ProgressinPlanning123(2018)1–31 ContentslistsavailableatScienceDirect Progress in Planning journal homepage: www.elsevier.com/locate/pplan Spatial inequalities and policies in South Africa: Place-based or people-centred? Alison Todesa,*,Ivan Turokb aSchoolofArchitectureandPlanning,UniversityoftheWitwatersrand,YaleRoad,Braamfontein,POWits2050,Johannesburg,SouthAfrica bHuman S ciencesResea rch Council,P leinParkB uil ding ,69-83PleinStr eet,P rivate BagX9182,Ca pe Town 8000 ,SouthAfrica A RTICLE IN FO A BSTRACT Articlehistory: Receiv ed11September2016 Thereisarobustinternationaldebateabouthowbesttotacklespatialinequalitieswithinnationsand Receivedinrevisedform25March2017 regions.Thepaperdiscussesthreecontrastingapproaches:spatialrebalancing,space-neutralandplace- Accepted27March2017 based.Theyvaryinthescopeandpurposeofgovernmentpolicy,fromredistributingeconomicactivity,to Available on line7A pril2017 facilita tinga ggre ga teg rowth ,and realisin gt heeconomic poten tialof less-develope dregions. Thepap er appliesthis framewor ktoanal yseS outhAfr ica’s fivedecad esofexpe rie nceofspatialp olicies.T heco ntext Keywords: isoneofstarkspatialinequalities,uneveninstitutionalcapabilities,andmountingpoliticalpressurefor Localandregionaldevelopment ch ang e. Under aparth eid,spatialt argeting washighly instrumenta lan dplayeda rolein reproduc ing Spatia l tar geting social d ivision s at consid erable financial cost . Since the end of apar theid t he re h as been much experimentationwith spatial initiatives, but without any overarching vision or policy framework. A cautionaryconclusionisthattherearerisksofextravagantspendinginmarginallocationswhenpolitical pressuresarestrong,publicinstitutionsareweakandeconomicdisciplinesarelacking.Anotheristhat place-basedpolicieshavepotential,butrequirestrongerverticalandhorizontalpolicyalignmenttostand anychanceoftacklingentrenchedspatialdivides.Enhancedlocalinstitutionsinvolvingprivatesector and commu ni tystakeh oldersarea lsoesse ntialfor spatialpol icies torespondt othespec ificchal lenges andopportunitiesencounteredineachplace. ©2017ElsevierLtd.Allrightsreserved. 1.Introduction processes, agglomeration economies and institutions (Barca, McCann, & Rodriguez-Pose, 2012; Glaeser, 2011; Pike, Rodri- Thereisarenewedbattleofideasaboutthebestwaytotackle guez-Pose, & Tomaney, 2012; Pike, Rodriguez-Pose, & Tomaney, spatial in e qu alities w ithin na tions and r egio ns ( Barc a, 2009; 2017; Stor pe r, 2013). T hese n otion s all stress the in fluence of McCann, 2016; OECD, 2009; World Bank, 2009). The contest geography on economic performance. The issues at stake are between different approaches has been spurred by heightened multi-dimensionalandcutacrossestablishedacademicdisciplines competitionforinvestmentinacontextofeconomicvolatilityand andpolicysilos. geopoliticaluncertainty.Apopularbacklashagainstglobalisation Somesimpledistinctionsareimmediatelyapparent.Ontheone in many lagging regions has added to the pressure for new hand,ithasbeenarguedthatgovernmentsshouldavoidsingling so lutions to une ven dev elop ment, refl ecte d in inte rnat ional out p ar ticul ar re gions f or sp ecial suppor t (‘spat ial ta rgeting’) commitm ent s to ‘le ave no-one b ehind’ in th e Sustainable bec ause it is more rel iabl e and e fficient f or marke t forces to Development Goals and other agreements. A series of other determine which places prosper (Cheshire, Nathan, & Overman, contemporary challenges also threaten regional prosperity, in- 2014; Glaeser, 2011; World Bank, 2009). Successful towns and cludinggrowingprotectionism,risingsocialinequality,disruptive cities will emerge more or less spontaneously and it is almost technol ogies, res ource scarcity and fi scal a usterity. M eanwhile, impos sible for go vernm en ts to turn-around local iti es whose newtheoriesofeconomicgrowthhavealsoinvigoratedthespatial economicbasehascollapsed.Theyshouldconservetheirresources policydebate,withtheiremphasisonendogenous(internal)causal and respond toplaces with proven demand for business growth andhouseholdpreferences.Astheseareasprosper,strongerlinks to p oorer reg ions throug h trade and migratio n (‘econ omic int egration ’) will s pread inc ome a nd n arrow the wealth gap * Correspondingauthor. (World Bank, 2009). Redirecting jobs and resources to under- (I.TuE-rmoka)i.l addresses: [email protected] (A. Todes), [email protected] perform ingre gionsw illmerelycu rbec onom icefficien cy, hamper http://dx.doi.org/10.1016/j.progress.2017.03.001 0305-9006/©2017ElsevierLtd.Allrightsreserved. 2 A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 growth and reduce aggregate welfare. Instead, governments space-neutralapproachcriticisethemforgoingagainstthegrainof should focus on‘spac e-neutral’(o r‘spatial ly-blind’) policieswhich marketforces ,hinderin gagglom eratio nt enden cies,ad din gtoth e targetpoorpeopleandequipthemwiththecapabilitiestoaccess taxburden,andjeopardisinggrowth(Cheshireetal.,2014;Nathan& opportunitieswherevertheyarise.Largecitiesaremostlikelyto Overman, 2014; World Bank, 2009). Proponents of place-based function as e ngines of grow th b ecause of t he advan tages of policies c riticise them for di fferent reasons – t hei r top-down, economicconcentrationanddensityforproductivityandinnova- standardised character, aiming to attract similar industries using tion (Lall, Henderson, & Venables, 2017; Wojan, 2016). In other identicalinstruments,andnotdoingenoughtobuilduponexisting words, there is a kind of inevitability to spatial outcomes that regionalassets,torenewlocalinstitutions,ortomoderniseoutdated govern ments sh ou ldn’t inte rferewithbe cau seit’sfu tiletoflyi nthe economi cstruc tur es(Bar caet al.,2012;Pik ee ta l.,2017).In deed,the faceofeconomicrealitybytryingtobuckthemarket(Gill,2010). European Union (EU) has steadily reduced support for physical Ontheotherhand,ithasbeenarguedthatgovernmentsshould infrastructure projects and the subsidies that governments are havee xpli citspa tialta rg etin gpoli ciesbec ause thereispoten tialfor permitted to o ffer firm s in orde rto preve nt wa steful beggar- my- growthinmanyregionsbesidesbigcities(Barca,2009;Dijkstra, neighbourbehaviourbetweenregionsandtostopthemfromtrying 2013; O EC D, 20 09; Sto rper, 20 10). Priva te firm s and markets to prop-up declinin g industr ies wit h sl im cha nces of su rvival cannot be relied upon to realise these opportunities because of (Garcilazo,2011;Turok,2004). inadequ ate infor matio n, risk-ave rsion, inertia or oth er failur es. TheEU’ snew place- basedpoliciesseektoboostdevelopment Collectiveactionbygovernmentsandcivilsocietycanimprovethe fromwithinbybolsteringindigenousstrengthsandbranchingout fortunes of places by creating conducive environments to guide intonewandrelatedeconomicactivities(Barcaetal.,2012;Barca, private investment decisions and to support productive activity. 2009; Hildreth & Bailey, 2014; McCann, 2016). They are not Governmentscanandshoulddomorethanrespondpassivelyto restrictedtothepoorestregions,andresourcesarenotallocated businesslocationchoicesandhouseholdmigrationpatternsafter accordingtostandardformulae.Strategiesaremorenuancedand theevent because inefficie ntfo rmsofurb andevelop mentma yget negotiated am ongstak eholderst oencourag egr eaterc reativity and locked-in. Policies should be sensitive to latent local assets and adaptabilityovertime.Theyarebroaderinscopethantraditional under-utilisedresources,andaddressthebindingconstraintsthat regionalpolicies,recognisingtheneedtoovercomediverselocal hamper investment and growth, such as weak institutions. A constraintsand unlock the unique potential of each place. Fresh growing literature o n ‘p lace-base d’ de velo pmen t suggests th at thinkingise ncou ragedb ym obilisin gdifferent so urces ofexp ertise, local and regional assets and know-how are foundations of energy and networks. A multi-level approach is important, national prosperity (Barca et al., 2012; Hildreth & Bailey, 2014). involvingnationalauthoritiesandsometimesinternationalbodies Places a re ‘sticky’ in that th ey attrac t and em be d prod uctive injecting ideas,res ourcesand disc iplines,such asensuring citizen investment, human capital and associated resources through accountability and regular evaluation to improve policy perfor- intenselocalinteractions.Thesesynergiescangenerateeconomic mance over time. External actors need to challenge entrenched dynamismbycreatingandstrengtheningcomparativeadvantage local interests that may have narrow, self-seeking agendas, through distinctive territorial capabilities, technologies and parochial tendencies or weak institutional capabilities (Barca, economic specialisations which are steadily upgraded and 2011;Boschma,2015;Tomaney,2013). enhanced overtime(Storp er,2013 ). Th epurpose ofthis paperisto examineSouthAfrica’s(SA)five This process of growth is always imperfect, uncertain and decadesofspatialtargetingpoliciesinthelightofthesedebates, uneven, as regions follow different paths. Effective governance, and using a three-fold framework of spatial rebalancing, space- leadershipandcollaborationamongstakeholderscanshapetheir neutralandplace-basedapproaches.Theintentionistocontribute developmenttrajectories.Jointproblem-solving,mutuallearning new insights into the theory and practice of spatial policies by andfocuseda ctiontoadd resss pecificlocalneed scanal lhelpto explo ringevid ence fro maqui tedi fferentco nt extbeyo ndEuro pe. fosterprogress.Placesfunctionasactiveagentstostimulateand SAisoneofthemostunequalandunevenlydevelopedcountriesin sustainproductiveactivity,andnotinertcontainersorreceptacles theworld.Italsoexperiencedaremarkablypeacefultransitionto for the locationde cisionso ffir ms (Barc a, 2011;Pik e etal.,2010, dem ocracy t wod ecadesago,i n dicatingsom esucces sinterms of 2017).Furthermore,economicintegrationthroughtradebetween national cohesion. Large parts of the country were deliberately regionsisnopanaceabecauseitmaywidenratherthannarrowthe under-developed historically and entire communities were forc- prosperitygapbyreinforcingthestrengthsofwell-endowedareas iblyremovedoffwell-locatedurbanlandtomarginalareasonthe anddepletingtheresourcesofotherterritories.Leadingcitiesmay periphery.Theblackmajorityofthepopulationwasdisempowered beco mediseng age d,or‘deco up led’,f romtheirsu rroundi ngreg ions educationa lly and le ft with i nf erio r healthcar e an d other public andmoreentwinedwithotherglobalcities,sothespreadeffects services.Acoercivemigrantlaboursystemwastheonlyoptionfor may neve rfilterthr ough (McCa nn, 20 16). E con om ic activ ityand manyme n toearna liveliho od.The entren ched so ciala ndspa tial politicalpowermaybecomeover-concentratedintheprimarycity inequalities have created enormous pressures for social redress andther esultin ginfl ationary pressuresandcon ge stio nmayac tas andspatialr ebala ncing.Th eycoincide withform idab leobs taclesto abrakeonnationalgrowth(Amin,Massey,&Thrift,2003;Martin, change, such as very uneven institutional capabilities and 2015).Migrationisalsonotasmoothorpainlessprocess.Itmay infrastructure in different places. Such immense geographical causesocialdislocationandinstability,sometimesinbothplacesof disparities cannot be resolved simply by devolving decision- originanddestination.Therefore,spatialpoliciescanhavevaluable making to disadvantaged regions or accelerating rural-urban nation alas wellasloca lbenefits, includi ngrealis ing theu ntapped migratio n.T hepaperanalys estheex peri enceofregion alpoliciesin potentialofless-developedareas,mitigatingthecostsofdisruptive the apartheid and post-apartheid eras, including industrial territorial divides, and relieving bottlenecks in over-heated decentralisation and special economic zones. Area-based initia- regions. tivesfocusedontheformerblackurbantownshipsanddecaying ThereisalonghistoryofspatialtargetingpoliciesinEuropean innercitiesarealsoassessed.Theevidencebaseincludesaunique countries. They have evolved from traditional efforts to steer collection of original programme evaluations, together with a investmen t and jobs f rom afflu ent to poor regio ns throu gh large range of s eco ndary lit erature and insights from interview s wi th financialinc entiv esan dmaj orinfrast ruc tures chemes ,towards more keyac tor s. complexregionaldevelopmentstrategiesmanagedbypartnerships Amixedpictureemergesfromtheanalysis.SAhasexamplesof oflocalandregionalstakeholders(Pikeetal.,2017).Advocatesofthe long-standing spatial policies driven by narrow ideological A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 3 objectivesthatgeneratedfewenduringbenefits,despiteoffering often weakest and least reliable in the places that most need someofthemostgenerousincentivesavailableintheworld.They support. One cannot assume the existence of an organised state neglected economic principles and lacked understanding of the capable of absorbing devolved powers and resources, devising underlyingproblemsorempathyforthecommunitiesconcerned. innovative plans and implementing agreed decisions. Special Therearealsomorerecentexamplesofwell-conceivedinitiatives efforts are necessary to build and maintain competent and thatwerehamperedbypoorgovernmentcoordinationandvision, accountable local and regional governments, with oversight and resultinginduplicationofeffort,inconsistencyanddissipationof supportfromcentralgovernment. resources. A third group of initiatives have had surprisingly The balance between national and local action ought to vary positiveoutcomesininauspiciousplaces.Thisisattributablemore dependingonthecircumstances.Additionalnationalinvestment, to determined implementation than to sound design. Capable enhancedtechnicalassistanceandamoreassertiveroleintackling organisationswithenergetic,well-connectedleaderswereableto governanceshortcomingsmayberequiredwherespatialdivides mobilisesubstantialpublicandprivateinvestment. are deep. A targeted approach that concentrates on a limited Sever alwiderles sonsfo rspa tialpol icyalsoemergefromSA’s num bero fp lacestobe ginwithi smor erealisticand cre di blethan eclectic ex perien ce. Spat ial t argetin gis no tinh erently ineffic ient spreadin g experti se widel yand th inly. National spa tialstrat egies and ineffective. However, it is notnecessarilystraightforward or and arrangements toalign sectoral policies and funding streams successfuleither,andthereisconsiderablescopeforimprovement. arealsoimportant.Theseandotherpolicyrecommendationsneed Generalisationsaboutthesuperiorityofparticularkindsofspatial tobefoundeduponamuchmoresubstantialbaseofresearchand policiesareproblematicbecausesomuchdependsonthenational evidence. andlocalcontext,andontheconditionsofimplementation.Spatial Thestructureofthepaperisasfollows.Chaptertwoprovidesa policiesarepotentiallyvaluablemechanismsforcoordinatingand review of the literature on spatial targeting. The third chapter focusing government action, yet they need to be grounded in examines the context of SA, including the changing spatial economicrealitiesandinstitutionalcapabilities.Spatialinitiatives economy and historical evolution of national spatial policies. aremostusefulifreinforcedbyotherstatepowersandresources, Chapterfouranalysesthehistoryandcontemporaryexperienceof andimplementedinpartnershipwithotheractors. regional policies, from efforts to deconcentrate industry under T hreeparticula ris sueswarran tmor eatten tionintheliterature aparthei d to rece nt sp ecial ec on omic zones. Th e fifth c hapter onspatialtargeting.First,allspatialpoliciesareconditionedbythe focusesonnationalarea-basedpoliciespost-apartheid,including structureanddynamicsoftheeconomy.Itmakesabigdifference schemestodeveloptownshipeconomies.Chaptersixconcludesby whethert hee conomyis op en tonewent er prise,c o mpe titionand drawing th e evide nce toge ther and r eflecting on the wid er diversity, or whether ownership, control and know-how are implicationsforcontemporarydebatesaboutspatialpolicies. concentrated and there are many barriers and rigidities in the system.Trans form ingsp atial traject oriesism uch morediffi cu ltin 2.Internationalexperienceandconceptsofspatialpolicies a relatively stagnant, unchanging economy. Second, social cohe- si onmakes iteasiert ointroducea ndsustain spatial policie s.SA’s 2.1.Introduction experienceisthatlaunchinginitiativesinverydeprivedenviron- ments is di fficu lt because commun iti es a re fracti ous and Lookingbackoverthehistoryofspatialpolicies,andattherisk institu tion s fragile . Develop ment efforts need strategie s for of over-sim plifi cation , i t is po ss ible to identify th ree br oad building trust, stability and shared agendas. Third, institutional approaches,eachwithadifferentrationaleandcharacter.Table1 capacityisrequiredatlocalandnationallevels.Statecapacityis summarises their essential features. Spatial rebalancing aims to Table1 DifferentApproachestoSpatialPolicy. Spatialrebalancing Space-neutral Place-based Goalsand Narrowtheprosperitygapbetweenregions Aggregateeconomicgrowthvia Eachregiondevelopstoitspotentialby objectives andreduceunemploymentinpoorerareas. agglomerationeconomiesandeconomic buildingamoredurableanddynamiclocal integration. economy. Mechanisms Steerinvestmentandjobsfromleadingto Facilitateeconomicdensity,scaleeconomies Strengthenlocalassets,know-howand laggingregions. andconnectivity. institutions. Attractforeigndirectinvestment. Removebarrierstomigration. Developmoreproductiveandinnovative enterprises. Policy Standardfinancialincentives. People-focusedpublicservices. Holisticandintegratedstrategiesbasedon instruments Improvedphysicalinfrastructure. Connectinginfrastructure. developingdistinctivehumancapabilitiesand Streamlinedbusinessregulations. Deregulationtoexpandhousingsupplyin novelactivities. fast-growingcities. Styleof Centralisedandpredictable. Nationalinstitutionsanduniversal Responsivecityandregionalgovernment. government Specialpurposeagenciestoexpedite programmes Externalpartnerships. implementation. tomeettheessentialneedsofpeopleand Multi-levelgovernance. fir ms. Economic Staticbenefitsofbusinessrelocationto Ongoingbenefitstoeconomicgrowthviathe Dynamicbenefitsderivedfromlocal rationale lower costreg ion s.Possibl ereduction in efficienc yandpro du ctivityde rivedfro m experime ntation, learning ,adap tationand overheatingandcongestionincoreregions. agglomerationeconomies. differentiation. Socialand Socialsolidarityandpoliticalstability. Popularacceptanceofurbanisationand Buildstrongandresilientfoundationsfor political unevendevelopment. sharedandlastingprosperity. rationale Whereisthe Relativelypoorregionsandlocalities. Largecities. Everyregionandlocality. policy focus? Source:Authors’creation. 4 A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 narrow the prosperity gap between regions, i.e. to promote Areas were designated within which companies could receive economicconvergence.Themainmechanismistosteerproductive sizeableincentivesfromcentralgovernmentinreturnforcreating investment and jobs from prosperous to poorer regions, and to andmaintainingjobs.Newindustrialestatesandadvancefactories attractforeigndirectinvestment.Financialinducementsarebased werealsobuilttoattractmobileindustry.InUnitedStates,amore on standard rules and eligibility criteria to ensure predictability comprehensiveapproachwasfollowedintheTennesseeValley,an andcertaintyforinvestors.Improvementsinphysicalinfrastruc- agriculturalregionalsodamagedbytheDepression.Agovernment ture also play a part in accommodating mobile manufacturing agencywasestablishedthatinvestedheavilyininfrastructure(e.g. plan tsan doth er majo rp rojects.Theshort -runeco nomicimpacts electric ity g eneration a nd fl ood con trol), te ch nical support (e.g. aremo stim porta ntfor policy-m aker sandfirm sbenefitf romthe developing and man ufact uring fertilizer s) and at tracting new lowercostsofdoingbusiness,suchaslabourandland. industriestotheregion.Italyintroducedaregionalpolicyinthe Thespace-neutralapproachaimstomaximiseaggregategrowth 1950s to reduce poverty and stem out-migration from the through increased e fficiency. It is as sumed th at an exp anding depres sed South o f the co untry . It inc luded major su bsidie s for economywillnarrowspatialdisparitiesbyspreadingincomeand industrial relocation and investment in new roads, irrigation investment from the buoyant core to surrounding regions. The schemesandotherinfrastructure. logicistofacilitateeconomicconcentrationandeconomiesofscale Thesepoliciessoughttorebalanceinequalitiesbytransferring in a fe w big citie s through labour migr atio n and con ne cting economic activit y from a ffluent to poorer re gio ns, and by inf ra struc ture .Policy instrum entsare designedt ores pondtoand stimulatin gdeman dinde pressedre gio ns,influe ncedbyK eyne sian reinforcethese ‘natur al’processes by people-ce nt redactio ns and economic i deas. Th e emphasis subsequ ently shift ed in many urban infrastructure, rather than to steer business in contrary countriestoattractingforeigndirectinvestment,usingthesame directi ons. The econ omic b enefi ts are su pposed to be more policy ins tru ments. Fo llowing simila r thinking, some cou ntries dynamicthanfromspatialrebalancing,andstemfromtheongoing soughttorebalancethenationalurbansystembysupportingthe cost savings and productivity gains associated with economic growthofsecondarycitiesandtownsinordertorelievecongestion densityandagglomeration.Theoutcomeisexpectedtobeahigher inthe biggestcities(Parr,1999).Inallcasestheyreliedonlarge rateofn atio nalgrowthand hig heravera ge incomes . fin anc ial tran sfers betwe en re gio ns orche strat ed by ce ntral The place-based approach is concerned with improving con- government. In many countries there were physical controls ditions in a wider group of reg ions by he lping to realise their placedonfir ms inthep rosperous areas intend edtorest ricttheir potentialthroughdevelopmentfromwithinratherthantransfers expansion and spur them to shift activities to less-favoured fromelsewhere.Thepriorityistostrengthenthedistinctiveassets, regions.Manufacturingplantswerethefocusofattentionbecause knowledgeandhumancapabilitiesofeachterritoryandsupport of their perceived locational mobility. Manufacturing also com- upgrading and diversifi cation into ne w an d more s oph isticated pr ised m ore manu al occupat ions than services, which help ed to activities.Policyinstrumentsaretailoredtothelocalcontextand absorb unemployment among less-skilled workers. Constructing geared to promoting indigenous enterprise and innovation infrastructure was also labour-intensive and provided an addi- throug h le arning and collaboration . The econo mic benefits are tionalstimulu stoj obcr eation. derived from the emphasis on creativity, experimentation and Amajorcriticismwasthesomewhatuninspirednatureofthis differentiation,i.e.undertakingmorecomplextasksandenhanc- approach, which often failed toengenderself-sustaining growth ing the qualitative character of growth rather than more of the (Barcaetal.,2012;Pikeetal.,2017).Thenewactivityrarelybecame same.Theidealoutcomeistheemergenceofadistinctivegrowth anchored in the region and did little to overcome structural traject ory forea chregion so ast omoderate ze ro -sum,or‘h ead-to- rigidities an d ob solesce nce. It e ncour age d investme nt in the head’,com pe tition betwe en re gio ns. region,ra ther thanall-round de velopmentan dtransform atio nof These are clearly generalisations that obscure variations theregion,i.e.anarrowsectoralratherthanabroaderterritorial withineachpolicyconceptandsimilaritiesbetweenthem.There emphasis. The same policy formula was followed in different are also distinctions between the spatial scales of different places, and generally managed bycentral government to ensure inte rvent ions that cu t across the three ap proach es. Regional consist ency . Beneficia ries were oft en matu re assembly pl ants or development policies tend to focus on the economic and branch factories seeking state subsidies, but lacking product infrastructure dimensions of development, while local, or area- design,marketingorothervalueaddingfunctions.Littletechnol- based, programmes stress the broader human development ogyorknow-howweretransferredtotheregionandfewspillovers aspects, such as community well-being and living conditions. weregenerated.Thenewjobsweresecureaslongasthefactories Thelogi cisth atr egionalpolic iesrelatemo recl oselyt ofunctional opera tedefficien tly, prod ucts alesw erehe alt hyan dt hes ubsidies economic units, particularly metropolitan labour markets or were sustained. Yet the strategic capabilities and skill-sets to commuting catchment areas. Local development relates more conduct their own research, development or marketing for the closely to t he scale o f every day l ife, including the ter ritory nextgen eratio nofp roductsw eremissing.I tw asdifficult forl ess- covered by peo ple’si mm ediatesoc ial n etworksan d jo urneysto favou red region s to mode rnise withou t the entrepre ne urial, shoppin g, s chool or recreation. This d istinction is no t fixed a nd technolog icaland fina ncialcompe tenciesto build companieswith definitive, just a s t he differen ces b etween reg io nal and a rea- theirownpro duc ts,proces sesandservic es. basedpoli cies sho uld n ot beexagg erated. Th ere w ere vario us econom ic j ustifications provided for this approach. One argument for redirecting activity to depressed 2.2.Spatialrebalancing regions was to offset overheating and congestion in the labour and housing markets of the core region (Armstrong & Taylor, Thefirstgenerationofspatialpolicieswasdrivenbysocialand 200 0; Kaldor ,1970). D ec entra lisati onwou ld reduce in fla tionary political pressures to reduce the gap between rich and poor pressures and enable interest rates tobelowered,which would regions. The dive rsio n of s tate reso urces wa s in tende d to benefitth ewh oleecon omy.An other arg um entwast hateco nomic demonstrate solidarity and promote national cohesion, as well and political power wasconcentrated in the leadingregion and astoimprov ematerial cond itionsin thetarge tareas(Pik ee tal., drai ning the rest o f the country of v ita l hu man an d finan cial 2017). A regional policy was introduced in the UK when heavy resources(McCann,2016).Theseideaswereneverfullyaccepted industriesinthenorthweredevastatedbytheGreatDepressionin in most countries. For example, in 1983 the UK government the1930s,causingmassunemploymentandsocialunrest.Assisted famouslydeclaredthattherewasnoeconomiccaseforregional A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 5 policy.1 The size of the incentives and eligible regions were geographicalfocusestablishesa‘special’status,includingstream- promptly sc aled do wn. There was a p arallel c oncern w ithin linedregulat ions,e xtrafinanc ia lbenefit s,enha ncedlogis ticsand Europe aboutever-increasingstatesubsidies beingextracted by greatervisibilityfordecision-makers. major corporations playing regions off against each other, or China has considerable experience of using SEZs to stimulate being misused by g overnm ents to p rot ect dom estic industr ies large-sca le in dustrialisatio n. Sleepy c oa stal to wns an d fishing fromexternalcompetition(Turok,2004). villagessuch as Shenzhen enjoyed inherentgeographical advan- Meanwhile, research within a political economy tradition tages when they were designated as priority locations to start suggestedthatregionalpolicywasreinforcingtendencieswithin manufacturingexportgoodsduringthe1980s.Thecombinationof multinatio nals todisagg regate their variousfunc tionsandse parate simplified pro cedure s, spe cial ta x b reaks, exe mptions fro m themphysicallyacrossregionsinanewspatialdivisionoflabour customs duties and large reserves of low cost migrant labour (Massey,1995). It made business sense for them to locate their provedenormouslyattractivetoforeigninvestmentandlaunched routinep roduct ion inlow ercost,p eriphe ral areas, esp ecially with China’s remarkable transfor ma tion fr om an agr arian society theavailableincentives.However,theadditionaljobsoftenproved (Miller, 2012; OECD, 2013; World Bank, 2014). The special transient because the plants lackedthe wherewithal toadaptto attributes of SEZs were reinvented over time as newareas were changing market conditions, or could be induced to move designatedinordertogivethemadistinctivecharacterandnovel elsewhere by mor e generous i nce ntives. The pursuit o f a ‘quick- advantages ov eroth er areas .Chin a designedSE Zregimes that were fix’ also m ean t tha t little wa s done to spur the for ma ti on and tailored to the cont ext in whic h they w ere introdu ced and gro wtho fnew, local ly-bas eden terpr ises . integrat edi ntow iderecon om icstrat egies. Thesta te’scommitm ent A different case was made for targeting particular industries to bold experimentation meant piloting far-reaching reforms of withi nthelagg ingr egion s.‘Gro wth poles’w ereabouti mplanting na tiona lregulationstom aximise theimp actofSEZs. propulsiveindustriesthatwouldgeneratelargemultipliereffects A similar determination to adapt zone regimes to their because of their backward and forward linkages. Having a focal particular regional circumstances and deliberately learn from pointonwhichtoconcentrateinvestmentandserviceswouldbe the experience has not been apparent to the same extent more effi cienta nd haveabigge rcatalyticef fect thanspre adingt he else where.Cons eque ntly, SEZsh avegenera ted very mixed results effort through outt here g ion.Th isconcep twas influe ntialinm any around the world (Farole , 201 1). Th e globalis ation of tra de and partsoftheworldfromaroundthe1960s,althoughitneverlived investment,enabledbythedisaggregationofmanufacturinginto up to its promise (Parr 1999; Storper,1991). Identifying the key global production networks and value chains, has been an ind us trie s in adv ance was a difficul ty. Pu rsuing a co nce rted import anttrendin theirfavou r.Ag ainstth ishasb eena tende ncy strategyto bu ildconste llatio ns offirmsar oundthem w asanother to see SEZ s as s ep arate initiativ es or p rivile ged encla v es, rather problem.As a result,the transplanted operations developed few thanpartofbroadereconomictransformationstrategiescovering local link age s and re ma ined ‘cathedr als in the desert’ (Ha rdy, skills dev el opment, regional clusters and s upply cha ins, and 1998). In several developing countries, the drive to launch new public-privatepartnerships. capitalcities,andtoshiftthelocusofpowerfromthecoasttothe Spatialtargetingtookonaquitedifferent,morelocalisedform interior ,were just ifi edon the basis th atthis wou ldst imula te the in many a dvanced econ om ie s from around the 1 960s with the economy of neglected regions, in line with a simple version of discovery of areas of concentrated poverty in the inner cities of growth pole theory. However, the new urban centres ended up large industrial conurbations. Many governments in Europe and mostly accommodating state administrative functions and the the US became more concerned with treating social and widereconomicimpactswerenegligible(Cain,2014).Manyturned environmentalproblemsinrun-downneighbourhoodsthanwith outtobeexpensivemistakesthatdivertedsubstantial resources steering growth (Cochrane, 2007; Musterd & Ostendorf, 2008; from mo re valuabl e econom ic a nd socia l projects elsewhere Oakley & Tsao, 2 006). Inten sified i nternatio na l competitio n and (Parnell & Simon, 2014). Nevertheless, the idea of using the under-investmentwerecausingmanyfactoryclosuresandleading physicalpresenceofgovernmenttostimulateregionalgrowthwas to vacant and derelict land and buildings. Working class an important principle with potential for more creative applica- communities were most vulnerable to the job losses and tion. It has resurfacedin the UK recently in recognitionthat the experiencedrisingpovertyandexclusionfromtherestofsociety highly centralised system of government spending and political (Jargowsky,1997; Wilson,1997). There were social protests and control militates against places outside London and acts as a street riots in many European and US cities as governments ‘counte r-regional policy’( Martin, 2015;M cCann,2 016) . seeme d to ret reat f rom their post -Wa r com mit ment to full Theconceptsoffreetradezones,exportprocessingzonesand employment, and as households slipped through the safety net enterprise zones differ from growth poles in their less discrimi- of the welfare state. Immigration from the former colonies nating approach to industrial selection. These special economic contributedtoanatmosphereofracialtensionandmistrust. zones (SEZs) date back at least 50 years and have become Thethreattonationalstabilityandpoliticallegitimacyforceda increasinglypopulararoundtheworld(Farole,2011).Ratherthan government response, especially as local authorities found it targetafew sectorsf orprior itya ttentio n,they embrac earan geof difficulttosk ewtheirst retchedres our cesto wardspoore rdistric ts. activit ie stha tfulfilc ert ainmini mumrequ ireme nts,such a shavi ng In the UK the Urba n Program me was launche d and the US ahighexportcontent.Theirfocusonparticularlocalitiesenables equivalent was called the War on Poverty. The focus was on investmentinmoderninfrastructuretobeconcentratedinplaces marginalisedneighbourhoodssufferingfrommultipledeprivation withbettergrowthprospectsthantherestoftheregion,suchas and social stress, including low income, poor education, sub- seaportsandairports.Hencesomeofthemshouldprobablynotbe standardhousing, ill-health, familybreakdownandrising crime. classified un der the spatial rebal an cing r ubric. Their res trict ed Extra res ources w ere provid ed to e nhance loc al sc hools, health facilities, social servicesand policing. Joint working was encour- aged bet ween teachers , soc ial worke rs, h ealth offi cials , police office rsandoth erprofess ionalst otreatthe proble msfacedb ypoor 1 Suchargumentshavere-emergedinrecentyearsfollowingthewideningofthe families is a mor e holistic way . M any are a-based in itiativ es also North-So uthdivide( Mart in,2015;Mc Ca nn,201 6).Th epointbe ing madeisth at the UK’srecentg rowth hasbeen based ontoona rrowa foun datio nofa fewin du strie sin made efforts to involve communities in decision-making through local partnerships, and to introduce new approaches to social afewregions,namelyLondonandtheSouth-East.Thisisunstableandwastefulof ta lent andres ourcesi nother regio ns. regen erationthrou ghco mm unity-bas edorg anisations. 6 A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 Yettherewasatendencytoperceivelocalproblemsindepen- genuinelyshapespatialoutcomes.Someeconomistsalsoargued dently of the economy or public policy more generally. Urban that balanced or equitable development is misguided because policy was treated as a special instrument designed to address growthisinevitablyunevenanddrivenbypowerfulmarketforces selected aspects of under-performance within poor neighbour- whichcannotbereversed(Cheshireetal.,2014).TheWorldBank hoods.It offered som ecompensationfor theirec onom icproblems (2009) pointe dly criticised theEU’sr egi ona lpolic yon thegro unds andfortheinabilityofmainstreampoliciestoensurecommunity thatitpromoteddispersedratherthanconcentrateddevelopment, well-being. However, there was little attempt to address the therebyunderminingproductivityandinnovation,anddepressing structural and systemic causes of poverty by rebuilding local growth. economies and creating jobs. Observers criticised area-based TheWorldDevelopmentReport(WDR)(WorldBank,2009)put projects for being too fragmented, reactive and neglectful of forward a clear alternative to spatial targeting. It advocated a sustaina bles olution s(C ochrane,200 7;Lawles s,201 2;Musterd & ‘spatially -b lind’o r‘place-neu tra l’appro ach,i.e.po lic iesthatapp ly Ostendorf,2008;Syrett&North,2008;Turok,2008).Atworstthey to all locations. These focus on meeting the essential needs of merely ameliorated poor living conditions, instead of being households,suchascleanwater,sanitation,energy,healthcareand catalyst s for socio- econom ic de velopment. There w asn ’t the education.T hese ‘pe ople- centre d’policies ensuret hatthefat eof politicalappetitetotestmoreradicalpolicyreformsortodeliver individualsisnotdeterminedbywheretheyhappentobeborn, solutionsatanythinglikethescaleoftheSEZsinChina. because they acquire the capabilities to get jobs elsewhere. The Some for ms of are a-ba sed targe tin g al so fo cus ed on physical outcome will beeffici ent andequitabl eb ecau sep eoplewillm ove structures,including the condition of the housing stock, derelict tothemostproductiveplaces,therebyreducingunemploymentin landandbuildings.Run-downareaswereperceivedtoencourage theiroriginalareasandsatisfyingtheincreasingdemandforlabour anti- socia lbehaviou randtod eterp rivate investmen t. Dampand atthe irdestin ation s.Th isspatial‘ adju stment’m echanism is saidto decayinghousingwaslinkedwithill-health,insecurityandother operate relatively smoothly and freely. Urbanisation reinforces socialproblems.Improvementsinphysicalstructureswerehighly agglomerationeconomiesandtherebyfuelsproductivity,knowl- visibleandrelativelyeasyforgovernmentstoengineer.Thishelps edge creation and aggregate growth, in line with the New to explain the higher levels of support for physical renewal and Economic Geography paradigm (Glaeser & Joshi-Ghani, 2013; investment in bricks and mortar. The assumption was that Glaeser,2011;WorldBank,2013). improved liveability would stabilise communities and reduce There is a secondary role for public policy to enable the thesocialmalaise.Yetindependentresearchquestionedthesocial expansion of the most successful cities byremoving bottlenecks and econo micbene fits ofphysicalim provem entsinthea bse nceof andbarrie rs. This mean srespondi ngto gro wthpress uresasthey broader economic progress and deeper social reforms (Lawless, emerge.Forexample,land-useplanningcontrolsthatrestrictthe 2011; Lupton, 2003; van Gent, Musterd, & Osterndorf, 2009). supplyofhousingshouldberelaxed(Cheshireetal.,2014).New Furthermore,well-locatedneighbourhoodswerepronetogentri- investmentisalsorequiredtoexpandthetransportnetworkand fication as t hey were up graded and exi sting reside nt s were otherurban in frast ructure.I nd uecours eci tieswillgr owtoap oint displacedbybetter-offincomers. wherethediseconomiesofscaleexceedtheadvantages.Thiswill There was a broader concern that urban initiatives treated lead naturally to deconcentration as core costs escalate and placesin isolat io n,neglec tingthere gion alconte xtandinte ractions cons traintson labo ursupplyandlan dp romp tfirms toreloca teto withsurroundingdistricts(Jargowsky,1997;Wilson,1997).They secondarycitiesandtowns.Thecentripetalforcesofconcentration were inward-look ing and didn’t do en ough t o conne ct resi dents will in du e cou rse genera te c entrifugal t enden cie s to disperse with wider opportunities (Andersson & Musterd, 2005; Syrett & activity.Itisfutiletotryandreviveformerindustrialcitiesorto North , 200 8; Turok & R obson, 200 7). There w as ins ufficie nt tackle p hy sic al der eli ctio n an d deca y in run -down ar eas. P eop le attention paid to the sequencing of actions and mitigation of matternot places, and people aremobile, sotheyshould be the gentrifica tion.R ath erl iketradition alr egionalp olicy ,therewas an focus o f su pport. The fair wa y to do this is thro ugh un ive rsal implicit assumption among senior decision-makers that urban education,healthandotherwelfareservices(WorldBank,2009). policyw aszero-sum (merely conce rnedwithshufflin gres ources These pro mote n atio nal co hesion and soci al stab ility b ecause and op por tunities be tween a reas), with no impact on national everyo ne benefi ts, where as spatial poli cies are divisive because prosperity(Turok,2008).Thelackofaclearlyarticulatedeconomic theyfavoursomecommunitiesoverothers. rationaleconsistentlyheldbackgovernmentsupport.Theoutcome Thecentralargumentof theWDRisthateconomicgrowthis wasmanyshort-term,piecemealinitiativespromotedbyseparate inevitably unbalanced, but it can still be inclusive (World Bank, agencies,oftenwithunwarrantedfanfare.Thisprecludedamore 2009). Inclusive growth can be attained through economic integratedapproach,includingstrongerlinksbetweenurbanand integration of leading and lagging regions, which enables trade, regionalpolicies,andacrossthesocialandeconomicdimensionsof resource transfers and migration. A similar logic applies to development. inequalities within cities. Run-down and deprived neighbour- hoodsreflec ttheso rtinge ffectsofthe hou singmark et(Cheshire 2.3.Thespace-neutralapproach etal.,2014).Peoplewithlowskillsaresusceptibletounemploy- ment and end up living in areas with the lowest quality, least A space-neutral perspective emerged partly because of the desirable housing. The appropriate solution is to improve their diffic ulties facing redistributiv e spatial polic ies in a m ore skillssot heycanc omp etemoreeffe ctivelyin t he labourm arket, competitiv e globa l environment with str essed pu blic fin ances. notto sp end large sumstry ingto revitaliset he par tsofth ecityin Mobilecapit aland talentmadeit more difficult toanch orprivate whi ch they l ive. T argeti ng dist ric ts for ent erp rise a nd eco nom ic investment within territories and threatened escalating public development will simply mean that jobs get displaced from subsidies a s gover nments b id fo r investme nt and sac rificed neighbouring areas ,withn oreduc tioni nloca lun employmen tand environm ent al and welfare stan dard s in a ‘race to th e bottom’. atsubstantial public cost (E iniö&Ove rm an,2 016). There were growing criticisms that scarce taxpayer funds were TheWDRalsoadvocatesgrowth policies atregional, national beingwastedonunviableprojectsinplaceswithlittleprospectof andinternationallevels(seealso,WorldBank,2013).Thepriority self-sustainingdevelopment,givenhoweasyitisforgovernments is to create spatially-blind institutions, including universal public tothrowmoney atsymbolicin itiativ es.T heyl ac kt he knowledgeor se rvi cesand lawstoensur eefficientla ndandlab ourmark ets.As foresight to anticipate the myriad individual decisions that growth takes off and urban centres emerge, transport and A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 7 communications infrastructure is needed to connect places and paths.Evidencefromaroundtheworldshowsthattherelationship facilitate agglomeration and economic specialisation. Spatial betweenurbanisationanddevelopmentisalsohighlyvariableand targeting policy (benefit ting a specific area) should only be thatcitie sdonotneces sari lydriveprospe ri ty(M cCann &Acs,2 011; introducedmuchlaterontotacklethemostintractablelocalised OECD, 2006; Turok & McGranahan, 2013). Local and national problems. It sho uld be “u sed sp arin gly s ince this i s where circum stances determ in ehowurbanisa tionun foldsa ndth ebalance misallocat ion is mos t li kely” ( Deichmann , Gill , & G oh , 2011; betweenitspo sitiveandn egat iveeffects.Cit y-level plan ning isvital p.167).Spatialtargetingisonlyrecommendedforhighly-urbanised inanticipatingandpreparingforgrowthbycoordinatingbusiness, countries with very divided cities and large regional disparities householdandpublicinfrastructureinvestmentdecisionstoensure (WorldBank,2009). acoherenturbanform. The messa ge for low and middle income countries is that Tosum up,eco nomists’recognitionofgeography’scontribution regionalinequalitiesareinevitablesincegrowthfocusesonafew todevelopmentisanimportantshiftfromthepreviousblindspot. placesw ithinherent adv antages.Ye twith accesst odecen tsc ho ols, Ho wever, the s pa ce- neutral ap proac h ne glec ts the in fluen ce of health facilities and other public goods, less-developed regions local institutions, the qualitative character of growth, and the willcatchupandpovertywillfall.Thesemarket-driventransitions enduring economic imbalances between places. These realities taketimeandrequirepatientpeople-centredpolicies.Thereisno complicatethenotionthatcitiesnaturallyfuelinclusiveeconomic alternative since special economic zones, big infrastructure growth. Government policy is given a circumscribed role, with projects an d oth er ‘quic k fixes’ ar e likely to become white unduef aithintheab ilityof ma rketfo rc estoreducein equa lities. elephants. Policies that siphon off resources and jobs from dynamic cities will dampen aggregate growth by undermining 2.4.Place-basedpolicies agglomeration,reducingproductivityanddeprivingurbanecono- miesofthepublicinvestmenttheyneedtofuelfurtherexpansion. Recognition of the pitfalls of relying on external resources to Se ve ral criticis ms can be level led a t t his analysis . First, the revitaliseregio nal eco nomies ha spromp ted serious reflection by space-neutral approach assumes that most government policies theEU,OECDandotherorganisations(Barca,2009;OECD,2006, have uniform geograph ical imp acts, desp ite evidence to the 200 9).T heem ergin gpoli cieshavebeen influen cedby thecon cepts contrary. Many policies that are supposed to be spatially-blind of endogenous growth, human capital, knowledge-based econo- haveunequalimpactsandoutcomesbecauseofdifferencesinthe miesandinstitutions(Barcaetal.,2012;Pikeetal.,2017).Place- conditions of implementation between areas, especially human basedpolicycontrastswiththenon-interventioniststanceofthe capital and institutions. The delivery of effective schools, clinics space-neutralapproachandtraditionaltop-downspatialpolicies. andbas icse rvicesalldep end oncapabl elo calautho ritiessta ffedby There is an a ssumptio n th at governm ents can influen ce local competentprofessionals.Thiscannotbetakenforgrantedinless- economic trajectories if their policies arecarefully-designed and developedregions.Nationalinnovationpolicieshaveveryuneven well-executed by resourceful, competent and accountable insti- impactsbe causeth e capacit yof firms and univ ersiti esto absorb tutions(Morga n, 2016).Region alandloca lide ntitiesalsom atter, thissupportvariesgreatlybetweenregions(Morgan,2016).Major implying that particular places and communities cannot be left government facilities are located disproportionately in big cities behindinthepursuitofnarrowideologies. becauserecruitingscarceskillsismucheasier.Militaryspending, A fundamental pillar of the place-based approach is that strategic installations and national research institutes also have development strategies should be tailored to and embedded unequalimpacts.Thesedifferencestendtoaccumulateovertime within their geographical context (Barca, 2009; OECD, 2009). andreinforcespatialdivisions,unlessdeliberateeffortsaremade Neglecting contextual diversity is a serious criticism of previous tooffsetandreversethem(Martin,2015;McCann,2016). models.Place-basedpoliciesshouldbuilduponlocalcapabilities Second,aspace-neutralapproachisunequivocalaboutthegains andresourceendowmentsinseekingtogrowandbranchoutinto fromeconomicintegrationbetweenregions.Itassumesthattrade newdirections.Localassetsandknow-howprovidethefoundation and factor mobility are equalising forces. This neglects historic on which jobs, incomes and prosperity are most likely to be evidence that if places have unequal endowments at the outset, sustained,buttheyalsoneedlong-terminvestment,upgradingand connecting them may enlarge spatial inequalities (Armstrong & injectionoffreshideasandtechniquestoundertakemorecomplex Taylor,2000;Pikeetal.,2017).Dominantregionsmaydrawresources tasks over time. Space and location shape the development frompoorerareasandpullfurtherahead.Compensatingtransfers possibilities of particular territories and the life chances of couldbeoutweighedbyawideningdivisionoflabourbetweenthem. individuals. The spatial context matters in ways that go well Local economies dominated by low value activities face systemic beyondphysicalgeographyandnaturalresources.Social,cultural barriers that hamper upgrading. Spatial inequalities may be andinstitutionalcharacteristics(suchasthequalityofregulations, cumulative, with divergence more likely than catch-up (Amin government capacity, local leadership and learning capabilities) etal.,2003;Martin,2015).Similarpointsapplywithincities,where also have a bearing on the rate and character of economic enclavesofconcentratedpovertymaypersistwithoutcountervailing development(Barcaetal.,2012;Morgan,2016). interventions.Perhapsthekeypointisthatanexclusivepolicyfocus Strategies should address the distinctive growth constraints onpeopleislikelytowidenthedifferencesbetweenplaces,because facingdifferentplaces,andaimtoutiliseandenhanceuniquelocal of the sorting effects of people with different incomes choosing attributes, untapped knowledge and ideas, and other under- wheretheyliveandleavingless-advantagedgroupsbehind,trapped employed resources. Regions may have geographical advantages inexclusionaryspaces. that could be exploited more effectively, such as a gateway Third, the approach is underpinned bya simple causal model location, regional service centre or tourist attraction. Astute whichholdsthaturbanisationinlowincomecountriesaccelerates development organisations may be able to improve or develop indust rialisa tion, which raise s p rod uctivity, creates w ealth and the market b y raising fina nce to provi de patient r isk capital, reducespoverty.Becausecitiesareenginesofgrowth,thebiggerthey support long-term business decisions, or stimulate productive are,thebetterfo rprosperity(Woj an,2016). Overtimet hebenefits activity informstha twouldn otoccursp ont aneously.T heymaybe extendoutwardstotheperipheryandspatialgapsarereduced.There abletofacilitatethediffusionofknowledgeandtechnology,and isakindoflaw-like,physicaldeterminismtothiswayofthinkingthat aligntheskillsandcompetencesrequiredbybusinesswiththose ignores the diversity of regional economies and trajectories. supplied by colleges and universities. These tasks require Countries,regionsandcitiesdonotfollowthesamehistoricgrowth integrated approaches rather than narrow sectoral programmes, 8 A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 piecemealincentivesordiscretepiecesofinfrastructure.Thisalso necessarilylessproductivethanbigcities,andmega-citiescanbe necessitate s governm en ts makin g the sp atial dimension of t heir particularly dif ficult to go vern (Di jkstra, Gar cilazo, & M cCan n, policiesandinvestmentprioritiesmoreexplicit(Barca,2009). 2013;OECD,2006).Largeandsmallcitiesmaybothunderperform Asecondprincipleoftheplace-basedapproachisthatdecision- becauseofinertiaorvestedinterestsholdingthemback.Govern- makingshouldnotbelefttonarrowlocalinterests.Nationalpolicies ments should not neglect smaller cities and towns, but rather and templates are unlikely to be responsive to local conditions engagewithlocalinstitutionstoincreasetheirimpact.Onestarting becausecentralgovernmentistoodistantanddistractedbyother pointistotacklephysicalimpedimentsorbureaucraticconstraints matters.Yetlocaldecision-makingisvulnerabletoparochialismand thathinderinvestment.Tryingtoidentifyspecialisedactivitiesin elite capture,orweddedtothe past.Selectedlocalinterestsmay which localitiesarebestsuitedis alsoworthwhile.Thisrequires engageinopportunisticbehaviourattheexpenseofthemajority. buildingonlocalcapabilitiesandsupportinginnovativeideasby Urbanlandisparticularlysusceptibletospeculationandotherforms combininginsightsfromdifferentactors.Unlockingthepotential of rent-seeking that inhibit coherent development. Established ofallcitiesandregionswillcontributemoretonationalprosperity property owners may resist additional housing or industrial thanfocusingonafewbigcities(Farole,Rodriguez-Pose,&Storper, development in their areas out of self-interest. Elected leaders 2011;Wojan,2016). maybepreoccupiedwithshort-term,partisanagendasandfactional confl ict s. The idea s of partnership , collab oration and mutual 2.5.Conclusion accountabilitycanreducesomeoftheserisks. Inaddition, som eofthe know le dgean dtechnologyrequiredfor Tackling uneven development is more difficult in a highly localdevelopmentisnotreadilyavailableandmustbeproducedby competitiveandvolatileglobaleconomysubjecttotechnological consulting external actors. Successful development needs deci- disruptions,migrationpressures,constraintsonpublicinvestment sion-makingtobeopentonewinsights,andbasedondialogueand and popular disaffection. Decision-makers face many complex debate(Barcaetal.,2012).Opennesstoexternalinputscanhelpto dilemmas about whether totarget people or places; tofocus on prevent insularity and cosy consensus. Engaging diverse stake- areas of greatest need or development potential; to build on holders across the business sector, civil society and government existingeconomicstructuresorbranchoutinnewdirections;to can res olve p rob lems, bu ild con fiden ce and ha rness wider promote change f rom the ce ntr e or as sist ini tiativ es to emer ge expertise and networks. Different constellations of commercial, fromthegrassroots;andtoplanforthefutureorrespondtomarket technical and political interests need to generate shared under- processes.Spatial policies havegraduallyshiftedover time from standings asa foundati onforinv esting th eirresour cesand know- steeringin vestme ntthroug hfix edincentiv esand hard infras truc- howinlocaldevelopment(Morgan,2016).Thisrequirescoopera- ture,towardsgrowingfromwithinbydevelopinglocalassetsand tionandcoordinationofdifferentpolicies,organisationsandactors know-how.Someobserversadvocateaspace-neutralapproach,in inpursuinganagreedstrategicagenda,andperhapsevenasocial which the priority is to target poor people and improve their contract. Promoting common values, building partnerships and capabilities, thereby playing down territorial differences. Others encouraging a sense of shared destiny and community can also favour a place-based approach, in which localities pursue their reduceopportunismandspeculation. developmentpotentialbyinvestinginunder-usedresources.The Mul ti-level plann ing, decision-making and co-financing are former assum es that gov ernments a re weak an d make p oor alsoimportanttoenablelocal,regionalandnationalperspectives economicdevelopmentdecisions.Thelatterassumesthatgovern- tobefactoredintopolicychoices.Nationalandregionalauthorities ments canacquirethe capabilitiestoshape economicoutcomes. ca n p rovide r esou rces a nd tech nical sup por t for loc al capacity Thepla ce-b asedap proa chisnewer and lesse asytodefi nebecause building.Theycansetparametersgoverninghowpublicfundsare it combines different policy mechanisms, depending on local spent, in order to prevent wasteful competition, duplication of economicandsocialrealities.Thereisasurprisinglackofrobust effort and malpractice (Barca, 2011; Boschma, 2015; Tomaney, empirical research to evaluate the validity of many of these 2013).Theymayinjectnewinsightsandadviceonpracticalpolicy argumentsandpropositions. instruments from development experience elsewhere into the local process. External authorities can also provide checks and 3.SpatialinequalitiesandpoliciesinSouthAfrica balancestoimprovetransparencyandaccountabilitytocitizens,to bolster c ivi c leaders hip and to b uild investor confid en ce. Last ly, 3.1.Introduction thereisanimportantroleformulti-levelarrangementstoenhance national understanding of local conditions in order to sensitise The territory of SA is one of the most unequal and visibly theirpoliciestoneedsontheground. polarisedintheworld(NPC,2012).Formanyyearsspatialdivisions Ofcoursemulti-levelgovernanceandcross-sectorcoordination were deliberately engineered and often brutally imposed. Geo- aredi fficultt oachievebe causepolicy fra gmentationa ndcompart- graph ical disparit ies original ly ar ose fr om the way in w hich me ntalised w orking are dee p-seate d in public bure aucracies colonialin stitutionsa ndpractice sexpl oitedt hec ountr y’s natural worldwide. Two way s of reducing the se division s are through resource s. The indig enou s black A frican2 p opul ation was forced transversal area-based or issue-based institutional arrangements. intoruralreservesandtheiraccesstourbanlivelihoodswasstrictly Eithercanprovideastrategicframeworkandintermediarybodies controlled.Residentialsegregationpoliciescreatedracially-divid- that – ove r time – b ring grea tercoheren ce to a range ofs ectoral ed cities an d towns, w ith unequal access to jobs a nd amenities. policies,instrumentsandinitiativessothattheycomplementand From 1948 the apartheid government enlarged these spatial reinforceeachother.Thepackagingofdifferentinterventionscreates synergies whic henh ance theirindivid ualcontr ibutions.Itisd ifficult enoughtoshiftthetrajectoryofalocaleconomywithoutpolicies contradictingeach other. Priorities are boundtovary in different 2 TheuseofracialcategoriesinthispaperreflectsprevalentrealitiesinSouth localitiesdepe ndin gonth eexisting leve lofcoo per ation ,th equality Africa,a ndi sn otme anttocond on eth em.Th eaparth eidcateg orieswer e‘ white’ oflocalin stitutionsa nd the nature ofthe c hallengesfac ed. (peopl eofE ur ope andesc en t),‘colour ed’(pe ople of‘mixed race’),‘Asia n’or ‘Indian’ (people of Asiandes cent)and ‘black’(pe opleofA fr icande scent). Inthis pa per,the Recognition that institutions matter for development means term‘bl ack ’isus edtorefe rtop eopleo fAfrica na ndAsian originan dt hose of‘mi xed that physical geography and urban size cannot determine race’, while t he te rm ‘bla ck Africa n’ refers o nly to pe ople of A frican d escent, locat ional per formance o n the ir own . Sm aller cit ies are not follow ingcu rren tCens usclas sification s. A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 9 fracturesthroughlawsimposingseparateurbandevelopmentand ThischapterprovidesanoverviewoftheSAcontext,examining establishing the rural reserves as self-governing, ethnically- theevolutionofunevendevelopmentandcontemporarytrends.It defined hom elan ds, or ‘bantust ans’ . Entire commu nities were also considers h owthe debateovers patia lpolicyhasch anged in forciblyremovedoffprimeurbanlandandrelocatedtoperipheral the apartheid and post-apartheid eras. Subsequent chapters areas. Land and space were used as cruel instruments of social examinetherangeofexplicitspatialpoliciesinmoredetail. separation and subjugation. Industrial decentralisation policies were introduced to create jobs in and around the bantustans in 3.2.Fromcolonialismtoapartheid ordertocontainmigrationpressuresandtherebykeeppeopleof different races apart. The logic resembled spatial rebalancing in During the eighteenth and nineteenth centuries, the spatial thateffortsweremadetosteermobilemanufacturingplantsfrom economydevelopedaroundDutchandBritishcolonialismcentred thecitiestowardstheselocations. on seaports, agriculture, trade, administration and military T hefir stdemoc raticg overnmentelectedin1994wasfacedwith act ivities in a few large t owns li nked to a netw ork o f smaller a stagn ant economy, a fiscal crisis, entrenc he dsoc ial a nd sp atial centres.T he di scove ryofg oldint helaten in ete enthcen tur yshifted inequalities, and a high risk of political instability. Economic the focus to Johannesburg and its hinterland (now Gauteng activity was concentrated in the metropolitan areas, but almost province),whichquicklybecamethedominanteconomichub.In half th e bla ck African p op ulat ion lived in and a rou nd the subsequen tdecad es,maj orreinve stm entandec onomicd iversi fi- bantustans with sparse opportunities. The government faced an cation occurred within the same region, as a few powerful agonising c hoice – whe ther to suppo rt pr ivate investm ent a nd corpor ations es tablishe d m anufac turing, ban ki ng a nd other employmentgrowthwhereitwasalreadyestablished(inlinewith tertiaryactivities linked to mining (Harrison & Zack, 2012). This space-neutra lideas), ortop r omo teredre ssandredis trib utio nby comman ding nex us bec am e conso lidated as a ‘min eral-en ergy favouring dev elopme nt in marginal ised are as. I n practice, it h as complex’ (Fin e & R ustomjee , 1996) that dom in ated the entire generallyeschewedthisdilemmaanditsspatialpolicyhasbeen nationaleconomy,buttressedbyveryconsiderablestatesupport. ambiguou sasaresu lt.Th ispartly reflec ts suspicio nsabo utsp atial Although industria ldiversifica tio nocc urredduring thetw entieth policyfollowingthedeplorablepracticesofthepast,alongwiththe century,ownershipandcontroloftheeconomyremainedhighly all-enc ompassin gfr ameworko ftherulin g par ty,wi thits assor ted concentr ated,withf ourc onglom er ates controllin gthelion’ sshare values,factions,interestgroupsandpatronagenetworks.Theweak ofeconomicactivityin1994(Black,2016;Philip,Tsedu,&Zwane, systems of long-term planning and coordination within govern- 2014). mentarealsoresponsibleforthispredicament(NPC,2012). The growth of mining also created powerful pressures that Asaresult,placeshavebeentreatedreasonablyeven-handedly undermined the black African peasantry and reinforced uneven in terms of investment in economic infrastructure. In contrast, development. Black Africans had previously been forced into expenditureonmostsocialprogrammeshasbeenskewedtowards wagedlabouronfarms,butthegrowingdemandforlabouronthe ruralareas,r efl ecting highe rlevelsofpo vert yand needfo rpublic mines led to mo re dra coni an m easures (Turok, 201 4). Th e 1 913 servic es (se e below). There has be en no exp licit polic y to wards Land A ct l aid the b asis for sp atial segre gation by con finin g the rural-urban migration or the management of urbanisation. In blackAfricanpopulationtoonly13%oftheland.Thismarginalised addition,urbanland,spaceandlocationhavebeendealtwithina peasantagricultureandremovedcompetitionwithwhitefarmers, muchmorepassiveandreactivemannerthanunderapartheid.A therebyforcingblackAfricansintowagedlabour.Menmigratedto range of spatial programmes have been introduced, but in a theminesonatemporarybasis,whiletheirfamiliesremainedin fragm ent ed and inconsistent way. The government’ s br oad ly the reserv es. Th ishelpedt okeep wage slow andboo stedprofi ts, neutral stan ce t owards SA’s territo ry h as avoided the serious alth ough con ditio ns in th e res erves soon de teriorate d from dislocat ionand damage toco mmunitie sinfl ictedint hep ast,and overpopu lation on u npr oduc tive land . The 1923 Native Urban ensuredbasicsocialstability.Yettherehasbeennosustainedeffort AreasActcastblackAfricansastemporarysojournersinthetowns torectify orre engin eerthes pati alleg acy ofsys tem aticexclu sion, and s tren gthe ned i nflux con tr ols, thereby reinforcin g the cheap or to shift the path-dependent pattern of economic growth migrant labour system. It also introduced the idea of residential towards, say, a more employment-intensive trajectory (Bhorat & segregationofdifferentracialgroupsbypreventingblackAfricans Mayet,2013;Black,2016).Indeed,somesectoralpolicies,suchas frompurchasingorrentinglandinwhiteareas. free housing, have paradoxically exacerbated inherited spatial From the mid-1920s, a new government representing white divis ionsbyc onfini ngpoorhouse holdstochea pperipher alland. working clas s and agric ult ural interests inst ituted measu res to They hav e t herefore r einfor ced the bur de n of an ‘apartheid tax’ promote indus triali sationande levatethe positiono fpoorwhi tes imposedonpoorblackhouseholds. andAfrikaners.Itputinplaceimportsubstitutionprogrammesand In the absence of an overarching rationale and policy createdagroupofstate-ownedenterprises,particularlyinsectors framework,spatialinitiativeshavehadvariousobjectives,includ- linked to the mineral-energy complex, such as iron, steel, ing attracting private investment, steering state spending, chemicals and electricity. Subsequent governments extended encouraginglocalenterpriseandimprovingpublicservices.There thesepoliciesandestablishedanIndustrialDevelopmentCorpo- has also been an ongoing debate between advocates of a space- ration (IDC) to accelerate industrialisation. The focus of these neutralposition(implyingafocusonbigcities)andproponentsof largely-successfulendeavourswasonnationaleconomicdevelop- various forms of spatial targeting (including more concerted ment.Therewasnoconsiderationgiventoregionalpolicyuntilthe programmes focused on rural areas or urban townships). In the 1940s,despitethehardshipanddeterioratinglivingconditionsin vacuum of national spatial policy, particular initiatives have thereserves. emerged th rough ad vocacy by spec ific groups and indiv iduals C ontrolsonthemovementofblackAfricanstothecitiesbegan withinor closetog overnmen t.T heyhave beenin fluen cedbothby to break d ow n d uring the 19 40s, and the S mu ts co alition internationalideasanddomesticexperiences.Manyprogrammes governmentbegantoexplorealternatives.Rapidindustrialisation, have taken on hybrid forms in practice, including elements of rising urban wages and strong migration pressures prompted a spatial rebalancing and a place-based approach. There has been range of public bodies to advocate industrial dispersal from the considerable experimentation, with much to learn from the cities.Severalgovernmentcouncilsinvestigatedthedesirabilityof accumulatedexperience,bothpositiveandnegative. regional policies as awayof limiting black African urbanisation. However,thereportswereequivocalandindustrialistsarguedfor 10 A.Todes,I.Turok/ProgressinPlanning123(2018)1–31 freerpopulationmovementtothecitiestomeettheirincreasing within the language of growth poles and growth centres. These demand for labour. The IDC established several projects in rural policieshadseveralimportanteffects,althoughthedepartments areas,closetonaturalresourcesorsourcesoflabour,butmostofits dealingwithregionalpolicywererelativelyweakwithingovern- invest ment s w ere in urban are as (Glaser , 1 987). H enc e pri or to ment (O ranje & Merr ifield, 2010) , and the ir effo rts we re over- 1948,regionalpolicywaslimitedandverypiecemeal. shadowed by more powerful processes. For example, national AnewNationalistgovernmentwaselectedin1948,represent- economic policy gave top priority to import substitution, which ingtheinterestsofthewhiteworkingclass,whiteagricultureand tendedtoreinforcetheconcentrationofindustrialactivityinthe Afrikanerbusiness.Itintroducedtheoppressiveapartheidsystem majorcentres,wherethemainmarketsforconsumerproductsand andextendedrace-basedspatialpolicies,includingtheelaborate intermediategoodswerelocated. syst emofbant ustans,toug hinflu xcontrols regulated thr oughpass Economic growt h beg an to slow in the late 1960s, reflecting laws, and segregated neighbourhoods. Most bantustans were deteriorating international conditions and the internal contra- isolatedfromthemaineconomiccentres(Fig.1),althoughseveral dictions of apartheid, including skill shortages and thin local bordered onl arge citie sandtow nssuch asD urb an,Preto riaand markets (Ge lb,1991).S A’srace-ba sedre gimeimpe deda djust ment East London. This created fragmented settlement patterns with toamoredynamicgrowthpathfoundeduponhigherproductivity longdistancesbetweenplacesofhomeandwork.Withincitiesand and superior skills (Bhorat, Cassim, & Hirsch, 2014). As material towns, racial segregation was imposed by law and separate circumstances in the bantustans worsened, many black Africans residen tial ar eas known a s ‘to wnships’ w ere cre ated for black devisedwayso fm ovi ngtothecit ies,andpop ularre sistan cetothe Africans, coloureds and Indians, generally on the edge of cities. apartheidsystemincreased.Thestaterespondedwithaseriesof Some th ree million peo ple who did not fit t he b luepr int were ‘reforms’, includi ng strong er a ttemp ts to stee r jo bs to t he forcib lyrem ovedfro murba nare asor evi cted fro m‘white’ farms homeland s. This di d not qu ell the po liti cal op positi on, and (Turok, 2014). Townships were developed as dormitory spaces mounting economic problems eventually brought about the through large public housing projects, with limited low-order collapse of apartheid, ushering in extended negotiations for retailandsocialfacilitiesandbasicpublicservices.BlackAfrican change.Aftertwodecadesofeconomiccontraction,international entrepreneurshipwasrestrictedbythestate,evenintheseareas. sanctionsandorganisedoppositionathomeandabroad,theruling Influxcontrolswe rein tensifiedt oli mit access bybl ack Afric ansto National Part y accepted a negotia te d tran sitio n to d emo cracy. townsandcitiesandthenumberofpeoplearrestedforpasslaw Many authors have argued that this settlement gave excessive offencesincreasedtoabout750,000ayearbythemid-1960s. concessions to established economic interests, which prevented Regional policy inthe form of industrialdecentralisationwas more far-reaching social progress in subsequent years (Bundy, used tosupportapartheid from1960, althoughit hadadditional 2014; Habib, 2013; Hart, 2013; Philip et al., 2014). The counter- objectives as well, as the following chapter explains. The 1975 argumentisthattheeconomywasinserioustrouble,civilwarwas NationalP hys icalD eve lopm entPlanw asthec ountry’sfi rsts patial possiblea nd the rew asarisk ofla rge -scalec apitalfl ight. The key plan.Itwasbasedonaspatialrebalancingkindoflogic,couched pointisthatthecircumstancesofthe1990spoliticaltransitionhad Fig.1. FormerHomelands,CurrentProvincesandMetropolitanMunicipalities.