This article deals with intentional homicides which are defined as “an unlawful death deliberately inflicted on one person by another person”.1 The data are presented as the homicide rate expressed as homicides per 100,000 inhabitants of a country or region.
Homicides refer to interpersonal violence. Civilian and military deaths during interstate wars, civil wars and genocides are not counted as homicides – but OurWorldInData presents the evidence on deaths in the linked articles.
Eisner (2003) describes the secular trend of homicide rates in Europe and has collected data for this research paper. From his database Eisner constructed data for several European regions and presented this in a table. Using estimates from the Institute of Health Metrics and Evaluation (IHME), we have extended this dataset through to 2016. European homicide rates have dramatically decreased over the last millennium and have remained steadily low over the past 50 years. Italy has historically had higher homicide rates than other European countries, but today those rates have reached Northern European levels.
The long-term raw data without distinguishing between different states is visualized in the following graph.
Note: This graph is Figure 2 in Eisner (2003) - Long-Term Historical Trends in Violent Crime. In Crime and Justice, 30, 83--142. It shows 398 local estimates from Eisner's Database and for the recent past it additionally includes the national series for Sweden, England and Wales, Switzerland, and Italy.
Homicide rates in Germany and Switzerland, local estimates and national series, 13th to 20th century – Eisner (2003)2
Homicide rates in England, local estimates and national series, 1200-2000 – Eisner (2003)3
The declining homicides in some regions and stagnating high homicides in other regions caused a huge inequality in violence in the 19th century as can be seen in the following map. Over the course of the last century this picture has changed: the south of Europe has followed northwestern Europe and now has similarly low violence. The remaining inequality is now between a more violent Eastern Europe and a less violent Western Europe.
Homicide rates in Europe, c. 1880 and 2000 – Eisner (2003)4
European homicide rates at subnational levels – UNODC5
At least with respect to Italy the convergence between the northwest and the south of Europe can be studied in the following graph.
Homicide rates in five Western European countries, 1900-2009 – Pinker (2011)6
Crude rates of homicide in East and West Germany by gender, 1970-1995 – Clark and Wildner (2000)7
Since 1995 (the end of the observation period of Clark and Wildner (2000)) homicide rates dropped significantly.8
In the last year of the study of Clark and Wildner - the year 1995 - there were 1.207 homicides in Germany. Over the following years the number dropped and reached 723 homicides in 2011. The number of homicides in which a gun was involved dropped from 226 (in 1995) to 78 (in 2011).
The interactive chart below allows to visualise the long-term trend of homicides for countries around the world. Time coverage varies between countries – for some countries data is available going back to 1800 while for other countries data for only 1 or 2 years are available.
Homicide rates in Japan have steadily decreased over the past century. It's interesting to note in the following graph that the homicide rate dropped significantly during and immediately after World War II, perhaps because the nation was focused on the war effort.
Homicide Rate (including homicide attempts) in Japan, 1888-2003 – Johnson (2006)9
The homicide rate in the US has fallen dramatically over the past few centuries. The following graph shows the fall in the homicide rate in relation to various armed conflicts in New England. In the early colonial days in America, the homicide rate was incredibly high. But as the colonies became more established and orderly, the rate fell, except for the time during the American Revolution when the colonies were understandably amidst upheaval. It's important to note that one cannot directly derive from this graph a causal relationship between these conflicts and the homicide rate.
European American adult homicide rate (per 100,000 persons per year) in New England, 1630-1800 – Roth (2009)10
Pinker (2011) put together many quantitative accounts of the change and distribution of violent crime in the US. Below, we can see the disparity in homicide rates between various areas in present day New England. Predictably, the more urban areas of new York, Boston and Philadelphia have in general had higher homicide rates than the less dense states of Vermont and New Hampshire.
Homicide rates in the northeastern United States, 1636-1900 – Pinker (2011)11
Homicide rates in the US have declined dramatically over history, but they still remain much higher than other industrialized states. Below, we can see that in the 1960s and 1970s, the US, Canada and England all experienced increases in violent crime, but the increase in the US was much more dramatic than in the other two nations. These rates have since fallen in the US, but homicide is still much more common there than in Canada and England.
Homicide rates in the United States (1950-2010) and Canada (1961-2009) – Pinker (2011)12
Homicide rates in the United States and England, 1900-2000 – Pinker (2011)13
The violent crime in the US is not shared equally among all the states; in general the more populous states tend to have much higher homicide rates.
Geography of homicide in the United States, 2007 – Pinker (2011)14
After a period of higher violence in the 1970s and 1980s the US experiences a marked decrease in homicides and other crime rates. Recent data is published in the US census here. From this data it can be seen that the downward trend continued in the recent economic crisis.
Estimated world homicide rates (per 100,000 people per year), late 1900s – Roth (2009)15
|Low- and middle-income nations|
|The Americas (a)||27.5||51||4.8|
|South Asia (b)||5.8||8.1||3.5|
|Eastern Mediterranean (d)||7.2||9.4||4.8|
|Western Pacific and Southeast Asia (e)||3.8||5.6||1.8|
|Western Pacific and Southeast Asia||1.1||1.3||0.8|
|The world (high, middle, low income)||8.8||13.6||4|
The visualisation below shows the annual number of deaths from homicide by country in the recent past, from 1990 onwards.
The visualisation below shows the annual number of deaths from homicide, differentiated by age category. This data can be viewed for any country using the "change country" option in the interactive chart.
The following maps show homicide rates globally, measured as the number per 100,000 people. This is available as estimates from the IHME, in addition to those published by the UN Office on Drugs and Crime's International Homicide Statistics.
Kellermann and Mercy (1992)16 studied 215,273 homicides in the United States between 1976 and 1987. In line with the results presented above only 14.7% of the homicides were commited by women. Similarly unequal was the distribution of gender among the victims. The study found that 77% of the 215.273 victims were men.
The UNODC Homicide statistic 2012 has global data on the gender of the sex of homicide victims.
The table below shows the percentage of female offenders for various regions throughout history.
Female offenders as a percentage of all offenders in various historical studies – Eisner (2003)17
|England, 1202–76||-||-||8.6||-||Given (1977, p. 48)|
|England, various counties, 1300–1348||-||5.1||7||9.8||Hanawalt (1979, p. 118)|
|Cracau, 1361–1405||1||1.0*||10||Schüssler (1998, p. 313)|
|Zurich, 1376–85||1.4||-||-||4||Burghartz (1990, p. 80)|
|Avignon, 1372||21.0†||-||-||23.7||Chiffoleau (1984, p. 250)|
|Arras, 1400–1436||13.7‡ 4.6§||-||-||-||Muchembled (1992, pp. 34, 89)|
|Constance, 1430–60||4.7||-||.0*||17.2||Schuster (2000, p. 71)|
|Douai, 1496–1519||-||-||1||-||Fouret (1984)|
|Amsterdam, 1490–1552||14||-||3||15||Boomgaard (1992)|
|Arras, 1528–49||5||-||-||20||Muchembled (1992, pp. 34, 89)|
|Brussels, 1500–1600||8.2||-||-||7.4||Vanhemelryck (1981, p. 314)|
|Cologne, 1568–1612||4.4||4.7||5.7*||22.9||Schwerhoff (1995, p. 91)|
|Seventeenth and eighteenth centuries:|
|Rural areas near Trier, late 16th to early 18th centuries||12.3||4.5||3.7*||5.2||Lacour (2000, p. 535)|
|Castile, 1623–99||1.4*||Chaulet (1997, p. 17)|
|Bavaria, 1600–1649||4.5||-||2.9*||12.4||Behringer (1995, p. 65)|
|Bavaria, 1685–89||5||.0||4.5*||13.2||Behringer (1995, p. 67)|
|Surrey, 1663–1802||18.2||7.9||13||23.9||Beattie (1975, p. 81)|
|Leiden, 1678–1794||7.8||41#||5.5*||47.3||Kloek (1990, p. 8)|
|Gent, 1700–1789||10.2||-||-||24.6||Roets (1982)|
|Stockholm, 1708–18||41||-||43||67||Andersson (1995)|
|Alençon, northern France, 1715–45||20||-||-||33.5||Champin (1972)|
|Armagh, Ireland, 1736–95||6.5||-||7.6||9.7||Garnham (1996)|
|Neuchatel, 1707–1806||6.2||-||-||14.7||Henry (1984, p. 660)|
|Nice, 1736–92||-||-||3||-||Eleuche-Santini (1979)|
|Rural northern Germany, 1680–1795||3.5||-||-||18.5||Frank (1995, p. 235)|
|Late nineteenth-century Germany||8||-||16||-||von Mayr (1917, p. 754)|
|Late twentieth century:|
|England and Wales, 1997||14.7||8.5||11.9*||23||Home Office (1998)|
|Italy, 1998||16||6.9||5.2||15.2||Istituto Nazionale di Statistica (2000)|
|Germany, 1997||12||7.5||10.0*||23||Bundeskriminalamt (1998)|
|United States, 1997||15||8||10||32||Federal Bureau of Investigation (1998)|
Similarly, the age distribution of violent offenders is not uniform, either. The following graph shows this distribution for various time periods and European regions.
Age distribution of violent offenders across time and space – Eisner (2003)18
There are two very different sources for historical data on homicide rates depending on the era discussed. For the last few decades, statistical offices have recorded homicide rates as part of the vital statistics. Eisner (2003) notes that "in most European countries, data series start during the second half of the nineteenth century, although Swedish national death statistics were introduced in the middle of the eighteenth century".
For earlier historical periods, researchers have reconstructed the long-term changes in homicide rates from historical records that often originated from some form of legal documentation of the crimes. Although there are gaps in the reconstructed data, and there are uncertainties in both the number of homicides and the size of the population, the historical record is more complete than one might assume. One reason for this is the great social importance of these crimes. Violent killings were of concern for a long time, and they were therefore often reliably registered. In some cases, or for prehistoric times, it is additionally possible to use the insights from forensic archaeologists, who can determine the causes of death from skeletal remains.
Usually homicide rates are calculated as homicides per 100,000 people per year.
The 'Historical Violence Database' Data on violence from medieval times to the present, mainly in Western Europe and the US, has been collected in the 'Historical Violence Database' at the Ohio State University. In addition to homicide rates for many countries this database also includes qualitative information - among much more inforamtion also on the place and the cause of death and on the victim and the perpetrator.
The website can be found here: Historical Violence Database at Ohio State University
- The World Health Organization presents data since 1950 here.
- The United Nations Office on Drugs and Crime (UNODC) has more recent data for more than 200 countries here.
For the USA there is data available from the Federal Bureau of Investigation (FBI) and the National Institutes of Health (NIH). And there is the National Violent Death Reporting System at the National Center for Injury Prevention and Control in Atlanta, Georgia that has detailed data here.
Institute of Health Metrics and Evaluation (IHME), Global Burden of Disease (GBD)
- Data: Death rates & absolute number of deaths from homicide
- Geographical coverage:Global, across all regions and countries
- Time span:Available from 1990 onwards
- Available at: Online here