I’ve worked on a lot of regulations, and it seems sometimes that the only thing people hate more than vague requirements are clear ones. And vice versa.
The world offers many opportunities to be dissatisfied. Regulations are no exception, and provide the double-hit of being difficult whether they are ambiguous and flexible performance standards or prescriptively clear. Performance standards, regulations couched as flexible, open-ended standards that accommodate a variety of technological solutions, may leave a hapless regulated entity in the dark as to what an agency requires. After all, as many have noted, the ultimate performance standard consists of the admonition “Be safe.” On the other hand, if an agency starts to issue regulations that spell out prescriptive design solutions as the answer to a safety concern, the regulations lose their flexibility even as the affected industry learns precisely what the agency requires of it. But it is not necessarily the performance standards themselves that lack sufficient specificity.
A lot of the time, it’s not the standard that’s the problem, but the demonstration. Applicants for authorizations, whether for a launch license or aircraft certification, must demonstrate how they satisfy the FAA’s regulations. If an applicant must satisfy a performance standard, the applicant may have to show the agency a lot more than if he only needs to run through a very specific checklist. If faced with a checklist, however, the applicant will have lost flexibility in designing his vehicle.
Examples of Performance Standards
The human space flight regulations in 14 C.F.R. part 460 are very performance based. For example, section 460.5, which governs crew flight qualifications, requires that each member of a flight crew demonstrate “an ability to withstand the stresses of space flight, which may include high acceleration or deceleration, microgravity, and vibration, in sufficient condition to safely carry out his or her duties so that the vehicle will not harm the public.” While broad enough to encompass diverse technologies by allowing the training to model the flight profile of both winged vehicles and capsules, it provides enough specificity to allow the operator to know that the FAA wants the demonstration to account for high acceleration or deceleration, microgravity, and vibration. Nor does it set specific levels on how much the crewmember must be able to withstand. Theoretically, the FAA should find acceptable far higher g forces when Superman acts as a pilot than when the average human fills the role.
The crew training requirement is thus a nice example of a good performance standard. What remains unclear in the regulations, however, is what the applicant must show the FAA: studies of the population as a whole, or tests of particular individuals slated to fly?
Similarly, suppose an applicant were to apply to the FAA for a launch license. One of the requirements for a launch license, whether from a federal launch range or not, mandates that a launch operator must demonstrate to the FAA compliance with 14 C.F.R. §417.129. That regulation offers another good example of a performance requirement. It requires a launch operator to ensure that for any upper stage left on orbit debris generation will not result from the conversion of energy sources into energy that fragments the vehicle. The provision does not tell the operator how to go about satisfying the requirement, and leaves the operator free to choose a different approach than other operators, or to improve his own process if he finds a more cost-effective one. (If a launch operator changes a process after demonstrating how an earlier one would work, the operator must obtain a license modification from the FAA).
One thing the regulation does not do is tell the operator what the FAA wants in the way of a demonstration. Does the FAA want a copy of the design, predictive analysis, computer modeling, or thousands of hours of testing? What kind of confidence levels are expected? Margins? Like other performance requirements, the regulation says only that it requires a demonstration. As one might guess, this demonstration must consist of more than a written assurance that, “sure, there’ll be none of that fragment-producing energy conversion.” But more what?
The results can be bad if the regulations don’t spell out what the demonstration must consist of. It’s difficult to plan or budget if one doesn’t know what the regulator expects. Particularly if an operator files an application precisely 180 days before launch, surprises about what the regulator requires may result in real-time schedule delays. There is also the problem of subjectivity and fairness. Different agency evaluators may be satisfied with different levels of proof, thus rendering the process more subjective than we might want. An applicant might not know that what one regulator finds an inadequate demonstration was approved by another one.
The Administrative Procedure Act, the federal law that governs how federal agencies regulate the private sector, provides two general categories of oversight. The one is adjudicatory, and it applies to individual entities. Adjudications consist not only of hearings of disputes (the administrative law version of trials), but of less adversarial determinations such as the issuing of licenses, permits, waivers, exemptions, and certificates to specific entities. The other oversight path happens through rulemaking, where an agency issues rules of general applicability that may apply to industry as a whole.
In the adjudicatory context, a little more transparency about what constitutes an acceptable demonstration might serve the regulated industry well. Where the law protects and treats as confidential designs and test results, the agency could scrub the data and disclose that it accepted computer modeling or nothing less than full bore flight testing as an acceptable means of demonstrating compliance with a particular regulation. Other operators may then start to learn what works.
One way this could happen is over time. An agency would issue its safety standards. The first applicant would likely have to engage in lengthy negotiations about what demonstration would suffice. If the next applicant with similar technology or circumstances knew that the agency had accepted computer modeling rather than wind-tunnel testing to satisfy a particular provision, the two entities would be treated fairly, subjectivity would decrease, and industry would gain greater certainty in its planning.
Another approach would be to spell out the demonstration requirements in advance by regulation. This might prove highly restrictive and result in repeated requests for waivers. Even so, the waivers should be publicly available so that other similarly situated operators would know that other paths might be available. The FAA does make waiver determinations publicly available the Federal Register. They may easily be found by an online search with the search engine of your choice.
Rulemaking provides a real opportunity for launch operators to influence regulatory requirements. The FAA publishes its proposed requirements in a Notice of Proposed Rulemaking and allows the public, which includes the affected industry, at least 30 days in which to provide comment. A certain amount of thought can pay off in providing comments. An operator shouldn’t just tell the agency that its requirement is too vague. The agency is as likely as not to mandate something too difficult or expensive if it starts getting more specific. Instead, operators should consider offering substitute language, and explain the technical or legal reasons for the proposed change. Like other agencies, the FAA must take the comments into account. Just as the FAA’s own proposal must be reasoned and supported by a technical foundation, so should the commenter’s counter-proposals. Agencies do want to have regulations that work. Rulemaking is an opportunity for everyone to walk that fine line between flexibility and transparency and try to get it right.