Not All Events Are Attended Equally: Toward a Middle-Range Theory of Industry Attention to External Events

This paper builds on prior theory and research on attention and identity to examine whether and how industries publicly attend to external events. Events are critical triggers of institutional transformation and industry evolution. However, they must first become the focus of public attention to have this effect. We draw on a paired case comparison of media coverage of eight nonroutine events affecting the natural environment and the U.S. chemical industry. We employ both deductive and inductive analysis to develop a model and hypotheses to explain two research questions. First, what determines the initial public attention to an event? Second, when and why do certain events attain high and sustained levels of industry attention? A key inference is that whether an event receives industry-level attention depends on either outsiders holding the industry accountable for the event, or insiders’ internal concerns with the industry image. We further infer that an event can be transformed into a critical issue for an industry, warranting sustained attention, if there is contestation with outsiders over the accountability for the event and its enactment, and internal contradictions and challenges to the industry’s identity. (Events; Attention; Identity; Institutions; Accountability; Environmental Protection) Introduction Highly publicized events are critical triggers of institutional transformation (Fligstein 1990, Sewell 1995, Hoffman 1999). Such public occurrences, here called critical events, are contextually dramatic happenings that focus sustained public attention and invite the collective definition or redefinition of social problems (Pride 1995). Variously referred to as shocks (Fligstein 1991), jolts (Meyer 1982), or discontinuities (Lorange et al. 1986), critical events have played a central role in fostering institutional change and industry evolution (Miles 1982, Leblebici et al. 1991). The publication of Silent Spring (Carson 1962) provides an example. Silent Spring triggered, within weeks of its release, a political and cultural struggle between the chemical industry, scientific academies, conservation groups, and various government agencies on the industry’s accountability for the ecological dangers of synthetic chemical production. The book’s author, marine biologist Rachel Carson, argued that chemical manufacturers, by barraging the environment with the synthetic pesticide dichlorodiphenyltrichlorethane or DDT, were poisoning the entire food chain and ultimately ourselves. For the U.S. chemical industry, this book release was no small affair. What was at first viewed as a possibly irritating event quickly enveloped into an issue of critical proportions. It became a threat to the image and identity (Dutton and Dukerich 1991) of the entire chemical industry and a challenge to the technological preeminence of synthetic chemical production (Florman 1976, Pillar 1991). This challenge triggered unprecedented public attention (Hilgartner and Bosk, 1988) from industry associations and individual companies. Ultimately, Silent Spring facilitated changes governing chemical industry action, clearing the way for increased government controls on pesticide application. Past organizational research on critical events has focused on the processes of sense-making (Isabella 1990, Thomas et al. 1993, Gioia and Thomas 1996) and the construction of accounts (Elsbach 1994). But despite the ANDREW J. HOFFMAN AND WILLIAM OCASIO Not All Events Are Attended Equally ORGANIZATION SCIENCE/Vol. 12, No. 4, July–August 2001 415 centrality of public attention to events in triggering institutional change, very little work in organizational studies has addressed why some events become the critical focus of attention while other events remain mostly unnoticed (Hoffman 1999). Not all events are attended equally. For example, why did Silent Spring receive substantial media attention in the trade journal Chemical Week while another major environmental book, The Limits to Growth, received limited coverage? The former event led to significant institutional change in environmental policies and practices for the chemical industry. The latter faded in public attention. Why? Existing theory on public attention (Hilgartner and Bosk 1988) focuses on competition for attention among broad social problems such as water pollution, the energy crisis, and the homeless, but does not explain the level of attention to specific events or why some events become critical problems while others do not. To address this gap in the literature, we undertake an analytical case comparison of public attention to eight environmental events by the U.S. chemical industry. We selected events relating to the natural environment for our study because they provide substantial variation in the level of public attention they have received by industry (Hoffman 1997, 1999). This variation allows for an exploration of the determinants of public attention and inattention that avoids sampling on the dependent variable (King et al. 1994). Building on methodologies of comparative studies of events (Skocpol and Somers 1980, Hicks 1994, Mahoney 1999), our objective is to construct a middle-range theory (Merton 1957, Eisenhardt and Bourgeois 1988) of industry-level attention to external events. In the following sections we first discuss the general theoretical framework on industry-level attention that guided our research study. Second, we discuss our data and methods used in the selection and analysis of the eight external events. We then present our cross-case analysis, out of which emerge our middle-range theory and explanatory hypotheses. Finally, we present the conclusions of our study and guides for further research. Theoretical Presuppositions and Research Questions We began our comparative analysis of cases with a set of orienting theoretical assumptions derived from attentionbased theories of organizational action (March and Olsen 1976, Weick 1979, Dutton 1997, Ocasio 1997). These theories view the environment as a source of constant input and stimulus for the organization, but posit that individuals and organizations have limited cognitive capabilities to deal with all available stimuli (Simon 1947, March and Simon 1958). At the level of individuals, attention encompasses the noticing and focusing of time and effort on both the environmental stimuli requiring action and the available repertoire of responses which define that action (Ocasio 1997). In this paper we focus on attention at the level of the industry. We introduce the concept of industry-level attention, which highlights how industry participants, in their communications and interactions with other industry participants, selectively focus their attention on a limited set of issues, situations, and activities that represent potential problems or opportunities for the industry. In particular we focus on industry attention to events external to the industry. In defining industry participants, we employ a field-level perspective (DiMaggio and Powell 1983) and include not only representatives from the producer organizations in the industry, but also those from industry associations, trade journals, and other members of the industry’s field. While ultimately thinking and attending are activities of individuals, cultural and social processes at the level of an industry shape whether, when, where, why, and how decision makers attend to issues and events (Douglas 1986). A critical principle of attention-based theories is the principle of selective attention (Simon 1947, Fiske and Taylor 1991, Ocasio 2001). This principle suggests that individuals, organizations, and industries will selectively attend to some external events while ignoring others. Attention-based perspectives further posit that selective attention is driven not by the objective characteristics of the situation or event, but by its enactment in the environment (Weick 1979, Ocasio 2001). According to Weick (1979, p. 164), “enactment emphasize(s) that managers construct, rearrange, single out, and demolish many of the objective features of their surroundings.” Enactment actively orders the environment through the imposition of schemas and causal maps on the objects of action. Selective attention to events is driven by salience (Fiske and Taylor 1991) and salience is shaped by how individuals, organizations, and industries enact events in the external environment. A second principle of attention-based perspectives is that of situated cognition (Suchman 1987, Ross and Nisbett 1991, Ocasio 1997). This principle posits that the attention of industry participants to particular issues and answers is situated within the particular channels of communication through which they interact. In this paper, for example, we focus on the internal channels of the chemical industry through the trade press as compared to the external channels of the broader field through the general news media. For both, we draw upon the concept of public attention (Hilgartner and Bosk 1988, Fine 1997, Rao ANDREW J. HOFFMAN AND WILLIAM OCASIO Not All Events Are Attended Equally 416 ORGANIZATION SCIENCE/Vol. 12, No. 4, July–August 2001 et al. 1999). Hilgartner and Bosk (1988) treat public attention as a scarce resource for which potential issues compete for time and space. In their framework, competition for attention occurs within public arenas or institutionalized channels of communication and social interaction (Ocasio 1997). But, where they look only at external attention to issues, we also consider attention by insiders and the linkage between them both. Relevant arenas may include the press, professional conferences, congressional committees, and academic journals. Each arena possesses limited carrying capacity, so only a few events or issues gain public attention, while most are ignored. According to the principle of situated attention, different public arenas will selectively focus attention on different issues and events in the external environment. A third principle of attention-based perspectives is that of the structural determination of attention. This suggests that how people think and how they attend to an event is a social and cultural process, shaped by the group, organization, industry, and organizational field (Ocasio 1995, 1997). Previous theory (Ocasio 1997) suggests that whether a given issue attracts public attention depends upon whether the claims surrounding it are supported by the following social structures of attention: the rules of the game, status of the players, their social identity and structural position, and the available technology and resources. We draw upon these theoretical categories in our inductive analysis of the determinants of industry-level attention. We rely on identity as a key component of the social structures of attention (March and Olsen 1976, Porac et al. 1989, Dutton and Dukerich 1991, White 1992, Ocasio 1995). We draw upon sociological (Douglas 1986, White 1992) and organizational (Albert and Whetten 1985, Dutton and Dukerich 1991) conceptions of identity. All of these perspectives emphasize the sameness of those who share a common collective identity, and the distinctiveness, real or imagined, between the collective identities of different social groupings. We thereby define industry identity as the common rules, values, and systems of meaning by which industry participants establish rules of inclusion, competition, and social comparison among industry members; create distinctions within and between industries; and delimit industry boundaries. Industry identity emerges both from cognitive awareness among industry competitors about the nature of industry rivalry (Porac et al. 1989) and from collective responses to external threats to the collectivity (White 1992). Industry identity embodies meaning and sense-making (Fiol et al. 1998) focused on answering the following questions for its members: Who are we? What are we? What do we do that makes us distinctive as an industry? While industry identity, like organizational identity (Whetten and Godfrey 1988), is often subject to contestation and change, it is an important influence upon actors’ collective behavior. For example, Florman (1976) and Hoffman (1997) describe the identity of the U.S. petrochemical industry in the 1950s as being embedded in beliefs in technological optimism. The self-perception was that member companies of this industry were improving the quality of life for individual Americans and the strength of the nation as a whole. Companies were proudly mobilizing America by fueling the record number of automobiles being produced and the economy’s expanding industrial base, and providing miraculous new materials that were revolutionizing fields such as medicine, food production, and fashion. Following previous theory and research (Dutton and Dukerich 1991, Dutton et al. 1994), we posit that the industry’s collective identity is shaped by its image. Industry image is defined as the industry’s internal perception of how outsiders think about them, their values, and their beliefs (Dutton and Dukerich 1991) as distinct from the industry reputation, defined as the status ascribed to the industry by outsiders (Fombrun and Shanley 1990). While image results from internal sense-making (Gioa and Thomas 1996), reputation results from external attributions. In sum, emergent norms of industry interaction, coupled with an examination of industry image and reputation, shape and constitute industry identity. We use the perspectives on industry attention and identity outlined above as an initial conceptual guide (Miles and Huberman 1994) in our examination of industry attention to external events. Building on this theoretical framework, the following research questions guide our theory development and hypothesis generation: Research Question 1: What explains whether and when some events receive public attention within an industry, while others are ignored? Research Question 1 seeks to explain variation in the levels of public attention and inattention to environmental events. In particular, we focus on how the social structures of attention and identification (March and Olsen 1976, Ocasio 1997) shape industry attention and inattention to specific external events. Research Question 2: When and why do certain events attain high and sustained levels of industry attention? Research Question 2 seeks to distinguish between shortterm levels of industry attention and sustained levels of attention. While many events may receive industry attention at the outset, only a small subset of these events receives continued attention and becomes a critical issue to the industry. Here we will explore the relationships between the process by which the event is enacted over ANDREW J. HOFFMAN AND WILLIAM OCASIO Not All Events Are Attended Equally ORGANIZATION SCIENCE/Vol. 12, No. 4, July–August 2001 417 Table 1 Paired Comparison Event Sample 1. The Cuyahoga River Fire, June 23, 1969, and The Declaration of a Health Hazard at Love Canal, New York, August 2, 1978. 2. The Burmah Agate Oil Spill, November 1, 1979, and The Exxon Valdez Oil Spill, March 24, 1989. 3. The First Earth Day, April 22, 1970, and Earth Day, April 22, 1990. 4. The Publication of Silent Spring, September 27, 1962, and The Publication of The Limits to Growth, March 2, 1972. time (Isabella 1990, Dutton and Dukerich 1991, Barr 1998) and the degree of sustained public attention accorded to an event within an industry. Data and Method Our approach to developing middle-range theories (Eisenhardt 1989, Eisenhardt and Bourgeois 1988) builds on methods of causal inference used by historical sociologists in comparative studies of events (Skocpol and Somers 1980, Ragin 1987, Quadagno and Knapp 1992, Hicks 1994). This research uses historical comparisons primarily for the purpose of making causal inferences about macrolevel structures and processes (Skocpol and Somers 1980, p. 181). Unlike grounded theory approaches (Glaser and Strauss 1967), which are more purely inductive, our comparative case methodology begins with a set of research questions and categories derived from previous theories on organizational attention and identity to draw specific causal inferences and testable hypotheses (Miles and Huberman 1984). Empirical Context of the Study The empirical context for our study deals with the emergence of events related to the natural environment and environmental protection. This is a rich area for research. Over the past thirty-five years, environmentalism has promoted rapid social change and has been propelled by formative, and at times sensational, events (Scheffer 1991, Goetlieb 1993, Hoffman 1997), while other events have received less notice. Central to this rich social history has been the involvement of the U.S. chemical industry. This industry has been singled out in public opinion polls as the preeminent environmental threat from the 1970s (Erskine 1971) through the 1990s (Cambridge Reports/Research International 1992). The volume of the industry’s waste streams exceeds that of the second most polluting industry sector (primary metals) by more than a factor of two (U.S. Environmental Protection Agency 1992). And in general, its role has been prominent in major environmental catastrophes such as Bhopal, Love Canal, and Seveso. Given this centrality in the environmental realm, the chemical industry is a prime candidate for our study. Sample of Events We began by developing a set of environment-related events for study between the years 1960 and 1995, an era marked by many as the “modern environmental movement” (Scheffer 1991, Gottlieb 1993). In the spirit of theory building, we sought variance across the events in our sample (King et al. 1994). These were not, however, meant to be representative of all possible types of occurrences. We first identified an event classification scheme presented by Hannigan (1995) which included milestones, catastrophes, and legal/administrative happenings. Next, drawing from a broad set of event candidates developed by Hoffman (1999, p. 371), we selected individual events from each category in complementary pairs for analysis, based on similar characteristics and attributes. We selected a total of eight events for case comparison. Given the sample size and selection criteria, this sample set may create possible biases if used for drawing inferences. We believe, however, that our sample set and methodology of comparison analysis offers advantages in theory development (Eisenhardt 1989). While it is unlikely that we could develop a sample that represents all possible types of environmental events, we feel that there is more explanatory power in choosing a small number of case comparisons as opposed to a more limited review of a larger number of cases. Descriptive differences uncovered through in-depth comparisons of seemingly similar events should reveal characteristic insights to the attentional processes that guide event enactment and interpretation. We first selected four events that were considered extremely important in the source literature: the publication of Silent Spring (1962); the First Earth Day (1970); Love Canal (1978); and the Exxon Valdez oil spill (1989). We then selected four comparable events for case comparison: the publication of The Limits to Growth (1972); the reenactment of Earth Day on its twentieth anniversary (1990); the Cuyahoga River fire (1969), and the Burmah Agate oil spill (1979). These eight cases are listed in their comparison pairs in Table 1 and briefly described here. (1) The Cuyahoga River Fire, 1969. On June 23, 1969, the Cuyahoga River caught fire for twenty-four minutes, causing $50,000 damage to two key railroad trestles in Cleveland, Ohio. The cause was attributed to oily wastes dumped into the river from waterfront industries and the event has been credited by many as a touchstone for the ANDREW J. HOFFMAN AND WILLIAM OCASIO Not All Events Are Attended Equally 418 ORGANIZATION SCIENCE/Vol. 12, No. 4, July–August 2001 genesis of the modern environmental movement (Opheim 1993). (2) Love Canal, 1978. In 1976, residents of a neighborhood of Niagara Falls mobilized to demand government action in investigating and remedying the appearance of chemical wastes in their neighborhood. On August 2, 1978, the New York Department of Health declared the area a health hazard and, with aid from the federal government, began buying homes and evacuating their occupants. It was determined that the neighborhood had been built on and around an abandoned waste site into which the Hooker Electro-Chemical Company had buried 21,800 tons of chemical waste from 1942 until 1953. (3) The Burmah Agate Oil Spill, 1979. On November 1, 1979, the freighter Mimosa rammed the Liberian tanker Burmah Agate while anchored off the port of Galveston, Texas. The ship and its leaking cargo burned for five days, out of which leaked 10.7 million (U.S.) gallons of crude oil. A large portion of this oil washed up on the beaches of Galveston over the ensuing weeks. (4) The Exxon Valdez Oil Spill, 1989. On March 24, 1989, the Exxon oil supertanker Exxon Valdez ran aground on Bligh Reef in Prince William Sound in Alaska. In all, 10.8 million (U.S.) gallons of crude oil coated about 1,200 miles of shoreline. Wildlife loss and charges of mismanagement resulted in unprecedented legal judgments against Exxon reaching over $5 billion, with many cases still pending. (5) The First Earth Day, 1970. On April 22, 1970, nearly 20 million Americans took part in a national event on college campuses all over the country. Festivities focused public attention on the mounting awareness of environmental degradation and targeted much of their protest against corporations. For many, this event marked the coalescence of a new “environmental movement” which involved constituents from all of society (Gottlieb 1993). (6) Earth Day, 1990. On April 22, 1990, Earth Day was reenacted on its twentieth anniversary. An estimated 200 million people participated in 140 nations. But, through funding of the days events and staging of special demonstration of their “green” activities, corporations were not villains, but prominent participants and organizing supporters. (7) The Publication of Silent Spring, 1962. As discussed in the introduction, on September 27, 1962, the Houghton-Mifflin Company published the book Silent Spring. Rachel Carson, the book’s author, charged that widespread application of the pesticide DDT and other synthetic chemicals was disrupting the “web of life,” posing a hazard to all living organisms, including humans. (8) The Publication of The Limits to Growth, 1972. Released on March 2, 1972, The Limits to Growth (Meadows et al. 1972) documented the results of a computer study based on the system dynamics model perfected at MIT. It concluded that mankind faced an uncontrollable and disastrous collapse within 100 years unless it moved speedily to establish a “global equilibrium” in which growth of the population and industry output were stabilized. Data Sources To capture data on event attention in the U.S. chemical industry in this study, we focused on how public attention is situated within one particular arena or communication channel which we see as central to these processes, the business press. While the business press is one among many public arenas within an industry, it offers some important advantages. Trade journals are one of the most critical communication and procedural channels through which industry attention is structured, providing both analyses of events and issues, and instruction to their readers on their relative importance (Clinton 1996). Research on the impact of trade journals shows that their structural position as a shared reference for knowledge transfer among industry constituents (Nederhof and Miejer 1995) makes them both a channel of communication in the early stages of industry-related policy process (Hollifield 1997) and a common reservoir for available information and interpretations. As such, the roles they play in attentional processes are multiple. First, they act as a common source of information, creating a historical record relevant to their readership based on both insiders’ and outsiders’ interpretations of data. Second, they act as an internal constituent of an industry, deciding which events to attend to and offering analysis and interpretation of their criticality to their readership. They are a dual force “for socialization of the young and attitude change in the old” (Webb et al. 1966, p. 78). Third, they act as conduits to other communication channels and public arenas. Trade journals actively scan other public media for their coverage of industry issues and events, recording outsiders’ accounts of industry activities and industry reputation, and thereby serving as linkages between outsiders’ and insiders’ public attention. The limitations in this data source are clear as well. Trade journals can be active agents engaged in processes of impression management, both by design and by cultural bias. They are organizational actors whose output is subject to the political pressures exerted by powerful figures and organizations within the industry (Molotch and ANDREW J. HOFFMAN AND WILLIAM OCASIO Not All Events Are Attended Equally ORGANIZATION SCIENCE/Vol. 12, No. 4, July–August 2001 419 Lester 1975). And journal coverage itself is the product of a fixed system (Schlesinger 1978) inhabited by individuals who are pre-selected for their biases toward the journal’s constituency, in this case industry. Trade journal coverage is, by definition, a biased interpretation of events and issues, where the bias is likely to reflect the interests and identities of its core readers and sources of information (Molotch and Lester 1975). The culture and social structure prevailing within the industry shape its content. As a source of our data set, we reviewed event coverage in the trade journal, Chemical Week, and supplemented that analysis with coverage reviews in the newspapers, the New York Times and the Wall Street Journal. Chemical Week was our internal industry source, representing a reasonable indicator of the interests, identity, and perspectives of the chemical industry as they react to external events. The New York Times and Wall Street Journal were our external sources. The New York Times was used to represent externally situated perceptions of these events located within the general public, and the Wall Street Journal was used to represent perceptions within the financial community. In each case, the level and content of media coverage was analyzed to draw inferences about what issues and events were being addressed by the media constituency, as well as to what extent and through what types of interpretation and presentation. Chemical Week was used to represent internal industry attention to external events. As our primary source, Chemical Week was one of several trade journals available. Two other prominent journals—Chemical & Engineering News (C&EN) and Chemical Engineering (CE)—serve this industry sector as well. As our rationale for choosing, we found Chemical Week to have coverage that was specialized to the interests of the chemical industry. Both C&EN and CE serve both the chemical and petroleum industries and C&EN targets academic and governmental audiences in its readership. Given this dilution in constituency, Chemical Week stood out as a central dedicated communication channel within the U.S. chemical industry. Data Collection and Analysis Our process of identifying and coding articles describing our event sample differed slightly by data source. The process ranged from initial broad-scale screening to final analysis of specific article content. For the New York Times and the Wall Street Journal, we began with the “Year in Review” compendia for each journal and collected gross data on number of articles covering a particular issue, including the date, page, and title of the article. Then each article was reviewed for content to uncover clues about the enactment of the event. Based on our theoretical interest on the social structures of attention (March and Olsen 1976, Ocasio 1997), we focused on particular players or constituencies mentioned, experts cited, companies blamed or praised, differing degrees and types of blame assessed, data presented, etc. For Chemical Week, we also used the “Year in Review” compendia and the Business Periodicals Index to identify and code the number of specific articles on a particular event. However, in this data we also went deeper, scanning the journals themselves for a more accurate and complete review of articles of importance. We began our review six months before the event and ended three years after the event, searching for similar content as with the New York Times and Wall Street Journal. With such content data, we began our analysis. We used the number of Chemical Week articles as a measure of industry-level public attention. We used New York Times and Wall Street Journal coverage as a measure of public attention by outsiders. These measures implicitly treat attention as a discrete activity, with each article as a separate occurrence of attentional processing within an industry’s communication channels (Ocasio 1997). The measures of attention and interpretations of industry events gathered from press publications were complemented with data and observations obtained from secondary sources (Erskine 1971, Evernden 1993, Scheffer 1991, Schmidheiny 1992, Gottlieb 1993, Cairncross 1995, Hoffman 1997). Case Comparison Method We began our analysis by undertaking between-case comparisons of industry-level attention based on existing theory, seminar discussions, and pilot tests on other types of events. We proceeded to review event coverage in relation to our developing model of event attention within the U.S. chemical industry. Our analysis was guided by a metatheory derived from research on attention and identity, as described above. We relied on an analytical process that combines induction with deduction (Miles and Huberman 1984) to develop an explicit model of how the industry structures attention to nonroutine occurrences. We used data reduction and data display methods to draw and verify our conclusions. The analytical process that followed involved repeated iterations, moving back and forth between our emerging model and the quantitative and qualitative data. Through successive iterations, we converged on a final model that best fit the empirical data and provided a coherent theoretical explanation of the industry-level attentional process. In addition to the theoretical model, we proceeded to develop a set of hypotheses inferred from the case observations. ANDREW J. HOFFMAN AND WILLIAM OCASIO Not All Events Are Attended Equally 420 ORGANIZATION SCIENCE/Vol. 12, No. 4, July–August 2001 Table 2 Industry Attention to Environmental Events Event Silent Spring Cuyahoga River Fire Earth Day Limits to Growth Love Canal Burmah Agate Exxon Valdez Earth Day Anniversary Date 1962 1969 197