Mises Wire

Facebook icon
LinkedIn icon
Twitter icon
< | < | <

An Approach to Assessing Crime Severity


Tags Legal SystemMedia and Culture


In 1977, the US Department of Justice conducted an interesting auxiliary study as part of the National Crime Victimization Survey. They asked 60,000 people how severe 204 crimes were (each person rated 25 criminal events). Yes, yes, interpersonal utility alert; but let's go on. On a scale where 10 = "a person steals a bicycle parked on the street", here are some of the study findings:

0.6 = a person trespasses in the backyard of a private home

1.5 = a person intentionally shoves or pushes a victim. No medical treatment is required.

3.2 = a person breaks into a building and steals property worth $10 (1977)

4.9 = a person snatches a handbag containing $10 (1977) from a victim on the street

6.2 = a person beats a victim with his fists. The victim requires treatment by a doctor but not hospitalization.

6.2 = an employee embezzles $1000 (1977) from his employer.

7.9 = a person intentionally hits a victim with a lead pipe. No medical treatment is required.

9.0 = a person, armed with a lead pipe, robs a victim of $1000 (1977). No physical harm occurs.

11.8 = a person stabs a victim with a knife. No medical treatment is required.

15.5 = a person breaks into a bank at night and steals $100,000 (1977)

17.7 = an employer orders one of his employees to commit a serious crime

17.8 = a person intentionally shoots a victim with a gun. The victim is wounded slightly and does not require medical treatment.

21.2 = a person kidnaps a victim.

22.9 = a parent beats his young child with his fists. The child requires hospitalization.

27.9 = a woman stabs her husband. As a result, he dies.

32.7 = an armed person skyjacks an airplane and holds the crew and passengers hostage until a ransom is paid.

39.2 = a man stabs his wife. As a result, she dies.

47.8 = a parent beats his young child with his fists. As a result, the child dies.

52.8 = a man forcibly rapes a woman. As a result of physical injuries, she dies.

72.1 = a person plants a bomb in a public building. The bomb explodes and 20 people are killed.


There are also the malum prohibitum offenses, such as:

0.5 = a person takes part in a dice game in an alley

1.3 = two persons willingly engage in a homosexual act

1.6 = a person is a customer in a house of prostitution

3.2 = an employer illegally threatens to fire employees if they join a labor union

4.6 = a person cheats on his Federal income tax return

6.5 = a person uses heroin

8.5 = a person sells marijuana to others for resale

20.6 = a person sells heroin to others for resale

33.8 = a person runs a narcotics ring


There are all sorts of things wrong with this survey (and a few inconsistencies and methodological flaws), but at root it has a deep appeal. As Benson has queried (here, p.80), "More significant is the question of how to determine the so-called 'punitive' or unmeasurable damages portion of restitution reflecting the harms associated with the invasion of a person's property rights." This study is one possible way to assess community standards of this sort. I have elsewhere suggested that mediation case studies could form the basis for arbiters, but casework takes time to build. Surveys could provide an independent assessment of severity. But if that's the case, why not just have, say, 12 people chosen at random to learn all the facts of the case and do an on-the-spot survey. Hmm, maybe there's something to the jury system after all.

Image source:

Add Comment

Shield icon wire