Ambio, Vol. 27, No. 2, pp. 112-117

tainable development' agency and between the new and existingagencies. But the internal conflicts would be less severe andthe external ones more manageable than before. Within the newagency, experts of different sectors of environmental protectionwould naturally have diverging opinions of what should be doneto move toward sustainable development. But this sort of professionalargument over the means to achieve ends is the normin any government agency. The point of administrativedecoupling is to transform the pull between long-term environmentalconcerns and short-term economic ones, which provedso elemental in sustaining the officials' decision-making dilemmain the two cases, from a conflict within an agency into aconflict between agencies. For such conflicts, there already existresolution mechanisms, such as negotiation, arbitration andthe courts.The recommendations run against the accepted wisdom ofmany recent panels on global environmental policy, such as theBrundtland Commission and the UN-Rio Conference, which ad-vise the institutional integration of environmental and economicconcerns (3, 30). Yet the decoupling of environmental and economicinterests is by no means extraordinary-it has been notjust recommended, but done before. The separation of irrigationand drainage is nothing less than a divorce between the admin-istrative system of irrigation from the environmentally tuned normativesystem that restricts drainage management. According toHabermas, one way for an organization to avoid a legitimationcrisis in the long run is by decoupling the administrative systemunder threat from the norms that need justification (31).Similarly, the recommendation to make a clear distinction betweenimplementation and regulation is not new. Wilson andRachal, for example, posed themselves the question "Can thegovernment regulate itself?" and discovered the answer wasnegative (32). Furthermore, the principle of creating autonomoussocial agents through institutional design is already much utilized.Institutionally determined autonomy enables judges tomake decisions according to their professional convictions, universityprofessors teach what they believe in and voters vote asthey please, without fearing social sanctions for doing so (8, 33).The creation of autonomous social agents with the mandate tomake long-term environmental policy would therefore be an expansionof an already well-established principle in social organization.DISCUSSIONEach of the two cases reveals a perspective on environmentalmanagement from which competing social interests come intofocus. In the Finnish case, that perspective is the social steeringof environmental management, and the lesson to be learned isthat organizational distribution of regulatory and implementingpowers has significant influence on the outcome of environmentalmanagement. The California case is about the operationalcontrol of environmental technology and shows how inappropriateadministration of technology can lead to significant politicaland economic conflicts of interest.But the proposed administrative reforms also raise the spectreof conflict between technocratic rationality and democraticbalancing of interests. What guarantees that the proposed autonomousdrainage agency in California, for example, will not developinto the kind of monolithic power base that many votersin the state think the irrigation bureaucracy today is? What guaranteesthat the values of the experts will represent the values ofthe population at large; that the experts have sufficient knowledgeon issues beyond their expertise; or that the experts canlegitimately resolve the distributional conflicts associated withthe various solutions? Representation and elitism are issues thatalready the early systems theorists tackled with when ponderingthe possibility of allowing a panel of experts to act as a surrogateof future generations (34).While the issue of technocracy versus democracy does not invalidatethe principle of decoupling conflicting interests in environmentalmanagement, it does indicate that the range of applicationof the design principle needs to be expanded. In thefollowing, I will outline a few reform ideas that are much moredynamic in nature than the administrative changes. To do this, Iintroduce two additional perspectives from which competing socialinterests in environmental management can be identified.Both perspectives overlap with the two earlier ones and are anunderlying theme not just of the two cases but of the past threedecades of environmental issues.The spatial-temporal perspective highlights the fact that themost pressing environmental problems today typically cover geographicallylarge areas and require attention spans over severaldecades, even centuries. The problem is, as both environmentaland biodiversity management analysts point out, that policiesmeeting sustainability criteria for local areas and shorter timescales often fail to do so for larger areas and longer time scales(1, 35, 36). So, too, in the California case. The in-valley agri-cultural management solutions that the latest governmental paneldeveloped were, by the panel's own admission (25), sustainableon a regional level over decades but not for the entire valley overcenturies (Table 2). Another significant perspective is the politicaleconomy of environmental management, which refers hereto the pull between environmental and monetary imperatives.Both laymen and politicians have been found to show compas-sion for the environment and even propose sensible plans to abateenvironmental pollution, but when actual decisions are taken,short-term economic factors override environmental concerns(37, 38). In the Finnish and Californian cases, this tension constitutedthe officials' cognitive dissonance.The institutional design challenge arising from the last twoperspectives is: How to guarantee autonomy to global, long-termand primarily environmental imperatives in the face of localized,short-term and primarily economic demands? A workable startingposition is to respect one of the fundamental findings of institutionaleconomics and economic history, namely, that theimmediate instruments of institutional change are individualsmaximizing at margins with the most profitable short-term alternatives(8). To make institutional change work for sustainabledevelopment, individuals should be able to maximize utilitybut do so within an institutional context that directs the sumand sequence of individual short-term decisions toward the longtermconservation of the environmental resource base. Governmenttaxes are an institutional instrument allowing individualutility maximization within the wider social context. But to servesustainable development goals, the emphasis of taxation shouldmove from labor to activities that burden the environment (39).And to expand the discussion from the narrow maximization ofindividual utility to the more comprehensive concept ofsatisficing individual preferences (40), we should not forget re-ligion as a model for developing environmentally sustainable institutions.Most religions offer immediate gratification, such asthat brought to the Catholic during absolution or the Hindu frombathing in the Ganga river, but have also survived over millenniaand convinced large numbers of believers of eternal life.Bringing up taxes and religion when discussing sustainable developmentis an attempt neither to secularize nor consecrate theconcern for the environment, but rather to remind of the existenceof institutions that successfully blend together an individual'sshort- and long-term interests, not only in terms of measurableutility but unquantifiable good as well.Another way of tackling the issue of technocracy versus de-mocracy is to develop political discourse procedures in environmentalmanagement. The institution of such procedures is likelyto result in surprise configurations of policy agreement. A casein point is the 1995 Conference of the Parties to the UN ClimateConvention in Berlin, in which the insurance sector and116 ? Royal Swedish Academy of Sciences 1998 Ambio Vol. 27 No. 2, March 1998http://www.ambio.kva.se