Robert H. WilsonPhilips International Visiting Professor at the EAESP/FGV,Mike Hogg Professor of Urban Policy at LBJ School ofPublic Affairs and Director of the Urban Issues Program atthe University of Texas at Austin.E-mail: rwilson@mail.utexas.edu

KNOWLEDGE AND POLICY

ACTION IN URBAN

DEVELOPMENT:

if we know so much, why is the

urban condition not improving?

RAE - Revista de Administração de Empresas • Jan./Mar. 2000São Paulo, v. 40 • n. 1 • p. 47-55

Administração Pública

47

RESUMOApesar da geração de riqueza potencial das grandes cidades no mundo, as condições precárias de vida de uma grandeparcela da população urbana persistem. Embora as cidades sejam muito estudadas, a política urbana é, aparentemente,ineficaz. Este trabalho assume uma abordagem da política pública na análise da relação entre conhecimento e açãogovernamental. As barreiras à melhoria da política urbana estão associadas com a incapacidade dos governos de atuar ecom a política envolvida na tomada de decisões em regimes democráticos. O artigo recomenda que uma visão pragmáticada geração do conhecimento deva ser utilizada.

ABSTRACTIn spite of the wealth generation potential of the world’s large metropolitan cities, poor living conditions for much of theworld’s urban population persist. Although the city has been widely studied, urban policy often remains ineffective.The paper adopts a policy process approach to analyze the relationship between knowledge and governmental action.Impediments to improving urban policy are found in the inadequate capacity of government to act and in the politics ofdemocratic decision-making. The paper recommends that a pragmatic view of knowledge generation be adopted.

PALAVRAS-CHAVEPolítica urbana, pesquisa urbana, capacidade governamental.

KEY WORDSUrban policy, urban research, governmental capacity.

48 RAE • v. 40 • n. 1 • Jan./Mar. 2000

Administração Pública

©2000, RAE - Revista de Administração de Empresas / EAESP / FGV, São Paulo, Brasil.

A sense of frustration can be heard inacademia and research communities in manycountries. Why does urban development policynot incorporate the current state of knowledgeabout cities and urban problems? Governmentofficials and others argue that academicresearch is irrelevant, if not politically naive.The fact that some research may be critical ofexisting policy further exacerbates the lack ofmutual respect and constructive interactionbetween researchers and policymakers.Knowledge appears not to be sufficient forimproving the urban condition. Although thebreach may be great, the state of the world’scities compels us to strive for better and moreeffective urban policy. This paper will examinethe challenges to bringing to bear research andknowledge on policymaking in order to achievemore effective government action in cities.

A cursory review of the relationship betweenknowledge and public policy reveals manyexamples of technological or scientif icunderstanding informing public policy todramatic effect. The history of public health,for example, convincingly demonstrates thatscientific research can be applied to publicpolicy in ways that substantially improve thehuman condition. The remarkable expansion inthe understanding of pathological agents andthe development of effective mechanisms toimpede these agents, ranging from vaccinationsto sewerage and drainage, lead to improvedliving conditions. International cooperationcontributed to past success in public healthresearch and policy. Certainly not all publichealth concerns have been resolved, forresources are inadequate in many parts of theworld and new public health challengescontinually emerge. Notwithstanding currentand future challenges, the integration ofscientific research in this field of governmentalaction has produced impressive results.

Moving from public health to the widerrange of urban issues, the contribution ofresearch and knowledge to improvement inurban policy is more difficult to identify. To besure, the field of urban research is extremelydifficult to demarcate. Among researchers amulti tude of problem definit ions anddisciplinary and methodological approachesexists. Further complicating this field ofinquiry, the nature of the problems variessubstantially across cities and countries. As

geographical spaces in which a great range ofhuman experience occurs, ci t ies pose aenormous challenge to human understanding,yet it is in these spaces that the engine forwealth creation and human development isincreasingly located. We are not modest instriving to develop a rigorous understanding ofurban development nor in expecting thisresearch to be applied to improving the urbancondition through public policy. But the stateof the urban condition in the world suggeststhere is much to learn and that effective urbanpolicy remains elusive.

I initiate this discussion with a simpleproposition: for government to act, someconceptual framework or understanding of thesociety, i ts problems, and the role ofgovernment in addressing societal needs isnecessary. Although numerous examples ofirrational government action come to mind,government action is generally premised onexpectation of the consequence of action, whichimplies some understanding of the causalrelationships and systems operating in society.1

The framework may be wrong or incomplete,but proposed actions are necessarily developedwithin some conceptual structure. As manifestin policymaking, conceptual structures may beimbued with the part icular interests ofindividuals or groups and may not incorporatebroad, public interests. In addition, there is noexpectation that rationality and good judgmentwill replace power in the exercise ofgovernmental action in the foreseeable future(Lindblom, 1990). Furthermore, governmentalaction can be taken without a recognition of therights of all citizens, may exclude certainsocietal groups, and may be taken on the behalfof particular private interests. Nevertheless,some understanding or expectation aboutimpacts of action is inherent. How does thisconception come to be formed in the policyprocess and how can it be changed or improvedby the application of knowledge and research?

To explore the relation of research andknowledge in urban development, I suggest aframework that focuses largely on urbanpolicymaking and governmental action. Inparticular, I wish to focus our attention on theway knowledge and information are utilized inpolicymaking. After presenting a simple modelof government action and the policymakingprocess, the capacity of government to act will

Knowledge and policy action in urban development

RAE • v. 40 • n. 1 • Jan./Mar. 2000 49

The history of public health, for example,

convincingly demonstrates that scientific

research can be applied to public

policy in ways that substantially

improve the human condition.

be examined. Knowledge and understandingwill not be effectively utilized if the capacityof government to act is inadequate. Thediscussion then turns toward the opportunitiesand impediments for the integration ofknowledge in policymaking.

URBAN POLICYMAKING

For purposes of this paper, a four-phaseconception of the policy process – agenda setting,policy formulation and choice,implementation, and evaluation –will be used.2 Information andknowledge enter the policymakingprocess in a variety of ways anddifferent types of knowledge areutil ized in the different phases.Although this conception of thepolicy process can be applied tomany different policy arenas, itsapplication to urban policymakingdraws attention to several distinctivefeatures. Urban policymaking affectsa relatively wide range of governmentalfunctions, making it a complex policy arena. Thequestion of private interests versus publicinterests is particularly problematic given theextensive externalities found in urban activities.Finally, in both unitary and federal governmentalstructures, intergovernmental relations arecritical since urban policy will frequentlyconfront questions of local resource disparities.

In the agenda-setting phase, issues ofconcern to a country and society are placedbefore governmental bodies which adopt actionsor policies to address the issues. Thegovernmental body is generally one withlegislat ive responsibil i t ies but in somecircumstances may have executive or evenjudicial roles. Issues become part of the publicpolicy agenda in a variety of ways.

Triggering events – natural or manmadedisasters, political crises, and, occasionally, thepublication of research – can focus public andlegislative attention on a problem. Or, issuescan be placed on the policy agenda by organizedgroups that bring pressure on an appropriategovernmental body. Organized groups –including citizen, business, occupational, andreligious groups – often utilize the media andother mechanisms to gain attention and placeitems on the public agenda.3

Applied academic research can be influentialin this phase. The analysis of such questions aspoverty, distressed communities, environmentalpollution or traffic congestion can be quitehelpful in terms of defining problems properly.The analysis of such problems, undertakenindependently of governmental authority, canprovide significant, alternative understandingsof problems. Such efforts can be reinforced bythe emergence of policy communities ornetworks within countries which provide

persistent and long-term attention to an issue.International institutions or actors may also beable to influence the agenda of a country. TheUnited Nations-sponsored Habitat II, held inIstanbul in 1996, provided the opportunity foran international policy community to legitimizethe question of housing as an important publicpolicy concern.

Political power can be important in thisphase of the policy process. Powerful groupscan control agendas, and even prevent issuesfrom being addressed as public policy concerns.In many countries, traditional oligarchies havebeen able to control national policy agendas.Disenfranchised communities without politicalinfluence may have difficulty convincing thepolicymaking process that their concerns arelegitimate and deserve to be placed on theagenda of a public decision-making body.Recent efforts to promote democracy andbroader access to democratic policymaking maywell affect the range of issues on the policyagenda in a country.

For ongoing governmental activities, theperiodic budget process, usually undertaken bya legislative body, may bring certain policies,agencies, or actions to the foreground of thepublic policy discussions. The extent ofdiscussion and debate during the budget process

50 RAE • v. 40 • n. 1 • Jan./Mar. 2000

Administração Pública

The remarkable expansion in the understanding

of pathological agents and the development of

effective mechanisms to impede these agents,

ranging from vaccinations to sewerage and

drainage, lead to improved living conditions.

on specific governmental endeavors will varyconsiderably – by level of government, salienceof the endeavor, and extent of interest amongaffected constituencies. Over the decades anumber of budgetary innovations have beenintroduced to improve accountability andenhance oversight of the budgetary process.Developing citizens’ ability to understandbudgeting and to participate in budget debateshas been identified as a key to enhancingdemocratic practice (Singer, 1996). To theextent that budgeting involves a process ofnegotiation and compromise among manyinterested parties, citizens groups need to havetheir independent sources of analysis andresearch. These mechanisms can lead to public

discussions of governmental policy and, thus,help shape the policy agenda. Academicresearch of budgeting questions may beimportant in this phase, but the media and otherintermediaries are likely to be more helpful indeveloping public understanding of budgetarymatters.

Once an issue is placed on the public agendaand pressure builds for government action, apolicy must be formulated and adopted .Formulation and adoption are distinct butrelated steps. Most issues produce competingpolicy formulations or approaches which canoriginate in many different quarters. Thelegislative body itself can formulate alternativesbut executive governmental agencies, politicalparties and organized interest groups canpropose alternatives and bring these to thelegislative body.

Research findings occasionally producedramatic results during this phase. The severityand breadth of the cholera epidemic in themiddle of the nineteenth century in Londonmade it a paramount public policy issue.

Through applied research, water wells wereidentified as the source of the epidemic and thisfinding suggested a policy to reduce, if notsolve, the problem. The experiencedemonstrated the importance of public healthresearch in problem definition and policyformulation and established the legitimacy ofan emerging profession.

For many issues, the specific legislationneeded for policy formulation requiresknowledge not only of the specific issue, but alsoof the relevant governmental apparatus. Thisexpertise is available in the governmentalapparatus itself and, consequently, governmentalagencies often design or assist in designing policyproposals. Expertise outside the legislative and

executive branches mayalso be important.Industries subject tosignificant governmentalregulation, such as theutility and real estateindustries, are a source oflegislative proposals. Theformulation phase in areassuch as environmental ort e l e c o m m u n i c a t i o n spolicy requires extensivetechnical and scientific

expertise. This phase does not necessarilyrequire new research, but can certainly benefitfrom the basic understanding of cause and effectrelations for specific problems, as was the casein the cholera epidemic.

The outcome of the legislative processconstitutes a policy choice or decision. Thischoice is generally the prerogative of alegislative body at local, provincial, or nationallevels. A great many factors affect the specificpositions taken by legislators. If they are electedofficials, the interests of their constituentscondition their views on policy formulation.Constituencies take many forms and include thevoting citizens, the general public, politicalparties, financial backers, and organized interestgroups. The ability of any single constituencyto affect policy decisions will depend on itspolitical power and its ability to cooperate withother interests through coalitions. As a result,the adopted policy frequently represents anamalgam of competing formulations andincorporates ambiguous goals. At this point inthe policy process, knowledge and

Knowledge and policy action in urban development

RAE • v. 40 • n. 1 • Jan./Mar. 2000 51

understanding of an issue become subordinateto the exercise of power and compromise.

Once a policy has been adopted, in thefollowing phase of the policy process it is putinto action. The implementation of policyalmost invariably involves actions by someagency of government thatcreates the mechanisms bywhich the pol icy isexecuted. Al though thepurpose of this phase mayseem clear, implementationcan be quite complex. Animprecise s ta tement ofgoals and objectives in theoriginal policy may forcethe implementinginst i tut ion to interpret the intent of thepolicymakers. The implementation of many,if not most, policies requires, or is affectedby, actions of agencies other than the primaryimplementing agency and even by other levelsof government . The in te rgovernmenta lcontext can be critical to successful policyimplementation.

Governmental institutions, particularly largebureaucracies, are not neutral, technical agentsin the implementation phase. Rather, theyapproach implementation with their own set ofconcerns, interests, knowledge, constraints, andconstituents. Professional norms of conduct,values held by administrators, the technicalcompetency of agencies, and bureaucraticpolitics all affect implementation. Particularlyin nonroutine policies, as in the creation of newprograms, the institutional context – especiallythe government agency’s ability and will torespond to change – will influence governmentperformance. For example, the departure ofcolonial powers from Africa or the displacementof military governments often left in placeadministrative systems that had limited abilityand interest in responding to issues previouslyexcluded from the policy agenda of the earliergovernments.

Governmental institutions often hold criticalinformation in the policy formulation processand may disagree with elected political leadersabout particular policy directions. Policymakersin cities and local governments are frequentlydependent on bureaucracies for analysis ofproblems, policy formulation, and evaluation.In this situation, the separation between policy

decisionmaking and policy implementation isobscured, if not rendered meaningless. Moreimportantly, it suggests that a potential avenuefor introducing better knowledge andinformation into policymaking is through thetraining of individuals in government.

The evaluation of policy and feedback intoa new cycle of policymaking occurs in a numberof ways. Evaluation is typically understood tobe a technical activity in which specific policyoutcomes are measured against the originalpolicy objectives in order to determine policyeffectiveness. The costs associated with thepolicy or program may be compared with itsbenefits in order to examine its efficiency.

Beyond this narrow and technical focus,evaluation also occurs through the changingatt i tudes of the public, in general , andpolicymakers, in particular. In a complexprocess of social inquiry, members of societycan develop new ways of understanding society,which can lead to developing the political willto force the changes in governmental action(Lindblom, 1990).4 Changes in attitude maylead to modifications or even reversals in policydecisions over time. The extent of democraticpractice and openness in a country will affectthe pace and potential of social inquiry. Changesin attitude may result from formal evaluationsof existing policies, but they may also be theresult of shifts in general public attitudes towardand understanding of issues.

Lindblom observes that in this process ofchanging social attitudes, knowledge is notenough. 5 Without vol i t ion, a t leas t indemocratic societies, government action willnot be focused on social problem solving.Knowledge can serve as an important supportfor social problem solving by helping to frameissues, analyze problems, and capture learningfrom past experience. But without volitionexercised in the policymaking system, policy

As manifest in policymaking, conceptual

structures may be imbued with the particular

interests of individuals or groups and may

not incorporate broad, public interests.

52 RAE • v. 40 • n. 1 • Jan./Mar. 2000

Administração Pública

choice and governmental action will not beresponsive to social concerns. The promise ofdemocratic and open societies is that elites willnot be able to control the formation of socialvalues and attitudes and, as a result, popularand broad-based vol i t ion for act ion canemerge.

The reservation of one-third of the seats forwomen in the Panchayati Raj elections in Indiarepresents a remarkable moment in the changeof social attitudes (Institute of Social Science,1995). This change was the result of efforts ofmany people, including academics, journalists,women’s organizat ions, and progressivepolitical leaders, and was supported by variousinternational efforts. Academic research andmedia reports on the conditions and problemsof women helped establish the legitimacy ofthe issue and were effective in changing socialattitudes about women’s role in society andgovernment. The adoption of democraticelection systems at the village level in Chinais similar. These dramatic changes are theoutcome of complex social processes, partiallythe result of improved understanding of socialconditions, but also a reflection of politicaladvocacy and broad change in social attitudes.

At the international level, understanding ofand attitudes toward urban problems havecertainly changed in recent decades. Thedeleterious effects of high pollution levels incities are now broadly recognized and theenvironmental sustainability of communitieshas gained acceptance as a policy goal. HabitatII and other international forums not onlysupport exchanges among researchers andothers to understand transnational trends andproblems but these meetings also givelegitimacy to issues that may have previouslybeen ignored in individual countries orcommunities. In other words, transnationalefforts can help set the policy agenda inindividual countries.

The pace of changing social attitudes andpublic policy is much slower than that ofexpanding knowledge gained from urbanresearch. Furthermore, the openness of thepolicymaking process to changes in socialattitudes is very much dependent upon politicalstructure and the extent of democratic practice

in a country. The great frustration inurban policy research communitiesderives, in part, from the slow paceof change in social attitudes, butthese processes may represent thebest hope for improving urbanconditions.

CAPACITY TO ACT

The policy process provides a frameworkfor understanding how government actioncomes about and how it can be modified overtime, but it does not explicitly address the issueof government’s capacity to act. A soundunderstanding of the urban dilemma may befound in the policy process but governmentmay not have the capacity to act upon thisunderstanding at a level sufficient to producethe desired impact. Without the capacity to act,the potential impact of knowledge will not berealized. This point may seem gratuitous orappropriate only in extreme cases. But whenaddressing government action in the area ofurban development , the point has broadapplication. In many countries of the world thecapacity of local government to act may beseverely lacking. Local governments arefrequent ly i l l -prepared to assumeresponsibilities for policymaking, resourcemobilization, and program implementation.

The question of governmental capacity isparticularly important in the many countriesof the world decentralizing their governmentalstructures. Decentralization is believed to leadto more effective governmental action and topromote democrat ic pract ice , but thedifficulties generated by these policies havecalled attention to the question of capacity inlocal government. In countries like Mexico,Brazil, and India, local governments, lackingadequate financial, human, and institutionalresources, have frequently not been preparedto assume the new responsibilities required bydecentralization. The degree of preparation oflocal government varies substantially from

Knowledge and understanding will not

be effectively utilized if the capacity

of government to act is inadequate.

Knowledge and policy action in urban development

RAE • v. 40 • n. 1 • Jan./Mar. 2000 53

country to country and even among regions ofthe same country.6

The intergovernmental dimension of urbandevelopment policy has become prominent inthe context of decentral izat ion. Localgovernments are linked to higher levels ofgovernment in several ways,including through constitutionaland statutory frameworks, fiscalrelations, joint responsibilitiesof program implementation, andpolitically. The specific set oflinkages varies substantiallyacross countries. The promise ofdecentral izat ion, in whichrela t ively r igid centra l izedsystems would devolve powersand cooperate with lower levels ofgovernment, has proven difficult to fulfill inpractice. Decentralization holds the potentialfor improved urban development policy, butthis policy may operate within a set of complexintergovernmental relations that can constrainif not impede effective local action.

PRODUCTION OF KNOWLEDGE ANDURBAN POLICYMAKING: APRAGMATIC VIEW

That policymaking requires different typesof knowledge and understanding is not a newobservation. Confucius, the ancient Greeks, andothers gave much consideration to knowledgeand social action. One view emerging fromclassical Greek thought argued that governingrequired two types of knowledge (Dahl, 1989).Moral knowledge is needed when issues of rightand wrong, fairness, and judgments are central.For other types of situations and issues, anempirical or instrumental knowledge ofmankind and society is also required. Over thecenturies, a series of epistemological questionshave been raised concerning these views ofknowledge, but a contemporary version ofPlato’s philosopher king, leading ascientif ically-directed society, seemsfundamentally incompatible with currentattitudes toward democracy.

The policymaking process described aboveadopts a more pragmatic and practical attitudetoward knowledge. The policy process isprimarily concerned with acting, not knowing.The process is imbued with interests and the

various actors in the process may not share thesame world view, thus generating conflict.Decisionmakers may be concerned with theaccuracy of their understanding of the worldbut public policy is first and foremost aquestion of action.

Drawing from another Greek, Peat t iesuggests Aristotle’s ideas about different kindsof knowledge, especial ly phronesis , arerelevant (Peattie, 1995). Aristotle’s term canbe translated as knowledge of what to do inparticular circumstances. I would suggest thatthis position requires, in policymaking, notonly an understanding of a particular problembut a theoretical or conceptual knowledge ofsociety and social organization in order toproduce effective policy. For example, toimprove an infrastructure system in a city,policymakers must have some expectationsconcerning the impacts of changes in relativeprices embedded in a policy instrument. Thereis an element of theory, of markets and marketbehavior, but this theoretical knowledge mustbe applied in existing infrastructure markets.Furthermore, effective policy must incorporatea practical understanding of governmentalorganizations and structures. Organizationaltheory may be helpful , but a specif icunderstanding of the organization or agencywhich will implement the action is essentialto success.

Those of us in academic settings tend toprivilege theoretical knowledge andunderstandings derived from the rigorousapplication of the scientif ic method andinvolving high standards of data quality.Without encouraging the abandonment of thistype of inquiry, we should recognize that suchknowledge may have limited direct applicationin policymaking for several different reasons.Disciplinary boundaries may impose anarrowness of focus, perfectly justifiable for

One means for introducing new ideas

or knowledge into policymaking

is through the training received

by public officials.

54 RAE • v. 40 • n. 1 • Jan./Mar. 2000

Administração Pública

academic inquiry, that renders the findingsirrelevant for the messy, if not boundless,social problems addressed in policymaking.Furthermore, the particular type of knowledgeavailable in academic fields may simply notbe on the policy agenda of government. Forexample, extensive, rigorous research of theurban informal sector exists, but in mostcountries the problems of this sector do nothold high priority for public policy. I wouldnot argue that scientific inquiry should beforemost concerned with policy relevance.Understandings developed in academicresearch communit ies , I bel ieve, areindispensable to the creation of appropriateframeworks for analyzing complex problems,even if policymaking does not explicitly adoptthese frameworks. It would be foolish, forexample, to disregard theoretical work inchemistry and biology as we develop policyconcerning environmental pollution. But ifindividuals in research communities look fordirect policy relevance, a better understandingof the use of knowledge in policymakingbecomes necessary.

The understanding of cities incorporated inthe policy process originates in many differentsettings. The very practical informational andknowledge requirements of policymakingmake theoretical knowledge of limited use tothe policy process. Rather, knowledge andinformation used in policymaking originate insuch places as governmental organizations,think tanks, for profit and nonprofit privateorganizations, and even the media.

Government action represents the outcomeof a process in which political power is acritical element, a process in which interestsand competing views of the world exist.Different interests come to policymaking withnot only unique, interest-based views of theworld and cities, but their own research andanalysis. The competition among alternativepolicies occurs in a political context. Thecorrectness of ideas and true understanding arenot unimportant, but democratic policymakingis contentious and scholarly research rarelyimportant to the outcome. More often than not,research is used to justify political positionsrather than change or modify positions in lightof the research.

Theoretical or instrumental knowledge mayalso be relatively unimportant to citizens

articulating their interests in the agenda settingprocess. In Bangalore, a relatively simplestrategy of giving visibility to citizen viewson public services has had a very importanteffect on improving the quality of the services(Paul, 1996). A nonprofit organization createda report card in which citizens graded publicservices in the city, and the results werepublished by the local newspaper. The gradeswere quite low and the service providers weresuff ic ient ly embarrassed that bet terperformance followed. In what sense wastheoretical knowledge useful in this context?The report card attempts to create pressure onagencies which have no pressure f romcompeting organizations. This initiative maydraw from theory on markets and organizations,but the application of this knowledge and thedissemination of the findings through an efficientchannel of communication have led toimprovement in services.

The policy process utilizes information frommany sources, including from the governmentbureaucracies and agencies. One means forintroducing new ideas or knowledge intopolicymaking is through the training receivedby public officials. These public officials mustobtain a view of urban development from somequarter, and the training institutions may be aneffective means for promulgating ideasconcerning urban development. This suggestsa long-term strategy, one in which researchersserve as teachers and trainers, not only in post-secondary insti tutions but also throughprofessional associations, political parties, andother channels.

A related strategy to improve the quality ofurban policy is the development of researchcompetencies in government. Not infrequentlygovernment agencies and officials will call uponacademic researchers for assistance inunderstanding problems. In many countries,government itself has started developing its ownresearch or analysis capabil i t ies. Suchendeavors generally appear first at the nationallevel, where the necessary resources are morelikely available, but local or provincialgovernments, with large urban populations,might be well served by developing internalresearch capacity. The academic communitymight find assisting such endeavors a usefulmechanism for disseminating research findingsand methods into the public sector.

Knowledge and policy action in urban development

RAE • v. 40 • n. 1 • Jan./Mar. 2000 55

It has been asserted here that for governmentto act in cities, some understanding of the urbancondition must exist in the minds of the policyactors. These frameworks can be influencedthrough a variety of mechanisms, butlegislators, advocates, civil servants, and othersare not likely to digest and utilize knowledgefound in formal research publications.Academic research will frequently not bepresented and disseminated in a fashion thatmakes it easily accessible to those in positionsof power or to communities that hold or seekpolitical influence. If influence on public policyis a goal, research findings must make cleartheir relevance to current issues and they mustbe placed in the channels of communicationlikely to affect the process of social inquiry.

EPILOGUE

Decentralized urban policymaking systemsrequire a better understanding of the urbandevelopment process and provide a challengeto research communities. Acquiring basic

descriptive information, much less anunderstanding of innovation, becomesincreasingly difficult as local governments aregiven more discretion for managing the affairsof their cities. Decentralization creates a newcontext for urban policy, and how localpolicymaking institutions and citizens areresponding to this context is unknown. Theimpacts of the broad expansion of elections inlocal government on interest representation andpolicymaking are relatively unexplored.International research communities shouldmonitor these changes and exchange findings.This collaboration will provide a basis forcritiquing local practice, for identifying goodpractice and disseminating information aboutthese practices. The collaboration will createand nurture policy communities which can bemobilized at times of triggering events whenthe opportunities for change are greatest.Investment in this research agenda is vital, eventhough its relevance for policymaking andsocial inquiry might not be apparent untilsometime in the future. �

1. We need not worry, for now, about the normativedimension of governmental action, as contrasted forexample by the liberal democratic concept in whichgovernment action is to create conditions where societalresources can be freely developed by individuals or amore aggressive concept in which government actionis intended to promote change and development.Conceptions vary widely around the world andfrequently citizens in a single country will dispute theappropriate purpose of government.

2. For a critique of the four-phase policy process model,see NAKAMURA, R. The textbook policy process andimplementation research. Policy Studies Review, v. 7,n. 1, p. 142-154, Autumn 1987 and SABATIER, Paul A.Toward better theories of the policy process. PS:Political Science and Politics, v. xxiv, n. 2, p. 147-157, June 1991.

REFERENCES

DAHL, Robert A. Democracy and its critics. New

Haven: Yale University Press, 1989. Chapter 5.

INSTITUTE OF SOCIAL SCIENCE. Status of

Panchayati Raj in the states of India, 1994.

New Delhi, India: Concept Publishing Company,

1995.

PEATTIE, Lisa R. An approach to urban research in the1990s. In: STREN, Richard, BELL, Judith Kjellberg (Eds.).Urban research in the developing world. Toronto: Centrefor Urban and Community Studies, University of Toronto,1995. Vol. 4: Perspectives on the City. p. 393.

SINGER, Paul. Budgeting and democracy. In: WILSON,Robert H., CRAMER, Reid (Eds.). International workshop

on local governance: second annual proceedings.Austin, TX: LBJ School of Public Affairs, 1996.

LINDBLOM, Charles. Inquiry and change: the troubled

attempt to understand and shape society. New Haven:Yale University Press, 1990.

PAUL, Samuel. A citizen report on public services:mixing barks and bites? In: WILSON, Robert H.,CRAMER, Reid. (Eds.). International workshop on

local governance: second annual proceedings. Austin,TX: LBJ School of Public Affairs, 1996.

NOTES

3. One critique of the four-phase characterization

of the policy process has been advanced by

Kingdon. (KINGDON, John W. Agendas,

alternatives, and public policies. Little: Brown,

1984). Instead of the sequential steps of agenda

setting and policy formulation, Kingdon proposes

three largely unrelated “streams”: a problem stream,

which consists of the knowledge about some

problem and the effects of past policy, a community

or network of advocates and specialists, and a

political stream. The opportunity for major policy

changes occurs when these three streams coincide

to create windows of opportunities and such

occurrences are rare.

4. An early formulation can be found in LINDBLOM,Charles, DAHL, Robert. Politics, economics, and

welfare: planning and political-economic systemsresolved into basic social processes. New York: HarperTorch Book, 1953.

5. For an analysis of this phenomenon in terms of asocial learning, see Lindblom (1990).

6. For a very useful four volume series, see Urban

research in the developing world, edited by RichardStren and published by the Center for Urban andCommunity Studies at the University of Toronto (VolumeI – Asia, 1994; Volume 2 – Africa, 1994; Volume 3 –Latin America, 1995; Volume 4 – Perspectives on theCity, 1995).