8/11/2019 A Unified Theory
1/2
At the start of the twenty-first century,cities emerged as the
source of thegreatest challenges that the planet
has faced since humans became social.Although they have proven
to be human-itys engines of creativity, wealth creationand economic
growth, cities have also beenthe source of much pollution and
disease.Rapid urbanization and accelerating socio-economic
development have generatedglobal problems from climate change
andits environmental impacts to incipientcrises in food, energy and
water availabil-ity, public health, financial markets and theglobal
economy 1,2.
Urbanization is a relatively new globalissue. As recently as
1950, only 30% of theworlds population was urbanized. Today,more
than half live in urban centres. Thedeveloped world is now about
80% urbanand this is expected to be true for the entireplanet by
around 2050, with some 2 billionpeople moving to cities, especially
in China,India, southeast Asia and Africa 2.
Cities are complex systems whose infra-structural, economic and
social componentsare strongly interrelated and therefore dif-ficult
to understand in isolation 3. The manyproblems associated with
urban growth andglobal sustainability, however, are
typicallytreated as independent issues. This frequentlyresults in
ineffective policy and often leads to
unfortunate and sometimes disastrous unin-tended consequences.
Policies meant to con-trol population movements and the spread
ofslums in megacities, or to reverse urban decay,have largely
proven ineffective or counterpro-ductive, despite huge
expenditure.
In New York City in the 1970s, for example,a strategy of planned
shrinkage intentionallyremoved essential services from some
urbanareas notably the Bronx to prompt peo-ple to move away and
allow for redevelopment.Instead, this strategy led to increases in
crimeand general socio-economic degradation.In North America in the
1950s to 1970s (andearlier in Europe), policies of urban
renewalintended to reduce high urban densities,by razing poorer old
neighbourhoods andcreating infrastructure, actually ended
upencouraging urban sprawl 3. Similar debatescontinue to play out
in rapidly develop-ing cities around the world today, fromBeijing
to Rio de Janeiro in Brazil, often lead-ing to similar
mistakes.
But cities supply solutions as well asproblems, as they are the
worlds centres ofcreativity, power and wealth. So the need isurgent
for an integrated, quantitative, pre-dictive, science-based
understanding of thedynamics, growth and organization of cit-ies.
To combat the multiple threats facinghumanity, a grand unified
theory of sus-tainability with cities and urbanization at its
core must be developed. Such an ambitiousprogramme requires
major internationalcommitment and dedicated transdiscipli-nary
collaboration across science, economicsand technology, including
business leadersand practitioners, such as planners anddesigners.
Developing a predictive frame-work applicable to cities around the
world
is a daunting task, given their extraordi-nary complexity and
diversity. However, weare strongly encouraged that this mightbe
possible.
UNIVERSAL FEATURES
Cities manifest remarkably universal, quan-tifiable features.
This is shown by new analy-ses of large urban data sets, spanning
severaldecades and hundreds of urban centres inregions and
countries around the worldfrom the United States and Europe to
Chinaand Brazil 4,5. Surprisingly, size is the majordeterminant of
most characteristics of a city;
history, geography and design have second-ary roles4,6.Three
main characteristics vary system-
atically with population. One, the spacerequired per capita
shrinks, thanks todenser settlement and a more intense useof
infrastructure. Two, the pace of all socio- economic activity
accelerates, leading tohigher productivity. And three, economicand
social activities diversify and becomemore interdependent,
resulting in newforms of economic specialization and cul-tural
expression.
We have recently shown that these generaltrends can be expressed
as simple math-ematical laws. For example, doubling thepopulation
of any city requires only aboutan 85% increase in infrastructure,
whetherthat be total road surface, length of electricalcables,
water pipes or number of petrol sta-tions 4. This systematic 15%
savings happensbecause, in general, creating and operatingthe same
infrastructure at higher densitiesis more efficient, more
economically viable,and often leads to higher-quality servicesand
solutions that are impossible in smallerplaces. Interestingly,
there are similar savingsin carbon footprints 7,8 most large,
devel-oped cities are greener than their nationalaverage in terms
of per capita carbon emis-sions. It is as yet unclear whether this
is alsotrue for cities undergoing extremely rapiddevelopment, as in
China or India, wheredata are poor or lacking.
Similar economies of scale are found inorganisms and communities
like anthillsand beehives, where the savings are closerto 20%9.
Such regularities originate in themathematical properties of the
multiple
A unified theoryof urban livingIt is time for a science of how
city growth affects society
and environment, say Luis Bettencourt and Geoffrey West .
O L I V E R M U N D A Y
Full content and enhanced
graphics at: nature.com/cities
SCIENCE AND THE CI
9 1 2 | N A T U R E | V O L 4 6 7 | 2 1 O C T O B E R 2 0 1
0
COMMENT
20 Macmillan Publishers Limited. All rights reserved10
8/11/2019 A Unified Theory
2/2
networks that sustain life, from the cardio- vascular to the
intracellular 9. This suggeststhat similar network dynamics
underlieeconomies of scale in cities.
Cities, however, are much more than giantorganisms or anthills:
they rely on long-range, complex exchanges of people, goodsand
knowledge. They are invariably magnets
for creative and innovative individuals, andstimulants for
economic growth, wealthproduction and new ideas none of whichhave
analogues in biology.
The bigger the city, the more the aver-age citizen owns,
produces and consumes,whether goods, resources or ideas 4. On
aver-age, as city size increases, per capitasocio-economic
quantities such aswages, GDP, number of patents pro-duced and
number of educationaland research institutions all increaseby
approximately 15% more than theexpected linear growth 4. There
is,
however, a dark side: negative met-rics including crime, traffic
conges-tion and incidence of certain diseasesall increase following
the same 15%rule4. The good, the bad and the uglycome as an
integrated, predictable,package.
Our work shows that, despiteappearances, cities are
approximatelyscaled versions of one another (seegraph): New York
and Tokyo are, toa surprising and predictable degree,nonlinearly
scaled-up versions of San Fran-cisco in California or Nagoya in
Japan. Theseextraordinary regularities open a window onunderlying
mechanism, dynamics and struc-ture common to all cities.
Deviations from these scaling laws, illus-trated by the spread
of data in the figure,measure how each city over- or
under-per-forms relative to expectations for its size 6.Relatively
large deviations (as much as 30%)are seen for quantities with small
numbers,such as patents and murders, whereas muchsmaller deviations
(with variances less than10%) are seen for economic properties.
Wealso find that quantities such as GDP aremore variable for urban
centres in developingcountries, such as China and Brazil, than
forolder cities in developed areas such as NorthAmerica or Japan.
It is unclear whether thisis a fundamental property of
developingnations or an artefact of data collection.
In biology, the network principles under-lying economies of
scale have two profoundconsequences. They constrain both the paceof
life (big mammals live longer, evolve slower,and have slower heart
rates, all to the samedegree 9), and the limits of growth
(animalsgenerally reach a stable size at maturity 10). Incontrast,
cities are driven by social interac-tions whose feedback mechanisms
lead tothe opposite behaviour. The pace of urban lifesystematically
increases with each expansion
of population size: diseases spread faster,businesses are born
and die more often andpeople even walk faster in larger cities, all
byapproximately the same 15% rule 4. More-over, this social network
dynamic allows thegrowth of cities to be unbounded:
continuousadaptation, not equilibrium, is the rule.
Open-ended growth is the primary
assumption upon which modern cities andeconomies are based.
Sustaining that growthwith limited resources requires that
majorinnovations such as those historically asso-ciated with iron,
coal and digital technology be made at a continuously accelerating
rate.The time between the Computer Age and
the Information and Digital Age was some20 years, compared to
thousands of yearsbetween the Stone, Bronze and Iron Ages.Making
major technological paradigm shiftssystematically faster is clearly
not sustainable,potentially leading to collapse of the
entireurbanized socio-economic fabric. Avoidingthis requires
understanding whether we cancontinue to innovate and create wealth
with-out continuous growth and its compoundednegative social and
environmental impacts.
ACTING ON EVIDENCE
The job of policy-makers is to enhance theperformance of their
city relative to base-lines for their size defined by scaling
laws.Although a scientific understanding of howcities work may not
be prescriptive for pol-icy-makers, recent work should help them
toencourage positive urban development.
Our research shows that cities are remark-ably robust: success,
once achieved, is sus-tained for several decades or longer 6,
therebysetting a city on a long run of creativity andprosperity. A
great example of success ismetropolitan San Jose, home to the
SiliconValley, which has been consistently over- performing
relative to expectations for its sizefor at least 50 years, well
before the advent ofmodern hi-tech industry. Unfortunately,
thereverse is also true: it is very hard to turnaround urban decay
swiftly. Ineffective
policy and unrealistic short-term expec-tations can condemn a
city to decadesof under-performance: witness formerindustrial
cities such as Buffalo, New York.
Todays rapid development and urbani-zation provides an
opportunity to collectdetailed data that will illuminate the
linksbetween economic development and its
undesirable consequences. Policy initiativesin developed and
developing cities shouldbe viewed as experiments that, if
carefullydesigned and measured, can help supportthe creation of an
integrated, predictive the-ory and a new science of
performance-basedplanning. Examples of this approach are
increasingly common, both amongposter children such as Barcelona
inSpain or Curitiba in Brazil, and as partof new initiatives in New
York or Lon-don. Ideally, by coupling general goals(such as lower
carbon emissions) toactionable policies and measurable
indicators of social satisfaction, suc-cesses and failures can
be assessed andcorrected for, guiding development oftheory and
creating better solutions.
Cities are the crucible of humancivilization, the drivers
towardspotential disaster, and the sourceof the solution to
humanitys prob-lems. It is therefore crucial that weunderstand
their dynamics, growthand evolution in a scientifically
pre-dictable, quantitative way. The dif-
ference between policy as usual and policyled by a new
quantitative understanding ofcities may well be the choice between
creat-ing a planet of slums or finally achievinga sustainable,
creative, prosperous, urban-ized world expressing the best of the
humanspirit.
Luis Bettencourt is a scientist at Los Alamos National
Laboratory and external professor at the Santa Fe Institute.
GeoffreyWest is distinguished professor at the SantaFe Institute
and senior fellow at Los AlamosNational Laboratory.e-mail:
gbw@santafe-edu
1. Schellnhuber, H. J., Molina, M., Stern, N., Huber,
V. & Kadner, S. (eds) Global Sustainability: A NobelCause
(Cambridge Univ. Press, 2010).2. UN-Habitat. State of the Worlds
Cities 2010/2011
Cities for All: Bridging the Urban Divide (2010);available at
http://www.unhabitat.org
3. Jacobs, J. The Death and Life of Great AmericanCities (Random
House, 1961).
4. Bettencourt, L. M. A., Lobo, J., Helbing, D.,Khnert, C. &
West, G. B. Proc. Natl Acad. Sci. USA104, 73017306 (2007).
5. Batty, M. Science 319, 769771 (2008).6. Bettencourt, L. M.
A., Lobo, J., Strumsky, D. &
West, G. B. PLoS ONE (in the press).7. Brown, M. A., Southworth,
F. & Sarzynski, A. Policy
Soc. 27, 285304 (2009).8. Dodman, D. Environ. Urban. 21, 185201
(2009).9. West, G. B., Enquist, B. J. & Brown, J. H.
Science
276, 122126 (1997).10. West, G. B., Brown, J. H. & Enquist,
B. J. Nature
413, 628631 (2001).
Data from 360 US metropolitan areas show that metrics such
aswages and crime scale in the same way with population size.
PREDICTABLE CITIES
2.0 1.5 1.0 0.5 0 0.5 1.0 1.5 2.0
log (city population/city population average)
l o g
( m e t r i c / m e t r i c a v e r a g e )
2.0
1.5
1.0
0.5
0
0.5
1.0
1.5
2.0
2.5
Income
CrimeMETRIC:
Patents
GDP
L B E T T E N C O U R T ; N
R O D R I G U E Z ; G W
E S T
2 1 O C T O B E R 2 0 1 0 | V O L 4 6 7 | N A T U R E | 9 1
3
COMMENT
20 Macmillan Publishers Limited. All rights reserved10
