Public Doesn't Mean Unrestricted: The Legal Edges of Redistributing Disclosure Data

Everything Quantgress scrapes is legal to collect and use for personal or research purposes. It's public government or public-domain disclosure data, most of it a U.S. government work with no copyright claim to begin with. That part was settled early. The harder question only shows up once you ask a second one: is it legal to hand that data to someone else, especially for money? "It's public" turns out to be the start of that question, not the end of it.

Contributor privacy survives aggregation, not existence

Corporate and PAC donation records come from OpenFEC, and OpenFEC's own API terms are permissive: build a commercial service on top, just attribute the source and don't misrepresent modified data as unmodified. That would be the whole story, except federal law adds a second, stricter rule that OpenFEC's terms don't override. 52 U.S.C. § 30111(a)(4) bars using FEC-disclosed contributor information "for the purpose of soliciting contributions or for commercial purposes."

The statute restricts exposing contributor identity, not the underlying fact that a company or PAC gave money. So api.py never serves the raw donor table, only corporate_donations_agg, rolled up by ticker and committee with contributor_name and sub_id dropped entirely. Quiver Quantitative ships the same shape publicly (ticker, total dollars, no individual names), which is the closest thing to a market precedent this has. It's a working assumption, not settled case law. The FEC's own commissioners have split on adjacent questions before, so it's treated as the safer default, not a guarantee.

A regulator's own terms of service can be stricter than the site's

FINRA's off-exchange short volume data looked, at first pass, like the same shape as everything else: a daily file, no login, framed on its own catalog page as posted "for public dissemination." But FINRA's site-wide Terms of Use separately ban bulk scraping and "developing or creating a database" from the content. FINRA's actual API Terms of Service go further, barring redistribution to anyone who isn't an "Authorized User" and specifically prohibiting "bulk distributor or service bureau" use. A public API reselling that data matches both restrictions about as directly as language like this gets.

Checking how the closest existing competitor handles this turned up nothing: no FINRA mention anywhere in Quiver Quantitative's terms of service or data-sources page. That's not evidence they've cleared it. It just means the answer isn't visible from outside. Until Quantgress gets a direct answer from FINRA, short_volume stays personal/research-use only, cut from any paid tier by default.

The pattern underneath both

Neither of these is "the government published it, so anything goes." One is a targeted federal privacy statute layered on top of an otherwise-open API; the other is a regulator's own contractual terms layered on top of an otherwise-public file. Public-domain status answers whether you can read the data. It doesn't answer whether you can redistribute it. Those turn out to be different questions with different sources of law, worth checking dataset by dataset rather than assumed once for the whole catalog.