Compare safety data

The One Question: How Do I Compare Safety Data Across Countries?

Comparing safety data between countries is not a matter of looking up a single number. The most common mistake is treating a reported homicide rate or road traffic death figure as a direct, objective fact. These figures are produced by different legal systems, definitions, and reporting practices. Before you rank or compare, you must align three things: the year, the definition, and the population covered. This page gives you a repeatable method to do that.

Step 1: Understand the Three Pillars of Safety Data

Safety data falls into three broad categories, each with its own strengths and weaknesses. You cannot compare a figure from one category directly with a figure from another.

1. Crime and Violence Statistics: These are usually counts of criminal offences reported to the police. The most commonly compared global metric is the intentional homicide rate, defined by the United Nations Office on Drugs and Crime (UNODC) as “unlawful death inflicted on a person with the intent to cause death or serious injury” (UNODC, 2023). The global total of intentional homicide victims was approximately 458,000 in 2021 (UNODC). The range across countries is enormous, from 0.3 to 42.0 per 100,000 population (UNODC, 2023). However, this only covers crimes that are reported. The “dark figure” of unreported crime varies hugely by country, crime type, and trust in police.

2. Traffic and Transport Safety: The World Health Organization (WHO) tracks road traffic deaths, defined as death from a road traffic crash within 30 days (WHO, 2023). The global figure is 1.19 million deaths annually (WHO, 2023). Country rates range from 3 to 37 per 100,000 population (WHO, 2023). This data is better standardised than crime data because the definition is clearer, but reporting completeness still varies, especially in countries with weak health information systems.

3. Perception-Based Safety Data: This comes from surveys like the Gallup World Poll, which asks respondents whether they feel safe walking alone at night in their neighbourhood. The share of the population answering “yes” ranges from 30% to 90% depending on the country (Gallup, 2023). This is not a measure of actual crime; it is a measure of fear, trust, and social cohesion. It often diverges sharply from official crime statistics.

Decision rule: Never compare a police-reported homicide rate with a survey-based fear-of-crime figure. Always check the source category first.

Step 2: Align Year, Definition, and Population

Before you put two countries side by side, you must confirm these three points.

• Year: Safety data lags. The most recent global homicide figures from UNODC are for 2021. A country’s 2023 crime report may not exist yet. If you compare a 2021 figure for Country A with a 2023 figure for Country B, you are not comparing the same moment in time. Check: What is the reference year for each figure? Are they the same?

• Definition: Even within a single category, definitions differ. For terrorism, there is no single United Nations definition (UN, 2023). The Global Terrorism Index (2024) reports 8,352 global deaths from terrorism in 2023, but its definition may exclude state-perpetrated violence or certain separatist conflicts. A country’s own definition might be narrower or broader. Check: Are both figures using the same conceptual boundary? For road deaths, does the country use the 30-day rule (WHO standard) or a different period?

• Population: A national average can hide enormous variation. A country with a low national homicide rate may have a capital city with a rate five times higher. A country with a high national road death rate may have safe highways but dangerous rural roads. Check: Does the figure cover the entire country, or just a region? Is it per 100,000 population, or total number? Per capita rates are the only way to compare across differently sized populations.

Checklist before you compare:

• [ ] Are the figures from the same or adjacent years (within 2 years)?
• [ ] Do the sources use the same definition (e.g., homicide vs. all violent death, 30-day vs. 1-year road death)?
• [ ] Are both rates per 100,000 population?
• [ ] Is the population base the same (national vs. urban only)?
• [ ] Are the sources named and verifiable (e.g., UNODC, WHO, national statistical office)?

Step 3: Use the Right Source for the Right Question

Not all sources are equal. Here is the hierarchy for safety data:

Question	Best Source	Why
How many homicides per capita?	UNODC (United Nations Office on Drugs and Crime)	Most standardised global dataset; includes definition notes and limitations.
How many road traffic deaths per capita?	WHO Global Status Report on Road Safety	Standardised 30-day definition; data from health ministries.
How many terrorism deaths?	Global Terrorism Index (Institute for Economics & Peace)	Standardised methodology; includes definitional notes.
What is the perception of safety?	Gallup World Poll	Only global survey asking the same question in 140+ countries.
What does the country’s own government say?	National Statistical Office	Primary source; may use different definitions; check methodology.
Should I travel there?	Your own foreign ministry travel advisory	Not a data source; reflects government risk assessment, not crime statistics.

Do not use the CIA World Factbook as a primary source for safety data. It compiles secondary figures and its methodology is less transparent than UNODC or WHO.

Step 4: Apply the “Claimed vs Real” Check

Every official safety statistic has a gap between the reported number and the lived reality. The size of that gap depends on the country and the type of data.

• Homicide rates: In countries with weak police reporting, the true number can be 2-5 times higher than official figures, especially in rural areas and conflict zones.
• Road traffic deaths: Official figures in lower-income countries may undercount deaths by 50% or more, because many deaths occur outside hospitals and are not recorded in health systems (WHO, 2023).
• Crime perception: In high-trust societies, perceived safety often exceeds actual safety. In low-trust societies, perceived safety is often lower than actual safety.

Decision rule: If a country has a low reported homicide rate but a low score on corruption indices (Transparency International) or weak rule of law (Freedom House), treat the official safety data with extra caution. The gap between official and real is likely larger.

Country Basics (for Context)

When comparing safety data, you need a baseline. These are the standard reference points for any country comparison. All figures should be attributed to the named source and year.

• Country: Sovereign state (UN member or equivalent recognition). The United Nations maintains the most widely accepted list (UN, 2024).
• Population: Number of people. Source: UN Department of Economic and Social Affairs (UN DESA) or national census. Censuses are typically every 10 years; estimates fill the gap.
• Capital: Seat of government. May differ from largest city (e.g., Canberra vs. Sydney, Brasília vs. São Paulo). Source: national government or UN.
• Currency: Official monetary unit. Source: IANA or national central bank.
• Language: Official language(s) as legally defined. May differ from most spoken vernacular. Source: UNESCO or national constitution.

The Final Decision Rule

When you have two countries’ safety figures in front of you, ask:

1. Are they from the same source? If not, definitions may differ.
2. Are they from the same year? If not, adjust for data lag.
3. Are they per capita? If not, convert.
4. Does the source explain its definition? If not, treat with caution.
5. Does the figure match what you know about the country from other indicators (corruption, rule of law, health system quality)? If there is a large mismatch, the official figure is likely unreliable.

Never use a single number to decide a country’s safety. Use the homicide rate, road death rate, and perception of safety together. If all three point in the same direction, you have a reliable picture. If they conflict, dig deeper.

What This Page Does Not Do

This page does not replace travel advice. For current safety recommendations, go to your own government’s foreign ministry (e.g., US State Department, UK FCDO, Australian DFAT). It does not provide visa information, real-time alerts, or personal security assessments. It gives you the tools to compare safety data responsibly, with full awareness of its limitations.