|This page in the original is blank.|
Environmental Measures: Developing an Environmental Decision-Support Structure
Given ever-present resource constraints, a substantial proportion of business managers' time and effort is devoted to maximizing output and profitability and minimizing risk and loss. In the increasingly competitive international markets where many firms operate, the choices about resource allocation have become ever more important to both the short term profitability and the long-term survival of the firm. In making choices, managers must continually trade off one feasible business opportunity against others, weighing the resources they expect to be consumed against the anticipated benefits for each. The ultimate quality of the decision made, that is, the relative benefits and costs achieved or realized, will depend heavily on the quality of information that the manager has available when the decision is made.
Information systems dedicated to supporting managers' decisions have long been an integral part of business. However, environmental decision making is relatively new as an intrinsic, fundamental part of ongoing business strategy development and definition. Consequently, the environmental information support that managers need to make decisions is rarely as well developed as the information available for other business areas. Most crucially, the quality of the available information may be poor. Moreover, upper-level managers, responsible for most strategy and policy formulation, may not be fully aware of the deficiencies of the information because details about the data's limitations are typically lost as the data are transmitted up through the firm's levels of management.
Few disciplines have been marked by such rapid change as those related to the environment. Many developing areas of corporate and academic research, for example industrial ecology and life-cycle analysis, were unknown 20 years ago, and today they consume significant resources. However, the relative infancy of
these and other environmental efforts stymies the development of the measurement techniques and dedicated instrumentation needed to carry them out properly. This lag exists in part because firms face such an enormous task in removing waste from emissions and effluent streams as rapidly and fully as possible. These activities rightly have had to consume the lion's share of corporate environmental investment resources in the short term. As a result, however, most of the development has had to be directed to end-of-the-pipeline waste control rather than to prevention and control of input or process-stage pollution. As a consequence, we see relatively sophisticated measurement and control efforts being made over incineration and waste-water treatment processes, while little or no effort goes toward processes to handle the flows going into the facilities other than, perhaps, engineering estimates required for recharges of facilities' costs.
Previous studies have examined some of the qualitative environmental inadequacies of typical business accounting systems (Todd, 1989, 1994) as well as systematic approaches to environmental profiling and information gathering of firms' productive activities (see, for example, Allenby and Graedel, this volume; Allenby, 1994). This paper focuses on the development of a taxonomy for evaluating the adequacy of environmental measurement systems and information available within the firm, targeting those areas with the most urgent need for measurement development or refinement. The objective is to provide firms with a basic structure for (1) evaluating the current state of environmental measurement and control, and (2) developing a business strategy that will allow them to move systematically from this stage to a rigorous and comprehensive system of environmental monitoring, control, and managerial decision support. Other environmental information systems, for example Allenby and Graedel's (this volume) environmentally responsible facility assessment matrix, could prove invaluable as basic building blocks in the structure. Beyond this immediate objective, the ultimate goal is to develop a metasystem to support pollution prevention efforts at the source. This paper first examines the types of decisions that managers must make, especially about the environment. Next, it considers the nature of measurement and some of the factors that reduce the quality of the measures. Finally, it describes the development of an environmental measurement taxonomy and the types of managerial decisions that are likely to be affected.
Managers' Business Decisions
Typically, managers must make decisions that affect three broad categories of operations: (1) long-term capital investment, (2) monitoring and control of operations, and (3) performance evaluation and motivation of employees. Environmental considerations are increasingly involved in each of these.
Among the most difficult decisions that managers must make are those that involve the long-term dedication of scarce capital to direct asset investment or research and development projects. Such investments are typically undertaken to
improve profitability or competitive position or to reduce risk. The specific goals may be, for example, to develop new products or markets, increase production yields, or reduce wastes. More and more frequently in recent years, these goals have had environmental aspects, such as developing processes that eliminate toxic or highly reactive compounds or reduce air emissions, such as volatile organic compounds. With these investments, the manager may be able to achieve several major objectives at once: increase profitability, improve compliance with environmental regulations, and reduce the firm's long-term imbedded risk by decreasing or eliminating sources of future liabilities.
An obvious difficulty with such otherwise desirable investments is that they are likely to be highly uncertain undertakings and involve very long payback times, amounting to 10-15 years or more. Indeed, a middle-level manager proposing such long-term investments is unlikely to be there to reap the rewards of the investment under the performance evaluation systems that most firms use.
Therefore, firms that seek to minimize imbedded risk and improve competitive positions must provide senior managers responsible for long-term decision making with the best information obtainable, given cost and benefit constraints. This information must not only be relevant to the decision but also be as reliable as possible. Although large sums of capital may be expended in a single decision, only rarely do firms ''invest" in major projects to improve the quality of information. Moreover, because environmental decision making is relatively new, firms may not be systematically gathering and aggregating information suitable for environmental decisions.
A second major use of managerial information is in the monitoring and control of ongoing operations. Decisions in this area are commonly regarded as being routine—for example, continuous monitoring of a chemical batch process, quality testing of products, and comparisons of actual and budgeted production output. However, managers responsible for this day-to-day routine monitoring are likely to be those with the greatest knowledge of specific processes and product markets, as well as with the environmental aspects of the operations. Therefore, they are in the best position to recognize and respond to changes in the environment, and also those of a more general business nature. Indeed, they may have most of the responsibility for achieving the goal of minimizing future environmental liabilities now. Therefore, the greater the precision, relevance, and validity of the measures they rely on, the likelier it is that they can take timely and effective action to minimize both short- and long-term environmental problems.
Performance evaluation and motivation of employees are among the most important uses of information for supporting managerial decisions. Most firms want to minimize waste and reduce environmental liability exposure. Generally, employees are encouraged to achieve corporate goals through the performance evaluation and rewards systems that are established specifically to accomplish this purpose. For example, because most companies want to increase profitability, many typical managerial reward systems award so-called incentive pay and bonuses to those managers who achieve planned profit targets. However, few com-
panies have chosen to incorporate environmental goals into incentive compensation contracts. Indeed, although many firms stress environmental, health, and safety awareness and goals at all levels, the managers making the day-to-day decisions may view the environment as being of secondary importance at best. The reason is simple: The rewards system by which managers chart their own progress speaks only to profits, not pollution prevention or long-term risk reduction.
The difficulty of measuring environmental "progress" is a common reason cited for the omission of environmental concerns in the direct incentive and rewards scheme. Part of the difficulty is that both the risks and the paybacks in this area are uncertain and long term. Typical pollution-prevention projects in the past commonly have been characterized by easily measured inputs and outputs with immediately apparent cost-benefit trade-offs. Projects involving complex measurement difficulties are frequently ignored. However, an emphasis on short-term problems is not unique to the environmental area. Recently, both accounting regulatory organizations and firms have had to come to grips with such issues as the measurement of pension and postretirement (health care) benefits. Indeed, a major impetus for firms' shifting to health care cost-containment schemes is that, once they had developed workable measures of the future liabilities and the related current costs of these programs, managers were forced for the first time to monitor and control them.
Thus, gains to long-term strategic management of firms' competitive positions, profitability, and risk are likely to be enhanced substantially by investment of time and resources in the development of improved measures of environmental input, output, outcome, and performance.
The Nature of Measurement
Measures, whether of the height and weight of a person or of emissions from volatile organic compounds, are likely to be flawed representations of the true natural state of the object being measured. The difference between the "true" natural state and that suggested by the measure, or proxy, is termed measurement error. The relationship between the true state and the measure is
xmeasured = xtrue + esystematic + erandom
where xmeasured is the flawed proxy for what we wish to know, xtrue; esystematic is a form of repetitive error; and erandom is a nonrepetitive and, by definition, unpredictable form of error. The measure of, for example, the amount of a certain toxic chemical in waste that is flowing to a water treatment facility, is the only directly observable component of the relationship. The systematic error, esystematic, arises from sources of bias in the way the toxic chemical is measured. A simple example would be if a malfunctioning instrument were used in the analysis and caused the measurement to be consistently higher (or lower) than the actual
amount of chemical present in the stream. Certainly, more than one source of systematic bias may be present in a given measure. erandom encompasses the remaining element of error. As the name suggests, this error source is unpredictable and nonrepeating. This portion is assumed to have a mean of zero over the long term, although individual observations may have large positive or negative random errors, depending on the nature of the measure.
The usefulness of a measure in decision making is directly related to its reliability. Reliability is the proportion of the measure that is free of error. For example, if 90 percent of a measure is either systematic or random error, the reliability is only 10 percent. Put differently, only 10 percent of the measure accurately reflects the object being measured. In many fields, for example medicine, much time and effort is expended on the development of highly reliable measures and tests. The reason is obvious: correct diagnosis and treatment and perhaps the patient's life depend on the accuracy of the informational inputs to the physician's decisions.
In business, however, the reliability of measures may vary widely. In traditional cost accounting, reasonably precise measures may be obtained for the direct materials and labor expended to manufacture a product. Other cost sources are measured with substantially less precision, including such overhead items as depreciation, the allocation of the historical cost of fixed assets to each of the periods in which the asset is used in production. Indeed, when precisely and imprecisely measured costs are summed, the result may not be useful for informing managerial decisions (Todd, 1994).
Such problems are likely to be particularly extreme in the environmental area. Several factors explain this, including the relative inexperience of firms' managers and financial staff in dealing with the much newer environmental measures, the high levels of complexity involved in much of the measurement, and uncertainty about the long-term environmental implications of many of the materials used in production.
As a consequence, many managers have come to rely on qualitative evaluations, rather than quantitative data, for environmental decision making. This certainly makes good sense where the knowledge base is low and measurement and monitoring technology is primitive or nonexistent. Indeed, the practice of medicine was in this state two centuries ago. However, if information determines the quality of both decisions and outcomes, it clearly is necessary to develop systems for evaluating and monitoring information quality and continually upgrading the information available to decision makers.
Common Sources of Environmental Measurement Error
Environmental measurement error may arise from a variety of sources. However, several circumstances account for a large proportion of such error.
In most cases, the most reliable environmental measurements that firms now
use are those related to government environmental regulatory oversight. Given possible public audits by either regulators or public-interest groups, managers have incentives to ensure that the numbers they report are reasonably accurate. However, such measurement is largely centered on end-of-the pipeline emissions and effluents, rather than pollution prevention and elimination. Consequently, managerial decisions have been flavored to some extent by this regulatory and measurement focus. That is, a manager who must comply with an environmental discharge regulation will no doubt seek the most readily achievable fix, for example installing a scrubber to remove the emission. The difficulty is that the short-term regulatory solution may not be followed by long-term process-waste removal unless both reliable measures and incentives are in place to encourage managers to do so.
Managers are likely to use less-precise information for internal managerial environmental decisions than for reports required by regulations. For example, managers frequently rely on estimates of individual process waste streams and component volumes developed either in process research and development or in early stages of implementation. Sampling to confirm the accuracy of the estimates may occur infrequently or not at all. Without such confirmation, managers cannot know the accuracy or reliability of estimates. Moreover, if either processes or planned inputs are changed, the estimates may well not be updated. In addition, as facilities age, the yield and waste proportions may change.
Such situations may result in a very large systematic error component of waste-stream measures, which renders the measures unreliable and decisions based on the measures suboptimal or even dysfunctional. For example, waste treatment facilities' costs are typically redistributed by means of recharges (or transfer prices) to production units using the facilities' services, based on estimates of waste volumes and components. More important, however, such inaccuracies may strongly influence long-term investment decisions on capital asset replacement. Monitoring and control functions will fail if control targets are not known with reasonable accuracy.
Random measurement error results from short-term fluctuations in any of the factors that can influence the measure. These may include (but are not limited to) human error as well as external sources, for example inputs that do not meet engineering specifications. Unless the measurement is ongoing, such error may not be detected.
A Taxonomy for Improving Environmental Measurement
Current managerial information systems, including accounting systems, typically focus on only a relatively small subset of the data available or obtainable about a company's productive efforts. This occurs because these systems have
been driven primarily by the external financial reporting imperative incumbent on all publicly held firms.
Regulatory bodies have defined both the type and the amount of financial information that firms are required to report. To minimize the cost of acquiring this information, firms have tended to adapt the information systems they use to support external reporting requirements. For example, external financial reporting rules require full-absorption (that is, fully aggregated) costs of goods sold and ending inventory valuation numbers. Full-absorption costs comingle not only current out-of-pocket expenses but also allocations of the historical costs of fixed assets.
Managers, rather than suffer additional delays while a new form of information is being specially processed or incurring additional costs of information development, will commonly use the readily available but, in many cases, highly unsuitable numbers for decision making. In the example just cited, a manager in the process of deciding whether, for the short term, to manufacture a product internally or buy it from an external vendor would want measures of the out-of-pocket, or short-term escapable, costs. All other factors are irrelevant. However, historical allocations built into the full-absorption cost may be so large as to swamp the out-of-pocket information, and it may not be possible to disentangle the irrelevant from the relevant decision information. In terms of the criteria for measurement quality in the foregoing section, a full-absorption number for a make-or-buy decision will likely contain a large proportion of systematic error that will be evident only after considerable investigation. If a manager wants to determine the proportion of a product's cost that has environmental implications, most information systems in use probably could not generate that information without considerable additional analysis.
In summary, existing managerial and accounting information systems may suffer from two broad classes of inadequacies related to environmental decision making: (1) the required information may not be available in the current system, and (2) the information available may be irrelevant or subject to substantial and indeterminate error as applied to the decisions to be made.
Therefore, for environmental decision-making purposes, including capital investment, monitoring and control, and performance evaluation and motivation, the firm must review or profile the environmental information generation, aggregation, and reporting that it uses. Table 1 suggests a simple taxonomy, readily adaptable to a wide variety of firms. Environmentally relevant functions or activities run across the top, indicated by numerals I to V. Depending on the nature of a firm's business and productive activities, the firm may have more or fewer such categories. However, to be comprehensive, the measurement profile should extend back to the relevant supplier/factor stage and forward to customer/end-use activities. This scheme is entirely consistent with life-cycle analysis approaches.
The second dimension of the taxonomy reflects the relative quality of the information available. Only four categories are suggested here, ranging from
TABLE 1 Taxonomy for Development of Information and Measurement Systems for Environmental Decision Making E
Measurement/ Information Quality
Waste Treatment/ Processing
Materials, production methods, products, by-products, waste generated, waste handling, waste treatment, regulatory profile, legal liability issues etc. . .
Raw materials (nonenvironmental), raw materials (environmental), direct labor, relevant overhead, energy, ancillary services, production methods, regulatory profile, legal liability issues, etc. . . .
Products (nonenvironmental), products (environmental), by-products (nonenvironmental), by-products (environmental), waste (nonenvironmental), waste (environmental), regulatory profile, legal liability issues, etc. . .
Waste handling, waste recycling, waste reprocessing, waste treatment (air), waste treatment (water), waste treatment (solid), ancillary services, regulatory profile, legal issues, etc. . .
Waste generated, waste recycling, waste reprocessing, waste treatment and disposal, regulatory profile, legal issues, etc. . .
Measures not available
necessary but currently nonexistent information to measures of the highest and most reliable quality. Again, this range can be adapted to meet a firm's needs in a particular case. For example, if a particular management decision is designed to identify which facilities are "environmentally responsible" (see Allenby and Graedel, this volume) and those that are substandard along one or more dimensions, qualitative, or category C, measures may be perfectly suitable. However, if research and development managers are trying to decide where to target scarce research and development resources according to which of a number of competing projects will yield the highest environmental benefit in terms of pollution reduction, then category B or even A measures may prove most useful.
The purpose of the taxonomy, then, is the early identification of information needs, sufficiencies, and inadequacies so that firms can undertake planned information system enhancements. In many cases, they can achieve the enhancements at very little additional cost, given adequate time and planning.
A handful of items that firms might consider are listed under the various activity categories. Traditional accounting systems, which concern themselves only with the productive activities within the firm, will not generate information about suppliers (category I). The one exception is vendor lists in the purchasing departments. More recently, however, many firms have begun to make more or less routine visits to suppliers to learn something about the suppliers' exposure to potential environmental liabilities, and thus the firm's possible liabilities as well. In such cases, they usually collect qualitative information. Nonetheless, the firm's environmental stewardship begins with the factor inputs and any business relationships that may be relevant.
Category II, inputs/processing, is usually the point where managerial accounting and information processes begin. For example, data on materials and supply inventories and on labor costs have long been accounted for relatively precisely. However, additional costs, usually designated overhead items, are normally pooled and reallocated to products rather than being explicitly accounted for. Energy costs, one of the most important costs for environmental measurement and control, are invariably treated in this way. Only rarely does the information system recognize the environmental relevance of such cost items. For example, raw materials may be toxic, highly reactive, or have other environmental implications, but only those with direct knowledge of the processes will have this information. Moreover, many items highly relevant to the environment, for example organic solvents, may disappear into overhead pools in the accounting process (Todd, 1994). Other category II costs, including ancillary services such as toxicity testing and legal advisory services, which may have high environmental relevance for a particular product, may not be linked to the product in any fashion. Indeed, the product-specific costs incurred will likely not be identified at all, becoming category D measures (or nonmeasures).
Beginning with the outputs category, specific internal accounting declines rapidly in quality insofar as environmental considerations are concerned. For
example, because most managerial accounting and information systems focus on a salable manufactured product, detailed records on waste generated are unlikely to be available. Moreover, the system does not capture the distinction between environmentally relevant products, by-products, and wastes.
As observed above, information on waste recycling, reprocessing, and treatment activities is usually based on engineering estimates that may be out of date and substantially in error. Measures for category V, customer end-use and recycling activities, are at best in an early stage. Nonetheless, these are receiving increasing regulatory and consumer attention and are likely to become significantly more important in the future.
A latent but vital dimension in this taxonomy is the identification of pollution sources. For example, among the system's environmentally relevant outputs that may require waste recycling, reprocessing, or treatment may be a hazardous input that is only partially consumed in the process, thereby generating a hazardous waste. The resulting waste costs, including any incurred by regulatory monitoring and oversight, are commonly treated by accounting systems as if they arose in category IV. In fact, however, the problem originates in categories I and II and can only be eliminated by intervention there. Rather simple refinements to the information gathering and processing system will make it possible not only to identify the point sources of such items but to aggregate the entire effect of their use through the whole of the system.
This paper recommends that firms develop comprehensive systems for evaluating the adequacy of environmental information currently available to support managers' environmental decision-making needs and establish procedures to systematically upgrade those systems. The rationale for such a process is that the quality of managers' decisions is likely to be heavily influenced by the quality of information available to them.
A simple prototype taxonomy is presented that firms may adapt to their own operating environments and that can be used to guide the identification of relevant environmental information and the quality of information currently available in the firm. The firm can use this knowledge to facilitate the deployment of assets and the adoption of development efforts in those areas where information is deemed to be most severely lacking. The ultimate goal is to provide essential support for efforts to prevent pollution and reduce energy consumption across the entire spectrum of a firm's activities.
Allenby, B. R. 1994. Integrating environment and technology: Design for environment. Pp. 137-148 in The Greening of Industrial Ecosystems, B. R. Allenby and D. J. Richards, eds. Washington, D.C.: National Academy Press.
Todd, R. B. 1989. Environmental Accounting: Patching the Environmental Fabric. National Research Council. Washington, D.C.: National Academy Press.
Todd, R. B. 1994. Zero-loss environmental accounting systems. Pp. 191-200 in The Greening of Industrial Ecosystems, B. R. Allenby and D. J. Richards, eds. Washington, D.C.: National Academy Press.
|This page in the original is blank.|