A Framework for Information Systems in Life-Cycle-Oriented Environmental Management

A Framework for Information Systems in Life-Cycle- Oriented Environmental Management Teresa M. Shaft College of Business Administration University of Tulsa Tuka, OK, USA

Rex T. Ellington University of Oklahoma N m n , OK, USA

Mark Meo Science and Public Poky Program University of Oklahoma N m n , OK, USA

Mark P. Sharfman College of Business Administratiun University of Oklahoma Norman, OK, USA

- K t E p V O d S

environmental management LCA information systems life-cycle management LCOEM life-cycle assessment

I Summary

Although business firms have improved their environmen- tal performance, a variety of forces are pushing businesses toward adopting environmental management throughout the entire life cycle of their products and processes. In this

Address comspondence to: Teresa Shaft University of Tulsa College of Business Administration 600 South College Avenue Tulsa, OK 74101 ahaftm~entum.utulsa.edu

0 Copyright 1997 by the Massachusetts Institute of Technology and Yale University

article we discuss the information systems elements of an environmental management approach we call “life-cycle- oriented environmental management” (LCOEM).This ap- proach requires the firm to manage the effects of its pro- cesses from the creation of inputs to the final disposal of outputs, that is, from cradle to grave.We present a frame- work of the classes of information systems needed, de- scribe their use in an LCOEM setting and define their interrelationships. We conclude with a discussion of the implications of LCOEM information systems.

Volume 1, Number 2 I

journal of Industrial €cology I35

1 APPLICATIONS

New . . . information flows will have to be developed as corporations internalize en- vironmental issues . . . . It is clear that de- velopment of appropriate . . . information subsystems is critical to completing a nec- essary feedback loop for environmentally appropriate behavior by corporations. (Graedel and Allenby 1995,87,90)

In response to the increasingly competitive nature of the marketplace, businesses must im- prove the quality of their goods and services while simultaneously reducing costs. Coupled with these competitive forces are pressures on firms to be better environmental citizens. By re- ducing their negative impact on the natural en- vironment, firms can cut costs, improve product and service quality, and use good environmental citizenship for competitive advantage.

To achieve these goals, some firms are adopt- ing what we call “life-cycle-oriented environ- mental management” (LCOEM).’ This approach to environmental management prevents pollu- tion by focusing on more efficient use of more benign resources. Through LCOEM, managers consider the totality of their firm’s effect on the environment from cradle to grave-that is, from the creation of basic inputs to production of products to final disposal and decontamination of outputs. As Graedel and Allenby point out above, information systems (1%) will play an in- creasingly important role in corporate environ- mental management. This is especially true for LCOEM because it is particularly information intensive, requiring detailed information on pol- lutants, emissions, resource usage, material char- acteristics, production alternatives, government regulations, and new technologies. In this article we discuss the information systems that are re- quired for LCOEM.

The information for LCOEM will come from within the “focal firm” as well as from suppliers, customers, and external data banks. By “focal firm,” we mean the firm that is defining the product chain or life cycle (i.e., the firm that defines who is a customer and who is a supplier). Any firm can be the focal firm; it is a matter of perspective. Internal and interorganizational ISs will be necessary to process the information re-

quired to establish and maintain LCOEM. This article defines the framework for these ISs.

Before discussing LCOEM ISs, it is necessary to define terms. An LCOEM approach has three necessary components: (1) an understanding that organizations have effects on the natural environment through all stages of the product life cycle; (2) a coincident reduction in resource utilization through more efficient production, design, and waste management; and (3) coordi- nated efforts of all parts of the organization, its suppliers, and its customers (Ellington et al. in press; Sharfman et al. forthcoming).

Our second task is to address the difference between life-cycle assessment (LCA) and life- cycle-oriented environmental management. LCA, in its various forms, is an analytic tool. Specifically, LCAs are techniques for assessing the effects of some product or process through- out its life cycle. LCOEM is a management ap- proach that is based on what we and others call “life-cycle thinking” (Henn and Fava 1993). LCOEM includes LCA but also such activities as design for the environment, product take- back, and ecolabeling (Graedel and Allenby 1995) plus accompanying organizational changes. In LCOEM, as firms examine the envi- ronmental impact of their products’ life cycles, they adopt processes and structures to manage that impact. Key to the maintenance of an LCOEM approach are the ISs necessary to sup- port the approach.

As argued elsewhere (Ellington et al. forth- coming; Sharfman et al. forthcoming), firms are under increasing pressure to manage their envi- ronmental “footprint” beyond their own bound- aries toward the entire product chain life cycle. The pressures on firms to adopt LCOEM prac- tices come from such disparate sources as the Dutch regulations on LCA, U.S. government regulations on chlorofluorocarbon (CFC) label- ing, German product take-back laws, and the provisions of what will become IS0 14041 (the standards for life-cycle assessment). Further, several leading-edge firms such as Baxter Healthcare, DuPont, and Interface Carpeting raise the standard for others in their industries by using life-cycle approaches for competitive advantage.

I36 lournal of Industrial Ecology

APPLICATIONS 1

The Importance of Information and Information Systems in LCOEM

As firms move toward LCOEM, the need to consider environmental issues throughout the product chain will create new intra- and interorganizational information flows. At mini- mum managers will face a need for increased data collection and analysis in the following ar- eas: resource utilization (physical and energy), waste-stream monitoring (solid, liquid, and gas- eous), and internal and interfirm component flows. If the incentives for increased environ- mental quality vigilance are regulatory, then firms are likely to face increased compliance re- porting. If incentives for LCOEM are market- based, managers are likely to face increased cost and pollutant accounting requirements. These new information flows will require ISs for LCOEM to be effective.

The idea that firms will create new ISs in re- sponse to the organizational change required by LCOEM is consistent with Markus and Robey’s (1988) “organizational imperative.” This prin- ciple characterizes the relationship between in- formation technology and organizational change as one in which ”human actors design informa- tion systems to satisfy organizational needs for information” (1988,587).

The creation of new 1Ss for LCOEM is par- ticularly important because of two data-related challenges. First, because environmental data by their nature are ambiguous and equivocal, wide disagreement exists even in the scientific com- munity about the dynamics of the environment, whether current conditions represent a “prob- lem,” and if there are problems, what the best solutions are (Bruntland 1987). These disagree- ments arise when environmental data are inter- preted differently, particularly when data characterizing the temporal, spatial, or func- tional dimension of one element in the environ- ment are mismatched with that of another element that is substantively different (Lee 1993). One of the sources of ambiguity in envi- ronmental information comes from errors in in- terpretation or analysis based on these mismatches. A historic example of this kind of error is found in the Colorado River Compact of

1922 in which long-term water allocations were mistakenly based on a (then) recent, anoma- lously wet period of years (Brown 1988). In this case, errors in water allocations occurred be- cause the political planning horizon did not match the length of the flow trends in the river. The implications of the misallocations based on that error are still being felt today. Perhaps to remedy this problem, firms may have to increase their internal, scientific capacity so they are pro- cessing their own data rather than that of others. In any case, to use environmental information effectively, firms will need to develop ISs that can better handle such ambiguity.

Second, the sheer volume of data we have mentioned will require new systems that have both the speed and the processing power to handle the amount of data. Also, the volume of data be- ing processed will necessitate that ISs present the data in ways that do not overwhelm users.

A Framework for LCOEM Information Systems

The adoption of LCOEM will entail a variety of changes and additions to a focal firm’s ISs. Some of the changes will be extensions of exist- ing systems, others will be as complex as the de- velopment of electronic information interchange (EII), where firms will interchange information (as opposed to raw data). Our approach is predi- cated on the idea that different types of ISs must work together to meet a firm’s LCOEM informa- tion-processing needs. We suggest that an effec- tive LCOEM approach will require a set of ISs including office automation, transaction process- ing, management information, decision support, and executive information systems. Expert sys- tems, a sixth type of IS, are expected to play a role eventually in LCOEM; however, a require- ment for the development of expert systems is that “(t)here are recognized experts that solve the problem today” (Prerau 1985,27). LCOEM is considered too new for human expertise to have developed sufficiently for expert system imple- mentation (Graedel and Allenby 1995).’ Hence, expert systems are not included in our framework. Our framework below defiies each type of IS, dis- cusses its role in LCOEM, presents an environ- mental example, and describes necessary

Shafr, Ellington, Meo, and Sharfmon, LCOEM Information Systems 137

I APPLICATIONS

extensions to support LCOEM. Note that the boundaries between systems are imprecise, and the specific capabilities of each type of system vary among organizations. At times, capabilities can overlap different systems. Following our dis- cussion of the types of ISs, we discuss their inter- relationships.

Office Automation Systems

Definition Office automation systems (OASs), some-

times known as office information systems, sup- port information workers (including professional staff and clerical workers). OAS capabilities typically include some combination of the fol- lowing: electronic and voice mail, word process- ing, computer and video conferencing, facsimile transmission, videotext, imaging, and desktop publishing.

Purposes and Tasks in LCOEM Many OAS capabilities are communication

oriented, and it is here that OASs will play a vi- tal role in LCOEM, supporting intra- and interorganizational information. Because tradi- tional “paper and pencil” systems would be swamped by the volume of communications that an LCOEM approach entails, OASs will be in- tegral to the success of LCOEM.

Data Inputs OASs are relatively informal ISs in that there

are few fixed inputs or outputs. Particularly prior to formalization of the interorganizational stan- dards required for electronic data interchange (EDI) and EII, the flexible nature of the OAS will provide product chain members with a means of communication. After the standards for ED1 and EII are developed, the standardized interorganizational information flows will be supported by other types of systems. The OAS will continue to serve an important role, how- ever, by supporting the informal communication among members of the product chain.

outputs The outputs of the OAS are relatively infor-

mal such as memos and electronic mail. Firms may wish to consider “groupware” and “netware,”

that is, software products that facilitate group- level communication and support electronic meetings and discussions.

Example Given the informal nature of an OAS’s out-

puts, we do not anticipate that specific LCOEM OASs will exist. Their role is to support the flow of interorganizational information among the member firms.

Transaction Processing Systems

Definition Transaction processing systems (TPSs), also

known as data-processing systems, support the day-to-day information-processing needs of an or- ganization. The TPS processes the most detailed data in an organization. Its primary requirement is to support the basic business functions of an orga- nization, that is, both internal and external trans- actions. Internal transactions typically include such activities as producing paychecks and payroll reports. External transactions include producing sales receipts and ordering, then paying for, raw materials and inventory. Increasingly, TPSs sup- port external transactions with EDI-that is, the computer-to-computer exchange of formatted data to support basic business transactions (e.g., ordering inventory). In traditional ISs, TPSs have two unique characteristics. First, TPSs perform necessary tasks (i.e., by law a firm must maintain a record of its activity). Second, historically TPSs have been the only ISs that interact with external business elements. To support LCOEM, however, all other types of ISs are likely to develop interorganizational functions.

Purposes and Tasks in LCOEM In terms of LCOEM activities, TPS will

gather data on resource utilization as well as on gaseous, liquid, and solid emissions from the fo- cal firm. These data will be used for internal management and planning plus external compli- ance reporting. The TPS will also monitor re- source uses and emissions from members of the product chain. The transmission of TPS data from product chain members to the focal firm will, for the most part, be voluntary (and there- fore potentially inconsistent). I t will be impor-

I38 journal of Industrial Ecology

A P P L I CAT I0 NS

tant to integrate as many of these data into the focal firm’s TPS as possible, however, to provide a complete picture of the product chain’s effect on the environment.

Data Inputs The data inputs for the TPS will come from

the focal fiim and the other product chain mem- bers. The TPS will require input at the start-up of the LCOEM approach to determine the data that must be gathered to support LCOEM activi- ties, to ascertain reporting requirements, and to develop reporting formats. Inputs include data from what we call “environmental transactions”; the more obvious environmental transactions include releases of pollutants, data from interac- tions with recyclers, and expenses for site remediation or cleanup. A second category of environmental transaction data comes from the financial transactions connected to environ- mental management. Under the rubric of “green” accounting (environmental cost ac- counting), firms are developing mechanisms for collecting data on environmental financial transactions, including such costs as training, on-site waste treatment, and reengineering for environmental goals (U.S. EPA 1995a).

outputs The outputs of the TPS will include summary

totals and breakdowns of resource utilization in- cluding energy, emissions, and environmental fi- nancial transactions from the focal firm and members of the product chain. The TPS will also generate data for use in regulatory compli- ance reports (see below). Finally, the TPS will support ED1 as a reporting mechanism between the focal firm and its suppliers, customers, and product disposers/recyclers. As the amount and complexity of emissions, resource use, and com- pliance data increases, the focal firm and the product chain members are likely to increasingly rely on ED1 to provide data in a format that all parties can use. Key to effective ED1 among the member firms will be common data formats and definitions (U.S. EPA 1995a).

Example TPSs provide the most common example of

an extant environmental management IS. Any

manufacturing firm in the United States that employs ten or more workers and meets mini- mum materials use standards must complete the U.S. Environmental Protection Agency’s (U.S. EPA) Toxic Release Inventory (TRI). By law, each release of any of 316 chemicals or 20 chemical categories to the air, water, or land, or transfers of these chemicals off-site for disposal must be recorded. These data are reported annu- ally to the U.S. EPA. Large users of such chemi- cals cannot manually tabulate the TRI. To extend the TRI to an LCOEM environment will be a process of collecting and integrating data from each product chain member’s TPS.

Management Information Systems

Definition Management information systems (MISs)

support the general information needs of manag- ers with a focus on providing the information to support the control function in an organization. MISS typically furnish managers with regular, routine reports to monitor the status of an orga- nization. These reports typically fall into one of two categories: (1) periodic reports, generated according to a time-based schedule (e.g., monthly, quarterly, etc.), or ( 2 ) exception re- ports, generated only when a variable exceeds a predefined limit (such as when emissions are above designated levels). Because MISS are re- porting and control oriented, they focus more on providing the information to identify potential problems and less on solving problems.

Purposes and Tasks in LCOEM MISs will be essential to LCOEM in support-

ing required monitoring and control. If an LCOEM standard has not been met, the MIS could help identify the source of the problem(s). Within the focal firm, the MIS will monitor per- formance. With other members of the product chain, the information from the MIS can be used in advisory ways to assist in problem iden- tification and performance monitoring, includ- ing liquid, solid, and gaseous emissions from all processes, levels of resource utilization, degrees of regulatory compliance, training and aansac- tion costs between and among product chain members, and so on.

Shaft, Ellington, Meo, and Sharfman, LCOEM Information Systems 139

1 APPLICATIONS

Although the traditional role of MISS is in- ternal, as a result of the tight interorganizational relationships required by LCOEM, MISS will serve a new interorganizational role. Unless con- tractual relationships develop, the interorgan- izational monitoring role of the MIS will be advisory; however, given that the focal firm will most likely have the best developed monitoring tools, its MIS could provide valuable informa- tion to other member firms.

Data Inputs At start-up, data from both the focal firm and

members of the product chain will be necessary to help develop the myriad of performance reports the LCOEM approach requires. Some perfor- mance standards may be set externally via regula- tions, while others will be the result of decision support systems-based analysis, and still others will be developed through consensus among the members of the product chains. Once standards are established, monitoring data will be necessary from all phases and members of the product chain. Much of these data will come from ongo- ing LCA efforts by all product chain member firms. Presumably, as the firm and the other chain members institutionalize LCA, LCOEM, and the attendant ISs, the TPS will collect LCA data rou- tinely and feed it into the MIS for analysis and/or report generation. From the MIS these data likely will be inputs to the DSS.

outputs The primary output of the MIS will be re-

ports concerning the environmental perfor- mance of the focal firm and other members of the product chain. These reports serve as inputs to the DSS, EIS, and strategic planning. Sec- ond, via exception reports, the MIS spotlights problems and variations from standards either by the focal firm or by members of the product chain. The third major output from the MIS will be EII. It is likely that chain members, including the focal firm, will share processed information. Given the anticipated volume of information, standard paper reports will be inefficient. Rather, members of a product chain could have access to other member’s computers and pro- cessed information. The EII concept is similar to

EDI, which automates transactions between firms. El1 differs from ED1 efforts in that EII sup- ports the exchange of information, as opposed to transaction data.

Example The 3M Company has integrated pollution

prevention (PP) into total quality management (TQM) to form a PP/TQM effort. The objective of the effort is to reinforce the concept that each employee’s job responsibilities include reduction of environmental waste (PCEQ 1993). As part of this effort an MIS computes a “waste ratio” that is reported to top management on a quar- terly basis. The regular reporting of the waste ratio allows management to monitor progress on eliminating waste from processes and to recog- nize teams or individuals for their achievements toward this objective. 3M’s goal of eliminating waste is consistent with an LCOEM approach. To fully implement an LCOEM approach, how- ever, 3M would have to be able to monitor (di- rectly or indirectly) the waste levels of other members of the product chain.

At present, no environmental EII examples exist, although we see models for EII in other fields; for example, semiconductor manufactur- ers and suppliers exchange computer-aided de- sign (CAD) and computer-aided manufacturing (CAM) specifications, production control and inspection/reliability information based on pho- tolithographic “masks,” that is, “hardcopy of the CAD circuit design,” provided by suppliers (Hart and Estrin 1991, 378). While El1 is more complex than EDI, Hart and Estrin (1991) pro- vide insight into how it can be accomplished us- ing interorganizational networks. LCOEM will require similar levels of information exchanges.

Decision Support Systems

Definition Decision support systems (DSSs) support the

problem-solving needs of individual or small groups of managers engaged in analytical deci- sion making for the firm. DSSs are narrowly fo- cused and therefore produce specialized reports on an as-needed basis. A problem-solving ses- sion with a DSS is typically user-initiated and

I40 Journal of Industrial Ecology

APPLICATIONS 1

interactive (as opposed to time- or exception- based batch processing). Given the specialized nature of DSSs, capabilities vary greatly from system to system.

Purposes and Tasks in LCOEM DSSs will be important to LCOEM to sup-

port the numerous decisions (e.g., choices among production processes, inputs, etc.) requir- ing the consideration of a number of variables, including environmental impact criteria. DSSs will be the locus for establishing the baselines and standards by which the entire LCOEM ap- proach will operate. These baselines and stan- dards likely will be developed through the use of LCAs. In LCOEM, the most valuable support will come in the form of quantitative models to analyze the environmental impact of the total product chain. Quantitative models will be use- ful in determining at what pollutant levels the firm and the product chain should operate. Such decisions consider many alternatives and mul- tiple criteria (e.g., improvements with respect to one type of waste may yield inferior results with respect to a different type of waste). For in- stance, methods like Ellington et al.5 (1992) approach to estimating greenhouse gas emissions could be incorporated into DSS. Of all the types of systems, the DSS will perhaps be the most useful in that it enables managers to interpret the complex data that must be considered.

Data Inputs To help set standards for the LCOEM, the

DSS will need data from supplier firms, the focal firm’s own input, transformation (e.g., produc- tion) and output processes, its customers, and disposers and recyclers of any products. Once standards are developed, the DSS will require access to internal and external data including data from the focal firm, suppliers, customers, and any product disposers/recyclers.

outputs The primary outputs from the DSS will be

analyses of factors in the business, regulatory, or social environments that are salient in the opera- tion of the LCOEM process. The DSS will also produce analyses of the LCOEM process as input

to the strategic planning process. Quantitative models in DSS will be used to set standards (e.g., emission standards for production process, re- source utilization levels, etc.) and to compare costs of various pollution prevention options.

Example Several organizations have developed DSS

software tools that will fit into LCOEM ISs. Batelle Memorial Institute developed life-cycle analysis and design (LCAD) to aid designers conducting in-depth life-cycle inventories of any product or proce~s.~ The Sandia National Lab has developed EcoSys, which also performs life-cycle inventories.‘ EcoSys focuses on identi- fying the relative environmental effects of vari- ous choices in the production process. These tools would have to be enhanced to include data from other members of the product chain to fully support LCOEM.

Executlve lnformation Systems

Definition Executive information systems (EISs) sup-

port the information needs of firms’ highest level executives. Characteristics of EIS include pro- viding online access to both internal and exter- nal information. Features of EISs include the ability to extract, filter, and compress data and to present data in a variety of formats (i.e., text, graphical, or tabular). Two unique qualities of EISs are the ability to “drill down” and the abil- ity to deal with both “hard” and “soft” data. Drilling down enables executives to start at sum- mary data, then move down through multiple data levels to access more detailed underlying information. The focus on both hard and soft data gives executives access to data from existing databases (hard data) and supports the need to collect and store less formal (soft) data, such as rumors, news, ideas, and so on. Soft data can re- quire special data collection efforts and can be labor intensive to incorporate into a system.

Purposes and tasks in LCOEM LCOEM is a strategic choice, sometimes

made as a means to gain competitive advantage (Graedel and Allenby 1995). The purpose of the

Shaft, Ellington, Meo, and Sharfman, LCOEM lnformation Systems I 4 I

1 APPLICATIONS

EIS within LCOEM is to provide online access to data in an organized, easy-to-use fashion for executives tracking their firm’s achievements with respect to environmental issues. EIS will provide ways to blend the managerial and tech- nical information necessary to successfully implement LCOEM. Managerial information would include hard data concerning relevant costs associated with pollution control and those associated with design for environment (i.e., process redesign to lower environmental im- pact). Soft data might include the ability to monitor political trends concerning environ- mental regulation or to assess the degree of co- operation from other product chain members. To obtain this type of data, a firm may subscribe to online services coupled with a “clipping” function to identify social, legal, or regulatory news salient to the firm. The EIS could assist top management in tracking data from these various sectors and in preparing new LCOEM plans.

Data Inputs The EIS will require data from all elements of

the business, regulatory, and social sectors to help managers ensure that strategic and LCOEM plans are consistent with the organization’s mis- sion, societal norms, and laws. The EIS will re- quire access to both internal and external data sources including data from all members of the product chain.

Examples of external data sources an EIS might access are integrated data for enforcement analysis (IDEA) and hazardous solvent distribu- tion data system (HSDDS), both of which are supported by the U.S. EPA. IDEA contains compliance and enforcement data including the Comprehensive Environmental Response Com- pensation and Liability Act of 1980 (CERCLA) (“Superfund”) Information System and TRI data. Accessing IDEA allows managers to track enforcement and compliance activities by the US. government. HSDDS gives managers ac- cess to current information on solvents, their use and disposal.

outputs Given their process nature, EISs are more

outcome than output oriented. Results of an EIS session could include new strategic objectives

regarding LCOEM. For instance, an executive might discover an unacceptable level of emis- sions. Drilling down to detailed information, the executive could pinpoint the cause of the prob- lem as a particular production process. Search- ing soft data stored by the EIS, the executive could locate information regarding an engineer- ing breakthrough (e.g., the use of different mate- rials or alternative processing). Coupling the information about the source of the problem with knowledge of the engineering break- through, the executive could order a demonstra- tion project.

Example Johnson and Johnson’s emergency manage-

ment software emergency information system/ chemicals (EIS/C) is an example of a system that integrates data from several sources to support managers (Ebbs 1991). EIS/C focuses on infor- mation needed to handle emergencies due to chemical spills. It combines an electronic map- ping function, data, and communications tech- nology. EIS/C can access regulatory, chemical, and emergency response data about hazardous materials by drawing on Johnson and Johnson’s PC-based regulatory environmental, chemical information system. EIS/C can collect real-time meteorological data and predict dispersion pat- terns for airborne chemicals and display them on the local area maps. Combining these features into a single system allows executives and local emergency management authorities access to the information to manage a chemical emergency.

Extending this system to an LCOEM setting would entail including location and chemical data plus emergency response information for other product chain members. When emergen- cies occurred at product chain members’ facili- ties, this information would be integrated with other ElS to advise the appropriate units (e.g., production, purchasing, etc.).

herrelationships among the Types of Information Systems

We present the interrelationships between the five types of ISs independent of LCOEM in figure 1, adapted from Laudon and Laudon (1994). Following this overview, we discuss the

142 journal of Industrial Ecology

APPLICATIONS 1

-u

interrelationships to support LCOEM. The TPS is responsible for gathering data about basic transactions, therefore it does not receive input from the other types of ISs, but generates infor- mation for the other ISs for problem solving, re- porting, and communication functions.

The OAS, because of its flexible nature, re- ceives input from the EIS, DSS, MIS, and TPS. Such inputs might include new strategic objec- tives derived from an EIS, analysis from a DSS problem-solving session, results of reports gener- ated by the MIS, or summarized data created by the TPS. Each of these inputs could be incorpo- rated into memos or e-mail. The OAS can out- put information to the DSS for use in a problem-solving session (e.g., solution param- eters regarding achieving a certain percentage reduction in emissions).

The MIS receives input from the TPS, mainly in the form of detailed data concerning transactions, which is transformed into informa- tion that is useful for monitoring performance. The information the MIS generates can be out- put to the EIS, DSS, and OAS. The information could identify a problem that stimulates an in- vestigation using an EIS, could initiate a DSS problem-solving session, or could be dissemi- nated throughout the firm via the OAS.

The DSS receives input from the OAS, MIS, and TPS for use in problem-solving sessions. The DSS output information to the EIS and OAS. DSS output could be beneficial to an ex- ecutive using an EIS to consider strategic alter- natives. The OAS could disseminate the results of a DSS analysis.

Figure I The interrelationships between the five basic types of information systems.

The EIS receives input from the MIS and DSSS The MIS provides information that ex- ecutives may access via drilling down with an EIS. The DSS provides analytical results. The EIS outputs information to the OAS for commu- nication and dissemination.

When a firm adopts an LCOEM approach, the relationships among the different types of ISs be- come more complex as a result of the interorgan- izational relationships that must be supported. We believe that the interorganizational ISs links will evolve from a small set, creating a loosely coupled interorganizational relationship (figure 2a) to a larger set, creating a more tightly coupled relation- ship (figure 2b). For simplicity we do not represent the systems for each non-focal firm; rather we rep- resent them in aggregate.

The loosely coupled relationship will exist early in the adoption of LCOEM. At this stage, the OAS of the focal firm and other firms in the product chain will be capable of two-way commu- nication. An OAS’s flexible nature gives it the ability to support informal information: therefore, this should be the first interorganizational link established. This interorganizational link can be established easily via commercial sources. Alter- natively, with the rise in intranets, a firm could extend nodes of its intranet to product chain member firms.

The other interorganizational link that must be established to achieve LCOEM is between the TPS of the focal firm and the TPS of the other members of the product chain. This link can be supported by a one-way relationship (from mem- bers to the focal firm). The simple ability for the

Shaft. Ellingcon, Meo, ond Shorfman, LCOEM Information Systems 143

A P PLI C AT1 0 N S

focal firm to receive data from member firms’ TPS will provide access to the data needed to monitor the product chain. However, the focal firm must be prepared to bear the brunt of the information processing in this loosely coupled configuration; that is, the basic data will require additional pro- cessing to create meaningful information for use in monitoring and decision-making activities. Two-way OAS communication and one-way TPS communication require fairly simple technologi- cal support and are not difficult to implement. We believe that more extensive interorganizational linkages will develop, however, to create more participatory information processing among members of the total product chain.

Figure 2b represents the more evolved state of interorganizational linkages. The two-way

Figure 2a The loosely coupled interorganizational configuration.

Figure 2b The tightly coupled interorganizational configuration.

link between the OASs remains. Note that the TPS of the focal firm and the member firms are connected via a two-way link, giving all mem- bers of the product chain the ability to send and receive data. All EISs, DSSs, and MISs also are linked via respective two-way connections, thus shifting the burden of information processing from the focal firm to a more equitable distribu- tion throughout the product chain. For instance, the focal firm may access reports and informa- tion prepared by member firms’ MISs.

In addition, DSS results from a member firm can be used by the focal firm (and vice versa), instead of recreating the results using TPS data. For instance, when considering different alter- natives using LCA, members of the product chain must consider the impact of alternatives at

144 journal of lndustriol Ecology

APPLICATIONS 1

different stages in the life cycle (i.e., environ- mental “savings” in one stage of the product chain might result in “costs” in other stages). It is unlikely that a single firm would have the re- sources, or knowledge, to develop models that properly analyze the impact of different altema- tives for all members of the product chain. In- stead, it would be more efficient to use models that reside in member firms’ DSSs (e.g., models of member firms’ own processes).

Similarly, the focal firm can access member firms’ EISs (and vice versa), allowing firms di- rect access to tools that help identify the cause of a problem rather than having to wade through TPS data to pinpoint the reason. For example, it would behoove the focal firm (in fact all mem- bers of the product chain) for other firms to have access to industry scanning systems concerning expected changes in regulations, laws, industry standards, or social norms. When a single firm has access to such information, that firm must convince other member firms when action is re- quired. Instead, all firms would have access to industry scanning information and are more likely work in unison as increased information exchange leads to increased coordination (Weitz et al. 1994).

Figure Zb, using dashed lines, depicts one- way links from member firms’ ISs to the focal firm. Based on the “within firm” relationships depicted in figure 1, we model similar relation- ships among different types of ISs between firms. For instance, an information flow from DSS to TPS is not modeled within a firm. Therefore, we suggest that an interorganizational flow from a DSS to TPS is not needed. Similar reasoning dictated the inclusion or exclusion of other interorganizational information flows. Establish- ing these links will require greater coordination and cooperation among the firms than is re- quired by the loosely coupled configuration. Standards must be agreed to regarding nearly every aspect of these systems. Also, these links imply a peer relationship between firms since all firms have equal access to each others’ 1Ss. This tightly coupled configuration is only possible when members of a total product chain have a great deal of trust in the other members of the product chain or a great deal invested in the product chain (likely both).

In depicting these two levels of interrelation- ships we are not implying an either-or scenario. Instead, they depict two ends of a continuum. The loosely coupled configuration represents the minimum linkage required to implement LCOEM. The tightly coupled configuration il- lustrates the maximum connection that may evolve. Although the tightly coupled LCOEM ISs may be the goal, for the forseeable future, it is likely that firms will implement something less ambitious.

Discussion

While the framework organizes the informa- tion-processing tasks of LCOEM, several impli- cations of adopting the approach must be considered. Developing any IS can be difficult; creating the systems we have described requires interorganizational cooperation, resulting in ad- ditional challenges. The first difficulty is stan- dardization. If systems are developed from scratch, standards must be created, then imple- mented. More likely, and more challenging, ex- isting systems will be modified to meet the needs of LCOEM. Questions arise such as: What (whose) standard will be used? How will the ef- fort be distributed across the member firms?

Even as standardization problems are solved, the issue of the size and scope of LCOEM ISs must be addressed. Developing systems that pro- cess data and provide information across an en- tire product chain will be expensive. Does the focal firm bear these costs or should the other product chain members have responsibility? One could argue that the ISs serve the needs of the focal firm, so it should bear the costs. It is important to note, however, that when firm A is a supplier to focal firm B, firm B is a customer of firm A. As one’s conception of the product chain moves back a stage, firm A becomes focal and firm B becomes part of firm A’s product chain. In this scenario, assuming firms face equal pressures to adopt LCOEM approaches, both firm A and firm B have incentives to develop (and pay for) the ISs we describe. Alternatively, when the focal firm represents a large portion of a supplier’s business, the supplier firm may wish to share costs to keep a large customer happy. Similar dynamics may exist when the focal firm

Shaft, Ellington, Meo, and Sharfman, LCOEM Information Systems 145

AP P LI CAT1 0 N S

is a supplier and a customer who is dependent may also wish to keep its supplier happy. For these mutual incentives to exist, however, firms must face equal pressures to adopt LCOEM ap- proaches. Given that the pressures may diffuse to firms at different rates, larger firms likely will develop the systems earlier, and hence suppliers or customers may have incentives to help de- velop the systems.

Along with the cost issues, there is the diffi- culty inherent in developing interorganizational information systems (IOISs). Although there is a well-developed literature in 101s (e.g., Banerjee and Golhar 1994), much of the work has been in the area of ED1 which has not spread as easily or as rapidly as expected (Bouchard 1993). We believe that LCOEM will require the exchange of large amounts of information (i.e., EII) not just transaction-based EDI. Relatively little is known about how and when firms trans- fer actual information (see Weitz et al. 1994 for a notable exception). It is essential that firms determine what conditions are necessary for EII to occur and when it will be successful.

Another issue related to the introduction of LCOEM IS is that firms along the product chain may have conflicting goals and objectives. LCOEM involves overarching goals that act as unifying .forces for the entire physical system. However, when the objectives of individual member firms are taken into account, conflicts will arise. A simple example is the conflicting cost objectives of the focal firm and its suppliers. For any given product a supplier’s objective is to maximize the profit on that product. When the focal firm buys that widget, its management wishes to minimize cost (subject to quality and other constraints). The two objectives are logi- cally inconsistent. Resolving these conflicting objectives will likely entail political negotia- tions. Further, we described the simple interac- tions between focal firm and supplier. As the focal firm attempts to incorporate all members of the product chain into LCOEM, the level of complexity increases.

A related implication concerns the general power, control, and cooperation issues in LCOEM. The presence of the overall environ- mental goals necessary for LCOEM should re-

duce a firm’s tendency to use political solutions to problems consistent with Sherif‘s super- ordinate goal research (Sherif et al. 1961). The larger goal diverts a firm’s attention from its in- dividual interests in political behavior to the collective need for improved environmental performance, thereby increasing cooperation. However, one can argue that the dependencies that suppliers have on buyers and vice versa may increase the likelihood of political behav- ior (Pfeffer and Salancik 1978). Developing techniques to elicit cooperation will be essential to the successful implementation of LCOEM.

For instance, research on interorganizational politics suggests that when possible firms will use coercive sources of power. Hunt and Nevin (1974) suggest, however, that in a marketing channel relationship (which is conceptually simi- lar to a product chain), firms down the channel are more likely to be satisfied with the inter- organizational relationship when noncoercive sources are used. It is unlikely that LCOEM ef- forts will be successful if product chain member firms cannot move beyond politics.

Conclusion

Our framework is speculative. Only a few firms presently use any elements of an LCOEM ap- proach. However, there are a wide variety of pressures leading firms to look beyond their own boundaries toward their product chain for solu- tions to environmental problems (Sharfman et al. forthcoming). Even if environmental “prob- lems” are not present, in order for firms to re- duce their impact on the natural environment, they must understand the entire effect of any product or service they create. Doing so requires an LCOEM approach at the core of which is ex- tensive information processing. LCOEM ISs will require extensive interchange of data and infor- mation among the members of the product chain. While the ED1 literature is rather well de- veloped, the EII literature is not. Further re- search is necessary to understand the complete implications of LCOEM. With the discussion in this article, we believe that firms can begin working together to build the ISs that LCOEM requires.

146 Journal of Industrial Ecology

APPLICATIONS 1

Notes

1. We have included in the appendix a glossary of all of the acronyms we use as well as a brief defi- nition of each term.

2. At the time of this writing, several groups are developing expert systems that could be inte- grated with LCOEM ISs. For example, the en- ergy and environment group at Sandia National Labs is working to extend EcoSys LCA software to include expert system capabilities.

3. Information about the LCAD project can be ob- tained from Battelle Memorial Institute, Environ- mental Science and Technology Division, 3230 Q Street, Richland, Washington, 99352 (509) 372- 4759.

4. A review of EcoSys and several other environ- mental DSS tools can be found in U.S. EPA (1995b).

5. Much MIS data are derived from TPS data; therefore, the EIS’s connection with the TPS data is indirect.

References

Banerjee, S. and D. Y. Golhar. 1994. Electronic data interchange: characteristics of users and nonus- ers. Information and Management 26 (2): 65-74.

Bouchard, L. 1993. Decision criteria in the adoption of EDI. Proceedings of the International Confer- ence on Information Systems. December, at Or- lando, Florida, 365-376.

Brown, B. 1988. Climate variability and the Colorado River compact: Implications for responding to climate change. In Societal responses to regional climate change, edited by M. Glantz. Boulder, CO: Westview Press.

Brundtland, G. H. 1987. Our common future: The world commission on envi*onment and development. Oxford: Oxford University Press.

Ellington, R. T., M. Meo, and D. E. Baugh. 1992. The total greenhouse warming potential of technical systems: Analysis for decision making. Journal of the Air Waste Management Association 42: 422- 428.

Ellington, R. T., M. Meo, M. P. Sharfman, and T. M. Shaft. Forthcoming. A management method based on total technical systems and total life- cycle analyses. International Journal of Enuiron- menrally Conscious Design and Manufacturing.

Graedel, T. E. and B. R. Allenby. 1995. Idushid ecol- ogy. Englewood Cliffs, NJ: Prentice-Hall,.

Hart, P. J. and D. Ektrin. 1991. Interorganization net- works, computer integration and shifts in inter-

dependence: The case of the semiconductor in- dustry. ACM tramtions on information systems 9 (4):370-398.

Henn, C. L. and J. A. Fava. 1993. Life-cycle analysis and resource management. In Environmental strategies handbook: A guide to effective policies and practices, edited by R. V. Kolluru. New York: McGraw Hill.

Hunt, S. D. and J. R. Nevin. 1974. Power in a chan- nel of distribution: Source and consequences. Joumd of Marketing Research 11: 186193.

Laudon, K. C. and 1. P. Laudon. 1994. Management mformation systems: A contemporary gerspectiue. New York: Macmillan.

Lee, K. 1993. Greed, scale mismatch, and learning. Ecological Applications 3: 560-564.

Markus, M. L. and D. Robey. 1988. Information tech- nology and organizational change: Causal S t N C -

ture in theory and research. Management Science

Pfeffer, J. and G. R. Salancik. 1978. The e x m l con- trol of organizations. New York: Harper and Row.

Prerau, D.S. 1985. Selection of an appropriate do- main for an expert system. AI Magazine, Sum- mer, 26-30.

PCEQ (President’s Commission on Environmental Quality). 1993. Total quality management: A framework for pollution preuention, Washington, DC: US. Government Printing Office.

Sharfman, M. P., R. T. Ellington, and M. Meo. Forth- coming. The next step in being green: The evo- lution to life-cycle oriented environmental management. Business Horizons.

Sherif, M., 0. J. Harvey, B. J. White, W. R. Hood, and C. W. Sherif. 1961. Inrergroup conflict and coopera- tion: The Robbers Cave experiment. Norman: Uni- versity of Oklahoma, Institute of Group Relations.

Ebbs, H. 1991. Industrial ecology: An enuironmental agenda for indusrry. New York: Arthur D. Little.

US. EPA (US. Environmental Protection Agency). 1995a. Green accounting at AT&T, Office of Pollution Prevention and Toxics. EPA742-R-95- 003 Washington, DC: US. Government Print- ing Office.

US. EPA (US. Environmental Protection Agency). 199515. Incolporatng environmental costs and con- siderations into decision making: Review of auailaMe took and sofrware. EPA/742/R-95/006, Washing- ton, DC: US. Government Printing Office.

Weitz, K. A., 1. K. Smith, and J. L. Warren. 1994. Developing a decision support tool for life-cycle cost assessments. Total Quality Environmental Management 4( 1): 23-36.

3 4 583-598.

Shaft, Ellington, Meo, and Sharfman, LCOEM Information Systems 147

AP P LI CAT1 0 N S

Appendix: Glossary of Terms

CAD-computer-aided design-software-based tools to support design efforts.

CAM-computer-aided manufacturing-software- based tools and controls to support production.

CERCLA-Comprehensive Environmental Response Compensation and Liability Act 1980-the ma- jor federal program to clean up contaminated sites, better know as the “Superfund.”

DSS-decision support system-tool to aid manag- ers in analysis and problem-solving tasks.

EDI-electronic data interchange-the computer- to-computer exchange of data between firms to execute business transactions.

EII-electronic information interchange-the com- puter-to-computer exchange of information to support interfirm collaboration.

EIS-executive information system-supports the information and problem-solving needs of stra- tegic managers.

EIS/C-emergency information system/chemicals- Johnson and Johnson’s emergency management software that integrates data from several sources to support upper-level managers.

HSDDS-hazardous solvent distribution data sys- tem-comprehensive US. EPA database on sol- vents, their use, and their disposal.

IDEA-integrated data for enforcement analysisaa- tabase that contains compliance and enforcement data including the CERCLA IS and TRI data.

101s-interorganizational information system-fa-

cilitates data or information exchanges between independent organizations.

IS-information system-set of procedures that col- lect (or retrieve), process, store, and disseminate information to support decision making and control.

LCA-life-cycle assessment-analysis technique that examines the environmental effects associ- ated with a given activity from the generation of raw materials or energy until the ultimate dis- posal or decontamination of outputs.

LCOEM-life-cycle oriented environmental man- agement-an approach to operating a firm that considers the environmental effects of products and processes from cradle to grave and coordi- nates environmental and production efforts with customers and suppliers.

MIS-management information system-supports the managerial control function, typically via time- or exception-based reports.

OAS-office automation system-computer-based system to support information workers, includ- ing professional staff and clerical workers.

PP/TQM-pollution prevention/total quality man- agement-3M’s program to reduce environmen- tal impact by considering environmental effects as part of overall product quality.

TPS-transaction processing system-gathers, stores, and disseminates data relating to basic business exchanges.

TRI-toxic release inventory-U.S. EPA mandated compliance report that lists any emission or us- age of several hundred regulated chemicals.

148 Journal of ‘Industrial Ecology