Statistics which will count toward the Violent Death Rate. Comment below if you think I am leaving out a category which should be included.

Person On Person Homicide: The most straight forward. This refers to any standard murder.

State Executions/Capital Punishment: Of course you can say that the buggers got what they deserve, and that’s fine; But the purpose here is to tally deaths that are violent, and this is a violent way to go out. Countries with Capital Punishment tend to be more violent too, so it’s as good an indicator as any of a “violent society”.

Justifiable Homicide By Law Enforcement: Again, pretty straight forward. This is when police use deadly force.

Justifiable Homicide By Citizen: Any killing later found to be justified. Wife stabbing an abusive hubby… burglar catching a lead sandwich during commission of a crime… you get the picture.

Citizen Deaths By Terrorism: This should be lumped in with normal homicides, but for reasons I can’t fathom, authorities often leave these out of their homicide stats. It happens to be one of my pet peeves. A city or country that is more likely to be a target by terrorists, is by definition, more dangerous… end of story. I don’t think motive should lead us into sub-categorizing homicide data. What’s next? We no longer consider drug related murders homicide? Folly!

Prison Inmate Homicides: In the US, these are typically added into the pool of regular homicide data, but in many other countries they are tallied separately. Where I can actually find some info (or at least estimates) I will count these toward the VDR.

Military Battle Deaths: Seems easy enough, but here we hit some gray area of interpretation as far as what constitutes a violent society. Surely… if you are habitually throwing large numbers of your populace into various wars, which then lead to high body counts, then that nation can easily be considered “violent”. But what if you are merely defending yourself (think Poland 1939) against an aggressor state? Shouldn’t it be they who are labeled the “violent” ones? Another oblique area would be reconciling countries that maintain extremely high kill ratios. In combat operations during Persian Gulf Wars 1 and 2, Coalition forces took out Iraqis in high numbers, without significant loss of life on their end. Here too, you could say that Iraqis are being unfairly penalized statistically for simply not having the same level of military technology… and that’s true. However, I think the purpose of this site is to answer the question “As a citizen of countryX how likely is it I will be violently killed?” not “As a citizen of countryX how likely is it that I will kill someone else violently?”

Civilian Deaths From Wartime Starvation: When a military directly causes the deaths of citizens with blockade, encirclement, etc. Picture The Siege Of Leningrad 1941-1943.

Military Deaths By Terrorism: Here too, I have no idea why they don’t just lump these deaths in with battle casualties, but they rarely do. A soldier being deployed, and dying by terrorism while on deployment IMO is no different than a soldier dying while taking small arms fire defending a trench.

Military Homicides: These can take place domestically, or abroad. Domestic military base homicides will almost always be counted by the local government though.

Military Executions: Not as common as they used to be, but where applicable, they will be added to the tally.

Statistics that will not be counted toward the Violent Death Rate, and my reasons. If you disagree be sure to comment, and I’ll consider changing the variables.

Citizen Suicides: I have seen papers where suicides were lumped in with homicides to determine how violent a society is. I don’t agree with this. Sure, a person jumping off a bridge is dying violently. But it is not a violence that was directed against them, it is a concerted choice. Some of the safest countries also have the highest suicide rates. In a perfect world, one in which we have conquered war and disease, suicide would be the only way anyone would ever check out. As more or less a sign that a society has mental health issues, or too much time on their hands, I would not say a society is more dangerous due to it’s suicides.

Diseases: Whether or not a person dies of a virus or disease, it is a constraint of the times and their society’s medical advances. Or plain bad luck. It’s not a sign of willful violence.

Accidents: Here too, you could argue that slipping into an industrial grinder is the very definition of a “violent death”. But what we’re looking for is preventable deaths that someone chose to inflict on another person. Accidental deaths, like disease, are products of their times. Example… automobile accidents before seat belts and other safety devices. Preventable, yes. Willful, no.

Military Accidents Or Suicides: This is on the fence, but at this time I am choosing to leave them out. People working very safe jobs also still die in accidents and commit suicide, so I don’t see why the military occupation should be singled out. At first, I considered comparing the military accident and suicide rates against their demographic civilian counterparts… and only then counting them if they went above the norm (the thought being that their deployment led to their death). But I have since discounted this logic, owing to the opinion that the military can simply be said to be a more dangerous job than some others, dangerous jobs also include such things as taxi driver and fisherman, and as such I don’t see cause to view the military differently.

Drunk Driving Deaths:
These sometimes get written up as manslaughter, so some homicide stats may contain a number of DD deaths. But on the whole, I am going to stay away from these. I did come up with a “Drunk Driving Inflicted Death Rate” for the United States (these are deaths caused to pedestrians and non-intoxicated drivers of DD operated vehicles), but finding it nearly impossible to get reliable stats for any other countries, I have given up on this category.

Notes on data gathering and estimates by country:

Australia:
-Australia does not count multiple victims as separate homicides for some reason, only the instances. For this reason, homicide statistics are lower than they should rightly be.
Belgium:
-See this http://www.brusselsjournal.com/node/1156 for an interesting argument about Belgium crime stats.
Colombia:
-Higher population totals can be found in some statistic tables. Almost 10% of Colombians world wide do not actually live in Colombia. For our purposes we will use the lower population totals that reflect only those living in Colombia.
France:
-population stats refer to “metropolitan France”, not the overseas departments and territories.
-rather infuriatingly, France keeps separate categories for “homicide” and “fatal assaults”… even though they are the same thing. This gives them a misleadingly low murder rate in most articles and statistic publications I’ve seen. I have added them together for the homicide/manslaughter column.
-will count “unknown” cause of death in Operation Enduring Freedom as battle deaths, rather than accidents.
-state executions post revolutionary period until 1959 will have to be estimated, since reliable figures can not be found. 2 per year will be used in all cases, accept for 10/year during the 1940’s, as that period was though to have higher execution totals.
-pre 1973 French homicide rates difficult to come by. Rates are therefore only estimates in all older French history. There are studies that show that homicide rates decreased in line with a jump in literacy between 1852 and the first world war; so we will use a declining rate during that time, and a blanket 6.0 homicide rate in the years previous to 1852.
Germany:
-1960 to 1986 homicide rate will be estimated at 2.5/100,000. Other years will be 1.5/100,000.
Iran:
-There are wide discrepancies in population data. This is at least partially due to the fact that some tallies do not count refugees. For our purposes, I will always try to include refugees, as I am considering anyone to be on a nation’s soil to be of that nation at that time.
-Post Revolution homicide rate seems consistently around 2.7/100,000 based on the few years in which official statistics are given. Estimates of the remaining years of 2.7/100,000 will be used for the unknown years.
Ireland:
-After 1921, there is a drop in population due to the formation of The Republic Of Ireland, and the North Of Ireland’s population into the UK.
Japan:
-Due to pressure to keep an almost impossibly high clearance rate, police in Japan are notorious for under reporting homicides. Investigations into unnatural deaths in Japan are minuscule by any country’s standard. Japan’s homicide rate should be assumed to be higher than official reports suggest.
New Zealand:
-New Zealand records their stats with years ending on June 30th. For example, the excel table will show “2008” refering to the period of June 30th 2007 to June 30th 2008.
Nigeria:
-Military deaths in my spread sheet will refer to “political issue” violent deaths, as recorded by NGA-Watch. The other economic conflicts that reseult in deaths will be tallied alongside traditional homicides, in the Domestic Manslaughter Rate.
Poland:
-For 1951-1959 population totals, averaging the growth between 1950 and 1960 censuses will be used, as specific tallys could not be found. An average of 465,590 new citizens per year of this growth period.
-About 6 million, or -21.4% of Poland’s population died between 1939-1945. Since exact numbers for each year of WW2 can not be found, averages for population and military + civilian war dead will be used.
-After WW2 growth rate, as well as return to Poland of refugees led to +3,817,000. Averages will be used for those years as exact tally can not be found. .0485507
-An estimated 2500 killed or executed during the Silesian Uprisisings of 1919-21 will be placed front heavy over the 3 years.
Russia:
-Since Chechnya did not successfully break away, all their deaths during the 2 Chechen wars should be counted into Russia.
-Casualties during the Chechen conflicts must rely on the loss estimates that observers have reported over the years. For our charts (first war), Russian military deaths will be the official tally of 5,500; Chechen militant deaths will be near the middle of the estimates at 6,000; Chechen civilian deaths are estimated from 30,000 all the way up to 100,000. But we will use the rather Bearish figure of 40,000., with the majority assumed to be resulting from the Battle of Grozny in early 1995. It should be noted that numbers used for all Chechen conflict casualties err on the low end of estimates.
-Since Soviet losses in WW2 do not specifically separate Russians, I have taken the 65% of Soviet population figure (according to the 1939 border) and applied that to the estimated military and civilian war dead. The new total then is 15,525,900.
-State executions refers to Stalinist purges, a loose estimate of 748,000 given by one historian have simply been parceled out evenly by year, since specific incidences are impossible to come by and tally by date.
-It is believed that during the calm period of the post war Soviet Union that homicide levels were believed to be under that of the United States, largely due to the enormous police presence. Lacking detailed statistics I have chosen a rate of 4/100,000 to be applied to the years between 1919-1980 (at which time an agreed upon homicide rate increase occurred). I agree that this detailing in incomplete, and that there most certainly years of much higher, and even probably lower rates during that time frame, but it’s the best I can do at this time.
UK United Kingdom:
-They have a very silly policy whereby they remove old homicides if the suspects have been acquitted, or if they pleaded to a lesser charge. Make sure you only go by RAW homicides.
-Between certain decades population rose…
1951-61 average 248,400/yr
1961-71 average 321,900/yr.
1971-81 average 42,950/yr.
1981-91 average 108,120/yr
1991-2001 average 136,130/yr
2001-2008 average 306,273/yr
-In the whole UK, they apparently don’t consider killing someone while driving in the homicide stats. That is considered death from reckless driving, and the stats for that will need to be added in.
-Prior to the 12th century Britain, all murders were considered a civil wrong.
-For WW1 stats: Deaths are… 752,258 Military (army, colonial armies, navy, merchant marine)+ 109,000 civilian starvations= 861,258 total. Difficulty determining exact figures per year, so total parceled out over the years 1914-1919.
-Scotland’s homicide rate in excel spreadsheet refers to it’s murders as a rate of the entire UK population estimate for that year.
-Studies between 1202-1276 in 5 rural counties showed homicide rates of 9-23/100,000. About 10-20% were attributed to robbers, most were the result of deaths in group brawls, 1-on-1 deaths were less common.
-Homicide rates for 3 counties near London in 1559-1603 had homicide rates of 6.7, 16, and 14 per 100,000.
-In England, the 14th century was a murder spike compared to the previous century. The Hundred Years War and The Black Death led to much destabilization of the traditional law and order.
-Oxford of the 1340’s had a homicide rate of 110/100,00… well over the 14/100,000 estimated in England at the time.
-military casualties of 4 per year from 1970-1974 will be added to account for the Dhofar Rebellion. Exact dates of deaths not found.
-For population stats 1901-1950 only census data every 10 years available. Will take intra-decade growth rate and parcel it out by year to arrive at yearly population count.
-519 casualties in the Malayan Emergency. Unknown how many casualties in specific years, so will be distributed evenly over 1948-1960
-Populations taken from census data England + Wales + Scotland + N. Ireland from 1801-1891. Pop every ten years, for interm population figures average yearly growth will be used.
-Pre 1841 Ireland data hard to come by. Will assume similar growth rate to Scotland for estimation purposes.
-For Afghan wars of the 19th century, a great number of the British dead were Indian troops. Only 1/3 of the casualties will be estimated as British.
-First Anglo-Burmese War has a loosely estimated casualty tally of 15,000. Since no indication whether this means wounded or dead, we will use total of 5,000 British killed, over the 3 year period evenly.
-Loose attempt to include British dead from the Gurkha War is difficult. 34,000 deployed at height, 500 dead seems a reasonable estimate. To be spread over the three years of 1814-1816.
-Pre-1857 homicide stats are difficult to come by on a year by year basis. It is thought by many historians that later 16th century England had a rate per 100,000 of about 15. We will use a gradual decline rate spread from 1580 (at 15.0) to 1856 (at 1.6).
-Since serial killer Harold Shipman’s victims were only pronounced homicides after the fact, a tally of his victims by year has been added to the total UK homicide stats. He begins to significantly contribute to homicide rate in 1984.
-For wars in the american colonies pre-1775, English combat deaths will only be counted if the soldiers appear have been deployed specifically from England. Otherwise, the population of the colonies will have to be added to the pop totals of the UK spreadsheet for those years, and I don’t really want to go off on that tangent.
-Executions from 1735-1965 will have to be done in parceled out fashion, according to the totals that are given from various periods by multiple sources. Undoubtedly, there were years where executions spiked or abated, but we do not have exact yearly tallies to go by. For pre-1735, a flat rate of 1/100,000 will be used: though they were certainly more violent times, the lower role of government in all walks of life probably means that disputes were handled more privately than through the state system.
-Using the 215 confirmed victims of Harold Shipman as a rough gauge of his body count (some think he may have killed twice as many); we find that…Of 15,500 homicide victims between 1975-1998 in England +Wales, he caused  1.38%. Between his most intense period, 1995-1997, he killed 95 of England + Wales’s 2,259 homicide victims, or 4.2%.
United States US:
-Justifiable Homicide stats for 1976-2008 fell in pretty well with a pattern, if homicide numbers went up, so did the number of justifiable homicides. During that 31 year period, the average ratio of homicides/justifiables was 30.3/1. In certain years the gap widened to around 33/1 (for ex: 2005, 2000, 1990), and in certain years the gap was trimmed to 24/1 or so (for ex: 2004, 1976). However, there seemed to be no pattern to the gap slimming or widening based on the homicide rate in general. Because Justifiable Homicide stats pre-1976 are hard to come by, I have divided the homicide rate of each period by 30.3 in order to come up with a rough estimate of it’s presumed number of Justifiable Homicides.
-2001 would have been 5.6 homicide rate per 100,000; but the 3,017 homicides of the 9/11 terrorist attacks bring it up to 6.7.
-Active duty average is 28 years old for US Military.
-Old, detailed drunk driving deaths are difficult to come by. We look at years 2003-2007:
Total fatalities 14,678/14,409/13,582/13,491/12,998
Death from those hit by drunk drivers 2438/2381/2583/2520/2207
As % of total drunk driving fatalities 16.6%/16.5%/19.0%/18.7%/17.0% respectively. A fairly consistent stat.
We will go ahead and reach a total for less known years (1982-2002) by simply taking 17.6% of all drunk driving deaths, under the assumption that since the wide use of motor vehicles this rate hasn’t changed much.
-The NHTSA only has drunk driving fatality stats for 1982 forward. The rate of alcohol impaired drivers of motor vehicle deaths being a % of total motor vehicle deaths was 48% in 1982, and 32% in 2007. Let’s assume the rate was as high as 1982’s in the 1970’s and early 80’s, as that era had an unusually high rate of alcohol consumption. And let’s further assume that in years before 1960 the rate was closer to present day’s 32%. Starting at 1974, the rate will be lowered by one point a year, until settling on 32% in year 1958. Also, no drunk driving deaths pre-1914 will be counted, as we can safely presume the novelty of the automobile prevented people from driving drunk and especially unlikely it would seem, hitting other people (cars being very rare). We will also take the 17.6% rate of non-driver deaths and lower that by 2% a year starting in 1928… is this period the sheer lack of cars on the road should have made non-driver deaths a rarity to the point of being negligible for our purposes.
-Drunk Driving deaths are considered any death involving a driver with a BAC level of .08 or higher.
-The homicide/manslaughter column includes convict executions.
-Korean war deaths of 36,574 are spread out fairly evenly over 1950-1953, owing to difficulty in finding accurate year-by-year casualty stats.
-War stats already counted into Excel sheet… Iraq 1 and 2, Vietnam, Korea,
-Since year by year population data in the 1800’s is hard to come by, we will take the known 10 year census stats and use each years average population increase each decade to arrive at the sum.
For the years 1882-1911, we will use the “top cities homicide rate” as provided in New York Times article of that period. There is some debate whether rural homicide rates were either lower or higher at this period in US history, we will hope that it was close enough to the city rates for our purposes.
-Allot of the estimated homicide rates pre-1900 come from figures of various combined cities. It should be noted that unlike recent American history, at that time rural populations actually had a higher homicide rate, not the other way around as is the case today.
-pre-1980 military homicides will need to be estimated in order to be included. For the 1980’s homicides averaged 109/yr, in a period that averaged approximately 2 million active duty; for years with similar active duty deployment we will use this number. For years of increased or lowered active duty numbers we will raise or lower the number appropriately. Due to the much lower homicide rates in general during the 1940’s and 1950’s, the 109 number will be used even during active duty # increases during the Korean War; using the assumption that soldiers being deployed in that conflict had a lower propensity to commit murders going in.
-pre WW2 military homicides will not be estimated and added to our table. This was America’s isolationist past, and as such, we can assume that domestic homicides would have been included in local totals, foreign base installations not having a significant number of American servicemen.
-It is believed that the US execution rate was around 10/100,000 in 1780. The category State Execution Rate reflects these deaths.
-For years previous to 1775, soldier and civilian deaths on now US soil will be counted in the British statistics.
-During Pontiacs Revolution an estimated 2000 civilians were captured or killed. For our purposes we will estimate 1000 killed.
-Although we don not have stats for lynchings before 1882, we know they were relatively common during Reconstruction (1865-1871), and before that, less so. I will use a graduated declining estimate for the period in question.
-Scholarly estimates put the execution rate of Colonial United States at around 10/100,000 in 1780, we will use this number, and lower it progressively until only .5/100,000 in 1853, by which time many states had abolished it’s use. Sentiment turning against the use of the death penalty very early in nation’s history I believe justifies this estimate. Executions became a statistical category as early as Vital Statistics of US 1880, which is the first year we will count state executions in the separate “Homicides + Manslaughter’s”category
-Although it may seem like a cop out; per-1859 homicide statistics will simply be given the rate of 5.1/100,000. Getting even vague year by year data is next to impossible. The fact that there were less firearms (and what firearms there were were of poor quality) I think we can reasonably estimate that although the early years of the United States history was brutal, that brutality primarily came from Indian conflicts, mass executions, and other domestic strife. There is little in literature of the era to make one think that murders were anything but uncommon.