Monday, August 9, 2010

The crime triangle - a veneer

Proper analysis is the first step to prevent crime

I like the crime triangle. It is popular among crime analysts. It helps analyze crime hotspots. It is also part of problem-oriented policing. Sometimes too it is part of the prevention practitioner’s toolbox. It has elegant utility and simplicity.

The crime triangle emerges from "routine activity theory" (RA) in the early 1980s. RA explained some behavior quite well, like predatory crime (stalking). It did so with a simple premise: crime converges at the intersection of likely offenders, suitable targets and an absence of guardianship (or, more recently, "handlers").

In lay terms, picture the three sides of a triangle with an offender, a victim, and a target/place. When those things come together, so the theory goes, crime goes up.

The crime triangle is useful. Break the triangle and you prevent the crime. Want to increase guardianship? Get property managers to keep better control of their properties. Improve bar management in bars that over-serve. Simple. Elegant.

So far, so good. Except for one thing; that is where it typically ends in the RA world. The crime triangle does not dig deeper into the causes of crime.

Why? Because routine activity (and its crime triangle progeny) is one of those crime and place theories.


RA is less a causal theory explaining why and more a descriptive symbol predicting when, where or how. It ignores why someone becomes motivated in the first place. RA assumes an endless supply of motivated offenders. They are motivated for some reason; we just don't know why. The only "explanation" of motive falls back to rational choice theory.

Rational choice assumes offenders are rational actors who weigh risks against rewards. From the window of RA (and the crime triangle), crime looks like a "normal" condition of life.

True, some criminal behavior is “normal” in the sense that as events, products, and social affairs change, so too do crime opportunities. But as a causal theory, that’s rather trifling. It’s a bit like saying with enough water, sun and moderate temperature, certain environmental conditions will produce rain.

Then again some criminal behavior is not normal at all and RA just doesn’t work. Consider the story of the Connecticut mass murder in the news today.

As the NY Times says, this is “the latest in a series of American workplace tragedies”. It is a sad story about a workplace shooter who killed numerous workmates and then himself. He may have snapped from perceived workplace injustice. Perhaps he was clouded in a drug stupor. Maybe he was insane.

Routine activity theory might suggest how to remove opportunity for future incidents like this. That is a good start. Baby steps.

But RA theory will never actually know why because it will never ask. In cases such as this, the risk and reward assumptions of crime-and-place theories look rather silly. What does a suicidal shooter “risk”? What “reward” was this shooter gaining? Vengeance? (If so, we’re back to motive.)

Motive is a key piece of the puzzle


What will the crime triangle tell us in cases like this?

1. Capable guardians - cameras, plentiful supervision, and so forth. Unfortunately this shooting occurred at shift change when there were lots of employees and supervisors about. As for CCTV, how often do we watch nighttime news clips of robbery/shootings on corner-store CCTV? Cameras don’t stop shootings.

2. The time/workplace environment – preventing guns in the workplace. Will metal detectors work? Perhaps, but how difficult is it for shooters to become bombers. What then? Bomb sniffing dogs? At some point Orwellian paranoia replaces civility. Where do we stop? Body cavity searches?

3. The offender – modus operandi (not motive. Remember, RA is motive-neutered). The crime triangle asks if "handlers" like armed security might have intervened (that actually might have helped). Or maybe we could have prevented the shooter from getting guns in the first place? Others can argue 2nd Amendment rights. I won't bother here.

Crime triangle questions just don't do it. Instead, we must also ask this: Why did the shooter shoot? What can we learn about motive to prevent such tragic events in future?

The crime triangle is a useful and elegant baby step. I like it and I use it. But it is veneer. It is short term. It is not enough.

Our analysis of neighborhood crime must include a more robust analytical dialogue. If our analysis does not encompass action to move social life forward, it is not robust. Ultimately, if our theory fails to include motive, we are cluttering our dialogue with junk and the analysis of junk.

6 Replies so far - Add your comment

Severin Sorensen, CPP said...


I agree with you, the crime triangle on its own is a limited but useful teaching construct. However, while the crime triangle holds a necessary place in teaching crime opportunity, the crime triangle is wholly insufficient to map out the underlying conditions of the crime set.

I much prefer the theoretical rubric mapped out by Paul and Patricia Brantingham in their book, Environmental Criminology, wherein they write, that crime must be understood as the confluence or coming together of offenders, victims, targets, laws, at specific times, places, and settings. Examination of laws, nuance of time, settings, and place types are all intertwined into the calculus of crime.

For example when considering crime time, there is time of day, day of week, week of month, month of year. There are additional times of extreme heat, of moon light, of precipitation, and other natural events. Additionally there are times of a community calendar (such as a sporting event, parade, etc), and historical time (such as the anniversary of a terrorism event, fire, or violent crime). There are even times that correspond to frequency of events and the likelihood of greater numbers of targets or victims. So in this light, even the nuance of time for crime impacts crime opportunity in meaningful ways and should be examined closely to consider crime opportunity.


Severin Sorensen, CPP
Sikyur LLC

Patrick Murphy said...

In the workplace violence scenario of recent, the truly dedicated offender will not be deterred. There was a mission with no regard for risk because the ending had already been predetermined in the shooters mind. Whether he died at his own or the hands of others it mattered not. He certainly could have gone to his car if that was needed to accomplish the mission.

I like the theories contained within the Brantingham's Environmental Criminology book. The offenders learn their craft over time by formulating some sort of picture perfect scenario/environment that leads to the most success. It almost sounds like Sales 101. But this would not explain behavior or thought process of "mission-based" crimes. Columbine and the recent Times Square bomber come to mind.

They guy driving around looking for someone to rob is looking for that confluence as mentioned by Severin. Fertile crime environments like apartment properties offer 24/7 potential victims. There need not be a sophisticated study conducted by a property manager to determine violent crime is occurring on their property. Regardless of factors such as time of day, weather, etc, the courts only want to know what management knew or should have known and what they were or were not doing about it in a reasonable manner.

GSaville said...

Thanks for the comments Severin.

As usual, penetrating and thoughtful.

I will, however, respectfully disagree that the Brantingham theory offers anything truly different. More expansive perhaps, but not theoretically different.

That is because I do not believe these are motive prevention theories; they are time/place/situation theories. Nuances of the offense time, settings, and opportunity are just different ways of saying the same theoretical thing. Place is the point - not motive.

Incidentally, from my perspective when it comes to the construct of "law" in environmental criminology theory, the whole thing starts to unravel. Dysfunctional behavior needs no law to be dysfunctional (witness glue sniffing, solvent abuse, and loitering all which, in many places, are not legally prohibited by law but are very much dysfunctional).

Thus I am left with the same conclusion: just as with RA, environmental criminology is a "descriptive symbol" (in methodology terms, a non-etiological typology). It is not a causal theory. In my view the spatial pattern theory is little more than variations on the crime-and-place theme.

It STILL offers nothing whatsoever on the whole dimension on "motive".

By the way, as I say in my blog, I don't think this is a bad thing. These theories ever intended to tackle motive. That's why I think the baby steps offered by such approaches are a very useful tactic to begin the analysis - just not to finish it.

Thanks again for the insights. Great debate.

GSaville said...

Excellent points, Patrick.

I especially concur with your comments that the truly dedicated offender will not be deterred. By the way, that's precisely why I think opportunity theories fall short. We know nothing about the "why".

I think you've hit on the salient point - place based theories cannot explain behavior or thought processes of "mission-based" crimes. (It's probably not their intention anyway).

When place based theories speak of "causes", they all go back to the rational offender theories (pain vs pleasure, etc).

As I mentioned last year in my blog on "preventing crime with self interest", the Nobel prize winning work on modern economic theory by Gary Becker changed many rational choice assumptions. It is pretty exciting stuff and I think it provides us with some penetrating insights into neighborhood crime.

Ultimately I think we're going to have to read outside crime-and-place theories for a more clear understanding. Check out John McKnight's "The Careless Society". My recent favorite is James Gilligan's "Preventing Violence". Very exciting possibilities there.

Combining this newer stuff with the crime-and-place research is probably where we need to go in the 21st Century.

Thanks again Patrick.

Unknown said...

Reflecting on a 35 year policing/regulatory enforcement career and exposed to both crimes against persons and property at the community and transnational organized crime levels, I’m not convinced criminals make rational choices.

I hopefully understand Greg’s point correctly that the crime triangle is a useful analytical tool, but doesn’t go far enough. We need to dig deeper. My view is part of “digging deeper” includes recognizing and understanding innate qualities of human behaviour that surface as patterns.

My particular fascination over the past several months is research breaking down human behaviour in the social sciences in much the same way a physicist studies the quantum world, looking for the relationships between things.

I recall Greg refers to the work of Thomas Schelling reknowned for the segregation game. I understand it is referred to as one of the all time classics – that if every trace of racism where to vanish tomorrow, something akin to a law of physics might still result in races separating, much like oil and water.

Mark Buchanan (author of Ubiquity and the Social Atom) refers to his time as editor of Nature magazine and his exposure to an increasing number of research papers seeking to find mathematical irregularities in the human world of the kind known to physicists. He believes researchers are finally taking Schelling’s way of thinking seriously. Perhaps we need to look more deeply into our innate adaptive, imitating and cooperative nature identified by Buchanan and then, as crime prevention specialists, learn how to channel these characteristics positively. At least it’s a place to start.

I attempt to apply this insight to my current responsibilities, which includes countering egregious abuse and fraud in health care payment systems. This is without much success, and likely will continue to be in the short run, given the dominance of the auditor/investigator mindset of the counter-fraud consulting industry. Nonetheless, the research of behaviour economist Dan Airely convinces me of what I intuitively already knew - everyone cheats a little bit. If this is to be believed, the million dollar question becomes finding strategies that influence the behavior of providers billing health care payment systems.

GSaville said...

Interesting points John.

Resonating for me are two points.

First, if everyone cheats a bit, then the rational choice theory folks have it right and all that matters for prevention is the waxing and waning of routine activities. Instead, I tend to think some don't cheat at all and opportunities matter little. Equally, some cheat constantly and make their own opportunities. In both cases motives matter a great deal. We need to find them for both the latter and the former.

Connected to that is my second point; I agree with your view that we need to look more deeply into our innate adaptive and cooperative nature. Kennedy's speech, "seek not what your country can do for you…" came at a time when social conditions were ripe for altruistic behavior. Learning how to channel these characteristics positively within the neighborhood (that's what the new economists have done to upgrade rational choice) is EXACTLY what motive preventers need to do.

That's why I think the combination of opportunity and motive within SafeGrowth is the best way to move forward.

Thanks for those points John. They provide much to think about.