Counting ConventionConcept demonstration

Counting Convention

Two true numbers shouldn't look like a mistake.

Public counts of human-trafficking cases often come from several places. One is a running total since a task force began. Another covers a single year for a single unit. One counts related charges and the other doesn't. Both are right, but nothing on the page says so, and readers end up comparing numbers that were never meant to match. This concept is a shared way of saying how you count. It isn't a new reporting system.

See how two counts differ ↗

An Advanced Magnolia concept. Not in use anywhere.

People. Partnerships. Possibility.

Why two official counts can both be right

Different numbers, no error.

Mock-up comparing two fictional arrest counts. A sample task force reports 120 arrests as a running total, including related charges and joint operations counted by every partner. A sample state unit reports 45 arrests in one calendar year, trafficking charges only, its own cases only. A note says both are correct. — Fictional numbers. They differ in period, charge scope and who's counted. The answer isn't a new number. It's saying which convention each number follows.

The concept

Say how you count. Then check before you publish.

Three phone screen mock-ups: a reviewer deciding whether two fictional agencies reported the same operation, a change record of seven decisions with one still open, and a draft summary that shows how 42 submitted victims became a draft total of 37. — Concept demonstration with sample data from three invented agencies and a made-up year. These are visual mock-ups, not working software. Select the image for a closer look.

Mock-up of a reviewer deciding that two fictional agencies' rows describe the same operation, so it counts once statewide while both rows are kept. — Concept mock-up · Fictional sample data.

Field	A-03 (Agency A)	B-02 (Agency B)
Operation	OP-17	OP-17
Date	2031-03-22	2031-03-22
Type	SEX	SEX
Victims	3	3
Trafficking arrests	2	2

Mock-up of a change record listing seven fictional decisions, each with its reason and marked as undoable, and one still open. — Concept mock-up · Fictional sample data.

ID	Row	Change	Reason	Status
D1	B-04	Type blank → "Type not reported"	Returned to agency	Can be undone
D2	C-02	Excluded from 2031	Prior period	Can be undone
D3	A-07	CSEC → SEX (original kept)	Local code	Can be undone
D4	B-02	Counted under A-03	Same joint operation	Can be undone
D5	A-05, C-04	Kept separate	Operation refs differ	Can be undone
D6	B-06	Total 4 → 3	Agency confirmed keying error	Can be undone
D7	Agency C	Cover sheet doesn't match rows	Asked Agency C	Open

Mock-up of a draft summary marked not for publication, showing one open item, 37 victims and 15 trafficking-charge arrests, and the arithmetic from 42 submitted victims to 37. — Concept mock-up · Fictional sample data.

In practice

One fictional year, three made-up agencies.

Three sample agencies send in their counts for a reporting year. Two of them worked the same operation and both reported it. A quick check flags the pair, and a reviewer confirms it's one operation, so it counts once in the combined total and stays in each agency's own numbers. Nothing is deleted. Another pair looks identical but turns out to be two different operations, so they stay separate. That's why nothing is merged automatically. A row from the prior year is set aside, one local category code is mapped to the shared one, and one agency's cover sheet doesn't match its rows, so that question goes back to the agency and stays open. The draft that comes out says what's still open, which arrest count it uses, and exactly how 42 submitted victims became 37.

How it would work

Define it once, count it once, show your work.

01 / A SHARED CONVENTION
A few plain rules everyone uses: a joint operation counts once in the combined total; a victim is counted in the period first identified; arrests come in two defined counts (trafficking charges, and related charges), and any published total says which it includes.
02 / CHECKS A PERSON DECIDES
Missing fields, dates outside the period, unknown categories, possible duplicates and totals that don't add up are flagged with a sentence explaining why. A reviewer decides each one and writes a short note. Flags are questions, not verdicts.
03 / A DRAFT THAT SHOWS ITS WORK
Every change keeps the original value and can be undone. The draft lists open items first, appends the definitions, and shows the arithmetic from what was submitted to what's reported.

The convention

The rules, on one card.

Mock-up of five illustrative counting rules: joint operations count once statewide but stay in each agency's totals; victims count in the period first identified; arrests on trafficking charges; arrests on related charges, with any published total saying whether it includes them; and four defined trafficking types, with local codes mapped rather than dropped. — Illustrative definitions, not any agency's rules or any law's requirements. The office that owns a report would confirm or change each one.

What already exists

Most of the pieces are already out there.

Federal task-force performance measures already define arrest and victim categories, and ask for "new" counts in each period.
The FBI's national incident reporting has had offense codes for commercial sex acts and involuntary servitude since 2013.
Ordinary spreadsheets can already enforce most of these checks with drop-downs, required cells and totals.

That's why this concept is a convention and a template first. Software would only make sense if a reporting office said spreadsheets weren't enough.

Questions we'd ask first

Before anything gets built.

Where do the combined counts come from today, and who adds them up?
Which counting rules does each source already follow, even if they aren't written down?
Is a shared template something contributors would actually use?
Which differences cause real follow-up work, and which only confuse readers?
Who would keep the definitions current?

Those answers come from conversations with the people who do this work. We haven't contacted anyone.

Sources for the section above: federal task-force measures, OVC ECM law-enforcement questionnaire, June 2025; FBI trafficking data collection since 2013, BJS, Human Trafficking Data Collection Activities, 2022.

Small print: This is a concept. It hasn't been reviewed by any reporting office or counsel, tested with any agency, or evaluated. The definitions are illustrative. It doesn't collect or show any information about real people or cases. It isn't affiliated with or endorsed by any government agency or other organization.

Closing

Clear about what the numbers mean.

The aim is published counts that say how they were made, so two true numbers can sit side by side without looking like a mistake.

Start a conversation.

Contact Advanced Magnolia ↗