Environment and Urbanization

147EnvironmentandUrbanization,Vol.10,No.1,April1998GuidetotheLiteratureI.INTRODUCTIONTODATE,MOSTdevelopmenttheoryandpracticehavefocusedoneither“urban”or“rural”issueswithlittleconsiderationoftheinterrelationsbetweenthetwo.Bycontrast,severalempiricalstudiesshowthatthelink-agesbetweenurbancentresandthecoun-tryside,includingmovementofpeople,goods,capitalandothersocialtransactions,playanimportantroleinprocessesofruralandur-banchange.Withintheeconomicsphere,manyurbanenterprisesrelyondemandfromruralconsumers,andaccesstourbanmar-ketsandservicesisoftencrucialforagricul-turalproducers.Inaddition,alargenumberofhouseholdsinbothurbanandruralareasrelyonthecombinationofagriculturalandnon-agriculturalincomesourcesfortheirlive-lihoods.Thispaperreviewssomeofthere-centliteratureonrural-urbaninteractions,withparticularattentiontothewaysinwhichtheyhavebeenaffectedbyrecentandcur-renteconomic,socialandculturaltransfor-mations.Thepaperisorganizedasfollows:thefirstthreesectionsdiscussdefinitionsofruralandurbanareasandactivities,reviewconceptualframeworksandconsiderhowrural-urbaninteractionsareconceptualizedwithindevelopmentplanning.Thelastfoursectionsreviewempiricalstudiesondiffer-entflowsconnectingruralandurbanareas(flowsofpeople,ofgoodsandofwastes),andonsectoralinteractions(agricultureinthecities,non-agriculturalemploymentinthecountrysideandrural-urbaninterlinkagesinperi-urbanareas).II.DEFINITIONSTHEDISTINCTIONBETWEEN“rural”and“urban”isprobablyinescapablefordescrip-tivepurposes;however,itoftenimpliesadi-chotomywhichencompassesbothspatialandsectoraldimensions.Incensusesandothersimilarstatisticalexercises,ruralandurbanpopulationsareusuallydefinedbyresidenceinsettlementsaboveorbelowacertainsize;agricultureisassumedtobetheprincipalactivityofruralpopulationswhereasurbandwellersarethoughttoen-gageprimarilyinindustrialproductionandservices.Inreality,however,thingstendtobefarmorecomplex:thewaysinwhichna-tionsdefinewhatisurbanandwhatisruralcanbeverydifferent;theboundariesofur-bansettlementsareusuallymoreblurredthanportrayedbyadministrativedelimita-tions,especiallywhentowns’useofruralresourcesisconsidered;populationmove-ment,especiallytemporaryandseasonalmigration,isnotusuallyreflectedincensusfiguresandcanmakeenumerationsofruralandurbanpopulationsunreliable;finally,alargenumberofhouseholdsinurbanareastendtorelyonruralresources,andruralpopulationsareincreasinglyengagedinnon-agriculturalactivities.a.DefinitionsofUrbanandRuralAreasDemographicandeconomiccriteriaonwhichdefinitionsofurbanandruralareasarebasedcanvarywidelybetweendifferentnations,makinggeneralizationsproblematic.InthePhilippines,urbanareasaredefinedRural-urbaninteractions:aguidetotheliteratureCeciliaTacoliCeciliaTacoliisaResearchAssociateattheInternationalInstituteforEnvironmentandDevelopment,whereshecoordinatesacollaborativeresearchprojectonrural-urbaninteractionsinAfrica,AsiaandLatinAmerica.TheauthorwishestothankDavidSatterthwaite(HumanSettlementsProgramme,IIED)andSimonCroxton(SustainableAgricultureandRuralLivelihoodsProgramme,IIED)fortheirhelpfulcommentsandsuggestions.Address:IIED,3EndsleighStreet,LondonWC1H0DD;e-mail:[e-mail]148EnvironmentandUrbanization,Vol.10,No.1,April1998GuidetotheLiteraturebythenationalcensusasallsettlementswithapopulationdensityofatleast500personspersquarekilometre.Theurbanstatusalsoappliestocentreswiththefollowinginfra-structure:•aparallelorright-angledstreetpattern;•atleastsixcommercial,manufacturingorsimilarestablishments;•atleastthreeofthefollowing:atownhall,churchorchapel;apublicplaza,parkorcemetery;amarketplaceorbuildingwheretradingactivitiesarecarriedoutatleastonceaweek;andapublicbuildingsuchasaschool,ahospitaloralibrary.Barangays(administrativeunits)withatleast1,000inhabitantswherethemajorityarenotoccupiedinfarmingand/orfishingarealsoconsideredurbancentres(PhilippineNationalStatisticsOffice,1992).InBenin,theNationalInstituteofStatis-ticsandEconomicAnalysisconsidersastownanyheadtownofadistrictwithapopulationof10,000inhabitantsormore,andwithatleastfourofthefollowing:postoffice,taxof-fice,publictreasury,bank,runningwatersupplies,electricity,healthcentreandsec-ondaryschool.Populationdensityandtheproportionofnon-agriculturalactivitiesarenotconsidered(Tingbé-Azalou,1997).Thisisoftenthecaseinsub-SaharanAfrica,wheresmalltownsaredefinedonthebasisofad-ministrative,demographicandinfras-tructuralcharacteristicsevenwhenthemajorityofthepopulationengagesinagri-culturalactivities(seeGadoandGuitart,1996,onNiger).Exceptionstorules,how-ever,includeSenegal’smainreligiouscen-tre,Touba,whichiseffectivelya“sacredsite”ruledbythereligioushierarchyandwhereIslamiclegislationprevailsoverstatelegisla-tion.Indeed,Toubaisstillclassedasavil-lagedespiteanestimatedpopulationofover300,000whichmakesitthecountry’ssec-ondlargestsettlement(Gueye,1997).Asiaremainsapredominantlyruralconti-nent,withtwo-thirdsofitspopulationlivinginruralareasin1990.However,ifbothIn-diaandChinaweretochangetheirdefini-tionofurbancentrestoonebasedonarela-tivelylowpopulationthreshold-asusedbyma