International Demands on Crime Statistics

1
International Demands on Crime Statistics
Geneva, November 3, 2004
Wolfgang RhombergUNODC, Research and Analysis
Section
2

• Not everything that can be counted, counts. And
  not everything that counts can be counted.
• Albert Einstein

3
ECOSOC res. 1984/48 25 May 1984

• Crime related database
• Surveys on Crime Trends and Operations of
  Criminal Justice Systems
• Collect data on the incidence of reported crime
• Collect data on the operations of criminal
  justice systems
• Improve analysis and dissemination of information
• Informed decision making in administration,
  nationally and internationally

4
Systems thinking approach

• Conceptual models
• Every model needs a purpose
• Modelling implies
• Omission
• Distortion
• Generalization
• Who are we doing it for?

5
Different conceptual models lead to different
results

• ICVS (victims)
• WCTS (police)
• - (media)
• - (insurance companies)

6

• 8th UNWCTS
• 4 Parts Police, Prosecution, Courts, Prisons
• 19 Tables 4 3 4 8
• 164 Variables/Year 48 30 36 50
• Web http//www.unodc.org/unodc/crime_cicp_surveys
  .html
• Available for download in 6 UN languages

7
8th UNWCTS Table 2 of the questionnaire(Crimes
recorded in criminal (police) statistics, by type
of crime including attempts to commit crimes,
2001 2002)lists the following 18 types of
offences 2.1 Total recorded crimes, regardless
of typeIntentional homicide2.2 Committed
2.3 Attempted and 2.4 Committed with a firearm
2.5 Non-intentional homicideAssault 2.6 Major
assault 2.7 Total assault2.8 Rape 2.9
RobberyTheft 2.10 Major theft 2.11 Total
theft 2.12 Automobile theft2.13 Burglary
2.14 Fraud 2.15 Embezzlement2.16 Drug-related
crime 2.17 Bribery and/or corruption2.18
Kidnapping
8
Policy impact

• Criminals ??
• Cost of crime ??
• Proceeds of crime ??
• Attrition rates ??
• CJS efficiency ??
• Crime infection rates ??
• Security and quality of life ??

9
Meta statistics

• On reporting
• Who? (and who not)
• What?
• When? (input vs. output)
• How?
• For whom?
• For what purpose?
• Feedback?

• On statistics, how
• accurate?
• complete?
• timely?
• reliable?
• consistent?
• comparable?

10
Who reports?

• Federal State
• Different Police Organizations
• Reporting chains, consolidation
• Statistical institutes - Ministries

11
Percentage of homicides reported by the World
Health Organization as compared to those reported
by the United Nations surveys of crime trends
and operations of criminal justice systems,
1986-2000
(United Nations surveys 100 per cent) Source
United Nations Office on Drugs and Crime.
Published in Forum on Crime and Society
12
What is being reported?

• Example given significant drug seizure
• What is significant?
• Which drug?
• Purity?

13
Statistic

• Is it relevant?
• Is the answer available?
• Is there a clear unit of measurement?
• Semantics?
• Less is more
• Small is beautiful
• Reductio ad absurdum

14
When and How

• Input statistics vs output statistics
• This years convictions vs last years arrests
• Media of reporting (double count)

15
Dynamic crime models
Offensive Eliminate the Causes
Defensive Minimize Likelihood of Significant
Impact
decrease
increase
Disagreement escalates to Violence
Anti Corruption Initiative
16
UN Crime Trends Survey
Crimes
Suspects
Prisoners
Personsconvicted
Counts/year
Counts/year
Counts/year
Counts on given day
17
Delays
Seemingly stabilized system becomes unstable
after a long time
18
Crime Conventions

• United Nations Convention against Corruption
• United Nations Convention against Transnational
  Organized Crime
• Protocol to Prevent, Suppress and Punish
  Trafficking in Persons
• Protocol against the Smuggling of Migrants by
  Land, Air and Sea
• Protocol against the Illicit Manufacturing of and
  Trafficking in Firearms, Their Parts and
  Components and Ammunition

19
Wish list

• Please consider
• Objectives, conceptual models
• Meta statistics
• Keep to essentials
• Reporting chains
• Alternative solutions
• Individual statistics
• Interpretation of data
• Crime conventions

20

• Thank you !

21
(No Transcript)
22
(No Transcript)