Paperwork Reduction Act — the hour counts that become the budget

The PRA section is how the Commission is allowed to ask the public for the information the forms collect. The issuer-count methodology is transparent and contestable. Startup: 2024 Reg D and Reg CF crypto offerings of $5 million or less → 99. Fundraising: 2024 Reg D, A and CF crypto offerings above $5 million and up to $75 million → 31. Safe harbor: 15% of 3,165 crypto projects launched in 2024 → 475 Form TRs under 400(b). Plus 24.75 Rule 200(e) TRs, 3.1 Rule 305(c) TRs, 7.75 Rule 305(d) TRs, and 130 new Form IDs.

Hours, as the Commission modeled them

• Form NOR: 99 responses, 3,960 company hours, $1.01 million. Three hours of notice is the tell — the 103 website is the real work, and it is in the 56-hour startup total, not in the NOR line.
• Form 1-CRYPTO: 31 responses, 16,679 hours, $3.53 million. 717.372 hours per filing, 75/25 internal/external split.
• Form 1-KC: 31, 13,950 hours, $2.95 million (600 hours).
• Form 1-SC: 31, 4,955 hours, $0.56 million (188.04 hours, 85/15).
• Form 1-UC: 31, 132 hours, $14,764 (5 hours, 85/15).
• Form TR: 511 responses, 11,180 hours, $2.37 million. 200(e) at 20 hours, 305(c) at 1.5, 305(d) at 20, 400(b) at 30.
• Form ID: +130 responses, +78 hours, $49,530.

Collections are mandatory for anyone who uses the exemptions. They are not confidential, except that an abandoned unpublished 1-CRYPTO stays non-public. There is no recordkeeping retention requirement layered on top — the forms are the collection. That last point matters: a comment asking the Commission to require issuers to keep the 103 website snapshot for five years is a new collection, and it has to be costed.

• The 75/25 and 85/15 splits are the Commission’s usual internal/outside-counsel assumptions. They are not a survey of crypto counsel rates.
• 717 hours for a 1-CRYPTO vs. 56 hours for a startup package is the disclosure-on-EDGAR vs. disclosure-on-a-website gap, quantified. Use it.
• 15% of 3,165 as the Rule 400 take-up is an assumption stacked on an assumption. If you run a foundation universe, give a better percentage.

• Contest 99 and 31 if you have a 2024–2026 census of crypto raises. The entire PV table moves with those two integers.
• Contest 717 hours if you have actual 1-A hours for a crypto issuer. CrowdCheck and DealMaker can win this RFC with invoices.
• Contest 30 hours for a 400(b) TR. A supporting analysis against a 103(b)(1) register is a legal memo, not a 30-hour form. 80–120 hours is a more honest opinion-practice number, and it raises the PV.
• If you want a website-snapshot retention rule, cost it. Otherwise PRA will kill it in the final.