One event, two basic measures

An intentional homicide count records reported victims. A homicide rate divides that count by population and usually expresses it per 100,000 people. A populous country can have a large count and a moderate rate; a small country can have a smaller count and a very high rate. Both measures are legitimate, but they answer different questions.

The UNODC-sourced World Bank series currently reports a global intentional homicide rate of 5.20 per 100,000 for 2023. Humain Alert uses the rate for comparability and keeps the unit visible.

Why a rate is not a complete safety score

Homicide is a severe but narrow outcome. It does not include non-fatal assault, conflict deaths classified outside criminal homicide, disappearance, intimate-partner violence that is not fatal, fear of violence, or the distribution of harm within a country. A low national homicide rate can coexist with acute danger for particular communities.

Coverage changes the regional story

Regional averages are especially vulnerable to incomplete reporting. In the current World Bank feed, the latest-year coverage available for some regional groupings is far below the number of countries in the group. Humain Alert now labels that coverage and withholds strong regional conclusions when it is weak.

QuestionUseful measureRemaining limitation
How many people died?Victim countPopulation size differs
How common was lethal violence?Rate per 100,000National averages hide concentration
Is risk changing?Multi-year trendReporting and definitions can change
Who is most affected?Disaggregated local dataOften incomplete or delayed

What responsible comparison requires

  • The same definition and unit.
  • The observation year beside the value.
  • Enough reporting coverage to represent the geography.
  • Caution around conflict-affected and weak-registration settings.
  • Local evidence before any policy or causal conclusion.

Why Humain Alert does not equate safety with morality

Homicide rates describe a lethal outcome shaped by institutions, inequality, organized crime, weapons, social conditions and many other factors. They do not measure the moral worth of a population. The site treats the indicator as an alert for further inquiry, not a label.

Primary sources and data

Evidence note: Sources were checked on 7 Sep 2026. Publication years differ because each institution follows its own release schedule. Humain Alert preserves those dates instead of presenting the series as simultaneous.