2010 October 12 Tuesday
Economic Cost Per Murder Calculated

By one measure each murder costs $17.25 million.

AMES, Iowa -- Murder takes an obvious toll on society in terms of the loss of human life, but what does it actually cost each time there's a murder? It's about $17.25 million according a recent study by an Iowa State University sociologist.

But it has to depend on who gets murdered. If a violent criminal kills another violent criminal then seems to me there might not be a net cost to the rest of society. By contrast, murder of a CEO or CTO of a high-tech firm would have a very high cost to the rest of society of the lost guy is not easily replaced.

Matt DeLisi, an ISU associate professor of sociology and director of the criminal justice program, led a team of five Iowa State graduate students on the study of 654 convicted and incarcerated murderers. Expanding upon earlier monetization estimates, they calculated the costs of five crimes -- murder, rape, armed robbery, aggravated assault and burglary -- in terms of the victim costs, criminal justice system costs, lost productivity estimates for both the victim and the criminal, and estimates on the public's resulting willingness to pay to prevent future violence.

A serial killer is especially costly. So if a brain scan could identity with very high probability that a known criminal will eventually commit murder would you see preventive exile to a deserted island as acceptable?

"That each murder costs more than $17.25 million still does not convey the true costs imposed by homicide offenders in the current sample," the authors wrote. "Since the mean homicide conviction was more than one, the average murderer in these analyses actually imposed costs approaching $24 million. For the offender who murdered nine victims, the total murder-specific costs were $155,457,083!"

Measures to catch criminals earlier in their criminal careers would save a lot of money.

The ISU researchers also calculated costs of rape ($448,532), armed robbery ($335,733), aggravated assault ($145,379) and burglary ($41,288).

Share |      By Randall Parker at 2010 October 12 09:56 PM  Economics Crime

Mthson said at October 12, 2010 11:51 PM:

That's fascinating to see the actual numbers.

My girlfriend was recently mugged somewhat violently because some liberal thought it'd be a good idea to put a rent-assisted, lower income housing building in the otherwise nice area of the city where she lives.

It made me think about crime... some countries on the list of countries by homicide rate do well by being soft (Scandinavia, Germany), and some do well by being hard (Singapore, most of the Muslim world). ( http://en.wikipedia.org/wiki/List_of_countries_by_intentional_homicide_rate )

While hardness could seemingly work anywhere, particularly if you have a diverse population, it doesn't seem like softness is doing much for the US crime rate. If we told liberalism in 1960 that liberalism was going to gain control of the culture, but during that period crime would skyrocket to 3rd world levels, even in some of the most liberal states with the highest tax rates, they would have been surprised.

Black Death said at October 13, 2010 7:29 AM:

The article cited (Murder by numbers: monetary costs imposed by a sample of homicide offenders
Matt DeLisi; Anna Kosloski; Molly Sween; Emily Hachmeister; Matt Moore; Alan Drury ) is behind a for-pay firewall ($30!), so I wasn't able to read any more than the abstract, which doesn't say much. However, the cost estimates are based mainly on work by Cohen and Piquero (http://webcache.googleusercontent.com/search?q=cache:euaGLaBl808J:www.soc.iastate.edu/staff/delisi/murder%2520by%2520numbers.pdf+Cohen+and+Piquero&cd=4&hl=en&ct=clnk&gl=us). This is what they found:

Similarly, DeLisi and Gatling (2003) suggested that ‘the manyqualitative consequences of these criminal victimizations are incalculable’(p. 291). To redress this, Cohen, Rust, Steen, and Tidd (2004) calculatedwillingness-to-pay (WTP) estimates which are the amount of money thatcitizens would be willing to pay to prevent crimes. Based on a nationallyrepresentative sample (n = 1300), Cohen et al. (2004) found that WTPestimates were between 1.5 and 10 times higher than previous estimates ofthe costs of crime because they encompassed collateral costs. These costsincluded prevention expenditures for personal security, avoidant behaviorsto safeguard against victimization, third-party costs of insurance, andgovernment welfare programs.


Of the $17.25 million cost they calculate, about $5 million is direct victim costs, incarceration and justice costs and lost productivity. The rest, or about 71% of the total, is WTP. Now this seems like a pretty nebulous concept to me - asking people an abstract question about how many dollars (Other Peoples Money - OPM) they would be willing to spend to prevent a hypothetical murder. Remember, these dollars aren't coming out of their own pockets - they are hypothetical dollars, or OPM. I don't really disagree with the concept that there are severe negative externalities associated with crime that are very difficult to estimate. It's just that I think that they are, well, very difficult to estimate. When numbers like $17.25 million per murder get bandied about and acquire lives of their own, I feel we ought to give some thought to how these numbers were derived. After all, this isn't theoretical physics. In a lot of these adventures, you can make the numbers come out any way you want.

gig said at October 13, 2010 8:44 AM:

If we ever adopt again some sort of sterilization program of poor people, this is exactly the kind of amoral study that would support the program.

Jehu said at October 13, 2010 9:16 AM:

Do a regression analysis on property values in the attempt to extract how much the local crime rates for the several crimes affects the value of the property? That'd give you an estimate of how much people ACTUALLY pay to avoid crime. Still not perfect obviously, as it wouldn't count the amount people pay in various hardening measures, like guards, gated communities, concealed weapons, police bond measures, et al. But it's a great place to start.

John C. said at October 13, 2010 11:40 AM:

"By contrast, murder of a CEO or CTO of a high-tech firm"

Almost nobody is irreplaceable, and what if the next CEO is actually better and significantly increases exports, is this gain taken off the 17 mil? A good year for even a smallish company might wipe out the loss for all murders that year. "Studies" like this should confine themselves to direct provable costs.

"So if a brain scan could identity"

Or we could just pay him say 75 grand a year which he loses if he is caught engaging in felonies and save ourselves a bundle. In some cases his fellow predators will off him in a couple of years anyway and we'd save even more!

no i don't said at October 13, 2010 2:06 PM:

"A serial killer is especially costly. So if a brain scan could identity with very high probability that a known criminal will eventually commit murder would you see preventive exile to a deserted island as acceptable?"

You mean we should convict somebody based on what some INTERPRET he MIGHT do??? It would sound funny if it didn't sound medieval, to say the least. In fact it sounds just as criminal. Reminds me of the Inquisition in my History class.

Besides the procurement of justice has nothing to do either with "revenge" or "saving money" it has to do with and should provide Social Readaptation.

Engineer-Poet said at October 13, 2010 6:08 PM:

It looks like one armed robbery creates damages equal to the value of a small apartment building.

This suggests a remedy:  when the cost of the crimes committed by residents of subsidized housing equal the value of that housing, it's deeded over to the victims with the victims given veto power over residents.  Sooner or later (probably sooner) nobody will take a public subsidy to rent to people who victimize the public.

Yes You Do said at October 13, 2010 6:09 PM:

"Besides the procurement of justice has nothing to do either with "revenge" or "saving money" it has to do with and should provide Social Readaptation."

Hey, it worked for Willie Horton!

Randall Parker said at October 13, 2010 6:26 PM:

Black Death,

At $42k per capita GDP it does seem a stretch to make a murder cost over $17 mil. I can see it for some murders, but not most.


How is the study amoral? Did it assign too low a cost to murders or too high?

John C.,

Lots of people are very hard to replace. I've seen companies function more poorly due to loss of a single key manager or engineer. It is hard to find people who can do some forms of highly intellectually difficult work.

California Kid said at October 14, 2010 12:14 AM:

What's the economic cost of an entire abandoned city ? What's the cost of having to rebuild it all as a ring-city on the outskirts ? What's the cost of having to do this repeatedly, throughout time and space ?

How much of the "economy" consists of the productive, decent population having to rebuild its entire living space over and over again ? I mean, just to stay alive and un-raped ? All with the connivance of the government, media, Hollywood, universities, etc.

That's the USA.

gig said at October 14, 2010 5:24 AM:

@ Randall

It is amoral because it inputs a value to life. That´s not Christianity, for sure. But once it is done, it is a small step to calculate the "net contribution" each individual does for society. That´s when sterilization of the poor becomes acceptable once again

Engineer-Poet said at October 14, 2010 5:42 AM:

If the poor's child-bearing wasn't subsidized, they'd be more than happy to be sterilized.  Drug-addicted women take offers of free sterilization quite readily, even more if they get paid.

What kind of person argues that drug addicts should pump out more babies on the basis of morality?  What's the morality of taking money from people who could support their own children and make them productive citizens and giving it to people who demonstrably cannot?

not anon or anonymous said at October 14, 2010 6:18 AM:

"At $42k per capita GDP it does seem a stretch to make a murder cost over $17 mil. I can see it for some murders, but not most."

Why? The value of a life has very little to do with monetary expenditure as measured by GDP. The value of a statistical life as measured by economists is about $3 mil to $10 mil, so $17 mil for the total costs of wrongful death is a plausible estimate.

It's true that "fair compensation" cannot make the victim whole in this case; but people routinely accept small risks to life in exchange for some sort of benefit. Moreover, many people buy life insurance in order to compensate third parties in the event of their own deaths. This suggests that (in principle) large damage awards could compensate a victim of wrongful death by reducing the need for premium-financed life insurance.

gig said at October 14, 2010 7:16 AM:

@ Engineer

I completely agree. We just need to build the case for stoping the subsides for the reproduction of poor people. As I wrote many times in Roissy, you can never win that argument by pointing at their low IQs. We should highlight how ugly poor women are and how costly poor people in general are.

Take a look at the link below:

How long before that beacon of enlightment, the British middle class, realyzes how much they have been lied to?

Michael L said at October 14, 2010 8:07 AM:

"value of a statistical life as measured by economists is about $3 mil to $10 mil" - so how exactly do you measure something like this?

I suspect that these numbers are actually not the "productive value" or "expected income" of any sort but rather the "expense required or acceptable to prevent a single expected death". Sort of like, if we expect that one person per year will die in an accident crossing a busy street at a particular place, then we should be build an underground passageway for pedestrians if that passageway costs less than 3 mil. And if it costs more like 15 mil, then it isn't worth the expense. This is reasonable and sound logic in the world with limited resources, but note that there is nothing magic about such numbers, they are purely policy decisions based on how much money we have to spend. In a poorer country their "prevent accidents" budget would be much lower, hence the maximum acceptable price would be similarly lower.

Engineer-Poet said at October 14, 2010 8:40 AM:
We should highlight how ugly poor women are and how costly poor people in general are.
Better yet, point out that the country is short of highly-educated workers (not peasants), and the single best predictor of a child's educational achievement is their parents' educational achievement.  Therefore, we should discourage people from reproducing before they have an education (and the poor, largely being that way because of lack of education and other lack of ability, will be discouraged).  We should also point out that criminality is predicted by family history, and give no subsidy to people who don't have the self-control to stay out of trouble.

To avoid this from becoming another huge subsidy to the school racket, credit by exam or work history should be mandatory.

Underachiever said at October 14, 2010 7:44 PM:

"By one measure each murder costs $17.25 million."

All the more reason to use fMRIs in the court system.

Randall Parker said at October 14, 2010 8:14 PM:


Why is it immoral to put a value on life?

Let me put it another way: Should we treat all people as being of equal value? For example, should we treat murderers and rapists as of equal value to emergency room physicians or firemen? Should we treat Bernie Madoff as being of equal value to Nobelist in Medicine?

If we do not see all as being of equal value then we have started to rank people by value. Whether we then put price tags on these rankings is just a question of whether to use currency or another measure of relative value.

Sterilization of the poor: In your mind is there a moral difference between forced sterilization versus voluntary sterilization in exchange for a cash payment? If we offered all high school drop-outs $20k to get sterilized would that be immoral?

Engineer-Poet said at October 15, 2010 8:02 AM:

Waiting for them to drop out is probably too late in many cases.  How many welfare "families" bear several children before 18?

There are reversible methods for women (plastic clips on the fallopian tubes).  I'd be happy to pay the academic failures a stipend through age 21 if they got their tubes closed before having any children.  They could get the clips taken off if they managed to be successful enough to support a family.  Since this is far easier for couples than singles, it would also be a pro-marriage initiative.

I predict that a program like this would result in a crash of welfare payments and caseloads for social workers.

sg said at October 16, 2010 4:38 PM:

Many large dangerous projects actually include the cost of deaths that will occur during construction based on the statistical incidence of deaths on similar projects. They use the statistical cost of life model.

Randall Parker said at October 16, 2010 6:20 PM:


There's also a price goal per additional life saved when it comes to safety equipment. For example, there's some marginal number of dollars spent per saved life in commercial aviation.

The US government does similar things with pollution controls. How many dollars per life saved with each level of increment of diesel engine exhaust for example.

bbartlog said at October 17, 2010 9:26 AM:

The difficulties involved in estimating the largest cost component (WTP) make this a little hard to take seriously. Also I have to wonder whether they took into account victim and perp demographics when estimating lost productivity due to murder; you're looking at a lot of black-on-black murders in the US, and for that matter even within the black population you're talking disproportionately about the least productive (and least likely to become productive) members. In short I expect this paper was written with the goal of generating an eye-popping number and I don't believe the result should be taken seriously.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Web parapundit.com
Go Read More Posts On ParaPundit
Site Traffic Info
The contents of this site are copyright ©