Reforming Standards-Setting: The Procedural Perspective
Most policy prescriptions concerning standards-setting reflect a procedural view inspired by legal formalism. The evolution of more detailed and formal administrative procedures in the public sector reflects deep distrust of discretionary decisionmaking. Administrative procedures are seen as a method of constraining this discretion. There are two related approaches: one emphasizes rules of participation, the other rules of analysis. Rules of participation circumscribe who can participate in a proceeding and in what manner they can do so. In the public sector, the Administrative Procedures Act (APA) sets forth the requirements for federal "notice and comment" rulemaking. The private sector is widely thought to be less desirable than the public sector primarily because the balance of participating interests is perceived as more "skewed." A popular policy prescription is to establish a private sector version of the APA and do something to "improve" the balance of participating interests. The Administrative Conference of the United States recommends, for example, that private standards be utilized by government only when written by "a broadly based and balanced array of relevant interests."
Rules of analysis, on the other hand, attempt to prescribe the criteria used by the ultimate decisionmaker. A series of controversial executive orders mandate that administrative agencies conduct some form of cost-benefit analysis. Those concerned with the rules of analysis also tend to consider the private sector worse than the public sector because
analytical procedures in the private sector are obscure and occasionally unorthodox. While few argue that analytic rules can assure desirable results, many agree that such rules represent an improvement over the informal approach of the private sector. The fashionable policy prescription: more cost-benefit analysis.
The case studies suggest that the procedural perspective provides only limited insight into the differences between public and private standards-setting. There is little reason to think that these prescriptions will be effective or desirable. Of course it matters who participates and what analytical procedures they use, but the differences between the two sectors are not as big as often imagined. Moreover, efforts to alter the balance of "interests" or to force more rational analysis probably will not make much difference. The procedural perspective is worth considering in more detail, however, because it remains so dominant in policy discussions about improving standards-setting.
Rules of Participation
The most popular view of standards-setting is based on what Richard Stewart calls "the interest group model of administrative law." The underlying notion is that standards-setting, and almost any other policy decision by an administrative agency, is a political process shaped largely by the "balance" of participating "interests." The process is often described in terms of "capture." The Administrative Procedures Act is supposed to provide some protection against capture by opening the process to all interested parties.
Although some have argued that activist judicial interpretations of the APA have resulted in "capture" by consumer and environmental groups, most legal commentators believe that the act has helped insulate agencies from "capture" by business. Stewart concludes that improving interest group representation is not a universal solution but is nevertheless an important "technique for dealing with specific problems of administrative justice." Under this view, there is widespread suspicion that private standards-setting is less desirable than the public alternatives because it is dominated by "business interests." After all, the activity is, by its own terms, private, not public.
Others favor the procedural approach simply because of the immense difficulties in judging actual outcomes. Herbert Simon has argued that agencies should pursue "procedural rationality" when "problems are immensely complex, and where crucial information is
absent." These conditions characterize much of the terrain of safety regulation.
Under either view, the private sector is widely assumed to be worse than the public sector. Surprisingly, the conventional wisdom is incorrect. Most private standards-setters boast a set of procedural requirements that require a "balance" of participating interests. Many also emulate the requirements of the APA by providing notice of their proceedings and allowing the opportunity to comment. The case studies suggest that these "balance" provisions do not necessarily produce the intended results, but many private participants are nevertheless more safety-conscious than expected. The case studies also suggest, contrary to most legal commentary, that the private sector is not much different from the public sector in its "notice and comment" requirements. In a few surprising ways, the private sector is probably better.
Notice Requirements and "Openness"
It has been alleged that private standards-setting is biased in favor of industry because consumers and other related interests are not even aware that some private standards are being written. A lobbyist for the Association of Flight Attendants objects that he cannot follow developments in NFPA 408 as easily as he can monitor the FAA. The case studies suggest that NFPA is not to blame. In fact, rules concerning "notice" barely distinguish the public and private sectors, much less explain any differences in their approaches to standards-setting. Pursuant to the APA, government agencies publish notice of proposed actions in the Federal Register . Private groups do much the same. In an unusual agreement with the National Bureau of Standards, NFPA actually publishes notice of its rulemaking activities in the Federal Register . UL and AGA notify the public through trade journals and in-house bulletins as well as ANSI's Standards Action .
The case studies suggest that any differences in "notice" rules are unimportant. In the case of woodstoves, for example, UL received comments from more people on its canvass list than the CPSC heard from at its public hearing. In fact, a CPSC commissioner commented at the public hearing that the agency did not do a very good job of publicizing its actions. By holding regional hearings in Texas and Florida, the CPSC heard from more members of the public in the space heater proceedings, but the number was still much less than those reached by the comprehensive comment system employed by AGA Labs. In the
two cases where more people participated in the public proceedings than in the private ones—grain elevators and aircraft fire safety—the difference in participation cannot be attributed to procedure. Both the public and the private proceedings were announced in the Federal Register . And in neither case, incidentally, did the additional participation seem to affect the public agency.
The Importance of Being "Balanced"
Even if private standards-setting is as "open" as its public counterparts, what really matters under the interest group theory of administrative law is who actually participates, not who theoretically could. In this respect, most private organizations lend credence to the theory by placing a premium on the appearance of "balanced" decisionmaking. NFPA is the extreme case, conducting what one NFPA member calls "standards-setting by town meeting." The membership at large virtually makes the final decision on whether to adopt a standard. (The Standards Council actually has the final say, a power it has exercised in response to efforts to "pack" votes on the convention floor.) At the committee level, NFPA goes to great lengths to classify its members according to their "interest." The Standards Council reviews assignments to ensure that committees are "balanced" and that no single interest dominates the proceedings. As a practical matter, this usually means that no more than two representatives on a committee are ever from a single category of interests. But the Standards Council looks at specific issues as well as overall numbers. On the Fire Sprinkler Committee, for example, "balance" also means equal representation from metal and plastic interests, and the Standards Council has recently decided that "balance" on another committee requires both union and nonunion representation. The importance played by NFPA members on the makeup of committees is evident in the increasing number of formal appeals over committee membership decisions.
Whether any of this really affects the quality of standards-setting is unclear. Three questionable assumptions underlie the application of "interest group theory" to private standards-setting. First, standards-setting is assumed to be a legislative-type process, in which voting is the ultimate method of decisionmaking. That is not the case with UL standards. It probably is not an accurate description of most "consensus" decisionmaking. Second, the balance of participating interests on the private side is thought to be weighted heavily in favor of business rather
than consumer interests. That is not true in many private standards-setting efforts. Various "business" interests advance the cause of safety. Finally, the underlying assumption is that adding more "consumers" or consumer advocates would have a beneficial effect on the process. The case studies provide several reasons to discount this hypothesis.
Appearances Can Be Deceiving
"Balance" requirements notwithstanding, there might be, as is often alleged, a lack of meaningful diversity in the interests actually represented in the private sector. In fact, the case studies suggest that the "balance" of interests in private standards-setting looks much better on paper than it does in reality. First, the requirement that interests be "balanced" seldom applies at the primary level of decisionmaking, where the standard is actually written. "Balance" tends to be required at the reviewing stage, not the writing stage. The general Z21 committee is "balanced," for example, but technical subcommittees, such as Z21.11.2, which are not, do all of the work. Similarly, UL's canvass lists are "balanced," but the standards themselves are written in-house. The significance of balanced participation depends, then, on the scope and significance of review. The more deference given to technical committee decisions, the less significant any "balance" requirements. Since review in the private sector tends to be highly deferential, the practical significance of current "balance" requirements appears minimal at best. (They may still have symbolic value, of course.) The real decisions are made by technical committees or individual engineers who operate outside these requirements.
Even when "balance" is required, the process is riddled with practical problems. The categories are too crude. For example, under the most elaborate classification system—NFPA's—insurance company representatives complain that brokers, who really represent the interests of their clients, not of the insurance industry, are classified as "insurance" representatives. It would take a classification scheme as complex as the Standard Industrial Classification system to begin to capture the important differences in business interests. Even then, unless participants were also classified by intraorganizational affiliation (for example, Engineering and Safety Department, General Counsel's Office, Marketing), it would be impossible to capture many of the significant differences in interests.
The problem is worse with consumer interests. The category "consumer" is so broad as to be practically meaningless. Everyone is a consumer. For a standard written by the Hydraulic Institute, both the Chemical Manufacturers Association and the American Water Works Association were considered "consumer" representatives. This kind of flexibility led an NFPA member active on several aviation committees to conclude that "classification of members in NFPA is a joke." Moreover, consumers care about more than just safety. "Good coffee," points out the manager of the J.C. Penney Testing Lab, "is as important as the quality of the handle on the pot." Cost can also be as important as safety. Not everyone wants or can afford the "Rolls Royce of coffee pots," or any other appliance for that matter. Styling is also important to consumers, as suggested by the marketing adage about 1956—the year Ford sold safety and Chevrolet sold cars.
Still, concerns about the "balance" of participating interests appear to be validated by the modest voice that organized consumer groups have relative to manufacturers in private standards-setting. UL and some ANSI-sponsored committees pay lip service to "consumer participation," but consumer groups play only a minimal role in the process.
The Opportunity to Comment
Whatever the significance of who participates in standards-setting, what may be most important is how that participation occurs. This raises the second part of "notice and comment" rulemaking. On the public side, participation is often in written form. Pursuant to the APA, agencies request written comments on rules proposed in the Federal Register . Some agencies choose to hold public hearings as well, but this is not required under the "informal rulemaking" procedures of the APA. Responses to written comments are routinely published in the Federal Register before a rule becomes final. On the private side, contrary to popular conceptions, participants, whatever interests they represent, generally have better access than in the public sector. Most organizations also offer an opportunity to respond to proposals with written comments. NFPA utilizes a procedure almost identical to that of the APA. ANSI mandates similar procedures as part of the larger requirement that private standards reflect a "consensus" of affected interests. Many of the APA-like rules in the private sector were adopted
in the wake of the FTC's investigation of standards and certification. There are examples in both sectors of seemingly legitimate concerns being ignored. This suggests that rules governing "comment" might not be very effective in either sector.
Surprisingly, the forms of participation are often more limited in the public sector than in the private, leading an attorney for the FTC to conclude that "government standards-writing is less open." "The only idea of process in the public sector," he complains, "is holding hearings." Such hearings were held in all of the public sector cases but aviation fire safety. (Written comments were solicited in all four cases.) Yet many participants in the CPSC proceedings complain that participation is stifled because observers cannot communicate with the commission during the many important briefing sessions conducted by the staff.
Direct dialogue between commenters and decisionmakers is more common on the private side. This has led some to conclude that private standards-setting holds the advantages sought in proposals for public "regulatory negotiation." There were opportunities for informal dialogue in all of the private sector cases. UL engineers met with the Industry Advisory Council several times while converting UL 1482 to published form. The minutes of the AGA/ANSI Z21.11.2 committee indicate that outsiders often attend meetings as guests. ANSI rules provide that "affected interests," a term subject to some dispute, can always attend committee meetings. In short, those who want to participate appear to have better access to the standards-setting process. UL's canvass method is a bone of contention because it allows only industry participation during the formative stages of standards-setting, relegating other parties to commenting in writing on a proposal developed in their absence.
Under the comment process specified by the APA, public agencies are expected to provide "reasoned responses" to all comments on the proposed rule. In theory, this should force the agency to face its critics. In reality the process is supervised by agency lawyers whose main concern is to ward off judicial review. Responses to comments are tailored to protect the agency, not to answer the commenter. The same appears to hold true on the private side, where the demands of the commenting process are greater. Most private standards-setters aspire to develop "consensus" standards, meaning that "unresolved negative comments" must be addressed in a convincing manner. But the response to comments in both sectors is prone to superficiality.
In the private sector, the process that unfolded in all four case studies was far from responsive. Like government agencies, UL delegates the task of responding to commenters to a special department that does not write the standards and whose main interest is in gaining approval, not changing the standard. In the case of woodstoves, the "response" to several comments was unresponsive; in a few cases, it was actually misleading. NFPA has a similar problem. According to an NFPA official, commenters often complain that "committees claim to 'accept a comment in principle' but then go off in another direction." Another tactic, used in the grain elevator proceedings, is to hold over controversial comments for study and then dismiss them summarily several years later.
Given their more technical orientation, however, private standards-setters might actually do a better job than their public counterparts in responding to comments. A detailed study of public and private standards for liquefied natural gas facilities concluded that "responses by the [NFPA] technical committee were far more specific than responses given by the [Department of Transportation]." In neither case does the requirement to respond to comments appear to have a significant effect on the quality of decisionmaking.
Rules of Analysis
The procedural perspective has evolved into a broader effort to force "comprehensive rationality" in administrative decisionmaking. This "analytic imperative," as Colin Diver describes it, is embodied in a combination of executive orders and statutes mandating that public agencies analyze the economic impact of proposed standards. There are no similar rules in the private sector. Cost-benefit analysis is almost universally avoided. Leland Johnson of the RAND Corporation has identified this as a major shortcoming of private standards-setting. In short, there appear to be significant differences in the rules of analysis governing standards-setting in the public and private sectors.
Beginning with outward appearances, systematic economic analysis seems to pervade public standards-setting and elude private efforts. The CPSC prepared preliminary and final economic impact statements in both cases, quantifying costs and describing possible benefits. The FAA's full-blown economic analysis, comprising the bulk of the Federal Register notice announcing the proposed rule, quantified costs and benefits in a detailed fashion. OSHA contracted for two risk-benefit anal-
yses of the grain elevator rule, with the second addressing criticisms of the first.
By contrast, private standards-setting appears much less structured and systematic. A member of the NFPA committee on chimneys and venting—a man with a doctorate in physics, widely acknowledged to be an expert on woodstoves—recalls his initial shock at the "casual and off-handed way in which many decisions are made." Decisions made at the NFPA committee meetings on grain elevators in 1985 were similarly informal, often based on anecdotes, if on any information at all. The very notion of estimating overall benefits or costs, let alone explicitly trading them off against one another, is opposed by many in the private sector. "I don't know of any time that we have consciously gotten into cost-benefit analysis," comments a UL vice president. An NFPA executive committee has spent seven years considering the appropriateness of cost-benefit analysis. Many members share the feeling expressed at the organization's 1985 annual meeting that "cost-benefit analysis is not relevant to the actions of NFPA."
Of course, decisions packaged as cost-beneficial do not necessarily embody the concept of comprehensive rationality. Agency subterfuge is one reason. Procedural requirements can spur defensive tactics instead of substantive changes. For example, many agencies faced with the analytical requirements of environmental impact statements successfully repackaged their decisions without changing the outcome. The same phenomenon is likely when cost-benefit analysis is mandated by statute, regulation, or executive order. What Stephen Breyer is quick to observe about public standards-setting is equally true of private efforts: the process "as it might exist in the world of the rational policy planner" almost never happens in reality.
In none of the public sector case studies did the economic analysis, so prominent in the official record, appear to shape the content of the standard. The people "downstairs" write the rules at the FAA; the people "upstairs" do the cost-benefit analysis. The tasks are done more or less at the same time, and no one interviewed on either floor thinks that the cost-benefit analysis affects the substance of the rule. (Of course, it might help justify what ends up being proposed, and it certainly delays the process to some extent.) Many OSHA officials are similarly skeptical of systematic attempts to quantify costs or benefits. They view such efforts largely as methods for justification, not decisionmaking. The risk-benefit analyses funded by OSHA were done after the grain elevator standard was written. In short, analyzing public and
private standards-setting in terms of their rules of analysis is an invitation to frustration. On the public side, it is clear that the rules do not tell the whole story, while on the private side it is not even clear what the rules are. The ideal of comprehensive rationality provides little basis for evaluating or comparing standards-setting systems in a manner that illuminates the observations in chapters 7 and 8.
Standards-setting is often analyzed through the lens of administrative law. Decisionmaking processes are evaluated by reference to the rules of participation and the rules of analysis—rules intended to promote such lofty goals as "balance" and "comprehensive rationality." Most public policy initiatives concerning private standards-setting reflect this perspective. The underlying assumption seems to be that the more private standards-setting looks like public standards-setting, the better. Recommendations from both the FTC staff and the Administrative Conference of the United States come close to requiring a private version of the APA.
The case studies suggest that some common conceptions about the differences between public and private standards-setting are incorrect. Notice is not necessarily better in the public sector, nor is the form of participation. More significantly, the case studies suggest that the procedural perspective is inadequate for understanding the standards-setting process. The fundamental flaw with this perspective is summed up in the fact that the public and private systems, which vary in myriad respects detailed in chapters 7 and 8, look practically alike in the language of administrative procedure. Under the veneer of "due process" and "consensus," the many institutional arrangements for standards-setting—from the trade association that writes its own standards, to the product certifier that writes them for others to use, to NFPA's "town meeting" and the various approaches taken by government—all appear to be more or less equivalent.
In short, the procedural perspective seems to gloss over, rather than illuminate, the most important aspects of standards-setting. Many of the procedural devices attendant to standards-setting, particularly those adopted on the private side in response to the threat of FTC regulation, are apparently like "grafts on a scheme of government to which they are intrinsically foreign." They may improve outward appearances, but
they make little difference to the underlying dynamics of decisionmaking. The procedural perspective, although popular, is elusive because it does not speak to the important differences between public and private standards-setting. Some alternative avenues of analysis, with related policy prescriptions, are contained in chapter 12.