Small Wins Redefining the Scale of Social Problems Karl E. Weick Cornell University ABSTRACT: The massive scale on which social prob- lems are conceived precludes innovative action because bounded rationality is exceeded and dysfunctional levels of arousal are induced. Reformulation of social issues as mere problems allows for a strategy of small wins wherein a series of concrete, complete outcomes of moderate importance build a pattern that attracts allies and deters opponents. The strategy of small wins incorporates sound psychology and is sensitive to the pragmatics of policymaking. There is widespread agreement that social science research has done relatively little to solve social prob- lems (Berger, 1976 Cook, 1979 Kohn, 1976). Com- mon to these assessments isthe assumption that social science is best suited to generate solutions, when in fact it may be better equipped to address how prob- lems get denned in the first place. A shift of attention away from outcomes toward inputs is not trivial, because the content of appropriate solutions is often implied by the definition of what needs to be solved. To focus on the process of problem definition is to incorporate a more substantial portion of psychology, specifically, its understanding of pro- cesses of appraisal, social construction of reality, problem finding, and definition of the situation. Whether social problems are perceived as phe- nomena that have a serious negative impact on sizable segments of society (Kohn, 1976, p. 94), as substantial discrepancies between widely shared social standards and actual conditions of life (Merton, 1971), or as assertions of grievances or claims with respect to al- leged conditions (Spector & Kitsuse, 1977, p. 75), there is agreement that they are big problems. And that's the problem. The massive scale on which social problems are conceived often precludes innovative action because the limits of bounded rationality are exceeded and arousal is raised to dysfunctionally high levels.People often define social problems in ways that overwhelm their ability to do anything about them. To understand this phenomenon, consider the following descriptions of the problems of hunger, crime, heart disease, traffic congestion, and pollution. To reduce domestic hunger we grow more food, which requires greater use of energy for farm equip- ment, fertilizers, and transportation, adding to the price of energy, which raises the cost of food, putting it out of the price range of the needy. To solve the problem of soaring crime rates, cities expand the enforcement establishment, which draws funds awayfrom other services such as schools, welfare, and job training, which leads to more poverty, addiction, prostitution, and more crime. To ward off coronary heart disease, people who live in cities spend more time jogging and cycling, which exposes their lungs to more air pollution than normal, increasing the risk of coronary illness. To ease traffic congestion, multilane highways are built, which draws people away from mass transit so that the new road soon becomes as overcrowded as the old road. To reduce energy use and pollution, cities invest in mass transit, which raises municipal debt, leading to a reduction in frequency and quality of service and an increase in fares, which reduces ridership, which further raises the municipal debt (Sale, 1980). When social problems are described this way, efforts to convey their gravity disable the very re- sources of thought and action necessary to change them. When the magnitude of problems is scaled upward in the interest of mobilizing action, the quality of thought and action declines, because processes such as frustration, arousal, and helplessness are activated. Ironically, people often can't solve problems un- less they think they aren't problems. If heightened arousal interferes with diagnosis and action, then at- tacking a less arousing "mere problem" should allow attention to be broader and action to be more com- plex. Responses that are more complex, more recently learned, and more responsive to more stimuli in changing situations usually have a better chance of producing a lasting change in dynamic problems. To recast larger problems into smaller, less arousing problems, people can identify a series of controllable opportunities of modest size that produce visible results and that can be gathered into synoptic solutions. This strategy of small wins addresses social problems by working directly on their construction and indirectly on their resolution. Problems are con- structed to stabilize arousal at moderate intensities where its contribution to performance of complex tasks is most beneficial. 40 January 1984 ��� American Psychologist Copyright 1984 by the American Psychological Association, Inc. Vol. 39, No. I, 40-49

Arousal and Social Problems The following analysis of small wins assumes that arousal varies among people concerned with social problems, but tends to be relatively high, which affects the quality of performance directed at these problems. Arousal is treated as a generic concept under which is assembled a variety of findings that cohere because of their mutual relevance to the Yerkes-Dodson Law (Broadhurst, 1959). Although arousal mechanisms are neither simple nor unidimensional, they do seem to be localized in at least two physiological sites (re- ticular formation, limbic system), are visible under conditions of sensory deprivation, produce differences in the quality of learning and performance, and have observable physiological effects. The specific effects of arousal on performance associated with the Yerkes-Dodson Law are that (a) there is an inverted-U relationship between arousal and the efficiency of performance with increasing lev- els of arousal, first improving and then impairing performance and (b) the optimal level of arousal for performance varies inversely with task difficulty. Even though these coarse propositions have been amended, tuned more finely, and differentiated, they remain basic principles in which an analysis of social problem solving can be anchored. Key assertions for the present analysis culled from previous investigations of arousal and perfor- mance include the following: 1. Arousal coincides with variation in degrees of activation and varies along at least two dimensions, energy-sleep and tension-placidity (Eysenck, 1982 Thayer, 1978a, 1978b). 2. As arousal increases, attention to cues be- comes more selective and this editing is especially detrimental to performance of difficult tasks (Eas- terbrook, 1959, although this generalization has re- ceived mixed support. See Baddeley, 1972 Pearson & Lane, 1983 Weltman, Smith, & Egstrom, 1971, for representative work). 3. At relatively high levels of arousal, coping responses become more primitive in at least three ways (Staw, Sandelands, & Dutton, 1981): (a) people who try to cope with problems often revert to more dominant, first learned actions (b) patterns of re- sponding that have been learned recently are the first Tom Peters's (1977) original description of small winswas a crucial point of departure for this formulation. Subsequent discussions with Peters, as well as with Linda Pike, Richard Thaler, Joseph McGrath, Sharon McCarthy, David Anderson, Marianne LaFrance, and students and faculty of the Psychology Department at Rice University contributed to my understanding of this phenomenon and I am grateful to all of them for their help. Requests for reprints should be sent to Karl E. Weick, Cornell University, Graduate School of Administration, Malott Hall, Ithaca, New York 14853. ones to disappear, which means that those responses that are most finely tuned to the current environment are the first ones to go and (c) people treat novel stimuli as if they are more similar to older stimuli than in fact they are, so that clues indicating change are missed. To invert this list, highly aroused people find it difficult to learn a novel response, to brainstorm, to concentrate, to resist old categories, to perform com- plex responses, to delegate, and to resist information that supports positions they have taken (Holsti, 1978). When these findings are focused on problem solving, they suggest that to call a problem serious is to raise arousal, which is appropriate if people know what to do and have a well-developed response to deal with the problem. This is analogous to the sit- uation of a simple task, the performance of which improves over a considerable range of activation be- cause selective attention does not delete the few cues that are essential for performance. High arousal can improve performance if it occurs after a person has decided what to do and after she or he has overlearned how to do it. To call a problem minor rather than serious is to lower arousal, which is also appropriate if people don't know what to do or are unable to do it. If we assume that most people overlook the fine-grain detail of problems, think only in terms offeree as a response (Nettler, 1980), and overlook minor leverage points from which the problem might be attacked, then it is clear they have neither the diagnoses nor the re- sponses to cope. This means that people need lower arousal to keep diagnostic interference at a minimum and to allow for the practice of relatively complex skills. To keep problem-related arousal at modest in- tensities, people need to work for small wins. Sometimes problem solving suffers from too little arousal. When people think too much or feel too powerless, issues become depersonalized. This lowers arousal, leading to inactivity or apathetic perfor- mance. The prospect of a small win has an immediacy, tangibility, and controllability that could reverse these effects. Alinsky (1972, pp. 114-115) persuaded a de- moralized neighborhood group to picket for rein- statement of Infant Medical Care, which he knew would be granted if they merely asked. Organizing for the protest, making the demand, and then receiving what they asked for energized people who had ba- sically given up. Examples of Small Wins Small wins have been designed and implemented in a variety of settings. For example, the Pittsburgh Steelers in the National Football League have won 88 games and lost 27 under their coach Chuck Noll (as of February 4, 1980). Those statistics become more interesting if they are partitioned on the basis of January 1984 �� American Psychologist 41