Division of Research January 1990 School of Business Administration KEEPING AN EYE ON THE MIRROR: THE ROLE OF IMAGE AND IDENTIFY IN ORGANIZATICNAL ADAPTICN THE CASE OF THE PIRT AUILTHITY OF NEW YCRK AND AND NEW JERSEY AND THE ISSUE OF HLMELESSNESS Working Paper #626 Jane E. Dutton The University of Michigan Janet M. Dukerich University of Texas at Austin FOR DISCUSSION PURPOSES ONLY None of this material is to be quoted or reproduced without the expressed permission of the Division of Research Copyright 1989 University of Michigan School of Business Administration Ann Arbor, Michigan 48109-1234

Keeping an Eye on the Mirror: The Role of Image and Identity in Organizational Adaptation This is an article about how individuals make sense of their organization's response to a non-traditional and emotional strategic issue. The research also concerns microprocesses involved in organizational adaptation. The case describes how the Port Authority of New York and New Jersey has dealt and is dealing with the issue of homelessness. We use this description to build a different view of how organizations adapt to environments which highlights the central role that an organization's identity and image play in this process. The basic claim is that an organization's image and identity are imortant reference points and motivating forces experienced by individuals in org ations. Changes in an organization's image over time, and their relationship to the organization's identity explain patterns of responses to issues. We contrast this new view with ideas of organizational adaptation, enactment and issue interpretation processes, and suggest new directions for theory and research. Daphne Futter was an important research assistant on this project. We thank Susan Ashford, Daj Bjorkegren, Arthur Brief, Tom D'Aunno, Martha Feldman, Ari Ginsberg, Stuart Hart, Susan Jackson, Michael Moch, Doug Orton, Lance Sandelands, Zur Shapira, David Skidd, Pringle Smith, Karl Weick and Janet Weiss for suggestions and comments. We also thank our respondents at the PA who gave so generously of their time so that we could understand their struggle with this issue.

3 ".... the homelessness problem is perhaps a blight on that professionalism that we like to display, and that we are so proud of, and I think this is of great concern there. Again, there may be some conflicting issues on spending money to help solve the problem, but I think that's a value. We build beautiful facilities, we take pride in that, and the homelessness issue is something that obviously affects the perceptions of us". (Facility staff member, The Port Authority, 1989) Models of how environments and organizations relate over time typically assign causal primacy to either environmental forces (e.g., institutional theory, resource dependence or population ecology), organizational forces (e.g., strategic choice) or some combination of the two for explaining how organizations and their environments co-evolve over time (e.g., Singh, House & Tucker, 1986; Hambrick & Finkelstein, 1987; Hannan & Freeman, 1984; Tushman & Romanelli, 1985). None of these theories treat in depth the processes by which environments and organizations are related over time. While the language of these theories implies a process for how environments and organizations are connected (e.g., strategies are chosen in response to environmental changes or environmental selection mechanisms favor one type of structural form more than others), views of the process through which these relationships are accomplished are currently limited (Sandelands & Drazin, 1989). In this paper we develop a framework for conceptualizing the process through which organizations adapt to their environments. The conceptual and empirical perspectives are ones that take seriously the assertion that organizations respond to their environments by interpreting and acting on issues (e.g., Daft & Weick, 1984; Dutton & Duncan, 1987). It is the pattern of issue actions, over time, that creates patterns of organizational action. The claims are built from a case study of how the Port Authority of New York and New Jersey has defined and responded to the issue of the rising numbers of homeless persons present in their facilities. The case is used to generate a framework for understanding how organizations and their environments interrelate over time. It employs the idea from research on organizational culture and strategy that organizations have competencies (Selznick, 1957) or identities that influence how individuals interpret events and developments, as well as how they behave towards them. The idea that organizational identity affects interpretations and actions has some support from other studies of organizational adaptation (Meyer, 1982; Miles & Cameron, 1982). The study

4 also builds on ideas from impression management in social psychology (e.g., Tedeschi, 1981) and from theories of reputation in corporate strategy (Weigelt & Camerer, 1988), suggesting that individuals are motivated to influence how others see and evaluate the organization. The paper transverses the levels of macro and micro organizational theory to explain how the Port Authority has dealt with the issue of homelessness. Issues as a Starting Point The perspective taken in this paper begins with the idea that some organizational actions are tied to sets of concerns that we call issues. Issues are events, developments or trends that are collectively recognized by organizational members as having some consequence to the organization. Thus, issues can arise from changes taking place internal to the organization (e.g., employees threaten to stage a strike, a new technology transforms an organization's product or service) or changes originating externally (e.g., a demographic trend, a regulatory act, a supply shortage). The definition of an issue by a collectivity is a social construction (Hilgartner and Bosk, 1988). Issue definitions can be contested. Issue definitions often emerge and evolve over time (El Sawy & Pauchant, 1988; Weiss, 1990; Dutton, 1988). Which issues gain attention and how they are interpreted are important concerns, as issues represent arenas that galvanize interest and focus attention in organizations because of the consequences associated with action or inaction. In some cases issues activate decisions, in other cases, issues incite neglect or intentional inaction (Bachrach & Baratz, 1973). A focus on issues as a starting point for sense-making and action in organizations charts a different course for seeing patterns of organizational action than a traditional decision making view. Researchers who look at decisions as creators of patterns in organizational actions (e.g., Mintzberg, et al.; Nutt, 1984) use an endpoint in a process (e.g., a choice or non-choice) as the defining referent, describing who and what were involved in producing a certain pattern of action. Typically, research is done by defining a choice or decision and then tracing backwards from that point to find relevant interpretations and actions for that decision. In contrast, a focus

5 on issues begins with the "issue" or some collective construction that some data, fact or event is of concern for the organization, and then traces forward from this recognition process to find relevant actions and interpretations. For organizations, some types of issues are routine and expected. Such issues can be easily classified by organizational members. The issues fit into existing categories that an organization has in use, and once classified, elicit a type of automatic, well-learned response (Starbuck, 1983; Starbuck & Milliken, 1987; Weick, 1988). Other types of issues are not as easily interpreted or processed, however. Issues may be problematic because they are nontraditional, e.g., they have not been encountered in the past, and thus do not easily fit into wellused categorization schemes. In the context of issues which are non-routine, and for which no standardized responses exists, the process through which an organization adapts to its environment is most transparent. Alternatively, issues may be problematic because of the feelings that they evoke, and the inappropriateness of these emotions given the feeling rules present in the organization (Sutton & Rafaeli, 1987). In models of issue diagnosis and organizational adaptation we know very little about how the level of emotion evoked by an issue affects individual and collective processes. Issues that are "hot", i.e., issues that evoke strong emotions, represent different types of stimuli and activate different responses from individuals and organizations, than cooler, less affectivelycharged issues. It was this logic —a desire to study how individuals and organizations make sense of and act upon a non-traditional and emotional strategic issue, and a wish to study issue processing — that drew us to a case study of how the Port Authority of New York and New Jersey is dealing with the issue of homelessness. The study was designed to generate new theory on how individual interpretations and organizational action on an issue were related over time. The case analyses revealed that an organization's identity and image are critical constructs for understanding the relationship between actions and interpretations of the issue over time. Both constructs emerged clearly from the theme analysis of the data (described in the

6 methodology section below). The organization's identity (what organizational members believe is the organization's central, enduring and distinctive character (Albert & Whetten, 1985)) filters and molds an organization's interpretation and action on an issue. At the same time, organizational members monitor and evaluate organizational actions on issues because these actions are used by others outside of the organization to make character judgments about the organization, and by implication, the organizations' members. Organizational members use the organization's image (the way organizational members believe others see the organization) to gauge how outside members are judging organizational members. As a result, individuals are motivated to take actions on issues that will create a favorable organizational image (and hence, minimize their own character damage1). At the same time, the organization's identity limits and directs issue interpretations and actions. These actions, in turn, gradually modify the organization's future identity. These claims are illustrated by charting and describing the pattern of actions, interpretations and events that capture how the Port Authority of New York and New Jersey has reacted to the issue of homelessness. METHODS A case study methodology was well-suited to our desire to generate and build theory for an area where little data or theory existed (Yin, 1984), where we could study a process as it unfolded over time, and where we could use controlled opportunism to respond flexibly to new discoveries in collecting new data (Eisenhardt, 1989). We selected the case of how the Port Authority of New York and New Jersey (PA) has responded to the issue of homelessness because the issue had far-reaching social implications and was tremendously visible to both organizational members and to outside constituencies. In this sense, the case matches the criteria of an extreme case, where the process of theoretical O1f course, not all individuals will be so motivated. For those organizational members who have little or no commitment to the organization, how others see the organization may be inconsequential. Thus, we are concerned with those individuals who have at least some stake in a positive organizational image.

7 interest is more transparent than in less extreme cases (Eisenhardt, 1989). A few details on the major data sources, the respondent sample, the interview process, and data analysis follow. Data Sources The story of how the PA and the issue of homelessness are related was built from five sources: 1) open-ended interviews with 25 members of the Port Authority during the period from September, 1988 to May, 1989; 2) collection and analysis of all reports, memos and speeches prepared within the PA on the homelessness issue during the time period from 1982 until March, 1989; 3) analysis of articles from regional newspapers and magazines (published from March, 1986 through November, 1988) that mentioned both the Port Authority and the homelessness issue; 4) analysis of text from conversations with the head of the Homeless Project Team (a temporary task force specifically set up to examine the corporate response to the issue of homelessness) whom the researchers encountered regularly (at least once every 2 weeks), and 5) notes from an all-day training session with PA facility staff, sponsored by the Homeless Project Team (HPT) during May, 1989. Sample of Informants. We interviewed individuals from four different groups of stakeholders in the homelessness issue. The distribution of informants in the four groups appears in Table 1. We interviewed the Executive Director and top level managers who had involvement with the issue (4), all members of the Homeless Project Team (6), line managers who currently or historically had responsibility for the facilities that were actively involved in trying to deal with the issue (6), staff personnel from the public affairs, corporate planning and the budget offices who had current or historical responsibility for developing and analyzing ideas for a PA response to the issue (5). The final group included police personnel and customer service managers who dealt "hands on" with the homeless in various PA locations (4). Insert Table 1 about here Interview Questions. The interview guide targeted data on five different clusters of variables (described in more detail in Table 1). The average interview lasted two hours with one

8 researcher asking questions, while the other took notes. More than half of the interviews were tape recorded and transcribed verbatim. Data Analysis. "Analyzing data is at the heart of building theory from case studies" (Eisenhardt, 1989, p. 11). For the analysis used in this paper, two analyses were critical: 1) The construction of the issue history (interpretations, actions and events from 1982-1989) and 2) use of the theme analysis to explain the pattern of interpretations and across over time. Both analyses emerged from an identifiable set of steps. Step 1. Devising and Coding Using a Contact Summary Form. Following the procedures recommended by Miles and Huberman (1984), we used a contact summary form for recording the "main themes, issues, problems and questions" in each interview, which was originated by one researcher and check coded by the other. Themes are defined as "recurrent topics of discussion and/or actions on the part of the actors being studied" (Bjorkegren, 1989). Like a recurring melody in music, themes capture central ideas or relationships displayed in the interview content (Bjorkegren, 1989). Step 2. Developing a Complete Theme List. The contact summary sheets for the 25 interviews generated 84 themes which were collapsed into seven major groupings based on a very general classification of theme substance. For example, "Organizational Reactions to Homelessness" and the "Identity of the PA" were broad theme categories. The first broad category included 14 different themes, each addressing unique ways that the PA was responding to the homelessness issue (e.g., denial of being in the social service business or negative reaction to other agencies' failures to take responsibility for the issue). Themes were used for two distinct purposes: 1) to isolate commonalities in how Port Authority members interpreted the homelessness issue; and 2) to suggest an explanation for the issue's history based on two dominant theme categories —the importance of organizational image and identity. Next, each theme was assigned a separate sheet on a coding fonn in preparation for Step 3.

9 Step 3. Coding the Interview Data onto the Themes. Each interview was coded, line by line, onto the theme list in order to document and evaluate the degree and breadth of support for a particular theme across the range of issue informants. Step 4. Constructing an Issue History. Questions on the meaning of the issue and on milestones in the processing of the issue were used to construct a history (1982-1989) of how the PA has interpreted and responded to the issue. Information from memos, speeches and minutes served as important supplements to interview data in constructing the issue history. Members of the HPT were consulted to validate the issue history once it was completed. The Site The PA was established by a compact between the states of New Jersey and New York as a government development corporation, established to administer the economic and social development of the Port District (an area encompassing roughly a 25-mile radius of the Statue of Liberty). While most public authorities in the United States were established to develop and operate a single public improvement project (e.g., a bridge, tunnel, airport), the PA was the first multi-purpose public authority (Caro, 1974). Today it owns and operates thirty-five facilities including the World Trade Center, the Bus Terminal, Journal Square Path Center, Kennedy, LaGuardia and Newark airports, Path Train Service, tunnels, bridges and marine facilities. The mission of the PA remains very broad —to protect the economic vitality of the NY-NJ Port District. The PA defines itself as in the business of transportation. The PA is the largest public authority in the United States, employing 10,000 people, total assets of approximately $5 billion, and with an annual budget of $1 billion. The PA is self supporting through the issuance of bonds, user fees and leasing revenues. It is run by an Executive Director and a Board of Commissioners who are selected by the Governors of the two states. The Issue The presence of "bums, winos and bag ladies" has always been part of the scene for transportation facilities. During the last several years the number of homeless persons living

10 and spending time at transportation facilities has dramatically increased. For the Port Authority, the rising numbers of homeless persons present at their facilities caused increasing problems with the delivery of quality transportation service. One of our respondents described the change this way: "Well, a lot of it had to do with the change in the type of people... And the Bus Terminal always had its share of down-and-out people, but you were able to move them along and get some kind of arrangement with them. But as the numbers increased, you couldn't do that. And the nature of the people began to change, and they began to get younger, and in some respects the people [the PA's patrons] became more afraid of them because they were rowdier, they were more inposing...." The way that the issue was interpreted changed over its lifetime in the organization. The relationship between interpretation changes and organizational action on the issue is the focus of the findings that we will present below. Over time, there were some pervasive themes in the way that the issue was experienced and interpreted by our respondents that distinguish this issue from other issues confronting the Port Authority. In particular, three attributes of the issue were mentioned by more than 10 respondents: 1) broad scope; 2) links to other negative issues, and 3) perceptions of a lack of control over the issue. Issue Breadth. For PA members, the homelessness issue was distinct in its broad scope and linkages to a number of regional issues affecting the PA. In particular, informants saw the issue as related to the shrinking in housing availability, particularly in the supply of SROs (single-resident occupancy housing), as related to trends in mental health policies (e.g., deinstitutionalization) and tied to education and job skill training. Informants expressed linkages between the homelessness issue and other societal issues by describing the possibility of classifying the issue in multiple ways and in terms of ties to other issues: "It's a health problem, it's a mental health problem, it's a problem of [de]institutionalization. It's a problem of reduced federal funding for housing. It's a problem of burgeoning development around New York City, and in which a lot of people are displaced. It has to do with education, it has to do with family structure, it has to do with housing, it has to do with who one identifies with role models".

11 Links to Other Negative Issues. PA members saw the homelessness issue as unique in terms of its strong ties to other negative issues facing the PA, particularly during recent years (1987-1989), the homelessness issue became tied to issues of drugs and crime. The links between homelessness and crime were reflected in PA members' struggle to separate "the desperates from the desperados" or the "preyers from the preyed upon". By developing classification schemes that distinguished the "true" from "imposter" homeless, PA members felt that they were better able to deliver services to the "truly" needy and to manage correctly the impression of the PA by its outside constituents. However, PA members felt this task was becoming increasingly more difficult because of the efforts by "not-so-desirables" to hide under the "homelessness umbrella". "It's become a way for some people who are not such good human beings, and perhaps, not such good society members, certainly, to hide in this general umbrella. So a crack dealer, not a user, but a dealer, someone who is perpetuating this plague on society, may be homeless and say he is homeless, but to be homeless, if he's arrested, he's almost guaranteed to serve no time. They don't want to deal with him. So there are layers of homeless, and when you put people into this category which takes more and more people, well, who's destroying the Bus Terminal bathrooms? Well, some of them are classic homeless people, for whom the bathrooms weren't made, but use them to wash up and all that stuff. They're taking their toll. But the crack dealer who flushes his paraphernalia down the toilet because a cop is approaching is taking a greater toll. It's very hard to separate out, and I think what's happening to society is that people are turning off to the whole notion of homelessness, not giving and not caring". Feelings of Lack of Control Over the Issue. The homelessness issue was frustrating to PA members in part because individuals felt the organization had little control over the issue and its solutions. Some individuals believed the lack of control came from limits on the organization's ability to access the means for controlling the issue, i.e., what one manager called "going beyond our internal ability to solve the problem". Others felt the lack of control stemmed from the "human ingredient" in the issue. Clients' and customers' constant demand to the PA to "do something about this issue" aggravated the feelings of lack of issue control. Individuals' pride in being a member of "a fixer" organization magnified frustration levels. One facilities manager described his frustration with the issue in the following way:

12 "I think that with all of the building and fixing and all of those good, concrete, reassuring things that we did and still do, and the feeling, the good feeling that we got from being in control, I think this has been undermined in a way by this homeless problem. I think that it [the homeless issue] said to us, 'Look, here is something that you really can't control, and you can't fix it, and you can't caulk it, you can't waterproof it, you can't dig it, you can't make it go away"'. Summary The common threads in how individuals saw the issue of homelessness for the PA —its ties to other broad issues, its close affiliation with negative issues of crime and drugs, and the sense of limited ability to control or fix the issue, illustrate the common themes in how the issue was interpreted by PA members. The ambiguous boundaries for the issue, its linkages to other issues, and the organization's sense of inability to control the issue contributed to the fluidity of the issue's interpretation over time. The next section describes the major actions and dominant interpretations for the homelessness issue during the period 1982-1989. INTERPRETATIONS AND ACTIONS ON THE HOMELESSNESS ISSUE The Port Authority's struggle with the homelessness issue can be mapped onto a set of five phases. Each phase is distinctive in terms of how the issue was interpreted and in the set of actions that the PA took to deal with the issue. While the five phases are represented as though there were clear, identifiable signs that an old phase had ended, and a new phase had begun, the distinctions between phases are more qualitative than quantitative, and more evolutionary than revolutionary. The path of understanding and responding to this issue can be thought of as an evolving series of interpretations, emotions and actions on the issue that were linked together by their commonality in time, forming a type of issue history. The issue history, in turn, provides important insights into the organizational processes at work in creating patterns of action. Three components of each issue phase are described below: 1) key events, 2) major issue interpretations, and 3) issue actions. The key events during each phase describe the major developments and changes (e.g., executive succession or availability of some new piece of data) that our respondents identified as significant during this phase of the issue's evolution in the PA. The discussion of key events provides important contextual information for comprehending how the issue was interpreted at that point in time, and how and why certain actions were taken.

13 While in certain cases the events appear to "cause" a certain action or interpretation, for now such causal inferences should be suspended. Our purpose is to provide a relatively complete description of how interpretations and actions co-evolved in the context of a series of unfolding events, and against the backdrop of a particular organization. It is the themes that were identified in the interviews that provide a basis for interpreting this pattern of events, interpretations and actions on the issue. Figure 1 presents a synopsis of the five issue phases portrayed in terms of a timeline. Each phase is described in more detail below. Insert Figure 1 about here Phase 1: Homelessness is a Police Issue (1982-1984) Homeless persons have always been part of the landscape for transportation services. The features that are important for the delivery of effective service to transportation agency clients also attract homeless persons. The facilities are warm in the winter, cool in the summer. They are clean, provide access to toilets and running water, and guarantee some degree of personal safety through the constant presence of police. Thus, for most transportation agencies, and the police who patrol them, having to deal with a certain number of "bums, winos, and bag ladies" is normal, and part of doing business in transportation. Key events. In 1982, several factors converged to make the issue of homelessness a more prominent one for the PA, particularly at the Bus Terminal. First, there was a notable recognition by PA members of a rise in the number of homeless persons present. Second, a $226 million dollar renovation that had just been completed at the Bus Terminal made the rise in numbers of homeless persons more visible. The renovation opened up new space that could be used by both passengers and homeless persons (an increase in square footage of about 40%). At the same time, in New York City, a large number of single room occupancy hotel rooms were closed down in 1982. As one respondent told us: "As the Manhattan real estate market picked up, these hotels were closed, and we had an increase in the number of homeless people, without many skills,

14 without abilities, and without much money, all ending up out on the streets. A fair number of them ended up in the Port Authority Bus Terminal". The renovation also accentuated the homeless problem for the PA by creating a strong contrast between the beautification of the Bus Terminal (accomplished by adding space and expensive works of art) and the stain on that beauty inflicted by the presence of homeless persons who "smelled, looked dirty", and to PA patrons and workers, represented a mark on the attempt to clean, spruce up, and beautify the Bus Terminal. For an organization that prided itself in being "the building of beautiful structures" homeless persons signified a stain on that identity. Major Issue Interpretation. During 1982, the homelessness issue was defined as a police or security issue: the presence of homeless persons was problematic for PA customers traveling through its facilities, and something had to be done. As one respondent described it: "The issue was 'How do we keep these people out of our facility'? Plain and simple. Because they were interfering with our patrons in the sense that they felt that they were not safe because of their presence" The police were (and continue to be) a major source of contact with the rising number of homeless at the Bus Terminal; they were also the ones who carried out the action on this issue. The police were the ones whom the customers confronted when they wanted someone from the PA to "DO SOMETHING ABOUT THIS PROBLEM!". The PA employs a full-time police staff of 1,500 officers (the 26th largest police department in the United States), 130 of whom were assigned to the PA Bus Terminal. At this time, it was the local police force at the Bus Terminal, in conjunction with facility level management, who developed a means for dealing with the issue; there was no coordinated, corporate-level response to the issue. Major actions. The existence of the loitering law in New York gave PA police the option of encouraging homeless persons to leave the Bus Terminal. In 1982, facility staff at the Bus Terminal took two additional actions to deal with the issue. First, they hired a consultant to train police on how to move persons out of the facility, in a manner that "acknowledged the difficult nature of the problem". The second action was the establishment of a relationship with the Human Resources Administration (HRA) of NYC (ignited by a chance meeting of the PA Director and HRA head

15 when running in Prospect Park) and the Manhattan Bowery Corporation to develop an outreach program to "give the police some place to send these people". At that time, the PA police assisted staff from the social service agency in getting homeless persons in the right locations so that they could be transported to City shelters. Summary. Early actions on the homelessness issue were facility based, limited in scope, and focused primarily on the Bus Terminal. The issue was primarily framed as a police and security issue, which, given the existence of the loitering law, helped contain the problem. Actions to engage the assistance of New York City's social service support were also part of the facility-based solution at this time. Phase 2: Homelessness is a corporate issue, but we are not in this business. (1985-1986) Demarcations between phases in relationship between the homelessness issue and the Port Authority are not clean and clear-cut. However, in the 1985-1986 period, PA members changed the way that they talked about the issue. This change could be attributed to a number of different events and the recognition that the problem could no longer be isolated and contained in the Bus Terminal. Key events. During the 1985-1986 period, an awareness grew that the homelessness issue was no longer just a Bus Terminal issue, where homeless persons were expected, and routines had been developed to deal with them, but that the homeless were present in several PA facilities. Key to making the issue visible at the corporate level was the presence of homeless persons at the World Trade Center and the airports. Here the homeless were unexpected and causing problems with central components of the PA's identity: "It wasn't until homeless people started to show up at the World Trade Center, and tenants, very affluent tenants in the World Trade Center, and the image of the WTC as being a place where homeless people were, began to raise its head, that people started to say, "wait, geez, this is a problem...." It started to show up finally in corporate documents as an issue. It never did before, because everybody knows the Bus Terminal is an aberration, but when it started to show up at the World Trade Center, and then ultimately, one or two people at the International Arrivals building at Kennedy Airport and at LaGuardia Airport, then it began to touch upon the heart and soul of the organization".

16 The departure of Peter Goldmark as the Executive Director for the PA, and the appointment and arrival of Stephen Berger was another key event during this time period. This leadership change was significant on several counts. First, persons closest to the issue argued that the momentum to recognize and deal with the issue of homelessness at the Bus Terminal had been built and approved through Peter Goldmark. The momentum dissolved with his departure, and advocates for the issue felt that they had to "start from scratch". Second, the new executive director had a different vision for the PA in terms of its "returning to its basic businesses". As described by others, there was a clear sense that he wanted to show others that "we could run this like a business and show the world what a public agency could do at running a business". One implication from this visionary change was an emphasis on using "business practices" and "business justifications". In 1986, the issue appeared for the first time in business plans for several line departments. Simultaneously, the Public Affairs Department became increasingly concerned about the issue as the rate and intensity of customer complaints were increasing. Stephen Berger openly expressed a strong personal aversion to straying away from the main businesses of the PA, and a correspondingly strong aversion to "getting into the social service business". Major interpretations. In 1985-1986 the interpretation of the issued shifted to a recognition that the issue was corporate wide, and one which could not be contained as a police issue at the Bus Terminal. At this time concern was focused on how to minimize the negative impact of the issue by removing and restricting the problem as it presented itself in the various affected facilities. Major actions. During this phase, three sets of actions were taken. First, a group of staff persons were asked to collect data, analyze it, and come up with recommendations for a corporate policy on the issue. For the police and staff at the facility level, this action was viewed as a sign that the issue was receiving corporate-level attention. As one upper level manager stated, the results from this analysis represented "the first time that it was explicitly recognized as a problem and put in writing". Second, actions at the facility level were intensified: 1) more

17 extensive outreach services were obtained (with daytime as well as nighttime assistance), through a new contract with the Volunteers of America (VOA, a not-for-profit social service provider); and 2) areas of the Bus Terminal were closed from access and patron benches were removed from the waiting areas. The latter actions were designed to make the Bus Terminal an undesirable place to be by "making it as unattractive and uncomfortable to the homeless as possible". As one informant told us, "I think some of it was motivated by aesthetics, that you didn't have the people sitting around and maybe they would find some place else to go". Similar types of outreach services and actions to make the facilities unattractive to homeless were implemented at two other locations in which issue was visible (the World Trade Center and Journal Square Transportation Center). The third action was an attempt by facility staff at the Bus Terminal to manage patrons' understandings and reactions to homeless persons by issuing and posting a lengthy description of the types of homeless persons that patrons were observing at the Bus Terminal. This action was the first of many attempts to try to actively improve the image of the PA by using a well-learned recipe: "educating others or helping them get smart on the issue". "We had a thing at the Bus Terminal called a concourse, which is really a notice to the customers about any changes in operational aspects. And we decided to write a long concourse describing to them what we had begun to learn from people like our consultant about homelessness, in an attempt to gain their sympathies. We thought if people began to know what we knew, then they would feel differently about these people and realize they were not a threat, and that, in fact, they were not filth and vermin. So we produced this thing. And certainly, that kind of assuaged our conscience for about two days, and then it really began to snowball". Summary. During this second issue phase, the PA had not significantly changed how they interpreted or how they acted in response to the issue. In fact, this phase can best be characterized as involving "Doing the same, but doing it harder". While there was a shift in the recognition that the issue was more than local, the PA responded in a way that preserved the fragmented, local facility-based response, with the overarching goal of "get the homeless out of here". At the time, persons at the Bus Terminal began to try to manage others' understanding of the issue of homelessness, a tendency that was to become more prominent as they became more

18 involved with the issue, and as the image of the Bus Terminal (and the PA by its affiliation in the Bus Terminal) deteriorated. This phase also marked the beginning of some serious soul searching about what the role of the PA would be with respect to this issue. As one respondent put it: "And then we were saying to ourselves... Can we get them out of there? Should we get them out of here? What are we supposed to do with them? Whose responsibility is this?" It was this type of concern that helped to usher in the third issue phase. Phase 3: Homeless is a Business Problem and a Moral Issue. (1987) In 1987, several events contributed to the way that the issue was framed and to a change in the level and type of PA response to the issue. Key Events. In late 1986, several events shifted the PA's view of its responsibility for the homelessness issue. First, there was the repeal of the anti-loitering law in NY, which significantly restricted the NY-based PA facilities' options for "moving out the homeless". For the police, the repeal of the anti-loitering law "tied their hands" in terms of options for removing homeless persons from the PA facilities, resulting in a real "blow to police morale". As one informant told us: "It's not that we never arrested people for loitering. But the loitering law's existence allowed us, without as much hoopla, to ask people to move on or to leave". At the same time, the nature of the homeless persons changed abruptly, brought about primarily by an influx of crack (a derivative of cocaine which is easily obtained, relatively inexpensive, and very addicting). Thus the problem worsened through its tight linkages with drugs and crime, the fact that the hands of the police and facilities' staff were restrained by legal changes, and the police staff were and had been without a contract. Without a contract and without clear legal guidelines for how to deal with the homeless, the police were in a difficult situation. In their eyes, they were forced to pressure the PA to grant certain concessions by generating unfavorable press coverage about the PA. They hired a PR

19 agency "to float stories about the PA". The stories were intended to create a more favorable union environment. "They [the PR firm] generate publicity all the time and the publicity is aimed at embarrassing the Port Authority and creating this climate of fear and stuff around its facilities, to promote the police position, you know...that they need more cops and that sort of stuff'. In late 1987/early 1988, the bad press about the PA was at its peak, particularly during the winter months. During this time period 65% of the articles written about the PA (from our sample of articles in NY-NJ newspapers) were negative in tone. The PA received negative press for its attempts to control the homelessness issue through a tightening of regulations. A sample excerpt from a local paper follows: In its last board meeting before Christmas, the Port Authority of New York and New Jersey played Scrooge to Jersey City's poor by outlawing begging and sleeping at the Journal Square PATH Transportation Center. (The Jersey Journal, December 11, 1987). In conjunction with the negative press, in 1987 the number of homeless persons congregating at the PA facilities surpassed 1,000 persons on some nights. This number represented an important threshold that, in the minds of PA members, made the issue no longer deniable for the organization. Key interpretations. The most significant change in the way the issue was defined during this time period involved acceptance of some organizational responsibility for dealing with the issue, and a gradual awareness that the issue was more than a police issue. An excerpt from an important internal memo on the issue (January, 1987) illustrates this interpretive shift. "It is important to recognize that the agency is not in a position to solve the problems of the homeless.... The PA's homeless policy is to encourage individuals to leave our facilities and find more appropriate shelter and services, and to minimize their return.... We seek to do this in a humane manner, through the assistance of social service agencies..." The shift in the way that the issue was now being defined was subtle. There was still extensive denial of responsibility for solving the problem in any way other than to alleviate the burden on the facilities staff, but there was new concern for choosing moral or humane solutions for dealing

20 with the issue. In addition, there was a recognition that some of the social service actions that were in place were having a positive effect in diminishing some of the burden of facility staff. Key actions. The repeal of the anti-loitering law provided a major impetus to the development (technically it was an updating) of rules and regulations for each of the facilities. The rules and regulations first appeared at the Bus Terminal, but this procedure spread rapidly to the other PA locations. These regulations were viewed as important to the police and facility staff as it "gave us a mechanism to deal with certain types of personal conduct for anyone in our facilities". Nevertheless, the options of the PA police were constrained, leaving many of them feeling "as if you're pumping out the ocean" in terms of resolving the problems with homeless persons. Space restrictions and closing off parts of the building were acknowledged as ineffective in minimizing the visibility of the homeless. PA actions during this time period indicated a type of resignation to the fact that the problem could not be solved through outreach or restrictions alone, and that the organization needed to take a stand. "And then we kind of gave up, you know, we gave up some space.... They just sort of took over the waiting room. That was it. You know, we just didn't know what to do. You know when you get 15 degree temperatures at night, and there's absolutely no place for them to go. And so, we said, well, how are we in good conscience going to throw them out of this facility?.... And this was the first time that people really began to look at it and say, 'Wait a minute, you know this is a real moral issue.' And this was when we decided to make the commitment. And while Grand Central and every place else was throwing them out, we weren't". In 1987, top management reluctantly admitted the need to develop a more coordinated, corporate-level response to the issue. It was during late 1987 that the Executive Director decided to form a centralized project team (The Homeless Project Team —HPT) whose major responsibilities would include 1) developing a PA policy on homelessness; 2) relieving the burden from the facility staff; and 3) reducing the amount of top management time spent on the issue. The formation of this team was seen by many as a key signal ihat the PA was ready to do something about this issue.

21 Another key symbol of top level commitment to the issue was the granting of a prestigious one-year fellowship opportunity to a Public Affairs employee for the expressed purpose of studying how the transportation industry was addressing the issue of homelessness. As one person described the significance of funding a fellowship that focused on this type of issue: "It was a very risky thing for the PA to do, because it is not typical of the transportation kind of issue or business or economic development issue that this kind of a conservative organization would generally grant". Summary. In 1987, the level and type of attention being paid to the issue changed. Important symbolic actions signaled to internal and external constituencies that top management was now interested in the issue: the formation of the HPT and the granting of the Cullman Fellowship. Early in 1987, the "Batten down the hatches" response dominated, evidenced by the increased use of rules and regulations restrictions on access and closings of parts of the facilities. While there was some evidence that assistance from social service agencies and the use of rules and regulations were providing some relief, the problem in terms of numbers of homeless persons worsened. Several events transformed this early response into a more resigned acceptance that something different needed to be done, and it needed to be done in a way that did not violate moral standards embedded in the PA's "way of doing things". At this time, the image of the PA was severely damaged by the rise in the negative press coverage about the Port Authority. With their hands tied by the anti-loitering law change, police-based solutions proved unsuccessful. In addition, the negative image of the PA as an inhumane institution really bothered some of our informants, and reaffirmed the importance of taking a more "humane stance" on the issue. In the case of homelessness, where the hotness of the issue varies with the coldness of the weather, this meant not endangering anyone "by throwing them out into the cold temperatures"

22 Phase 4: Homelessness is an issue of regional image and no one else will deal with it. (1988) When tracing the unfolding of any interpretation —action process over time, it is difficult to isolate clear causes and effects, or exact points of punctuation. Rather, as mentioned earlier, a series of events seem to co-align —some planned, some unplanned —which seem to be followed by a qualitative shift in how people interpret the issue and actions on the issue. The 1988 period represents a period of significant action on the homelessness issue. Key events. Three events are important for understanding the unfolding of interpretations, emotions and actions on homelessness during this time period. First, there was the launching of a $5.8 billion capital plan for the PA, aimed at updating facilities and improving the image of services in the region to help make it internationally competitive. This campaign introduced resource constraints and created expectations for positive press coverage. As one informant described it: "We had embarked on this capital campaign at the airports and all of our facilities. We needed the resources to handle the program. It gave us the impetus....so we need to control other priorities as much as possible, particularly at the airports. From an organizational standpoint, we are focused on the major initiatives. We expected all of this positive press about the capital plan, and instead, all we have gotten is negative press about homelessness. It overshadows the positive". The other two events were reactions to PA actions on the issue taken during this phase. In order to do something "different" the PA decided to commit capital funds to build drop-in centers (i.e., centers designed to provide social services to the homeless) at two locations near their facilities. The two events related to this action were: 1) New York City (NYC) resisted taking over the running of one of the drop-in centers the PA had built; and 2) organized opposition to the opening of the second drop-in escalated. The implications of both events are discussed in the description of PA actions provided below. Key interpretations. The dominant interpretations of the homelessness issue (and the PA's relationship to it) for the first half of 1988 was publicized and memorialized in a speech given by the Executive Director of the PA in January to the Partnership for the Homeless (in New York City). The speech was seen by many informants as clear evidence that the PA was

23 going to "do something" about the issue. In the words of some of our informants, the speech was a coming out party of sorts —where commitment to the problem became publicly known. This speech contained several critical points for understanding the actions and future interpretations on this issue: First, there was continued denial that the PA was "in the social service business". Second, the speech described the homelessness problem as a regional responsibility, noting that the failure to solve it would have devastating consequences for the region. The speech contained a symbolic association of the homelessness issue with the fiscal crisis for New York City, which communicated the seriousness of the issue for the entire region. The speech illustrated that the interpretation of the issue has considerably broadened in scope, representing an attempt to involve others in the PA's efforts to deal with the issue. At this point in time, PA leadership and membership acknowledged that no one else would solve the issue, leaving the PA with no choice but to get involved in a significant way. "And so, once it became clear that we were really going to have to become more aggressive, I think at that point there was a kind of watershed which said, 'We are going to have to do some things which clearly stretch our mandate, which commit both dollars and cents beyond what is appropriate, and what is probably on some level defensible, because the agencies that have this responsibility are just not prepared to act"'. Informants were distinctly emotional when they described the realization that "the PA was forced to do this because no one else would". On the one hand they expressed anger, frustration and disappointment that other organizations had shirked their responsibilities by not solving the problem (14 of 25 respondents mentioned this type of reaction). Illustrative quotes display these emotions: "It's frustrating —wanting to make a difference and having to deal with people who have the mandate, but don't take the responsibility" "Just because other institutions have failed, does the responsibility go to the recipient of that failure?" "It's been dumped on us, and nobody's giving us any help that I can see". "There is absolutely no reason on earth, why we should be running a dropin center, anywhere, or at any time".

24 The negative emotions associated with the realization were revealed in informants' descriptions of how uncomfortable the Board of the PA felt with financial commitments to the homelessness issue. Top level managers expressed this feeling bluntly: "The board is very unhappy and I think rightly so. They feel that we're spending money which we are, which is money that is desperately needed for other things in terms of our mandate". Emotional reactions, however, involved more than unease and anger at the PA's new role. Some informants described hurt and frustration brought on by accusations being made about their personal character based on outsiders' judgments of the PA's action on this issue. Many of the PA members felt good about what the PA was doing with the homeless, but the image externally was that the PA was acting inhumanely. For individuals, this discrepancy was distressing and hurtful. The image of the PA symbolized certain attributes that were generalized to the individual who worked there. "You know, the guy that's running the Lincoln Tunnel doesn't have a full perception of how the Bus Terminal or the homeless impact of what he does from a day to day basis. But the minute he leaves and he goes to the cook-out in his neighborhood and he meets somebody and this person says, 'What do you do for a living?' 'Oh, I work for the Port Authority.' They say, 'how can you stand that Bus Terminal, what can you do?' That's the name. That's the symbol of the Port Authority. It's the standard bearer. And you know, so personally everybody that's involved in any aspect of working for the Port Authority is identified with that place and with that issue". Another facility manager described a case where the press had "bashed" the PA and made derogatory comments about his personal character based on the PA's refusal to set up tables in its facilities during Thanksgiving in order to serve the homeless. In fact, although the press did not pick up this story, the PA had paid for 400-500 Thanksgiving meals served at a local soup kitchen. The manager was deeply troubled because of the inaccuracy (in his mind) of the external portrait of the PA and the misinterpretation of his actions. "When you see your name in print and they call you callous and you know that in your heart you are probably one of the more compassionate people about this issue, it's hard not to get angry". At this point in time, clearly the issue was "hot" (emotionally charged) individually and organizationally, and PA actions heated up.

25 Key actions. The most dramatic actions during this time period involved financing and building two drop-in centers. In early 1988, the Board approved expenditures for building and operating two drop-in centers to service the Bus Terminal and the World Trade Center as well as commitment to opening them within a one-year time period. The total expense (operating and capital expense for these facilities was close to $2.5 million. The completion and opening of The Open Door Drop-in Center (adjacent to the Bus Terminal) by May, 1988 was viewed unanimously as a significant accomplishment, symbolizing the PA's commitment to the issue, and reaffirming PA members' identification of the PA as an organization that could "get things done". As an upper-level manager expressed it, "There have been more major achievements than anybody would ever imagine, because of the circumstances and the speed with which we have put this thing together". In October, 1988 New York City (actually the Human Resources Administration) balked on its informal agreement to take over the financing of the operations of the Open Door, and the PA's altered its stance on the issue. First, some members of the HPT and upper management expressed a hesitancy about solving the problem too well by getting into the building and managing of drop-in centers. In their minds, the incident with The Open Door taught them that they should not solve the problem too well "or we just get burned". As one informant told us, "Next time we will live with the problem much longer". There was a sense by members of the HPT and top management that the process that had been used to get the Open Door up and running created "expectations that PA would fund and operate facilities or created the impression that somehow the homeless at the Bus Terminal were the PA's problem". PA members became committed to eliminating this impression. Actions in the next issue phase were partly attempts to alter this false set of expectations. The financing and building of the second drop-in center (The John Heuse House) to service the homeless in lower Manhatten (near the World Trade Center) was also seen as a significant milestone in the processing of this issue. While this second drop-in center officially opened in December, its approval had been a rocky one, due to the formation of organized

26 opposition of downtown business interests. The opposition was strongest during the Spring when the City's Board of Estimate had to approve the operating contract with the church that would run the center. Summary. The 1988 period signified a critical phase in the PA's relationship to the homeless issue. It marked a turning point in the sense that the PA now viewed and justified action with a sense of resigned heroism —a sense that no one else will do it, so the PA will step in, in its usual, excellent way. The resigned admission that they had to do something on the issue was accompanied by a great deal of emotion about the unfavorable image that the PA was receiving in the press, the sense of outrage that those responsible were not doing their job, and a sense of embarrassment and anger at the negative press coverage of PA action on homelessness. The formation of the HPT helped to congeal a set of initiatives that had already begun in earlier phases. They were important catalysts to implementing the building and setting up the operations of the two drop-in centers. It was the time when instrumental involvement with the issue significantly escalated, evidenced by the expenditure of $2.5 million to fund initial operating and renovation expenses for The Open Door and renovations for the John Heuse House. Phase 5: Homelessness is an issue of regional competitiveness and we are quiet advocates on this issue. (Late 1988-Early 1989) While the PA's relationship to the issue of homelessness is still evolving, the data collection ended in May, 1989. Key Events. While the time of active data collection was nearing an end, one event did stand out in the minds of informants. In February, Newsweek published a particularly damaging article entitled "The Nightmare of 42nd Street" (describing the Bus Terminal) that added fuel to the fire surrounding efforts to improve the PA image. In this article the Bus Terminal was portrayed as a dangerous place for both commuters and the homeless: "a vortex of hopelessness, crime and despair" (Newsweek, February 27, 1989). In fact, one day after this article was published, the Board convened an emergency group to "try to do something dramatic to turn

27 around the PA image". The formation of this group signaled the heightened frustration with the tarnishing of the PA image through the strong association of the PA with the Bus Terminal, and with the Bus Terminal and the issue of homelessness. The Newsweek article and information collected during this time period also led to the acknowledgement and articulation that the problem with the Bus Terminal was far broader than homelessness —but involved issues of loitering and drug abuse. Key Interpretations. During Spring, 1989, there was an increasing awareness that although there had been some significant "victories", the homelessness problem was not going away. The PA was still getting "bashed" in the press, although the intensity of bashing was less than it had been during the previous two years. Informants acknowledged that the winter had been mild, making the visibility of homeless persons in PA facilities unusually low. At the same time, several of the PA initiatives put into place (e.g., rules and regulations, social service assistance) were producing some positive results. Top management claimed that the number of complaint letters received per week was significantly less than there had been a year ago (from 7 letters a week at the Bus Terminal to an average of one letter a week). Completion of the PA funded drop-in centers for the homeless signaled an increasing acknowledgement that the PA was getting more and more into the business of homelessness. As one informant put it, "yeah, we're two feet deep into the business of homelessness, and we don't want to be". Another informant displayed the ambivalence that accompanied this change in level of involvement: "We may be throwing a lot of resources at this, but our heart just isn't in it". A shift occurred in terms of defining the role of the PA on the homeless issue. The PA now framed its role as helping others "create capacity" for the special kind of homeless persons who frequented transportation facilities. So while PA management still adamantly denied that they were in the housing or social services businesses, they accomplished some of the same objectives "by increasing the capacity of other agencies that are better equipped to substantively address this issue".

28 Key Actions. The PA continued to implement its formulas for dealing with the issue that it had developed over the previous six years. It established outreach services that were offered at the airports. It also financially backed a deal with Jersey City to set up a drop-in center and SRO that would be run by Let's Celebrate (originally a soup kitchen and pantry operator) for its facilities at Journal Square. The drop-in center concept was consciously modeled after the John Heuse House arrangement, which in contrast to The Open Door arrangement was viewed as a more successful and appropriate model for PA involvement. The PA encountered delays and resistance to these facility solutions, but the resistance was treated as "normal" and "part of the process". The sense of urgency or outrage that had accompanied previous setbacks with the Open Door and John Heuse House Centers were notably absent. As one informant told us, "You learn that those people who fight you the hardest, may turn around and be your biggest advocate". Awareness that the PA was getting more involved in the issued (spending more money, adding services at more facilities), was accompanied by a conscious attempt to downplay or minimize the PA's public association with the issue. PA policy was explicitly designed to favor the role of "QUIET advocate for the single homeless male", (the typically homeless person for transportation facilities). Consistent with this thrust was a desire to not take the credit for any action or solutions for the homeless issue. For example, one staff member in remarking that the coverage of an incident by a local paper had been "balanced" and "good" explained that this meant not mentioning that the Port Authority had played any role in bringing about the successful solutions that the article described. Or as a top manager explained: "I don't want any credit. Let them take the credit. Let the bastards who fought us six months earlier take the credit. It's easy to give the credit. I prefer to work behind the scenes". Part of the quiet advocate role involved educating others about the special needs of homeless persons at transportation facilities in the context of a transportation forum. At the same time, the PA began to more actively recruit and link up with other transportation agencies on the issue. For example, members of the HPT began to meet with their

29 counterparts at the MTA (Metropolitan Transportation Authority). "We are trying to broaden the circle of people who participate, working with the business community as a team". At this point in time, a major thrust of PA action in the issue was to actively seek partners on the issue, although the form of these partnerships, and the sorts of solutions this implied were not known. However, the PA's position at this point was that the agency would offer its "special expertise and viewpoint on the issue to the City and to businesses who needed it". An unintended consequence materialized from publicity over the Culman Fellowship and other efforts to manage outsiders' impressions of the PA's stand on homelessness. The PA was increasingly viewed as a "leader" on this issue. Organizational members described the PA as "on the cutting edge of what a transportation agency can do on this issue", or "the most creative solutions are being offered by the PA to this problem". However, some managers were quick to see that this reputation represented a double-edged sword: "I think there is another temptation, which is a peculiar PA temptation. There's a tendency in a lot of places around this organization that wants people to get involved in something, and they want to be leaders in it. I just want to deal with this problem, not become a leader on it". Summary. The relationships of the PA to the homelessness issue took a new turn in 1989. While still not solidified (as one respondent called it, "We are still like an ameoba with this issue"), the PA's position was increasingly deliberate and active in what actions it highlighted and what actions were downplayed. During 1989, for the period of time that data were being collected, the PA more actively managed the context in which the issue was affecting them. These efforts included a search for partners for designing new collective solutions to this regional crisis. Efforts involved presenting information about the issue and information about the PA's action on the issue in a way that would minimize image damage by disassociating the PA from the issue. The efforts took place within the constraints of taking actions consistent with the PA' identity (e.g., that complemented their perceived expertise). At the same time, the PA was increasingly recognized as a leader on how to deal with homelessness within the transportation industry. PA members expressed tremendous pride in the PA's method for dealing

30 with the homeless. In their eyes it was the "most humane approach" used by transportation agencies in the region. DISCUSSION AND IMPLICATIONS The story of the PA's relationship to the issue of homelessness is still unfolding today. Despite its complexity, the pattern in this story is sufficiently prominent to extract, examine, and build on several important themes. Two central themes that emerged from analysis of interviews, media coverage and internal memos focus on the role that the organization's identity and image played in creating the pattern of how individuals in organizations interpreted and responded to the homelessness issue. More specifically, we found that the PA's identity played a key role in judging whether or not the organization should take action ouon the issue, the relevant skills and procedures for issue action, and in defining the criteria for judging issue action success. Each of these claims is illustrated in more detail below. At the e time, the organization's image (how organization members thought others saw the PA) served as the gauge against which action on the issue was evaluated and justified. In addition, the organization's image was an important mirror for judging and prompting issue action because of close link between insiders' view of the organization and insiders' and outsiders' inferences about the characters of organizational members. Over time, actions taken on issues reposition an organization in its environment by modifying tasks, the allocation of resources, and the assignment of personnel. The pattern of action on issues reinforces or transforms the organization's identity and image through individuals' sense-making efforts, and the process of adaptation continues. The Importance of Organizational Identity An organization's identity refers to members' perceptions of the central, enduring and distinctive characteristics of the organization (Albert & Whetten, 1985). The identity of the organization is important to individuals as it embodies characteristics, which, by a member's affiliation with the organization, are generalized to the individual (Ashforth & Mael, 1989). Thus, organizational identities provide meaning to the individuals who belong to them. The

31 power of the organization's identity to grant meaning and identity to individuals is probably most evident when the identity of the organization is spoiled, and this generalizes to an individual's identity. Sutton and Callahan (1987) showed this in the case of managers in firms that had filed for Chapter 11. Bridges (1988) describes similar consequences for employees of E. F. Hutton, Johns Manville, A. H. Robins and Union Carbide. In all of these instances, employees had to cope with a loss of faith in oneself that arose from a loss of faith in the organization. The Identity of the Port Authority. PA members consistently described five attributes as distinguishing their organization. First, members believed the PA was a professional organization, consisting of a skill base that is uniquely suited to technical experts. Second, was the sense that this professional organization was first class, high-quality, a provider of superior service. Third, PA members prided themselves in being fixers, or a "can do" organization. This third characteristic was expressed in several managers' reactions to the homelessness issue: "The people are obviously needy and to stand by and not to be able to fix it —well, that is sad and very frustrating. We are an agency of fixers". Fourth, the agency prided itself in its high commitment to the welfare of the region. Part of this dimension of the PA's identity was a sense that the PA "spoke for the region", and symbolized the region's successes and shortcomings. Finally, PA members consistently referred to their organization as ethical, scandal-free, and altruistic. The PA's identity played a critical role in how PA members defined the homelessness issue. It offered well-used recipes for issue-level action, and defined criteria for judging action success. Identity and issue processing. The PA's identity was a critical force in defining homelessness as an issue to which the PA should not respond. Particularly in the early phases of the issue's processing, justification for non-action on this issue centered around not having the social service skills for dealing with homeless persons. The strength of sentiment behind this belief was expressed in stories some informants told us about incidents where technical experts were diverted from using their technical skills. For example, we were told several stories by different informants about architects holding AIDS babies, engineers having to change babies'

32 diapers, or sanitation engineers having to clean the filth in bathrooms —all related to the issue of homelessness. Whether the substance of the story was accurate or not is less important than the values that the stories conveyed. The stories communicated a great disdain about the inappropriate diversion of technical skills for the delivery of social services. This disdain was a strong defense for not responding to the homelessness issue, particularly in the 1982-1986 period. In addition, the intractability of the homelessness issue, and PA members' sense that the issue could not be controlled, were blasphemous to an organization which considered itself to be a "fixer" and "doer". This sense of not being able to control the homeless issue further defended and delayed PA involvement. However, these defenses were no longer sustainable when the problem worsened, and when the image of the organization was severely damaged by the issue's visible appearance in PA facilities. The PA's identity was associated with a set of routines or standard procedures for dealing with the issue, that once activated, engaged familiar ways of doing things that PA members identified as "typical of the PA". In this sense, an organization's identity is closely tied to an organization's culture by providing a set of skills and a way of using and evaluating these skills that produce typified ways of doing things (Swidler, 1982). The organization's identity is one of the vehicles through which "preconceptions determine appropriate action" (Weick, 2988, p. 306). For example, when the homelessness issue was no longer deniable, the PA went to work to "Get smart on the issue". The phrase describes the PA's ideal approach to a problem —investigating and analyzing the issue from all angles. They learned a lot about the unique attributes of homeless persons at transportation agencies. The engagement of learning routines for understanding all angles on a problem was seen by some as typical of the PA, and indicative of its "professional" character. The idea of searching for and engaging partners for dealing with the issue, and framing the issue as related to the region's future were also seen as actions that "typified the PA approach to things". Finally, the PA's identity offered implicit guidelines for evaluating the effectiveness of PA action on the issue. A good example of this role was the use of speed of completion of the

33 two drop-in centers as a criterion for judging the success of the HPT and the success in dealing with the issue. Informants told us with great pride that the PA was the only agency that could ever approve, design and implement the opening of the two drop-in centers within such a short period of time. The efficiency of this accomplishment was used as an important barometer of the PA's success with the issue, even though PA members admitted that the actual problem in terms of number of homeless at PA facilities had not really lessened. A summary of the relationship between the PA's identity and interpretations, emotions and actions on the issue are summarized in Figure 2. Insert Figure 2 about here The Importance of the Organization's Image While an organization's identity describes what individuals believe is the internal reflection of the organization, the organization's image refers to those attributes which organizational members believe others outside the organization use to distinguish the organization. Thus, the organization's image captures the external reflection of the organization as seen by those inside. The organizational image is different from what economists or corporate strategy researchers call reputation (e.g., Fombrun & Shanley, 1989; Weigelt & Camerer, 1988; Wilson, 1985). These researchers tend to view reputation as a composite of attributes ascribed to a firm by those outside of the firm. By image, we are referring to shared attributes or core features that those inside the firm believe that others outside of the firm hold. In certain cases image and reputation will be highly consistent, as when the reputation of a company like Merck Pharmaceutical gains so much publicity that the image and reputation become one in the same. However, in most cases, there is divergence between how those outside and those inside view the essence of the firm. In the case of the Port Authority and its dealings with the issue of homelessness, it was the PA image, as we have defined it, that played a key role in motivating individuals to escalate and modify their actions on the issue.

34 The Image of the Port Authority. The PA believed its image had suffered tremendously as a result of its involvement in the homelessness issue. One theme that recurringly appeared in the interview data was that homelessness negatively impacted the image of the PA. Related common themes were that homelessness made the PA vulnerable to bashing from the press, and that the homelessness issue was linked to other negative issues, such as crime and drugs. PA members felt the organization was seen as ineffective, dirty, dangerous, inhumane, and rich. This image was based on three sources of information: the number of complaint letters received, actual confrontations with PA employees by customers and tenants, and negative articles about the PA in the press. Several quotes are illustrative of the effect of homelessness in the PA image. "...we began to get our first series of letters about them, and the letters were written by people who referred to the homeless people as filth, vermin, and that we should do something about that. And this gave the impression that it was a disgusting kind of thing". "They [homeless] were interfering with our patrons in the sense that they felt that they were not safe because of their presence". "It's gotten more violent, in the sense that now people begin to feel there's a relationship between homelessness and violence". Perhaps most frustrating to our respondents was the fact that nothing they could do pleased everyone. On one side were customers and tenants who were furious at the PA's ineffectiveness in removing the homeless from the facilities: "We have letters here from people that are moving their businesses out of New York because they won't have their staffs traveling through the Bus Terminal. A guy says, 'I've worked in Manhattan for 20 years. I've recently taken a job in Paramus, New Jersey, and the sole reason that I took the job in Paramus is because I refuse to come into your Bus Terminal any more. I can't be subjected to that; my day begins every day with, you know, this vision and I can't put up with it anymore'". On the other side were people who complained about the inhumaneness of the PA. "I think people have the perception, and the media in particular had the perception, that cops were indiscriminately tossing people right and left, or arresting them, or getting them out, just because they didn't like the way they looked. I know cops at the Bus Terminal that fed homeless people; that was the only meal they got, every single day for two and a half years".

35 It was this contrast between what PA members thought about themselves and the organization, and what they believed others thought of them which led to a high degree of frustration. "The PA bashing goes on for lots of different reasons. One is the image that was created, and which I think is inappropriate —that we have a lot of money to throw around. Which we don't. That we are very powerful, and that we stand aloof of issues like this. In fact we don't". The Image and Issue Processing. For PA members, the image of the organization served as a gauge or reference point for determining the need for and evaluation of action. Members of the PA were aware of its image as it related to homelessness and of the need to manage it. "We discussed it [the homelessness issue] in the same terms that we discussed marketing of a facility; or we strategized about, you know, the public perception of this issue and our activities about this issue". When the Port Authority's image began to "slip" in 1987, attributed in part to the association of homelessness with the Bus Terminal, and the Bus Temninal with the PA, top management became concerned because the tarnish to the PA's image from its association with 'homelessness was not being overridden by the positive actions that the PA was doing. Frustrations began when PA renovations at the Bus Terminal in 1982 made the homelessness problem seem even more severe. In the minds of those at the PA, not only could positive actions of the PA not overcome the image damage to the PA, but the stain or blemish from homelessness spread to the entire region. As described by one top level manager: "The quality of life of the region is severely impacted by having as a kind of visible ornament, a large number of people who are described as homeless.... It creates an environment of extraordinary depression in a transportation mix which is already congested, difficult and harassed. In some ways, like the graffiti on the subways, it is both a fact and a symbol that the environment is out of control". Some PA members believed that the PA as an organization, and the NY-NJ area as a region, was unable to compete effectively in an international transportation market because of image damage to the PA.

36 Impetus to taking action to improve the damaged image of the PA brought on by its association with homelessness was more than organizationally-based. As the story revealed, damage to the organization's image hurt some people personally as well. Spoiled organizational images transferred, soiling the individual characters of organizational members (Sutton & Callahan, 1987). Researchers tell us that this link between organizational image and individuals' motivation becomes more tightly coupled when actions that affect the image are public and irrevocable. As Weick (1988) notes, it is in these situations that the actions "become harder to undo" and "harder to disown" (p. 310). As a result, individuals are strongly motivated and committed to take actions that will restore the organization's image, and hence restore the integrity and in this case, the implied humanity of PA employees. The close link between an individual's character and the organization's image implies that individuals are personally motivated to preserve a positive organizational image and repair a negative one, through association and disassociation with actions on issues. This explanation is complementary to Sutton and Callahan's (1987) description of how filing for Chapter 11 caused such energy to restore the self image (and remove the stigma) for the managers in the eyes of critical organizational audiences. Like Sutton and Callahan, the pattern of actions that we observed in the Port Authority's struggle with the issue of homelessness were attempts to actively manage outsiders' impressions of the organization, however, subject to the constraints of doing things that were consistent with the organization's identity. In the case of the PA and the issue of homelessness, the organization continues to try a variety of impression-image management tactics (Sutton and Callahan, 1987). It keeps trying different tactics to see if they effectively transform the image problem, while not violating attributes that define its core identity. In this sense, the pattern of actions represents a type of continuous experimentation and learning process, that became more deliberate over time. While throughout the dealings with the issue the PA denies responsibility for the problem, it steps in when there is no alternative, and takes some action (building the drop-in centers). However, as the significance of the actions on the issue increase (e.g., as human and monetary resources

37 invested in the issue increase), the PA becomes more strategic in which actions it wants to associate with, and which actions it wants to conceal. As of the time we stopped collecting the data in mid 1989, the PA was actively educating and sharing information with other transportation agencies on what could be done on the issue of homelessness (its advocacy actions), but intentionally maintaining a low profile in the development of programs and services for the homeless (its quiet actions). In the minds of the homelessness project team and most of upper management, the costs of being associated with taking responsibility for the homelessness issue far outweighed any gains from being seen as the builder of superior drop-in centers. The pattern of actions that we observed over time was partially a trial and error imagemanagement pattern, that became more assertive and less defensive over time (Tedeschi and Melberg, 1984). The facility-based solutions were largely reactive ones, based on attempts to conceal, contain and eliminate the problem. However, as the problem became more severe (as the number of homeless rose and its link to crime and drugs became stronger), as the image deterioration amplified emotional reactions to the issue (bad press, not getting credit for its capital plan), the organization went into "high gear" on the issue in an instrumental sense and "low gear" in a public sense. In a way that was consistent with its high-quality, efficient and fixer-doer identity, the organization proposed and funded major outreach facilities for the homeless near three of its affected facilities. The links between the organization's image and interpretations, emotions and actions on the issue are summarized in Figure 3. Insert Figure 3 here Links to Organizational Adaptation The ideas of organizational image and identity help to make sense of the pattern of interpretations and actions on the homelessness issue for the Port Authority. In 1982, the initial denial of the need to respond to the issue of homelessness and the fragmented, facility-level response were hard-wired responses to the issue based on the organization's identity. The organization's identity contained certain kicks to action (and defenses to non-action) that

38 resemble Starbuck's (1983) idea about action generators. The limited response and the intensification of the problem contributed to a very negative image of the PA, which fueled a stronger, but identity-consistent response. Over time, however, the level of resource allocation to the issue substantially increased, the definition of the issue was considerably broadened, and the organization tried to disassociate itself from certain actions on the issue, while actively managing outsiders' understanding of the issue. The escalation of resource commitment can be seen as prompted by the image damage that the organization incurred in the 1986-1987 period, the intensification of the problem itself, and the mobilization of resources towards the problem directed by the HPT. These actions, in turn, while altering some dimensions of the organization's image (now the organization could count the number of persons directly helped by the PA to help offset its "Calcutta in NY image"), were also strengthening the humane-altruistic dimension. Thus, the actions taken to restore the image were at one time, altering this image, but also modifying the identity of the organization, by strengthening the centrality of some characteristics and suggesting new possible characteristics that made the PA distinctive (e.g., its creative actions on this issue). It is this latter relationship that helps to illustrate the importance of the feedback loops between organizational actions, image and identity, as depicted in Figure 4. Insert Figure 4 here The view of organizational adaptation that arises from this case study is one that is consistent with ideas embedded in some current theories, but also unique in its specification of a link between micro and macro-level phenomena. It implies that micro-level (e.g., individual commitment and motivation to work or not work on an issue) are related to interpretations of organizational actions. Like theories of environmental enactment (e.g., Weick, 1977), this view puts individuals' interpretations of organizational actions as a centerpiece to the adaptation process. However, the view departs from enactment models by suggesting that it is the organization's image (its external reflection) and its identity (its internal reflection) that are the

39 important reference points for interpreting problematic issues (like homelessness) and for galvanizing interest behind actions on the issues. In this way the model is consistent with Bartenuk's (1984) study of strategic change in a religious order in that the organizational context (in this case the image and identity) mediate the effect of the environment (in this case the rise in the number of homeless persons) on organizational action and/or change. The PA's experience with the homelessness issue also emphasizes a number of themes that have been raised in the organizational literature but which have direct implications for thinking and studying adaptation process. First, it suggests that environments, and issues which sometimes arise from them, are socially constructed. Objective facts matter far less than how the issue is bounded and interpreted by organizational members (Smircich and Stubbart, 1984; Weick, 1979). Second, it suggests that patterns of action and change are issue-specific, and evolving over time (Dutton, 1989; El Sawy & Pauchant, 1988; Weiss, 1990). The pattern of interpretation and action on this issue formed a type of natural history —with new actions and interpretations arising from previous episodes. This discovery merely reinforces an old story — that traces of processes are time-dependent. Consequently, students of adaptation processes must use process methods and process theories to capture the patterns of change in a meaningful way (Mohr, 1981). The paper will end with the usual caveats about the dangers of generalizing results from the study of one organization (after all, the PA is an unusual organization), and one issue (after all, the issue of homelessness is an unusual strategic issue). However, consistent with the spirit that Glaser and Strauss (1967) intended, new ideas for theory building have been generated. The idea that an organization's identity and image are central to understanding how issues are interpreted, how reactions are generated, what emotions are evoked and how this pattern of behaviors are related to one another in an organizational context is very simple. It suggests that organizations, like individuals, keep one eye on the mirror when they act and make sense of actions. Strategy, organization theory and management researchers might better understand how organizations behave by asking where they look, what they see, and whether or not they like it.

40 Table 1 Description of Informants and Clusters of Variables Informants Top level managers (including Executive Director) 4 Members of Homeless Project Team 6 Line managers (Facility Managers) 6 Staff involved specifically in homelessness 5 Police and customer service managers 4 N =25 Clusters of Variables: 1. Issue Interpretations: a) Emotionality of issue b) Issue's distinctiveness and similarity to other strategic issues c) Perceived "hotness" d) Interrelationships with other issues 2. Personal involvement in the issue: a) Time spent on the issue b) Amount of direct contact with homeless persons c) Nature of change in involvement 3. Organizational processing and actions on the issue: a) When issue first noticed b) Major milestones in the processing of the issue c) Major setback points d) Major success points 4. Perceived effectiveness of issue processing: a) Costs and benefits for PA's involvement in the issue b) Evaluation of HPT's handling of issue 5. Structural context of the issue: a) Shared values at PA b) Institutional mission c) Task context

Figure 1 Snapshot of the Issue History 3 (1987) TIME 1 1982-1984 2 (1985-1986) 4 (1988) 5 (1989) INTERPRETATIONS @is a policesecurity issue )is a corporate issue but not in SS business U)is a business problem and need moral solutions KS) @is an issue of regional image and no one else will solve it (most emotional) @linked to other regional problems, )of transportation are unique and need advocates EVENTS *BT remodeling *)spreadlng to Repeal of Burned by NYC *SROS removed soul of organization anti-loitering law Ca. pitalplan launched: 'Leadershlp A ->1000/night -Appearance in - Bad press business plans ACTIONS -Police training ~Outreach services mom. P- w W- -~~~~~~~~~~~~~~~~~~~~~~~~I *Collect data on@ 'Restricting access laaa, --- r~ ~ ~~ * o o >~~~~~0 |. w.. _ o.. a -Find humane alternatives: Drop-in centers proposed 'Formation of HPT ~Culman Fellowship......... P o _..... o. lw....... *Go public on the issue 'Rules and regulations *Authorize and build 2 drop-in centers @ = Homelessness *Quiet advocate for Transp () 'Seeking partners-creating capacity *Approve 3rd drop-in center *Educate others about( da. i-a

42 FIGURE 2: ORGANIZATIONAL IDENTITY AND ITS RELATIONSHIP TO INTERPRETATION, EMOTION AND ACTION Elements of PA Identity: --- Issue Behaviors Issue Interpretations I * Legitimates attention to issues (e.g., "This is not our issue, we are not a social service agency") Issue Emotions I * Evokes emotion if identity improperly used (e.g, engineers holding AIDS babies) -H Issue Actions I * Provides recipes (skills and procedures) for issue action (e.g., "Get smart on the issue," develop partnerships") * Guidelines for evaluating action success (e.g., "Speed of completion of dropin center")

43 FIGURE 3: ORGANIZATIONAL IMAGE AND ITS RELATIONSHIP TO INTERPRETATION, EMOTION AND ACTION PA Image on Homelessness: - Issue Behaviors: Issue Interpretations * Weighting image gain/loss relative to investment in issue Ineffective (e.g., not getting benefits from big investment in capital plan relative to loss from association with homelessness) j^ pA Rich * Reference point for determining f/ Image from \ / importance of action Homelessness Homelessness (e.g., when image severely ~\. ~ / y deteriorated, issue interpretation was broadened and seen as more corporate):rous Inhumane Issue Emotions * Generalization of organizational image to individual character (e.g., PA is inhumane, so its members are inhumane) Issue Actions l * Defines what actions to associate or disassociate from (e.g., go public on being advocate for transportation homeless, stay private on building drop-in center)

44 FIGURE 4: SIMPLIFIED DEPICTION OF ROLE OF ORGANIZATIONAL IDENTITY AND IMAGE ON PA'S RESPONSE TO HOMELESSNESS I I I I I I I I Issue interpreted through lens of organization's identity Hard-wired, identityconsistent, response Identityconsistent response and intensification of problem — contribute to deterioration of PA image Image deterioration contributes to more pronounced identity —consistent response (~ = Homelessness

45 REFERENCES Albert, S., and Whetten, D. 1985. Organizational identity. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior, 7: 263-295. Greenwich, Conn.: JAI Press. Ashforth, Blake, and Mael, F. 1989. Social identity theory and the organization. Academy of Management Review, 14, 1: 20-39. Bachrach, P., and Baratz, B. 1962. The two faces of power. American Political Science Review, 56: 947-942. Bartenuk, J. 1984. Changing interpretive schemes and organizational restructuring. Administrative Science Quarterly, 29: 355-372. Bjorkegren, D. 1989. It doesn't have to be that way. Paper presented at the Organizational Behavior Teaching Conference. Bridges, William. 1988. Surviving corporate transition. New York: Doubleday. Calicchio, D. 1987. PA takes a tough stance against the homeless. The Jersey Journal, December 11. Caro, Robert. 1974. The power broker: Robert Moses and the fall of New York. New York: Vintage Books. Daft, R., and Weick, K. 1984. Toward a model of organizations and interpretation systems. Academy of Managment Review, 9(1): 284-296. Dutton, J. E. 1988. Understanding strategic agenda building in organizations and its implications for managing change. In L. R. Pondy, R. J. Boland and H. Thomas (Eds.), Managing ambiguity and change: 127-144. Chichester: John Wiley & Sons. Dutton, J. E., and Duncan, R. B. 1987. Creation of momentum for change through the process of strategic issue diagnosis. Strategic Management Journal, 8: 279-295. Eisenhardt, K. M. 1989. Building theory from case study research. Academy of Management Review, forthcoming. El Sawy, O. A., and Pauchant, T. C. 1988. Triggers, templates and twitches in the tracking of emerging strategic issues. Strategic Managment Journal, 9: 455-473. Fombrun, C., and Shanley, M. 1989. What's in a name? Reputation-building and corporate strategy. Academy of Management Journal, forthcoming. Glaser, B., and Strauss, A. 1967. The discovery of grounded theory: Strategies for qualitative research, London: Weidenfeld. Hackett, S., and McKillip, P. A nightmare on 42nd Street. Newsweek, February 27, 1989. Hambrick, D., and Finkelstein, S. 1987. Managerial discretion: A bridge between polar views of organizational outcomes. In L. L. Cummings and B. M. Staw (Eds.), New directions in organizational behavior: 369-406. JAI Press.

46 Hannan, M., and Freeman, J. 1984. Structural inertia and organizational change. American Sociological Review, 49: 149-164. Hilgartner, S., and Bosk, C. L. 1988. The rise and fall of social problems: A public arenas model. American Journal of Sociology, 94(1): 53-78. Meyer, A. 1982. Adapting to environmental jolts. Administrative Science Quarterly, 27: 515 -583. Miles, R. H., and Cameron, K. 1982. Coffin nails and corporate strategies. Englewood Cliffs, NJ: Prentice-Hall). Miles, M. B., and Huberman, A. M. 1984. Qualitative data analysis. Beverly Hills, CA: Sage Publications. Mintzberg, H., Raisinghini, A., and Theoret, L. 1976. The structure of unstructured decision processes. Administrative Science Quarterly, 21: 246-275. Mohr, Larry. 1982. Explaining organizational behavior: The limits and possibilities of theory and research. San Francisco: Jossey Bass. Nolan, J. Terminal disease. New York Post. January 22, 1988. Nutt, P. 1984. Types of organizational decisions. Administrative Science Quarterly, 29, 414 -450. Sandelands, L. E., and Drazin, R. 1989. On the language of organization theory. Organization Studies, forthcoming. Smircich, L., and Stubbart, C. 1985. Strategic management in an enacted world. Academy of Management Review, 10: 724-736. Starbuck, W. H. 1983. Organizations as action generators. American Sociological Review, 48: 91-102. Starbuck, W. H., and Milliken, F. 1988. Executives' perceptual filters: What they notice and how they make sense. In D. Hambrick (Ed.), The executive effect: Concepts and methods of studying top managers. JAI Press: Greenwich, Conn. Sutton, R., and Callahan, A. L. 1987. The stigma of bankruptcy: Spoiled organizational image and its management. Academy of Management Journal, 30(3): 405-436. Swidler, Ann. 1986. Culture in action: Symbols and strategies. American Sociological Review, 51: 273-286. Tedeschi, J. T., and Melburg, V. 1984. Impression management and influence in the organization. In Sociology of organizations. JAI Press: Greenwich, Conn., 3: 31-58. Tushman, M., and Romanelli, E. 1985. Organizational evolution: A metamorphosis model of convergence and reorientation. In L. L. Cummings and B. Staw (Eds.), Research in organizational behavior, 7: 171-222.

47 Weick, K. E. 1988. Enacted sensemaking in crisis situations. Journal of Management Studies, 24(4): 305-317. Weick, K. E. 1977. Enactment processes in organizations. In B. Staw and G. Salancik (Eds.), New directions in organization behavior: 267-300. Chicago: St. Clair. Weick, K. E. 1979 The Social Psychology of Organizing. Reading, Mass.: Addison-Wesley. Weigelt, K., and Camerer, C. 1988. Reputation and corporate strategy: A review of recent theory and applications. Strategic Management Journal, 9: 443-454. Weiss, Janet A. 1989. The powers of problem definition: The case of government paperwork. Policy Sciences (forthcoming). Wilson, R. 1985 Reputations in games and markets. In Roth, A. E. (Ed)., Game-theoretic Models of bargaining. Cambridge University Press. Yin, Robert. 1984. Case Study Research. Beverly Hills, CA: Sage Publications.