Multiple Puzzles in Criminology

Crime is a complex problem. There are many aspects of this problem, and the pieces fit together like a big puzzle.  The difficulty in solving this puzzle was noted by Lombroso (1899/1911: p. 1), who noted that “Every crime has its origin in a multiplicity of causes, often intertwined and confused, each of which we must, in obedience to the necessities of thought and speech, investigate singly.”  We have to look at each piece singly, but we also have to look at how the various pieces of the puzzle fit together, and this is a difficult proposition. I have been trying to look for patterns related to the criminological puzzle and working to find some order in the chaos.  I think I have succeeded in discovering several important pieces to the puzzle, and am promoting my solutions on this site.

Some of the major pieces of the puzzle are related to the age crime curve and the problems relating criminal propensity, sanctioning processes, and crime rates. The solution to these problems are intertwined and difficult to separate. In essence, they need to be solved at once, but this is difficult at best. Therefore, I have tried to provide separate explanations of the pieces, and then put them together.

There are several sections of the criminological puzzle that can be solved by thinking differently about the methods needed to develop a solution.  Some of the major pieces of the criminological puzzle are presented below.

   1. Criminal propensity is a Massively Multivariate, Complex, and Normal (MMCaN) latent trait
   2. Criminal sanctioning is asymmetric
   3. Crime is a sigmoid cumulative distribution function of criminal propensity and sanctioning.
   4. Crime varies with age because of a developmental lag between strength and mental capacity.
   5. The criminal propensity of individuals is both relatively stable and highly variable over time.
   6. Several factors control crime at the population level, and these might not be what you think they are.

1. Crime is a sigmoid function of criminal propensity and sanctioning

The importance of this section of the puzzle is difficult to understate. If we do not deal with the sigmoid nature of crime, we cannot solve the age crime curve or the prisonization problems. The pieces of this section of the puzzle have been around for over a century, yet it does not appear that anyone has out them together.

De Candole (1829) noted that we cannot look at crime rates without examining both the individual propensity for crime and the societal reaction to crime. His message seems to have been lost over time, but the importance of this fact must be understood. The nature of criminal propensity and sanctioning are intertwined in a complex mathematical relationship which is known as the probit model. The essence of this model is as follows,

   1. Criminal propensity is a function of many independent factors
   2. Because of the central limit process, criminal propensity is normal
   3. Societal sanctioning is asymmetric
   4. If propensity is normal and sanctioning is asymmetric, and crime is sigmoid

An understanding of the processes that create crime rates is essential. However, the underlying processes are difficult to understand. The basic pieces of this section of the criminological puzzle have been around since the mid 1800s. Quetelet (1848) noted that crime was normal and that sanctioning was asymmetric, but he did not make the connection that if these two facts were true, crime was sigmoid. I will try to explain why this should be expected in a simple way, so that the sigmoid nature of crime rates can be understood.

2. The Age Crime Curve and the Developmental Lag between Strength and Mental Capacity

One of the enduring puzzles related to criminal behavior is the way that crime varies over the life course.  The common name for this pattern is the age crime curve. Crimes are rare in childhood, increase to a peak in late adolescence, and then drop rapidly in adulthood.  The reasons for the age crime curve are not well understood.  I will demonstrate that if one uses the first section of the puzzle, realizing that crime is sigmoid, it is possible to solve the age crime curve puzzle by proposing a developmental lag between the development of physical strength, and the development of mental capacity.

3. The Stability/Variability Puzzle

Another question for criminologists is “how can criminal propensity be both stable and variable at the same time?”  There seems to be a fair amount of rank order stability in criminal propensity.  People who commit crimes today have a strong tendency to be committing crimes in the near future.  Factors such as aggression or low self-control, which have a fairly strong association with criminal behavior, tend to have a strong rank order stability over time.  The rank order stability coefficients of criminal propensity are in the r=.8 range in the short term and in the r=.5 range over several years.  The rank order stability of crime correlates suggests that criminal propensity is fairly stable.  However, the stability does not lead to predictability.  The best predictors of criminal behavior tend to fall short of the r=.4 level of prediction.  The reasons for this poor predictability are not well understood.  Somewhere within the stability of criminal propensity, there must be a large amount of variability. I will explore the reasons for this stability/variability paradox.

4. The Factors Affecting Crime at the Population Level

One of the puzzles in criminology is that the relationship between age and crime seems to hold at the individual level, but not at the population level. I will demonstrate that other factors could be involved. It could be that the age structure of society has multiple effects on the crime rate.

A Different Approach to the Solution to the Criminological Puzzle

The prevailing method of examining criminological theories has been to look at each theory singly. While this is helpful, it does not address some of the bigger issues such as the multiple factor nature of crime, and how biological development affects the macro structure of crime over the life course. By building mathematical models, it is possible to address some of these larger issues. This can then be used to bring the focus back to the individual theories.

This site is a work in progress. If you have any questions or suggestions, please contact me.

References
Lombroso, Cesare (1899/1911). Crime: Its Causes and Remedies. Translated by Henry P. Horton, with an introduction by Maurice Parmelee. Boston: Little, Brown, and Company.