18 posts / 0 new
Last post
OhioMike
OhioMike's picture
Looking for primary sources on the nature of criminal violence

Hello all,

I have recently been looking for primary source material, surveys, case studies etc on what actually happens during criminal violence. I can find a lot of people talking about what happens but the best of that material seems to be from police and is based on experience rather than true data. And SO much of the rest of what is available is people talking about limited personal experience and is some cases just parroting back things they were told by others. My class is occasionally asked to teach self protection (to local women/teens groups for the most part) and I want to accurately reflect the differences between social, asocial, and domestic violence. So I am trying to find the data in as raw a form as I can so that I let the data speak to me rather as much as possible.

I have been able to locate the BJS which has a lot of data related to preverlance rates for attackers, victims, locations, etc. But it is very limited on descriptions of the actual encounters, which attacks are most common etc.

Thanks,

Mike 

Iain Abernethy
Iain Abernethy's picture

I’m not sure about the USA, but there in the UK the Office of National Statistics is a useful resource:

https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice

Some general thoughts on the use of crime data:

Martial artists have a bad habit of wanting to focus in on the physical actions of the criminals alone. However, it is much more important to look at the wider trends i.e. what is most likely to happen to what groups of people?

I use the crime data to ensure my self-defence courses match the group I’m working with. The course will be different for the elderly, young men, young women, etc. True self-defence training needs to place the greatest emphasis on personal security issues; not physical technique. We therefore need the crime data to inform us what security precautions would be both appropriate and commensurate with the actual level of the threat (i.e. don’t engender harmful paranoia).

As an example, people may want to learn physical knife defences as a result recent trends in the UK, but the data can be used to show that the physical risk of knife crime remains minuscule for the vast majority of the population outside of certain demographics in given locations. The correct approach is therefore addressing the disproportionate fear (by putting the real-world risk into context) and giving advice that relates to reality of knife crime as it effects the people being instructed. That advice will vary depending on the specifics of person being instructed and the location where they live. You therefore need the crime data for that.

If we ignore the big picture stuff and try to seek “X% left to right slash, Y% right to left slash, Z% upward thrust, etc” figures for knife crime then we are unlikely to give our self-defence students any advice that will actually help them. It also important to know that all violence is frenzied and detailed figures are often unavailable as result (although injury figures can help get more detail).

The statics on detailed actions, where they exist, can be comforting because they seem to imply an order in the chaos. However, that “X% left to right slash” is one of 4 things that probably happened in a single second. They are therefore of very limited value.

Finally, I’m not a big fan of the “he does this, you do that” approach to violence. Knowing what the enemy’s likely actions are is important; but structuring training as a series of reactions to defined actions is highly problematic. It essentially puts the enemy in charge because they are dictating your actions and it promotes reaction over action (extremely unlikely to work). Far better we emphasise pre-emption and being proactive. That is way more likely to work than teaching a set of defences to the enemy’s actions. Using crime / injury figures to support a “they do, you do” centred approach is not a good use of the data and it is not good self-protection.

We need to know what we are facing and the collective experience of many is way more useful that the subjective experience of any given individual. Knowledge of crime statistics and trends is therefore vital for anyone purporting to teach self-defence. That knowledge needs to be applied to a solid approach to self-protection; and that should mean the greatest emphasis being placed on personal security as it relates to the individual being instructed. We need to avoid the trap of focusing in on the physical alone. We also need to avoid sanitising the chaos of violence as it happens with a comforting “statistical lens”. Finally, we need to avoid statistics of specific physical actions / injuries being used to support (intentionally or unintentionally) reactive training and thinking.

In conclusion, we need to understand the problem we are seeking to provide solutions to. Crime statistics are vital in this regard. However, having understood the problem, we also need to ensure that the solution offered is as effective as possible for the individuals being instructed.  

All the best,

Iain

AnthonyC1118
AnthonyC1118's picture

Hey Iain and co,

A great resource for crime statistics in the US is the UCR (Uniform Crime Report). It's a database compiled by the FBI using data from participating police agencies across the country.

https://www.fbi.gov/services/cjis/ucr (Main Site)

https://ucr.fbi.gov/crime-in-the-u.s/2018/preliminary-report (2018 crime data for the US, January - June)

TW Smith
TW Smith's picture

Excellent Piece Iain, 

I will refer to it in upcoming podcast too. 

Iain Abernethy
Iain Abernethy's picture

AnthonyC1118 wrote:
A great resource for crime statistics in the US is the UCR (Uniform Crime Report). It's a database compiled by the FBI using data from participating police agencies across the country.

Fantastic! Thank you for sharing that resource. Much appreciated!

TW Smith wrote:
Excellent Piece Iain, I will refer to it in upcoming podcast too.

Thank you! I’m pleased you like it and I look forward to hearing the podcast!

All the best,

Iain

Gerth
Gerth's picture

Hi All, 

I’m probably missing the obvious but, following the office of national statistics link, how do you drill down into the ‘violent crime’ figures to see the breakdown? Or do you have to pay to see that level of detail? 

Thanks

Gareth

Iain Abernethy
Iain Abernethy's picture

Gerth wrote:
how do you drill down into the ‘violent crime’ figures to see the breakdown? Or do you have to pay to see that level of detail?

It’s all free and it sounds like you are not following the links? For example:

Main Page:

https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice

Where you can click on “Crime in England and Wales: year ending March 2019”:

https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/crimeinenglandandwales/yearendingmarch2019

 You can then click on the links in the “Table of contents”, for example:

6. No change in the most common types of violent crime, but a fall in homicides”:

https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/crimeinenglandandwales/yearendingmarch2019#no-change-in-the-most-common-types-of-violent-crime-but-a-fall-in-homicides

And then on further links for the breakdown of the figures, statistics and tables. Lots to explore!

I hope that helps.

All the best,

Iain

sarflondonboydo...
sarflondonboydonewell's picture

When I did research a good few years back I found the following publication very useful;

the home office research paper 254 'the nature of personal robbery' especially chapter 3 and 4.

Also look at what is known as the 'cardiff model' in reducing violence linked to the night time economy. 

I have somewehere the Ancident and Emergency treatment figures that were used as part of the intial reseach which look at what type of injuries were being treated by Cardiff hospital. In essence it is what we know that around 80% to 90% were facial injuires and injuries around the neck area as of a result of blows ( fists mainly).  I concluded from that and other research  that handling a punch is evidentially the more than likely form of assault. A few years back research carried out from memory from a scandianvian hospital regarding knife wounds was very enlighting. If I can find the research paper I will post the link.

Tau
Tau's picture

sarflondonboydonewell wrote:

Also look at what is known as the 'cardiff model' in reducing violence linked to the night time economy. 

I have somewehere the Ancident and Emergency treatment figures that were used as part of the intial reseach which look at what type of injuries were being treated by Cardiff hospital. In essence it is what we know that around 80% to 90% were facial injuires and injuries around the neck area as of a result of blows ( fists mainly).  I concluded from that and other research  that handling a punch is evidentially the more than likely form of assault. A few years back research carried out from memory from a scandianvian hospital regarding knife wounds was very enlighting. If I can find the research paper I will post the link.

It is interesting that you cite Accident and Emergency. I would like to take this opportunity to shamelessly plug my book, if only for one of the bonus sections. In it I discuss the nature of violence as observed from the perspective of a health care professional specialising in emergency care. I don't believe I add anything to the information already out there but I do strongly believe that I endorse that information from looking from a different viewpoint.

Steve Gombosi
Steve Gombosi's picture

In the US, the best source is probably the Uniform Crime Report compiled by the FBI and published by the Bureau of Justice Statistics of the US Department of Justice. There's a reporting tool available at https://www.ucrdatatool.gov/. I don't think it's broken down as well as the UK statistics, and it's dependent on the quality of voluntary reporting by local and state law enforcement agencies (which is highly variable). The nature of the US Federal system means that this sort of thing is necessarily more loosely organized than the UK (crimes like "assault" are defined in state law and may be differently defined from state to state). Also, since we don't have anything like the NHS the data on injury types, etc. isn't as uniform or available as it is in the UK.

OhioMike
OhioMike's picture

I agree completely, a good portion of my desire to find the data is to look for the Who, Where, When type of information since that is more useful in my mind for formulating strategies to avoid violence rather than tactics to deal with someone once they have become violent. . A lot of my interest revolves less around the "he does this, you do that" than around understanding the stages of violent encounters and understanding better the transitions between them. 'Seeing the interview' to paraphrase "The Little Black Book of Violence".  But the interview process varies based on type of crime (asocial, social, domestic) and victim (male, female, young, old, etc.) hence my interest in the raw data.  

That being said if we intend to teach physical skills for self production (not that those are always necessary) it is reasonable to ask the question of What form the violence takes in order to make certain that we are not focusing on a narrow subset of criminal violence in order to fit martial arts based solutions. 

Thanks,

Mike

Josh Pittman
Josh Pittman's picture

Hi, everyone

In Iain’s cynical and empiricist spirit (just kidding, Iain!), I, too, have been trying to find this kind of statistic in order to fact-check the various claims I’ve heard about most common types of attack.

Trying to find the Accident and Emergency report that sarflondonboydonewell referenced, I came across this CDC report on US satistics: https://www.cdc.gov/nchs/data/nhamcs/web_tables/2016_ed_web_tables.pdf. There are two relevant tables, as far as I can tell, one on ED admissions by the intent of the injury (assault accounts for 3.6% of admissions) and by the mechanism (cutting/piercing injuries account for 5.2% of admissions, firearms for 0.2%, and blunt force trauma for 13%). Perhaps someone more talented in data interpretation can make conclusions from these numbers relevant to us.

This other report, specifically about deaths, has some stastics on death by firearm. Naturally, those data include accidental death, but perhaps they have some value to us: https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_09-508.pdf. Firearm injuries account for 16.4% of all injury deaths in 2017, and 36.6% of those injuries were homicide (compare to 60% suicide).

The Bureau of Justice Statistics report, referenced by Steve Gombosi, is here: https://www.bjs.gov/index.cfm?ty=tp&tid=316. It looks promising, but I haven’t delved into it yet.

Thanks for starting this thread, Mike!

Steve Gombosi
Steve Gombosi's picture

One of the more interesting things I gleaned from perusing the UCR is that nearly half of assaults (47%, IIRC) seem to involve multiple assailants, at least in the US. Again, the US statistics might be skewed because of the voluntary nature of the reporting. It's quite possible that many simple assaults don't get reported to local law enforcement at all and that local authorities are not uniformly diligent in reporting assaults to DOJ (since the Federal government doesn't have authority to require this). Similarly, CDC data might be skewed because, given the lack of universal healthcare in the US, some individuals might only seek treatment for the most serious injuries and not all medical personnel are going to report the injuries they do treat.

My completely unsubstantiated guess is that this might cause the overall incidence of assaults (particularly minor ones) in the US to be underreported by quite a bit.  I think the UK statistics (both the crime reporting and the medical data) might be more accurate overall.

Iain Abernethy
Iain Abernethy's picture

Steve Gombosi wrote:
One of the more interesting things I gleaned from perusing the UCR is that nearly half of assaults (47%, IIRC) seem to involve multiple assailants, at least in the US.

That’s comparable with figures in the UK. The Office of National Statistics has multiples assailants recorded at 43% for all violent crime where the assailants are not known to the victim. It’s around 24% when the assailants are known to the victim; and way down at 1% when it comes to domestic abuse. What is clear is that training methodologies that focus exclusively on one-on-one situations are ignoring a huge part of the problem.

This video has more on this point.

All the best,

Iain

Steve Gombosi
Steve Gombosi's picture

Iain Abernethy wrote:
That’s comparable with figures in the UK. The Office of National Statistics has multiples assailants recorded at 43% for all violent crime where the assailants are not known to the victim. It’s around 24% when the assailants are known to the victim; and way down at 1% when it comes to domestic abuse. What is clear is that training methodologies that focus exclusively on one-on-one situations are ignoring a huge part of the problem

There's a rule of thumb that I've heard from police in the US: "Where there's one, there's two." It's just a reminder to be on the lookout for accomplices when in a confrontation. It's probably something we should all keep in mind. Even bad guys have friends.

Iain Abernethy
Iain Abernethy's picture

Steve Gombosi wrote:
There's a rule of thumb that I've heard from police in the US: "Where there's one, there's two."

It's wise to make that assumption and I like that turn of phrase.

All the best,

Iain

OhioMike
OhioMike's picture

The BJS just released their report and data from 2018 for their victimization survey, this has a lot of hard data on the victims in the US.

https://www.bjs.gov/index.cfm?ty=pbdetail&iid=6686

Interesting highlights from the my first look at the data. 23 violent attacks per 1000, about a 10% increase from the past several years, of that 8.6 of them were more serious that simple assault, of that 8.6 incidents 4.8 involved a weapon, and of that 4.8 incidents 1.7 times that weapon was a firearm. So about 40% of the time an attack is going to be more serious that a couple of guys punching it out in a parking lot. And while about 20% of violent encounters will involve a weapon of some sort, thankfully only about 8% of total violent incidents will involve a gun. 

Other troubling data is that women now exceed men in the rates of criminal victimization, in particular the rates excluding simple assault, and the rate of sexual assault have doubled in 2018 compared to the last several years.  Dometic violence is also about 10% higher than previous years.

I think I will move my analysis of the data to a fresh topic, there is a lot of data in the report and I will try and update the group as I get thru it.

Thanks,

Mike

Iain Abernethy
Iain Abernethy's picture

Hi Mike,

OhioMike wrote:
The BJS just released their report and data from 2018 for their victimization survey, this has a lot of hard data on the victims in the US.

Thank you for sharing this information. This thread is proving to be a good resource. I’m also loving that we have international data for the purposes of comparing and contrasting.

If we are to be able to mitigate a problem, we need to know what that problem is. Statistics are vital in this because personal experience can never give the big picture.

All the best,

Iain