to nearly 70 percent o US population and generate

more than 70 percent o the nations GDPa share that is 20 percentage pointshigher than that o European middleweight cities in their region.

17 There are signi cant di erences in the role middleweight cities play across individual WesternEuropean countries. In Germany, middleweights accounted or 49 percent o the countrys

GDP in 2010. French middleweights, in contrast, accounted or only 31 percent o theirnations GDP, refecting the higher concentration o economic activity in the single megacityo Paris with 30 percent o French GDP. The United Kingdom in turn is unusually dependenton greater London, which is home to one-quarter o the UK population, generates one-thirdo the nations GDP, and whose inhabi tants earn 44 percent more than the Western Europeanaverage.

Exhibit 5Middleweights drive the wide per capita GDP gap between large cities andrural areas in the United States

1 We include 229 cities in the United States, 156 cities in Western Europe, and 41 cities in Japan.


43

3034

UnitedStates

JapanWesternEurope

Per capita GDP gap between large cities and the rest of the economy, 2010%

Total gap

Gap in top 2 cities by GDP Gap in top 30 cities by GDP Gap in other middleweights 1

51

77

63

363542

25

1219


24/54


25/54


When we compare US metropolitan economies with those o China, we ndthat the relatively higher weight o US cities also holds, although Chinese urbanareas have made substantial gains over the past ew years. While Chinas GDP

at market exchange rates today amounts to two- ths o total US GDP, the GDPlevel o Chinese large cities is slightly lower at 37 percent o the level o the USlarge cities. Yet this gap has narrowed dramatically in the past three years. In2007, the GDP o large Chinese c ities amounted to only 20 percent o the value o their US counterparts. This change refects the continuing urbanization and rapidper capita GDP growth in China during the years when the United States was inrecession, as well as the appreciation o the renminbi against the dollar. The percapita GDP in PPP terms in Chinas large cities is about one-quarter that o thelevel in US cities, while Chinas overall per capita GDP is only one-sixth that o theUnited Statesand has risen rom around one- th in 2007 (see Box 2, Patternso urban growth in China and India).

In the so-called century o cities, the United States is in a strong initial position.But how sustainable will the clout o US cities prove to be in the uture? In thenext two chapters, we will rst look at the past to understand pat terns o growthamong US metropolitan areas, and then discuss prospects or the uture.

Exhibit 6The disparity in per capita GDP of US citiesrelative to the average is no higher than inWestern European cities

0

-40

140

-20

204060

-60

-80-60-40-20

020406080

-80-60-40-20

020406080

-80-60-40-20

020406080

-80-60-40-20

020406080

-80-60-40-20

020406080


City per capita GDP deviation from regional average, % PPP

46.8

34.035.2

34.0 33.0

36.0

Regional per capita GDP($ thousand)

% of cities

Within 20% of regional average

Beyond 20% of regional average

United States

Western Europe

Japan

Germany

France

United Kingdom

10 66 24

19 58 23

94 6

25 64 4

7 80 13

61 2415


26/54

16

Box 2. Patterns of urban growth in China and India

The pat terns o urban growth vary widely between the two largest Asian nations. WhileChina is already in the middle o a rapid wave o urbanization, India is at a much earlier

stage o this process (Exhibit 7).

China is experiencing a period o massive urbanization, on a scale never be orewitnessed. 1 The portion o the population living in cities has grown rom 36 to50 percent over the past decade, with the greatest growth occurring in clusters o cities located along the eastern coast. Chinas population in large cities is expectedto expand rom approximately 650 mil lion in 2010 to about 890 million in 2025.

This increase is a lmost equivalent to the current population o large c ities in theUnited States. MGI expects large cities to account or more than 90 percent o ChinasGDP growth in the next 15 years. Since 2007, Tianjin, Guangzhou, and Shenzhenhave joined Shanghai, Beijing, and Chongqing as megacities, and in the period to2025, we estimate that our more Chinese cities wil l join their ranks. During this period,nearly three-quarters o Chinas growth is expected to come rom rapidly growing

middleweight cities.In comparison with China, India is still at relatively early stages o urbanization. Only20 percent o the population lives in large cities, o which there are only 234. MGIestimates that large cities, scattered across the nation, will generate nearly 50 percento the nations GDP by 2025. 2 In India, it appears that state borders are limiting mobility,leading to an urban economic concentration in state hubs rather than city clustersacross the nation. Moreover, Indias economic development policies have traditionallyavored small-scale production and discouraged larger-scale operations in cities.

This is another actor slowing Indian urbanization that stands in contrast to both theUnited States and China where more mobi le populations have moved in search o better jobs and other economic opportunities.

1 Preparing for Chinas urban billion , McKinsey Global Institute, March 2009 (www.mckinsey.com/mgi).

2 Indias urban awakening: Building inclusive cities, sustaining economic growth , McKinsey GlobalInstitute, April 2010 (ww w.mckinsey.com/mgi).

Exhibit 7While India is still at an early stage of urbanization, China will continue tosee rapid growth across cities of all sizes including rising megacities


Contribution to country GDP growth, 201025%

Rank 6 to 15

100

Top five cities

51

Total growth

18

Rank 16 to 30 7

All other large cities

10

Small citiesand rural areas

14

100

6

51

11

15

17

In India, small cities and ruralareas continue to matter only234 cities make it into Cityscope

710 large cities in China willgenerate more than 90 percentof the nations GDP growth

In China, three citiesbecame megacitiesbetween 2007 and2010, and four moreare expected to doso by 2025


27/54


17

Given their weight, the per ormance o large cities in the United States has largelydetermined the growth trajectory o the entire economy. The 259 large US citieshave contributed almost 85 percent o US growth since 1978. 19 The largest 30metropolitan areas contributed roughly hal o US growth between 1978 and2010, with the remaining 229 middleweight cities contributing a third. Discerninghow these cities have grown in the past can be a use ul basis or assessing thei ruture prospects.

THE GROWTH OF US CITIES H AS VARIED WIDELY OVER THE PAST THREE DECADES

The GDP growth o a city consists o growth in its population and increases inits per capita GDP. When looking at large cities on these two dimensions, it isthe diversity o their per ormancenot the similaritythat is striking (Exhibit 8).Houston, Texas, or instance, sustained its growth between 1978 and 2010through robust population expansion at a compound annual growth rate o morethan 2 percent over that period. Yet the citys per capita GDP growth, at anannual rate o 0.5 percent, has been only one-third o the US average. In contrast,San Jose, Cali ornia, has experienced relatively modest growth in its population

at just over 1 percent a year, but its per capita GDP has grown at an impressivecompound annual growth rate o around 3 percent. Detroit, Michigan, and NewOrleans, Louisiana, have the same population today as they had 30 years agoand have experienced per capita GDP growth as slow as that o Houston. 20

19 We use the longest comparable MSA GDP and population time series rom Moodys Analytics. The uninterrupted GDP time series is available or a ll la rge cit ies s tart ing in 1978. Furthermore,we use 2010 GDP data to refect city per ormance up until the recent past. Neverthele ss, it isworth noting that 2010 urban GDP is still l ikely below the long-run trend o potential output.

20 It is important to note that the evolution o per capita GDP, which is the variable we ocuson in this report, can di er rom the development o per capita income. However, the broad

patterns o per capita GDP evolution do not di er rom those o per capita income. For a moredetailed discuss ion on the determinants o sub-national income growth, see P. Bauer et al.,Knowledge matters: The long-run determinants o state income growth, Journal of Reg ional Science , 2011; also see Michael Storper, Why does a city grow? Specialization, human capital,or institutions? Institut dEtudes Politiques de Paris and London School o Economics, July2008.

2. There is no single pathto rapid growth among USlarge cities


28/54

18

Exhibit 8


The growth patterns of individual cities have varied widely

Tallahassee

McAllen

Anchorage

Colorado Springs

Bakersfield

Oklahoma City

New York

2.0

Per capita GDP growth

3.0

1.2

1.1

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

2.9

2.8

2.7

2.6

2.5

0

-1.1

Population growth

5.04.83.83.63.43.23.02.82.62.42.21.81.61.41.21.00.80.60.40.20-0.2-0.4-0.6

Dover

Fort Collins

2.4

2.3

2.2

2.1

2.0

1.9

1.8

1.7

1.6

1.5

1.4

1.3

Detroit

Minneapolis

Phoenix

Miami

AtlantaSan Francisco

Boston

Philadelphia

Dallas

Houston

Chicago

Los Angeles

Raleigh

New Orleans

Buffalo

Austin

Las Vegas

Charlotte

Kansas City

Orlando

Pittsburgh

Portland (OR)

Riverside

San Jose (CA)

Size of bubble =

GDP, 2010US average

SOURCE: Moodys Analytics; McKinsey Global Institute analysisNOTE: For metropolitan regions, we use the first name of the region: e.g., New York for New York-Newark.

New England

Southwest

Great Lakes

Far West

Mid East

Alaska, Hawaii, Puerto Rico

Rocky Mountain

Plains

Southeast


29/54


The diversity o economic per ormance o cities in the United States is evidentwhen we look at rankings within its top 30 cities (Exhibit 9). US megacitieshave continued to tower above all others in terms o their populations and theireconomic clout. The megacities o New York and Los Angeles maintained their

rst and second positions between 1978 and 2010, and Chicago, Illinois, which isexpected to cross the ten million population threshold and become a megacity by2025, held on to its third spot. Even though the growth rates o these three citieshave been quite di erent, their strong starting positions helped to maintain theirrank.

But we are seeing dynamic change among US middleweights. The rest o the top30 cities in terms o their GDP have shi ted around over the past three decades,evidence that the ortunes o established cities can change in just a ew decades.O the top 30 cities, ve dropped out o the groupNew Orleans, Louisiana;Milwaukee, Wisconsin; Columbus, Ohio; Indianapolis, Indiana; and Bu alo,New Yorkand were replaced by Riverside, Cali ornia; Portland, Oregon; Tampaand Orlando in Florida; and Sacramento, Cali ornia. Even within the top 30,rankings have shi ted. Cleveland, Ohio, or instance, dropped rom 17th place to27th, while Phoenix, Arizona, rose by 15 places, rom 28th to 13th.

Exhibit 9There have been significant shiftsin the economic weight of the topUS cities over a long period

SOURCE: Moodys Analytics; McKinsey Global Institute analysis

127Washington, DC 146Detroit 148Houston 149Philadelphia 156Chicago 263Los Angeles 307New York 566

51New Orleans 53Baltimore 63Cleveland 66St. Louis 68

Atlanta 69Minneapolis 69Pittsburgh 72Seattle 75Miami 77Dallas 94Boston 113San Francisco

39Columbus (OH) 39Milwaukee 43Denver 48Cincinnati 49Kansas City 49San Diego

Buffalo 35Indianapolis 35Phoenix 36San Jose (CA) 37Hartford

123456789

101112131415161718192021222324

30 Cincinnati 95Kansas City 95Hartford 96Cleveland 98Sacramento 102Orlando 105Pittsburgh 116Portland (OR) 117Tampa 117St. Louis 125San Jose (CA)

732New York 1,180

296Philadelphia 314Dallas 325Houston 341Washington, DC 392Chicago 496Los Angeles

181Phoenix 182Seattle 211Miami 236

Atlanta 250San Francisco 283Boston

127Riverside 132Denver 142Baltimore 145San Diego 173Detroit 179Minneapolis

CityRankReal GDP, 1978$ billion 2010 City

Real GDP, 2010$ billion 2010

Rank risenor declined

000304-31-1500

150-94

177

8-11

-8

15

-511

2928272625

-8

26

-10-2

26

Rose at least 3 spots

Declined at least 3 spots

Entered top 30

Dropped from top 30


30/54

20

FOUR FEATURES OF US URBAN GROWTH STAND OUT

Given the diversity, are there any longer-term patterns that could help shed lighton the growth prospects or US cities as they come out o the Great Recession?Or are there lessons to be drawn that can help us to understand the likely

evolution o rising emerging market middleweights? We have observed oureatures that stand out.

Differences in population growth, not in per capita GDP growth,explain most of GDP growth performance dif ferences acrossUS cities

We nd that di erences in the growth o populations were responsible or thelions share o the di erences in the GDP growth per ormance o individual citiesin the United States rom 1978 to 2010 (Exhibit 10). Cities that have success ullyattracted and accommodated expanding populations have thrived economically.US cities that achieved GDP growth rates o 25 percent or more above the USaverage did not, on average, experience higher growth in per capita GDPinact, both groups o cities posted identical average 1.6 percent increases in per

capita GDP. Instead, the population o rapidly growing cities grew signi cantlyaster. These cities had about 2.5 percent average population growth, more than

two and a hal times the 0.9 percent achieved by other large cities. In additionto Houston, which we have already mentioned, Dallas, Texas; Atlanta, Georgia;and Phoenix, Arizona, are among the middleweight cities that have bene tedrom rapidly expanding populations. Conversely, the slowest-growing cities in

the United States had average population growth o only one-quarter the urbanaverage although, at the same time, their per capita GDP grew. 21

21 Population growth translates into GDP growth when accompanied by employment growth orrising productivity.

Exhibit 10Differences in GDP growth are linked to differences in population growth

SOURCE: Moodys Analytics; McKinsey Global Institute analysis.

1 US Metropolitan Statistical Areas with more than 150,000 inhabitants, excluding three Puerto Rican cities due to lack of data.2 Cities outperforming US GDP compound annual growth or per capita GDP level by at least 25 percent.3 Cities underperforming US GDP compound annual growth by at least 25 percent.

Absolute GDP growth2010 $ trillion

Number of cities

Average GDP growthper city2010 $ billion

2.5

1.6

1.7 1.6

0.90.91.3

Fast-growingcities

2.5

High per capita GDPcities

1.6

0.3

Other cities

2.6

Low-growthcities

4.1

1.6 1.6

1.21.0Population

Per capita GDP

All largecities 1

2.6

UnitedStates

2.8


Winning cities 2

8.2 2.4

47

52.5

1.6

13

123.8

256

7.1

27.9

182

2.5

13.7

14

0.6

40.4


31/54


Although ci ties have taken diverse development paths and there are cities withdi erent growth decompositions within each category, the evidence suggeststhat expanding population has been a key component o overall GDP growthin US cities in the past. There are many objectives that ci ties may have or their

development e orts, but or those cities that seek GDP growth, actors thatimpact population growth are among the most critical ones to address. 22

A favorable sector mix contributes to the fast growth of topperforming citiesbut much less than observers often expect

The most success ul cities have excelled at making the most o their economicsituation and have achieved growth above and beyond the underlying growthtrends o the sectors in their local economy. On the contrary, underper ormingcities grew less than what we would expect based on the composition o theirlocal economy. The sector mix o cities has some predictive power or explainingthe increase in their GDP levels. The astest-growing citiesthose that achieved

GDP growth rates 25 percent higher than the US average between 1978 and2010were expected to experience 15 percent higher growth rates on averagethan large cities overall due to their avorable sector mix (3.2 percent annualcompound growth rate versus 2.8 percent annual compound growth rate). Yetthis group o cities outpaced the average growth rate o large cities by more than45 percent (4.1 percent compound annual rate versus 2.8 percent), indicatingthat a large share o growth or the astest-growing cities did not come only roma avorable sector mix. On the basis o their sector mix, the worst-per ormingcities in terms o GDP growth rates over the same period would have beenexpected to grow at a rate 10 percent below the average US city (2.5 percentannual compound rate versus 2.8 annual compound growth rate). But these

cities actually lagged behind the average by more than 40 percent (1.6 percentannual compound rate versus 2.8 annual compound rate) (Exhibit 11). So citiesthat outper orm their peers do not do so only because they have a avorable mixo ast-growing sectors but because they a lso have aster-than-average growthwithin their sectors. 23

The degree o industry specialization or diversi cationanother indicator o the composition o the local economydoes not appear to be a preconditionor growth or the stability o growth (E xhibit 12). It would be easy to point to a

concentration in a single industry or the r ise and all o Detroit, Michigan, andits big three US automotive companies, or the dramatic shi ts in the ortunes o

the gambling and enterta inment center o Las Vegas, Nevada, as evidence thatoverspecialization leaves cities vulnerable. But over long periods, ast-growingcities do not seem to di er in the degree o concentration o their industry. Nor isthere a signi cant correlation between the degree o economic diversity in a cityand its growth rate and the volatility o that growth. Not even the size o a cityexplains di erences in GDP growth.

22 See, or example, Edward Glaeser, Joseph Gyourko, and Raven E. Saks, Urban growth and

housing supply, Journal of Economic Geography 6, No. 1 (Jan. 2006):71-89.23 This nding di ers rom the patterns that emerge when we compare national growth across

developed economies. See How to compete and grow: A sector guide to pol icy , McKinseyGlobal Institute, March 2010 (www.mckinsey.com/mgi). The nding is closer to those at statelevel in Jed Kolko, David Neumark, and Marisol Cuellar Mejia, Public policy, state businessclimates, and economic growth, NBER, Working Paper No. 16968, 2011.


32/54


33/54


Broad economic trends contribute to the diversity of experienceacross US cities

Industries rise and decline. Cities that have thrived or many years on the back o a particular economic activ ity can nd themselves stranded. Economic tectonic

shi ts a ect not only countries, but also the weight o regions within them. USindustrialization and the subsequent transition to a service-oriented economyover the past century go a long way toward explaining the rise o ormidablemanu acturingor Iron Beltcities such as Detroit, Michigan, and Pittsburgh,Pennsylvania, in the rst hal o the 20th century and their subsequent declineover the past 40 years into what is now called the Rust Belt. The United Stateshas undergone the emergence o a wave o rapidly growing middleweight citiesin Southern and Western statesthe Sun Beltthat originally bene ted rom anabundance o cheap land and technical advances in air conditioning and havesince ound di erent ways o prospering. In the Western United States, cities havealso bene ted rom a shi t in the global economic balance away rom Europe and

the Atlantic and toward Asia and the Paci c. For example, trade has moved to theWest Coast, with container shipping growing at a 3.9 percent compound annualgrowth rate in the por ts o the West Coast rom 1980 to 2010, compared with only2.4 percent on the East Coast. 24

Such shi ts are evident when we compare the actual growth rate o cities withpredictions based on their industry mix (Exhibit 13). 25

24 Container tra c historical data rom Drewry Shipping Consultants.

25 Edward Glaeser, Jose Scheinkman, and Andrei Shlei er, Economic growth in a cross-sectiono cities, Journal of Monetary Economics , Vol. 36, No. 1 (August 1995): 11743.

Exhibit 13Some cities have grown significantly more strongly thanthe composition of their sectors might suggest


Minneapolis

Miami

Los Angeles

Las Vegas

Houston

Detroit

Portland(OR)

Denver

Dayton

Dallas

Chicago

Charlotte

Boston

Baltimore

Austin

Atlanta

10 70-10 5030 90 110 130 150 170 190 210 230 250

Non-agricultural GDP growth, 197820102010 $ billion

20

-20

40

60

100

80

120

140

160

0

180

Growth outperformance: actual over expected GDP growth 1Index: 100 = actual growth on par with expectation

26024022020018020 1401201008060400

Raleigh

Washington, DC

Tampa

St. Louis

Seattle

160

San Jose (CA)

San Francisco

San Diego

Riverside

200

Pittsburgh

PhoenixPhiladelphia

Orlando

New York620

240

220

1 Expected non-agricultural GDP growth computed by applying US GDP growth rate at sector level (19782010) to each sector at city level, using five-year intervals to account for changes in the city sector mix.

Rust Belt

Size of bubble =non-agriculturalGDP, 2010

Sun Belt


34/54

24

The diversity of growth patterns among strongly performingmetropolitan areas suggests that there is no single path toeconomic success

When looking more closely at the group o US cities whose economic

per ormance has outstripped that o their peers, the diversity in the citycharacteristics is almost as striking as or all large cities. We de ne top-per orming as cities that have either grown their GDP at least 25 percent asterthan US average growth between 1978 and 2010 or cities where per capita GDPin 2010 is at least 25 percent higher than the US average. These cities vary widelyin the growth mix o per capita GDP and population that has driven their growthsince the late 1970s (Exhibit 14).

This group includes large, establ ishedor alpha middleweightcities such asBoston, Massachusetts, and Washington, DC, that already enjoy a per capitaGDP that is higher than average and tend to continue growing at moderate rates

by leveraging their established economic base.

Exhibit 14The most successful cities over the past three decadeshave diverse profiles, showing there is no single pathto success

1.5

1.41.3

1.2

1.1

1.0

0.9

0.8

0.7

0.6

0.5

0.1

0

Population

5.03.53.02.01.51.00.50 2.5

Per capita GDP

3.0

2.9

2.8

2.7

2.6

2.5

2.4

2.3

2.2

2.1

2.0

1.9

1.8

1.7

Fort Collins

Visalia

Deltona

Tallahassee

McAllen

Wilmington

Provo

Sioux Falls

Green Bay Springfield (MO)

Modesto

Ogden

Fayetteville (AR)

Boulder

Palm Bay

Lafayette (LA)

Manchester (NH)

Santa Rosa (CA)

Cape Coral

Colorado Springs

Trenton

Boise City

1.6

Anchorage

Durham

MadisonDes Moines

Bakersfield

Oxnard

AlbuquerqueAlbany (NY)Raleigh

Washington, DC

Tucson

Tampa

Seattle

San Jose (CA)

San Francisco

San Diego

San Antonio

Salt Lake City

Bradenton

Riverside

Portland (OR)

Phoenix

OrlandoNew York

Nashville

Miami

Las Vegas

Jacksonville (FL)

Houston

Hartford

Denver

Dallas

Charlotte

BridgeportBoston

Austin

AtlantaSacramento


Sample of cities outperforming US GDP compound annualgrowth or per capita GDP level by 25 percent, 19782010Compound annual growth rate (%)

US average

Outperformed US per capita GDP level by atleast 25 percent in 2010


35/54


Also among the top per ormers are cit ies that we dub gazellescities that havebeen able to increase their per capita GDP and populations signi cantly asterthan the US average by building on their strong university and research presencein growing tech industries. Many o these cities are located in the Sun Belt, and

include Austin, Texas. Raleigh, North Carolina is another gazelle.

Cities such as Dallas, Atlanta, and Salt Lake City, Utahwhich we might calla ordable metropoliseshave outper ormed national average GDP growth,despite slower-than-average growth in their per capita GDP, because thei rpopulations have expanded rapidly. This has refected housing developmentpolicies that have kept the cost o housing relatively a ordable (Exhibit 15).

Some metropolitan areas have been able to grow rapidly within a broadergeographic economic agglomeration. One example o such a city is Bridgeport-Stam ord in Connecticut, which has built a major nancial services sector closeto the city o New York but outside New York States tax jurisdiction. Anotherexample is Boulder, Colorado, a relatively small city that bene ts both rom a highquality o li e but also rom being close to the larger city o Denver, Colorado. Yetanother example is San Jose, Cali ornia, which is integrated into Silicon Valley andwhose GDP and per capita GDP have both grown aster than the nations. Smallercities can prioritize to make the most o any unique strengths they may have.Portland, Oregons selling point, or example, has been its high quality o li e.

Exhibit 15Affordable metropolises offer an attractive trade-off between the cost of living and working opportunities


10498

110118117

100113

145

191

140156

148212

215289

308255

250

397

110

AtlantaHoustonDallasSalt Lake City

PhoenixTampaOrlandoDenver Washington, DC

ModestoRiversideMiamiBostonBoulder OxnardBridgeportSanta RosaSan DiegoSan Jose

Medianhome price 1$ thousand

Median householdincome 1$ thousand

1 Simple average for 19792010.2 Calculated on average values for 19792010.NOTE: Not to scale.

Rank amongwinning cities

Multiple of homeprice to income 2

Average multiple

19792010

Affordablerelative toUS average

US city average

42384039

3832

3543

55

3538

3548

4552

5444

4159

35

2.52.6

2.83.03.13.13.2

3.4

3.4

4.04.14.34.4

4.75.65.75.8

6.0

6.7

3.1

123456789

10

1922303235

40465253


36/54


37/54


27

US cities will continue to be an important contributor to the global economy,but they will need to navigate some choppy economic waters as the economyrecovers rom the Great Recession and then brace themselves to cope with somemajor trends that will continue to hinder growth and that will vary in their severityrom city to city.

We expect the collective GDP o the large cities o the United States to rise byalmost $5.7 trilliongenerating more than 10 percent o global GDP growth to2025. 26 The top 30 US cities alone are likely to contribute 7 percent to global GDPgrowth during this period.

The two US megacities are expected to grow strongly rom 2010 to 2025buteven so US middleweights will outpace them. On their current growth trajectories,New York is poised to maintain its position as the second-largest city by GDP inthe world, behind Tokyo. Los Angeles is projected to become the ourth-largestcity by GDP in 2025. While New York and Los Angeles combined are expectedto post 2.1 percent GDP compound annual growth rates to 2025, the top 30 USmiddleweights are expected to grow on average at compound growth rates o 2.6 percent, above the overall expected US compound growth rate o 2.5 percent.

However, not all US cities, including middleweights, will be able to keep up withthe rapid growth o many rising cities in emerging markets. Both Beijing andShanghai are expected to overtake Chicago, the third-largest US city, in termso their GDP over the next 15 years. Within the United States, Miami, Florida, isexpected to drop out o the top 30 cities globally measured by GDP.

In the near term, the major challenges acing cities will be deleveraging andpersistent unemployment. In the longer term, a number o broad trends includingaging and the increasing power o ast-growing cities in developing countrieswill continue to shape the opportunities and challenges acing US cities. In theace o multiple potential pressures on their growth and prosperity, cities need to

think about how to make the most o their advantages and how to mitigate theirweaknesses.

26 Projections rom MGIs Cityscope 1.5 database.

3. To regain growth momentum,US cities need to navigatechoppy waters


38/54


39/54


Exhibit 16Several cities with exuberant constructionbooms have experienced the steepest pricedeclines during the downturn

SOURCE: US Census Bureau, Federal Housing Financing Agency, Moodys Analytics, McKinsey Global Institute analysis

-15

-10

-5

0

5

10

15

20

25

30

35

40

45

50

55

60

0 10 2 0 3 0 4 0 5 0 6 0 7 0 80 9 0 10 0 110 12 0 1 30 14 0 1 50 16 0 1 70 18 0 19 0 2 00 21 0 2 20 230 24 0

Minneapolis

RiversidePhoenix

Miami

Price index growth in boom% increase in FHFA 1 house price index, 19952007

Atlanta

Houston

Philadelphia

Dallas

Los Angeles

New York

Houma

Longview

Myrtle Beach

Naples (FL)Cape Coral

Boise City

Austin

Charlotte

Las Vegas

Orlando

Price index decline in downturn% decline in FHFA 1 house price index, 200710

Washington, DC

1 Federal Housing Finance Agency.2 Building activity quartiles defined as cumulative building permits granted between 1995 and 2010 per inhabitant in 1995. A

higher quartile indicates higher building activity during the period.

Bubble cities

Highest quartile

Third quartile

Second quartileLowest quartile

Building activity 2Size of bubble =non-agriculturalGDP, 2010

Sample average

Exhibit 17Unemployment rates vary widely across the United States

SOURCE: US Bureau of Labor Statistics: Household Survey (CPS); Moodys Analytics; McKinsey Global Institute analysis

910%

89%

78%

>10%

67%56%

< 5%

US unemployment, December 2011%


40/54

30

US CITIES WILL NEED TO NEGOTIATE A PATH THROUGHEVOLVING LONGERTERM TRENDS

As well as carving their path through the near-term stresses o the recover y, UScities will need to grapple with a number o key long-term trends that will, to a

greater or lesser extent depending on the city, a ect their growth prospects.

Relying on population as a source of growth will beincreasingly challenging

Demographic trends suggest that expanding through population growth maynot be as easy or cities in the uture as it has been in the past. The US working-age population is not growing as ast as it was, labor mobility is slowing, and thepopulation is aging. Cities that manage to attract incomers and migrants will be ina stronger position or growth than those that dont.

The US working-age population grew at around 1.0 percent per annum over the

past three decades and is expected to expand by around 0.5 percentage pointsannually over the next 15 years. 31 Boosting population growth through attractingmigrants to jobs is thus likely to be harder. There is evidence that the traditionallymobile population o the United States is becoming less so (Exhibit 18). In the1950s and 1960s, one in ve Americans changed residences every year; today,that has dropped to one in ten. Short-distance moves have been alling since1990 and long-distance moves either across county lines within a particular stateor rom one state to another declined sharply in the past decade. And mobilityhas declined among people at a ll levels o education.

31 Historical growth o the working-age population, aged 15 to 64, is based on Moodys

Analy tics, with data rom the US Bureau o the Census and US Bureau o Economic Analysis.It is worth noting that while the US population is aging rapidly, other nations share thisphenomenon. In France, or example, the population aged over 65 is expected to grow3.7 times as ast as the population as a whole. In China, this segment o the population isexpected to grow 3.1 times as ast. In the Uni ted States, the corresponding growth rate is2.7 times.

Exhibit 18

0

2

4

6

8

10

12

14

16

18

20

22

199019801970 119601950

Long-runaverage =18%

20092000

Mobility in the United States has been declining since 1990and is now at a 50-year low

SOURCE: US Bureau of Labor Statistics; McKinsey Global Institute analysis1 19701981 data are interpolated due to data constraints.

Annual domestic migration rate, 19482009% of residents who have changed addresses during the past year

That sharehas droppedto one in ten

In the 1950s and 1960s, onein five Americans changedresidences every year


41/54


US cities are aging and the share o working-age population is declining, andthese actors are likely to prove a signi cant headwind to growth in the economyoverall. One in six people will be aged over 65 by 2025, and the share o working-age adults in the total population is projected to decline rom 67 percent o the

overall population to 63 percent.32

Beyond slower population growth, the demographic composition o many citiespopulations will shi t, and there ore the mix o demand or local services willshi t as well. Today, seniors outnumber children in only one o every 20 citiesby 2025, one-third o cities will have more seniors than children (Exhibit 19). InMiami, Florida, which already has a relatively high share o seniors at 16 percent,this proportion will rise to 23 percent. And in the old industrial city o Pittsburgh,Pennsylvania, the share o people over 65 will rise rom 17 percent to 23 percentby 2025. 33 In contrast, 30 percent o the population o Brownsville and Laredo,both in Texas, is expected to be aged below 15thats twice the share o the

65-plus age group in Brownsville and three times that share in Laredo. In mostcities, policy makers will need to ensure that services rom local school systemsto those used by older citizens adjust to changing demand.

32 For a discussion o the aging o the US baby-boom generation, see Talkin bout my generation: The economic impact of aging baby boomers , McKinsey Global Institute, June2008 (www.mckinsey.com/mgi).

33 For more on the demographic transition in large US metro areas, see The State of Metropolitan America, Brookings Institution, 2010.

Exhibit 19One in 20 US cities has more seniors than childrentodayby 2025, this will be the case in one in three cities

SOURCE: McKinsey Global Institute Cityscope v 1.5

0

5

10

15

20

25

30

35

40

45

0 5 10 15 20 25 30 35

McAllen(TX)

Bradenton (FL)

Punta Gorda (FL)

Jacksonville(NC)

Laredo(TX)

Provo (UT)

BarnstableTown (MA)

Naples (FL)

Children (aged 0-14)

Seniors (aged 65+)

0

5

10

15

20

25

30

35

40

45

0 5 10 15 20 25 30 35

Children (aged 0-14)

Seniors (aged 65+)

Provo (UT)

Punta Gorda (FL)

Laredo(TX)

Jacksonville(NC)

McAllen(TX)

BarnstableTown (MA)

Bradenton (FL) Naples (FL)

% of population

Share of children =share of seniors

Top 30 cities by2010 GDP

2010 2025


42/54

32

The industries fueling growth wil l continue to evolve

Over the past decade leading up to the Great Recession, the nancial and real-estate sectors were major sources o growth or many cities, yet these industriesare likely to take some time to recover. Looking orward, we should expect the

types o industries that thrive and grow the astest in the years ahead to continueto shi t in response to new innovations and consumer trends, as well as broadglobal economic shi ts.

US cities that can take advantage o the rising wave o middle-class urbanites inemerging economies are likely to bene t. The rising economic power o emergingcities in Asia and Latin America is expected to continue to avor Southern andWestern US cities. Companies rom in rastructure suppliers to consumer goodssuppliers are increasingly looking at emerging economies not just or lowerproduction costs but as an alternative to slower-growing consumer markets inthe United States. Urban centers that have good connections or build those

connections to global growth hubswhether physical connections such asairport hubs and ports, business connections (e.g., some electronics valuechains), or personal connections (e.g., cities with universities that attract oreignstudents)will be in a better position to take advantage o the growing emergingmarket opportunity.

With service sectors accounting or more than 80 percent o GDP and nearlyall o net job growth, a strong per ormance in those sectors will continue to becritical or growth in all US cities. We can expect broad-based service sectorssuch as retail, hotels, and restaurants to uel growth as demand returns. 34 Withan aging trend, demand or health care will continue its upward growth trajectory,

as will other services to retirees. Some cities, such as Naples, Florida, havesucceeded in growing by at tracting mobile retirees to their communities. The roleo manu acturing in metropolitan growth is likely to continue to vary widely acrosscities. 35

34 Growth and renewal in the United States: Retooling Americas economic engine , McKinseyGlobal Institute, February 2011 (ww w.mckinsey.com/mgi).

35 The role o manu acturing in the US economy is subject to extensive debate today. See, orexample, Helper, S., et al., Why does manufacturing matter? Which manufacturing matters?

A poli cy framework , Metropolitan Policy Program at the Brookings Institution, Februar y 2012.


43/54


Increasing global competition among cities for investment andskilled workforce

The shi ting economic balance toward emerging markets and the rising pro le o a broader range o middleweight cities will mean more competition or US cities

o all sizes in attracting both inhabitants and businesses. One example o thistrend we are already seeing is that some US baby boomers choose to relocateto places like Mexico to take advantage o the countrys lower cost o living andelderly assistance.

Many cities and states across the United S tates already work hard to attractnew businesses to their jurisdictions. International competition comes not onlyrom rapidly growing emerging markets, with the advantage o their relatively

low wages, but also rom developed markets that are able to create tailoredvalue propositions. The Dutch Innovation Plat orm has brought together thegovernment, key business leaders, and other leading representatives o society to

develop an explicit plan or how to attract 50 signi cant international businessesto the Netherlands. France has an Ambassador or International Investment whoreports annually on jobs created. And or decades now, Singapore and Irelandhave set the bar or e ective agencies attracting oreign direct investment. Bothcountries have built capable organizations that have many o the hallmarks o high-per orming, private-sector sales orces. To succeed in the increasinglycompetitive global arena, cities need to look beyond their US peers and have agood understanding o the speci c priorities among cutting-edge rms in theirtarget sectors and the capacity to o er a competitive business environment orthese businesses.

Recent MGI analysis ound that there could be a shortage o up to 1.5 millionworkers with bachelors degrees or higher in 2020, while nearly 6 million Americans without a high school diploma are likely to be without a job. 36 Demandwill be particularly strong in technical and analytic skillsneeded, or example,to handle the rising volume o big datameaning that cities with young andtechnically educated workers will have an advantage. 37

36 An economy that works: Jo b creati on and Americas future , McKinsey Global Institute, June2011 (www.mckinsey.com/mgi).

37 Big data: The next frontier for innovation, competition, and productivity , McKinsey GlobalInstitute, May 2011 (www.mckinsey.com/mgi).


44/54

34

A CITY SCORECARD: HOW PREPARED ARE US CITIES FOR GROWTH A ND RENEWAL?

Cities that build their strategies on an in ormed understanding o their startingpoint, set targets grounded on that act base, and are e ective in executing

strategy are those that are likely to outper orm their peers over the next 15 years.So what do we know about how prepared cities are or growth?

Each o the top 30 cities that contributed hal o US GDP growth over the pastthree decades aces di erent prospects or short-term recovery and long-termprosperity. In act, the top 30 cities have di erent underly ing strengths andchallenges to uture growth. Our city scorecard maps how these citiescriticaleconomic engines o the past and utureare positioned to address the prevailingshort-term and long-term trends acing the US economy (Exhibit 20).

Exhibit 20Each city will leverage different strengths and facedifferent challenges to regain growth

SOURCE: Moodys Analytics, Core Logic, American Community Survey (US Census Bureau), McKinsey Global Instituteanalysis

1 Reference is US average.2 Reference is average across all MSAs.

Top 30 citiesranked by2010 GDP

Recovery index2010 GDP/pre-crisis peak,200608

Homes innegativeequity, 2011

Unemployment,2011

Net migration,2010Net migration/city population

Educationattainment, 2010People aged 25+with bachelorsdegree or higher

Senior population,2010Populationaged 65+

New York 101 10 8.4 0.1 35.6 13.1

Los Angeles 98 24 11.4 -0.2 30.6 11.1

Chicago 97 25 9.5 -0.2 33.4 11.4

Washington, DC 103 29 5.8 1.2 47.1 10.0

Houston 104 11 8.4 1.0 28.4 8.6

Dallas 100 12 8.1 0.8 30.9 8.8

Philadelphia 100 8 8.6 0.2 32.3 13.3

Boston 101 16 6.8 0.3 42.1 13.1San Francisco 97 10 9.6 0.4 43.5 12.7

Atlanta 96 35 10.0 0.3 34.4 9.0

Miami 94 47 11.0 1.0 28.7 16.0

Seattle 99 16 8.8 0.4 36.9 10.9

Phoenix 93 55 8.4 0.5 28.1 12.3

Minneapolis 100 17 6.5 0.0 37.5 10.7

Detroit 91 42 11.8 -0.8 27.0 13.2

San Diego 98 29 10.0 0.6 34.1 11.4

Baltimore 103 19 7.4 0.2 34.6 12.7

Denver 102 22 8.7 0.9 37.7 10.1

Riverside 92 47 13.7 1.0 19.5 10.4

San Jose 102 18 10.1 0.3 44.5 11.1

St Louis 99 17 9.0 0.0 29.2 13.4

Tampa 94 48 11.0 0.7 25.9 17.3

Portland 98 18 9.2 0.6 33.4 11.4

Pittsburgh 103 6 7.2 0.2 28.3 17.3

Orlando 97 54 10.4 0.7 28.0 12.4

Sacramento 93 42 12.0 0.3 30.0 12.1

Cleveland 96 27 8.0 -0.4 27.0 15.2

Hartford 109 Not available 8.8 0.0 34.2 14.4

Kansas City 98 14 8.4 0.3 32.2 12.0

Cincinnati 99 22 8.9 -0.1 28.5 12.3

Reference 100 1 22.7 1 9.0 1 0.3 2 27.9 1 13.1 1

>20% better than reference

Within 20% of reference

>20% worse than reference%


45/54


35

Large US cities have time and again demonstrated that they collectively have theresilience and capacity to adjust to new situations. There is no reason that thiswill not be the case in coming decades. Although the challenges urban Americaaces appear to be severe, US cities actually have a stronger starting position

than their counterparts in other developed economies. 38 The strong, diverse poolo metropolitan areas in the United States is another advantage.

As in the past, the trump card or urban America as i t navigates toward growthand renewal will be the diversity o strategies and experiences o individualcities. There is no single blueprint or all cities to ollow. So how can all ci tiesseek to make the most o the hands that they have been dealt, and what shouldcompanies be doing to make a contribution?

KEY APPROACHES FOR POLICY MAKERS

Know thysel and tailor strategy accordingly. Cit

mgi urban america full report[1]

Documents