Posts
The Consent Machine
Eleven Lines That Changed the Rulebook
On March 7, 1996, President Clinton signed the National Technology Transfer and Advancement Act of 1995 into law. Public Law 104-113. Section 12(d) is the provision that matters for this series. It is eleven lines long. It reads, in operative part:
The framing of the statute is efficiency. Government was writing thousands of its own technical specifications — Military Specifications (MIL-SPECS), Federal Acquisition Regulations appendices, agency-unique testing protocols — that duplicated work private industry had already done. The cost was real. The DoD alone estimated hundreds of millions in savings from transitioning to commercial standards. The argument was: why should taxpayers pay government engineers to write steel specifications when ASTM already wrote them?
The argument was correct as far as it went. It did not go far enough to ask the question this series asks: when you make the private standards mandatory, what have you done to the private organizations that produce them? You have given a private membership club — funded by its members, governed by its members, accountable to no one outside its membership — the force of federal law. The efficiency gain is real. The accountability question was not asked.
Before NTTAA, a private standard was a recommendation. After NTTAA, the same document, written by the same private committee, funded by the same industry membership dues, became the mandatory technical baseline for federal procurement, federal regulation, and — through subsequent state adoption — the built environment of the entire country. The document did not change. Its legal status did.
OMB Circular A-119's implementation guidance is explicit about what the mandate accomplishes: it reduces the cost of developing standards, decreases the cost of goods procured, and supports international trade. All of this is accurate. The circular does not address what the mandate does to the internal governance of the private bodies that now hold the rulemaking function. That question — what happens inside the machine when the stakes change — is the subject of this post.
How Consensus Is Manufactured
The ANSI Essential Requirements are the procedural constitution of the Standard Architecture. Any organization seeking ANSI accreditation — and the legitimacy that accreditation confers — must satisfy them. They require openness (any materially affected party may participate), balance (no single interest dominates), due process (defined procedures, documented decisions, appeals rights), and consensus (broad agreement, not unanimity, but something more than a simple majority). These are genuine procedural commitments. ANSI audits compliance. Appeals are heard. The process is not theater.
What the process is, however, is expensive to use — and that cost is not acknowledged in the procedural requirements.
The structural result is predictable from these numbers alone, without any assumption of bad faith. The organizations that can absorb the cost of sustained committee participation are the organizations with the most at stake in the outcome — which are, systematically, the producers whose products the standards will certify and whose competitors will be screened by the compliance threshold. A fire suppression system manufacturer has clear incentive to maintain committee presence through multiple revision cycles. A renter living in a building governed by fire codes has no mechanism to do so.
This is not corruption. It is selection — the same selection pressure that operates in any system where participation is nominally open but practically expensive. The ANSI balance requirement does not eliminate that pressure. It attempts to counteract it after the fact, by requiring that no single interest dominate. But "balance" and "equal capacity to shape outcomes" are not the same thing, and the Essential Requirements only address the first.
Who Is Actually in the Room
ANSI accreditation requires that standards committees maintain balance among interest categories. ANSI defines these differently for different bodies, but the core categories are consistent: producers (those who make the regulated product or material), users (those who use it in commerce or construction), general interest (academics, government representatives, insurance professionals, testing laboratories), and sometimes a separate consumer category. No single category may constitute more than one-third of a committee under strict balance requirements.
The formal numbers often comply. The functional reality is that "balance" between categories does not mean balance in technical depth, institutional memory, or sustained engagement — the three things that actually determine whose preferences appear in the final document.
The producer category is typically populated by employees of manufacturers, trade associations, and industry groups whose products are directly affected by the standard being written. These participants are employed to engage with the standards process. Their employer bears their travel cost. They have institutional memory across multiple revision cycles. They arrive at committee meetings with technical staff support and pre-drafted proposals.
The user category in construction-related codes is often populated by contractors, engineers of record, and AHJs (Authorities Having Jurisdiction — local building officials). These participants have real expertise and genuine stakes. They are also typically small businesses or local government employees whose participation requires personal time investment and, often, personal expense.
The general interest category is where the structural asymmetry is sharpest. Academic representatives typically serve on fewer committees and rotate more frequently, losing institutional memory. Insurance representatives have aligned interests with producers on many questions (standards reduce claims) and divergent interests on others. Testing laboratories — often including UL itself — occupy general interest seats while having direct financial interests in the testing requirements the standard will mandate.
The consumer category, where it exists as a separate classification, is the thinnest. Consumer advocacy organizations lack the funding to maintain sustained presence across hundreds of technical committees in multiple standards bodies simultaneously. Individual consumers have no practical mechanism for engagement. The "consumer interest" seat is often filled by a single person serving on multiple committees, against a producer delegation that may arrive with five or six technical specialists for a single committee.
What the balance requirement actually produces: Formal compliance with the one-third threshold, genuine representation from sophisticated institutional actors in all categories, and a systematic gap in the representation of diffuse public interests — the interests of people who live in buildings, use products, and breathe air but are not employed by anyone to care about how the rules are written.
The question is not whether industry experts dominate standards committees. They do, openly, by design — because technical expertise is the stated basis of the system's legitimacy. The question is whether technical expertise and producer interest are the same thing. They overlap significantly. They are not identical. The gap between them is where standards get written.
Public Participation As Architecture
Every ANSI-accredited standards development process includes a public comment period. Proposed revisions are published — some freely accessible, others requiring document purchase to understand in context — and any person may submit a written comment. The comments are received, logged, and addressed. The committee must document how each comment was resolved. If a comment is rejected, the rejection must be explained. The appeals mechanism allows a persistent commenter to escalate.
This is a real process. It is not theater. It has, historically, produced real changes in proposed standards when technically well-supported comments identified genuine problems. The question is not whether the process functions. The question is whether it functions as a democratic check on private rulemaking — which is the claim it implicitly carries.
The structural barriers are specific and documented. To write a substantive comment on a proposed revision to NFPA 70, a participant needs: the current edition of NFPA 70 (purchased), the draft revision text (available, but requiring navigation of the NFPA website and often conditional on account creation), sufficient technical background to understand what is being changed and why, and enough time to write a technically coherent objection or proposal. The comment must engage the technical substance. A comment that says "this seems wrong" carries no procedural weight. A comment that says "the proposed amendment to Article 210.8(A)(5) would fail to address arc fault scenarios in multi-wire branch circuits because..." requires the knowledge base of a working electrician or electrical engineer.
The cycle: The 2023 edition of NFPA 70 was developed through a revision cycle that began in 2020. The process involved 19 Code-Making Panels reviewing proposals and public comments across approximately three years of committee work, public input periods, and balloting.
Who participates: NFPA reports that its standards development process involves tens of thousands of volunteer participants across its full portfolio. The breakdown of those participants by interest category — the balance data — is reported at the committee level but is not aggregated into a publicly accessible analysis of system-wide producer vs. public interest representation ratios.
What the comment record shows: The public input period for a major code like NFPA 70 receives thousands of proposals and comments per cycle. The majority originate from industry participants — contractors, manufacturers, trade associations, testing laboratories — who have the technical vocabulary and institutional capacity to engage the process at the required level. Academic and consumer-interest comments are present but a small fraction of the total.
What this means: The comment record is a market for technical influence, and like all markets, it favors participants with resources to invest. The process is open. The outcome reflects who showed up prepared. The two facts are compatible — and their compatibility is the design of the system, not a failure of it.
Why the Cycle Runs Every Three Years
The NFPA updates NFPA 70 on a three-year cycle. ASTM updates individual standards on schedules ranging from annual to five-year review cycles. UL updates product safety standards on cycles tied to technology change and incident data. These cycles are described, by the organizations that operate them, as responsive to technical change — a way of keeping rules current with evolving materials, construction practices, and safety knowledge.
This is accurate. It is incomplete. The three-year cycle also structures the revenue model of organizations whose primary funding source is the sale of standard documents, training materials, certification programs, and the professional development infrastructure that each new edition generates.
The organizations operating these cycles are, formally, nonprofits. NFPA is a 501(c)(3). ASTM is organized as a nonprofit. UL operates as a nonprofit safety science organization. Their nonprofit status is genuine in the legal sense — they do not distribute profits to shareholders. It does not mean they are indifferent to revenue. It means their revenue funds their operations, including the expert staff and volunteer committee infrastructure that produces the standards. The revenue model and the mission are not in conflict. They are, however, in tension — and the tension is visible in the revision cycle structure.
A standard that was technically stable and needed no revision would, under a strictly need-driven model, not be revised. A stable standard generates one document sale per jurisdiction adoption. A standard revised every three years generates a document sale per jurisdiction per cycle, plus a training market, plus a certification update market. The institutional incentive for regular revision is structural, regardless of whether any individual within the organization consciously prioritizes revenue over technical accuracy.
The NFPA does not revise NFPA 70 every three years because electrical systems change significantly every three years. It revises NFPA 70 every three years because that is the cycle. The cycle funds the organization. The organization is genuine in its safety mission. The cycle is also its business model. These facts coexist. The series documents their coexistence without adjudicating which is primary.
Efficiency by Structural Choice
The Standard Architecture is often described, by its defenders, as the least-bad option in a world of technical complexity. Congress cannot write electrical codes. Agency engineers cannot maintain expertise across thousands of product categories. The private consensus process, whatever its limitations, harnesses genuine expertise and produces rules that are more technically sophisticated than anything a purely political process could generate. This is true.
What the "least-bad option" argument does not address is that the design of the system — the specific choices about committee composition, funding models, revision cycles, document pricing, and participation mechanics — reflects choices made by the organizations that benefit from the design. The ANSI Essential Requirements were written by ANSI. The balance requirements were determined by the organizations subject to them. The funding models were developed by the organizations that collect the revenues. The revision cycle structure was established by the organizations that sell the documents.
None of this is concealed. It is simply not the subject of regular examination. The Standard Architecture functions as invisible infrastructure — like electrical wiring inside the walls it governs. You do not think about who designed the conduit until something fails. This series is the examination that precedes the failure analysis.
| System Feature | Official Rationale | Structural Effect | Who Designed It |
|---|---|---|---|
| Balanced committees | Ensures diverse stakeholder input | Formal balance; functional producer advantage due to participation cost asymmetry | ANSI (private, member-governed) |
| Document sales model | Funds expert volunteer committees | Creates paywall on incorporated-by-reference law; barriers to public comment | SDOs (NFPA, ASTM, UL — member-governed) |
| Three-year revision cycle | Keeps standards current with technology | Generates recurring revenue independent of technical need; burdens small-jurisdiction adopters | NFPA (private, member-governed) |
| Incorporation by reference | Avoids duplicative government standard-writing | Transfers rulemaking function to private bodies; removes democratic accountability checkpoint | Congress (NTTAA 1995) + OMB (A-119) |
| ANSI accreditation | Ensures procedural due process | Provides legitimacy cover for the system; ANSI itself is a private member-governed body | ANSI (private, member-governed) |
The table above is not a finding of bad design. It is a description of whose choices produced the current design — and therefore whose interests the design reflects. The answer, in every row, is: the organizations that operate within and benefit from the architecture. The one exception is NTTAA — a congressional act — and even there, the legislative record shows that the private standards community was the primary advocate for the mandate that made its work product binding on the federal government.
What This Post Establishes
The NTTAA is the structural pivot. Before 1995, private standards were influential. After 1995, they were mandatory — for the federal government and, through downstream adoption, for the entire built environment. The eleven lines of Section 12(d) are the legislative mechanism by which a private system of expert deliberation became, by statutory mandate, the default rulemaking infrastructure of the United States. The efficiency argument was real. The accountability question was not asked at the time. This series asks it.
The consent is procedurally real and functionally asymmetric. The ANSI process produces genuine expert deliberation. It also systematically advantages participants with institutional resources — employers who cover travel costs, technical staffs who support proposal drafting, trade associations that coordinate positions across committee seats. The formal balance requirement addresses the first dimension of this asymmetry. It does not address the second. The consent that emerges from the process is real. It reflects the composition of the room. The composition of the room reflects who can afford to be there.
The revenue model and the rulemaking function are in structural tension. Organizations that fund expert committees through document sales have institutional incentives for revision cycles that are regular enough to generate document revenue, regardless of whether the underlying technical landscape requires revision at that frequency. This is not a claim that revisions are manufactured. It is a claim that the incentive structure does not require anyone to consciously manufacture them — the system produces revision pressure through its financial design alone.
The design reflects the designers. Every structural feature of the consent machine — committee balance rules, document pricing, revision cycles, ANSI accreditation requirements — was designed by the private organizations that operate the system and benefit from its design. This is not concealment. It is simply the natural outcome of a system in which the regulated community was asked to design its own regulation, and did so competently and in good faith, while also designing it in ways that serve the regulated community's interests. Understanding whose interests the design serves is prerequisite to evaluating whether the design is adequate.
Next: Post III · The Committee Room goes inside a specific technical committee — the composition, the meeting structure, the proposal process, and the documented patterns of whose preferences survive the revision cycle and whose do not. The architecture in practice.
