The Blind Men and the Elephant: The Explanation of Gentrification

Author(s): Chris Hamnett

Source: Transactions of the Institute of British Geographers, New Series, Vol. 16, No. 2 (1991),
pp. 173-189
Published by: The Royal Geographical Society (with the Institute of British Geographers)
Stable URL: http://www.jstor.org/stable/622612 .
Accessed: 20/02/2014 09:57

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

The Royal Geographical Society (with the Institute of British Geographers) is collaborating with JSTOR to
digitize, preserve and extend access to Transactions of the Institute of British Geographers.

http://www.jstor.org

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 173

theexplanation
Theblindmenandtheelephant:
ofgentrification
CHRIS HAMNETT
inGeography,
SeniorLecturer Faculty WaltonHall,
TheOpenUniversity,
ofSocialSciences,
MiltonKeynesMK7 6AA

2 January,
MS received
Revised 1991

ABSTRACT
reviews
Thispapercritically themajortheoriesofgentrification
which haveemerged overthelast10yearsandthedebate
whichhassurrounded them. debatehasattracted
Itarguesthatthereasonwhythegentrification andhas
so muchinterest,
beenso hardfought,isthatitisoneofkeytheoretical ofcontemporary
battlegrounds human geography which highlights
theargumentsbetween andagency,
structure andconsumption,
production capitalandculture,
andsupplyanddemand. It
alsoarguesthateachofthetwomajorexplanations whichhavebeenadvanced to account (therentgap
forgentrification
andtheproduction arepartial
ofgentrifiers) eachofwhich
explanations, isnecessarybutnotsufficient. that
itargues
Finally,
anintegrated
explanation mustinvolve
forgentrification bothexplanationoftheproductionofdevaluedareasandhousing
andtheproduction andtheir
ofgentrifiers consumption
specific andreproductionpatterns.

KEYWORDS:Gentrification, Production,
Theory, Residential
Housing,
Consumption, change

INTRODUCTION novel and potentially interestingcity-specific

researchtopics.Hence the largenumberof one-off,
The gentrification phenomenon, and thedebateover locallybased case-studies.
its significance, processes,explanationand effects A second,andmoreconvincingexplanationis that
have occupieda remarkably largeamountofspace in gentrificationhas posed a major challengeto the
the scholarlyjournalsover the last 10 years (see traditionaltheoriesof residential locationand urban
Hamnett,1984; Smithand Williams,1986 forrecent social structure(Hamnett,1984). Neighbourhood
bibliographies). In theAnnalsalone,therehave been changewas viewedbyHoytandBurgessas a one-way
articlesby Ley(1980; 1986; 1987),Schaffer andSmith process where 'the wealthy seldom reverse their
(1986),Smith(1987b) and Badcock(1989). stepsand move backwardsintotheobsoletehousing
Gentrification has now been identified in a large which they are giving up' (Hoyt, 1939, p. 118).
numberof cities in North America,Europe and Gentrificationundermines thedominantassumption
Australia,butdespiteitsexpansionduringthe1970s thatfilteringis a uni-directionaldownwardsprocess
and 1980s, it is stilla relativelysmallscale and very in whichlowerincomegroupsmove intoprogress-
geographically-concentrated phenomenoncompared ively deterioratedhousing,and it challengesthe
to post-warsuburbanization and innercitydecline. explicitassumptionunderlying Alonso's 'structural'
Berry(1985) dismissivelyrefersto it as Islandsof theoryof theurbanland marketthatthe preference
renewalin seas ofdecay.It is therefore importantto forspace and low densitiesare farmore important
ask why so muchattentionhas been devotedto the thanaccessibility to the centralcity.Finally,gentri-
subject.At least fivepossible explanationscan be ficationundermines existing'stage theories'or evo-
identified.These are outlinedin ascendingorderof lutionarymodels of urbanresidentialchangewhich
importance.First,and somewhatinstrumentally, it see middleclass suburbanization as the finalstage
can be suggestedthatgentrification has provideda of a progressionfrom the pre-industrial to the
convenientsubjectfora new generationof urban industrialcity.Ley (1981) has commentedthatas a
geographersand sociologists on the lookout for resultof:

Trans.Inst.Br. Geogr.N.S. 16: 173-189 (1991) ISSN: 0020-2754 Printedin Great Britain

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
174 CHRIS HAMNETT
therevitalizationprocessofthepastdecade,sections of The fifth, and arguably,the most importantex-
thepost-industrialinnercityhavebeguna transformation planationfortheprominence ofgentrification incon-
fromthehomesoflabouring classestowarda zoneof urban is that it
temporary geographical literature,
privilegereminiscent oftheinner-most residential
ringin
represents one of key theoreticaland ideological
Sjoberg'smodelofthepre-industrial city. Ifpresent
trends
continue,thesocialgeography ofthenineteenth-century battlegroundsin urban geography,and indeed in
industrial human geographyas a whole, betweenthe liberal
citymayevenappeartourbanscholars ofthe
futureas a temporary interludeto a more humanists who stressthekeyroleof choice,culture,
historically
persistentpattern ofhigher-statussegregation adjacent consumptionand consumerdemand,and the struc-
tothedowntown core(Ley,1981,p. 145). turalMarxistswho stressthe role of capital,class,
productionand supply.Gentrification is one of the
The thirdreason forthe emergenceof gentrifi- main arenasof conflictbetweenthe proponentsof
cationas a centralresearchissueliesinthepolicyand culture,preference and humanagency,and thepro-
politicaldebates dis-
regardinggentrification-related ponents of the imperatives ofcapitalandprofitability.
placement.Whereasgentrification has been seen by Indeed,two ofthemajorcombatants, David Leyand
someas thesaviouroftheinnercities,heralding a halt Neil Smithhave been closelyengaged in widerde-
to decadesofwhitemiddleclassflight and residential bates aboutepistemologyand explanationin human
abandonmentand offeringan increasedtax base geographyas a whole[seeDuncanand Ley,1982 and
(Sumka,1979; Sternlieband Hughes, 1983) others Smith,1982; 1987c).
regardit as a threatto innercityworkingclass areas To the extentthatthisinterpretation is correct,
(Ley, 1981; Hartman, 1979; Marcuse, 1986; LeGates gentrification is a frontier (Smith, 1986) not just
and Hartman,1986) and a preludeto the wholesale physically, economically, sociallyand culturally, but
conversionofpartsoftheinnercityintoa bourgeois also theoretically, ideologicallyand politically.It
playground(Schaeffer and Smith,1986). comprisesa contestedboundaryzone betweenradi-
A fourth, and relatedexplanation,is thatgentrifi- cally different theoriesand explanations.And it is
cation can be seen to constituteone of the major arguably this aspect of gentrification, above all
'leading edges' of contemporarymetropolitanre- others,whichhas keptthegentrification debateat the
structuring. Justas suburbanization and innercity forefront of urbangeographicalliterature forover a
declinecomprisedtheleadingedges ofurbanrestruc- decade. The gentrification debate is one played for
turingin the 1950s and 1960s, so gentrification is high theoretical and ideologicalstakes.Not surpris-
arguedto represent one oftheleadingedges ofurban ingly,it has also been fiercelycontested,withthe
restructuring in the 1970s and 1980s. By slowingor proponentsofproductionand profitability snipingat
reversinginnercitymiddleclasspopulationloss and the advocates of consumptionand choice and vice
housing decay, gentrification representsa partial versa.As Schaeffer and Smith(1986) clearlystated:
reversalof previoustrends.Fromthis perspective,
gentrification, like suburbanization beforeit, high- thedebateovercauseshascometocenter ontheissueof
lightsthe importanceof capitalswitchingbetween production basedvs. consumption basedexplanations
differentsectorsoftheeconomyanddifferent partsof .... Eachof thedifferent positionsin thisdebate...
thecity(Smith,1979; Harvey,1978; 1980; Badcock, involves a larger theoretical
commitment concerning the
1989; King,1989a, 1989b, 1989c). This argumentis way inwhichurbanspaceis continually patterned and
developed by Smith and Williams (1986) who repatterned (Schaeffer andSmith, 1986,p. 350).
suggest,among other things,that gentrification has
to be seenas partofthechanginginternational spatial And Rose (1984) notesin hersympathetic critiqueof
divisionoflabour,and theemergenceofglobalcities Marxistanalysesof that:
gentrification
withcontroland commandfunctions as partofa new
urbanheirarchy dominatedby flowsof financecapi-
Marxistworkon gentrification has insistedthatthe
tal.Thisis leadingto a restructuring ofboththeurban
place
'correct' to begintheorising aboutthisprocessis
heirarchyand of intra-urban space. Schaefferand withtheproduction of thecommodities of gentrified
Smith(1986) thusrejectthe claimsof 'minimalists' I usetheword'correct' to drawattention to
dwellings.
suchas Berrywho see gentrification as a smallscale thefactthatthistypeofinsistence ona singleanalytical
process.They argue that:'we are witnessingnot a startingpointinthe'sphere ofproduction' is politically
curious anomaly but a trenchantrestructuring of grounded inthissphere
. . . activities are... (seenas) ...
urbanspace' (Schaeffer and Smith,1986,p. 362). theprimary motorsofchangewithin capitalistsociety

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblind menandtheelephant 175
.... To a large extent,Marxist to
approaches gentrifi-argue that the role of the state is important an for
cationhavedefined theirobjectives self-consciouslyin understanding ofgentrification in certainareas,there
opposition topositivist approaches...A crucialelement is considerabledebateovertherelative
importance to
in the Marxistapproach,in contrast, to positivist be to individualactorsand theirmotivations
is thatit seesthatgentrificationis ... not given
approaches, andto thestructural roleofthestate.In somerespects
reductible to thebehaviour ofindividuals (Rose,1984,
therefore, the debate over the role of the state in
pp.49-50).
gentrification reflectsand embodiesthewidergentri-
ficationdebatebetweentheproponentsof structure
Itis thisaspectofgentrification, thatofintellectual and agency(Gregory,1981).
battlegroundbetween competing and radically The paper is dividedintonine sections.The first
opposed theoreticalperspectives,that I intendto sectiondefinesgentrification and outlinesthecriteria
focuson in thisarticle.Althoughseveralalternative forexplanation.The second and thirdsectionsout-
explanatory emphaseshavebeenidentified (Hamnett, line and assess Ley's approach.The fourthand fifth
1984; Ley, 1986; Smith,1986), notably those of sectionsoutlineand assess Smith'sinitial'rentgap'
changesindemography, andurbanamenity; thesis.The sixthsectionstressesthe importanceof
life-style
landandhousingmarket dynamicsandinurbanecon- the 'productionof gentrifiers' and theirlocational
omicactivityand employment structures,in essence preferences, the seventhsection examinesSmith's
they collapse into two main competingsets of reformulations and his attemptto incorporatecon-
explanations.The first, primarily associatedwiththe sumptioninto his theoreticalframework, and the
workof Smithhas stressedtheproductionof urban eighthoutlinestheelementsofan integrated theory.
space,theoperationof thehousingand landmarket, The final section summarizesand concludes the
theroleofcapitaland collectivesocialactorssuchas argument.
developersand mortgagefinanceinstitutions on the
supplyof gentrifiable property.The second,which
Smithhas termedthe consumptionside argument, GENTRIFICATION: A DEFINITION AND
focuseson the productionof gentrifiers and their CRITERIA FOR EXPLANATION
associated cultural,consumptionand reproductive
As a preliminary to the outlineand analysisof the
orientations (Ley,1980; 1981; Mullins,1982; Moore,
competingarguments, we firstneedto definegentrifi-
1982; Rose,1984;Williams,1984;Beauregard, 1986).
cation,and establishthecriteria fora comprehensive
This paper arguesthatboth of the two principal
explanation, against whichvarioustheoriescan be
theoreticalperspectiveson gentrification are partial
assessed and evaluated. Hamnett (1984, p. 284)
abstractionsfromthe totalityof the phenomenon, defined as:
and have focusedon different gentrification
aspectsto theneglect
of other,equallycrucialelements.LikeAesop's fable Simultaneously a physical,economic, socialandcultural
oftheblindmenand theelephant,each of themajor phenomenon. Gentrificationcommonly involvesthe
theorieshas perceivedonly partof the elephantof invasionby middle-class or higher-income groupsof
gentrification. The two theoretical perspectives are previously working-class neighbourhoods or multi-
ratherthancompeting.Thishas sub- occupied 'twilight areas'andthereplacement ordisplace-
complementary
mentofmanyoftheoriginal occupants. Itinvolvesthe
sequentlybeen slowly appreciated,and the initial
tendencies have been watered down to physicalrenovation or rehabilitation
ofwhatwas fre-
exclusionary
someextent.The gradualemergenceofan integrated quentlya highlydeteriorated housingstockand its
upgrading to meettherequirements ofitsnewowners.
theoryofgentrification (Hamnett,1984; Beauregard, Intheprocess, housing intheareasaffected, bothreno-
1986) has arisenfromtherealizationthatproduction vatedand unrenovated, undergoes a significant price
andconsumption arebothcrucialto a comprehensive appreciation. Sucha process ofneighbourhood transition
explanation. commonly involvesa degreeof tenuretransformation
In arguingthisthesis,onlylimitedattention is paid from renting toowning.
to the debatesover the role of the statein gentrifi-
cationand to the genderdimensionsof theprocess Smith(1987b,p. 463) stated:
(Rose, 1984; 1989) but it is contendedthat,while Thecrucial pointaboutgentrification is thatitinvolves
important,these are essentiallysecondaryto the notonlya socialchangebutalso,attheneighbourhood
central issue of productionversus consumption. scale,a physicalchangein thehousingstockand an
AlthoughCybriwskyet al. (1986) and Smith(1989) economic changein thelandandhousingmarket. It is

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
176 CHRIS HAMNETT
this combinationof social,physical,and economic thenewliberalism was tobe recognised
lessbyitspro-
changethatdistinguishes as anidentifiable ductionschedules
gentrification thanbyitsconsumption styles(Ley,
ofprocesses.
process/set 1980, p. 239).

It is clearfromthesedefinitions thatgentrificationLeyalso
arguedthat:'an understanding oftheemerg-
involvesbotha changein the social compositionof
ing urbanlandscaperequiresa priorgraspof wide-
an areaand itsresidents, anda changeinthenatureof
rangingprocessesof change in societyitself'(Ley,
thehousingstock(tenure, price,conditionetc.)andan 1980, p. 240). In an attemptto identifythese,Ley
adequate explanationof gentrification will have to drew on the workof Daniel Bell on
post-industrial
coverbothaspectsoftheprocess;thehousingandthe
societyand Habermason advanced capitalism.He
residents. Movingfromthesedefinitions to thespeci-
acceptedtheywere unlikelytheoretical bedfellows,
ficationof the criteriafor explanationof gentrifi- but arguedtherewas: 'a deepercomplementarity in
cation,I suggestthatanycomprehensive explanation theirpositions.Bothsee a decisivetransition between
ofgentrification mustexplainfourkeyaspectsofthe nineteenthand late twentiethcentury society,
process.First,why gentrification is particularly
con- betweenthe industrial
period(earlycapitalism)and
centratedin a smallnumberof large citiessuch as (latecapitalism) .... (p.240).
post-industrialism
Paris,London,New York,San Francisco,Toronto, Ley'sthesisinvolvedthreekeypropositions focus-
SydneyandMelbourne(andwhyitis morelimitedin ing respectively on economics,politicsand culture.
older industrial cities).Secondly,why gentrificationAs we shallsee later,theorderis important. First,at
occursin someareasand housingand notothers,and the level of the
economy,the decliningrole of
the characteristics of the areas involved.Thirdly,it unskilledlabour in the
productionprocess and the
mustexplainwhichgroupsbecome gentrifiers and growing of technologyin thefactory, in
importance
why, and fourthly, it must explain the timingof theofficeand in administration is a majorbreakwith
gentrification. In other words, a comprehensive thenineteenth Thishasbeen associatedwith
century.
explanationmust address the questionsof where, a major transformation of the labourforce,with a
whichareas,who,whenand why. declinein blue collarworkersand a growthof white
I shallargueintheremainder ofthispaperthateach collarworkers, in theprofessional, man-
particularly
ofthemajorexplanationsaddressesor answerssome and technical
agerial,administrative occupations.Ley
of thesequestions,but not others.Indeed,it can be linkedthisto the shiftfroma
goods producingto a
arguedthatwhileeachis ofconsiderable explanatory service-producing society, and to the decline of
value,theyareincapablein isolationofansweringall manufacturing industry and theriseofofficework.
these questionsby virtueof theirfocusand range. The second propositionwas thatpost-industrial
Thereforethey constitutepartialexplanationsof
societyis distinguished fromindustrial societyby the
limitedvalidity.As Smith(1979) is inparta response activeroleof As a consequenceof this,
government.
to previouswork by Ley (1978) and others,Ley's
Ley arguedthat'decisionmakingand allocationof
workwillbe consideredfirst. resourcesis now referred to the politicalarenaand
notonlyto themarketplace .... The politicization of
varied interest groups is challenging the formerly
CULTURE AND CONSUMPTION IN THE
firmhold of thebusinesslobbyon politicaldecision
POST-INDUSTRIAL CITY
making'(Ley,1980,p. 241). Thirdly, Leyarguedthat
In 1980 David Leypublished'Liberalideologyandthe at thesociocultural leveltherehas beena re-assertion
post-industrial city'.In thispaperhe setoutwhatcan of the role of individuality and a growthof a more
be seeninretrospect to be a keytheoreticalstatement sensuousand aestheticphilosophyamongthegrow-
regardingthe originsand causes of gentrificationing numbersof the North Americanserviceclass,
althoughitsfocuswas ontheriseoftheElectorsAction particularly on theWestcoast.He concludedthat:
MovementinVancouverand theirpolicyoflimiting
realestateand freewaydevelopmentand creatinga we maysee fromthisframework theappearance ofa
livablecity.In whatlaterprovedto be a red rag to theoretically groupofactors... (who)form
significant a
Marxistanalysts,Ley arguedthat: theoreticalcounterpoint to nineteenth century notions
ofcapitalandlabor... a classinemergence .... Witha
A new ideologyof urbandevelopment was in the secureeconomicbase,theyrepresent thepresent day
making. Urbanstrategy seemedto be passingfroman counterparts of Veblen'sleisureclass,displaying the
emphasis ongrowth toa concernwiththequality oflife; canonsofgood taste,intent upontheaesthetic. Their

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 menandtheelephant
Theblind 177
is ... consumption
lifestyle in
and statusorientated local housingmarket,quickening the transition pro-
ofself-actualization
pursuit '(Ley,1980,pp.242-3). cess and fuellinginflationary land values, through
speculationand by increasing the expectationsof
Ley'sreference to 'a class in emergence'is import- homeownersto receive windfallprices for their
ant,andhe notedthatas thepost-industrial thesiswas homes'(Ley,1981,p. 138,emphasesadded).
developed by sociologists it was not locationally The causalprimacyis quiteclear.Leysees property
specific.But he argued that 'these traitsare not activityas stimulatedby the marketpower of the
uniformly distributed; thereis a geographyof the growingwhitecollarlabourforce,whichis a product
post-industrialsociety... it mightfitcircumstances of changesin economicand employmentstructure.
more closely in San Franciscoor London than in He has reiterated thisview in a morerecent(1986)
ClevelandorGlasgow'(Ley,1980,pp. 242-3). Thisis paper.As he putit:
a key pointwhichhas an important bearingon the
question of where gentrification is found and Ley
job growth(in)thewhite-collar complexofdowntown
proceeded to apply the thesis to Vancouver, looking headoffices, andindirectly,
services, (in)public
producer
at changes in industrial,occupationaland demo- andagencies
institutions in... nodalcentres...leadsto
graphicstructures and in thelifestyles and innercity the'production' of professionals,managers and other
housing market which had occurred. Ley did not quaternary employees working downtown, who then
referto thetermgentrification
explicitly inthispaper, providethedemandbaseforhousingre-investment in
in
but 1981 he made a clearlink betweenthegrowth theinner
city ... this
population, as itgives
politicaland
ofthetertiary and quaternary sectors,thegrowthof economicexpression to itsownpredeliction to urban
and managerialoccupations,changesin amenity, will restructure the built environment and
professional accelerate
the 1986,
of housingdemandin Vancouver and gentrificationprocess (Ley, p. 532).
the structure
As he putit:
gentrification.

it is possibleto followthetransmission of largescale AN ASSESSMENT OF LEY'S THEORY OF

adjustments intheeconomy tothepatternofjobcreation POST-INDUSTRIAL URBANISM
in Vancouver, withtrendsfavouring whitecollarjob
growthin thecentral ThesecontextualThereis muchin Ley's thesisthatMarxistanalysts
businessdistrict.
factors liebehind thedemographic changes inthemetro- would stronglychallenge,not least the political
politanareaand thehousingdemandpressures which emphasishe accords to a new elite of tastemakers
them
accompanied (Ley,1981,p. 128 emphases added). and opinionformers, the importanceof cultureand
consumption,his acceptanceof the idea of post-
Butthesehousingdemandpressuresare locationally industrialism (Walkerand Greenberg,1982),and his
specific.Discussing the growing number of small, seemingrelegationof the productionof the built
young,highincomehouseholdsand theirimpacton environmentand nineteenth-century notions of
theinnercityhousingmarket, Leyarguedthatcultural labourand capitalto a secondaryroleinurbanaffairs.
factorsare important:'The neighbourhoodsthem- ButLeywas notadvocatingan autonomoustheoryof
selvesincludea measureoflife-style, ethnicandarchi- consumption-determined urban developmentand
tecturaldiversity,valued attributesof middle-class change,or a straightforward consumerpreference
moversto thecentralcity... thesedesiderataof the theoryof gentrification as some of his criticshave
cultureofconsumption shouldnotbe under-estimated argued,and nordoes hisworkrestjuston Bell'scon-
in interpreting the revitalizationof the innercity' ceptof thepost-industrial city.On thecontrary, the
(Ley,1981,p. 128). importancehe accordsto cultureand consumption in
Ley had less to say on thestructure and operation the post-industrial city are clearlyrooted in the
ofurbanlandand housingmarketand thesupplyand deeper changes in the structure of production,the
productionof gentrifiable propertiesand areas and, changing division of labour, and the rise of a
wherehe does, it is more focusedon the demand locationallyconcentrated serviceclass.
aspects of the equation.Ley noted the role of thereal While Leyargues thatthis classplayeda keyrolein
estate industry, but he accorded it a secondaryor politicsandculture, he also identifieditas a productof
reinforcing roleinthegentrificationprocess.Referring thechangesinthedivisionoflabourand thespatially
to therevitalization oftheinnercityareaofKitsilano, uneven nature of these changes. He thus linked
he statesthat:'Thereis littledoubtthattheactivityof together changes in the organization of pro-
therealestateindustry addedto theinstability of the ductionand theeconomy,politicsandculture, intoan

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
178 CHRIS HAMNETT

approachto gentrification andurbanchangebased on contradictory. Inhisview,theactionsofproducersas

theproductionofgentrifiers andtheircultural charac- well as consumersneed to be takenintoaccountin
teristicsand requirements. Withoutthis,he would explainingthegentrification phenomenon. As he put
havebeenguiltyofadvocatinga non-materialist, con- it:'To explaingentrification accordingto thegentri-
sumption-based, theoryofgentrification as hiscritics fier'sactionsalone,whileignoringtheroleofbuilders,
have suggested.But,inmyview,theyhave misinter- developers,landlords,mortgage lenders,govern-
pretedhis stresson cultureand consumptionas a ment agencies, real estate agents and tenantsis
narrowdemandandpreference-based approachwhen, excessivelynarrow.A broadertheoryof gentrifica-
in fact,itis based on changesin thesocialand spatial tion must take the role of producersas well as
divisionof labour and on the supply of potential consumersintoaccount'(Smith,1979,p. 540).
gentrifiers.These changesunderpinthedevelopment Smithis entirelycorrectin thisrespect,and this
of a new cultureand the residentialand political is somethingthat Ley largely failed to do. But
demandsthatfollowfromit. what Smiththenproceededto do was to argue for
Lookingat Ley's earlyworkin general,it can be producerdominance:
arguedthatitsstrength liesinitsfocuson thechanges
in the social and spatialdivisionsof labour,and the itappearsthattheneedsofproduction - inparticular the
concentration ina limitednumberof'post-industrial', needtoearnprofit - area moredecisive initiative behind
service-dominated cities,of a professional and man- gentrification thanconsumer preference.Thisis notto
agerial elite.He accords a considerable stress to the say in some naive way that consumption is theauto-
role of changesin cultureand consumptionand the maticconsequence ofproduction, orthatconsumer pref-
or demandsofthenew elite, erenceis a totally passiveeffect causedbyproduction.
residential requirements
but he locates thisin the contextof changesin the Suchwouldbe a producer's sovereignty theory, almost
as one-sided as itsneo-classical counterpart.Rather, the
natureand structure ofeconomicorganization. Ley's between
relationship production and consumption is symbi-
thesisis strongestin theexplanationit offersof the otic,but it is a symbiosis in whichproduction dominates.
typeofcityin whichgentrification is likelyto occur, Althoughit is of secondaryimportancein initiating the
and the characteristics of the gentrifiers. It also actualprocess, andtherefore inexplaining whygentrifi-
implicitlydeals with the timingof gentrification cationoccurred in thefirst place,consumer preference
through itsanalysisofthegrowthoftheserviceecon- anddemandareofprimary importance indetermining
omyin the1970s and 1980s.Whereitis weakeris in thefinalform andcharacter ofrevitalized areas(Smith,
its explanationof the areas in whichgentrification 1979,p. 540emphases added).
occurs,whichLeyseeslargelyas a productofdemand
forinnercitylocationsand theamenityand cultural Smithconcludedthat:
facilitiestheyofferto the gentrifiers. The supplyof
potentialgentrifiable houses is assumed to followon The so-calledurbanrenaissance has been stimulated
fromthe demandsand marketpower of potential morebyeconomic thancultural Inthedecision
forces. to
gentrifiers to outbidotherusers.But Ley's stresson rehabilitateinner citystructure, oneconsumer preference
the marketpower of the new elitesuggeststhathe tendstostandoutabovetheothers - thepreference for
sees thepowerto outbidotherusersas a majordeter- profit,or,more accurately a sound financialinvestment.
minantof theurbanlandscape:perhapsas important Whether ornotgentrifiers articulatethispreference, itis
fundamental, forfewwouldevenconsider rehabilitation
as thenew elite'scultureofconsumption. A theory
ifa financiallossweretobeexpected. ofgentri-
fication musttherefore explainwhysomeneighbour-
THE SUPPLY-SIDE ANALYSIS: hoodsareprofitable to redevelop whileothersarenot?
GENTRIFICATION AND THE 'RENT-GAP' What are the conditions of Consumer
profitability?
sovereignty explanations tookforgrantedtheavail-
Ley's approachto the explanationof gentrification abilityof areasripeforgentrification whenthiswas
stressedtheproduction ofgentrifiers andtheircultural preciselywhat had to be explained(Smith,1979,
andconsumption requirements as itskeyelement.The pp. 540-1).
supplyofgentrifiable properties and theoperationof
theurbanlandand housingmarketswereaccordeda Smiththenproceededto lay out his theoryof the
secondaryrole.Smith(1979) completely reversedthis rentgap. This is by now verywell known,and I do
explanatoryemphasis,arguingthat the 'consumer notintendto detailhisargument infull.Suffice to say
preference'argumentswere takenforgrantedand thatitlocatesgentrification withinlong-term shifts of

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblindmenand theelephant 179
investmentand disinvestment in the builtenviron- tal,eventuallyproducestherentgap. Whenthisgap
ment,and focuseson the relationship betweenland growssufficiently
large, (orforthatmatter,
rehabilitation
and propertyvalue,particularly on theway inwhich renewal) canbegintochallengetheratesofreturnavail-
able elsewhere and capitalflowsback'(Smith,1979,
disinvestment producesthe possibilityof capitalre-
investment. Smitharguesthatin thenineteenth cen- p. 546).
tury,most citieshad a classicalland value gradient, Hence,the subtitleof Smith'spaper:'A back to the
highestat thecentreandfallinggraduallytowardsthe citymovementby capital,notpeople'.
periphery. But,as thesuburbanization ofindustryand
populationproceededfromthe turnof the century
AN ASSESSMENT OF SMITH'S RENT GAP
onwards,landvaluesintheinnercityfellrelativetothe
CBD and thesuburbsand a 'valley'in thelandvalue THEORY OF GENTRIFICATION
gradientopened up which intensifiedduringthe This is an elegant argument,and Smithwas quite
decades of sustainedsuburbanization in 1940s, 50s correctto attemptto shiftthe emphasisaway from
and60s. Thisdevalorization oftheinnercityprovided the early consumerpreferenceand demand argu-
thebasisforsubsequentprofitable reinvestment. mentstowardsa consideration of thesupplyof gen-
The keyforSmith,is therelationship betweenland trifiablepropertyand the role of mortgagefinance
value and propertyvalue.When depreciationof the and profitability.Butit is now clearthat,despitethe
existingstructureshas proceeded farenough, the importance ofhisrentgap thesisforan understanding
pointis reachedwherethecapitalizedgroundrentof of theunevenpatternof investment, disinvestment
site or neighbourhoodis less than its potential and reinvestment in thebuiltenvironment, his rejec-
groundrentin its 'highestand best use'. This is the tion of alternativeexplanatoryapproaches,- particu-
rentgap, and accordingto Smith,gentrification or larlytherole of the new class,and its consumption
redevelopment,can occur when the gap is wide and culturalcharacteristics, and his unwillingness to
enoughto ensurea profit. accordindividualactorsanysignificant rolerendered
his initialapproach of only limitedvalue for the
Oncetherentgapiswideenough, maybe
gentrification In Smith'sthesis,indi-
ina givenneighbourhood
initiated explanationof gentrification.
byseveraldifferentvidual aremerelythepassivehandmaidens
actorsinthelandandhousing market.Andherewecome gentrifiers
backto therelationship betweenproduction andcon- ofcapital'srequirements.
sumption, for the empiricalevidence suggestsstrongly The logicalplace to startis withSmith'srejection
thattheprocessis initiated notbytheexercise ofthose ofconsumerdemandtheoryand Ley'spost-industrial
individualconsumer preferences much beloved ofneo- thesis.Smithacknowledgedthat only Ley's post-
classicaleconomists, but by someformof collective industrial thesisis broadenoughto accountforgen-
socialactionat theneighbourhood level(Smith, 1979, trificationinternationally,buthe rejectedit as being
p. 545emphasis added). contradictory. If individualpreferenceschange in
unison,theycannotbe individualpreferences or the
Smith'soppositionto any explanationof gentrifi- constraintsare strongenough to force
cationbased on individualconsumerpreferences is overriding
themintothesamemould.Thereis sometruthin the
clearcut,andreferring to theimportance ofmortgage
second argument.Consumer preferencesdo not
fundingin thisprocess,he arguesthat:
emerge out of thin air. They are partlysocially
All theconsumer preferencein theworldwillcometo created,manipulated andshaped,andtheyarenecess-
noughtunlessthislongabsentsourceof funding re- arilymade on thebasis of theavailableoptionsand
appears;mortgage capitalis a prerequisite.
Of course, constraints and notalwaysin thecircumstances ofan
thismortage capitalmustbe borrowed bywilling con- individual'sown choosing.WhereSmithis wrongis
sumers exercising somepreference oranother.Butthese in
arguingthat,forthe conceptof individualpref-
preferences are not prerequisites sincetheycan be erenceto be valid,individualsin different countries
created
socially (Smith,1979,pp.545-6). mustmakedifferent choices.Ifsimilargroupsin dif-
ferentcountriesarefacingsimilaroptionsat thesame
Smithsummarizes his thesisas follows:
time,itis scarcelysurprising thattheremaybe similar
is
'gentrification a structural of
product thelandand outcomes. But this does notmean thatindividuals are
housing markets. Capitalflowswheretherateofreturn is totallydeterminedin theirchoices as Smith(1979,
highest,and themovement of capitalto thesuburbs p. 540) seemsto imply,or thatall 'preferences are ...
alongwiththecontinual depreciation ofinnercitycapi- sociallycreated'.

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
180 CHRIS HAMNETT
Smith'ssolutionis to redefine preference in terms correct(see Moore, 1982) to argue that:'the pro-
of 'collectivesocial preference', but this does not cess is not initiatedby the exerciseof individual
explain where collective social preferencescome consumerpreferences ... but by some formof col-
from.All it does is to displace the problem of lective social action at the neighbourhoodlevel'
explainingthe originsof preference up the scale to (p. 545). Smith'stendencyto consistentlydismiss
a more ideologically acceptable, if theoretically the role of individualgentrifiers in favourof col-
mysteriouslevel. It should also be stressedthat lective social actors is clearly seen where he
onlya minority of people decideto live in theinner identifiesthreetypes of developerswho typically
city and become gentrifiers. Many more decide to operatein gentrifying neighbourhoods.They are:
move out to the suburbs.There remains,therefore,
(a) professionaldevelopers whopurchase property, re-
theproblemof explainingwhy some people do one itandresellforprofit;
develop (b) occupier developers
thing, and some do another. This cannot be who buyand redevelopproperty and inhabit it after
explainedin termsof capital flows,disinvestment completion; (c) landlorddeveloperswho rentit to
and reinvestment. Althoughthe gentrification pro- tenantsafter Thefragmented
rehabilitation.... structure
cess does involve capital flows,it also involves of... ownership hasmadetheoccupier developer, who
people, and this is Achillesheel of Smith'ssupply is generally an inefficient
operator in theconstruction
side thesis. industry into an appropriate vehiclefor recycling
Not onlydoes Smithrelegateconsumerpreference devalued neighbourhoods (Smith, 1979,p.546).
and demandto a subsidaryrolein favourof thepro- What Smithis arguingis that,contraryto all his
ductionof residential space,he arguesthatthefocus otherassertions on thecentralimportance ofproducer
of a theoryof gentrification must be one of the
interests,andthesecondaryroleofconsumer choice,is
reasonswhysome neighbourhoods are profitable to thattheindividualhouseholdsare themselvesone of
redevelopwhileothersare not.Smithis correctthat themostimportant and indeed,appropriate forcesin
consumer sovereignty/demand-led explanations theproduction ofgentrified neighbourhoods. Onlyby
took for grantedthe availabilityof areas ripe for themas developersisheabletocircumvent
classifying
gentrification. But, as we shall see, Smithfellinto thisawkwardintrusionof individualrenovationfor
an almostidenticaltrapby takingforgrantedthe
intohisproducer-dominated thesis.To
existenceofa pool ofgentrifiers andtheconditionsof consumption
theextentthatindividualproducers/consumers play
demand.He assumedthatiftheconditionsofprofita- a
key role in the gentrification process (and thisis
bilitywerefavourablethatgentrification (or,forthat trueinLondon),Smith'sdistinction between
certainly
matter, renewal)wouldtakeplaceand thatthepoten- and is an artificial
one andhe
consumption
tial gentrifierswere on hand to play a role in the production
failsto explainwheretheindividualdevelopergentri-
revalorization process.Only laterdid he attemptto fierscome from,or why some individualsbecome
rectify thislacuna,itselfa productofhis tendencyto while othersdo not. In Smith'sanalysis
assume thatdemandwas of secondaryimportance gentrifiers, individualsseem to gentrifybecause of the value
to supplyin theexplanationofgentrification. of theircharacteristics, tastes and
Smithis correctin arguingforthe centrality of gap, irrespective
as
demands,butas Rose (1984) perceptively pointsout:
mortgagefinanceinurbanresidential restructuring
Harvey(1974),Williams(1976; 1978),Boddy(1976), gentrifiersarenotthemerebearers ofa processdeter-
Wolfe etal. (1980), Hamnett and mined independently of them. Their constitution, as cer-
Dingemans(1979),
(1986; 1987) have also shown. But taintypesof workers, and as people,is as crucialan
Randolph element intheproduction ofgentrification as is thepro-
althoughabsenceofmortgagefinance rendersgentri-
ofthedwellings
fication on allbut a small its
scale, presence
duction they occupy. They mayormay
impossible not makethepotential process happen in particular
does not, of itself,creategentrification. Mortgage situations(Rose,1984,p.56).
contingent
financeis a necessarybutnota sufficient conditionof
large-scalegentrification. Nor is itadequateforSmith Rose's statementis a powerfulindictmentof
to argue that although'mortgagecapital mustbe the economisticand deterministic characterof the
borrowed by willing consumersexercisingsome rentgap theoryof gentrification with its overrid-
preference or another... preferences can be socially ing stresson the productionof gentrifiable areas.
created' (Smith1979, p. 546). This is correct,but It cannot be too stronglyemphasized that gentrifi-
Smithimpliesthatall preferences are sociallycreated cation does not occur independently of individual
which is nonsensical. Nor is it empirically gentrifiers. Althoughtherentgap maybe necessary

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblindmenandtheelephant 181
forgentrification to occur,it is not sufficient.
It does their ownhousing, areessential
totheparticularhistories
not necessitatethatgentrification willtakeplace. In- whichunfold ina place'(Clark,1988,p. 244).
deed,rentgap theorysaysnothingaboutwhygentri- Badcock (1989) in his studyof Adelaide,South
ficationshouldtakeplaceratherthansomeotherform Australiafoundconvincingevidencethata sizeable
ofrenewalor redevelopment. The rentgap theoryof rent
gap had developedby 1970 in the Cityand in
gentrification is thussubstantially under-determined.someofthe Victorianresidential suburbs
surrounding
Gentrification is not 'to be expected'wheretherent and thatsubstantialgentrification had subsequently
gap exists;itis a contingent phenomenon.Gentrifica- occurredwhichfilledin the rentgap. But he also
tioncould occurbut so could renewal,deterioration concludedthat'theprocessesresponsibleforthisrent
or abandonment. as Smith
gap are nowherenear as straightforward
And given that the gap between potentialand would have it' (Badcock,1989, 132). He argued
p.
actual groundrentsis predicatedon the existence that was thethirdbestresponseofcapi-
gentrification
of potentialground rent,Smith says very little tal to
existingconditionsin Adelaide,and was, in
about the processes by which such potential some ways a
sub-optimalinvestmentstrategy
ground rentscome into existence.It is possible, (p. 133). In otherwords,gentrification was not an
forexample,thatin gentrifying areas,the potential inevitableoutcomeoftherentgap.
ground rent is, in part,a resultof demand from It is clearfromthesetwo studiesthattheexistence
potentialgentrifiers (Moore, 1982). As Munt (1987) ofa rentgap is not a sufficient conditionforgentrifi-
argues:'As gentrifiers can afford numerousinner-city cationto occur.On the contrary, the existenceof a
residentiallocations,itfollowsthatthesizeoftherent rentgap can lead to a varietyof different results
gap inparticular locationsdependson theirattractive-
includingredevelopment or further decline. More
ness, and hence on demand,whichis absent from generally,it appearsthatSmith'stheoryis of value
Marxist gentrification theory'(p. 1177). Ley goes insofaras it explainsthe existenceof areas within
furtherto argue that the rent gap is not even a cities where
gentrification may take place. It says
necessaryelementof gentrification. In his view, all tendsto occurin
nothingabout why gentrification
thatis necessaryis the potentialforprofitand the somecitiesratherthanothers,or aboutthecharacter-
abilityof gentrifiers to outbidexistingor potential isticsand
originsof the gentrifiers themselves,and
users fordesirableinnercitysites.Ley also argues rather than suburbanized. As an
whytheygentrified
that most developersare riskaverse and will not
analysisof the cycles of investment and disinvest-
riskenteringan area untildemandis proven.'From mentin the built environment it remainsa major
the developerspoint of view, demand is the bot- contribution, but its role in explainingotheraspects
tom line. In shortcapital followsdemand,though of is limited.
gentrification
this is not to say that local marketscannot be The principalreason why Smith'stheorywas
manipulatede.g., blockbustingor that demand is unable to address these otherquestionswas that,
produced by broader economic contexts' (1990, givenitsfocuson theproduction ofthebuiltenviron-
personalcommunication). ment,it was 'limitedto the specificationof pre-
These problemswiththerentgap thesishavebeen conditionsforthe
productionofgentrified dwellings
documentedin two recentempiricalstudies.Clark without
consideringthe productionof 'gentrifiers',
(1988) foundclearevidenceofa rentgapinhispioneer- theoccupantsof suchdwellings'(Rose, 1984, p. 51).
ing analysisof the evolutionof land and property BecauseSmithfocusedhisexplanationon theproduc-
values in Malmo,Sweden,buthe arguedthatit was tionof the rent
gap, and conflatedand dismissedas
theoretically explicableeitherin termsof Marshall's 'preferences',changes in occupational structure,
neo-classicalformulationor in terms of Smith's andreproductive behaviour, he ignored
demographic
Marxistone,and thattherentgap was in no sensea
key materialchangesinfluencing the productionof
determinant of gentrification or a completeexpla- and equatedmaterialist
gentrifiers, explanationswith
nationforit.Infactredevelopment ratherthangentrifi- therent
gap.
cation occurredin all cases in Malmo. Clark thus
rejectedtheidea of:
THE PRODUCTION OF GENTRIFIERS AND
somepredetermined development withthe'needsof THEIR LOCATIONAL CHOICES
capital'as primemoverand therent-gap as time-set
triggeringmechanism. Theactionofagentswithecon- In theearly1980s, Ley's thesisregardingtheroleof
omicorpolitical andofindividuals
interests, interestedin changesinthesocialandspatialdivisionoflabourand

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
182 CHRIS HAMNETT
in occupationalstructureand the rise of a 'new theoretical and empirical workby Marxistshas been
middle class' and the links to gentrification was exclusively preoccupied withthoseaspectsofgentrifi-
paralleled in differentways by several other cationwhichcanbe directly relatedto theoperationof
workers who made theoretical links between thelaw ofvaluein thebuiltenvironment ofcapitalist
cities.... Thishascreated notonlyananalytical gapbut
changes in the social and spatial restructuring of
alsoanepistemological errorofconsiderable
labour processes,corporateorganizationand what importance
(Rose,1984,p.52).
Rose (1984) termedthe productionof gentrifiers.
One such link was made by Mullins (1982) who She arguesthatitis essentialto move beyondthis
argued that dramaticchanges had taken place in verylimitedconceptionto explorethelinksbetween
the Australianinnercity.The declineof innercity and changesin the social and spatial
gentrification
manufacturing and the skilledworkingclass resi- of labour processes and the repro-
restructuring
dent population had been accompanied by the ductionoflabour
powerandpeople,whichhavebeen
emergence of corporate centres for monopoly largelyignoredbyeconomistic approacheswhichsee
capitalismand middleclass officeworkers.Mullins socialprocessesas eitherderivablefromtheeconomic
linkedthis to gentrification, arguingthat 'whereas or epiphenomenal.Beauregard(1986) has similarly
the workingclass of an earlierformof innercity
arguedthattherentgap alone is a totallyinadequate
lived therebecause of employment reasonscentred
explanationof gentrification. 'The explanationfor
on manufacturing industry,"educated labour" is gentrification begins with the presenceof "gentri-
comingto residein the innercity(for)uniquecon- fiers",the necessaryagentsand beneficiaries of the
sumptionreasons'(p. 45-6). But,as Mullinsnoted: gentrification process,and the directionstakenby
'the development of officeemploymentcannot their
reproductionand consumption'(Beauregard,
whollyexplainthe residentialincreaseof innercity 1986,p. 41).
educated labour simplybecause the bulk of these His argumentinvolved three key components.
workers... residein the suburbsand commute.... First,that the demand for inexpensive,inner-city
Other processes must have been involved in this
housingis nota newphenomenonandcannotsimply
residentialdevelopment'(Mullins,1982, p. 53). A be
explainedby therentgap. Secondly,that'thegen-
similarlinkwas also made by Moore (1982, p. 1) areoften, seldomalone,the"agents"
trifiers though ofthe
who argued that'gentrification representsthe pro- gentrification and thusprovidethemotivations
process,
cess wherebyan important fractionofthenew classis and thatshapeit',and thirdly, thatwithout
aspirations
establishing a residential identity concomitant with this the ceases to exist.Different
group process types
its social identity,with the overall contextof the of
housingmightbe rehabilitated, but as character-
centralcitybecomingmoreand morea whitecollar isticsof arebroadlysimilaracrossa variety
gentrifiers
city'. ofdifferent isdefined
areas,'gentrification bythepresence
Mullinspointedto thekeyroleofproductionand
of gentrifiers' (Beauregard,1986, p. 41 emphases
consumption ofparticular leisure-orientated
artsser- added).
vices withinthe innercity,whichare producedand This is an argument radicallyat odds withthatput
consumedby a limitednumberof educatedworkers. forward Smith.The causal
by primacyis exactlythe
This explanationforgentrification, whichis linked reverse.Whereas Smithassumed the existenceof
to the productionof gentrifiers and to theircultural and saw the productionof
potential gentrifiers,
requirements is similarto Ley'sthesis,and identifiesa areasas the
appropriate keytotheprocess,Beauregard
specificreasonforthelocationalconcentration ofthe identifies
gentrifiers as thekeyto explainingthepro-
new class in theinnercities:theirculturalneeds and cess.Gentrification withoutgentrifiers does notexist.
theconcentration ofculturalfacilities.The locational Like
LeyandMullins,Beauregardpointsto thecrucial
question is of crucial importance.What Mullins role played by the changesin industrialand occu-
realizedwas thatthegrowthofa new middleclassor and suggeststhatit is withinthe
pationalstructure,
serviceclassis necessary, butnotsufficient to explain 'urban
professionaland managerialfraction of labor
gentrification. A sufficient explanationmust also thatgentrifiers are situated'.And like Rose (1984),
accountforwhysomeofthisgroupresideintheinner Mullins(1982), Moore (1982) and Williams(1984),
cityratherthanelsewhere(see also Moore, 1982). Beauregardarguesthat:
The argumentregardingthekey role of the pro-
ductionof potentialgentrifiers was developed by In orderto explainwhytheseprofessionals andman-
Rose (1984) who arguedthat: agers... remainwithinthecityand also engagein

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblindmenand theelephant 183
we mustmoreawayfromthesphereof
gentrification thefactofurbantransformation.Giventhemovement of
productionandfocusupontheirreproduction andcon- capitalintotheurbancore,andtheemphasis on execu-
sumption .... Whatis itaboutan urbanresi-
activities administrative
tive,professional, and managerialfunc-
dence,in additionto proximity to work,whichis aswellasother
tions, thedemographic
activities,
support
especially
compatible withthereproductionand con- andlifestylechanges... helpto explainwhywe have
sumption activities
of thisfractionof labour?(1986, quichebarsrather
proliferating thanHowardJohnstons,
p.43). trendyclothesboutiques andgourmet foodshopsrather
thancorner stores'(Smith,
1986,p.31).
Beauregardconcludesby arguingthat:
This view represents the totalmarginalization of
therentgap argument provides one the to
only of necessary consumption' influencing thecolourand designof
conditions andnone
forgentrification ofthesufficient
ones.... theicingon thecakeofurbanrestructuring and gen-
Manyareasof central citieshaverentgaps greatly in trification.
It ignoresthe argumentsput forwardby
excessofthoseareasthatgentrify. Thusthetheory can- Moore,
Beauregardand Ley regardingthe import-
noteasilyexplainwhyHoboken...becomesgentrified,ance of cultureand
butNewark... doesnot(1986,p.39 emphases consumptionin explainingwhy
added). thenewclass theinnercityratherthanmove
gentrify
out to the suburbs.While Smithaccepts thatit is
Thisisa crucialpointwhichgreatlyweakensSmith's
to theroleofchangesin thestruc-
claims.To sumup,itisclearthattheexistenceofa pool important explain
tureof productionand thechangingspatialdivision
ofnewmiddleclasspotentialgentrifiers is a necessary
of labourin producingprofessionaland managerial
pre-requisitefor gentrificationto takeplace.So is the workersintheinner
existenceof a stockof potentiallygentrifiable areas city,he failsto addressthereason
a fractionof thisgroupshouldlocateintheinner
and houses. But neitherof these are sufficient for why
city.And whenhe discussestheroleofgentrifiers he
gentrificationto occur.Thatrequiresa fragment ofthe
resolutely dismissesany idea thattheymightplaya
expanded professionaland managerialgroup who crucialrolein the
wishto live in theinnerareas,and a concentration of process:
appropriate facilitiesand environments. Without as withtheoriginal themythology
frontier, hasitthat
these prerequisites, it is highlyunlikelythat gen- gentrificationisa processledbyindividualpioneersand
trificationwill occurnotwithstanding the actionsof homesteaders whosesweatequity, andvisionare
daring
developersand theavailability ofmortgagefinance. pavingthewayforthoseamonguswhoaremoretimid.
But... itisapparent thatwhereurbanpioneers venture,
thebanks, realestatecompanies thestateorothercollec-
SMITH'S ATTEMPT TO INTEGRATE tiveeconomic actorshavegenerally gonebefore(Smith,
CONSUMPTION INTO GENTRIFICATION 1986,pp.18-19).
In 1986 Smithattempted to locatetherentgap thesis
But thisis not borneout by evidencefromLondon
within a wider analysis of gentrification which
and New York (Zukin,1987) whichindicatesthat
includedthe de-industrialization of capitalistecon-
individualpioneersdo play a key initialrole even if
omiesandthegrowthofwhitecollaremployment, and
theymaybe oftenovertaken by thebanks,realestate
changesin demographyand consumptionpatterns. and Munt (1987) arguedthatin
This appearedto heralda significant agents developers.
wideningof his
Battersea,London,'a gradualprocessof infiltration
approach,and Smithnoted that'althoughprevious ... preceededany largescale develop-
by gentrifiers
attemptsat explanationhave tendedto fastenon one ment'(p. 1177). Contraryto Smith,thereis a strong
or theothertrend,theymay not in factbe mutually
case that where the collective economic actors
exclusive(p. 21). Thisis an important concession,but
venture,urban pioneers have often gone before
Smith'sview regretably remainedfirmly production
(Goetze,1979).
based, viewingdemographicand culturalprocesses In 1987 in a majorpaperentitled'Of yuppiesand
as epiphenomenon or surfacefroth. As he revealingly
housing:gentrification, social restructuringand the
putit: urbandream',Smithattemptedto tacklethe social
and head on.
changesin demographicpatterns pref- restructuring consumptionarguments
and life-style
erencesarenotcompletely but... theimport-
irrelevant, Looking firstat theevidencefor theexistenceofa 'new
anceof demographic issuesseemsto be middleclass',Smithacceptedthattherehas been an
and life-style
inthedetermination
chiefly ofthesurfaceform takenby undeniableoccupationaltransformation, with 'pro-
muchoftheurbanrestructuring
ratherthanexplaining fessional,managerialand upperlevel administrative

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
184 CHRIS HAMNETT

personnelin expandingsectorsheavilyrepresented Smith'sargumentis a fascinatingone. Having

among gentrifiers' (Smith,1987a, p. 154), but he acceptedthatdemandcan play a role in alteringthe
arguedthatthisdoes notprovetheexistenceofa new natureofproduction, hethenavoidstheconsequences
middleclassinMarxistterms(i.e.inrelationto owner- of thisadmissionby arguingthattheconundrum of
shipand controlof themeansofproduction). Thisis does notturnon wheredemandcomes
gentrification
correct,but as the social restructuring thesisis pri- from,buton whyittakesthelocationalformit does.
marilyconcernedwithoccupationchangeandnotthe This questionis fundamental fortheexplanationof
theoretical
validityofMarxistclasscategories,thisis But it is only halftheissue.The con-
gentrification.
largelyirrelevantand Smithappears to accept the undrumof gentrification turnson boththe expla-
existenceof a new 'class' in empiricaltermsifnot in nationofwheremiddleclassdemandcomesfromand
termsofMarxistclasstheory.As he putsit: on its manifestationin the centraland innercities.
Smithhoweveridentifies the secondquestionas the
Thereis no doubt thatemployment structure has keyone. He arguesthat:
changed dramatically andthata profound socialrestruc-
turingistaking place... andthatitisaltering...theclass Therecan be littledoubtthata continued and even
configurationofsociety. thissocialrestructuring accelerated
Equally, centralization of administrative,
executive,
isheavily implicated inthegentrificationprocess(Smith, professional, managerial andsomeservice activities
may
1987a, p. 161). makea central domicile moredesirable fora substantial
sectorofthemiddle class.Butdo thesearguments really
But while Smithacceptedthe 'overarchingimport- amountto an explanation ofthegeographical reversal
of thelocationhabitsby a proportion ofmiddle-class
ance' of the new work on social restructuring for menandwomen? ... theargument thatsocialrestructur-
explaininggentrification he arguedthat:
ing is the primary impetusbehindgentrification is
underdetermined
substantially 1987a,p. 164).
(Smith,
theyalso bringcertain intrinsic
dangerswiththem.If
istobeexplained
gentrification andforemost
first as theresult Smithis correctinarguingthatsocial
restructuring
oftheemergence ofa newsocialgroup... thenitbecomes aloneis notan adequateexplanationofgentrification.
toavoidat leasta tacitsubscription
difficult tosomesortof
But,as we have seen,the proponentsof the social
consumer preference model, no matterhowwatered down.
Howelsedoesthisnewsocialgroupbring restructuring thesisdo not argue thatit is. On the
aboutgentri-
fication kindsandlocations contrary, they all point to the crucialrole of the
exceptbydemanding specifc
ofhousingin themarket' (Smith, 1987a,p. 163 added specificculturaland consumptionrequirements of a
emphases). fragment of thenew class,and argue that are
they met
byaninnercitylocation.Thereisa causallinkbetween
theproductionofa new professional and managerial
Smith'sfearsare very clear,and they shape his
labourforce,theculturaland consumption character-
attempt to resolve his of
problem acceptingtheexist- isticsof
enceofa newsocialgroupwithoutgivingthema key partofthatgroup,andthecreationofpotential
Thereare two stepsto theargument,
gentrifiers. not
role in the gentrification process.His 'solution'is one, but Smith
only acknowledges the firstand dis-
ingeneousand highlightswhat is perhapsthe key missesthesecond.Not Smithconcludes
surprisingly,
problemin theexplanationof gentrification: namely that:
itsspatialmanifestation. He states:
I woulddefend
therent-gap ... notas initself
analysis a
Thereisnoargument butthatdemand canattimes - and orcomplete butas thenecessarycenter-
definitive explanation
especiallythosetimeswhendemandchangesdramati- pieceto any theoryof gentrification. It is the historical
cally- alterthenature ofproduction.Buttheconundrum ofcapitalinvestment
patterns anddisinvestment inthe
doesnotturnonexplaining
ofgentrification where middle central andinnercitycitiesthatestablishestheoppor-
classdemand comesfrom. Rather,itturnsonexplaining the (notthenecessity)
tunity forthisspatialreversalinthe
essentially
geographical ofwhycentral
question andinner first (Smith,
place 1987a,p. 165emphases added).
areasofcities,
whichfordecadescouldnotsatisfy
thedemands
ofthemiddle class,nowappearto do so handsomely. If,
indeed,demandstructures havechanged, we needto This statementrepresents a substantialretreatfrom
explainwhy thesechanged demands have ledtoa spatial Smith's
initial
position,andpresupposes whatBadcock
re-emphasis on thecentralandinner city(Smith,1987a, (1989,p. 126)has termed'a considerable relaxationof
pp.163-4emphases added). thetheory'soriginalassumptions'. Smithnow seems

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblindmenand theelephant 185
to view therentgap as a key whichtranslates more process.Butsuchdifferences aside,thefactthatSmith
generalprocesses, i.e. the production of gentrifiers had to undertake thisreinterpretation is indicativeof
intoa spatialreversal.ButSmith'sargumentthatthe thelimitations oftherentgap theoryofgentrification
rentgap is thenecessarycentrepiece to anytheoryof and Smith'sfundamental unwillingnessto concede
is too
gentrification large a claim. As Smith points out, thatindividuals have any significant role in shaping
the rentgap establishesthe opportunity, not the theirenvironment. Yet the closestSmithcan bring
necessity, fora spatialreversalto occur.The rentgap himselfto go is to accepttheroleof collectivesocial
may providethe means,but it does not providea actors and the functionalrequirements of differen-
motiveforgentrification. Forthis,we needtolookinto tiatedconsumption innewmodeofregulation. Itisnot
what is, forSmith,the heartof darkness:locational thatSmithrefusesto grantindividualagencydomi-
preferences, lifestyles and consumption. nance- thisisnottheargument - butthathe seemsto
GiventhatSmithfindsanyemphasison individual refuseto acceptitevenexistsat anything otherthana
lifestylesand consumption unacceptable;in 1987 he superficial level.His oppositionto anyformofagency
outlineda way oftryingto integrate production-side explanationrevealshimas a structuralist forwhom
and consumption-side argumentsvis-a-visgentrifi- individualagencyis reducedto theroleof flickering
cationin termsof a historicalanalysisof societalre- shadowscastby thelightofcapital'sfire.
structuring. This entailed rejectionof Ley's ideas
about post-industrialism as a 'shallow empirical
abstraction... incapable of sustainingtheoretical TOWARDS AN INTEGRATED THEORY OF
scrutiny'(Smith,1987a, p. 166) whilereinterpretingGENTRIFICATION
the substanceof the consumptionsocietyargument
in termsofthe'regulationist' analysisofAglietta.Itis It has been arguedthatboththesocial restructuring
argued thatas the intensive regimeof accumulation thesisassociatedwith Ley and the rentgap thesis
began to frayat the edges in the 1970s and 1980s, advanced by Smithare partialattemptsto explain
therehas been a switchtowardsa new (post-Fordist) gentrification. Ley'sapproachfocusedon changesin
regimeofaccumulation associatednotwithmasspro- the social and spatialdivisonof labour,changesin
duction and consumption,but with differentiatedoccupationalstructure, the creationof culturaland
productionand consumption. In thisnew regimeof environmental demandsand theirtransmission into
accumulation, theaccentis on product-differentiation thehousingmarketvia thegreaterpurchasing power
andnichemarkets. Gentrification is explainedinthese of the new class. He largelytook forgrantedthe
termsas a resultofthedesireofgentrifiers to differen- existenceofpotentialareassuitableforgentrification
tiatethemselvesfromothersocial groups.As Smith and saw the processprimarily in termsof housing
notes: marketdemand.Smithon theotherhandfocusedon
the productionof gentrifiable housingthroughthe
It is thisquestionof cultural differentiation in a mass mechanismof therent
gap. He tookforgrantedthe
market whichismostrelevant togentrification. Gentrifi-existenceof a
cationis a redifferentiation of thecultural supply of potentialgentrifiers and
socialand the of a of the new class
economiclandscape... gentrification andthemodeof ignored question why segment
it are an opted to locate in the innercity.Mullins,Moore,
consumptionengenders integral partofclass andRose arguedthatan understanding of
constitution; theyarepartof themeansemployed by Beauregard
newmiddleclassindividuals to distinguish themselves theproductionofgentrifiers and theirsocialand cul-
fromthe... bourgeoise above and theworking class turalcharacteristicswas of crucialimportance foran
below(Smith, 1987,pp.167-8). understandingof gentrification. They developed
Ley's thesisconsiderably and arguedthatgentrifiers
What Smithhas done is to reinterpret, in termsof are centralto the gentrification process. Without
regulationist theory,Ley's work on post-industrial them,the process cannotoccurat all. But gentrifi-
consumption.But Smith's interpretation of con- cationis notsimplya productofchangesin thesocial
sumptionand itsrolein gentrification is clearlyvery and spatialdivisionoflabour,crucialthoughthishas
different fromthatsuggestedby Ley and others.By been. A specificlocationalorientationtowardsthe
stressingthe importanceof consumptionwithinthe innercityor specifichousingareas withinit,is also
framework of capitalaccumulationhe attemptedto necessaryand a supply of gentrifiable areas and
circumvent thetheoretical dangersinherent ingiving housingdefinednotjustintermsoftheexistenceofa
individualgentrifiers a key role in thegentrificationrentgap, but also in termsof relativedesirability or

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
186 CHRIS HAMNETT
attractivenessto the potentialgentrifiers (Munt, and spatialdivisionsof labourwhichhave led to de-
1987,pp. 1195-6). industrialization of advanced capitalisteconomies
There are fourrequirements forgentrification to and thegrowthoftheservicesector.This,inturn, has
occuron a significant scale. The firstthreeare con- been associatedwiththerapidexpansionof thepro-
cerned,respectively, withthesupplyofsuitableareas fessionaland managerialserviceclass,and the con-
forgentrification, thesupplyof potentialgentrifiers,centration of keyfinancial,legal and otherfunctions
and theexistenceof attractive centraland innercity in a relativelysmallnumberof majorcitiessuch as
environments. They comprisethe necessarysupply London and New York and Parisand a numberof
side elementsof theequation.The finalrequirement other major cities such as Vancouver, Toronto,
involvesa culturalpreference forinnercityresidence Sydneyand San Francisco.It is in these citiesthat
bya certainsegmentoftheserviceclass.Itis therefore gentrification has beenmostmarked.
possibleto conceiveofa rangeofpossibleoutcomes The explanationforgentrification musttherefore
depending on the combination of thesefour elements. begin with the for
processesresponsible theproduc-
The range of outcomesare shown in Table I. The tionand concentration ofkeyfractions oftheservice
important pointto emergefromthe schemais that class in a numberof major cities.These processes
gentrification onlyoccursunderone combinationof have producedthepool of potentialgentrifiers, and
circumstances. None oftheothercombinations leadto theprimary emphasismustbe on theexplanationof
gentrification, althoughLeywouldarguethatitcould the expansionof thiskey group.This is nota con-
occurwithouta rentgap as longas thenewclasshave sumptionbased explanation.It is firmly based in the
thepurchasing to or
power displace replace other land changes in the of
structure production thesocial
and
users. and spatialdivisionof labourin advancedcapitalist
TABLE I. Conditions schema
forgentrification countries.Itis thennecessaryto explainwhygentrifi-
cationoccursinsomeofthesecities.As we have seen,
two conditionsare necessary.First,it is necessaryto
Rent No Rent
have a supplyof potentiallygentrifiable innercity
gap exists gap exists
property.This is whererentgap theorycomes in,
No potential No gentrificationNo gentrification
explainingwhya supplyofdevaluedinnercityprop-
gentrifiers
ertyexistsas a resultof priorsuburbanization and
dencentralization. The potentialvalue of thisprop-
Supplyofpotential value.But,as we have
exists
gentrifiers
ertyis greaterthanitscurrent
seen,theexistenceofa rentgap does notnecessarily
No innercitydemand No gentrificationNo gentrification
Withouttheexistenceofa pool
lead to gentrification.
Innercitypreference Gentrification Gentrification? of potential gentrifiersand available mortgage
by a sectionof the will not occurhowevergreat
'new class' finance,gentrification
therentgap and howevergreatthedesireof devel-
opers to make it happen. And where appropriate
Butthisis merelya classification ofcircumstances. innercityhousingstockdoes not existin sufficient
It does not,of itself,providea basis of a theoryof quantity,as for example in cities such as Dallas,
gentrification. And,as we have seen,thekeyquestion Pheonix and othernew southernand westernUS
forsucha theoryis itsstarting point.It is inadequate cities,gentrification may be verylimited,however
to argue thatgentrification is the resultof a combi- large the new serviceclass. In older north-eastern
nationofcircumstances withoutattempting to assign Americancitiessuch as Baltimore, Philadelphiaand
some theoreticalpriorityto those circumstances. I WashingtonD.C., on the otherhand, thereis an
have no doubt that,as Beauregardhas argued,that abundantsupplyofnineteenth centuryrowhousing,
'the explanationforgentrification begins with the much of it devalued and run-downand home to
presence of gentrifiers' and that 'gentrification is workingclassandminority populations.In suchcities
definedby the presenceof gentrifiers' has
(Beauregard, gentrification proceededapace.
1986). But thisdoes not mean thatcultureand con- Secondly,therehas to be some effective demand
sumptionareassignedfirst placeintheexplanationof for innercity propertyfrompotentialgentrifiers.
gentrification. As Ley, Mullins and others have This may resultfromfinancialinabilityto afforda
pointed out, the appropriate place to startis withthe suburbanhome or, as is morecommonlyargued,it
changesin thestructure ofproductionand thesocial may stemfroma preference to live in theinnercity

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblindmenand theelephant 187
close to centralcityjobs and socialand culturalfacili- 'choice,consumption and culture'side of thedebate
ties. This, in turn,depends on both the growthof has,in fact,alwayshad one footveryfirmly planted
serviceclass job opportunities downtown,and on in the realitiesof changes in the materialbase of
demographicand lifestyle changeswhichhave seen productionand itsculturalmanifestations.
largenumbersof womenenterthelabourforceand In some ways,theconflict has been betweentwo
growing numbers of both single households and dual of
interpretations production.The one looking at
careerchildlesscouples.Forthesegroups,witha high changesinthesocialandspatialdivisionoflabourand
disposableincome,innercitylocationsoffer proximity theproductionofgentrifiers, and theotherlookingat
to employmentand to restaurants, arts and other the productionof the builtenvironment. But,until
Not surprisingly,
facilities. a significant
proportion of recently,Smith has consistentlyinterpretedthe
themappearto have optedforinnercityresidencein former approachin termsofindividualconsumption,
thosecitieswherecitycentresocialandculturalfacili- cultureand choice,and has generallyrejectedwhatit
ties exist.Withoutthis effectivedemand,based in had to offer.And this,as we have seen, has been
largeparton a positiveorientation towardscentraland considerable.Smithhas recentlyacceptedthatit is
innercity living,gentrification is unlikelyto occur important to integrateproductionand consumption,
howeverlargethe armyof potentialgentrifiers and but thisintegration has stillbeen in termsofa frame-
howeverlargetherentgap. workwhicheitherascribesprimacyto questionsof
We are therefore facedby threesetsofconditions productionor re-interprets consumption in a collec-
all of whichare necessary,and none of whichare tive non-problematic way. Smith'sconceptionof
sufficient.Butitis clearthattheexistenceofa poten- individualactionis a limitedand circumscribed one.
tial pool of gentrifiersis logicallyand theoretically He accepts that collectivesocial actors can make
priorto the housingpreferences and lifestylesof a gentrification, but not a multiplicity of individual
sub-group of the serviceclass. And, while the exist- actors. If the criticism ofLey'sposition beenmore
has
ence of a supplyof appropriateinner-city houses is limited, it is partlybecausehe has said less and been
necessaryforgentrification to occur,theexistenceof farless assertivein hisclaimsfortheoretical primacy.
a rentgap willnot,ofitself, producegentrification. It It is also clearthathis initialrecognitionof the key
is thusdifficult to acceptSmith'sview thatthe rent roleofa new groupofpotentialgentrifiers withtheir
gap is 'the necessarycenterpieceto any theoryof specificculturaland locational requirementswas
gentrification'.Necessaryit may be, but if gentrifi- broadlycorrect. His sinshavebeenofomissionrather
cationtheoryhas a centrepieceit mustreston the thancommission.The supplyof dwellingsand the
conditionsfortheproductionofpotentialgentrifiers. role of developer/speculators in the process have
gone largelyunexaminedby Ley. They are seen as
CONCLUSIONS beinglargelyderivedfromthedemandsfromthenew
class.
I have attemptedto show thatthe debate over the Smith'sclaim,that gentrification is a structural
explanationofgentrification hasbeenbroadlyshaped productof the land and housingmarketsalone,can
by theconflict betweenthosewho have arguedthat now be seen to have been misplacedas Smithnow
the keyto the problemlies in global changesin the partiallyaccepts.This is not to say thattherentgap
structureof productionand the social and spatial thesiswas wrong.The pointis ratherthattherentgap
divisionoflabour,andin theconcentration inspecific explains,at best,halftheproblem,and probablyless.
citiesofa sectionofthe'newmiddleclass'or 'service' The existenceofrelatively cheapand devaluedhous-
classwitha particular demographic composition, and ingis a necessary, butfarfromsufficient elementofan
culturaland consumptionorientation. On the other explanation.Equally,explanationof the production
handSmithhasconsistently arguedforthekeyroleof of potentialgentrifiers, theirculture,consumption
investmentand disinvestment in the builtenviron- andreproduction isnecessarybutinsufficient. A com-
mentand foran approachbased on the primacyof prehensiveand integratedexplanationof gentrifi-
profitability.This conflicthas manifesteditselfin a cationmustnecessarilyinvolve the explanationof
varietyof ways. In a conflictbetween so-called wheregentrifiers come fromand whytheygentrify,
'supply'and 'demand'explanations,choice and cul- how the areas and propertiesto be gentrified are
tureversuscapitaland so on.Yet,I havearguedmany producedand how the two are linked.And thereis
of these dualismsand polaritieshave been more a strong case that, notwithstanding the role of
apparentthanreal and what Smithwould label the institutional and collectivesocial actorssuch as real

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
188 CHRIS HAMNETT
estateagents,developersand mortgagelenders,the School of Social Sciences,The AustralianNational
key actorsin the gentrification process have been University,Canberra.Versions were given at the
individualgentrifiers themselves.It is necessaryto Institute of British Geographers Conference,
acceptthatindividualagencyis important in theex- Universityof Glasgow, January1990 and at the
planationof gentrification and to seek to integrate workshopon EuropeanGentrification, Department
productionand consumptionnot in termsof struc- ofGeography,University ofUtrecht, January 1990. I
turalcausesorindividualeffects, butintermsofstruc- am gratefulforcommentson the draftpaper from
turesand individualagency. Blair Badcock, University of Adelaide, Steve
Because Smithdevelopedrentgap theory,he has Borassa, Urban ResearchUnit, ANU, Tim Butler,
been vigorousin itsdefence,makingtacticalretreats Polytechnicof East London, David Ley and Neil
and concessions where necessary,but essentially Smith.The usualdisclaimers apply.
seeking to ensure its continuingcentralityin the
explanationof gentrification. But while Smithhas
acceptedthatchangesinthesocialandspatialdivision REFERENCES
of labourand the concentration of professionaland
managerialemployment inthedowntownareofcon-
siderableimportance, and has attemptedto integrate BADCOCK,B. (1989)'AnAustralian viewoftherentgap
Ann.Ass.Am. Geogr.79: 125-45
hypothesis',
theconsumption of
patterns gentrifiers into histheory, BEAUREGARD,R. A. (1984) 'Structure,
he has done thisin such a way thatit becomes a agency,and urban
redevelopment',in SMITH, M. P. (ed.) Citiesin trans-
functionalrequirement oflatecapitalism, ratherthana formation,vol.26,UrbanAffairs AnnualReviews(Sage,
recognitionof the role of individualpreference and London)pp. 51-72
agency.ButSmith'sinterventions inthegentrificationBEAUREGARD,R.A. (1986) 'The chaosand complexityof
debatehavenotbeencounterproductive. On thecon- in SMITH, N. and WILLIAMS, P. (eds)
gentrification',
trary,only by challengingthe so-calledchoice and Thegentrification
of thecity(Allen and Unwin,London)
preference theories and his advocacy of a logical, pp. 35-55
coherentalternative,has the debate over explan- BERRY,B. J.L. (1985) 'Islandsof renewalin seas of decay',
ationsadvancedas faras ithas.PreciselybecauseLey in Pederson,P. (ed.) Thenewurbanreality (The Brookings
Institute,
Washington, D.C.)
and Smithpioneeredradicallydifferent theoriesand
BODDY, M. (1976) 'The structure of mortgagefinance:
interpretationsof gentrification,it has been possible
buildingsocietiesand theBritish socialformation',
Trans.
to advanceourunderstanding of theprocessby see- Inst.Br.Geogr.N.S. 1: 58-71
ing how the two partialexplanationsfittogether.If CLARK,E. (1988) 'The rentgap and transformation of the
theirworkhasbeenshowntobe limitedincertainkey builtenvironment: case studiesin Malmo 1860-1985',
respectsand theyhave had to amendtheirexplan- GeografiskaAnnaler70B: 241-54
ations,thisis thepricepaid by theoretical pioneers. CYBRIWSKY, R. A., LEY, D. and WESTERN, J. (1986)
Neithermay have recognizedthe elephantof gen- 'The politics and social constructionof revitalized
trification
at first,
but theyeach identified a keypart neighbourhoods:Society Hill, Philadelphiaand False
of its anatomy,and otherresearchershave subse- Creek,Vancouver',in SMITH, N. and WILLIAMS, P.
oftheCity(Allen& Unwin,London)
(eds) Gentrification
quentlybeenableto piecetogethera moreintegrated pp. 92-120
explanation.As Clark(1988) concluded: DINGEMANS, D. J. (1978) 'Redliningand mortgage
in
We shouldstopaskingtheone-dimensional question: lending Sacramento,Ca.', Ana. Ass. Am. Geogr.69:
'Whichtheoryof gentrification is true,the rent-gap 225-39
theory, thepost-industrial
restructuration theory,the DUNCAN, J.and LEY, D. (1982) Structural marxismand
consumer demandforamenities or theinsti- humangeography:a criticalassessment.Ann.Ass. Am.
theory,
tutionalist andstart
theory?', asking'Ifitisso thatthere
is Geogr.72: 30-59
forallthesetheories,canwe arrive at GOETZE, R. (1979) Understanding neighbourhood
change
empiricalsupport
an understandingofthewaysinwhichtheystandina (Ballinger,Coalbridge,Mass.)
logicalrelation
ofcomplementarity?'
(p.247 emphasis in GREGORY, D. (1981) 'Humanagencyand humangeogra-
original). phy'.Trans.Inst.Br.Geogr.N.S. 6: 1-18
HAMNETT, C. (1984) 'Gentrification and residential
locationtheory:a reviewand assessment',in HERBERT,
ACKNOWLEDGEMENTS
D. T. andJOHNSTON, R.J.(eds) Geography andtheurban
Thispaperwas written whiletheauthorwas a visiting environment.Progress in research and applications,
researchfellowat theUrbanResearchUnit,Research vol. 6 (JohnWiley,London)pp. 283-319

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions
 Theblindmenandtheelephant 189
HAMNETT, C. and RANDOLPH W. (1986) 'Landlorddis- ROSE, D. (1989) 'A feminist perspectiveon employment
investment and housingmarkettransformation: the flat restructuring and gentrification:thecase ofMontreal',in
break-upmarketin London', in WILLIAMS, P. and WOLCH, J.and DEAR, M. (eds) Thepowerofgeography:
SMITH, N. (eds) Gentrification ofthecity(George Allen how territory shapessocial life(Unwin Hyman,London)
and Unwin,London)pp. 121-52 pp. 118-38
HAMNETT, C. and RANDOLPH, W. (1987) Cities,housing SCHAFFER,R. and SMITH, N. (1986) 'The gentrification of
andprofits(Hutchinson, London) Harlem?',Ann.Ass.Am. Geogr.76: 347-65
HARTMAN, C. (1979) 'Commenton neighbourhoodre- SMITH, A. (1989) 'Gentrification and the spatial con-
vitalizationand displacement: a reviewof theevidence', stitutionof the State: the restructuring of London's
]. Am. Plan.Ass. 45: 488-94 Docklands', Antipode 21: 232-60
HARVEY, D. (1974) Class monopolyrent,financecapital SMITH, N. (1979) 'Toward a theoryof gentrification: a
and theurbanrevolution.Reg.Stud.8: 239-55 backto thecitymovementby capital,notpeople',J.Am.
HARVEY, D. (1978) 'The urbanprocessundercapitalism: Plan.Ass. 45: 538-48
a framework foranalysis',Int.J.Urb.Reg.Res.2: 101-31 SMITH, N. (1982) 'Gentrification and uneven develop-
101-31 ment',Econ.Geogr.58: 139-55
HOYT, H. (1939) The structure and growthof residential SMITH, N. (1986) 'Gentrification, thefrontier,and there-
neighbourhoods in Americancities (Federal Housing structuring ofurbanspace,inSMITH, N. andWILLIAMS,
Administration, WashingtonD.C.) P. (eds) Gentrification of the city (Allen and Unwin,
KING, R. J. (1989c) 'Capital switchingand the role of London)pp. 15-34
groundrent:3. Switching,between circuits,switching SMITH, N. (1987a) 'Of yuppiesandhousing:gentrification,
betweensubmarkets, and socialchange,Environ. Plann.A social restructuring and the urban dream',Societyand
21: 853-80 Space5: 151-72
LEGATES,R. T. and HARTMAN, C. (1986) 'The anatomy SMITH, N. (1987b) 'Gentrification and therentgap', Ann.
of displacementin theUnitedStates',in SMITH, N. and Ass.Am. Geogr.77: 462-78
WILLIAMS,P. (eds) Thegentrification ofthecity(Allenand SMITH, N. (1987c) 'Dangersoftheempirical turn',
Antipode
Unwin,London),pp. 178-203 19: 59-68
LEY,D. (1978) 'Innercityresurgenceunitssocietalcontext', SMITH, N. and WILLIAMS, P. (eds) (1986) Gentrification of
mimeo, paperpresentedto theAAG AnnualConference, thecity(Allenand Unwin,London)
New Orleans STERNLIEB,G. and HUGHES, J.W. (1983) 'The uncertain
LEY, D. (1980) 'Liberalideology and post-industrial city', futureof the centralcity', Urban AffairsQuart. 18:
Ann.Ass.Am. Geogr.70: 238-58 455-72
LEY, D. (1981) 'Inner city revitalizationin Canada: a SUMKA, H. (1979) 'Neighbourhoodrevitalization and dis-
Vancouvercase study',CanadianGeogr.25: 124-48 placement:a reviewoftheevidence',]. Am.Plan.Ass.45:
LEY, D. (1986) Alternativeexplanationsfor innercity 480-7
gentrification: A Canadian assessment.Ann. Ass. Am. WALKER,R. and GREENBERG,D. (1982) 'Post-industrial-
Geogr.76: 521-35 ism and politicalreformin the city:a critique',Antipode
LEY, D. (1987) Reply:therentgap revisited.Ann.Ass.Am. 14: 17-32
Geogr.77: 465-68 WILLIAMS,P. (1978) 'Buildingsocietiesandtheinnercity',
MARCUSE, P. (1986) 'Abandonment,gentrification and Trans.Inst.Br.Geogr.3: 23-34
displacement: thelinkagesinNew YorkCity',inSMITH, WILLIAMS, P. (1984) 'Economic processes and urban
N. and WILLIAMS, P. (eds) Thegentrification ofthecity change:an analysisofcontemporary patternsofresiden-
(Allenand Unwin,London)pp. 153-177 Austr.Geogr.Stud.22: 39-5 7
tialrestructuring',
MOORE, P. W. (1982) 'Gentrification and the residential WOLFE, J.,DROVER, G. and SKELTON, I. (1980) 'Inner
geography of the New Class, mimeo,Scarborough cityrealestateactivityin Montreal:institutional charac-
College,Univ.ofToronto teristicsof decline',CanadianGeogr.24: 348-67
MULLINS, P. (1982) 'The 'middle-class'and theinnercity', ZUKIN, S. (1982) Loftliving:cultureand capitalin urban
J.Australian Polit.Econ.11, 44-58 change, (JohnsHopkinsUniversity Press,Baltimore)
MUNT, I. (1987) 'Economicrestructuring, cultureandgen- ZUKIN, S. (1987) 'Gentrification: cultureand capitalin the
a case studyof Battersea,London, Environ.
trification: urbancore',Am. Rev.Sociology13: 129-47
Plann.A 19: 1175-97
ROSE, D. (1984) 'Rethinkinggentrification: beyond the
uneven developmentof marxisturbantheory',Society
and Space2: 47-74

This content downloaded from 146.96.33.32 on Thu, 20 Feb 2014 09:57:57 AM

All use subject to JSTOR Terms and Conditions