THE WORLD IN NUMBERS

The World Is SpikyGlobalization has changed the economicplaying field, but hasn't leveled it

I POPULATIONUrban areas house half uf all theworlds people, and continue to growin both rich and poor countries.

Tlie world, according to the titleof the New York Times co\uTi\-

nist Thomas Friedman's book., is flat.Thanks to advances in technology, theglobal playing Held has been leveled,tbe prizes are there for the taking, andeveryone's a player—no matter whereon the surface of the earth he or shemay reside. "In a flat world," Friedmanwrites, "you can innovate without hav-ing to emigrate."

Friedman is not alone in this belief:for the better part of the past centuryeconomists have been writing aboutthe leveling effects of technology.P'rom the invention of the telephone,the automobile, and the airplane to therise of the personal computer and theInternet, technological progress hassteadily eroded the economic impor-tance of geographic place—or so theargument goes.

But in partnership with colleaguesat George Mason University and thegeographer Tim Gulden, of the Centerfor International and Security Stud-ies, at the University of Maryland, I'vebegun to chart a very different eco-nomic topography. By almost anymeasure the international economiclandscape is not at al! flat. On the con-trary, our world is amazingly "'spiky."In terms of both sheer economic horse-power and cutting-edge innovation.

PEAKS, HILLS, AND VALLEYSWhen looked at through the lens of economic production, many citieswith large populations are diminished and some nearly vonish. Threesorts of places make up the modern economic landscape. First arethe cities that generate innovations. These are the tallest peaks; theyhave the capacity to attract global talent and create new products andindustries. They are few in number, and difficult to topple. Second arethe economic "hills"—places that monufacture the world's establishedgoods, take its calls, ond support its innovation engines. These hillscan rise and fall quickly; they are prosperous but insecure. Some, likeDublin and Seoul, are growing into innovative, wealthy peaks; othersare declining, eroded by high labor costs and a lack of enduring com-petitive advantage. Finally there are the vast valleys—places with littleconnection to the global economy and few immediate prospects.

surprisingly few regions truly mat-ter in today's global economy. What'smore, the tallest peaks—the cities andregions that drive the world econ-omy—are growing ever higher., whilethe valleys mostly languish.

The most obvious challenge tothe flat-world hypothesis is the

explosive growth of cities worldwide.More and more people are cluster-ing in urban areas—the world's demo-graphic mountain ranges, so to speak.The share of the world's populationliving in urban areas, jnst three per-cent in 1800, was nearly 30 percentby 1950. Today it stands at abont .50percent; in advanced countries threeout of four people live in urban areas.Map A shows the uneven distribution

of the world's population. Five mega-cities currently have more than 20million inhabitants each. Twenty-fourcities have more than 10 million inhab-itants, sixty more than 5 million, and150 more than 2.5 million. Populationdensity is of course a crude indicatorof human and economic activity. But itdoes suggest that at least some of thetectonie forces of economics are concen-trating people and resources, and push-ing up some places more than others.

Still, differences in population den-sity vastly understate tbe spikiness ofthe global economy; the continuingdominance of the world's most produc-tive urban areas is astounding. Whenit comes to actual economic output,the ten largest U.S. metropolitan areascombined are behind only the United

z <

4 8 THE ATLANTIC MONTHLY OCTOBER 2005

LIQHT EMISSIONSEconomic activity—roughly estimated hereusing light-emissions data—is remarkablyconcentrated. Many cities, despite theirlarge populations, barely register.

States as a whole and Japan. New York'seconomy alone is about the size of Rus-sia's or Brazil's, and Chicago's is on apar witb Sweden's. Together New York,Los Angeles, Chicago, and Boston havea bigger economy than all of China. IfU.S. metropolitan areas were countries,they'd make up forty-seven of the big-gest 100 economies in the world.

Unfortunately, no single, compre-hensive information source exists forthe economic production of all theworld's cities. A rough proxy is available,though. Map B shows a variation onthe widely circulated view of the worldat night, with higher concentrations oflight—indicating higher energy use and,presumably, stronger economic produc-tion—appearing in greater relief U.S.regions appear almost Himalayan onthis map. From their summits one mightlook out on a smaller mountain ratigestretching across Europe, some isolatedpeaks in Asia, and a few scattered hillsthroughout the rest of the world.

i

Population and economic activityare both spiky, but it's innovation—theengine of economic growth—that ismost concentrated. The World Intellec-tual Property Organization recordedabout ;iO0,000 patents from residentinventors in more than a hundrednations in 2002 (the most recent yearfor which statistics are available). Nearlytwo thirds of thetn went to Americanand Japanese inventors. Eighty-five per-cent went to the residents of just fivecountries (Japan, the United States,South Korea, Germany, and Russia).

Worldwide patent statistics can besomewhat misleading, since differentcountries follow different standards forgranting patents. But patents grantedin the United States—which receivespatent applications for nearly all majorinnovations worldwide, and holds themto the satne strict standards—tell a simi-lar story. Nearly 90,000 of the 170,000patents granted in the United States in2002 went to Americans. Some 35,000went to Japanese inventors, and lLOOOto Cermans. The next ten most inno-vative countries—including the usualsuspects in Europe plus Taiwan, SouthKorea, Israel, and Canada—producedroughly 25.,000 more. The rest of the

broad, flat world accounted for just fivepercent of all innovations patented inthe United States. In 2003 India gen-erated 341 U.S. patents and China 297.The University of California alone gen-erated more than either country. IBMaccounted for five times as many as thetwo combined.

This is not to say tiiat Indians andChinese arc not innovative. On thecontrary, AnnaLee Saxenian, of theUniversity of California at Berkeley,has shown that Indian and Chineseentrepreneurs founded or co-foundedroughly 30 percent of all Silicon Val-ley startups in the late 1990s. Butthese fundamentally creative peoplehad to travel to Silicon Valley and beabsorbed into its innovative ecosystembefore their ideas became economi-cally viable. Such ecosystems matter,and there aren't many of them.

Map C—which makes use of datafrotn both the World Intellectual Prop-erty Organization and the U.S. Patentand Trademark Office—shows a worldcomposed of innovation peaks and val-leys. Tokyo, Seoul, New York, and SanFrancisco remain the front-runners inthe patenting competition. Boston, Seat-tle, Austin, Toronto, Vancouver, Berlin,Stockholm, Helsinki, Lotidon.. Osaka,Taipei, and Sydney also stand out.

Map D shows the residence of the1,200 most heavily cited scientists inleading fields. Scientific advance iseven more concentrated than patent

THE AGENDA THE ATLANTIC MONTHLY 4 9

0 PATENTSJust a feiu places produce most ofthe world'sinnovations. Innovation remains difficult with-out a critical mass of financiers, entrepreneurs,and scientists, often nourished by world-classuniversities and flexible corporatiom.

THE QEOCRAPHY OF INNOVATIONCommercial innovation and scientific advance are both highlyconcentrated—hut not always in the some places. Several citiesin East Asia-—particularly in Japan—are home to prolific busi-ness innovation but still depend disproportionately on scientificbreakthroughs made elsewhere. Likewise, some cities excelin scientific research but not in commercial adaptation. Thefew places that do both well are very strongly positioned in theglobal economy. These regions hove little to fear, and much fogain, from continuing globalization.

production. Most occurs not just in ahandful of countries but in a handfulof cities—primarily in the United Statesand Europe. Chinese and Indian cit-ies do not even register. As far as globalinnovation is concerned, perhaps a fewdozen places worldwide really competeat the cutting edge.

Concentrations of creative and tal-ented people are particularly

important for innovation, according tothe Nobel Prize-winning economistRobert Lucas. Ideas flow more freely,are honed more sharply, and can beput into practice more quickly whenlarge numbers of innovators, imple-menters, and financial backers arein constant contact with one another,both in and out ofthe office. Creativepeople cluster not simply because theylike to be around one another or theyprefer cosmopohtan centers with lotsof amenities, though both those thingscount. They and their companies alsocluster because of the powerful pro-

ductivity advantages, economies ofscale, and knowledge spillovers suchdensity brings.

So although one might not haveto emigrate to innovate, it certainlyappears that innovation., economicgrowth, and prosperity occur in thoseplaces that attract a critical mass of topcreative talent. Because globalizationhas increased the returns to innovation,by allowing innovative products andservices to qtiickly reach consumersworldwide, it has strengthened the lurethat innovation centers hold for ourplanet's best and brightest, reinforcingthe spikiness of wealth and economicproduction.

The main difference between nowand even a couple of decades ago is notthat the world has become flatter butthat the world's peaks have becomeslightly more dispersed—and that theworld's hills, the industrial and ser-vice centers that produce mature prod-ucts and support innovation centers,have proliferated and shifted. For the

better part ofthe twentieth centurythe United States claimed the lion'sshare ofthe global economy's innova-tion peaks, leaving a few outposts inEurope and Japan. But America hassince lost some of those peaks, as suchindustrial-age powerhouses as Pitts-burgh, St. Louis, and Cleveland haveeroded. At the same time, a number ofregions in Europe. Scandinavia, Canada,and the Pacific Rim have moved up.

The world today looks flat to somebecause the economic and social dis-tances between peaks worldwide havegotten smaller. Connection betweenpeaks has been strengthened by theeasy mobility of the global creativeclass—about 150 million people world-wide. They participate in a globaltechnology system and a global labormarket that allow them to migratefreely among the world's leading cities.In a Brookings Institution study thedemographer Robert Lang and theworld-cities expert Peter Taylor iden-tify a relatively small group of leadingcity-regions—London, New York, Paris,Tokyo, Hong Kong, Singapore, Chi-cago, Los Angeles, and San Franciscoamong them—that are strongly con-nected to one another.

But Lang and Taylor also identify amuch larger group of city-regions thatare far more locally oriented. People inspiky places are often more connected

5 0 THE ATLANTIC MONTHLY OCTOBER 2005

O SCIENTIFIC CITATIONSThe worlds mast prolific and influentialscientific researchers overwhelminglyreside in U.S. and European cities.

7

a <

to one another, even from half a worldaway, tban they are to people andplaces in their veritable back yards.

The flat-world theory is not com-pletely misguided. Il is a welcome

supplemetit to the widely accepted view(illustrated by the Live 8 concerts andBono's forays into Africa, by the writ-ings of Jeffrey Sachs and the UN Mil-lennium project) that the growingdivide between rich and poor countriesis the fundamental feature of the worldeconomy. Friedmati's theory more accu-rately depicts a developing world withcapabilities that translate into economicdevelopment. In his view., for exam-ple, the emerging economies of Indiaand China combitie cost advantages,high-tech skills, and etitrepreneur-ial energy, enabling those countries tocompete effectively for indnstries andjobs. The tensions set in motion as theplaying field is leveled affect mainly tbeadvanced countries, which see not onlymanttfacturing work but also higher-end jobs, in fields such as softwaredevelopment and financial services,increasingly threatened by offshoring.

But the flat-world theory blinds usto far more insidious tensions amongthe world's growing peaks, sinking val-leys, and shiftitig hills. The innovative,talent-attracting "have" regions seemincreasingly remote from the talent-exporting "have-not" regions. Second-tier cities, from Detroit and Wolfsburgto Nagoya and Mexico City, are enter-ing an escalating and potentially devas-tating competition for jobs, talent, andinvestment. And inequality is growingacross the world and within countries.

This is far more harrowing thanthe flat world Friedman describes, anda good deal more treacherous than theold rich-poor divide, We see its effectsin the political backlash against glo-balization in the advanced world. Therecent rejection of the EU constittitionby the French, for example, resulted inlarge part from high rates of "no" votesin suburban and rural quarters., whichunderstandably fear globalization andintegration.

But spiky globalization also wreakshavoc on poorer places. China is see-ing enormous concentrations of tal-ent and innovation in centers such asShanghai, Shenzhen, and Beijing, allof which are a world apart from its vast,

impoverished rural areas. Accordingto detailed polling by Richard Burk-holder, of Gallup, average householdincomes in urban China are now tri-ple those in rural regions, and they'vegrown more than three times as fastsince 1999: perhaps as a result, urbanand rural Chinese now have very dif-ferent, often conflicting political andlifestyle values. India is growing evenmore divided, as Bangalore, Hyder-abad, and parts of New Delhi andBombay pull away from the rest of thatenormous country, creating destabi-lizing political tensiotis. Economic anddemographic forces are sorting peoplearound the world into geographicallyclustered "tribes" so different (andoften mutually antagonistic) as to cre-ate a somewhat Hohbesian vision.

We are thus confronted with a dif-ficult predicament. Economic progressrequires that the peaks grow strongerand taller. But such growth will exac-erbate economic and social dispari-ties., fomenting political reactions thatcould threaten further irniovation andeconomic progress. Managing the dis-parities between peaks and valleysworldwide—raising the valleys withoutshearing off the peaks—will be amongthe top political challenges of the com-ing decades. —RiCHAHD FLORIDA

Richard Florida, the author of The Flighl of theCreative ClasiS, is the Hirsr Professor of Public Miiyat George Mason University.

THE AGENDA THE ATLANTIC MONTHLY 5 1