1. Overview

A. Background and objectives

To investigate victimization rates across nation, unveil factors of crime vulnerability and figure out people’s perception and attitude toward crimes

B. History

2017: Survey V (Living Safety Survey 2016)
- Survey form and survey methods maintained based on survey III after revision
- Special topic - ‘Road Rage/Retaliation Victimization’
- Inspection and management through a professional private social survey firm
- Sample size: 6,100 households
2015: Survey IV (Living Safety Survey 2014)
- Survey form and survey methods maintained based on survey III after revision
- Inspection and management through a professional private social survey firm
- Sample size: 6,960 households
2013: Survey III (Living Safety Survey 2012)
- The 2nd additional change in survey method since there vision in 2009
- Survey was renamed
- Survey in cooperation with Statistics Agency Dept., Korea National Statistical Office (KOSTAT)
- Sample size: 6,300 households
2011: Survey II (Korean Crime Victim Survey 2010)
- The 1st additional change in method since there vision in 2009
- Survey in cooperation with Statistics Agency Dept., Korea National Statistical Office (KOSTAT)
- Sample size: 7,550 households
2009: Revised Survey I (Korean Crime Victim Survey 2008)
- Survey methodology fully revised
- Designated as ‘Approved Statistics No. 40301’
- Special topic ‘Voice Phishing / Identity Theft’
- Survey in cooperation with Statistics Agency Dept., Korea National Statistical Office (KOSTAT)
- Sample size: 4,710 households
2006: Survey V (2005 Survey on Crime Victimization in South Korea)
2003: Survey IV (2002 Survey on Crime Victimization in South Korea)
2001: (2000 International Crime Victimization Survey - South Korea)
1999: Survey III (1998 Survey on Crime Victimization in South Korea)
1997: Survey II (1996 Survey on Crime Victimization in South Korea)
1994: Survey I (Korean Crime Victimization Survey)
1991: Survey on Crime Victimization in Seoul (Pilot Survey)
1990: Survey - What is ‘Crime Victimization Survey’?
# This report intends to use the title ‘Korean Crime Victimization Survey’ which states the purposes and details of the survey.

C. Survey design

Target population: The households within the region under administrative power in the Republic of Korea at the time of the survey (May 29, 2015) and members of the family with age 14 or older.
Survey population: Among the enumeration districts (EDs) of the 2010 Population and Housing Census Results, all households and their members aged 14 or older from general enumeration district (1) and apartment enumeration district (A).
Stratification:
- Enumeration district on the 2010 Population and Housing Census: Stratified into 7 metropolitan cities and 9 provinces; 9 provinces further divided into eup, myeon, dong.
- New apartments since November 2010: Comprised of 17 strata based on city/province.
Sampling: Extraction of sample enumeration district by probability proportional to size sampling (PPS sampling); 10 households extracts through systematic sampling in the extracted enumeration district.
Weighted values: Design weight, nonresponse adjusted weight, post-stratified weight.
Enumeration size: 610 enumeration districts (10 households per district).

D. Survey cycle and duration

Duration: 2 years
Period: May 1 - Jun 16, 2017

E. Method

Data collection: Face-to-face interview or self-administered at home

F. Survey details

Types of survey form and contents: Basic survey form (for household heads / members) + incident survey form.
- Basic survey form: Household formation, community & neighboring environment, daily life & anti-crime activities, screening questions on crime victimization for the past year, 2016 special topics (road rage/retaliation victimization), and demographic information including monthly household income and education level.
- Incident survey form: Incident time & place, modus operandi, bodily damage, a victim’s response, reporting to the police, etc.

G. Survey system

Research Team of Korean Institute of Criminology : Overall survey planning for survey design, survey & management support, data analysis & report preparation
KANTAR Public: Survey and management.
Statistics Agency Dept., Survey Management Bureau in the Korea National Statistical Office (KOSTAT): Advisory and inspection on overall survey statistics processes.
Survey System Management Dept., Statistical Service Policymaker in the KOSTAT: Support on the development and utilization of NARA Statistics for standardized data input and processing.

H. Release

Release method: Report publication.
Release period: 2 years.
Title of publication: Korean Crime Victim Survey 2016.
- Downloadable on the website of the Korean Institute of Criminology (http://www.kic.re.kr)
Website of the Crime and Criminal Justice Statistics (CCJS) of the Korean Institute of Criminology (http://www.crimestats.or.kr)
- Statistical analysis through the Korean Crime Victim Survey database.
- Microdata.
KOSIS: Korean Statistical Information Service of the KOSTAT (http://kosis.kr)

2. Major Results

A. Victimization rates in 2016

1) The crimes measured in the Korean Crime Victim Survey are divided into violence crime and property crime. The former includes robbery, assault, sexual violence and harassment while the latter includes theft, fraud and property damage.
2) Among a total of 46,702,347 respondents aged 14 or older in 2017, 1,617,001 people were estimated to experience violence/property crime in 2016. The victimization rates were 3.46%, decrease from 2012 (3.98% in 2012, 3.46% in 2014). In violence crime such as robbery, assault and sexual violence, victimization rates decreased from 0.69% in 2012 to 0.33% in 2014 and 0.36% in 2016. In 2016, they slightly increased. In overall, they have declined by approximately 0.4% for the past five years. In property crimes such as theft and fraud as well, victimization rates declined from 3.44% in 2012 to 3.09% in 2014 and 3.11% in 2016. Compared to 2012, they dropped slightly in 2014. In 2016, they increased by 0.02%. However, they were still low compared to 2012.
3) According to analysis of victimization rates focusing on the number of crimes per 100,000 population, it continued to drop since 2012 (4,600 in 2012, 3,743 in 2014, 3,556 in 2016). It decreased by about 1,000 incidents per 100,000 people during the period. When analyzed by type of crime, the number of property crimes continued to drop from 3,832 in 2012 to 3,290 in 2014 and 3,168 in 2016. The number of violence crimes per 100,000 population has also decreased by about 50% for the past five years with 768, 372 and 388 in 2012, 2014 and 2016 respectively.

B. Special topic in 2016: Road rage and retaliation rates and victimization

1) The percentage of people who suffered from road rage (e.g.: Signal violation, violation of central line, speed limit, etc.) in 2016 was 8.84% while 2.07% experienced retaliatory driving such as assault, threat and property damage. In fact, 9 of 100 experienced road rage. 2 of 100 suffered from retaliatory driving. Among a total of 3,689,070 road rage incidents, 713,966 incidents were caused by retaliatory driving. Therefore, it appears that approximately 18,443 road rage and 870 retaliatory driving incidents occur per 100,000 population.
2) In terms of the characteristics of road rage and retaliatory driving, they occurred more often in women. Those who often violate traffic laws experience road rage or retaliatory driving frequently, which explains that violence on the road is related with a driving habit. Regarding the frequency of road rage and retaliation experience per victim, ‘twice or more’ was responded by 54% in road range and 48% in retaliatory driving. In terms of repeat victimization rates, retaliation is lower than road rage. However, unlike general violence crimes (e.g., robbery, assault, etc.) and property crimes (e.g., theft, fraud, etc.), a victim of the violence on the road often tends to suffer from it again. Therefore, there should be in-depth discussion on this matter.
3) In terms of a type of road rage, ‘sudden lane change driving zigzag (39.28%)’ and ‘signal violation driving over speed limit (30.81%)’ were high. Road rage mostly occurred in ‘city road’. In terms of a time zone, ‘15:00-20:00 (22.17%)’ was the highest, followed by ‘18:00-21:00 (21.83%)’. In terms of a vehicle type, ‘mid- and full-size domestic car’ was the highest. According to analysis on a vehicle type regarding road rage based on car registration as of December 2016, the percentage of road rage was relatively high in commercial and imported vehicles.
Regarding damages from road rage, bodily damage (e.g., bruise, etc.) and property damage (e.g., scratch, breakage, etc.) were about 8%.
4) In term of a type of retaliatory driving, ‘overtaking & sudden breaking (52.15%)’ and ‘sudden braking and profanity (36.75%)’ were most common. Regarding the circumstances before retaliation, ‘sudden lane change and honking/flashing the hi-beam (41.10%)’ was the highest, followed by ‘abrupt lane change without signal (15.53%)’ and ‘abrupt lane change (11.26%)’. According to analysis of the time and place of retaliation, ‘city road’ was mostly reported. In terms of the time of retaliation, ‘18:00-21:00’ was the highest with 20.72%, followed by ‘12:00-15:00 (20.10%)’.
In terms of a vehicle type in retaliation, ‘mid- and full-size domestic car’ was the highest with 46.45%, followed by ‘mid- and full-size imported car (22.55%)’, ‘commercial car (8.66%)’ and ‘domestic compact car (8.33%)’. In retaliatory driving, ‘mid- and full-size imported car’ accounted for 8.97%. In terms of a vehicle type in retaliation, however, it was as high as more than 20%. Lastly, regarding damages from retaliatory driving, light bodily damage was 3.14% while property damage (e.g., scratching, breakage, etc.) was 4.54%. In terms of bodily damage, retaliation was higher than road rage.

C. Damages and results by crime victimization type

(1) In terms of the time of violence (e.g., robbery, assault, sexual violence, etc.), ‘fall (38.80%)’ was the highest in 2016. Specially, it mostly occurred in ‘nighttime (18:00-24:00, 47.05%)’. More than 40% involved a bodily attack, which was higher than the previous year. On the contrary, the percentage of carrying a weapon at the time of violence gradually decreased since 2012 (25.14% → 7.75% → 7.68%). Meanwhile, in more than 70% of cases, there were efforts to protect the victim by himself/herself or by others around such as ‘shouting, calling the police’ and ‘asking for help’. The figure was much higher than the previous years (2012: 46.45%, 2014: 45.80%).
(2) In terms of bodily damage at the time of violence, about 80% (78.33% exact) suffered from a bodily injury in 2016. The figure was the highest since 2012 (2012: 31.04%, 2014: 64.15%). In terms of a type of bodily damage from violence in 2016, ‘bruise, scratch (83.08%)’ was the highest.
More than half of the violence was caused by ‘acquaintance’.
(3) Regarding the percentage of reporting the crime to the police in 2016, ‘reported the crime to the police’ was 31.79%, higher than the previous years (2014: 16.26%, 2016: 27.44%). Compared to 2014, the reporting rates increased. In terms of the reasons for the reporting, ‘to have the assaulter arrested and punished’ and ‘to prevent such crime from occurring again’ were most responded.
(4) Regarding the time of property crime occurrence in 2016, ‘fall’ was the highest, just like violence crime. Specifically, it frequently occurred during ‘12:00-18:00’ and ‘18:00-24:00’. In terms of the place of the
crime’s occurrence, ‘victim’s (or assaulter’s) home’ and ‘residential area or nearby street’ were responded. In terms of the percentage of property crime at a densely populated commercial area, it has gradually increased from 7.34% (2012) to 9.62% (2014) and 14.31% (2016). In contrast, the percentage of property crime in school has continuously dropped since 2012.
(5) In terms of theft victim items, ‘cash (check and gift certificate included)’ and ‘bicycle (accessories included)’ were highest in 2016. In particular, the loss of ‘bicycle (including accessories)’ has continued to increase since 2012 (17.57%→17.86%→23.71%). Meanwhile, average damage in 2016 was the lowest for the past years with KRW 566,000. The percentage of ‘finding the lost item partially or wholly’ was as low as 9%.
(6) In term of property damage in 2016, ‘vehicles damage(including components)’ has continuously increased since 2012. The repair cost or repurchase price after such damage was the highest with KRW 280,000 for the past years. In addition, the percentage of claiming the damage to the assaulter is on the rise.
(7) In terms of fraud in 2016, ‘borrowing money despite no intention or capability to pay pack’ accounted for a great portion. In fact, this case has continuously increased since 2012. The percentage of ‘unable to get a product after payment at a store (including an online shopping mall)’ or ‘getting a fake product’ has also increased. In terms of fraud schemes, ‘soliciting narration (58.32%)’ and ‘online shopping malls (19.47%)’ were mostly reported. In particular, the percentage of fraud through ‘online shopping malls’ has gradually increased since 2012. In contrast, the percentage of ‘voice phishing’ has decreased. The recovery of the damage from fraud (including partial recovery) in 2016 was higher than the previous year.
(8) In terms of difficulty or pain suffering from property crime such as theft and fraud, ‘depressed’ and ‘scared’ were high in 2016. In fact, the percentage of ‘depressed’ has been on the rise since 2012.
(9) The percentage of reporting property crime to the police was 21.53% in 2016. When asked why they didn’t report the crime to the police, ‘not much damage’ and ‘no evidence’ were mostly responded. When asked how much they were satisfied with the action taken by the police to the respondents who had reported the crime to the police, ‘neither good nor bad’ was mostly responded in 2016.

D. Factors of crime vulnerability

(1) To investigate crime vulnerability-related factors, this study analyzed differences by the features of local community, household and personal characteristics. The features of local community include relationship with neighbors, participation in neighborhood activity, physical & social disorder and police activity while household characteristics include housing type, residence type, monthly household income, security level and frequency of leaving home vacant. Regarding personal characteristics, socio-demographic characteristics such as gender and age, lifestyle and self-control variables were examined.
(2) According to analysis on differences by the features of local community, property damage was high in areas where physical disorder was serious.
In contrast, violence crime was relatively lower in the areas where participation in neighborhood activity was high, social disorder was minor, and the police was active.
(3) In terms of differences in crime damage by household characteristics, in residential burglary which appears to be very closely related with household characteristics, the crime damage was low in apartments where a security level was relatively high. On the contrary, the damage was high in non-residential buildings with a low security level such as stores, factories and inns.
(4) According to analysis on differences by crime damage by sociodemographic characteristics, property damage was higher in women (gender), 50s (age), married (marital status, compared to singles, loss of spouse by death, divorcees), graduate student or higher (education) and simple labor job (occupation). In terms of differences by socio-demographic characteristics by dividing property crime into personal larceny and fraud, the results were pretty same with the overall property crime damage. The personal larceny was relatively higher with people who lost their spouse by death or divorce than singles and married. In addition, the damage from fraud was higher in 30s (age). In contrast, violence crime had no significant relationship with any socio-demographic variables.
(5) According to analysis on the effects of lifestyles and self-control level on different types of crime victimization, the percentage of property crime victimization was higher in those who go out in luxury costume or with a lot of accessories or those who usually put on luxury brand clothes. In addition, property crime victimization was lower in those with high self-control level. In terms of lifestyles, violent crime victimization showed a statistically significant relationship with ‘returning home late’ only. Therefore, the crime victimization was more often in those who returned home late. As stated above, factors of crime vulnerability revealed a different pattern across types of crime victimization.
(6) In terms of annual trends in factors of crime vulnerability, neighbor relationship which represents bonding and solidarity with neighbors, participation in issues in neighborhood which measures a level of participation for solving the problems of a local community and police activity which represents the efficiency of crime control have mostly increased over time. The percentage of responses on community physical and social disorder has decreased. A security level has mostly increased. The percentage of heaving their home empty has declined. In terms of personal characteristics, the frequency of returning home late or drunken and luxurious outfit have increased over time. The percentage of using public transportation has decreased and started to rebound. In other words, community and household factors of crime vulnerability have mostly improved over time, while crime vulnerability by personal characteristics have increased over time.

E. Awareness and fear of crime

(1) Regarding awareness of crime, people tend to believe that an increase in crime in the future would be more outstanding in nationwide than in their community. According to differences in crime awareness by direct and indirect crime victimization experience, the percentage of believing that crime would increase was higher in those who experienced crime victimization than in those who didn’t regardless of a type of crime victimization experience (direct or indirect experience) and scope of crime (nationwide, community they live in).
(2) Regarding general fear, respondents seemed more scared in ‘walking down a narrow alleyway alone at night’ than in ‘being alone at home at night’. In terms of fear by a crime target, they were most afraid of ‘crime against their child’, followed by ‘spouse (lover)’ and ‘themselves’.
In terms of fear by a crime type, ‘residential burglary’ was the highest, followed by ‘assault’, ‘fraud’ and ‘sexual violence’ ranking the fourth among 8 types of crime.
(3) According to analysis on differences in fear of crime by personal characteristics, city size, housing type and exposure to the media, all types of fear of crimes revealed the similar patterns in general. In other words, fear was greater in women (gender), 10s and 20s (age), single (marital status), elementary school graduates (educational background), housewife/student (occupation), high average monthly household income, experience of crime victimization(including indirect victimization) and serious physical/social disorder. In case of participation in neighborhood activity, as the participation increased, general fear of crime diminished, but specific fear and fear of property/violence crime increased. As police activity and self-control level increased, all types of crime related fear crimes declined.
(4) When anti-crime activities were divided into passive and active crime-preventive ones, people were more engaged in passive activities (locking up, being with someone when walking at night, not taking a taxi alone, etc.) than in active ones (carrying a self-protection device, participation in local community’s anti-crime activities, etc.).
(5) According to differences in anti-crime activities by personal characteristics, experience of crime victimization, community size and housing type, gender, educational background, occupation, direct crime victimization experience and housing type had effects on crimepreventive activities in a mostly same manner regardless of the type of anti-crime activity (direct/indirect anti-crime activity). However, the effects of age, marital status, average monthly household income and city size on anti-crime activity differed across the type.
(6) According to trends of crime awareness, fear of crime and anti-crime activity over time, the percentage of believing that national crime incidence rates and crime occurrence in their community mostly dropped over time. This year, general fear, fear by crime target and fear by crime type were the lowest. Meanwhile, passive anti-crime activities declined over time. Active crime-preventive activities significantly increased in 2015 and decreased again this year, almost reaching the level of 2013.

3. Suggestions for the directions of the Korean Crime Victim Survey

This study summarized comprehensive assessment on this year’s survey processes and major issues which should be considered for a future revision of the crime victim survey through literature review on crime victim surveys in foreign countries. The results are as follows:
First, regarding a survey target, the members of the households aged 14 or older are investigated. For this, there should be a review on the appropriateness of the age limit and a separate victim survey on young people. In terms of analysis on current victimization rate and factors of crime vulnerability, there should be debate on the problems which could occur when using the data obtained by investigating all members of a household aged 14 and older. Considering the fact that it is not easy to solve a budget problem, in addition, this study proposed a necessity of applying a strategic survey method which could get maximum effects with the minimum samples, instead of expanding a sample size.
Second, this study suggested a necessity to review crime victimization types and logics used for crime classification. The crime victimization types which have been measured in current crime victim surveys have faced a limitation. Therefore, it is needed to measure diverse crime victim patterns which occur these days. For this, there is a necessity to expand the range of victimization experience. Furthermore, there should be a discussion on the logics used to classify victimization according to hierarchical rules of crime severity and measurement of series victimization. Furthermore, overall review on specific survey items and improvement of the survey form are required.
Third, revising survey methods should be discussed. In fact, it’s been continuously pointed out during the crime victim survey. It is urgent to develop a data collection method, considering the efficiency of survey response and characteristics of the survey. In a mixed survey in which more than two different surveys are adopted, in addition, the effects of survey methods on survey results can differ. Therefore, ‘para data’ should be actively utilized. To enhance the usability of survey materials, it is necessary to secure appropriateness and reliability on data collection through adding more theoretical variables.
This study investigated major crime victimization as well as damages from road rage and retaliatory driving which have recently emerged as a major social problem from a very comprehensive perspective. In other words, it examined specific crime victimization such as violent and property crimes, relationship between assaulters and victims, factors of crime vulnerability, fear of crime and correlates of them. This study also summarized issued to be discussed for a revision of the survey in the future based on the results of comprehensive assessment on survey process and crime victimization surveys in foreign countries. The Korean Crime Victim Survey started in 1994, and was fully revised in survey design and methods in 2009 and minor changes thereafter to secure validity and reliability of the data. There should be continued efforts in order for the Korean Crime Victim Survey to provide more accurate and scientific statistical data.