Heterogeneity and Data Analysis: Heterogeneity #2, Mixture of types (cont.)

Heterogeneity #2, Mixture of types

Evaluations of Closed Circuit television (CCTV) as described by Tilley (2000) might be subject to a meta-analysis.  However, as Tilley’s lists below indicate, such an analysis would mix together studies of situations in which different mechanisms (or a mix of mechanisms) and different contexts apply.  What meaningful recommendation could emerge form the meta-analysis, even if all results were in the same direction?

I was asked by officials at the Home Office to look at the effectiveness of the introduction of close circuit television in car parks as part of the Safer Cities Programme which was aiming to deal with local crime problems in 20 cities in England…  First I thought about mechanisms. How might close circuit television affect rates of car crime?

Here is a list of mechanisms:

a) The ‘caught in the act’ mechanism. CCTV might reduce car crime by increasing the chances that current offenders are seen on screen detected committing their crimes and arrested, taken away, punished and deterred.

b) The ‘you’ve been framed’ mechanism. CCTV might reduce car crime by leading potential offenders to avoid the perceived risk that they might be caught and convicted because of the evidence on tape.

c) The ‘nosy parker’ mechanism. CCTV might lead to increased usage of car parks since drivers feel more safe. Their increased usage might then increase natural surveillance deterring potential offenders worried that they might be seen committing their crimes.

d) The ‘effective deployment’ mechanism. CCTV might enable security staff to be deployed more quickly where suspicious behaviour was going on. They then act as visible guardians.

e) The ‘publicity’ mechanism. CCTV and signs announcing its installation might symbolise efforts to take crime seriously and to reduce it. Potential offenders might want to avoid the perceived increased risk.

f) The ‘time for crime’ mechanism. Offenders might calculate that car crimes taking a long time risk their being caught on camera and they might decide only to commit those car crimes that could be completed very quickly.

g) The ‘memory jogging’ mechanism. The presence of CCTV and associated notices may remind drivers that their cars are vulnerable and lead them to lock them and operate security devices and remove easily stolen items from view.

h) The ‘appeal to the cautious’ mechanism. Cautious drivers sensitive to the possibility that their cars may be vulnerable to crime may use car parks with more security devices and displace less cautious drivers to other car parks. The high level of security of the car park users may make it difficult for offenders successfully to commit their crimes.

Having thought about mechanisms I then thought about context. Are all car parks and all car park crime problems the same? Well, here are some of the variations that I identified.

1. The ‘criminal clustering’ context. A given rate of car crime may result from a small number of very active offenders or a large number of occasional offenders. A mechanism leading to the offender being disabled holds promise according to the offender/offence ratio as in (a) above.

2. The ‘style of usage’ context. Long stay car parks fill up early in the morning and empty after work in the evening. If the dominant CCTV mechanism turns out to be increased confidence and usage, as in (c) or (h) above, then this will have little impact because the pattern of usage is already high, with little movement dictated by working hours not fear of crime. If, however, the car park is little used, but has a very high per user car crime rate, then increased usage mechanisms may lead to an overall increase in the number of crimes but a decreased rate per use.

3. The ‘lie of the land’ context. Cars parked in CCTV blind spots will be more vulnerable if the mechanism is increased chances of apprehension through evidence on video tape as in (b), but not if it is through changed attributes or security behaviour of customers, as in (g) or (h).

4. The ‘alternative targets’ context. The local patterns of motivation of offenders, together with the availability of substitute targets, provide the context for potential displacement elsewhere.

5. The ‘resources’ context. In isolated car parks with no security presence and no police near to hand the deployment of security staff or police as a deterrent as in (d) is not possible.

This is not, of course, necessarily a comprehensive list of contexts or mechanisms. What it brings out, though, is that even in relation to a relatively simple measure in a relatively simple setting the range of mechanisms and contexts is quite wide. It is unlikely that closed circuit television will have the same effect on car crime rates in all circumstances. The mechanisms and contexts are just too varied. Added to this, of course, CCTV itself varies substantially in its technical capacity, which will affect its potential to trigger some of the mechanisms which have been identified here. The issue for the evaluator is that of working out how to test, or arbitrate between, a variety of theories that explain how and where CCTV might have its impact on car crime.

(continuing a series of posts—see first post; see next post)

One thought on “Heterogeneity and Data Analysis: Heterogeneity #2, Mixture of types (cont.)

  1. Pingback: Heterogeneity and Data Analysis: Heterogeneity #2, Mixture of types « Intersecting Processes

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s