HOW MASS‑SHOOTING DATA COUNTS INJURIES AND FATALITIES

Empty hospital corridor with closed doors, fluorescent lights, and a fire exit sign

(AND WHERE PEOPLE WHO DIE LATER FIT IN)

1. ON‑SCENE FATALITIES

These are the easiest for all systems to count.
If someone is declared dead at the scene, every dataset — GVA, Everytown, Mother Jones, FBI — includes them.

Reporting dynamic

On‑scene deaths are immediately visible to police, media, and databases.
They become the “official” fatality count within minutes.
This creates a bias toward instant, visible outcomes.


2. PEOPLE WHO DIE LATER (HOURS, DAYS, OR WEEKS)

This is where the definitions diverge.

GVA (Gun Violence Archive)

  • Yes, they update counts when victims die later.
  • They track hospital updates, coroner reports, and follow‑up news.
  • A shooting with 3 injured can become “4 killed” days later.

Everytown

  • Yes, they update when later deaths are confirmed.
  • Their definition includes both killed and injured, so the event remains a mass shooting either way.

Mother Jones

  • Yes, but only if the death fits their narrow criteria
    (public, indiscriminate, non‑domestic, non‑gang, etc.).
  • If someone dies later but the event was excluded for other reasons, it stays excluded.

FBI

  • Yes, but only for “mass killings” (3+ fatalities).
  • If later deaths push an event over the threshold, it becomes a mass killing retroactively.

Reporting dynamic

Later deaths are under‑reported because:

  • media attention fades
  • follow‑up stories are rare
  • databases rely on secondary reporting
  • some victims die weeks or months later from complications

This means the true fatality burden is always higher than the initial numbers.


3. INJURIES (NON‑FATAL GUNSHOT WOUNDS)

GVA

  • Counts all gunshot injuries
  • Includes grazes, through‑and‑throughs, and serious wounds
  • Injury count is updated as hospitals release information

Everytown

  • Counts injuries in their definition
  • Injury totals are updated when new information emerges

Mother Jones

  • Does NOT count injuries
  • Injuries are irrelevant to whether an event qualifies
  • A shooting with 20 injured and 0 killed = not a mass shooting in their dataset

FBI

  • Does NOT count injuries for mass killings
  • Injuries only matter for “active shooter” reports, not mass‑shooting statistics

Reporting dynamic

Because injuries are:

  • medically complex
  • slow to confirm
  • often under‑reported
  • sometimes not disclosed due to privacy

…injury counts are always incomplete in early reporting and often remain lower than reality.


4. PEOPLE WHO ARE INJURED AND DIE MUCH LATER

This is the biggest blind spot.

GVA

  • Will update if the death is reported publicly
  • But if the death occurs months later, it may never be captured

Everytown

  • Same dynamic — depends on public reporting

Mother Jones

  • Only updates if the event was already included
  • If the event was excluded for other reasons, the later death does not change anything

FBI

  • Only updates if the later death pushes the event into “mass killing” territory
  • Otherwise, it is ignored

Reporting dynamic

Victims who die:

  • after multiple surgeries
  • from infection
  • from organ failure
  • from complications
  • after long ICU stays

…are often not counted in mass‑shooting fatality totals.

This creates a systematic undercount of deaths.


5. WHY THIS MATTERS

Because the system’s logic is:

“If the injury didn’t kill you immediately, it doesn’t count the same.”

This mirrors the interpersonal minimization pattern:

“At least he doesn’t hit you.”
“At least no one died.”

Both collapse violence into instant, visible harm, erasing:

  • long‑term medical consequences
  • delayed fatalities
  • chronic disability
  • psychological trauma
  • community‑level impact
  • the coercive meaning of the act

A person who dies three weeks later from a gunshot wound is just as dead,
but the system often treats them as statistically invisible.


6. THE STRUCTURAL TAKEAWAY

Mass‑shooting data is shaped by:

  • definitional thresholds
  • reporting delays
  • media attention cycles
  • hospital privacy
  • coroner timelines
  • database update policies

The result is a dataset that under‑counts fatalities,
under‑counts injuries,
and erases delayed deaths.

The violence is real.
The reporting is partial.

We Believe You


Apple Music

YouTube Music

Amazon Music

Spotify Music

Explore Mini-Topics



Leave a Reply

Discover more from Survivor Literacy

Subscribe now to keep reading and get access to the full archive.

Continue reading