Back to the toolkit
PRESS KIT

Journalist Toolkit

Bringing the Family Court Story to Reporters

Before you use this — please read the use policy, particularly if you are currently in a family court matter. Pitching your individual case to a journalist is not what this toolkit is for and will hurt the broader effort.

What this is

A self-contained kit any citizen, advocate, or researcher can use to bring the family court accountability story to a journalist in their region. It contains a media one-pager (the data summary that any reporter, anywhere in the country, can use as the foundation for a state-specific or national piece), a pitch email template, and guidance on what makes a pitch land.

It does not target specific outlets or specific reporters. The reform community has many people who know their local press; this toolkit is the raw material those people can adapt.

The Pitch Email

Keep it short. Reporters delete long pitches unread. Aim for around 200 words and a clear ask.

The Media One-Pager

Format on letterhead before sending. Treat it as a press release, not a personal note.

MEDIA ONE-PAGER

THE FAMILY COURT BLACKOUT — A STORY ABOUT THE DATA NOBODY HAS

One Line

American family courts are the largest publicly funded decision-making institution that publishes no performance data, and the silence is structural, not accidental.

Five Numbers That Matter

  • 18.9%
    the share of total child support obligations actually collected in FY2024 by the federal Title IV-D enforcement program, despite $6.6 billion in combined federal-state administrative spending. (Source: Office of Child Support Services, FY2024 program statistics.)
  • $25.8B
    paid to families in FY2024 through Title IV-D; $6.6 billion — administrative cost. The program counts every dollar collected as its own contribution, including dollars that would have been paid voluntarily without state involvement.
  • 66¢
    on the dollar — the federal reimbursement rate for state child support enforcement administrative costs, with no ceiling. Federal incentive payments reward enforcement volume, not enforcement accuracy.
  • 0
    the number of states that publish ruling-pattern data by family court judicial officer in a form a researcher could meaningfully analyze.
  • ~3,000
    the approximate number of county-level family court jurisdictions in the United States. The fragmentation is itself a barrier to accountability.

The Structural Story

  • A class of decision-makers (judges, guardians ad litem, custody evaluators) who hold absolute or quasi-judicial immunity from civil liability.
  • Operating under a legal standard ("best interests of the child") that is intentionally undefined and unappealable on most factual determinations.
  • Inside a professional ecosystem where the same actors generate, review, and police each other's recommendations, and economically benefit from prolonged conflict.
  • Funded by a federal incentive structure that rewards enforcement activity rather than enforcement accuracy.
  • With essentially no published outcome data on any dimension that matters.
  • And — the part that survives every partial reform — no common schema. Where data is released, it is released in incompatible, undocumented, locally invented formats, which means the release can be simultaneously complete and useless. Common exchange standards for exactly this data have existed for two decades and are voluntary everywhere they apply. The recordkeeping policy that keeps them voluntary is written by, and on the advice of, the same professional class the data would measure.

Why Now

For thirty years, the diagnostic capability to evaluate this system at scale was concentrated in expensive consultants and the people running the system. The cost of an honest data-driven look at family court patterns has now fallen, with widely available tools, to nearly zero. Citizens in multiple states are now actively requesting public records that, when aggregated, produce for the first time an empirical picture of who decides what, on what basis, with what consistency.

Sources & Background

  • Two long-form structural pieces:
  • The Schema That Never Gets Written — on the absence of a common records schema, and the request clause that goes after it
  • Family Court Record prototype (synthetic-data accountability dashboard demonstrating the methodology)
  • Office of Child Support Services FY2024 program statistics
  • Congressional Research Service report RS22380 on Title IV-D incentive structures
  • Stump v. Sparkman, 435 U.S. 349 (1978), on judicial immunity

What Makes a Pitch Land

A few practical observations for anyone bringing this to a reporter:

Lead with the pattern, not the case. A reporter's filter is "is this a structural story or a personal grievance." Pitches that lead with an individual case get categorized as the second within fifteen seconds. Pitches that lead with the data gap get read.

Offer access, not opinions. What journalists want is documents, datasets, and interview-ready sources. They do not need someone else's analysis of what it all means; they need the raw material to do their own analysis. The most valuable thing in any pitch is "I have data nobody else has, and I can connect you with people willing to talk on the record."

Stay calm and structural in tone. Reporters notice when a pitch reads angry, and they discount the source accordingly. The structural argument is more powerful than any individual outrage, and the tone of the pitch should reflect that.

One outlet at a time. If you pitch widely and one outlet picks it up, the others will treat it as old news. Rolling exclusivity — one outlet at a time, with a stated timeline — is the right cadence.

Two follow-ups, then move on. Once at one week, once at two weeks. After that, no response is a no.

The Schema Angle — A Second, Harder Story

The pitch above is the story about missing data. There is a second story underneath it that is generally easier to prove, harder to deny, and — for a reporter who has already covered records fights — considerably more interesting: the story about why the data cannot be compared even when it is released.

Every records reporter has lived this. You win the request and receive an undocumented spreadsheet, or a PDF of a database, or forty counties reporting the same fact under forty field names. The instinct is to file that under bureaucratic incompetence. It is worth interrogating that instinct, because the pattern is too consistent and too favorable to the institution to be pure entropy — and because the decisions that produce it are documented decisions, made in legal review, with a paper trail that can be requested.

The reportable version has three legs, and all three are obtainable through ordinary records law:

  • The standards exist and were declined. National exchange standards for justice-system data have existed for roughly two decades; the National Center for State Courts publishes court reporting standards; federal agencies have operated under a machine-readable open-data default since 2019. Ask each agency for any record of its evaluation of, participation in, or decision to decline adoption of a common exchange format. The answer — including "no responsive records" — is the story.
  • The advice is on paper. Request the legal guidance issued to staff on answering records requests: when data constitutes a "record," when running a query counts as "creating" one, the standards for asserting exemptions. Much of it may be properly withheld as privileged. Request the privilege log anyway. The volume and date range of institutional legal guidance on how to answer the public is itself a publishable fact.
  • The contract gave it away. Request the case management system contract, limited to the provisions on data ownership, export format, and extract fees. Agencies that cannot describe their own records because a vendor owns the data model made a procurement decision, on the record, with public money.

Why this lands with editors when the underlying reform story sometimes does not: it is a story about the reporter's own beat. Every newsroom in the country has been handed unusable public data. Framing that experience as a designed outcome rather than a nuisance converts a familiar frustration into an investigation — and it generalizes well beyond family court, to any agency the outlet covers.

The full mechanism, the request clause, and a model minimum schema are in The Schema That Never Gets Written. Hand a reporter that page and your own schema-request responses together; the pairing of "here is the argument" with "here is what my agency actually said when asked" is what makes it assignable.

What Reporters Will Want From You

If a reporter takes the bait and asks for more, have the following ready as a single shareable folder:

  • The two published structural articles
  • This media one-pager
  • Your public records request log (what you've sent, what's come back, what's been refused)
  • Any responsive records you've received, with a brief summary of what they show
  • Your schema responses — the data dictionaries, retention schedules, and extract specifications you asked for, and every refusal to supply one. Put the refusals in the folder deliberately. A reporter can publish "the agency declined to describe the structure of its own public records" without waiting on a single dataset.
  • A short list of people in the reporter's coverage area who have been through the system, whose cases are fully concluded, and who have agreed to speak on the record

That last item is the single most consequential asset. Many pitches die at the "do you have anyone willing to be quoted" question. If you don't have those names, the reporter will have to find them; if you do, the story is dramatically more likely to get assigned.

A Final Reminder

The story is about the system, not about any individual judge, evaluator, or attorney whose name happens to appear in your records. Bring reporters to the structural pattern. The names will surface from the data on their own — that is not your job to supply, and supplying them as the lead of a pitch turns a reform story into a vendetta story, which gets killed in editorial review every time.