Annual federal IV-D distribution · FY2024
$26.7B
distributed by HHS OCSE to states in FY2024 to run the family-court enforcement machine — flowing through every contested custody case as fee-paid appointments, billable continuances, ordered evaluations, and revenue-tracked support arrears.
- Non-primary custody avg
- 18%Census SIPP 2022
- IV-D per child in case / yr
- $2,182$26.7B ÷ 12.2M kids
- Judges of record
- —Loading…
- IV-D collected, all states / yr
- $26.7BHHS OCSE FY2024
Explore the Map
See IV-D revenue, arrears, custody patterns, and judge data across all 50 states. Click any state to drill into counties and individual courts.
Open heat map →
Submit Your Case
Your case data is anonymous and builds the statistical record. Every submission adds to the pattern proof that courts cannot ignore.
Submit anonymously →
Find Your Judge
Search every verified family court judge in the database. See their court, role, and whether other users in your county have submitted cases involving them.
Search judges →
Pattern analysis updates automatically as cases are submitted. Charts below require a minimum of 10 cases per judge and 30 for chi-square significance.
Custody distribution vs Census benchmark
Days parents separated from children
Gag orders per judge — annual estimate
Pro se vs represented — avg custody %
Income imputed vs actual — by state
Continuances per case — billing signal
⬡ MAP
Heat map:
IV-D Revenue
LowHigh
Select a complaint type, your state, and who the complaint is about. The AI will draft a complete, ready-to-file letter and show you exactly where to send it, what to attach, and what to expect.
Select your state to see your representatives and their contact info. The AI drafts a letter using your state's IV-D financial data and case context, pre-populated with the specific numbers that matter.
This section presents mathematical proof that family court outcomes are predetermined by template — not determined by the facts of individual cases. Each analysis uses statistical methods that courts, legislators, and journalists can verify independently.
Based on real submitted cases in your state. This is a statistical estimate — not a legal prediction. Use it to plan, budget, and understand what you are likely to face.
Your case details
State cost comparison — median contested case
Zero-knowledge submission. We collect no name, email, IP address, or identifying metadata. Your token is generated in your browser — we store only a mathematical hash. For maximum anonymity use Tor Browser from tails.boum.org.
How your submission appears publicly
Judges and attorneys are public officials — their names and courts stay visible so outliers can be identified. Your submission is never shown as a standalone case — it only aggregates into per-judge, per-attorney, per-state, and per-issue statistics. Per-judge stats require 10+ submissions before display, so no single submission dominates.
Small-jurisdiction note: if your county is small (few existing submissions on record), your submission is automatically released in a larger, less-frequent batch — timestamps can't be used to single you out. No action needed from you.
Why your submission won't appear today.
New submissions are queued for 1 to 3 weeks (randomized per-submission) and released in batches — then bucketed to week-granularity when published. This breaks same-day correlation between a court event and a submission, and the 3-week tail makes targeted deanonymization much harder. You can always view, edit, or delete your entry with your token while it's pending.
🔑
Write down your token the moment it appears.
When you submit, you'll get a one-time token. That token is the only way to edit, update, or delete your submission. The zero-knowledge design means we cannot look it up for you, recover it, or identify which submission is yours without it — ever. Write it on paper, or take a photo of the paper. Do not store it only in a digital note that could be lost with a crashed device or cleared cookies.
🔒 Optional: verify with your court order
Why upload? A verified submission carries more weight in the aggregate data and gets a ✓ badge.
Our AI reads the order, extracts only public/statistical fields (judge, county, custody split, support amount, flags) and auto-fills the form below.
We never store the file. Names of parties, children, attorneys, addresses, and case numbers are stripped before anything is saved — the AI is prompted to refuse them. A cryptographic fingerprint (not the file) is kept to stop the same order from verifying multiple submissions.
Before custody court opened
State-side designation & Title IV-D
Case basics
Hearing master / magistrate / recommending officer
In many states, a non-judge officer (magistrate, hearing master, referee, special master, support magistrate, probation officer) holds the hearing, writes findings, and recommends an order. A judge then signs it — sometimes without reading it. If that's how your case was handled, fill this out. Otherwise skip.
Timeline — delays & days without children
Parenting schedule & custody
Financial — income, support, arrears
Time lost to court appearances
Court doesn't just cost attorney fees. Every hearing takes you off the clock.
Enter how many full and half days you've missed for this case and how you covered them — we'll estimate the dollar cost based on the monthly income you already entered above, and surface it in aggregate so the true cost of a family-court case is visible.
Of those court days, how did you cover the time away from work?
Enforcement, orders, gag orders
ACF / IV-D Benefit Programs (payors): Federal law requires states to offer alternatives to enforcement. Did the court or agency offer — and did you get to use — any of the following?
Mediation / out-of-court settlement
Court-appointed professionals
Campaign finance & conflicts
Description & impact
Judges are public officials. Their rulings, outcomes, and conduct in official capacity are public record. Campaign finance data is from public state election records. The corruption index is a composite of eight weighted pattern indicators. Community reports document patterns — not personal attacks. Official judicial conduct commission links are provided for each state.
Attorneys listed are licensed professionals subject to state bar oversight. Bar numbers and verification links are included. Bias scores reflect submitted pattern data. Campaign finance records are public. Submit bar complaints through official state bar links provided.
Court appointees represent mandatory costs no family can refuse. Tracking who orders them, AFCC memberships, recommendation patterns, and costs reveals the financial ecosystem. Add appointees from your case below.
Add an appointee from your case
Appeals expose judicial error rates and the financial barrier to correcting them. Transcripts alone run $1,800–$5,200. The system is designed to make error correction prohibitively expensive for average families.
Outcomes
Average cost by component
Submit an appeal record
AFCC connects judges, attorneys, GALs, evaluators, and parenting coordinators who refer, appoint, and testify for each other — creating closed financial ecosystems billing $40,000–$80,000 per family. Membership is public record.
Campaign finance — attorney donations to presiding judges (public record)
The clearest paper trail of conflict between judges and the attorneys who appear before them is campaign contributions. In the 39 states that elect some or all family-court judges, every donation is disclosed — but the records live in 50 different state portals, not in FEC. The table below fills as submitters contribute data. For a direct lookup, use the state portal links in your state's profile page.
Why this table is sparse right now: FEC (federal) does not track state judicial races. State-level campaign-finance data requires per-state portals (TX Ethics Commission, OH Secretary of State, FL Division of Elections, etc). Two ways to populate it:
· Submitters: in your case submission, note whether your attorney or opposing counsel donated to the presiding judge — those entries show up here
· FollowTheMoney.org integration: when you open any state profile (state.html), judicial campaign-finance data is pulled from the National Institute on Money in Politics
· Submitters: in your case submission, note whether your attorney or opposing counsel donated to the presiding judge — those entries show up here
· FollowTheMoney.org integration: when you open any state profile (state.html), judicial campaign-finance data is pulled from the National Institute on Money in Politics
AFCC ecosystem map
AFCC judges — appointment rates per case type
AFCC vs non-AFCC appointees — % recommending mother primary
In 39 states, at least some family-court judges are selected by election. Their campaign donors are often the same attorneys who appear in their courtrooms. Vote-them-out is the single most direct accountability tool available in those states.
Lobbying money — who funds the status quo
Generate a voter guide for your county
How judges are selected, by state
Judges in database with an inferred next election
Pick your state to see judges with an inferred next election year. Year is derived from when they took office plus the state's term length — actual ballot date depends on whether the seat is elected, on retention, or up for reappointment.
Track bills that could fix the system — and who is blocking them. Use the contact buttons to reach your representatives directly.
Every number here represents a family. This section tracks what happens to children — not just parents. Including a separate submission form for adults who were children in contested custody cases.
Children's stated preference vs court order
Days parents separated from children — before any order
Primary reason for separation from children
Foster care pipeline — CPS to adoption with federal revenue
Average days without children — by state
Adult children submission — were you a child in a contested custody case?
Your voice is the most powerful data on this platform. Courts dismiss parent concerns as self-interested. They cannot dismiss the voices of the children those proceedings affected. Same zero-knowledge architecture — your identity is never collected.
Adult child accounts submitted
The financial architecture of family court is structural. Every professional is paid per appointment. Every continuance is a billing event. Every order is a revenue node. This section maps total extraction per family.
IV-D collections distributed by state — top 20
Arrears owed by state — top 20
National IV-D money flow — admin spend vs collections, 8 fiscal years
Current-support distributed vs arrears backlog
Federal incentive pool estimate by state — top 15
How judges, hearing officers, and appointees earn bonuses
Every professional in family court has at least one compensation mechanism that rewards volume or outcomes. None of this is secret — it's written into federal law, state budgets, and appointment contracts. The problem isn't that these structures exist; it's that the submitter rarely knows which ones apply to the officer deciding their case.
How the federal government pays states to run child-support cases
Plain version: the federal government runs a yearly bonus pool (~$534M in FY2024) and splits it among states based on 5 scorecards. The better a state scores, the bigger its check. This page shows each scorecard, the rules it uses, and where every state currently stands.
Why this matters to you: every scorecard is a reason your state pushes cases a certain way. States aren't just trying to collect support — they're also trying to hit specific bureaucratic targets that earn them federal money. When a case feels aggressive, automated, or inflexible, this is often the reason why. You're looking at the incentives behind the behavior.
How to read each card below:
Why this matters to you: every scorecard is a reason your state pushes cases a certain way. States aren't just trying to collect support — they're also trying to hit specific bureaucratic targets that earn them federal money. When a case feels aggressive, automated, or inflexible, this is often the reason why. You're looking at the incentives behind the behavior.
How to read each card below:
- Rules table — how federal law converts a state's performance on this measure into an "applicable %" (100% = max bonus, 0% = no bonus).
- Colored strip — every reporting state dropped into its current tier. Teal = top tier (maxing this measure), amber = mid-tier, orange/red = below floor.
- Hover the colored segments to see which states are in each tier.
Every state, every measure — side-by-side
Same data, different view. Each row is one state; each pair of columns is a measure. The left column in each pair is the state's actual number; the right is the applicable % that number earns. Green = maxing the measure. Red = below the floor, earning zero on that measure.
States graded on actual outcome data from submitted cases and official government sources. Click any state for detailed breakdown.
Compare states side-by-side
Pick up to 5 states to filter all charts below. Leave empty to see all 50.
Reading the bands — IV-D current-support-collection reward tiers (42 USC 658a)
below 50% = $0 incentive
50-60% floor = 50% applicable
60-70% = 60-80% applicable
70-80% = 80-100% applicable
80%+ = 100% MAX
The colored bands behind each chart show which IV-D incentive tier a state sits in. Hover a state bar to see the state's exact rate + tier. Click any chart to expand it full-size.
Reading the bars: bar color matches the IV-D applicable-percentage tier the state actually earns on that measure — red = below floor ($0), orange = 50% floor, gold = 60-80% applicable, lime = 80-100% applicable, teal = 100% MAX. In FY2024, only 1 state (ID) hit 80%+ on current-collection; most sit in 60-70% (earning 60-80% applicable — not 100%). Order establishment is where almost every state maxes out (85%+ is typical). That uneven pattern across measures is exactly how the incentive pool gets distributed.
IV-D measure #3 — Current Support Collection Rate
IV-D measure #4 — Arrears Collection (cases paying / cases with arrears)
IV-D measure #1 — Paternity Establishment %
IV-D measure #2 — Support Order Establishment %
IV-D measure #5 — Cost-Effectiveness Ratio ($ collected per $1 spent)
Custodial-mother share by state — Census P60 vs submission-derived
OCSE / ACF multi-year fraud indicators
ACF vs Census coverage gap
The evader-list paradox — public shaming vs "100% collection" claims
Full state report card
| State | Grade | Collection rate | Cases with arrears | Caseload | $ collected / $ spent | IV-D ($M) |
|---|
Compute a cryptographic fingerprint (SHA-256) of any document — entirely in your browser. The file never leaves your device. Save or email the resulting proof bundle to preserve a record of the document's exact byte-for-byte content at a specific moment. Later, anyone can re-hash the same file and confirm it matches the record.
Document Fingerprinting — local SHA-256 proof
Pick any document. We compute its SHA-256 hash in-browser using the Web Crypto API — nothing is uploaded. A proof bundle containing the hash, filename, size, and the moment you generated it becomes a portable record you can email to yourself, your attorney, or a notary. If the document is altered, the hash will differ — re-running the hash later proves the bytes changed. The strength of the timestamp comes from where you preserve the bundle (email headers, a court filing, a lawyer's file) — not from this website.
Your saved document fingerprints
Storage: Entries here are stored only in this browser's local storage. Clearing browser data will erase them. For durable records, always download the proof bundle and email or archive it.
Verify a file against a saved hash
Already have a proof bundle (from yourself or someone else)? Upload the document and paste the original hash — we'll tell you whether the file matches byte-for-byte.
What this actually protects against (and what it doesn't)
Genuine use cases
- Tampering detection: If anyone modifies a document even by a single byte, its SHA-256 will differ. A preserved earlier hash proves the change occurred.
- Income documentation: Hash pay stubs + W-2s when you receive them and email yourself the bundle. If a court later imputes a different income, the archived email shows the documents were in your possession and unchanged as of the email's server-stamped arrival date.
- Court orders: Hash every order you receive. If an order is later claimed to have been amended, the archived hash proves the original text was different.
- Screenshots of communications: Hash texts / emails while fresh, then email the bundle to a fixed address. The email headers become a third-party time witness.
- CPS reports: Hash received copies immediately. Creates a record that fixes the document's state at the moment it was generated.
What this does NOT do
- It does not prove a document is authentic. SHA-256 proves the bytes haven't changed — not that the document itself is real, unforged, or legally valid.
- It does not create a legally admissible timestamp on its own. The timestamp you see above is generated by your own browser's clock. To establish a credible date, you must preserve the proof bundle somewhere an adversary can't alter — typically your own email inbox (the provider's server-side arrival time becomes the witness), a certified-mail envelope, or a notary.
- Automatic blockchain anchoring is opt-in. Nothing leaves your browser unless you click Anchor to Bitcoin — that button submits only the hash (never the file contents) to public OpenTimestamps calendar servers, which aggregate the hash into a Merkle root that gets published to Bitcoin within 6–24 hours. Without clicking it, no network request is made and the timestamp is purely local.
This platform is a press organization. Our data is available for journalistic use. We have had our methodology reviewed by press freedom counsel. All data derives from community submissions and official public records. Publishing analysis of public officials' official conduct is protected First Amendment activity.
Data downloads
Scope your story — national vs state vs county
Pick any state (and optionally a county) and see the nationwide numbers, state numbers, and county numbers side-by-side. Use this to ground a local piece in national context — or to show how far a single state/county deviates from the national pattern.
Story angles — click any to expand with the data
Expert contacts for journalists
Press contact & inquiry form
Media inquiries
If you're a journalist investigating a family-court pattern — judge, attorney, appointee, state program, or any of the statistical claims below — send your inquiry via the form. We answer within a few days.
Alternative channels: if you prefer to use an encrypted channel first, mention that in the "reply" field (Signal, ProtonMail, Matrix, etc.) and we'll set up a keyed session before sharing any source-sensitive material.
Data provenance & legal framework
Sources: All statistical claims trace to (a) community submissions anonymized via token-only architecture, (b) HHS OCSE Preliminary Data FY2017–FY2024, (c) FEC federal contribution records, (d) state judicial commissions' public filings, (e) state campaign-finance portals, (f) Congress.gov official records. Source URLs are linked on each relevant page.
Verification: Every chart has a methodology trail — click "Expand" on any chart to see the exact data behind it; use the downloads above for the underlying rows.
Published as protected journalism. Analysis of public officials acting in their official capacity is First Amendment-protected activity (Near v. Minnesota, 1931; New York Times v. United States, 1971; Hustler v. Falwell, 1988). This platform does not publish private-party names, addresses, case numbers, or children's information.
Every external data source used by Unsealed. For each: what the site shows from it, how often it refreshes, what identifying data (if any) is transmitted when the user triggers the call. Nothing here fires automatically — these only run when you click a tab/action that needs them.
Data sources the server fetches on your behalf (via PHP proxies)
| Source | Server endpoint | What it populates | Status |
|---|---|---|---|
| CourtListener Public court bios + disclosures |
/api_courtlistener.php24h SQLite cache |
Judge panel "CourtListener bio" card: DOB, birth city, religion, gender, political affiliations, judicial positions history, financial disclosures filed. Fetched when you click a judge profile. | live |
| FEC Federal campaign donations |
/api_fec.php24h cache |
Judge panel "FEC federal donations" card: same-name donation records, amount + committee + date. Flagged as same-name matches (not verified same person). | live |
| FollowTheMoney State campaign finance |
/api_ftm.php24h cache |
Judge panel "FollowTheMoney state judicial campaigns" card. Covers elected state judges; empty for purely-appointed judges. | live |
| Congress.gov API v3 Federal delegation + bills |
/api_congress.php24h cache |
Contact Representatives tab: live senator + house member list per state with party + district. /state.html profiles. Feeds the AI letter-drafting prompt with real rep names. | live |
| OpenStates v3 State legislators + bills |
/api_openstates.php24h cache |
Contact Representatives tab: state legislature upper + lower chamber rosters per state. | live |
| Census ACS Single-parent households |
/api_census.php24h cache |
State Grades "ACF vs Census coverage gap" panel. Ratio = ACF caseload ÷ Census single-parent HH count. Flags states pulling in married/dual-custody families. | live |
| Evader scrapers State "most wanted" registries |
/api_evaders.php24h cache |
Evader-list paradox panel in State Grades. Live-scrapes Texas; seed/manual counts for 14 SPA-rendered states (CA, FL, VA, MA, etc.). | live |
| Anthropic Claude Letter drafting + order OCR |
/api_ai.php/api_order_verify.phpno cache (per-request) |
(1) Complaint / rep-letter drafting — your prompt + ACF stats + judge record go to Claude. (2) Order-upload OCR — your PDF sent once, Claude returns public fields ONLY (no party names), PDF never stored server-side. | live |
| Submission API User-submitted cases |
/api_submit.php · /api_token.php · /api_stats.phpSQLite |
Accepts zero-knowledge case submissions, emits tokens, aggregates anonymized stats, allows token-based edit/delete. | live |
Static data files on the server
| File | Source of truth | What it feeds |
|---|---|---|
acf_state_data.json | HHS OCSE Preliminary Data FY2017–FY2024 (four Excel workbooks ingested locally) | Heat map, State Grades, /state.html, rep letter prompts, cost-calculator baseline |
judges_verified.json | Every entry verified against the court's own official directory (linked per-row) | Map pins, judge autosuggest, complaint cascade (state→county→judge), /state.html, judicial elections tab |
state_evader_lists.json | Each state's own child-support enforcement agency in April 2026 | Evader-list paradox panel |
state_evader_counts.json | Manually verified seeds where the page is a JS-rendered SPA; scraped live for Texas | Paradox math tile (publicly named ÷ state's own implied non-payers) |
appeals_courts.json | Each court's official website (50 state supreme courts + intermediate courts + all 13 US Circuit Courts) | Amber hex pins on the map, Appellate courts block on /state.html |
legislation_feed.json | Server-side cron populated via Congress.gov API | Legislation ticker + Legislation tab. Currently empty pending cron setup. |
Not collected / not sent
Confirmed absent: Google Analytics, Plausible, umami, hotjar, mixpanel, Segment, Meta Pixel, TikTok pixel. No cookies set. No fingerprinting (canvas, WebGL, font enumeration). No IP address logging (nginx access_log off globally). No user-agent logging. No referrer retention.
What a visit actually touches: the HTML (one request to your own server) + vendored libraries d3/topojson/Chart.js (served from /lib/ on your own server, not from Cloudflare CDN anymore). No third-party origin is contacted unless you click a tab that fires a proxy call — and even then, only the server-side PHP makes the external call, not your browser.
Known holes (on purpose)
- Press contact scraping: asked about earlier — not wired anywhere yet. Nothing currently populates a "media contacts" field.
- Per-county per-metric heat data: ACF only publishes state-level aggregates. County heat coloring therefore inherits the state value; it's not a county-specific number.
- Cost calculator: refuses to compute until a state has ≥5 real verified submissions. Currently shows the state's ACF baseline instead of a projected number.
- Foster care counts, gag order counts, custody-split %: removed from heat map pills. No verified public source at state granularity.
Warrant Canary — Updated weekly
As of the date below, this project has received NO subpoenas, national security letters, court orders, legal demands for user data, or gag orders preventing disclosure of such demands. We have not been compelled to modify our data collection practices.
Why we cannot hand over your data under subpoena
This platform uses zero-knowledge architecture. We are technically incapable of identifying who submitted what.
- We do NOT collect: name, email, IP address, device fingerprint, cookies, browser headers, or any identifying metadata. Our server discards all request logs before writing to disk.
- We DO collect: the statistical record you fill out — state, year, financial data, outcome data, court name, judge name, attorney name (official capacity), and a one-way hash of your submission token.
- Submission tokens: generated in your browser, shown once, never stored. Only a mathematical hash is stored — impossible to reverse.
- Subpoena response: "We have no IP logs, no user accounts, no names, no identifying metadata. The database export is the entirety of our data. No responsive records identifying individuals exist."
How to submit with maximum anonymity
- Tor Browser (torproject.org) — strongest option. Routes through multiple servers, prevents IP identification.
- Mullvad VPN — paid with cash, no email required. Strong second option.
- Public WiFi — library or coffee shop you rarely visit. Do not use home or work network.
- Description field: Do not include details that uniquely identify your case. Judge names, attorney names, and court appointee names in official capacity are appropriate and important.
- Your token: Store written on paper — not in a cloud note or screenshot.
Legal protections
- First Amendment: Documentation of public officials' official conduct is core protected speech. Courts cannot suppress reporting on judicial proceedings. Near v. Minnesota (1931).
- Shield laws: 40 states protect journalists and advocacy publishers from being compelled to reveal sources.
- Gag orders: Gag orders bind parties to a case — not the press. We are not a party to any case we document. A judge ordering us to remove documentation of their own conduct is a prior restraint — unconstitutional under settled law.
- Icelandic jurisdiction: Our operating entity is registered in Iceland under the IMMI press protection framework. US civil subpoenas have no reach to Icelandic editorial operations.
Resources
- Reporters Committee for Freedom of the Press: rcfp.org · (800) 336-4243
- Electronic Frontier Foundation: eff.org
- Freedom of the Press Foundation: freedom.press
- National Coalition for Family Justice: ncfjustice.org
- National Parents Organization: nationalparentsorganization.org
- If served with legal papers related to this platform: Call Reporters Committee immediately. Do not respond without counsel. You have rights.