Research Purpose and Method

Research Purpose
- Concentrated poverty neighborhoods have received political attention as a major factor in the crime due to theories of environmental criminology and social disorganization. While there is standardized research about this topic overseas, there is nearly no domestic research on this topic.
- This research aims to connect the missing links within domestic academia and to empirically analyze how just how serious crime is in concentrated poverty neighborhoods. The significance of this research is even greater given the government’s active push of urban regeneration projects.

Research Method
- The most controversial issue surrounding the research process is defining the term ‘concentrated poverty neighborhood.’ In the United States and Canada, an area densely populated by the poor (high-poverty area or a concentrated poverty neighborhood) is defined as, ‘a region where more than 40% of the population has an income below the government-defined poor threshold.’
- In contrast to the United States, a lack of official statistics means that proxy variables must be used. As such, the definition of a concentrated poverty neighborhood for Korea is based on the ‘ratio of basic assistance recipient households per town/township/neighborhood.’ However, different criteria are applied in each chapter, taking into account the research target and the purpose and nature of the data. For example, the nation-wide analysis in Chapter 3 uses 15% while the Seoul-only analysis uses 10%; other criteria such as income criteria (Chapter 5 and Chapter 7), housing type (Chapter 8), density (Chapter 13) and the like are defined and analyzed differently.
In general, as only city/county/district level data exists for the ratio of basic assistance recipient households, town/township/neighborhood level data was secured through the Division of Basic Livelihood Security of the Ministry of Health and Welfare.
- With cooperation from the Police Department, this study analyzed crime statistics at the town/township/neighborhood level and conducted research employing various methodologies such as GIS analysis, Seoul and case study area surveys, on-site investigations, in-depth interviews, and a legal system analysis with researchers from diverse fields such as criminology, police, architecture, social welfare, government, and law.

Current status of concentrated poverty neighborhoods and basic research trends based on basic assistance recipient household statistics

National Status
- The average ratio of basic assistance recipient households and individual recipients for town/township/neighborhoods are 7.20% and 4.32%, respectively, with median values of 6.86% and 4.03%, respectively. The recipient household ratio being higher than the individual recipient ratio demonstrates that the ratio of 1-member family households is higher among households receiving assistance than non-recipient households.
- For recipient households, the maximum value is 44.22% at OO-dong, Sasang-gu, Busan and the minimum value is 0.05% at O-myeon, Gapyeong-gun, Gyeonggi-do. For number of recipients, the maximum value is 31.59% at OO-dong, and the minimum value is 0.02% at OO-dong, Songpa-gu, Seoul.

Current metropolitan status
- According to the average number of recipient household and number of recipients in the 17 metropolitan municipalities nationwide, Jeonbuk(1st, 10.79% and 6.84%), Jeonnam(2nd, 9.70% and 6.16%), Gyeongbuk(3rd, 8.59% and 5.16%), Busan(4th, 8.54% and 5.11%), Daegu(5th, 8.47% and 5.22%) exceed the nationwide average of 7.20% and 4.32%.
- In general, the average value among the provinces is higher than the figures for the metropolitan cities. The former has a value of 7.97% and the latter 6.78%; in terms of number of recipients, the level of poverty in the provinces is higher than in the special/metropolitan cities.

Types of concentrated poverty neighborhoods
- Typically classified into 4 types. There are traditional city-type dense areas such as slums, multi-family areas, and multi-household areas, in addition to worn-out single-house areas in rural regions, permanent rental apartment areas provided since 1989 during President Roh Tae Woo’s administration, and dosshouse areas occupied by extremely poor single-member households.
- According to previous research, rural areas do not have serious crime issues if crop theft and elderly-targeted fraud is excluded, and neither do permanent rental apartment areas despite issues such as juvenile delinquency and the stigma of such areas being raised in the past. Livelihood and housing security issues are the primary concern in dosshouse areas, making safety issues a secondary concern.

Previous research trends on the relationship between poverty and crime
- Overseas studies: A review of research on the relationship between poverty and crime published between 1980 and 2012 show that relatively older studies report poverty having little impact on crime due to a lack of a clear affect that poverty has on crime, while an increasing number of recent studies suggest that poverty is not only a major factor that gives rise to crime, but also significantly impacts the damage done from crime. In particular, the impact of the geographical concentration of poverty on crime and damages has been reported by both American and British research.
- Indirect studies exist in Korea regarding the relationship between concentrated poverty neighborhood and crime. Some research applied a quantitative analysis using income and other variables as independent variables and the various types of crime as dependent variables, while others uated the effectiveness of crime prevention measures in concentrated poverty neighborhood by applying the CPTED crime prevention method after making an uation of the risk of crime in order to assess its effectiveness.

Analysis of the characteristics of concentrated poverty neighborhoods through official crime statistics

State of crime in concentrated poverty neighborhoods
- With the exception of murder, the average number of crimes in the five major crime categories of robbery, rape, sexual assault, theft, and violence is actually less in concentrated poverty neighborhoods.
- In terms of population-based rates of crime, crime in concentrated poverty neighborhood is notably higher in all five categories.

Crime rate by concentrated poverty neighborhood
- After dividing the basic assistance recipient household ratios in intervals of 5% to analyze rates of crime, there was a significant difference between the five major crimes except for theft, and generally, there was a high rate of crime in the 15~20 and 20~25% intervals.
- Crime rate was lower in all areas in concentrated poverty neighborhoods with a 20-25% ratio of basic assistance recipient households.

Poverty concentration levels and population, social, and residential environment characteristics
- There was no significant correlation between the ratio of basic assistance recipient households and crime rate, except for murder, and even the correlation with murder is low, confirming that there is no linear relationship between the level of poverty concentration and crime rates.
- There was a difference in population, social, and environmental characteristics in different recipient household ratio intervals, but environmental characteristics known to be highly related to crime were not distributed the same level in accordance with recipient household ratios.

Crime rate considering environmental characteristics
- After correcting for the effect of population, social, and residential environment variables, the types of crime with significant differences by poverty concentration level were robbery and violence, with no significant differences resulting from murder, rape, sexual assault, and theft, demonstrating that, while the poverty concentration level has an effect on the occurrence of crime that differs depending on the type of crime, this is not a large effect.
- The effect of environmental factors differed by the type of crime, but the variables having a significant impact on all crime types were social variables such as the ratio of 1-member households and the ratio of individuals with criminal records, in addition to residential environment variables such as the ratio of lodging businesses and the ratio of small houses, confirming that it is necessary to intervene in terms of environmental factors in order to reduce crime in concentrated poverty neighborhood.

Analysis of the geographical characteristics of concentrated poverty neighborhoods using GIS

Significance and methodology
- Analysis of the geographical heterogeneity of poverty’s impact on crime using a geographically weighted regression model, which has been rarely used in previous studies on the link between crime and poverty.
- The poverty index used is an absolute poverty indicator that counts the number of basic assistance recipients per 1,000 people, with the six variables used for the poor resident separation indicators being: index of dissimilarity (D), weighted index of dissimilarity (Dw), exposure (interaction index: xPy; isolation index: xPx), spatial proximity index (SP), relative concentration index (RCO), and absolute centralization index (ACE).
- The number of crimes (total number of incidents of the five major crimes) per 100,000 people between 2010-2014, estimated based on the town/township/neighborhood population for the local police jurisdictions nationwide, was used as the dependent variable, and land price, population density, early divorce rate, young and elderly support cost, pub concentration, population transfer ratio, revenue, financial self-reliance, and non-residential floor space ratio as extracted from various government statistical materials were used as control variables.

Analysis results
- In terms of basic statistics for the main variables (as of 2014), the crime rate for 230 city/county/districts: 352-4595 per 100,000 people, the assistance recipient ratio for 230 city/county/district: 5.33-85 per 1000 people, the heterogeneity and weighted index of dissimilarity is between 0 and 0.5, and the interaction index and geographical proximity index are relatively small.
- In terms of the geographical distribution of major variables (as of 2013), the assistance recipient ratio is concentrated in Gyeongbuk, Jeonnam, and Jeonbuk, with the poverty interaction index being concentrated in the capital area and Gyeongsang. The crime rate shows a geographical pattern similar to the index of dissimilarity in terms of temporal changes in the geographical distribution of major variables, most major variables remain constant between 2010-14. Geographical statistical models are more suitable than geographical panel models.
- Binary variance analysis between the crime rate and poverty index/resident segregation index per city/county/district size shows that there is a linear relationship between crime rate and index of dissimilarity rather than between the crime rate and assistance recipient ratio.
- Non-geographical statistical analysis results show that there is a statistically significant correlation between the crime rate and the index of dissimilarity, geographical proximity index, relative concentration index, and absolute centralization index, although the assistance recipient ratio is not statistically significant.
- Geographical statistical analysis model results show that, in consideration of the impact of poverty variables on crime rates according to the geographically weighted regression model that takes into account the geographical heterogeneity of coefficients, there a significant relationship only in the southern regions of Jeonnam, Jeonbuk, Gyeongnam, and Gyeongbuk in terms of the basic assistance recipient ratio. For index of dissimilarity, a statistically significant regression coefficient is deduced on a national level.

Analysis of characteristics of concentrated poverty neighborhoods in terms of the results of crime damage investigations

Significance
- Using national crime damage investigation data, the characteristics of poor households’ surroundings were analyzed to examine how poor and non-poor families differ regarding the fear of suffering damage from crime and crime prediction.
- In consideration of the possibility that the regional characteristics of metropolitan areas or provinces are being reflected addition to the effects of poverty, check these effects by applying the relevant models.

Analysis results
- Some characteristics of poor households are a low level of education for the head of the household, a high proportion of single-member households in terms of the makeup of the household, a high proportion of unemployment as compared to a high proportion of white-collar, service, and sales workers in non-poor households, and twice the length of an average residency period for non-poor households.
- Among physical disorder, social disorder, relationship between neighbors, and police reliability as characteristics of the surrounding neighborhood, physical disorder and social disorder levels were higher and the relationship between neighbors was better for poor households. In terms of lifestyle, the frequency of use of public transportation and alcohol intoxication were actually lower in poor households, with poor households being more prepared in terms of crime prevention activities than non-poor households.
Poor households in the metropolitan areas were more active in crime prevention compared to those in the provinces.
- Non-poor households exhibited higher levels of fear with regard to the damage done from crime as compared to non-poor households, and this also applied to crime prediction, despite regional differences. As for the types of damage done by crime, home invasions and rates of assault were higher in poor-households, whereas fraud and vandalism rates were higher in non-poor households.
- When broadly dividing up the damage done by crime into bodily damage and property damage, bodily damage was more prent in the provinces and among the younger population and bachelors, whereas property damage was more prent among non-poor households, women, and married people.
- In conclusion, the effects of poverty were not much greater than expected when analyzing the data on crime damage investigations, although different effects are shown when combined with local variables. Poor households are not general or uniform, and different properties and characteristics are exhibited depending on whether the area in question is a metropolitan area or a province.

Analysis of the subjective perception of crime and safety policies

Significance
- Household visit surveys of 1300 households in 14 traditional densely populated neighborhoods, 8 permanent rental apartment neighborhoods, and 4 dosshouse neighborhoods.
- Comparison between low crime rate areas (low density, low-danger area) versus concentrated poverty neighborhoods classified into high crime rate areas (high density, high-danger area) and low crime rate areas (high density, low-danger area), on the 26 crimes classified under the official statistics on crime rate.

Analysis Results
- Low crime rate areas in concentrated poverty neighborhoods ranks in the middle between high crime rate areas in concentrated poverty neighborhoods and low crime rate areas in non-concentrated poverty neighborhoods.
Residents in low crime rate areas in concentrated poverty neighborhoods show the highest levels of dissatisfaction with society and the fear of punishment for crimes.
- In terms of how people perceive their own neighborhoods, high crime rate areas in concentrated poverty neighborhoods exhibited the lowest levels of neighborhood satisfaction, social capital, efficacy of collective action, interaction with neighbors, and interaction with nearby people, followed by low crime rate areas in concentrated poverty neighborhoods and non-concentrated poverty neighborhoods, in that order.
- For the past year, the ratio of personal experience of crime was 12.0% for high-dense/high-risk areas, 8.2% for high-dense/low-risk area, and 7.7% for low-dense/low-risk areas, with the high-dense/high-risk area having the highest figures compared to other areas. The ratio of experience with damages from assault and intimidation showed no difference, while the ratio of experience with damage from house vandalism and house invasion was rather high. In contrast, the ratio of experience with damage from physical contact was higher in low-dense/low-risk areas. Even in terms of damage ratios over an entire lifetime, having had no relevant experiences in low-dense/low-risk areas were the highest at 84.0%, followed by high-dense/low-risk areas at 76.7% and high-dense/high-risk areas at 73.3%.
- Checking whether individuals have experience perpetrating crime throughout their lifetimes, 95% have no experience perpetrating crimes. By region, only 7.1% in high-dense/high-risk areas, 4.4% in high-dense/low-risk areas, and 3.1% in high-dense/low-risk areas responded that they have experience with perpetrating crime.
- Notwithstanding the relationship between concentrated poverty neighborhoods and crime rates, there was no difference in perception of the country as a whole, although residents in concentrated poverty neighborhoods felt less secure with regard to the stability of their neighborhood and the areas nearby. A general fear of crime was higher among residents in low-crime areas in concentrated poverty neighborhoods rather than in high-crime areas in concentrated poverty neighborhoods.
However, the result of studying how safe one felt from the six types of crime, including pickpocketing/snatching/robbery, assault/intimidation, house vandalism, house invasion, sexual harassment, and sexual assault showed that respondents from concentrated poverty neighborhoods gave lower scores to local security criteria as compared to non-residents, with the lowest uation coming from residents from high crime rate areas.
- Regarding the adequacy of security equipment installed in their current residential neighborhood, more residents replied 'appropriate' for CCTVs and streetlamps while more answered ‘not appropriate’ than ‘appropriate’ for emergency bells regardless of the crime rate in concentrated poverty neighborhoods. On crime prevention measures, demand for CCTV installations was higher in high crime rate areas in concentrated poverty neighborhoods with few apartment residents, whereas demand for streetlamps was higher in all other regions.
- With regard to crime prevention related safety policies, there was a big difference depending on the awareness of the policy in question. Generally, residents in low crime rate areas in concentrated poverty neighborhoods had a relatively higher perception of crime prevention related safety policies as compared to those in other areas. In particular, awareness of subway safety policies exceeded 50% among residents in low crime rate areas in concentrated poverty neighborhoods while rating less than 25% among the resident in the other two regions.
- Overall, the respondents had the most experience with using secure courier services at 3.1%, followed by autonomous security services and reliable return-home services. Regarding the most necessary measures for security for concentrated poverty neighborhoods, 'CCTV installation in public places' was the highest, followed by 'maintaining streetlamp brightness' and ‘maintaining clean streets’, and, after an analysis, this order was found to be the same for all other regions.

Multilevel analysis of the relationship between poverty and crime

Significance
- Analysis on the relationship between poverty and crime using the same results of the survey of the 26 districts in Seoul as set forth above. For the individual level analysis, variables such as economic tension (stress, objective poverty), relative economic level compared to others (relative poverty), perception of punishment, and perception of law were used; for the social level analysis, variables such as family ties, disorder, and collective efficacy were used.
- Used the concept of social network and official social control to analyze the discussions on individual-social interaction.

Analysis results
- Poverty (objective standard of poverty being monthly income, relative standard being the level of poverty felt by an individual when comparing themselves to others) is not directly related, or is low in relation to the damage done by crime. Meanwhile, poverty (objective poverty) increases the probability of crime perpetration by about 1.2 times. Poverty is an individual unit variable, although it can indirectly affect the damages or harms of crime through variables related to social/environmental variables or individual-social interactions.
- Among the individual unit variables, economic tension (stress) is a significant factor that increases the probability of crime perpetration by 1.7-1.8 times.
Relative deprivation is a factor that increases the probability of crime perpetration by 1.2. However, including the individual-social interaction variable eliminates this relationship. Perception of law under individual unit variables and efficacy of collective action in terms of social unit variables are closely related to the possibility of crime perpetration.
- Among social unit variables, family ties reduce the possibility of damage done by crime by 1.1 times. The formation of family ties increased the perception of law and affected the efficacy of collective action, indirectly affecting the possibility of crime perpetration. Disorder level is not related to the likelihood of crime perpetration, but increases the likelihood of damage done by crime.
- Among individual-social interaction variables, social networks are related to the likelihood of damage done by crime, but a longitudinal analysis study to determine the temporal priority in terms of causality should be conducted in the future. Although the formation of social networks in poor areas could increase the efficacy of collective action, there is also the possibility of increased damage or harm due to more opportunities for damage or harm done by crime as a result of increased exchanges.

Analysis of criminal behavior in concentrated poverty neighborhoods

Significance
- The purpose of this study is to investigate the residency status of criminals with electronic tagging in concentrated poverty neighborhoods (dosshouse, hostel, residence-denial status, etc.), and to examine the differences between general criminals and the likelihood of recidivism from those living in concentrated poverty neighborhoods.
- Collected and analyzed data on 6,544 total criminals with electronic tagging on that date (as of October 2017) with the help of the Crime Prevention Policy Bureau of the Ministry of Justice on October 2017.

Analysis results
- Of the 6,544 total criminals (with electronic tagging), 1,378 were judged to be living in concentrated poverty neighborhoods by probation officers. 5.5% of the offenders living in concentrated poverty neighborhoods committed another crime, whereas 4.6% of the general offender population committed another crime.
- Examining whether residency in concentrated poverty neighborhood increases the likelihood of recidivism, the concentrated poverty neighborhood residency variable was shown to be the major variable that increases the logistic regression model’s explanatory power.
- Examining the causes of recidivism, only age, number of offenses, and crime type (sex crime) were found to have significant impacts. In other words, in terms of criminals living in concentrated poverty neighborhoods, the younger the offender, the higher the likelihood of recidivism, after controlling for all other variables.
- Keeping all variables the same, sex offenders showed a statistically higher probability of recidivism compared to other crime types. Also, the more number of offenses, the higher the likelihood of recidivism.

Case Study on Urban Regeneration and Housing Policy in English-Speaking Countries

Urban Regeneration Policy in the US
- In the United States, the public housing programs that began in 1937 as part of the Roosevelt administration’s New Deal resulted in poor residents losing their homes in city centers, which led to adverse effects such as urban riots. Model city programs, community development block grants, urban development execution grant programs, and natural residential neighborhood growth programs, which were attempted from many angles during the 1960s to the 1980s, failed for various reasons.
- In the 1990s, urban movements such as New Urbanism and sustainable communities came into being, and housing policies for low-income families were implemented. The HOPE VI program, a public housing supply program aimed at integrating residents which began in 1993, was transformed in 2010 into a resident selection program that was designed to build houses where various income groups could live in deprived areas, and to provide improves in all areas of health, security, employment, mobility and education.

Urban Regeneration Policy in the UK
- Until the 1960s, large-scale housing construction was carried out indiscriminately, and in the 1970s, concentrated poverty neighborhoods emerged in city centers. In the 1980s, as the manufacturing industry declined, factories in cities were closed and relocated to the outskirts.
Large-scale redevelopment projects were implemented, but they further intensified the imbalance between regions.
- This promoted a comprehensive and integrated rehabilitation plan for social and economic aspects in the 1990s. Since the late 1990s, it has focused on five core tasks: employment, housing, education, crime and health. As interest in security increased after the World Trade Center attack in 2001, designs for crime prevention and terrorism prevention were emphasized.

Impact of Housing Programs on Crime
- In order to examine the impact on crime by housing programs aimed at dispersing poverty neighborhoods, a total of 26 studies were selected after a systematic review of relevant academic literature published since 2013.
A total of 10 studies examined the impact of Moving to Opportunity (MTO), a United States program which focuses on relocation and resettlement, on low income adolescents' delinquencies, problematic behaviors and mental health. A total of 6 studies examined the impact of low-income housing voucher programs on local crime rates. The remaining 10 studies looked at the impact of HOPE VI, a housing diversification program, on drug abuse and crime.
- Out of the ten studies on the MTO program's effectiveness, two studies showed negative effects, two studies showed positive effects, the remaining six studies showed both positive and negative effects.
- Out of the six studies on low-income housing voucher programs, three found that they were ineffective against crime, while the other three showed negative effects.
- Out of the total of 10 studies related to migration to new areas by HOPE VI program participants, four studies showed negative effects and five studies showed positive effects, while the one remaining study reported both positive and negative effects.

Case Study on Designing Crime Prevention Environments in Japan

Background
- Japan has been trying to prevent crime through state-led crime prevention environment related legislation from the urban design stage. If the domestic CPTED project had a point or line approach centered on a specific area, Japan is currently expanding its scale to include roads, complexes, and urban planning across the board.
- In legislative terms, Japan is taking the lead in establishing regulation (e.g. Tokyo safety and security town development regulation, Kanagawa crimeless safety and security town development promotion regulation, Osaka safe town development regulation, etc.), and the specifics are more detailed than Korea has.

Public Private Partnership
- Safety and security town developments, which are being implemented through public private partnerships (police-local government-local residents), are accomplished through local self-reliance with the recognition that residents must secure local safety and security by themselves.
- The support and cooperation from the National Police Agency (e.g. environmental designs for crime prevention, promotion of a crime-resistant community rehabilitation plan, securing infrastructure for self-crime prevention activities, and support for effective self-crime prevention activities) aligns with the operations of Korea’s Crime Prevention Officer (CPO) teams.

Case Study of Eirin district in Osaka City
- Although the Eirin district is a region perceived as a high-crime rate region where many poor and basic assistance recipients live, as a result of effective police policy and other measures, the crime rate is not as high as people think.
- This is a case that demonstrates crime rates are not necessarily high in areas expected to be high poverty areas, such as areas with high concentrations of welfare recipients and day laborers.

Direction of Crime Statistics and Crime Prevention Policy
- In Japan, there are relatively many different types of single houses and villas as opposed to large apartment complexes. Examining the statistical data on burglaries, the type of crime most prent in terms of housing crimes for single houses and villas, the crime rate among apartment houses (4.4% for 4th floor or higher, 11.9% for 3rd floor or lower) is much lower than single houses (42.0%), confirming that this shows the effectiveness of the policies and systems implemented by the Japanese government to lower the rate of home burglaries.
- Japan applied CPTED from the government-led urban planning stage, and implemented housing design, urban regeneration and self-directed crime prevention activities according to the characteristics of each area through 'safety and security town development', which adopted English speaking countries' crime prevention environment design methods.

Significance and Case Study of Domestic Urban Regeneration Projects

Significance of Urban Regeneration Projects
- Domestic urban regeneration projects can be categorized into urban vitality promotion development projects, new town village projects, urban regeneration projects and urban regeneration new deal projects based on the relevant laws, but the projects that are actually being undertaken are classified into urban regeneration new deal projects and new town village projects.
- However, when we look at the relevant project guidelines and manuals, CPTED project related contents are incomplete, and mainly hardware-oriented CPTED design elements are simply listed by category, making it difficult for local governments to reflect CPTED in project master plan as they prepare for a new town village project or urban regeneration new deal project.

GamOOdong Case Study Analysis Results
- According to the analysis of maintenance and utilization status of the major CPTED facilities for the GamOOdong new town village project, maintenance was generally adequate and utilization status was relatively high. However, several problems were found in terms of maintenance.
- According to an investigation of the social activities of residents at the designated site, various social activities were observed especially in the vicinity of safe roads and in areas where CCTV was installed.
- According to the survey results, awareness of the crime prevention environment improvement project was high at 83.8%. However, there was a big difference in awareness among sub-projects.
- Satisfaction with safety facilities (safety railings, embankments, etc.), CCTV, safety lighting fixtures, and safety bells was relatively high, whereas that of graffiti and garbage recycling was very low. In the case of graffiti and garbage recycling, not only was awareness low, but maintenance problems were also found during the on-site investigation.
- As a result of analyzing the correlation of each indicator using the structural equation model, it was found that, as the satisfaction level with the everyday life security of the neighborhood increased, the fear of crime was decreased and the level of neighborhood satisfaction was increased. In addition, the quality of the walking environment had a positive effect on neighborhood satisfaction and sense of community. When the level of neighborhood disorder was low, neighborhood satisfaction was increased, fear of crime was decreased, and sense of community and neighborhood satisfaction level appeared to increase. Therefore, in the course of promoting the crime prevention environment improvement project, it is necessary that the physical plan is accompanied by improvements to the quality of the walking environment, the disorder of the neighborhood, and the satisfaction with everyday life security.

Case Analysis of Physical Characteristics through Risk Assessment

Significance
- An onsite case study investigation using a checklist was conducted for a total of 10 sub-districts in which one region with high ratio of basic assistance recipient households and another with a low ratio were selected from the following cities: Seoul, Daejeon, Daegu, Gwangju, and Incheon.

Analysis Results
- In the case of the physical environment of the streets, Seoul fared moderately well at night lighting, distinction between sidewalks and roadways, and physical disorder, and there were difficulties in finding vulnerable aspects.
The target areas of Daegu were good at controlling CCTVs and the passages between buildings, but vulnerable in terms of physical disorder. There were good resting facilities on the streets in the target areas of Daejeon, but there were many blind spots due to curved streets. Incheon was good at horizontal structure, but vulnerable in night lighting, distinction between sidewalks and roadways, and resting facilities. Gwangju was vulnerable in terms of CCTV, control of passages between buildings and resting facilities and generally, it was difficult to find aspects that were satisfactory.
- As a result of examining the physical environment of individual buildings in the five surveyed target areas, Seoul and Incheon had a high proportion of multiplex houses while Daejeon and Gwangju had a high proportion of single houses. Daegu had a high proportion of shopping districts and row houses.
- In the case of boundary installations and internal monitoring of buildings in each target area, and considering the possibility of internal monitoring when a boundary facility is installed, in Seoul, all buildings with border facilities were comprised of low walls or transparent gates that could be monitored internally. Daegu and Incheon also had high internal monitoring capabilities. However, Daejeon and Gwangju had high walls or opaque gates, making internal monitoring difficult. The ratio of crime prevention facilities installed on the boundary of buildings was generally very low.
- In this field survey, there was no distinction in the physical environment between concentrated poverty neighborhoods and comparison areas. The physical environment dictated by the type of building was the most significant distinction between the target areas and the comparison areas. The ratio of single houses in the target areas was high, while the ratio of multiplex houses and shopping districts in the comparison areas were relatively high.
This distinction leads to a differentiation in the structure or form of the buildings, and it also seems to affect the installation rates for crime prevention facilities that handle surveillance and access control.

Analysis of Welfare and Security Policies on Domestic and Overseas Concentrated Poverty Neighborhoods

Domestic Central Government
- The roles of the Ministry of Justice, the Ministry of the Interior and Safety, the National Police Agency, the Ministry of Gender Equality and Family and the Ministry of Health and Welfare were examined. In the case of the Ministry of Gender Equality and Family, numerous preventative programs are carried out, but follow up and management programs are carried out to an even wider extent. In the case of the Ministry of Health and Welfare, many institutions in charge are not directly related to crime prevention, but instead deal with the fundamental aspects of crime prevention by enhancing national security through a stronger social safety net that is granted by social security institutions.
- While the Ministry of Gender Equality and Family focuses on protecting women, young adults, and the victims of domestic violence and sexual violence, the Ministry of Health and Welfare is mainly in charge of providing welfare and safety support systems for children and the elderly.

Domestic Local Governments
- While it is difficult to say that the characteristics and direction of local governments are strong given most of them are financially supported by the central government, plans and programs to promote social security, such as crime prevention, are pushed ahead and implemented under limited circumstances, in consideration of the characteristics of the relevant local government.
- Recently, crime prevention environment design has occupied a very important position among the major projects of local governments. In order to prevent and respond swiftly to crime, accidents, and natural disasters, CCTV installations and functionality are being improved rapidly. A total of 229 integrated control centers will be established by 2019, and it is anticipated that crime prevention and the overall safety of society will be secured.

Domestic Case Study
- The welfare and security projects in the case areas had a common factor in that they ran local projects that used household visits and local resources.
The function and role of the public sector was extended beyond simply implementing the central government's system into planning and implementing communications and coordination with the local private sector.
- Communication and cooperation with the private sector differed slightly depending on the resources available in the area. When a large hospital in the region is used as a base, support for the healthcare sector was more active. In the case of a municipal district where a permanent rental apartment complex was located, the local agenda and specialization projects were planned and operated according to the geographical characteristics of the rental apartment complex. In particular, areas with permanent rental apartment complexes had a higher frequency of visits and access than areas with single or villa households.
- Another common factor was that the social welfare budget ratio was higher than the national average. However, the ratio of the public order and security budget in the studied areas was lower than the national average of 1%, and there were differences between local governments in terms of the importance of welfare and security policies.
- Since concentrated poverty neighborhoods deal with complex issues such as crime, everyday life security, and welfare support, collaboration between the regional public offices has already been discovered. In addition, from the perspective of the central government, a more formalized system to handle the complex needs of poor people (neighborhoods) can be expected if regional projects are developed and supported through communication and cooperation between the related governments (departments).

Overseas Case Study
- In 2017, the Swedish government implemented a new crime prevention policy called ‘Combating crime together'(Tillsammans mot brott). This policy focuses on ensuring that all of society is responsible for crime prevention. A city government’s active attention, such as the city government's leadership and prioritizing crime prevention in the agenda, appeared to be important in successfully prevent crime. In addition, cooperation between the city’s crime prevention coordinator and the police is important, as well as cooperation with local businesses (business groups and management groups) and civil societies (civic groups).
- The crime rate in the Netherlands is declining. Establishing cooperative relations among various organizations such as the central government, the local governments, the police and the district attorney is a very important success factor in reducing crime rate. Long-term strategies to reduce the crime rate, willingness from leaders, and strong leadership to implement hem are important. In addition, cooperation between government departments, private, and civil society is important, and this kind of cooperation is implemented effectively in the Netherlands.
- In Seattle in the United States, the Neighborhood Matching Fund (NMF) was introduced in the late 1980s. The Neighborhood Matching Fund is a program in which the city matches the funds raised by local civic groups or institutions when such groups or institutions carry out community development projects.
The goal of the Neighborhood Matching Fund program was not crime control but community development. The program was successful because community development programs such as the NMF strengthens solidarity among community members and attempts to improve community conditions related to crime, disorder, violence and the like.

Comprehensive policy alternatives

Prioritized environmental improvement and benchmarking of best practices for medium-density concentrated poverty neighborhoods
- Develop and apply various CPTED business models in connection with urban renewal projects. Develop a variety of CPTED business models linked to urban renewal projects and reflect them in the New Town Village Project program, urban renewal New Deal project manual, and the like.
- To more effectively promote urban regeneration projects in the process of carrying out crime prevention environmental improvement projects, a tailored plan should be established for each of the project regions based on the results of site analyses, rather than simply listing existing CPTED design factors like a department store.
- It is necessary to benchmark both the best practices overseas as well as the successful CPTED projects of the Seoul Design Policy Division. Story design, rationalization of business management through role division, improvement of project understanding and expertise, a clear definition of the problem, and strengthening communications between stakeholders should be embedded.

Demand-tailored strategy for each policy target group
- Because objective poverty level, rather than relative poverty that compares one’s self to others, directly increases the likelihood of crime and indirectly affects the likelihood of damage from crime and potential harm, societal measures against absolute poverty should be devised.
- Poverty-related crime prevention policies must focus on creating countermeasures at an individual unit level, social/environmental unit level, and individual-social interaction level. These include strengthening policies that alleviate economic tensions, strengthening policies that improve family ties, improving individual perceptions of the law (positive attitude towards law), and developing initial attachment programs (for example, voluntary/local community support projects, etc.) that allow local residents to respond to disorder.

Supplementing residence movement policies
- Provide housing counseling services to low-income households before they move to new neighborhoods.
- Essential to collaborate with the police and housing management agencies.
- Develop special programs for low-income males so that they can adapt well to new environments when moving, as such moves may negatively affect their behavior and emotions.

Strengthening welfare and safety policies
- uations required for the welfare and safety project groups that relate to crime prevention effects.
- Strengthen program promotion.
- Construct and use a comprehensive information system related to welfare and safety. Mainly with regard to the before and after welfare security support system as it relates to stigmatization, the government and the local community should actively endeavor to prevent leaks of personal information.
- Develop and support local projects through connections and collaborations between the relevant central government ministries. Prepare a comprehensive manual on public-private partnerships at the central government level, and operate projects tailored to the region’s characteristics.
- To improve security in poor areas, it is necessary for experts to support ocal residents so that the residents can plan and promote security and welfare improvement solutions, and instill in them a sense of pride of their neighborhood.

Institutional improvements
- Continued attention is required to improving the standards for construction meant for crime prevention. In particular, strengthen the implementation and standardization of safety facilities (products) and lighting installation and management.
- Prepare plans to strengthen CCTV management. Establish specific installation and management standards such as lighting standards. These criteria shall include placement, performance, direction, height, awareness, poles and guide signs, priority, monitoring, inspection cycle, replacement criteria, etc.
- Establish a national standardization and certification system on security window performance.
- Strengthen the management policy for empty houses and redevelopment areas.

Diversification of the management plan for potential offenders in concentrated poverty neighborhoods
- If the residence of an offender with electronic tagging is located in a concentrated poverty neighborhood, identify any vulnerable areas, the CCTV installation status, and the status of any welfare facilities or children safety facilities near the offender. This must lead to further coordination with the relevant organizations.
- Strengthening the probation-police partnership is necessary.
- With a higher likelihood of recidivism from individuals with insecure housing such as dosshouse area residents and homeless people as compared to those with deposit-backed or monthly rental housing, it is necessary to provide intensive guidance, management, and support measures to residence-denial criminals even at the probation stage. They should move to a normal area and preparations must be made so that they can be located in normal residential areas, rather than in concentrated poverty neighborhoods that provide mere shelter.