Editorial Standards & Methodology

Why This Document Exists

The Political Transparency Project ("PTP," "we") aggregates and publishes political and campaign-finance data drawn from primary public sources. Because we report on persons holding or seeking public office, donors to political campaigns, and the financial connections between them, our credibility depends on the rigor of our methodology and the transparency of our editorial process.

This document explains:

1. What we publish and what we don't
2. The standards we hold our content to (the "Three Tests")
3. How we source and verify claims
4. How we use artificial intelligence, and where we don't
5. How we handle corrections, retractions, and editorial disagreements
6. Who we are and how to reach us

This is a living document. Material changes are dated.

---

1. The Editorial Posture

PTP publishes in a non-partisan editorial voice. Both major parties are treated equally in scope, attention, and rigor. No editorial position will favor or disfavor any candidate or party on the basis of partisan affiliation alone.

We are, however, thesis-driven rather than flat. PTP's thesis at the publication level is:

Money, votes, and odds tell a more honest story than the narratives political actors tell about themselves. We surface the receipts; readers draw the conclusion.

This thesis governs the publication. It does not produce editorializing against any individual candidate. Per-candidate content is descriptive, sourced, and party-reversible (see § 2 below).

PTP does not engage in:

• Election advocacy (we do not endorse candidates, parties, or ballot measures)
• Campaign fundraising (we are not a vendor service to candidates and we do not facilitate donations)
• Voter contact (we do not contact voters on behalf of any campaign or cause)
• Donor prospecting (we do not provide donor lists or fundraising-target services)
• Investment advice (we are not a broker-dealer or investment adviser; politician trade data is reported for transparency, not as financial recommendation)

---

2. The Three Tests — Every Published Claim

Every published claim on PTP, whether on the website, in the API, in social-media posts, or in newsletter emails, must pass three tests before publication.

Test 1 — Descriptive, not accusatory

The claim must describe what happened (a vote, a donation, a disclosure) without characterizing motive, character, or intent. Examples:

• ✅ "Senator X received $50,000 from the pharmaceutical industry in 2025 and voted in favor of HR 2400."
• ❌ "Senator X is bought and paid for by Big Pharma."

We do not call any person corrupt, bought, crooked, or a scandal. We let the documented facts speak.

Test 2 — Primary-sourced

The claim must link to a primary public record. Acceptable primary sources include:

• FEC filings (campaign finance)
• Congress.gov (bills, votes)
• House Clerk and Senate Office of Public Records (lobbying, financial disclosures, STOCK Act PTRs)
• State election agencies (state-level campaign finance)
• Official government press releases and statements
• Court filings
• Federal Register and equivalent state agency notices

If the only available source is a news article or commentary, the claim is marked as secondary-sourced and the news article is itself cited.

Open Secrets data (CC BY-NC-SA) is used only for the public site, not the commercial API, consistent with its license.

Test 3 — Party-reversible

If we swap the party affiliation of the subject in our statement, would we still publish it? Would it still sound balanced?

• ✅ "Representative Y voted against the Affordable Care Act repeal" — works regardless of party.
• ❌ "Republican Representative Y voted for tax cuts that benefit the wealthy" — fails because the framing implies motive.

We rewrite to pass Test 3.

---

3. Sourcing Discipline

3.1 Primary-source linking

Every factual claim on PTP that is not patent (e.g., "Senator X is a member of the United States Senate") links to the primary source. Links open the underlying FEC filing, Congressional record, or other government document at its source.

3.2 Source-link archives

Where the underlying source URL is at risk of change or removal, PTP retains an archived copy and may link to a Wayback Machine archive in addition to the live source.

3.3 Document hash

Where chain-of-custody matters (e.g., revolving-door evidence, where it is built), PTP stores a hash of the underlying primary document so that any subsequent change at the source can be detected and noted.

3.4 Data lag and accuracy disclaimers

Public records are sometimes lagged by 45-90 days behind the underlying transaction (FEC filings, LDA filings, PTR filings have statutory deadlines that are often missed; we report when filings are missed). We mark publication of data with the date of underlying filing and the date of our ingestion. If the underlying record is later corrected or amended, we update accordingly and mark the revision.

3.5 Editorial cross-references

When making a claim that relies on a derived metric (such as the Money-Vote Gap, see § 5), we link to the methodology page describing how the metric is computed, the inputs that produced this specific instance, and any caveats.

---

4. The Money-Vote Gap (MVG) — Methodology Summary

The MVG is PTP's flagship analytical metric. It measures, per industry, the alignment between a politician's voting record and the position pattern of the donor industry that funded them.

A summary of the methodology — sufficient for an attentive reader to evaluate the metric — is published at poltrapro.com/methodology. Key features:

• 30-industry taxonomy drawn from FEC PAC categorization and standardized for cross-industry comparability
• $50,000 donor threshold below which industry alignment is not computed (no statistical power)
• Sample-size floor of 5 votes per industry-candidate pair (no MVG displayed if fewer)
• Chamber-median normalization so a 73% Yes rate is interpreted against the chamber's actual median Yes rate for that industry
• 4-band deviation display (significantly more often in favor / somewhat more / in line / less often)
• Confidence tiers based on number of supporting filings and consistency

The full methodology — including computation logic, edge cases, classification process, audit results, and known limitations — is on the methodology page.

---

5. Use of Artificial Intelligence

5.1 What AI does in PTP

PTP uses AI tools for specific assistive tasks in the publication process:

• Classification: tagging bills and votes by 30-industry taxonomy
• Extraction: parsing Lobbying Disclosure Act filings to infer donor positions on specific bills
• Drafting: initial drafts of candidate bios, summaries, and social-media post content
• Search and synthesis: ranking sources and surfacing relevant data points for editorial consideration

5.2 What AI does NOT do

• AI does not generate factual claims that are not anchored in primary-source data
• AI does not impute motive, character, or intent to any person
• AI does not write final published content that has not been reviewed and approved by a human editor
• AI does not engage in autonomous social-media posting — all PTP posts route through human approval
• AI does not perform political analysis that produces "takes" or opinion content
• AI does not generate synthetic audio, video, or images depicting any real person

5.3 Human approval gate

Every PTP-authored post, biographical narrative, summary, and social-media item passes through a human approval gate before publication. The approval workflow is documented in PTP's internal Content Operations Standards.

5.5 Limits and known issues

We are transparent that AI use introduces specific risks:

• AI may produce text that sounds confident but is factually wrong
• AI may amplify biases present in training data
• AI may miss nuance that a human editor would catch

PTP's response: human approval, source-linking discipline, audit sampling, and a published correction policy.

5.6 Disclosure

When AI is used to generate content that appears on PTP, we disclose this:

• The methodology page describes the role of AI in the publication process
• Each candidate's AI-generated summary is labeled as such
• Social-media posts include implicit disclosure (via methodology link in profile)

---

6. Corrections and Retractions

PTP commits to a transparent corrections process.

6.1 How to report an error

Email PolTraPro@proton.me with:

• The URL where the error appears
• A description of the error
• Your reasoning, including primary-source references where applicable

6.2 Editorial review

Upon receipt of a correction request:

• We acknowledge within 5 business days
• We investigate by re-examining the primary source and any underlying methodology
• We make the correction promptly if the error is verified
• If we disagree with the correction, we respond explaining our reasoning

6.3 Correction marking

Significant corrections are marked on the affected content with:

• A "Corrected" banner or footnote
• A summary of what was corrected
• The date of the correction

We do not silently rewrite history. If a correction is material, the prior state is preserved in an editorial note.

6.4 Retractions

If the underlying primary source is itself revealed to be erroneous (e.g., a candidate amends a financial disclosure or a vote tally is corrected), PTP updates and notes the underlying source change.

If a piece of editorial content is determined to be unsupportable (e.g., the methodology behind an analytical claim is shown to be flawed), we retract the content with a public note explaining the retraction.

6.5 Disagreements and external complaints

PTP welcomes substantive critique. We do not retract content because a subject is unhappy with accurate reporting of public records.

If a person believes content about them on PTP is defamatory, false, or misleading, the appropriate remedy is:

• A request for correction (per § 6.1)
• For legal disputes: pursuit of available legal remedies, subject to our editorial defenses (truth, fair-report privilege, fair-comment, anti-SLAPP)

We do not maintain a private removal queue. We do not take down accurate reporting on request.

---

7. Editorial Independence

7.1 No partisan funding

PTP does not accept money from:

• Political campaigns
• Party committees
• Aligned PACs or super PACs
• Single-issue ideological organizations
• Foreign governments or foreign-government-funded entities

We accept funding from:

• API customers (journalists, researchers, indie devs, advocacy organizations focused on process reform)
• Donations from individual readers (Ko-fi)
• Sponsorships from organizations whose work is consistent with our non-partisan posture (case-by-case)

7.2 No quid pro quo

We do not exchange editorial coverage for funding. Our API customer list and donor list do not influence which candidates are covered, what data is displayed, or how the MVG is computed.

7.3 Editorial firewall

Sponsorship or partnership arrangements, if any, are disclosed in this Editorial Standards page and on the relevant content. No sponsor reviews editorial content before publication.

7.4 Transparency about ownership

PTP is operated by PolTraPro LLC, a single-member Arizona LLC owned by Sean Ragan. We will disclose any change in ownership, governance, or funding structure that affects editorial independence.

---

8. Who We Are

8.1 Founding

PTP was founded by Sean Ragan in 2026. Sean serves as founder, editor-in-chief, and managing member.

8.2 About PTP

PTP is a small, founder-led organization. We aim to publish data and analysis that voters, journalists, and researchers can use as primary-source reference. We do not claim to be a substitute for working journalists, partisan analysis, or expert commentary. We are a complement: the source-linked, methodology-disclosed data layer that strengthens whatever further work others do.

8.3 Contact

Mailing address: Service-of-process address available on written request via PolTraPro@proton.me.

---

9. Affiliations and Disclosures

9.1 No paid partnerships in 2026

As of the date of this document, PTP has no paid partnerships, sponsorships, or commercial arrangements with any third party that would affect editorial coverage.

9.2 Use of competitor data

PTP uses public-record data from sources that include OpenSecrets, FollowTheMoney, Ballotpedia, and Congress.gov, consistent with each source's license. We are not affiliated with any of these organizations.

9.3 Membership and certification

PTP is not currently a member of:

• The Online News Association (we may apply)
• The Institute for Nonprofit News (we may apply)
• The Society of Professional Journalists (we may apply as a press organization)
• Any trade association

If PTP becomes a member of any press, journalism, or transparency organization, the membership will be disclosed here.

---

10. Press Exemption Reference

This document is published in support of PTP's claim to press exemption under federal and state electioneering communication laws, and to support PTP's position as a "newspaper, magazine, book, or other similar communication" within the meaning of 11 C.F.R. § 104.15(b) (the news exception to the FEC sale-or-use prohibition).

PTP regularly publishes news, commentary, and data on candidates and public officials in editorial form, applies the editorial discipline described in §§ 2-6, and operates as an independent publication consistent with traditional press functions. PTP is not owned or controlled by any candidate, party, or political committee.

This is not legal advice but a public statement of our editorial framework.

---

11. Changes to This Document

This document is updated as PTP's editorial practices evolve. Material changes are dated and summarized at the top of the document. The version history is preserved in PTP's public repository.

---

12. Acknowledgments

This statement draws on:

• Society of Professional Journalists Code of Ethics
• Online News Association editorial standards
• The Trust Project's indicators of news quality
• Standards developed by The Markup, ProPublica, and other transparency-oriented news organizations

We aim to meet these standards as a small, founder-led publication.

---