The Bet Behind the Smaller Dataset

The CFPB’s 2026 Section 1071 rule answers the question in this series with a policy bet: less information at launch may produce a more reliable and less disruptive data regime over time. This final chapter connects the rule’s changes in institutional coverage, product scope, business definitions, data fields, demographics, and collection procedures, then asks what that bet sacrifices and how it can be judged. The conclusion is necessarily conditional. The supplied record supports the Bureau’s reasoning and its acknowledged trade-offs, but not a claim that the narrower regime has already improved lending or data quality.

The most important fact is that the CFPB is not treating every form of “less” as equally valuable. Some reductions remove lenders or products from the dataset. Some remove context from applications that remain covered. Some replace rigid collection procedures with flexible ones. The Bureau’s theory is that all of these changes serve one objective: begin with a population and information set that lenders can collect consistently, then use experience to decide whether expansion is justified. The opposing theory is that the first dataset will be least useful if it omits the places, prices, and groups where unequal access is hardest to see.

Less is a coordinated launch strategy, not one deletion

The first part of the bet is a smaller reporting population. The 2026 rule raises the threshold for a covered financial institution from 100 to 1,000 covered originations in each of two look-back years. It excludes Farm Credit System lenders and does not create a separate categorical exemption for community banks or credit unions. The practical effect is to concentrate the initial obligation in higher-volume institutions that the CFPB believes are more likely to have standardized systems, dedicated compliance resources, and enough activity to produce useful patterns.

The second part is a narrower product universe. The rule begins with loans, lines of credit, and credit cards viewed as core products. It removes merchant cash advances, agricultural lending, and transactions of $1,000 or less from the definition of a covered credit transaction. The Bureau says merchant cash advances have varied structures and evolving legal treatment; agricultural loans depend on distinctive collateral, weather, disease, and seasonal repayment; and very small transactions can be ancillary to deposit relationships and expensive to report relative to their informational value.

The third part is a narrower applicant boundary. A covered small business now has gross annual revenue of $1 million or less, rather than $5 million or less under the 2023 rule. The Bureau presents this as a simpler way to focus on businesses most likely to be truly small and to face credit barriers. But the line also removes some firms in the $1 million-to-$5 million range, including businesses that may be transitioning from startup to expansion. A simple definition reduces uncertainty for lenders; it does not perfectly identify which borrowers need transparency or affordable credit.

The fourth part is a smaller data payload. The CFPB removes the discretionary fields for application method, application recipient, denial reasons, pricing, and number of workers, along with LGBTQI+-owned business status. It retains the three-digit NAICS code, time in business, and number of principal owners because it views those fields as more integral to interpreting statutory information, identifying industry and startup patterns, or validating demographic submissions.

The fifth part is procedural. The rule removes the 2023 requirements to monitor low response rates and make peer comparisons as possible signs of discouragement. It also replaces rigid timing rules with a standard requiring collection at a time and in a manner reasonably designed to obtain a response. The statutory duty to request the information remains, as does the firewall that limits access to demographic data by people involved in credit decisions. The applicant’s right to refuse is made more prominent.

Seen together, these changes form a consistent design principle. The CFPB is reducing the number of reporting entities, the kinds of transactions, the size of the applicant universe, the amount of contextual information, and the prescriptive detail governing collection. It is not saying that each excluded element is worthless. It is saying that, at the beginning of a long-term system, marginal value must be compared with implementation complexity, response quality, privacy, applicant friction, and the possibility of reduced credit availability.

The Bureau’s cost analysis makes the incentive explicit. It estimates total quantified annualized cost savings relative to the 2023 baseline of roughly $176 million to $192 million using a 3% discount rate, or $178 million to $195 million using a 7% rate. It estimates annualized quantified costs of roughly $534,000 to $557,000 at 3%, or $649,000 to $676,000 at 7%. The largest ongoing savings come from institutions that would have reported under the baseline but will no longer be covered.