Previous Post
Next Post

When smart people want to know how much time, money and money to spend protecting an asset—be it a nuclear missile silo or a corporate server room—they come to me. Or someone like me. As a Risk Analyst, I identify the most likely threat or threats to any given asset, so that its stewards can efficiently allocate resources for its defense. So let’s apply my experience in this new(ish) field to We The Armed Intelligentsia. How much should we be spending on self protection? What risk does violent crime pose to us average Joes?

Risk is defined as the combination of threat, vulnerability, and consequence (R=T*V*C), where threat and vulnerability are expressed over some period of time, usually one year. The result is something called an “annualized loss expectancy,” or how much you expect to lose every year from whatever it is you’re concerned about.

The basic idea: one way or another, you’re going to lose money. It’s better to invest the money now to prevent the bad stuff than to spend the money recovering from its loss. For the purposes of the current question, the annualized loss expectancy would be what we should budget for self protection annually.

OK, so, let’s start with Threat.

What’s the probability that you will be the victim of a violent crime? For our purposes, we’re including a large swath of crimes which have the potential to escalate into murder or injury, including murder, forcible rape, robbery, aggravated assault, burglary and the generic “violent crime” category as well.

According to the FBI crime statistics for the last two years (for cities with a population of over 100,000, the only reliable raw data I could find), you can expect to be a victim of a violent crime once every 100 years (a 1.081% annual probability), on average, across these United States.

Now, Vulnerability

Vulnerability in this sense isn’t “how vulnerable are you to bullets.” It means “GIVEN that you get into a violent confrontation, what’s the probability that you get injured or die?”

This can be difficult to determine; there’s no good database out there with this information. The CDC claims that .006% of the population for the year 2009 was injured by violent crime, while 0.000061% of the population were murdered.

Since this data already includes the “threat” percentage, we have a final value for the combination of threat and vulnerability (since everyone injured by a violent crime had to be involved in a violent crime). We (OK I) can work backwards from this number to determine the “conditional probability” or vulnerability to injury or death from violent crime, by dividing the respective numbers by the threat.

Consequence is a bitch.

Obviously death is a massive consequence. For the individual, there’s no higher price to pay. To put human life into economic terms, analysts use something called the “Value of a Statistical Life” or VSL. The figure that represents the impact of a fatality on the economy of the nation or the finances of a family.

The average college graduate makes just under $50,000, the average age of murder victims is 25, and the average cost of a funeral is about $6,500. So, for a family, the VSL of an income-generating individual is (6500+[(55-25)*50,000]). I make that $1,506,500. Combined with the murder rate, the final economic risk from murder is $91.90.

Injury is a consequence that we can experience ourselves, not something that we have to worry about leaving behind for our families. According to a paper titled “Valuing the costs of violent crime: a stated preference approach” by Giles Atkinson, Andrew Healeyy, and Susana Mouratoz, the costs related to an injury from a violent crime were around $6,000. So, the economic risk posed by injury from a violent crime would be (6000*.006) = $36.

Now that we have final figures for the risk from injury and murder as a result of a violent crime, we can add them together to get our grand total risk. According to the limited information available to me on the internet, the annualized economic loss expectancy from violent crime for an individual is $127.90. Therefore, $127.90 is the amount we should be spending each year on self defense.

The biggest issue with this risk value is the VSL calculation. In reality, there’s no possible way to put a price on the life of a human being. Analysts often do it out of convenience so the “actual” cost of a catastrophic event can be calculated. But putting a price on something so important seems impossible.

At the end of the day it’s your money and your life, and personally I’m going to spend a whole lot more than the price of a Hi-Point 9mm on protecting myself. Of course, if you average my expenditure with the expenditure of thousands of people who spend nothing on self-defense . . .

Previous Post
Next Post


  1. I’ll spend whatever it takes, because I spend more money just playing with guns than for protection. I also love the game of RISK and since I’m the local champ.

  2. good analisys , but I look at it from a simpler point of view. cost of defense stuff-$$$$
    not having my wife and daughter not end up like the pettits – priceless.

    • That’s the problem with risk analysis involving human lives. How do you value something that is inherently priceless? Whenever I do an analysis involving fatalities at work, I always break it out as a separate figure from the economic stuff. I just can’t bring myself to use VSL in my day job.

      I agree, it’s better to be penniless and alive than rich and dead. But some people don’t understand that concept, and need it spelled out that they actually have to spend SOME money on personal defense.

      • There is another problem with your analysis.

        This: “Obviously death is a massive consequence. For the individual, there’s no higher price to pay.” is not necessarily true.

        • Yes it is, in fact, absolutely and necessarily true. An individual may wrongly place another value (or values) higher than his own life, but the moment he ceases to live, those values die with him. An individual who no longer wishes to live, is for all intents and purposes, already dead.

        • “An individual may wrongly place another value (or values) higher than his own life, but the moment he ceases to live, those values die with him.”

          What do you mean by “wrongly”?

  3. If you just evaluate the consequence by income then you’re ignoring the dramatic improvement my existence makes to society.

  4. $127.90? Comparing that to what ive actually spent, I feel as silly as if I’d just bought Sam Waterston’s “Robot Insurance” from Saturday Night Live.

  5. It’s good to do this sort of analysis, but it’s important to know that the general notion of threat is trumped by local circumstance and immediate actions. I can find the probability of my make/model of car being stolen in my ZIP Code, but leaving the doors unlocked and keys in the ignition multiplies the risk tremendously. I grew up in a “100-year” flood plain but we got flooded every 2-4 years after a neighbor downstream did some creative infill to increase the size of their orchard.

    Here’s a dated but relevant web site that offers a more behaviorally oriented approach to assessing your own risk:

    I guess the point is that you have to assess YOUR SPECIFIC risk when you decide how much to spend–there’s a reason insurance companies do this.

  6. If the average murder age is 25 years old, would that lower the V for someone who is 45 years old?

    In other words, is their a sliding index to your numbers. If someone who works from home 95% of the time, they have in effect, reduced their vulnerability, perhaps, as opposed to someone who walks or rides public transportation to work, and against those people who drive their own car to a secured garage.

    I am just extrapolating on what I thought was a good general rule – can one adjust their risk assessment by looking at other factors and adjusting accordingly? Can a Risk Analysis look at “perceived threat”. What I mean is, as a 6′ 300 lb guy, it would be less likely to be attacked, at least physically, than someone who is 5′ 7″ 150 lbs. Right?

    • Tailoring the analysis to your age, education, and specific income will make the answer more relevant to you. I was only presenting the average, you can adjust the formula yourself and figure out your own risk.

      Deterrence is one of those factors that isn’t really well understood by the risk analysis community yet. We know it’s a major factor in attacker preference (attacking the 80 lb girl or the 300 lb guy, for example), but we can’t actually model it yet without a ton of information.

      I could go on for hours about these two points, but the short and dirty answer is you’re right, these are indeed factors that affect the final risk calculation.

      • Cool stuff – I love data… Wish you could build a spreadsheet that we can input the variable data tailored to ourselves to figure out our Risk Analysis.

  7. Nice picture Foghorn! Seems like yesterday when I was waiting to pick up my spanking new TLE. I got mine NIB for $820 OTD! It was money well spent. 5+ years of carry and 6K rounds later, she’s sexier than ever.

  8. If my wife saw that I should only spend $127.90 annually on….. ummmm…. “gun stuff”…. well… the “C” would be severe! 🙂


Please enter your comment!
Please enter your name here