Transcript Illicit and risky behaviour (crime, drug use, anti

Risky and illicit behaviour
Steve Pudney
Institute for Social and Economic Research
Email: [email protected]
1
2
The range of issues
Possibilities include:
• Crime & anti-social behaviour - victimisation
• Crime & anti-social behaviour – behaviour
• Truancy, school exclusion, etc.
• Smoking, alcohol abuse, illicit drug use
• Diet and health impairment
• Problem gambling
• Unsafe sexual behaviour
• Foundations of behaviour: attitudes to risk and intertemporal
discounting; role of peer influences & other social
interactions; self-image and attitudes to society
• Experience/perceptions of the law enforcement system
3
Existing data sources
• Repeated cross-sections:
– Victimisation surveys, e.g. BCS
– Health surveys, e.g. HSE
– School surveys
• Longitudinal surveys:
– Cohort surveys: e.g. NCDS, BCS70, LSYPE, EYTCS
– Household panels: e.g. OCJS
– Limited coverage in BHPS
• Specialised surveys:
– Focused ‘outcome-based’ surveys, e.g. Arrestee Survey,
drug treatment records
– Medical surveys
4
Top-level issues
•
•
•
•
•
•
What are the most important research questions to be
addressed, now and potentially in the future?
Which issues should receive priority within each area
covered?
What is the case for covering these issues in the UKHLS
rather than existing cross-section or longitudinal studies or
small-scale special-purpose surveys?
Can these issues be covered without undue threat to UKHLS
response or questionnaire length?
How important is continuity of measurement relative to the
existing BHPS, and comparability with other UK national
surveys?
To what extent is cross-national comparability an important
consideration?
5
Definitional issues
• Optimal data collection frequency?
– e.g. sub-annual, annual, less frequent, occasional
• Reference period?
– e.g. actions in last month/year/ever
• How much detail?
– occurrence / frequency / quantity / expenditure
– range of consequences
• Which individuals?
– restrict questioning to specific ‘high-risk’ subpopulations?
– each adult? each child?
6
Data collection issues
• Which methods of data collection are workable?
– how to guarantee confidentiality in household survey context?
– e.g. postal self-completion versus computer-based selfcompletion (CASI or A-CASI);
– other non-traditional tools of data collection?
– automatic checks for unreliable responses?
– questions about worries over other family members’
behaviour?
• Relative pay-offs to different levels of detail
– e.g. exact amounts? grouped responses? unfolding bracket?
– e.g. degree of disaggregation over types of risky behaviour