siena rand

The National Crime
Victimization Survey:
32 Years of measuring crime in
the United States
Michael Rand
Bureau of Justice Statistics (BJS)
www.ojp.usdoj.gov/bjs
National Crime
Victimization Survey (NCVS)
n
Initiated in 1972
(as the National Crime Survey)
n
Purposes:
– Measure “dark figure of unreported crime”
– Obtain information on characteristics of
crime victims and crime events
– Provide estimates of year to year change
Impetus for NCVS - 1967 Report
“ If we knew more about the character of both
offenders and victims, the nature of their
relationships and the circumstances that create a
high probability of crime conduct, it seems likely
that crime prevention and control programs could
be made much more effective.”
The President’s Commission on Law Enforcement and
the Administration of Justice. Task Force Report: Crime
and Its Impact: An Assessment, 1967
Crimes measured by NCVS
n
Rape
n
Sexual Assault
n
Robbery
n
Aggravated Assault
n
Simple Assault
n
Pocket picking/purse snatching
n
Burglary
n
Motor vehicle theft
n
Theft
Core components of NCVS
methodology
n
n
n
n
n
n
stratified multi stage clustered address
sample
interview respondents 7 times over 3 years
6 month reference period
bounding interview
interview every household member for them
self
two part interview: screener and incident
report
Types of changes to the survey
n
n
n
n
Methodological (changes in survey
design)
Adding/changing questions to obtain
more information about victims and
crimes
Expand the scope of offenses
measured
Cost reducing
NCVS Redesign
n
Decade-long research/implementation
(1979-92)
n
Improved detection of hard-to-measure
crimes:
– Rape and Sexual assault
– Domestic violence
n
New cues to help victims recall events not
reported to the police
n
Change in “series” protocol
n
Change in survey name (NCS to NCVS to
reflect redesign
Violent crime rates, 1973-2003
& NCVS/NCS ratio
Rate per 1,000 persons 12+
60
NCS
50
NCVS
Violent crime (1.49)
40
30
Simple assault (1.75)
20
10
0
1973
Aggravated assault (1.23)
Robbery (1.03)
Rape (2.57)
1977
1981
1985
1989
1993
1997
2001
Recent and upcoming changes
n
n
n
n
Crime against people with disabilities
Identity theft
Stalking
Survey automation
Impact of the redesign
n
n
Higher victimization rates for most
crimes measured
Measurement of sexual assault (other
than rape or attempted rape)
The NCVS has taught us much of
what we know about crime in the
United States
NCVS Sample
n
Nationally representative stratified multistage
sample drawn from Decennial Census
n
Household (address)-based survey
n
One of the largest ongoing gov’t surveys
n
sample interviewed every 6 months
– 42,000 households
– 76,050 people
NCVS crime coverage
n
n
Measures crimes both reported and
not reported to police
Considered “Omnibus” Crime Survey:
– Measures crimes of violence and
theft
Data Collection
n
n
n
n
n
U.S. Census Bureau Field Representatives
(FRs) conduct all interviews
FR’s: 76% female; 24% male
FRs report to 12 regional field offices
throughout the U.S.
First interview is conducted in person;
subsequent interviews are attempted by
telephone
Currently both paper and pencil and CATI
NCVS Interview
n
Two part interview
– Crime screener
n Screens
months
for any crimes during previous 6
– Incident report
n Details
of incident: what, when, how, by
whom, consequences, etc
NCVS Procedures
n
n
n
n
n
Households interviewed 7 times at 6
month intervals (3 years)
First interview used to “bound” the next
interview
All household members age 12+
interviewed by self-response
First “screen” for possible crimes, then
obtain details
“Series” protocol for frequent or ongoing
victimization
NCVS Crime screening
n
n
7-9 multi-part questions determine
whether person (or household) was
victim of crime during previous 6
months
Questions focus on different types or
aspects of crime to help elicit response
Response rate
n
2002 response rates
– 92% households
– 87% individuals
n
Ranks among highest response rates of all
Census surveys
n
Senior field representatives are sent to
households that refuse in-person
interviews or in CATI
Quality Assurance
n
Regional office quality edit (with feedback to
interviewers)
n
Independent clerical edit and coding
n
Verification of data keying
n
Computer consistency edits
n
Statistical weighting to improve estimates
NCVS Summary:
Strengths
n
Strengths
– Large sample
– Ability to aggregate data over years
– Improved methodology since redesign
– Consistent methodology enables trend
analysis
– Unreported crime
– Detailed victim, offender and incident data
– Platform for supplemental studies
NCVS Summary:
Limitations
n
Limitations
– Administered to persons age 12+
– Typically cannot include very sensitive
questions
– No local data; cannot be disaggregated
beyond 4 regions
– “Omnibus” survey means cannot focus on
specific types of victimization
– Survey procedures may inhibit reporting of
some victimizations
Supplements to the NCVS
n
Conducted for 6 months
Limited to about 10 minutes
Future supplement
n
Previous supplements
n
n
– Stalking (January 2006)
n School Crime
n Police and Public Contact
n Workplace Risk
Annual Bulletin
n
Latest rate estimates
n
Year-to-year change
n
Crime trends
n
Basic characteristics
Special Reports
n
Topical
n
Detailed analyses
n
Aggregate over years
§
Victimization of College
Students
Weapon use in Violent
Crime
§
Serious violent crime levels declined since
1993.