1978 Bingham - Innovation, Bureaucracy, And Public Policy - A Study of Innovation Adoption by Local Government

8/12/2019 1978 Bingham - Innovation, Bureaucracy, And Public Policy - A Study of Innovation Adoption by Local Government

1/29

University of Utah

Western Political Science Association

Innovation, Bureaucracy, and Public Policy: A Study of Innovation Adoption by LocalGovernmentAuthor(s): Richard D. BinghamReviewed work(s):Source: The Western Political Quarterly, Vol. 31, No. 2 (Jun., 1978), pp. 178-205Published by: University of Utahon behalf of the Western Political Science AssociationStable URL: http://www.jstor.org/stable/447811.

Accessed: 30/11/2012 15:17

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].

.

University of Utahand Western Political Science Associationare collaborating with JSTOR to digitize,

preserve and extend access to The Western Political Quarterly.

http://www.jstor.org

This content downloaded by the authorized user from 192.168.82.203 on Fri, 30 Nov 2012 15:17:35 PMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/action/showPublisher?publisherCode=utahhttp://www.jstor.org/action/showPublisher?publisherCode=wpsahttp://www.jstor.org/stable/447811?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/447811?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=wpsahttp://www.jstor.org/action/showPublisher?publisherCode=utah


2/29

INNOVATION, BUREAUCRACY, AND PUBLIC POLICY:A STUDY OF INNOVATION ADOPTIONBY LOCAL GOVERNMENTRICHARD D. BINGHAM

University f Wisconsin-MilwaukeerT HE OMPLEXITIES ofmodernife,ncreasingffluence,ocialproblems,and many other factorshave dramaticallyincreased the size and respons-ibilitiesof local governments.' To cope with their added responsibilitiesmunicipal bureaucracieshave adopted new technologies nd proceduresto provideadditional servicesand/or to reduce spiralingcosts. Yet there have been few sys-tematic examinations of influenceson innovation adoption by local governmentalagencies since Mohr's studyof health agencies in four states and one Canadianprovince conducted in the early 1960s.2 Earlier studiesof innovation in local gov-ernmentwere largelyconcernedwith diffusion r geographic spread3 and usuallyignoredreasonsfor nnovationadoption. Thus, therehas been no nation-wide tudyof the determinants f innovationadoption in cities,norany comparisonof innova-tion adoption among typesof local governments e.g. school districts s. municipalgovernments).This is a studyof innovationadoptions by housingauthorities, chool districts,public libraries, nd municipal governmentsn American cities with 1960 popula-tions of 50,000 or more utilizing he innovation-decisionmodel suggestedbyDownsand Mohr.4 For each unit of analysistwo innovations are examined: one productinnovation,definedas an innovation requiringthe adoption of a physicalproduct,and one process innovation,an innovation requiringa change in method. TableI liststhe innovationsstudied.The goals were to identify nd weigh the factorswhich determine nnovationadoption in local government, o examine the relationshipbetweenpublic policiesand innovationadoption in local bureaucraticorganizations, nd to studythe link-age sequence or processofinnovationadoption.

PROBLEMS WITH THE INNOVATION LITERATUREWhile the literatureconcerningthe adoption of innovation by local govern-ment is limited,there is a substantialbody of literatureconcerningadoptions byothertypesoforganizations. The multi-disciplinarytudyofinnovation has createdsome confusion. The diffusion f innovations s discussed in hundreds of articles,many using theirown terminology.Fortunately, everal authors have successfully

NOTE: The authorwishes o thankJamesJ.BlascovichndBrettW. Hawkins or ommentson an earlier ersion f thismanuscript. he researcheported erewas supportedya grant rom heNationalScienceFoundation.The conclusionsnd recommendationscontained n thispaperare those fthe author loneand do notnecessarilyeflectheofficial osition f the National ScienceFoundation.The author lsowishes o thankIrwinFeller,J.David Roessner,nd RobertYin for uggestionselated othevisibilityofinnovations.For a summaryee NicholasHenry, ublicAdministrationnd PublicAffairsEnglewoodCliffs:Prentice-Hall,975), pp. 282-87.2LawrenceB. Mohr, Determinantsf Innovation n Organizations, merican oliticalScience Review 63 (March 1969): 111-26.For example, . StuartChapin,CulturalChange (New York: Century,928); EdgarC.McVoy, Patterns f Diffusionn the UnitedStates, American ociologicalReviezw5 (April,1940): 219-27; Robert . Crain, Fluoridation: he Diffusionf an Innova-tionAmong ities, ocialForces 4 (June1966): 467-76.4GeorgeW. Downs,Jr., nd LawrenceB. Mohr, Conceptualssues n theStudy f nnova-tion, AdministrativecienceQuarterly1 (December 976): 700-714.

http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp


3/29

Innovation,Bureaucracy,and Public Policy 179TABLE 1. INNOVATIONS XAMINED N THIS STUDY

Type ofgovernment Process innovation Product innovationPublic housing The application of computer The use ofprefabricated omponentsauthorities systems o housing authority in high-rise onstructionmanagementPublic school The use of a system findi- The adoption ofvideotape recordersdistricts vidualized instruction n by secondaryschoolselementary choolsPublic libraries The application of computer The adoption ofa theft etectionsystems o public library systemmanagementMunicipality The use ofautomaticdata ContractingwithMainstem, nc. forprocessing equipment by vehicle maintenancemanagement

police departments service

drawn together hisdisparatebody of literature.5Because of differing oals, differ-ent concepts and measures,and diversemethodologies, t is difficult o generalizeabout innovation adoption fromone type of organization to another. In otherwords,researchfindings o nottendtobe cumulative.Anothermajor probleminvolves the relationshipof innovation to the conceptof newness. The debate over newness as part of the concept of innovation ismulti-disciplinary.Many investigators o not require that a processor productbenew to be innovative,onlythat itbe new to theorganizationunder study.6Walker,forexample, suggeststhat innovations need onlybe new to the government r or-ganization consideringthem.7 However, others ike Mohr require that the innova-tionsbe completelynew.8Whether Walker's or Mohr's conceptionis used can make a criticaldifferenceforresearch. For example, Hawley9 found that cities witha high concentrationofpower (as measured by a ratio of managers, proprietors, nd officials o the em-ployed labor force) showed greatersuccess in urban renewal endeavors than thosewitha diffuse ower structure.Clark, however,reportedthe opposite.10He foundthat decentralized decision making correlates with urban renewal expenditures.ReconcilinghisfindingswithHawley's, Clark suggested hatfragility,ooselydefinedas newness,may play an importantpart. Clark suggestedthat fragilepolicies orinnovations are oftendifficult o adopt - that active opposition by even a smallgroup of dissidentsmay preventaction on an issue. Clark reconciledhis differenceswith Hawley by suggestingthat Hawley studied urban renewal adoption when itwas new and divisive while he studied renewal later, when the issue was less' EverettM. Rogers and F. Floyd Shoemaker,Communicationof Innovations: A Cross.-Cul-tural Approach, 2nd ed. (New York: Free Press, 1971); Gerald Zaltman, Robert Dun-can, and JohnnyHolbek, Innovationsand Organizations (New York: Wiley, 1973).6 Everett M. Rogers,Diffusionof Innovations (New York: Free Press of Glencoe, 1962), p.13; Kenneth E. Knight, A Descriptive Model of the Intrafirm nnovation Process,JournalofBusiness40 (1967): 463; JeraldHage and Michael Aiken,Social Change inComplex Organizations (New York: Random House, 1970), pp. 13-14; Neal Gross,Joseph B. Giacquinta, and Marilyn Bernstein, mplementing Organizational Innova-tions: A Sociological AnalysisofPlanned Educational Change (New York: Basic Books,1971) p. 16.'Jack L. Walker, The Diffusionof Innovations Among the American States, AmericanPolitical Science Review 63 (September 1969): 881.' Mohr, DeterminantsofInnovation, 112.9Amos H. Hawley, CommunityPower and Urban Renewal Success, in Community truc-ture and Decision-AMaking: omparative Analyses,ed. TerryN. Clark (San Francisco:Chandler, 1968), pp. 393-405.'TerryN. Clark, Community, tructure, ecision Making,Budget Expenditures, nd UrbanRenewal in 57 American Communities, American Sociological Review 33 (1968):576-93.



4/29

180 WesternoliticalQuarterlyfragile. f newness s an importantriterionfinnovation,heconclusionsfmanypast innovation tudiesmustbe questioned.Aiken ndAlford'studies furbanrenewal ndpublichousing, or xample,maynotbe innovationtudies tall becausetheywereconducted fter he ssueshad losttheir ragility.The operational efinitionf innovations criticallymportant. ecker ndWhisler's efinitionf nnovation,thefirst r early seofan idea byone of a setoforganizations ith imilar oals, 12s theone usedhere. Thus,an ideamust eadopted first r early to be considered nnovative.A variety f social sciencestudies n thepasthave indicated hat, vertime,nnovationsre adopted n ac-cordancewith a curveresemblingn S. 13 Roughly,nnovationsre adoptedslowly ver theearlyperiodof innovation.This period s followed ya secondperiod frapid adoptionwhile he nnovations hot orpopular, nd is then ol-lowedby a slowingn the rate of adoption nd waning nterest thusthe Sshapedcurve. Thisprocess ften overs period fyears 20 to30years rmore.The definitionf nnovationdoptedfor his tudyhus ssures hat ll innovationsfallon theearly rlowerportion fthe S curve.Adoption hus ncurs omerisktothe nnovatorss experiencesnotwidespread.Much ofthepreviousnnovation esearch as investigatedhecharacteristicsof individual innovatorsideology, motivation, ompetence, rofessionalism,opinion eadership,nd thelike.'4 Such analysishas wideapplication, ssumingthat the innovation ecisions re made byindividualscting lone. However, finnovationdoptiondecisions re groupdecisions, eneralizationrom tudies findividualsmaybe misleading.Social psychologicalesearchdemonstrateshatindividual nd groupdecisions ftendiffer,speciallyfrisk s involved.15ndi-vidualsmaymake eitherriskier r more cautiousdecisionsn groups hantheywouldalone. Whether roupdecisions re moreriskyr more cautious s notofmajorconcern ere. The point s,decisionsmade ngroups re oftenmore xtremethan thosemade by individuals ctingalone,hence concentrationn individualcharacteristicsaynotbevery nlightening.

CONCEPTUAL RAMEWORKExplanations f innovationdoptionhave beensought n threeevels: indi-vidual, organizational,nd extra-organizational.16onclusions re seldom nte-grated cross evels,however,makingtdifficultoweigh he relativemportanceof each level on innovationdoption. Since the individual evelof analysis asbeenwidespread,nd sincethegeneralizabilityfthis ype f researchs question-able, thepresentnvestigationxamines heimpactof organizationalnd extra-organizational actorsupon innovation doption. Innovationdecisionsmaybeaffectedya widerangeofvariables.Expanding pontheRoweand Boisetypol-

Michael Aiken nd RobertR. Alford,Communitytructurend Innovation:The CaseofUrbanRenewal, Americanociological eview 5 (August 970): 650-65;MichaelAiken ndRobertR. Alford, Communitytructurend Innovation: he Case ofPub-licHousing, Americanolitical cienceReview 4 (September970): 843-64.12SelwynW.BeckerndThomas . Whisler,The Innovative rganization: Selective iewofCurrentheoryndResearch, ournalfBusiness0 (1969): 463.For example,VirginiaGray, Innovationn the States: A Diffusiontudy, AmericanPolitical cienceReview 7 (December 973): 1173-85.14SeeRogers ndShoemaker,ommunicationf nnovations.15 A fine eview f this iteratures found n thespecial ssue rticles n theJournal fPer-sonalitynd SocialPsychology0 (December 971). See,for xample, ean G. Pruitt,Choice Shifts n GroupDiscussion:An Introductoryeview, pp. 339-60; DorwinCartwright,RiskTakingby ndividualsndGroups:An AssessmentfResearch m-ployingChoice Dilemmas, pp. 361-78; Dean G. Pruitt, Conclusions:Toward anUnderstandingfChoiceShiftsnGroupDiscussion, p.495-510.1 LloydA. Rowe and WilliamB. Boise, Organizationalnnovation:Current esearch ndEvolving oncepts, ublicAdministrationeview34 (May/June974): 284-93.



5/29

Innovation, Bureaucracy,and Public Policy 181ogy, xplanatoryariables eems ofall ntofour road andoverlappingategories:(1) thecommunitynvironment;2) policydemand; (3) theorganizationaln-vironment;nd (4) the haracteristicsftheorganization.Within hecommunitynvironmentwo classes fvariableswerebelieved oinfluencedoption:demographicnd cultural alues. The importancefdemo-graphicvariables n communityolicydecisions as been stressed yeconomistsand political cientistsormanyyears.17Manystudies upporthecontentionhatlocaldemographicnd socioeconomicariables re determinantsfcity olicy ut-put.18There is also an attitudinal omponent f communities hichcolors hethings itiesand/orpublic organizationso. Votingbehavior s oftenused toidentifyommunityttitudes19s are thepolitical ulture ypologieseveloped yElazar.20Policydemand ndicates communityemandfor specificervice.Demandmay,forexample, e basedupontheperceptionfa deficiencynservice elivery- often alled a perfomance ap- or it maybe based upon a need to reducecosts, mprovemanagement rocedures,tc. This study, orexample,uses ndi-cators fsuccess f the chool ystemdropout ate ndpercent fgraduates oingon tocollege) as demandvariablesn testinghemodelofadoption feducationalinnovations.The organizationalnvironmentefersoa setofexternal elationshipseyondthecommunityn which n organizationsembedded.Forthepublic rganization,it consistsf tsrelationship ithother overnmentalnits, heprivate ector,tc.Various works n economics nd political cience uggest hattheorganizationalenvironmentf ocalgovernmentalnits onsists ffour lements: 1) intergovern-mental relations;21 (2) professionalrelationships;22 3) private sector influence;and (4) slack, rexcess esourcesrovided y xternalources.23The fourth ategoryf variables ncludes he characteristicsftheorganiza-tion. Much of thediffusion esearch uggestshatorganizationalharacteristicssuch as (1) size,24 2) structure;253) professionalismwithin the organization;26and (4) formalizationnd complexity27ften ffectnnovationdoption.Thus,*7 orexample, olomon abricant, he TrendofGovernmentalctivityntheUnited tatesSince 1900 (New York: National Bureauof EconomicResearch,1952); G. RossStephensnd HenryJ. Schmandt,RevenuePatterns fLocal Governments, ational

Tax Journal 5 (1962): 432-37; RichardD. Bingham,ublicHousing nd UrbanRe-newal: AnAnalysisfFederal-Local elationsNewYork: Praeger, 975).18 A fine eview fthis iteratures foundn BrettW. Hawkins, olitics nd UrbanPolicies(Indianapolis:Bobbs-Merrill,971).9See,for xample, arl A. Lamb,As OrangeGoes (NewYork:Norton, 964); Aiken ndAlford,UrbanRenewal, . 653; Robert . WrinklendJerry. Polinard,Populismand Dissent:The WallaceVote nTexas, Social ScienceQuarterly4 (1973): 306-20.20DanielJ. Elazar,American ederalism:A View From The States,2nded. (New York:Crowell, 972), pp. 93-139; Bingham,ublicHousing nd UrbanRenewal, p. 47-50.2 Alan K. Campbelland Seymour acks, Metropolitan merica New York: Free Press,1967), pp. 53-66; IrwinFeller,Donald C. Menzel,and Alfred ngel,DiffusionfTechnologyn StateMission-Orientedgencies, eport o theNational cienceFoun-dation,DA-39596 (State College: Center or heStudy fSciencePolicy, he Penn-

sylvaniatateUniversity,974),p. 159.22Maw Lin Lee, A Conspicuous roductionheory fHospitalBehavior, outhern co-nomicJournal 8 (1971): 48-58; Rogers nd Shoemaker,ommunicationf nnova-tions, p.354-56;Feller,Menzel, ndEngel, iffusion,p.210-11.23For a review ee J.David Roessner,Innovationn PublicOrganizations, paper pre-sented t theNationalConferencen PublicAdministration,yracuse, .Y., May 5-8,1974, p. 3.24 Mohr, Determinants, p. 126.26EdwardC. BanfieldndJames . Wilson, ity oliticsNewYork:RandomHouse,1963);see also Robert . Lineberrynd EdmunP. Fowler, Reformismnd PublicPoliciesnAmerican ities, American olitical cienceReview 1 (1967): 701-16.26Feller,Menzel, ndEngel, iffusion,p.210-11.7Zaltman, uncan, ndHolbek,nnovations,p. 138-43.



6/29

182 WesternoliticalQuarterlythe tudyncludes othprimarynd secondaryttributesf theorganizations de-fined yDownsand Mohr.2A modelfortheadoption f nnovationsy ocal governmentalnits Figure1) was developed ncorporatinghefourcategories fvariablesdiscussed bove.According o themodel,thecommunitynvironment ay directlyffect nnova-tionor itmay ndirectlyffectdoptionbyaffectingolicydemand, heorganiza-tionalenvironment,nd/or hecharacteristicsftheorganization. olicydemandmay ffectnnovationdoption irectlyr ndirectlyy ffectingheorganizationalenvironmentnd/or heorganizationalharacteristics.he organizationalnviron-ment anaffectnnovationdoption irectlyr ndirectlyy ffectingrganizationalstructure. inally,organizationaltructurean onlyaffectnnovation doptiondecisions irectly. he hypothesizedodel sthus learly ecursive.

Figure.AModel orheAdoptionf nnovationsnLocalGovernment.

i l. iCommunity Policy Organizational Organizational InnovationEnvironment Demand Environment Characteristics AIn explainingheadoption fthecomputeroprovidemanagementnforma-tion ohousing uthorityecision-makers,ariables epresentingarge entral itieswith ow SES inhabitantscommunitynvironment) ereexpected o determinethenumber fpublichousing nits n thecity.A largenumber fhousing nitsindicates managementroblem f increasing omplexitydemandformore n-formation andling apacity) requiring ntergovernmentalssistance organiza-tional nvironment)nd a large ndprofessionalousinguthoritytafforganiza-tionalcharacteristics). ll ofthesevariables re thushypothesizedo affect om-puter se (the evel f nnovation) irectlynd/orndirectly.Temporalordering asalsobuilt ntothemodelto nsuretsrecursive ature.The variables elected o representhedemographicharacteristicsf thecom-munitynvironmenteretakenfrom he 1960censuswhile nnovationdoptionwas measuredn 1974. The other lements n themodeldate from ariousyearsbetween 960and 1974. Housing uthoritytaffize,for xample,was as of1968.

METHODS AND MEASURESThe types fgovernmentalnits obe studiedwere hosen or hediversityftheir rganizationaltructure. orexample, ousinguthoritiesregenerallyemi-autonomous ublicbodiesappointedby themayorwithcouncilapprovalwhilepublic ibrariesxhibit morediversifiedrganizationaltructure. f therespond-ing ibraries,8 percentregoverned ythecity rthe ocalschooldistrict,7per-centare countyibraries,9percent re city-county,nd 6 percent reregional.The selection f thespecificnnovations as basedon a telephoneurvey frelevant rofessionalssociation taffmembersnd headsofappropriate overn-mental nits hroughouthecountry. heseofficials ere sked: In your pinion,whatdo your onsider obe themajor nnovationsn (publichousing, ublicedu-

28Downs ndMohr, Conceptualssues.



7/29

Innovation, Bureaucracy,and Public Policy 183cation, etc.) over the past ten years? The processand product innovations istedin Table 1 were those mostfrequentlymentioned. Space does notpermit detaileddescriptionof each innovation. Such a description,however, is available fromanother source.29Once the innovations were selected,factorsdescriptiveof the community n-vironmentwerecomputed (factoranalysiswas used forpurposesofdata reduction).A factor analysis30of 18 census variables31was used to identifythe underlyingdimensionsdescriptive f thecommunity nvironment.Four factorswere identifiedand were labeled SES (socioeconomic status), Suburb, Ethnic/Ghetto,and Size.Factor scores for each of the 310 cities were then computed for each of thesefactors.32The aforementionedfactors,however,only representthe physical dimensionof the city's environment. Missing still are dimensionsthat suggest the culturalvalues of the population. Two measures of votingbehavior were selected as indi-cators of local political attitudes. The vote forWallace forPresident n 1968 wasused as an indicatorof anti-integrationistr racistattitude while the vote for Gold-water in 1964 was used as an indicator of a more general political conservatism.33To provide a thirdattitudinal ndicator,Elazar's classification f political cultureswas adopted. Each of the 310 cities was assigned a classificationbased upon thelocation of the city n relationto Elazar's map of dominantpolitical cultures34theMoralistic dominant cultures formedthe lower end of the scale withTraditionalis-ticat thehighend).These threeattitudinalvariables were thensubjected to principal componentsfactor nalysis. All threevariables loaded significantlynd negatively n one factor.The factoroperationalizesa conservative/liberal olitical culture dimension withinthe community nvironment. Since all of the loadingswere in a negativedirection,a positivevalue on the factor score fora given city s indicative of a liberal com-munitypolitical attitude. For this reason the factorwas labeled Liberalism. Factorscores foreach of the cities were also computed for this factor. Thus, in the statis-tical analysis that follows the communityenvironment s representedby the fourdemographicfactor coresand one attitudinalfactor core.The variables chosen to operationalizetheremainingcomponentsof the model- policy demand, organizational environment, nd organizational characteristics-are specific o theparticulartypeofgovernmentalunitstudied.Data concerning innovation adoption in housing authorities,public schools,and public librarieswereobtainedbymailed questionnaire. Response rates exceeded90 percentfor all questionnaire surveys. Data on the police use of ADP were ob-tained fromthe InternationalCity Management Association (ICMA) while vehiclemanagementdata were obtained from the vendor (Mainstem, Inc.).The operationalizationof each of thedependentvariables is discussed n detailin subsequentpages. Operationalization generallyextendedbeyond a simple use/29Richard D. Bingham and Thomas P. McNaught, The Adoption of Innovation by LocalGovernment, a report to the National Science Foundation (Milwaukee: Office of

Urban Research,Marquette University, 975).30R. J. Rummel, Applied Factor Analysis (Evanston: NorthwesternUniversityPress, 1970).31 U.S. Department of Commerce,Bureau of the Census, County and CityData Book 1962(Washington: GovernmentPrintingOffice, 962), pp. 476-575.32 Norman H. Nie, Dale H. Bent,and C. Hadlai Hull, SPSS: StatisticalPackage forthe SocialSciences (New York: McGraw-Hill, 1970), pp. 226-27.33 Voting statisticson a city by city basis were not available; therefore, he 1964 and 1968votingpercentagesare figures or the county n which the city s located. Data obtainedfromRichard M. Scammon, ed., America Votes 7 (Washington: GovernmentalAffairsInstitute,1968), and Richard M. Scammon ed., America Votes 8 (Washington: Gov-ernmentalAffairsnstitute,1970).3Daniel J. Elazar, American Federalism: A View fromthe States (2nd ed.; New York:Crowell, 1972), pp. 93-126.



8/29

184 WesternPolitical Quarterlynon-use ichotomynan attemptoassessevels f nnovationdoption.The schoolprocess nnovation,orexample, s thepercent felementaryeachers sing ndi-vidualized nstructionather hansimplyhe totalnumber fteachersnvolved.Initially,he ndependentariablesn each innovationmodelwerecorrelatedwith thedependent ariable. These zeroorder orrelationsre found n theap-pendixbut are not discussed ue to space limitations. artialcorrelation as thebasic toolof analysis,workingmodelright omodel eft. Initially, artialswerecomputed etween ach ofthe variables epresentingheorganizationalharacter-istics nd innovationwhilecontrollingorall of thevariablesrepresentinghecommunitynvironment,olicydemand, nd theorganizationalnvironment. llvariables epresentinghecharacteristicsf theorganization otsignificantt the.01 level whenpartialedwith the dependent ariablewerethendroppedfromanalysis.Next,partialswerecomputed etween hevariables f theorganizationalenvironmentnd thedependent ariable ontrollingor thecommunitynviron-ment, olicydemand, nd theremainingariables epresentingheorganizationalcharacteristics.he organizationalnvironmentariableswere lso partialedwiththe ignificantrganizationalharacteristicsontrollingor he ommunitynviron-ment nd policydemand.Variables n theorganizationalnvironmentotsignifi-cantlyrelatedto innovation r to at leastone of the remaining rganizationalcharacteristics ere also dropped. This procedurewas followed,working romright o left n testinghemodel,until ll of thepossible inkagesweretested ndunimportantariables ropped.The procedure ortestinghemodel s clearly onservative:t relies nlyonhighest rderpartialcoefficientstatisticallyignificantt the .01 level. The re-cursivenatureof themodel,based both on theconceptual rameworknd tem-poralordering,llowsfor nferencesboutpossible ausalrelationshipsnthemodel.The ratherargenumber frelevantndependentariablesmadetrue ausal model-ingvirtuallympossible. ue to spacelimitationsartialsnotsignificantt the 01level re notreportedrdiscussedlthoughheyre availablefrom nother ource.85

FINDINGSHousing nnovation

The housing nnovationstudiedwerethe adoptionof computer ystemssan aid tomanagementy publichousing uthoritiesnd theuseofprefabricatedcomponentsn theconstructionfhigh-rise ublic housing rojects. Operationaldefinitionsortheindependent ariablesrepresentinghe modelcomponentsrediscussed elow ndare also shown nthe ables n the ppendix.Policydemand s representedytwo variables:number funits nmanage-mentand number f high-riserojects onstructedrom1960 to 1970. A largenumber f units n management as expected o present demandfor omputerassistance n housing uthoritymanagement hilea heavyhigh-riseonstructionprogramwas expected o lead to theuse ofprefab omponentso cutbothcostsand constructionime.In theorganizationalnvironment,articipationn thefederalManagementImprovementrogram MIP)86 representedederalntergovernmentalssistancewhile ity ercapitarevenue nd a cityADP score37epresentedocalintergovern-3BinghamndMcNaught,TheAdoptionf nnovation.'A HUD programPublicHousingManagementmprovementrogram) roviding25millionoparticipatingousinguthoritieso evaluatend mplementew pproachestopublic ousing anagement.37The ADPscoresthe um f substantialumberf pecificDPapplicationseportedymunicipalovernments.he data were btainedrom he GMASurvey DP/70,SurveyfAutomatedataProcessingADP) inLocalGovernment.



9/29

Innovation,Bureaucracy,and Public Policy 185mental ssistance,r assistanceothehousing uthorityrovided ythemunicipalgovernment. ityper capitarevenue ttemptedoassess heavailabilityfexcess,or slack,fundswhichmight e usedto upgradehousing uthorityomputer ap-ability.Such slack would most ikely e madeavailable nequipment rtechnicalassistance. he cityADP score lsorepresentsform fslack. It wasexpected hata citywith highly eveloped omputerapabilitywouldbe likely oextend om-puter ssistance othe ocalhousinguthority.Use ofthehigh-riseonstructionethodsassumed otellussomethingboutthe nfluenceftheprivate ector n innovationdoption.Thisvariable uggestscontractornfluence,r dependence n contractornitiative,n the use ofinnova-tion. A construction ethod calefor ach authority asdeveloped yassigningscore f 1 toauthoritiessing nly onventionalontracting,toauthoritiestiliz-ingbothconventionalnd turnkey,nd 3 toauthoritieselyingolely ponturnkeycontracting.38Another spectoftheorganizationalnvironment as theorganization'sro-fessional elationship ithotherorganizations.t was expected hata reformedlocal government39ould ead to innovativer progressiveehavior romts bu-reaucraticgencies.Whilemosthousing uthoritiesre semiautonomousgencies,professionalspillover rom professionality dministrationnda goodgovern-ment rientation ere xpected.Within heorganization,rganizationalharacteristicsererepresentedythenumber f full-timemployees size), a dichotomousariable ndicatingwhetherthehousinguthority asindependentf thecity overnmentrwasa departmentof the cityadministrativetructureorganizationaltructure), professionalismvariablebased upon activityn theNationalAssociation fHousingand Rede-velopmentOfficials40professionalism)nd the age of the housing authority(formalizationnd complexity).HousingComputerUseThe questionnaireentto thehousing uthoritiesisted15 computerpplica-tionsutilized n publichousing hroughouthecountrynd askedtheexecutivedirectorso specifyhoseusescurrentlynoperationn their articularuthorities.Of the250 authoritiesesponding,omputer se scores sumoftheaffirmativee-sponses)ranging rom to 12 were obtained. The mean scorewas 1.7,withastandard eviation f2.56. Of therespondinguthorities,44had a computer sescoreof0.To test hehypothesizednnovationmodel fFigure1,partialswere omputedtestingll ofthepossibleinkagesn themodel s explained arlier.Partials oundstatisticallyignificantt the .01 level in testing hehousingprocess nnovationmodel re shown nTable 2.Of the fourselectedorganizationalharacteristics,nlyorganizationalizeshoweda significantndependentnegative) relationship ith nnovationwhilevirtuallyll ofthe variablesn theorganizationalnvironmentere ndependentlyrelatedto housing uthorityomputer se; onlyreformismr=-.14) was not.Demandwas also independentlyelated ocomputer se. The sizeofthemanage-mentproblemwas thereforendependentlyelated o the evel ofcomputer se.:' Turnkey ousing alledfor hedesign nd constructionfrental ousing y private on-tractors ho sold the finishedroduct o the ocalhousinguthority.he program ascalled turnkey ecause the developer turned he key over to the authoritynpurchase.39 eformmeasuresnclude ouncil-managerorm fgovernment,t-large epresentationndnon-partisanlections.The reformismcorevariesfrom to 3 and is thesum of thevalues ssignedothe hree eform easures.4 A dichotomousariable.A citywas given scoreof 1 if t had a NAHRO (professionalassociation)national fficer, ember f the boardofgovernors, national ommitteemember, regionalfficer,rchapterfficernthe uthoritytaff.



10/29

186 WesternPolitical QuarterlyTABLE 2. PARTIALS FOUND STATISTICALLY SIGNIFICANT IN TESTING MODELOF HOUSING PROCESS INNOVATION

rpOrganizationalCharacteristicswith nnovation (n= 134)Organizational size withcomputeruse .................. ................................... -.50OrganizationalEnvironmentwith nnovation (n= 136)Federal intergovernmentalssistancewithcomputeruse ........................................ .37Local intergovernmentalssistance (fiscal) withcomputeruse ........................... .20Local intergovernmentalssistance (equipment) withcomputeruse .................... .45Demand with nnovation (n = 140)Units in managementwithcomputeruse .................................................... ..38Demand withOrganizationalCharacteristics n = 140)Units in m nagementwithorganizationalsize ................................ ........................... .99Community nvironmentwithOrganizationalCharacteristics n= 142)Size w thorganizationalsize ..................................................................................... -.63Community nvironmentwithOrganizationalEnvironment n= 145)Liberalismwith local intergovernmentalssistance (fiscal) ............................... .25Ethnic/Ghettowith local intergovernmentalssistance (fiscal) .............................. .52Size with local intergovernmentalssistance (fiscal) ............................................. -.27Size withlocal intergovernmentalssistance (equipment) ........................................ .23SES with ocal intergovernmentalssistance (equipment) ..................2......6......... .26Suburb with local intergovernmentalssistance (equipment) .............--............... -.28Community nvironmentwithDemand (n = 24 )

Size with units n management ................................................................. - .94Ethnic/Ghettowithunitsin management ................................................................ .56SES with units n management ... ............................................................. -.24A partial was also computed between the demand variable and organizationalsize. As expected, a significant oefficientwas obtained. Obviously, large housingauthoritieshave themostemployees. The effect f the community nvironment norganizational size was also computed but only one of the coefficientswas signifi-cant. Size was negatively elatedto a largehousingauthority taff. The substantialnegativecoefficient etween Size and numberofemployees,however, s substantively

meaningless. The relationshipbetweencommunity nvironment nd the organiza-tional environmentwas also examined. A number of communityenvironmentvariables were related to local intergovernmentalssistance. Liberalism and Eth-nic/Ghettowere both positivelyrelated to per capita revenue while Size was nega-tivelyrelated. City ADP score was also sensitive to the communityenvironment.Size, SES, and Suburb were all independently elatedto the cityADP score.A numberof the characteristics f thecommunity nvironment ppear to stim-ulate developmentof a large number ofpublic housingunits. The mostsubstantialresult s also the most obvious. The sizable coefficient etween Size and policyde-mand merely ndicates that large cities have a lot ofpublic housing. It also appearsthat highlyethnic cities and those with lower socioeconomic characteristicswerealso more likelyto have more public housingthan theircounterpartswithout thesecharacteristics.It is now possible to describe the systemof process innovation adoption inpublic housing. Within the organization,the drive forefficiencyppears to be amajor factor n theadoption of the innovation. This fact s suggestedbythenatureof the negativepartial relationshipbetween numberofemployees nd thecomputeruse score. If one were to compare twohousingauthorities, ach with thesame num-ber of unitsin management,the authoritywithfeweremployees n 1968 would bemore likelyto have a highlydeveloped computerbased management informationsystem n 1974.



11/29

Innovation, Bureaucracy,and Public Policy 187Whiletheorganizationalnvironmentoes notdirectlyffect hestructuralnature ftheorganization,t least s it ismeasured ere, t doesaffectnnovation

adoption.The existencefa highresourceevelwithin hecitygovernments animportanteterminantf innovation.Althoughmost uthoritiesre semiautono-mousagencies, t appearsthatstrongnformal ies inkmunicipal esources ithauthoritynnovation.This suggestshatthe nfluence fthecity nproviding e-sources or uthorityseis a majordeterminantfauthoritynnovation. utwhatgeneratesheseresources?The resultsuggest hat lack n theorganizationaln-vironmentstheproduct f thecharacteristics both ttitudinalnddemographic-of thecommunityn which theorganizations located. Size apearsto be themajordeterminantfresourcevailabilityutmany f thecity haracteristicslsocontributeoa highresourceevel. The relationshipetween olitical ttitude ndslack s particularlyoteworthy. pparently,ttitudesoncerninghe distributionof resources orthepublicweal contribute o slackavailability.Whilecityre-formismhowedno substantial irect ink with nnovationdoptionor withthepertinentrganizationalharacteristicf theauthorities,twas suspected hatre-formism ight avean indirectink nthe nnovationrocess.Thisappeared o bethecase. A partialwas computed etween eformismnd cityADP score ontrol-lingforthecommunitynvironmentnd demand. This computationroducedsignificantartialof r=.18. Professionalismn citygovernmentpparentlynde-pendentlyontributeso a higherevelofcomputerapabilitynthecitywhich,nturn,eads to a higherevelofcomputer sebythehousinguthority.Intergovernmentalnfluence lso substantiallyontributed o adoption. Al-thoughMIP participation as limited oonly welve f the uthoritiestudied, heinfluencefthe vailabilityfsuch program asstartling. he availabilityfthisprogramwas extremelymportanto thedevelopmentf sophisticatedomputerbasedmanagementnformationystems.Demand also independentlynfluencesrocess nnovationdoption.A largenumber fhousing nitspresenteduthorityfficials iththemanagementrob-lem or goal. As theauthorityrew, heproblems fmanagementncreased ndbecame moresophisticated;ontrol ecame moredifficult,nd informationndcommunicationshannelsmorecomplex.Computer ased information as thusadopted oprovide p-to-dateataformanagementecisions.In general, he modelsuggestshat nnovationdoption n publichousingsa predictable rocess.Decisions oinnovate ppeartobe baseduponvery ationalconsiderations efficiency,eed. nd resourcevailability.HousingPrefabComponentsThe hypothesized odel fortheadoptionof product nnovationsypublichousinguthoritiessmuch he ameas thehypothesizedrocessnnovationmodel.The product ependent ariable stheuse ofprefabricatedomponentsnhigh-risepublichousing onstruction.41f the 170housing uthoritiesonstructingt leastonehigh-riseetween 964 and 1974, lmosthalf 48.2 percent)used one or moreof theprefabricatednnovationselected.The prefab cores angedfromO to4;however,nly ix of thehousing uthorities3.5 percent) ndicated se ofall fourof theprefabnnovations.The productnnovationmodelwas tested sing hesamemethodologys wasused in theprocessmodel. Onlyconstruction ethodwas significantlyelated oinnovation.Contractorctivityppearsto be theonlyvariablerelated o prefabuse. The communitynvironment,emand, nd theorganizationalharacteristicsdo notcontribute,irectlyr indirectly,o the useofprefabricatedomponentsn41Prefab nnovations erekitchen/bathroomtilityores,precast oncretexterior,recastinterioranel walls, nd precast anelfloors.Prefab coreswere implyhesumofthedifferentrefabnnovationssed.



12/29

188 WesternPolitical Quarterlyhigh-rise onstructionn public housing. It does appear, however,that ntergovern-mental influencewas, in the long run,responsiblefor nnovative behavior. It wastheturnkey rogramwhichopened thedoor fordeveloper nfluence nd innovation.Turnkeycontracting, hen,appears to hold thekeyto productinnovation in publichousing construction.In examining the adoption of innovationby public housing authorities, hereis nothing to suggest that public housing authorities are particularly nnovative.Hayes and Rasmussenemphasizethe factthat ocal governments evote little nergyto innovation and are essentiallyhostile to change.42 A studyof two innovationsin public housingdoes nothingto dispel this dea. In fact,theoverwhelmingdomi-nance of factors n the organizational environment s determinants f innovationsupport the contentionconcerningthe conservativenature of the public bureauc-racy. While demand and desire forefficiencyre relevant, t is the organizationalenvironment hat has themajor impact. Intergovernmentalnfluences nd resourceavailability seem to be the prime determinantsof innovation in public housing.PublicSchool nnovation

Two variables, the percent of elementaryteachers using programs of indi-vidualized instruction43 nd the number of videotape recordersper ten secondaryteachers,were theprocessand product nnovations tudied.Policy demand is representedby two variables: the percentof studentsgoingon to college and thedropoutrate. Measures ofacademic achievement (testscores)are not generally vailable; therefore he above twomeasures of theperceivedsuc-cess of the school systemwere obtained fromquestionnaireresponses.Within the organizational environment, ederal intergovernmental ssistancewas representedby the total intergovernmental evenuethe school district eceivedfrom the federal government, nd state intergovernmental ssistance was repre-sentedby total school district evenues receivedfrom hestate. Locally raisedreve-nue was the third major categoryof the organizational environment to be ex-amined. Specificvariables in thiscategory nclude cityper capita revenue,per stu-dent school expenditures, otal school district xpenditures, nd total school capitalexpenditures.44The capital expenditure variable obviously pertains only to theproduct innovationmodel. Again, professionalrelationships n the organizationalenvironment re represented ythecityreformism core.The organizationalcharacteristic, ize of theorganization, s represented ythenumberof elementary eachers n the school system in the individualized instruc-tionmodel) or by the number of secondaryteachers (in thevideo recordermodel).Organizational structurewas representedby a school board reformism core de-veloped in the same manner as the cityreformismcore.45 In order to examine thestructural mphasis placed on audiovisual material,a second variable, audiovisualstaffper 100 teachers,was also included in the productmodel. Professionalismnthe organizationwas representedby the percentof teachersin the systemwithouta BA degree (low professionalismndicator) and the percent of teachers in thesystemwithan MA or higher. In an attemptto measureorganizationalformaliza-42Frederick 'R. Hayesand JohnE. Rasmussen,ds.,Centers or nnovationn theStatesand Cities San Francisco: an Francisco ress, 972).4 Individualizednstructions defineds theadoption f one ofthe threemajorformalizedprogramsf individualizednstruction:ndividually uided Education IGE), Indi-vidually rescribednstructionIPI), and Program orLearningn AccordancewithNeeds (PLAN).44Expenditureata is usedas a surrogateorrevenue s in public gencies evenuend ex-pendituresre (nearly) qual.Reformisms representedyan elected choolboard, t largerepresentation,ndnonparti-sanelections.The reformismcore aries rom to 3 andis the umofthethree eformmeasures. Professionalisms represented y thetwo measures f teacher ducationalqualifications.



13/29

Innovation, Bureaucracy,and Public Policy 189tion and complexity, scale was developed from the questionnaireto indicate cen-tralization/decentralizationfrequestsfor audiovisual equipment.The scale rangedfrom1, if the individual school initiatedtherequest,to 3, if the centraladministra-tion determinedneed. Similarly,a decentralized/centralized cale for equipmentpurchasingwas developed witha score of 1 assigned to school districtsnwhich thecentral administrationwas responsibleforequipment purchasingto a score of 3 ifeach school purchased itsown VTR equipment.Individualized Instruction

As with the investigationof innovation adoption in public housing, it wasnecessary o use partialsto test the model. In correlating heorganizationalcharac-teristicswith innovation,onlythe relationshipbetwenorganizationsize was relatedto a commitment o individualizedinstruction s is shown in Table 3. It should bepointed out that the proponents of individualized instructionargue that extrateachers are not needed for individualized instruction.Nevertheless,the findingsheresuggestthatwhile added personnelresources re notnecessary or mplementa-tion of the innovation,those school districtswithhigherteacher resources showeda higherlevel of commitment o individualized methods than thosewithout.TABLE 3. PARTIALS OUNDSTATISTICALLYIGNIFICANT IN TESTINGMODEL OF SCHOOL PROCESS INNOVATION

rpOrganizationalCharacteristicswith nnovation (n = 172)Organizational size with individual instruction .............................. ............. .19OrganizationalEnvironmentwithOrganizationalCharacteristics n = 180)Federal intergovernmentalssistancewithorganizationalsize ............... .............. .35State intergovernmentalssistance withorganizationalsize ...................................... .72Total school district evenue withorganizationalsize ............................................. 70Cityper capita revenue withorganizationalsize .............--............i........ --. -.23Demand withOrganizationalCharacteristics n= 178)Percentgoingto collegewithorganizationalsize .......-......-.............. ......39Dropout rate with organizationalsize .................................. .......---- -.19Demand withOrganizationalEnvironment n= 182)Percentgoing to college with state intergovernmentalssistance ............................ -.22Percentgoing to college withtotal school district evenue ............................-.......... -.21Dropout rate with total school district evenue ..................................-............. .18Community nvironmentwithOrganizationalCharacteristics n= 183)SES withorganizationalsize .......................................... ...................... .... . .... . ... -.21Size with organizationalsize ..................................... ...........-- 77Community nvironmentwithOrganizationalEnvironment n = 186)Size with federal intergovernmentalssistance ..-............................. ...................... .48Size with state intergovernmentalssistance ...................................... ........... -- .89Size with total school districtrevenue .....-..............-................................--......---- .90Size with cityper capita revenue ..................................-. ................. ..23Liberalismwithcityper capita revenue .......................................................... ......... .26Ethnic/Ghettowith total school district evenue ......................-...... ......... .19Ethnic/Ghettowith cityper capita revenue ..........-.............................-... ... .57Community nvironmentwith Demand (n = 238)SES withpercentgoing to college ..........-.......................... ...- .44SES withdropoutrate .............................................................................................. .... -.20Suburb withpercent going to college ........ ................................ .17Suburb with dropoutrate ........................................................ -.31Liberalismwithpercentgoing to college .................................................................. -.16Liberalism with dropout rate ........ ....................................................................14



14/29

190 WesternoliticalQuarterlyThe fact thatpersonnel esourcesre a majorcontributoro theadoption findividualizedducation uggestshat lackresourcesntheorganizationalnviron-ment renecessaryoprovide hepersonnelesources. huspartialswere omputedbetween heorganizationalnvironmentnd organizationalize. Four of the sixvariableswererelated opersonnel esources.ntergovernmentalnfluenceppearsto be extremelymportant.The independent elationshipsetweenfederal ndstatentergovernmentalssistancendorganizationalizewere ubstantial. imilar-ly,theexistencef slackresourcesn theschool nvironmenttotal expenditures)providedhenecessaryundingorhigh eacher evels.Surprisingly,heexistencefslack n thecity as opposedtoschool) environ-mentwasnegativelyelated oorganizationalize. Elementarychools n citieswithlow per capita expendituresorcityfunctionsexcluding ducation) appear tohave highpersonnelevels, therthings eingequal. This relationshipeemstoreflect trade-offn financial rioritiesn thecity: a significantorrelationfr-.28 was obtainedbetween itypercapitarevenue nd total chooldistrictx-pendituresontrollingor he ommunitynvironmentnddemand.The impactof thedemandvariables n innovationdoptionwas also con-sidered. Demandwas not ndependentlyelated o innovation.There was an in-dependent elationshipetweendemand and organizationalize,however. Themodeltestinghusfarsuggestsn amenity ature o individualizednstruction.School districts ithhighperformanceevels endtohavea largenumber fele-mentaryeachers or heschoolpopulation nd also tendto exhibit commitmenttoprogramstilizingndividualizednstruction.The percent oing n tocollegewasnegativelyelated o state id and to totalschool districtxpenditures hile thedropout atewas positivelyelated o totalexpenditures. hese relationships,owever, o notfitthemodel. For example,under ontrols,istricts ith highpercentageoing n tocollege eceiveess tateaid and have lower xpenditureshan their ounterpartsithfewer ollege-boundstudents.Size is clearly heoverwhelmingeterminantf the vailabilityf ntergovern-mental unds ndofthefinancial lack vailable n theorganizationalnvironment.While thedemandvariableswererelated o theorganizationalnvironmentn awaynotconcomitant ith hemodelwhen ontrollingor ize, t s Size that s the

primaryeterminantfintergovernmentalundingnd financiallack.Severalvariables n thecommunitynvironment eresignificantlyelated oorganizationalize. AgainSize was dominant uta lowcity ocioeconomictatusalsocontributedohigh eacher esources.Also computedwerethe relationshipsetween ach of the communityn-vironmentariables nd demand.Threeof thevariableswere ignificantlyelatedto both thepercent ollegeboundand thedropoutrate and in theexpecteddirections. itieswith high ocioeconomictatus ndwith uburbanharacteris-ticsgenerally ad a highcollege-boundopulation.Thesesame cities lso tendedto have lowdropout ates. Liberalism lso showed ignificantelationshipsn theexpecteddirection.In summary,heschoolprocess nnovationwas clearly redictable lthoughthemodeltesting uggestshat here s a substantial ifferenceetween amenityinnovationsnd innovations asedon need. Earlier twasnotedthat headop-tionof computer-based, anagementnformationystems as broughtbout, npart,by owpersonnelevels, r a needfor fficiency.n thismodel, heoppositeoccurred a highpersonnelevel n terms f number felementaryeacherswasindicativef commitmento ndividualizednstruction.Within heorganizationalnvironment,ntergovernmentalransfersppeartoprovide, r at leasthelp provide, hefinancial esources ecessaryorhighperson-nel levels. While therewas no direct elationshipetween heorganizationaln-



15/29

Innovation, Bureaucracy,and Public Policy 191vironmentnd thecommitmento ndividualizednstruction,henegative elation-ship between ityper capita revenue nd total schoolexpendituresndicatesprobable rade-offetweenity ndschool evenue.The relationshipetween hepolicydemandmeasuresnd organizationalizeagain fits model of amenitynnovation.Otherthings eingequal, theschooldistricts iththefewest erceived ducational roblemsre more ikely o have ahigher atioofelementarychool teachers nd are thusmore ikely o adopt ndi-vidualized nstruction ethods.Bothmeasures f demandare related o thesame three actorsn thecom-munity nvironment factorswhichlargely apturecommunity ealth. Thuswealthyedroom ommunitiesave fewerducational roblems.As expected, ize was the most mportanteterminantf intergovernmentaIassistance, chool and cityexpenditureevels,and the numberof elementaryteachers. The environmentalharacteristicsf thecommunity,owever, o notoperate ndependentlyodeterminehe evelof nnovationdoption.VideotapeRecordersAswith heothernnovationstudied, series fpartialswascomputed otestthemodel, nd theywere omputedn the ame equence s reportedarlier.Noneoftherelationshipsere ignificant.husa quantitative odelfor he doption fvideotaperecordersn secondarychools s a blankpage. None of theselectedvariables howed ignificantndependentelationshipsiththe nnovation.Librarynnovation

Public ibraries oseda specialproblemn theanalysisn that herewere339librariesn the 310 largestAmerican ities. Some cities, hen,had twoseparatelibraryystems,sually dual systemfcity nd countyibraries. his forced hedevelopmentf newunit fanalysis the libraryity. allas,for xample,withtwo libraries, equired wodependent ariables n both theprocess nd productinnovationmodels. Since twas hypothesizedhat ity haracteristicsre determi-nants f nnovation,twasexpectedhat he haracteristicsfDallas would,npart,determinehe evelofcomputer se nboththecity ndcountyibraries. hus thelibrary ity was developed o thatthecommunitynvironmentalharacteristicsof Dallas were associatedwithboththemajor ibrariesocated n thecommunity.Insteadof310cases orcities, his nalysishus overed 39casesor libraryities.The library rocessvariable, ike thepublic housing rocess ariable, elatesto computer se in librarymanagement.Again,the ibraryomputer se score smerely he sumoftheyesresponse o eight pecifictems oncerningibraryom-puteruse on thequestionnaireent opublic ibraries.nformationor heproductdependentvariable (dichotomous) use of a library heft etection ystem-was obtainedfrom he six major manufacturersf theft etection ystems orlibrariesndwas confirmedyresponseso the ibraryuestionnaire.Four variableswereselected o represent policydemand for nnovation:number fvolumes n the ibrary,umber flibraryranches,nnualcirculation,and thepopulation fthe ibraryervice rea. While ll four fthevariablesweresizerelated, achwasselected omeasure specificmanagement roblem.Measur-ingneed,however,maybe somewhatllusory. n thearea oflibraryutomation,Markuson t al. distinguishetween enuineneedsand artificial eedsfor uto-mation.Genuineneedsarise from uchthings s an increased olumeofactivity,or a need for mprovedervices ousers.Artificialeeds, n theotherhand,maybe created ythe vailabilityfa computer.ressurerommanagement,rprestigein librarymanagement.4646BarbaraE. Markuson t al., Guidelines orLibraryAutomation: Handbook orFederaland OtherLibrariesSanta Monica,Calif.: Systemsevelopmentorporation.972),p. 8.



16/29

192 WesternoliticalQuarterlyWithinthe organizational environment, egional intergovernmentalssistancewas measuredbya dichotomousvariable,OCLC membership.47 ocal intergovern-mental assistance was measured by total libraryrevenue and per capita libraryrevenue.48 Professionalism n the organizational environmentwas again indicatedby cityreformism.Two variables were selected to represent rganizationalcharacteristics fpub-lic libraries: the size of the library taff organizational size) and a scaled variablerepresenting ibraryorganization. If the librarywas a city ibrary t was scored 1,a city-countyibrary2, a county ibrary3, and a regional library4. The organiza-tionalinformationwas obtainedfrom he ibrary uestionnaire.

Library omputerUsePartialswere first omputedbetween theorganizationalcharacteristics nd thelibrarycomputer use score controllingforcommunity nvironment, emand, andthe organizational environmentand are shown in Table 4. Both organizationalcharacteristicsmaintained positive relationshipswith the computeruse score.TABLE 4. PARTIALS OUNDSTATISTICALLYIGNIFICANTN TESTING MODELOF LIBRARY ROCESSINNOVATION

rpOrganizational Characteristicswith Innovation (n = 203)Organizational size withcomputeruse ...........-................................................... .16

Organizational structurewith computeruse .1....................... .................. ... .16Demand withOrganizationalCharacteristics n = 208)Numberof libraryvolumeswithorganizationalsize ..-............................ ........ ......64Annual circulationwithorganizationalsize ........ .......78........ .78Population of circulation area with organizationalsize ............................. .74Number of branch librarieswithorganizationalsize ....... ............................ ............ .48Number ofbranchlibrarieswithorganizationalstructure........... .. ...... .20Community nvironmentwithOrganizationalCharacteristics n = 211)Suburb with organizationalstructure ...................................... ................... -.17Ethnic/Ghettowith organizationalsize ............. .. .............................................. -.17Community nvironmentwith Demand (n = 229)

Ethnic/Ghettowith numberof libraryvolumes ............................................. .... .43Ethnic/Ghettowith annual circulation .................................................................... .22Ethnic/Ghettowithpopulation of circulationarea .................. .................. .33SES with number of libraryvolumes ....................................................................... -.20Size with numberof libraryvolumes .......................... .... ......................... .89Size with annual circulation .......-......................... ................................ .80Size withpopulation of circulation area ................ ............. ........................................ 79Size with numberof branch libraries ......................................................................... .73

Demand is also an influentialdeterminantof automation- but an indirectdeterminant.All four variables showed a strong ndependentrelationshipwiththesize of the organizationalthoughdemand was not strongly elated to the organiza-tional structure.Only therelationshipbetween the number of library ranchesandstructurewas significant nder conditionsof control.47 OCLC (Ohio College Library Center) is a nonprofit hio corporationdesignedas a vehiclefor greater cooperation among libraries in Ohio and regional library systems utsideOhio. The Center operates a computernetworkconnectedwith hundreds of librariesacross the country. Objectives of the Center include making libraryresourcesthrough-out the network vailable to users of the individual libraries and loweringthe per unitcostsin libraries.4'Again,data is expendituredata and is used as a surrogate orrevenue.



17/29

Innovation, Bureaucracy,and Public Policy 193Four of thevariablesn thecommunitynvironmentffectedhe ibraryom-puter sescore, ut ndirectly.izewaspositivelyelated o all four fthedemand

variableswhileEthnic/Ghetto as independentlyelatedto all of the demandvariables xceptnumberf ibraryranches.Onlytworelationshipsfsignificanceere notedwhen the characteristicsfthe ommunitynvironmentere orrelated ith heorganizationalharacteristics.Ethnic/Ghettoas negativelyelated oorganizationalizeand Suburbwasnega-tivelyelated o the ibrary rganizationaltructure.There are a number fgeneralizationshat an be madebaseduponthe find-ings. The first oncerns he relationship etweenorganizationaltructure ndcomputer se. The more regional he ibrary,hemore ikelyt s that he ibraryhas a high omputer sescore.But twasnotpossible o dentifyhe haracteristicswhich ead to a sharing rcooperative pproach o ibraries.None of thevariablesin the modelwererelated n any causal wayto libraryrganization.Onlytherelationshipsetween uburb and organizationnd number f branches nd or-ganizationwere ignificant.he negative elationshipetween uburb ndorgani-zation tellsus onlythatthe central fficesfcounty nd regional ibrariesre notlocated n the uburbs.The relationshipetween umber f branchesndorganiza-tionmerelyndicates hata regional ibraryystems most ikely o have a largenumber fbranches.Yet thismay, n fact, ead to computer se. While numberof brancheswasnot ndependentlyelated ocomputer se, twas related hroughbothof theorganizationalharacteristicsnd at thezeroorder evel. The fact hatcounty rregionalibraries avea largenumber fbranches,rrespectivef servicearea population, total number of volumes in thesystem,nd theannual circulation,may be the key to the relationshipbetween organization and computer use. Itmay be that countyand regional librariestend to be more likelyto automate be-cause oftheproblems ssociated witha largenumberofbranches.Why is organizational size positivelyassociated with computer use? Whilethis positive relationship s suggestiveof an amenity nnovation,the positive rela-tionship between organizational structureand computer use is suggestiveof anefficiencynnovation. It appears that Markuson is correct. Library automationseems to be theproductofbothgenuineand artificialneeds.Anothersurprising indingwas the sterilityf the organizational environmentin determining nnovation. The organizational environmentmade verylittledif-ference. Apparentlyresourcesprovided by the organizational environment re notsubstantialenough to have much effect n computeruse in public library ystems.TheftDetection SystemsA surveyof all of the manufacturersof librarytheftdetection systems sixmanufacturers)revealed that only twenty-sevenublic libraries n American citieswith 1960 populations of 50,000 and above had purchased such systems.This isonly 3.2 percentof the librariesunder study. The dependentvariable, then,was adichotomousvariable - uses theftdetectionsystem.Testingof the libraryproductmodel produced results similar to the housing and school product models- thatis, there were no significant esults. Thus there s very ittle than can be said aboutlibrary doption oftheft etection ystems.Innovation inMunicipal GovernmentIn an attemptto provide as broad a basis as possible for the investigation fthe adoption of innovation by local government,two verydifferentnnovationswere selected in two different reas of local government. The process innovationwas the adoption of computersas an assist to police operations,while contractingwithMainstem, Inc. for vehicle fleetmanagementserviceswas the selectedproductinnovation.



18/29

194 WesternoliticalQuarterlyTo measure commitmento computer se,a scale was developed ummingthe 24 possibleADP/computer pplications urveyed y the ICMA.49 The city

product nnovationwas a simpledichotomous ariable cityusesMainstem. fthe citycontractedwithMainstemforfleetmanagementervice t was givenascore f1; if tdid not ontractor he ervicetwasscored .Policydemandforpolice computer se was measuredbyperceivedneed toimprove oliceperformance the 1969 FBI crime ndexand thechange n thisindex from1964 to 1969. Therehas beensomequestion boutthe use ofcrimestatisticss an adequatemeasure fthecrime ate. The crime tatisticsontainedintheFederalBureauof nvestigation'searly niformrimeReport refrequent-lyconsidered eak ndicators fcrime even ofthe imited rimes hat hey ur-porttomeasure.Whilethese tatistics aybe weak ndicators fthecrime ate na givencity, here s increasingvidence s to their sefulness. recent tudy ySkogan, or xample, uggestshat heFBI crime tatistics aybe usefulndicatorsof therelative istributionfcrime cross ities.50 emandforMainstem ervices,on theotherhand,wasexpected obe efficiencyelated. t wasexpected hathighpublicworks ostswouldgeneraten interestn Mainstem s a method oreducevehicle ndequipmentmaintenancexpenditures.In theorganizationalnvironment,ederalntergovernmentalssistance egis-tered hereceipt f LEAA money o improve omputer apabilities.State nter-governmentalnfluencenpolice omputerdoptionwasmeasured y wovariables.First, scale scorefor ach citywas developed ysumminghenumber fsignifi-cant trainingnd selectiontandards orpolicecandidatesmandatedbythestatein which the city s located. These staterequirementsnclude: (1) mandatorytrainingtandards;2) candidatenterview;3) high chooldiploma; 4) physicalexamination; nd (5) backgroundnvestigation. he secondvariable was thenumber f hoursofpolicetraining equired ystate aw. Data forbothvariableswereobtainedfrom heNationalAssociation f State Directors fLaw Enforce-mentTraining.Local intergovernmentalssistance asagainmeasured ythecityADP score.Local resourcesn thepolicecomputermodel ncluded ity ercapitarevenue,totalpoliceexpenditures,ercapita policeexpenditures,nd a dichotomousari-able indicating olice departmentsavingtheir wn computer scoreof 1) andthose haring computer ith ther ity epartmentsscoreof0). Citypercapitarevenue nd the ncreasen thepublicworks udget ince1965 werethe ocalre-sourcemeasuresntheMainstemmodel. Once againcity eformismas indicativeofprofessionalelationships.The organizationalharacteristicize was representedythe ize of thepolicedepartmentnd thesize of thepublicworks PW) department.Organizationalstructureppliedonly opublicworkswith wo tructuralariableselected:varietyof functionserformedythePW departmentnd a dichotomousariable ndicat-ingPWresponsibilityor ehicle leetmaintenance.Professionalismn theorganization asmeasured ytheminimumalary ndthepresence f a union npolicedepartments,hile npublicworks rofessional-ismwas operationalized ycivil service overage nd bythenumber fAmericanPublicWorksAssociationAPWA) members orkingor he ity.Formalizationnd complexityn theMainstemmodelwas representedytheage of thePW department,henumber f divisionsn thedepartment,nda mea-sure fdepartmententralization.49Thepolice omputersescores the um f24 specificDP/computerpplicationsb-tained romCMA Survey DP/71, ApplicationfAutomatedata ProcessingnPolice epartments1971.Wesley . Skogan,TheValidityfOfficialrimetatistics: nEmpiricalnvestigation,SocialScienceQuarterly5 (June1974): 25-38.



19/29

Innovation,Bureaucracy,and Public Policy 195Police ComputerUse

While computeruse by police departmentshas undeniablygrown since 1964,adoption of thecomputerto police workis byno means standard. Of the 213 citiesformingthe population of interest,126 or 59.2 percent did not use the computerat all. The computeruse score developed foreach cityranged from0 to 23 witha mean of 3.9 and a standarddeviation of5.7.To testtheprocessinnovationmodel,partialswere first omputedbetweentheorganizational characteristicsand the police computer use score controllingforcommunity environment,policy demand, and the organizational environment.None oftherelationshipswas significant.Only two of the partials between the variables of theorganizational environ-ment and the police computer use score were significant Table 5). Of the fourindicators of local resources,only the coefficient etween total police expendituresand computeruse was substantial.

TABLE 5. PARTIALS FOUND STATISTICALLY SIGNIFICANT IN TESTING MODEL OFPROCESS INNOVATION IN THE COMMON FUNCTIONS OF CITY GOVERNMENTr,p

OrganizationalEnvironmentwith nnovation (n = 131)Total police expenditureswithpolice computeruse ....................... ........................... .49CityADP scorewithpolice computeruse ...................................................... . .46Demand with nnovation (n = 132)1969 crime ndex withpolice computeruse ....................-.....-.......................... 34Change in crime index withpolice computeruse ..............-..................-........ .24Demand withOrganizationalEnvironment n = 172)1969 crime ndex with total police expenditures ..................-.............................. ...581969 crime index withcityADP score ...1................ ........... .............. .19Change in crime ndexwith total police expenditures ................................ -- .62Change in crime index with cityADP score ........................................... - .23Community nvironmentwith nnovation (n= 137)Ethnic/Ghettowithpolice computeruse . . ............................ - ....................... -.28Size withpolice computeruse _-------------....................-................................. .30Community nvironmentwithOrganizationalEnvironment n = 117)Size with total police expenditures ........ ...... ...................................... 38SES withcityADP score ..................................................................................... .27Suburb with cityAD P score .................................................. ................................... -.30Community nvironmentwith Demand (n = 249)Ethnic/Ghettowith 1969 crime index ............. ......... .............. ..........54Ethnic/Ghettowith change in crimeindex ............................................................... .32Size w th 1969 crim index ..................................................................................... . .96Size withchange in crime ndex .......................................................... ... .89

The onlyothersignificant ariable in the organizational environmentwas themeasure of equipment slack. A positive relationshipwas found between the cityADP score and police computeruse. Furthermore, hereprobablyis a connectionbetween thecityADP score and police expenditures.When a partialwas computedbetween the cityADP score and total police expenditurescontrollingforthe com-munityenvironment nd demand, a significant egative coefficient f r=-.47 wasproduced. Thus it appears that a well-developed citycomputer system s indeedindependently elated to lowerpolice expenditures.The next set of partials examined the relationshipbetween the demand vari-ables and police innovation. Both relationshipswere significant.The demand vari-able influencedcomputter se indirectly s well as directly.When the relationship



20/29

196 WesternoliticalQuarterlybetween he two crimemeasures nd the organizationalnvironmentas com-puted, ll relationshipsere ignificant.The relationshipetween hecommunitynvironmentariables nd the de-pendent ariablewas also partialed ut. Two variables, thnic/Ghettond Size,were ignificantlyelated o nnovation. or thefirstime, hecommunitynviron-ment xerts n independentnfluencen the evelof nnovationdoption.Severalof thevariables n thecommunitynvironment erealso related ovariables n theorganizationalnvironment.sexpected, ize was related opoliceexpenditures.wo variables, ES and Suburbwere ignificantlyelated othecityADP score.As hypothesized,omeof the characteristicsf thecommunitynvironmentwerealso related opolicydemand. Ethnic/Ghettond Sizewerebothpositivelyrelated o demand.Whilethestrong elationshipsetween ize and the crimen-dices wereexpected,t was surprisingo discover ery ittle elationshipetweenSES and crime. t is lowSES characteristicshat renormallyssociatedwithhighcrimerates. The same was trueof Suburb. Central ity, r negativeoadings fSuburb,werealso expected obe related o crime.Again,thiswas not thecase:bothSES and Suburb oaded in a negative irection,ut therelationships erenotsignificant.The model fortheadoptionof computer ystemsypolicedepartmentss amodelbasedupon perceived eedand a drivefor fficiency. ithin heorganiza-tionalenvironment,hecityADP scorewaspositivelyelated o thecomputer sescorewhiletotalpoliceexpenditures erenegativelyelated.The factthat lackresourcesn terms fa welldeveloped ity omputerystem ere vailable, oupledwith he owexpenditureeed for fficiencynpoliceoperations,ppears o stimu-latepolice nterestncomputerse.Demand was again an importanteterminantfcomputer se. Evenundercontrols,hereportedrime atedefinitelyppearsto be related o theuse ofthecomputer.Police departmentsn citiesreporting ighcrimerates nd/or highincreasencrimen the1960'swere ikelyo havehigh omputercores y1974.MainstemUse

Of theAmerican itieswith1960populationsf50,000ormore, ome20citiescontracted ithMainstem or heir ehicle leetmanagement rogram,.5percentof the ities tudied. nitially,artialswere omputed etween ach of the rganiza-tional characteristicsnd theMainstemvariable (Table 6). Onlyone variable,PW departmentoesvehiclemaintenance, as significantlyelated o Mainstemuse. The negative oefficientndicates hatcities n whichthedepartments notinvolvedwith vehiclemaintenancere morelikely o utilizeMainstem ervicesthan re cities nwhich ehiclemaintenancesa PWfunction.Coefficientserethen omputed etween heorganizationalnvironmentndMainstem se. Only ocal intergovernmentalssistancewas significantlyelated oMainstem.The partialbetween he twodemandvariables nd the locationof vehiclemaintenance as next omputed.Bothtotal ndper capitaPW expenditureseresignificantlyelated o the ocation f thevehiclemaintenance unction. hisfind-ingwasobviouslyxpected.Partialswere omputedhowingherelationshipetween ommunitynviron-ment nd thecityADP score.Therewasveryittle ifferenceetween herelation-shiphere nd thesamerelationshipn thehousing,ibrary,ndcity rocessmodels.SES againshowed rathertrongndependentelationshipith ity omputer se.Suburbwas negativelyelated o computer se and Size showed slight ositiverelationshipboth ignificantnly t .05levels).Therewasalso a direct elationshipetweeneveral fthe ommunitynviron-



21/29

Innovation, Bureaucracy, and Public Policy 197TABLE 6. PARTIALS FOUND STATISTICALLY SIGNIFICANT IN TESTING MODEL OFPRODUCT INNOVATION IN THE COMMON FUNCTIONS OF CITY GOVERNMENT

rpOrganizationalCharacteristicswith nnovation (n = 71)Publicworks epartmentoesvehiclemaintenance ithMainstem se ....-.......... -.30OrganizationalEnvironmentwith nnovation (n = 91)CityADP score withMainstem use .................................... .... .. ........................... .29Demand with Innovation (n = 93)PercapitapublicworksxpenditureithMainstemse ................................... .25Demand withOrganizationalCharacteristics n = 94)Totalpublicworksxpenditureith ublicworks epartmentoesvehicle maintenance. ... .... ............................28Percapitapublicworksxpenditureith ublicworks epartmentdoesvehiclem intenance ...................................................................... .37Community nvironmentwithOrganizationalEnvironment n = 100)SES w thcityADP score ..........................................................................................34Community nvironmentwithDemand (n= 157)Size with otalpublicworks xpenditure.............................. ...... .... .61SES withpercapitapublicworks xpenditure..................................... .. .19ment variables and demand. As expected,the coefficient etweenSize and totalPWexpenditureswas substantial. The relationshipbetween SES and per capita PWexpenditureswas also significant.In summary, hree variables representing emand, the organizationalenviron-ment,and organizationalcharacteristics irectlynfluencethe use of the Mainstemvehicle fleetmanagement program in largerAmerican cities. It appears that thesocioeconomic characteristicsof the community, s captured by the SES factor,formthe base forMainstem adoption. A high socioeconomic level in the citypro-vides a high level of PW (and other) supporton a per capita basis. As was dis-covered in earlier models and affirmedhere, SES is also significantlyelated to ahighlydeveloped citycomputer capability.Initially t was expected that cities withvery ow ADP scoreswould be Main-stemusers based upon the rationale that cities with low-levelcomputer capabilitieswould be forcedto contract out with Mainstem to obtain the kind of manage-ment informationnecessaryfor effective ehicle fleetmanagement. This was notthe case however in fact, ust theoppositeoccurred. Cities withhighADP scoreswere themostprobable Mainstem customers. This suggests n alternative xplana-tion. It is quite possible that a high ADP score indicates that cityofficialswerefamiliar with and used computer compiled information n theirday-to-dayopera-tions. It is perhaps this experience that makes local bureaucrats amenable andeven enthusiastic bout the typeand level of information hat Mainstem can pro-vide. Thus the only satisfactory xplanation of the ADP-Mainstem relationship sto assume that a high cityADP score is representative f a certain bureaucraticexperience - an experience whereby bureaucrats see positive benefit to the kindsof information hat Mainstem provides.The relationshipbetween per capita PW expenditures 1969) and Mainstemuse was as hypothesized.High cost PW in 1969 showsan independent relationshipwith Mainstem use. That is, controllingfor the vehiclemaintenance function ndthe othervariables in the model, a high cost of public works n the late 1960s ledto the adoption of Mainstem in the 1970s - presumably as a cost reduction mech-anism. Again, then,themodel fortheadoption of theproductinnovationsupportsthe view that innovation adoption decisions are based on rational considerations.Innovation adoption appears to be based upon a desire to reduce cost.



22/29

198 WesternoliticalQuarterlySUMMARY

Model testingwas highly uccessful ortheprocess nnovations ut showedlimited uccess n explaining roduct doptions.Table 7 summarizeshefindings.A D in Table 7 indicates hat herewas a direct ighorderpartialbetween neormoreofthe ndependentariablesn a given ategorynd innovationdoption.An I, on theotherhand, ndicates nly supportingelationshipnthechainorprocess.Thus fortheprocessnnovations,ize (demographicharacteristic) asshownto be a majordeterminingactorn innovationdoption lthough he n-fluencewas largelyndirect. ize influencednnovationsnly ndirectlyxceptforthedirectnfluencefSizeon computer sebypolicedepartments.hesefindings,then, re somewhatt oddswith arlier olicy tudieswhich ndicate hat ize s adirectdeterminantf output. The findings ere are not necessarilyonflicting,however; heymerely ointout differencesetweenpolicy nd bureaucraticn-novations. Size has been shownto be a directpolicydeterminant. n terms fbureaucraticdoptions, owever, heinfluencef city ize is largelyndirect.The same conclusionsre basically rueconcerninghe otherdemographiccharacteristicsf thecity. Again,forprocessnnovations,he nfluencef socioeco-nomicvariables n innovationdoption s substantial,ut ndirect.In examiningnnovationdoption, hestudy howeda distinction etweenneedand amenity doptions.The need-amenityistinctionouldonlybemadeona posthocbasishoweverndmayormaynot pply o the nnovationtself. t wasdeterminedfter hefact nd is baseduponadoptionpatterns.Programsf ndi-vidualizednstruction,or xample, o not ppeartohave beendeveloped xplicitlyforeither ow or highexpenditurechool districts. he innovationtself,hen, snotparticularlymenityriented. t isonly heposthocorafter hefact xamina-tion fadoption atterns hich uggestshis istinction.A wordofcaution s warranted, owever. t is possible hat, ver the entiretimeperiodof the S curveofadoption, heamenity atternmaydisappear.Itmaybe thathighSES cities rovide hebest laboratories or ryingutnew andunproven ducationalprograms.The samemaybe true of the other pparentamenitydoptiondiscovered ith he ibraryrocessnnovation.Regardless f thepossible ong-termutcomes, owever,heposthoc exami-nation fadoption atternseems osupporthe deaof need-amenityontinuum.Several nnovationsppearedto be adopted n an attempto upgrade ubstandardperformancertosignificantlyeduce hegovernmentalnit's peratingxpenses.On theotherhand,othernnovationdoptionpatterns uggestn attempto im-prove lready dequate performance. ther nnovationshowed atternsf dop-tion uggestingoth needandamenityase.Conservative r liberal communityttitudes,t leastas measuredhere,have little o do withtheadoption f innovation y ocal bureaucracieslthoughin twocases (again bothprocess nnovations), ommunityttitudeswererelatedto funding evels and funding rioritieswhich, n turn, nfluencednnovationadoption.Withinthe organizationalnvironment,irector indirect ssistance y thefederal r stategovernment as an importanteterminantfprocess nnovationadoption. Of equal importance, owever,was the intergovernmentalssistanceprovided y ocal government. he obvious xample s the mportancefa well-developedcitycomputer apabilityn stimulatingomputer se by other ocalunits. The significantnfluencef federalprogramsn innovationdoptionwasclearly ndicatedby the importance f the HUD MIP programn stimulatingcomputer se.Did formof governmentreformismr professionalism)ffect nnovationadoption? Only in a very mallway. Otherthings eing equal, reformeditiesweremore ikely o have highly eveloped omputer apabilitieswhich, n turn,



23/29

TABLE 7. DIRECT AND NDIRECTLINKAGES N INNOVATIOCOMMUNITY POLICY ORGANIZATIONALENVIRONMENT DEMAND ENVIRONMENT

Intergov- PrivateDemo- Cultural ernmental Professional Sectorgraphic Values Relations Relationship InfluenceProcess Innovations

Housing ComputerUse ................... I I D, I D I *Individualized Instruction ............... I I I I *Library ComputerUse .................. I-Police Computer Use ..................... D, I D, I D I *

Product InnovationsHousing Prefab Use ..........D........... - DVideotape Recorders ...................... - _LibraryTheftSystems ................... -.Mainstem Use ........-...................... I D, I D I *

- No significantelationship* Not tested,data not available



24/29

200 WesternoliticalQuarterlyled tohigherevels fadoptionnotherocal units.Overall,however,heeffect freformismslimitednd indirect.Resource vailability asan extremelymportanttimuluso nnovationdop-tion. Resource vailability nquestionablyrovides hemajor mpetus o innova-tionadoption.The characteristicsf theorganization, ith he xceptionforganizationize,did notgenerallyffectdoption evels. Withtheexception fthetrend owardlibrary onsolidationnd regional rganization,nd itsrelationshipo computeruse, it ma

1978 Bingham - Innovation, Bureaucracy, And Public Policy - A Study of Innovation Adoption by Local Government

Documents