Geography of Crime in China since the Economic Reform of 1978 : A Multi-scale Analysis [1 ed.] 9781443884174, 9781443875356

Citation preview

Geography of Crime in China since the Economic Reform of 1978

Geography of Crime in China since the Economic Reform of 1978 A Multi-scale Analysis By

Yijing Li

Geography of Crime in China since the Economic Reform of 1978: A Multi-scale Analysis By Yijing Li This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Yijing Li All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-7535-X ISBN (13): 978-1-4438-7535-6

I would like to dedicate this book to my dearest son, Hengyi, my husband, Liang, and the world’s greatest parents.

TABLE OF CONTENTS

List of Figures............................................................................................. ix List of Tables .............................................................................................. xi Preface ...................................................................................................... xiii Acknowledgements .................................................................................. xiv Introduction ............................................................................................... xv Chapter One ................................................................................................. 1 Introduction 1.1 Economic Reform in China 1.2 Social changes in China during the economic reform 1.3 Crime changes in China since the start of the economic reform 1.4 Aims of the research 1.5 Structure of the book Chapter Two .............................................................................................. 15 Theoretical Background 2.1 Western studies into the geography of crime during periods of social change 2.2 Practical differences in Chinese society 2.3 Research aims and hypotheses Chapter Three ............................................................................................ 41 Data and Methodology 3.1 Study area description 3.2 Justification of the Methodology 3.3 Data Chapter Four .............................................................................................. 72 Analysing and Modelling Crime Change on a National Scale 4.1 Introduction 4.2 Data 4.3 Methods 4.4 Results 4.5 Conclusions and discussion

viii

Table of Contents

Chapter Five ............................................................................................ 106 Analysing and Modelling Crime Change on the Provincial Scale 5.1 Introduction 5.2 Data 5.3 Methods and research design 5.4 Results 5.5 Conclusions and Discussion Chapter Six .............................................................................................. 145 Analysing and Modelling Crime Change on a City Scale 6.1 Introduction 6.2 Data 6.3 Methods and research design 6.4 Results 6.5 Conclusions and Discussion Chapter Seven.......................................................................................... 169 Crime on the Neighbourhood Scale: A Case Study in Shenzhen City 7.1 Introduction 7.2 Data 7.3 Methods and research design 7.4 Results 7.5 Conclusions and Discussions Chapter Eight ........................................................................................... 238 Summary and Conclusions 8.1 Introduction 8.2 Results and conclusions 8.3 Discussion and future work Bibliography ............................................................................................ 262 Appendix A.............................................................................................. 281 Appendix B.............................................................................................. 302 Appendix C.............................................................................................. 326 Appendix D ............................................................................................. 362 Appendix E .............................................................................................. 386 Appendix F .............................................................................................. 407 Index ........................................................................................................ 412

LIST OF FIGURES

Fig. 1-1 Largest interprovincial migration streams 1995-2000 (Chan, 2001)........... 5 Fig. 1-2 The crime rate from 1950 to 2003 in China ................................................ 7 Fig. 2-1 The Gini index and crime rates in China from 1981 to 2001 .....................34 Fig. 3-1 Scalar structure for research design ...........................................................41 Fig. 3-2 Map of Guangdong Province in China ......................................................42 Fig. 3-3 An administrative map of Shenzhen ..........................................................44 Fig. 3-4 Locations of representative neighbourhoods..............................................45 Fig. 3-5 Procedure for the neighbourhood case studies ...........................................53 Fig. 3-6 Conceptual framework for macro-scale research .......................................58 Fig. 4-1 Total crime level in China from 1978 to 2008 ...........................................77 Fig. 4-2 Levels of the main crime types in China from 1978 to 2008 .....................78 Fig. 4-3 Differences between AC and EC over the past 30 years ............................81 Fig. 4-4 Example of data property check (EC and AI) ............................................81 Fig. 4-5 Long-run relationships between different types of crime and other variables ............................................................................................................93 Fig. 4-6 Short-run relationships between different types of crime and other variables ............................................................................................................97 Fig. 5-1 Administrative map of China at a provincial scale...................................105 Fig. 5-2 The criminal case process in China..........................................................106 Fig. 5-3 The three law enforcement agencies' roles in China ................................108 Fig. 5-4 Crime data comparison at national scale.................................................. 113 Fig. 5-5 Four-quadrant chart for crime level comparison among provinces from 1988 to 2008 ........................................................................................... 119 Fig. 5-6 Global Moran’s I value from 1988 to 2008 ..............................................120 Fig. 5-7 Spatial-temporal distribution of crime counts ..........................................122 Fig. 5-8 Spatial-temporal distributions of crime rates on those arrested ...............123 Fig. 5-9 Crime rates comparison among selected provinces .................................124 Fig. 5-10 Data property check for the four selected provinces ..............................125 Fig. 5-11 Long-run relationships comparison........................................................135 Fig. 5-12 Short-run relationships comparison .......................................................138 Fig. 6-1 Administrative map of cities in Guangdong Province, China ..................145 Fig. 6-2 Data interpolation example - CR series in Guangzhou ............................150 Fig. 6-3 Guangdong Province crime rates from 1988 to 2008...............................151 Fig. 6-4 Crime rates in the four cities from 1997 to 2008 .....................................152 Fig. 6-5 Robbery cases in the four cities from 1990 to 2008 .................................153 Fig. 6-6 Data property check examples in four cities ............................................155 Fig. 7-1 The administrative map of Shenzhen .......................................................171 Fig. 7-2 Neighbourhood selection in Shenzhen city ..............................................172 Fig. 7-3 Land-use sketch map for TL ....................................................................173 Fig. 7-4 Land-use sketch map for HB ...................................................................174

x

List of Figures

Fig. 7-5 Land-use sketch map for HL....................................................................174 Fig. 7-6 Focus group discussions in TL neighbourhood ........................................176 Fig. 7-7 (a), 7.7(b) and 7.7(c) Neighbourhood snapshots .....................................177 Fig. 7-8 Crime rates and proportions in Shenzhen: 1985 to 2007 .........................187 Fig. 7-9 Change in crime proportions for Shenzhen: 1985 to 2007 .......................188 Fig. 7-10 Residents’ responses to victimisation by seriousness of crime ..............207 Fig. 7-11 Neighbourhood insecurity comparison ..................................................214 Fig. 7-12 TL neighbourhood map ..........................................................................220

LIST OF TABLES

Table 2-1 Crimes committed by the migrant population (%) in selected Chinese cities in 1994 .....................................................................................................33 Table 2-2 Crime-related independent variables on macro scales .............................37 Table 3-1 Information collected through the three instruments ...............................62 Table 3-2 List of data sources and drawbacks .........................................................67 Table 4-1 Descriptions of variables used at the national level .................................74 Table 4-2 Summary of statistics for the three crime categories ...............................80 Table 4-3 Results of ADF unit root test ...................................................................83 Table 4-4 Granger Causality test for LC..................................................................86 Table 4-5 Larceny long-run equilibrium model .......................................................89 Table 4-6 Acquisitive crime long-run equilibrium model........................................90 Table 4-7 Expressive crime long-run equilibrium model ........................................91 Table 4-8 Linear regression analysis on the I(0) variables and first-order differenced crime rate .......................................................................................92 Table 4-9 The short-run VECM model - larceny .....................................................93 Table 4-10 The short-run VECM model - acquisitive crime ...................................95 Table 4-11 The short-run VECM model - expressive crime ....................................96 Table 4-12 Results of variance decomposition for LC - Economy ..........................98 Table 4-13 Results of variance decomposition for LC - Social harmony ................98 Table 4-14 Results of variance decomposition for LC - Social improvement and citizens’ welfare..........................................................................................99 Table 4-15 Results of variance decomposition for LC - Culture .............................99 Table 5-1 Descriptions of variables at the provincial scale ................................... 110 Table 5-2 Correlations between police recorded data (TC) and those from the procuratorate ............................................................................................. 113 Table 5-3 Crime rates for all provinces from 1988 to 2008 ................................... 115 Table 5-4 Inequality measure for the spatial distribution of arrested crime counts..............................................................................................................122 Table 5-5 Inequality measure for the spatial distribution of arrested crime rates .................................................................................................................123 Table 5-6 Non- I(1) variables in ADF tests ...........................................................127 Table 5-7 Significant Granger cause variables ......................................................127 Table 5-8 Variables correlated significantly with crime rates in the long-run .......129 Table 5-9 Variables correlated significantly with crime rates in the short-run.......131 Table 5-10 Variables explaining crime rate variation after variance decomposition .................................................................................................132 Table 5-11 Long-run relationships.........................................................................134 Table 5-12 Short-run relationships ........................................................................137 Table 6-1 Descriptions of variables at the city level ..............................................147

xii

List of Tables

Table 6-2 Non- I(1) variables in ADF tests ...........................................................156 Table 6-3 Variables correlated significantly with crime rates in the long-run .......157 Table 6-4 Short-run relationships ..........................................................................159 Table 6-5 Variables explaining crime rate variation after variance decomposition .................................................................................................160 Table 6-6 Granger cause variables.........................................................................163 Table 6-7 ADF test for non-I (1) variables ............................................................164 Table 6-8 Long-run relationships significant at 5% level ......................................165 Table 6-9 Short-rum relationships significant at 5% level .....................................165 Table 7-1 Recoding table for the responses to questions on NCI ..........................181 Table 7-2 Definition and descriptive statistics of variables ...................................181 Table 7-3 Neighbourhood sample population characteristics ................................192 Table 7-4 Sample demographic differences between neighbourhoods ..................193 Table 7-5 Summary of residents’ victimisation experiences..................................194 Table 7-6 Number of residents who have been victimised in the past two years ...195 Table 7-7 Regression results of residents’ victimisation model .............................199 Table 7-8 Case summary of respondents’ selections (PS and NPS, %) .................202 Table 7-9 Regression results of residents’ perceptions on safety models ..............204 Table 7-10 Response categories (%) .....................................................................206 Table 7-11 Regression results of residents’ responses to serious crimes model (RSC) ..............................................................................................................209 Table 7-12 Regression results for residents’ responses to medium level crimes model (RMC)..................................................................................................210 Table 7-13 Regression results of residents’ responses to minor crimes model (RMNC) ..........................................................................................................212 Table 7-14 Generic form of Contingency table for Kappa statistics (Monserud & Leemans, 1992) ...........................................................................................215 Table 7-15 Contingency table in TL ......................................................................216 Table 7-16 Contingency table in HL .....................................................................217 Table 7-17 Contingency table in HB .....................................................................217 Table 7-18 Unofficial and official data comparison ..............................................221 Table 8-1 Co-integration relationship analysis results ...........................................239 Table 8-2 Long-term relationships on a city scale .................................................247 Table 8-3 Short-term relationships on a city scale .................................................248 Table 8-4 Summary of dimensions affecting crime on a macro scale ...................254

PREFACE

From a geographer’s point of view, this book has been produced against the background of the rapid social change in China since the economic reform of 1978. The research for it undertook spatial-temporal analyses of crime in China on different geographical scales, utilising both quantitative and qualitative data, and applying GIS (Geographical Information System) and spatial data analysis approaches, and provided the opportunity to test relevant theories developed in western countries in the context of China. Despite a general increase trend on crime being observed, there are obvious effects of scale, either spatial scale or temporal scale, were considered a crucial precondition in analysing crime changes and relevant influential socioeconomic conditions during the social transition in China since the economic reform. Some indicators turned out to have consistent effects on crime as suggested by criminological theories, either criminogenic or crime-mediating, while some others did not due to the variations from spatial scale and temporal scale. It is suggested that the adaptation of the relevant theories to China needs to take into thorough consideration the targets’ contextual features. A unique contribution of this book is the attention to describe the crime change during social change in China top-down. Another hallmark of this interdisciplinary book is that it analyzed and displayed the social environmental conditions relevant to crime change from a spatial view. Overall, it provides a holistic view on the crime issues in China by both statistical way and case studies in detail, which could serve as both a reference and a good example to follow. The potential audiences who might be interested in this volume could be either academic researchers doing criminological studies, spatial data analysts, and sociologists; general readers who are interested in the topic of social transition in China since the economic reform; investigators trying to follow the similar procedures in other neighbourhoods; police officers hoping to get references from this book on local crime control and prevention, or local officials or policy makers hoping to derive illuminating information on policy making.

ACKNOWLEDGEMENTS

There are many people I would like to acknowledge on this page for their help during my PhD studies. First and foremost I would like to offer my sincerest gratitude to my supervisor, Prof. Robert Haining, who has supported me with his unlimited patience and professional knowledge, while allowing me enough room to explore knowledge in my own way. He has always been available for discussion of my progress, and for advice on the problems I have come across during my research. He has been very supportive in helping me to get the crucial financial support for my fieldwork and the completion of this thesis. His enthusiasm and rigorous work have inspired me in almost every aspect of life. As a supervisor, he could not have done more to help. I would like to extend my biggest thanks to my tutor at St John’s College, Dr. Sue Colwell, who has been very helpful throughout the four years of my PhD, helping me to manage my college life, giving me suggestions on life issues and on applying for various funding opportunities to support my research. I would also like to give great thanks to my colleagues in the main building of the Department of Geography: Dr Molly Warrington, Dr. Mike Bithell, Gae Matthews, Maria Constantinou, Dr. Gabriel Amable, Mr Robert Carter, Mr Shane Harvey, Dr. Jie Ding, Dr. Andy Wang, Shino Shiode, Andre Silveira, Dr. Yu Wang, Dr. Han Meng, Dr. Siyuan He, Ziyue Chen, Feng Mao and Dr. Xingjian Liu. Without your encouragement and help, my research life in Cambridge would not have been so enjoyable. My research has gained support from the Cambridge Overseas Trust, RGS-IBG in the form of a Hong Kong Research Grant, the Shenzhen Land Resources and Planning Bureau, Dongmen Police Station in Shenzhen, Prof. Hui Lin from Hong Kong Chinese University, Dr. Jianguo Liu from Hong Kong Urban University, Prof. Chunlei Li from China Public Security University, Prof. Shihong Du from Peking University, and Dr. Peng Chen from Tsinghua University. I have also been blessed by a warm and cheerful group of friends, both in my research and in my daily life: Echo Yue Ouyang, Lulu Li, Huajing Zhao and Kathy Liu. Last but not least, I would like to express my earnest gratitude to my parents, husband and son; only with your support could I have got this book published on time.

INTRODUCTION

This study was proposed against the background of rapid social change in China since the economic reform of 1978. It undertook a spatialtemporal analysis of crime in China from a geographical perspective, applying GIS (Geographical Information System) and spatial data analysis approaches, and provided the opportunity to test theories developed in western countries in the context of China. This work firstly investigated the changes in crime levels and types in China since 1978 on different environmental scales (national, provincial and city scales), using official statistical data and quantitative analysis techniques. It found that crime levels in China have, on average, increased dramatically, regardless of scale, and that crime patterns have varied by region over time. Secondly, it examined the on a macro scale the relationships between crime levels (rates) and sets of selected social, economic, demographic and cultural indicators, as suggested by different criminological theories, using statistical techniques including co-integration tests and regression. Based on the results, some indicators turned out to have consistent effects on crime, as theories suggested, whether they be criminogenic or crimemediating, while others did not, and their influences varied by region, spatial scale and temporal scale. Thirdly, it conducted a comparative study on a micro scale in selected neighbourhoods in Shenzhen city, and an in-depth case study in a particular neighbourhood (named TL), using primary data and a combined methodology using both qualitative and quantitative techniques. The results indicate that residents’ victimisation experiences, their perceptions on neighbourhood safety and their responses to crimes are related to their individual and neighbourhood characteristics. Through the case study in TL, important disparities between official and unofficial crime data were detected, and crime hotspots varied both spatially and temporally by landuse function.

CHAPTER ONE INTRODUCTION

1.1 Economic Reform in China China is now the 2nd largest global economy, with GDP increasing from 268.3 billion (US dollars) in 1978 to , billion in 2014, that is, by more than 38 times (National Bureau of Statistics of China, 2015). This big leap is thought to have been the achievement of the economic reform introduced in 1978. In the years before 1978, the planned economy was thought to be the hallmark of China due to its great contribution to China’s economic recovery and primary development from the 1950s up to 1978. Nevertheless, the maladjustment between economic policy and social development became more and more obvious as time went by. In order to adjust the policies on development, Deng Xiaoping proposed a Reform and Opening-up Policy in 1978, which was composed of internal reform and external opening-up. It placed great emphasis on economic development to enhance people’s income levels and living standards. The policy addressed problems associated with weak central planning whilst encouraging enterprise and foreign investment. The internal reform was first tried out in 1979 in the vast rural areas and the Special Economic Zones (SEZs), with Shenzhen city as the most significant representative. The opening-up to the outside world was also started in the special economic zones, allowing them to take priority in foreign trade. Later, in 1984, another 14 cities, including Guangzhou and Shanghai, were selected as foci for international business. The pilot reform in rural areas saw great success, followed by the reform in urban areas, which started in 1985 and was further adjusted in 1992. Essentially, this economic reform was a reversal of Mao’s policy of economic self-reliance; however, the planned and centralised management of the macro economy was applied indirectly and depended on market mechanisms. Over the course of 30 years, the Reform and Opening-up Policy has helped China become a modern, industrial socialist nation with distinct Chinese characteristics. The rigid state controls over the economy established after the Communist Revolution have been relaxed, and

2

Chapter One

market mechanisms have been encouraged (Zhuo and Messner, 2008). China’s economy has abandoned its former isolation in favour of deep engagement with world markets (Rawski, 2008), and this socialistic market economy reform attracted more and more foreign investment, advanced technologies and management experience. China has evolved from a closed country, with more than 200 million people living in poverty, to a thriving and prosperous economic entity tightly linked to the global economy. China’s successful transition from planned economy to market economy has brought about immense changes in its economic strengths and international status. China’s economic strengths have been continuously enhanced by achieving the number one outputs of economic products, dramatically solving the shortages of many necessities. Between 1978 and 2008, the import and export trade also saw a hundredfold increase, from 20.6 billion (US dollars) to 2,174 billion, resulting in China becoming the world’s third largest trading country. In the meantime, Chinese domestic living standards have improved and many people are now leading a more comfortable life, supported by the guarantee of a better welfare system. The GDP per capita has increased from 381 (US dollars) to 3,381, with per capita income increasing from 190 (US dollars) to 2,360. Meanwhile, the Engel index has decreased from 57.5% in urban and 67.7% in rural areas, to 36.3% and 43.1% respectively (China Statistical Yearbooks 1978-2008). The basic characteristics of China have been permanently altered (Deng and Cordilia, 1999), and the changes are profound, complex, and far-reaching. In all, China has experienced momentous shifts, from poverty to growing prosperity, from village to city, from planned economy to market economy, from public toward private ownership, and from isolation to global engagement (Rawski, 2008).

1.2 Social changes in China during the economic reform As economic reforms have been implemented, massive and protracted social change has transformed China from an underdeveloped nation into a growing economic power, with obvious consequences in every aspects of social life: economy, society, politics, culture and demography. First of all, economic reform has stimulated a dramatic economic upsurge in the past three decades, and the introduction of a market economy has encouraged the reform of national enterprises and the introduction of foreign capital. This has led to a rapid shift from stateowned to private enterprise, accompanied by an ever-larger class of entrepreneurs. These newly emerged elites are business owners or managers with high positions in private and joint venture enterprises

Introduction

3

(Duckett, 2001; Ding, 2000). They are able to enjoy a high standard of living owing to their high level salaries. At the other end of the social hierarchy, a new urban poverty stratum is emerging from laid off and retired labour (Bian, 2002), as well as disadvantaged people (e.g. migrants). These two extremes aggravate the inequality problem in urban areas. Social achievements under the policy of Reform and Opening-up are also remarkable: infrastructure and public services have been improved significantly; people’s living standards have risen dramatically; urban areas have grown as more people are drawn to them. In the meantime, the social structure has changed profoundly from Mao’s pre-reform stratification system, which advocated egalitarianism, homogeneity and minimal socioeconomic inequalities. The new social order has redefined the composition of social strata, and social inequality is becoming more severe as the gap between rich and poor widens. When the economic reforms began to take hold in cities in the mid-1980s, urban residents’ incomes surged ahead of peasants’. At the same time, authorities in rural areas reduced peasants’ benefits, driving them to abandon their farms and migrate into cities to seek employment. Together, this resulted in mass migrations from rural to urban areas. In the meantime, since the mid-1990s, market-oriented policies have forced thousands of state-owned enterprises to close down, leading to tens of millions of workers being laid off without welfare protection. In 1997 alone, 10 million workers were laid off (Cai et al., 2008), hence high rates of unemployment. The mass migrations, redundant workers and reduced employment opportunities have worked together to worsen the urban unemployment situation, and further aggravate inequality. Overall inequality has unambiguously risen in China since 1987, and it is suggested that rising inequality is driven by income differences between localities, as some provinces are better situated to take advantage of market liberalisation and the reforms related to foreign trade and investment (Benjamin et al., 2008). Politically, China’s one-party political system and its tradition of centralised government are well known. During the economic reforms, China began its transition to a market economy under a hierarchical political system, which itself has been reshaped in response to the forces unleashed by economic transition (Naughton, 2008). A“bottom up” political reform, started in the late 1970s, introduced steady changes into the political and legal system. Mao’s pre-reform rural people’s communes collapsed; central control and bureaucratic planning proceeded in the form of reorganising ministries and state-owned enterprises; more democratic and “objective” decisions were demanded by academics and reformers,

4

Chapter One

and social control became a unique mix of Chinese tradition and a formalised legal system. Legal reform has emerged since the government’s recognition of “the structural strains in China's socio-legal system” (Li, 1996). It stemmed from the institutional mismatch between a developing market economy and a one-party political monopoly. Attention has been drawn to the need for judicial independence, protection of individual rights, and judges’ exercising of power in sentencing and punishment. In this sense, educational levels and cultural values have had dramatic effects on political and legal debates. Traditionally, Chinese society is influenced by the religious heritages of Buddhism, Confucianism and Taoism. Chinese culture emphasises collectivism, family, shame, informal codes of conduct, and respect for authority (Anderson and Gil, 1998; Ren, 1997; Clark et al., 1989; Zhang and Messner, 1995). Since the economic reform, the invasion by western culture in technology, films, fast food, literature and, most importantly, ideas, has challenged traditional beliefs despite governmental control. Now “getting rich” has become an important goal in modern Chinese society, and the advocating slogan “to be rich is glorious” motivates the way people behave towards economic interests. Traditional morals have given way to a series of money-based bonus systems (Dutton, 2005), allowing pecuniary orientation to permeate all aspects of society. Meanwhile, in order to meet the higher demand for technological development, the general education level has been enhanced. Intellectuals have helped to introduce advanced knowledge and updated information from the western world, contributing a lot to the technological reform necessary for economic development. Despite dramatic changes in the aforementioned four aspects, demographic change due to rapid urbanisation should also be mentioned. In pre-reform periods, the implemented registration system limited population mobility by giving each urban citizen a registration card, held by neighbourhood police stations (Cheng and Selden, 1994). However, economic reform has placed great pressure on this system due to the large labour surpluses in rural areas and the need for more labour in the growing cities. As a response, the Chinese authorities have loosened restrictions on mobility, and reduced migration control has allowed people to move freely across administrative boundaries. Together with increasing economic liberalization, this has led to millions of peasants moving into cities hunting for opportunities and new life-chances, creating a new category of migrants called the floating population, which accounts for the rapid increase in the overall volume of migrants. According to statistics, the proportion of the urban population in China has grown from 20% to nearly

Introduction

5

50% of the total, and is especially high in some developed areas. Taking Shenzhen as an example, its urbanised population proportion had reached 100% as early as 1995. However, migration rates and directions varied among cities according to the level of their development. For example, in some more developed cities, such as Beijing, Shanghai and Guangzhou, wealthy people moved out of the city centres to nearby rural areas. On the other hand, in most of the other cities the direction of migration is from rural areas to urban areas. As for interprovincial migration, the predominant movement is from the periphery (central and western regions) toward the core (eastern region). Between 1995 and 2000, for example (Fig. 1-1), interprovincial migration showed a significant concentration into a single province, Guangdong. Of the 30 largest interprovincial streams, 11 went to Guangdong, including all the top six flows (Chan et al., 2001).

Fig. 1-1 The largest interprovincial migration streams 1995-2000 (Chan, 2001) (Source: based on data from the State Council and National Bureau of Statistics, 2002).

6

Chapter One

1.3 Crime changes in China since the start of the economic reform 1.3.1 Crime levels (rates) Economic reform has stimulated economic development and migration, creating an unprecedented rise in the living standards of the Chinese people (Liu, 2001), widened the inequality gap and driven reforms in the political and legal systems. These changes are partly attributed to the emphasis on and promotion of the idea that “it is good to be rich”. In such an environment, some people are tempted to “get rich quick” by whatever means available, even if it involves criminal behaviours, which can be seen in soaring crime levels. Pre-reform China enjoyed very low crime rates and had earned the reputation of being a “crime free” society (Fairbank, 1987; Rojek, 1996). According to Dutton (1997), China had exceptionally low crime rates from 1949 to 1979, before the economic reform, with less than 0.5 million criminal offence cases registered every year. This can mostly be attributed to Mao’s “mass-line” policy (Rojek, 2001), in that crime control was not merely the responsibility of the police and government; it was meant to be the outcome of cooperation through voluntary guardianship by the masses, and the pervasive socio-political control of the Communist Party (Dutton, 2005). Since the implementation of economic reform and the open-door policy in the late 1970s, crime has been on the increase in China (Liu and Messner, 2001; Deng and Cordilia, 1999; Rojek, 1996). According to statistics, registered annual criminal offences proliferated six times to more than three million by the late 1990s. By the start of the 21st century, the relatively high level of criminal offences had levelled off at over four million per year - a figure approximately 80 times the rates observed in the 1950s and 1960s (China Law Yearbooks, 1983-2009). Based on official statistical data, the change in crime rates from 1950 to 2003 is shown in Fig. 1-2. A general rising trend can be seen from the graph, especially since the late 1980s, when the achievements of the economic reform became more apparent and established. Crime rates in pre-reform China were on average less than 10 per thousand population, and this low level had earned China a reputation for being a low crime society. The fluctuations from the 1950s to the mid 1960s should mostly be ascribed to certain historic events, such as the suppression of counter revolutionaries between 1951 and 1953, the initial completion of socialism, the era of the “great leap forward” and the “great natural disaster” from 1958 to 1962. Since the initial implementation of economic reform in 1978 and its

Introduction

7

wide spread in the early 1980s, crime rates have surged, but increases have been neither steady nor uniform. Crime rates did not go up steadily in line with the development of China; instead, there were two stable periods, in the second halves of the 1980s and the 1990s. The first peak in crime rates appeared around 1991, after a sudden increase from five per thousand to more than 20 per thousand, although it fell back quickly to 13 per thousand in 1994. Another steep increase occurred in 1998 after a stable period at 13 per thousand, and then they rose to more than 35 per thousand in 2003. It is hard to explain this abnormal increase, and the underlying socioeconomic conditions need to be considered to get a better understanding of this issue. For example, demographic conditions partly explain the surge in crime in the early 21st century. The birth boom in the late 1980s contributed a lot to the increase in juveniles in the early 2000s, who are a disproportionately crime-prone cohort (particularly males). But this cannot tell the whole story behind rising crime levels, and further study into the contribution of other social and economic changes is needed.

Fig. 1-2 The crime rate from 1950 to 2003 in China

1.3.2 Crime types According to the “Criminal Law of the People’s Republic of China”, the main crimes in China are categorised into Criminal Offenses and Public Security Cases by the degree to which they break the respective laws. There are specific criteria to decide if an offence is regarded as “criminal” or “minor”. The former category defines serious offences consisting of, for example, homicide, assault, robbery, rape, abduction,

8

Chapter One

larceny, fraud, and smuggling. The latter category defines less serious crimes as minor offences. The crime types analysed in this research are larceny, acquisitive crimes (robbery, fraud, abduction, and smuggling) and expressive crimes (homicide, assault, and rape). In the past 30 years, it has been argued that acquisitive crimes and assault constitute the bulk of offences recorded by the police. Bakken (2000) shows that robbery and rape have been on the rise in China, while the homicide rate has been increasing slightly. Theft of various properties was top of the list for offences, including theft of cars, theft of facilities, industrial materials, oil-field or power supplies, and petty theft in crowded locations, such as marketplaces and at tourist locations. There have also been reports of thieves operating from motorbikes, carrying out, for example, street robbery and purse seizing, which are rampant in the cities, especially Guangzhou and Shenzhen. In addition to the rise in common types of crime, corruption has become a growing concern to the Chinese public and the government. Violent crimes are thought to be linked to financial motivations, extortion, and personal conflicts. During the rapid urbanisation process, violent crimes have become more and more common (Chen, 1997; China Law Yearbooks, 1986-1997). It is believed that the high rates of violent crime could be attributed to increased opportunities and targets brought about by urbanisation and migration. Organized crime (or “gang crime”) is a growing problem in China, and it can be summed up in one phrase: “crime perpetrated by highly involved local evil forces” (Zhao et al., 1997). The present socio-economic conditions in China make gang crime attractive. Chinese police statistics and other sources indicate that 72.7% of the criminal groups identified are still engaged in conventional crimes like theft, robbery, burglary, murder, rape, human smuggling and extortion; 16.3% are involved in prostitution, gambling, and drug abuse, while 11% specialise in smuggling and drug trafficking. Organised crime is still in its primary stage and has not developed on the scale of the Triad in Hong Kong or Macau; however, similar criminal organisations are taking shape in the southeastern coastal areas and border provinces (Zhao et al., 1997). Drug-related crimes have re-emerged and gradually spread nationwide in China since the late 1970s. This has been ascribed to the decades long history of drug use and the weakening of regulation and control. Some synthetic drugs are popular among the young, and are often sold and consumed at entertainment locales such as nightclubs and karaoke bars. This situation has become increasingly serious, and drug-related crimes such as theft, robbery, and prostitution have also been on the increase. As the market-oriented economy took hold in China, the last few years

Introduction

9

have seen defrauders committing white-collar crime or corruption by finding loopholes and imperfections in the economic development and market administration. Yu (2008) used western concepts and theories to analyse China’s white-collar crime, and drew on materials from the media, in order to investigate the historical and institutional roots of corruption. As for the characteristics of offenders, it is reported that the majority are laid-off workers and members of transient populations, largely migrant rural labourers (Dutton, 1997). It is said that more than 50% of the criminal cases in some cities are committed by migrants from the countryside (Chen and Yu, 1993). It seems that, as China’s economy develops, the disparity among social classes is expanding, and exacerbating the crime problem in terms of both its level and type. If we want to propose a comprehensive and systematic strategy to deal with China’s crime problem, the integration of punishment and prevention strategies should be considered, especially focusing on the grassroots level to nip crime in the bud. The ultimate goal of crime prevention is to encourage social stability, which also underpins economic reform.

1.4 Aims of the research This study was proposed against the background of rapid social change in China since 1978, when the Reform and Opening-up policy was launched. Since then, tremendous changes have been experienced in China’s demography, economy, social order, culture, politics and legal system. These changes are believed to have had significant impacts on crime levels and types in the following three decades. Research on the links between economic reform and crime in China has mostly been undertaken by sociologists and criminologists (Kong, 2005a, b; Liu, 2001; Deng and Cordilia, 1999), and comparatively, research in this field from a geographical perspective is lacking. The general goal of this study is to undertake a spatial analysis of crime in China from a geographical perspective. This involves investigating the changes in crime levels and types in China since 1978 on different geographical scales; developing an understanding of the relationships between crime rates and various demographic, social, economic and other independent variables in different geographical contexts, and finally bringing forward pertinent suggestions on crime control to policy-makers and governmental agencies based on wellfounded analysis. Within this context, the following objectives are going to be met in order to achieve the overall goals:

10

Chapter One

1) Undertake a spatial-temporal analysis of crime levels (rates) and types in China on national, provincial and city scales using official statistical data and quantitative analysis techniques. 2) On different geographical scales (national, provincial and city scales), examine the relationships between crime levels (rates) and selected social, economic, demographic, cultural and political indicators, as suggested by theory, using statistical techniques including regression. The aim is to identify those variables that best explain the spatial and temporal variation in rates of different crimes at specified scales. The size of China, and in particular the differences in the way different parts of the country have changed since 1978, make it an interesting case to study. 3) Carry out a case study in selected neighbourhoods in Shenzhen city, Guangdong Province. Using primary data and a combined methodology of both qualitative and quantitative techniques, the aim is to better understand the effects of social cohesion and criminogenic conditions on neighbourhood crime levels and to assess residents’ views on crime in relation to official views and official data. The first objective focuses on the changing levels of different crime types. Time-series analysis will be utilised on each geographical scale, followed by comparisons among provinces or cities. GIS will be used for visualisation. The statistical data required at these macro scales (national, provincial, and city) will be secondary data, obtained from public sources such as the Statistical Yearbook and Law Yearbook. The second objective is to explore the relationships between crime levels of each type and selected variables at provincial and city levels. The questions I aim to answer are: what variables can best explain crime patterns at specific time points? Do the explanatory variables stay significant over time? Spatial cross-sectional and space-time series regression analysis will be the main techniques used. The data used for this objective will be secondary data from public statistical sources, e.g. China Statistical Yearbook and the Local Statistical Yearbook of each province or city. The last objective is to explore the effects of criminogenic conditions and social cohesion in selected neighbourhoods in Shenzhen. Using qualitative techniques (interviews and focus groups) and a quantitative technique (questionnaire survey) I will obtain individual-level primary data on residents’ perception and experiences of crime in their neighbourhoods. The question I propose to assess is: To what extent are the factors and issues that are highlighted in western criminology found to be relevant with respect to insecurity levels at the neighbourhood level in Shenzhen?

Introduction

11

The whole project aims to examine crime in China since 1978 from a geographical, rather than sociological or psychological, perspective. It will also be innovative in applying GIS and spatial data analysis methods to criminal issues in China. It further provides the opportunity to test theories that have been developed in western countries in the context of China. The expected significance of the research is as follows: 1) Historically, research into crime change during the post-1978 period in China has mainly been carried out by sociologists and psychologists; this research is conducted from a geographer’s perspective, using GIS and spatial data analysis approaches. 2) Comparative criminological studies (Bennett, 2004), have mostly been based on cross-national data (Bennett, 1991) to analyse the relationships between crime rates and social change (Lafree, 1999; Neapolitan, 1997) among nations or provinces (states) in developed countries. There is a lack of comparisons across provinces, cities or neighbourhoods in China. 3) Theories in this field were developed in western industrialised countries (Farrington, 1999) using data geographically biased towards western developed countries (Neapolitan, 1997). China, in this sense, has almost been neglected, which renders the theories incomplete. This research provides an opportunity to test these theories in a different society. 4) Guangdong Province had been ranked as quite a dangerous place in China, especially the cities of Shenzhen, Dongguan and Guangzhou, which are considered the most dangerous cities in China. It is essential to do a case study in one of these cities to understand this abnormal phenomenon in-depth. The results from this part of the research may also provide useful support for local crime control. To sum up, this research will analyse changes in crime levels and types in China since 1978, and discuss their relationships with sets of social, and economic indicators on four different geographical scales (national, provincial, city and neighbourhood). Moreover, it will not only test existing theories in the context of Chinese society, it will also address criminal issues from a geographical perspective and provide a researchbased reference point for crime control measures applicable at the local level.

1.5 Structure of the book The book has been divided into eight chapters, and brief introductions of each chapter are given below.

12

Chapter One

1.5.1 Chapter 1: Introduction This chapter briefly introduced some background information about China’s economic reform and the social changes, as well as the crime changes, that have been associated with the process. It also listed the research problems, my research motivation and aims, the objectives that will advance knowledge in the field of the geography of crime, the methodology that is going to be applied, and the structure of the book.

1.5.2 Chapter 2: Theoretical background This chapter will present relevant areas of the literature on the geography of crime and environmental (spatial) criminology, mainly developed in relation to western society and discussed in the literature on criminogenic and crime-mediating conditions. Based on the conclusions drawn from the previous research results, as well as the reality of Chinese society, a practical framework of selected variables will be proposed as a guide for undertaking research. This chapter also summarises the research aims and lists the hypotheses that need to be tested in the following chapters.

1.5.3 Chapter 3: Data and methodology This chapter describes the study areas on each scale to help build up a general impression of each target before going into the specifics of the research. It lists the data used at each level and provides the data sources where appropriate. As stated, both secondary data from official statistical sources and primary data collected through qualitative and quantitative methods will be used as the basis of analysis for this research, and the data quality issues that are of importance will be discussed further. This chapter justifies the combined methodology, both quantitative and qualitative, with which I will explore my research, but detailed descriptions of how I actually carried out the research will be given in the corresponding chapters.

1.5.4 Chapter 4: Analysing and modelling crime change on a national scale After a brief introduction to the crime problem on a national scale in China since the economic reform, data sources and their availability for this piece of research are discussed. On top of this, the research design and

Introduction

13

analytical methods used in this part of the work are explained, followed by the results obtained at the national level and concluding with a discussion of results.

1.5.5 Chapter 5: Analysing and modelling crime change at the provincial scale On a provincial scale, general knowledge on China’s administrative system and the registration and judicial systems is necessary in order to understand the results in this section. The structure of this chapter follows a similar pattern to that of Chapter 4, starting with an introduction to the target areas, particularly the selected case study provinces, and followed by the accessible data for this part of the research. It applies similar methods to those used at the national level in four selected provinces: Beijing, Shanghai, Guangdong and Henan. Results for these provinces are presented both individually and comparatively.

1.5.6 Chapter 6: Analysing and modelling crime change on a city scale The structure for the city scale analysis follows a similar pattern to that in Chapter 5. It is scaled down; four cities in Guangdong province were chosen as the target areas for study. After a brief introduction to the target areas, data used for this part of the research are provided in this chapter as necessary information. The four selected cities are Guangzhou, Shenzhen, Dongguan and Foshan. Regression results for each city, as well as comparisons among these cities, are presented.

1.5.7 Chapter 7: Crime at the neighbourhood levelA case study in Shenzhen city The neighbourhood level study takes Shenzhen city and chooses three neighbourhoods, labelled TL, HL and HB, as case study areas.1 Based on primary data collected through a questionnaire survey, three models have been built (one of residents’ victimisation experiences, one of their perceptions on neighbourhood safety and one of their responses to crimes) to identify the explanatory variables. Furthermore, the TL neighbourhood

  1

Under the confidentiality agreement with the interviewees in selected neighbourhoods, I was only allowed to use these labels as representatives for their names.

14

Chapter One

was chosen for an in-depth investigation into the similarities and disparities between official crime data and residents’ fear of crime. This was done by comparing the spatial pattern of recorded crimes and residents’ mental maps of unsafe places.

1.5.8 Chapter 8: Summary and conclusions This chapter summarises the findings from all of my research; not only the main findings, but also the potential problems incurred by data and methodology limitations. It concludes by evaluating how the results extend the current literature and answering the hypotheses listed at the beginning. It also discusses the implications of these results and the future direction for research in this field.

CHAPTER TWO THEORETICAL BACKGROUND

This chapter had given an introduction to the development of the geography of crime, explained the key concepts within the theoretical literature, and briefly presented the literature on crime-related socioeconomic variables. It then discussed the practical difficulties that would arise when applying criminological and sociological theories, which are well tested in western society, into the case of Chinese society under the rapid social change during the period of economic reform. This means taking into account the background reality of Chinese society and the data availability issues relating to the variables selected. The aims of this chapter are: to draw the theoretical picture that will underpin this research; to set the scene for Chapter 3, which will discuss research design, methodology and data, and to bring forward the research hypotheses which will be tested in the following empirical chapters.

2.1 Western studies into the geography of crime during periods of social change 2.1.1 Theories on the geography of crime developed by western research Traditional criminological theories are mostly concerned with criminality; with crime understood as an expression of the offender’s deviant behaviour. It was not until the early 1980s that the geography of crime emerged as a distinct sub-discipline of geography, and geographers became directly involved with criminological research. Geographers’ contributions to the study of crime have been pioneered by research conducted in western industrialised nations such as America, Britain and Australia (Herbert, 1976; Capone and Nicholls, 1975; Davidson, 1976; Harries, 1974; Plyeet al., 1974; Scott, 1972). Ecological studies of crime events that are undertaken by geographers are also an important part of the field (Schmid, 1960; Lottier, 1938). These studies have examined the spatial patterns of crimes using environmental

16

Chapter Two

approaches, and have analysed the ecological association between crime and a variety of socioeconomic and demographic conditions from the perspective of “environmental criminology” (Brantingham and Brantingham, 1981), which over time has merged with the more general field of the geography of crime. 2.1.1.1 The concept of “Place” Conceptually, geography itself becomes important in criminology because geography is concerned with the scientific study of place. The term “place” is of significance to criminologists, as it is the setting in which crime occurs, where the environment provides opportunities for motivated offenders. It follows that the measurement of certain properties of places can help in understanding crime patterns (Herbert, 1989). Place is more than a position in space; rather it refers to a territory to which people attach values (Herbert, 1989) and a territory that forms a subjective environment, thereby acquiring significance in relation to the occurrence of crime. The geographers’ task is to understand the ways in which “places” have emerged and the characters that they have assumed. In this sense, environmental criminology (Brantingham and Brantingham, 1981) and the geography of crime can be merged through their shared emphasis on “place” in the study of crime. Environmental criminology studies criminal activity and victimisation, and how space impacts on offenders and victims (Bottoms and Wiles, 2002) from four dimensions: the law, the offender, the target, and the place. The place studied within this framework is defined as the location both in time and space where the other three dimensions intersect to produce a criminal event, therefore the characteristics of the place are taken as being important for understanding the occurrence of an offence. 2.1.1.2 Environmental criminology According to the environmental perspective, criminal events must be understood as confluences of offenders, victims or criminal targets, and laws in specific settings at particular times and places (Brantingham and Brantingham, 1991). This is based on the premises that, criminal behaviour is significantly influenced by the environment in which it occurs, and that the environment acts as a trigger for the crime event and shapes its course fundamentally. Crimes concentrate around crime opportunities and other environmental features that facilitate criminal activity. As for the origin of crime at the individual level, Brantingham and Brantingham (1981) proposed that crime is an event or series of actions

Theoretical Background

17

that occur when a motivated offender encounters a suitable target in a situation which activates that readiness potential. The crime opportunities mentioned above are taken as necessary conditions that help to translate criminal inclinations into criminal actions. Tempting opportunities entice motivated offenders into criminal action (Felson and Clarke, 1998). Three theories, referred to as new opportunity theories, are important: routine activity theory, crime pattern theory, and rational choice theory. (1) Routine activity theory (Cohen and Felson, 1979) states that a crime occurs when a motivated offender converges in space and time with a suitable crime target in the absence of a capable guardian against crime. The capable guardian is any person (or surveillance instrument) whose presence or proximity might discourage the occurrence of a crime; the crime target may be a person or an object, encountered by the motivated offender, and endowed with high value, inertia, visibility and accessibility. As Brantingham and Brantingham (1991) described, crimeV tend to occur in predictable locations defined by the intersection of crime opportunities and an offender’s awareness space. (2) Crime pattern theory is a central component of environmental criminology (Felson and Clarke, 1998), telling us about the interactions between residents and their environment in local areas. The “paths” people take in their daily activities, the “nodes” to and from which people travel, and the “edges” of areas where people live, work or enjoy entertainment, are the main concepts of the theory and carry crime opportunities. (3) Rational choice theory (Cornish and Clarke, 1986) focuses on the offender’s decision making and emphasises person-situation interaction. It assumes that offending is purposive and takes into account the benefits, risks, available information, time and effort associated with committing the crime, all of which are influenced by situational and environmental conditions. From this point of view, the theory is closely linked to situational crime prevention, which is explicitly designed to reduce crime opportunities. Situational crime prevention theory (Clarke, 1980) focuses on understanding the precise situational dynamics of specific crimes. This knowledge leads to suggestions on crime prevention through effective urban planning and environmental design, such as better land-use intensity planning (Angel, 1968), more diverse and mixed land uses, more clearly defined demarcation between public and private space, and the clarification of the function of space to promote territoriality and a sense of ownership of space by residents (Jacobs, 1961). This information is also

18

Chapter Two

applied in crime control through problem-oriented policing (Goldstein, 1990) and intelligence-led policing (Ratcliffe, 2003). Decision-makers involved in preventative patrolling by police may undertake a detailed environmental examination in the hope of discovering the conditions that result in a “problem crime area” (Clarke and Eck, 2003), thereby identifying an effective strategy for dealing with the problem. Other theories in line with environmental criminology that are also informative to crime control and prevention include geographic profiling theory (Rossmo and Rombouts, 2011), which is applied to identify the probable location of an offender, and CPTED (crime prevention through environmental design) theory (Jeffery, 1971), which is used as reference for built environment design, because “the proper design and effective use of the built environment can lead to a reduction in the fear and incidence of crime, and an improvement in the quality of life” (Crowe, 2000, p.46). Crime sites in Brantingham’s model (1981) are defined as places where offenders’ awareness space converges with suitable targets, and are restricted by target availability, spatial-temporal factors, and the socioeconomic and demographic conditions of the broader environment, which usually refers to the neighbourhood. High-crime neighbourhoods are typically distinguished by poverty, residential instability, population heterogeneity and family disruption (Sampson and Groves, 1989). Residents of such neighbourhoods tend to show little trust in one another, are unwilling to help their neighbours, and are thought to demonstrate a lack of collective efficacy (Bursik and Grasmick, 1993; Sampson and Lauritsen, 1997). 2.1.1.3 Spatial studies of crime at the ecological level However, as suggested, crime can form very different patterns at different levels of analysis (Brantingha et al., 1976; Galster et al., 2007), and environmental criminologists are less concerned with aggregate crime trends, and more concerned with particular problems occurring in specific places (Wagers et al., 2008). Micro-level analysis, based on environmental criminology, focuses on the effects of specific elements of the immediate environment, such as building type, landscaping and lighting, and security facilities, on specific crimes (Brantingham and Brantingham, 1991). Compared with aggregate ecological analysis, micro-level analyses can be interpreted as the breaking down of areas into smaller constituent parts. However, attention should be paid when using neighbourhood units to undertake statistical comparison between crime and information collected for aggregate units, as it will probably lead to scale-dependent or even misleading results (e.g. the modifiable areal units problem illustrated in Chapter 3).

Theoretical Background

19

Ecological level analyses may be undertaken on national, provincial, urban, census tract or neighbourhood scales (Brantingham and Brantingham, 1991). The geography of crime that is captured is the aggregation of micro-level patterns and processes. Average or typical crime patterns are observed, which combine and filter the patterns and processes associated with individual events (Brantingham and Brantingham, 1976), accumulating these individual experiences to the aggregate level. Amongst studies exploring crime patterns at the aggregate level, especially the spatial distribution of crime, early research explored the relationship with area social disorganisation factors. These factors include poverty, population turnover and racial and ethnic heterogeneity (Shaw and McKay, 1942). The embryo of the spatial perspective on crime emerged from the 19th century cartographic school, followed by the Chicago school’s explanation of high crime occurrence in certain areas by sociological variables in the 1920s and 1930s (Shaw and McKay, 1931). After interest in the spatial perspective waned in the 1930s, it resurfaced in the 1950s and 1960s through the factor analytic school, and in the 1970s researchers began to realise that crime could be better explained and understood by exploring its geographical components through place-based analyses (Eck and Weisburd, 1995; Weisburd, 2002; Weisburd et al., 2004). Since the 1980s, when the geography of crime came of age, there has been a resurgence of interest in analysing the geospatial dimension of crime. The geography of crime took particular interest in the importance of spatial structures, environmental associations, and the special qualities of place (Herbert, 1989). It was rooted in the cartographic reports of officials and statisticians, demonstrating that crime patterns were regionally uneven. The modern geography of crime not only inherits these traditions in a critical way, but also explores a wider range of methodological and conceptual concerns. For example, traditional research focused on regional variations in different types of crime and of justice, while modern research pays more attention to criminal definition (e.g. “hot spots”, referring to geographically bounded spaces associated with heightened victimisation risk and a proportionately greater number of criminal incidents than other similarly sized areas of a city (Eck, 2005)) and the relationships between socioeconomic indicators and crime. It has also been substantiated by new theories focusing on identifying crime patterns, exploring the relationships between crime and environmental or socioeconomic characteristics, and assessing the effectiveness of policing and crime reduction programs. The earliest “spatial studies” of crime were carried out by Quetelet and Guerry in the 1830s (Goodchild et al., 2000). Much research has been

20

Chapter Two

done on the spatial distribution of crime on different geographic scales, from the level of a nation, state or province, down through those of a county, city, neighbourhood, right down to the level of an individual address (Brantingham and Brantingham, 1999; La Vigne and Groff, 2001; Sherman et al., 1989). Spatial theories and research involving spatial measurements provide clear insights into the conventional sociological wisdom of the importance of context (Clinard and Abbott, 1973). Pioneer research was conducted in America, for example by Harries (1974), in Australia by Scott (1972) and in Britain by Herbert (1976). With developments in criminological theories and spatial analysis techniques, the spatial distribution of crime has gained more and more attention recently. For example, the study by Zazoe (2002) in Jedden revealed that crime rates for female foreigners was higher in the northern district, and more frequent police patrolling was recommended; Loukaitou-Sideris and Liggett (2002) explored the relationship between transit crime and corresponding social and physical environmental characteristics near stations, with the aim of evaluating the geography of crime at the neighbourhood level; Baumer et al.’s study (2003) found that residents of areas with higher homicide rates are more likely to support the death penalty; Freisthler (2004) used spatial regression models to examine the relationship between access to alcohol and neighbourhood rates of child maltreatment in three counties in California; Gruenewald et al. (2006) utilised spatial statistical models to test the relationship between alcohol outlets and violent assaults, finding that population and place characteristics are correlated with rates of violent crime across spatial areas; Ceccato and Haining (2007, 2008) used regression modelling and spatial statistics to model spatial variation in homicide rates in Sao Paulo, Brazil, and in acquisitive crime rates in the transitional Baltic States of Estonia, Latvia and Lithuania; Buonanno et al. (2009) exploited provincial-level variations in civic norms and associational networks in Italy to investigate their negative effects on property crime rates, and Cracolici and Uberti (2009) used exploratory spatial data analysis (ESDA) to explore the spatial structure and distribution of four types of crimes in Italian provinces, highlighting those socioeconomic variables associated with different levels of crime. 2.1.1.4 Geography of crime during periods of social change There is specific literature on the geography of crime during periods of great social and economic change. Crime has been acknowledged as an inevitable by-product of social change ever since Emile Durkheim’s classic studies (1893, 1897), which explored the increased levels of crime

Theoretical Background

21

as a result of anomie in societies undergoing profound social transition. Demographic and socioeconomic factors, such as the divorce rate, urbanisation, proportions of the population in younger age groups, the proportion of young males, and the proportion of people who are educated or from different cultural backgrounds, are all expected to affect the crime rate. For example, Shaw and McKay’s social disorganisation theory (1942) argues that neighbourhoods with greater residential instability, lower socioeconomic status, and more ethnic heterogeneity are more likely to experience crime (Sampson et al., 1999) because of the absence of social cohesion and social control (Bursik and Grasmick, 1993; Sampson and Groves, 1989; Sampson et al., 1997). Fleisher (1963, 1966) was the first to work on criminality from an economic view, but the early literature on the economics of crime was pioneered by Becker (1968) and Ehrlich (1973), based on the opportunity cost of committing a crime. Becker (1968) built the first model of criminal choice, and opened the door to a new field of verifying and studying the variables that affect crime from an economic perspective. Theories relate crime closely to economic, socioeconomic and demographic variables, and factors that may affect an individual’s propensity to commit crime, such as poverty, income inequality, cultural and family background, level of education, age, gender and urbanisation (Fajnzylber, 2002; Pyle and Deadman, 1994; Deadman and Pyle, 1997; Hale, 1998; Masih and Masih, 1996). Studies into crime and criminality have traditionally been carried out by disciplines such as sociology and psychology (Georges, 1978). For instance, sociologists highlight the role of socialiation on the basis of factors such as the family, peer groups and educational experiences; economists emphasise indicators of economic activity. It was not until the late 1970s that the spatial dimension to crime began to gain much attention. Regional scientists and geographers drew attention to the importance of spatial variables, including distance and travel costs, in understanding criminal behaviour (Hakim and Rengert, 1981; Hakim and Buck, 1989; Sorenson et al., 1997). They see crime as a result of numerous spatiotemporal interactions of multiple conditions at varied levels, and published a well-developed array of literature probing the consequences of social change on crime, based on levels that vary from the national to the state (provincial), city and neighbourhood (Durkheim, 1897; Messner and Rosenfeld, 1997; Chamlin and Cochran, 1995; Savolainen, 2000; Bernburg, 2002; Kim and Pridemore, 2005; Zhang et al., 2007). GIS is one of the most effective ways of handling and displaying geographically referenced crime data. With the advent and widespread utilisation of powerful GIS software since the 1990s (Wortley and

22

Chapter Two

Mazerolle, 2008), researchers can now place the “pins” on the map rapidly and automatically, query data rapidly to illuminate trends and patterns, and put similar cases or connections together out of stacks of reports (Spencer and Ratcliffe, 2005). If GIS can be integrated into the analysis of relationships between crime and social change, it will be particularly valuable for visualiding their spatial distribution and patterns.

2.1.2 Criminogenic conditions during periods of socioeconomic change 2.1.2.1 Conditions at macro scales In the course of exploring criminogenic conditions during periods of great social change, researchers tend to attribute the increased crime to the anomic conditions induced by poverty, inequality, criminal opportunities, cultural conflicts, weakened social control and social disorganisation (Durkheim, 1897; Cloward and Ohlin, 1960; Kim and Pridemore, 2005). I. Poverty and economic inequality At macro scales (national and regional), poverty and economic inequality are posited in Merton’s strain theory (1938) as the economic conditions that lead to higher crime rates. Income that falls below a level necessary to maintain basic subsistence (Blau and Blau, 1982; Messner, 1982) and the income inequality gaps among social classes both contribute to the development of emotional frustration (referred to as strain), which has direct crime-producing effects. This is part of a long tradition dating back to work by western scholars in the 19th century on the links between poverty and crime (Marx, 1867; Engels, 1845), as well as anomie and crime (Durkheim, 1893). Criminological studies on criminal opportunities (Kick and LaFree, 1985) and social disorganisation (Braithwaite, 1979; Cao and Maume, 1993) also emphasise the relationship between inequality and crime (Blau and Blau, 1982). Hsieh and Pugh (1993) suggest that 97% of empirical studies support a positive relationship between crime rates and income inequality. Fajnzylber et al. (2002) suggest positive correlations between robbery, homicide and income inequality in their cross-national study. Kim and Pridemore (2005) also found that high levels of poverty and increasing inequality played a role in the increase in Russian homicide rates.

Theoretical Background

23

II. Weakened social organisation The roots of social inequality can be traced back to changes in social structure (Merton, 1938), the weakening of social control (Hirschi, 2002) and social disorganisation (Shaw and McKay, 1942). Merton (1938) argues that societies share similar cultural values, but people’s access to the approved “means” of achieving accepted goals are unequal. When individuals do not have the institutional means to reach culturally prescribed goals, the inequality among them worsens, and deviant behaviours, including crime, ensue. Crime rates among lower class individuals (e.g. laid-off workers, immigrants, and people living a poorer life) tend to be higher. Hirschi (1969) suggests that criminality is expected in the absence of social control. Rapid social change weakens social control over individuals’ behaviours, and further frees them to engage in deviant behaviour. Shaw and McKay’s social disorganisation theory (1942) attributes higher crime rates to a breakdown in social bonds through the processes of value and norm conflicts. The factors eroding social bonds could be urbanisation, high levels of poverty, ethnic or racial heterogeneity, high levels of residential mobility, family disruption, and even a high level of population density (Paulson and Robinson, 2004). Derived from the above, conditions used in empirical studies usually include urbanisation, unemployment and family cohesion, migration rate, ethnic or racial heterogeneity and population composition (age and gender). Take urbanisation, unemployment and migration as examples. During the course of urbanisation, rural residents flock into cities to seek opportunities to be successful. This not only worsens the unemployment situation, it also brings into cities traditional rural violence, both of which cause crime rates to increase. In the meantime, geographically unbalanced affluence and the availability of goods also draw migrants to urban areas, with the result that the desire for unattainable goods and the impersonal nature of city life inspire criminal thoughts. According to Shelley (1981b), urbanisation changes the dominant crime from violent crime to property crime, while increasing the contribution of juveniles and females to crime levels. Studies have suggested that urbanisation brings about an irreversible increase in recorded crime rates (LaFree, 1999; LaFree and Kick, 1986; Neapolitan, 1997). Clinard and Abbott (1973) see increased crime in developing countries as an inevitable developmental stage which is the consequence of rapid social change and a replica of that in developed countries during the industrial revolution (Bennett, 1991; Cao and Maume, 1993; Heiland et al., 1992; Shelley, 1981b). Krohn (1976) points out that urbanisation and unemployment are fairly effective in

24

Chapter Two

explaining property crime, in that it is predicted to increase in accordance with the level of social change. For another instance, crime research shows that minority ethnic groups, males, and younger age cohorts are more likely to commit crimes than other demographic groups (Senna and Siegel, 1993). III. Criminal opportunities Besides the aforementioned social conditions, Kick and LaFree (1985) also ascribe the shift in crime during social change to the increased opportunity to commit offences, drawing on related theories (Cloward and Ohlin, 1960; Cohen and Felson, 1979). As Cloward and Ohlin (1960) assert, higher crime rates correlate with increased opportunities for crime created by changes in society’s structures, institutions, culture, and social processes. These conditions weaken traditional family ties, change lifestyles, and increase the opportunities for deviant behaviours (Gyamfi, 1998). They also broaden human choices and give people enormous opportunities to commit crimes (Felson, 1983). The increased prosperity and availability of goods or targets also contribute to the concentration of crime in urban areas. On the other hand, the urban growth and migration brought about by social change are also thought to enhance the opportunities. As routine activity theory (Cohen and Felson, 1979) suggests, crime rate trends are an indicator of the dispersion of activities away from safer settings; it can increase if there are more targets or fewer guardians present. The more that lightweight durable goods are available, and the more time people spend among strangers away from their homes, the greater the risk of being victimised. This theory emphasises changes in technology and organisation on a societal scale, and these structural changes are thought to have dramatic impacts on society in increasing crime opportunities. IV. Cultural conflicts Cloward and Ohlin (1960) point out the importance of cultural conflicts in explaining crime. “Cultural norms may promote or condone violence in certain situations, the attitudes and values of individuals within cultures that have higher rates of crime or violence should be distinguishable from those with lower rates” (Paulson and Robinson, 2004, p.21). They suggest that crime is caused by differences that exist in group values, norms, and attitudes found within mainstream society. This is coherent with Sellin’s (1938) culture conflict theory. Sellin points out that

Theoretical Background

25

conflicts between two different cultures might cause clashes when they intersect with each other, as in the case of new migrants in a host society. Sutherland (1937) also asserts that cultural conflicts resulting from social and economic change generate a pervasive individualism and other conditions conducive to criminality. Messner and Rosenfeld (1997) explore the correlation between the increase in crime and cultural conflict within the context of institutional anomie theory. They argue that capitalist culture encourages an anomic cultural environment by promoting economic success at the expense of pro-social non-economic institutions such as family, education, polity, and religion. Studies testing this view, like Kim’s (2005) study in Russia, have drawn the conclusion that, as former communist countries move toward a free market, it is likely that their citizens will start to adopt capitalist ideologies, including an emphasis on individual economic success at the expense of non-economic social institutions (Merton, 1938; Polanyi, 2001). V. Climate conditions Climate and weather are also included as regional factors. Individuals may become more irritable in hot weather due to physical discomfort and hence commit criminal acts (e.g. domestic violence). Warm weather facilitates acquisitive criminal activity too, as windows and doors are less likely to be locked (Hoch, 1974). 2.1.2.2 Conditions on micro scales At the micro (neighbourhood) scale, theories in line with environmental criminology (Brantingham and Brantingham, 1981) concentrate on the offender’s motivation to commit crimes, and tend to explain criminogenic conditions in terms of neighbourhood residential mobility, population heterogeneity, mutual surveillance and demographic composition (age and gender). In light of environmental criminology, criminal events must be understood as confluences of offenders, victims or criminal targets, and laws in specific settings at particular times and places (Brantingham and Brantingham, 1991), in which the environment plays a fundamental role in stimulating crime and shaping its course. Theories relating to offenders and their motives, such as strain theory, rational choice theory, routine activity theory (Cohen and Felson, 1979) and lifestyle exposure theory (Hindelang et al., 1978), have been used in empirical studies. Strain theory (Merton, 1938) argues that crime is caused by a disjunction between one’s goals and one’s means of achieving those goals. Rational choice theory argues that an offender will choose specific

26

Chapter Two

targets through a decision-making process. He/she will weigh the effort and risk against potential payoffs, such as the value and accessibility of the target, the ease of escape, as well as the mobility of objects. According to routine activity theory, when people who are inclined to commit a crime converge with a suitable crime target, in the absence of a capable guardian, there is an increased probability of crime. It applies on both micro and macro scales, and it is claimed that social change disperses people over space and time, away from families, household settings, property, guardians, and handlers. It will further increase the probability that a motivated offender and a suitable target will converge without the presence of capable guardians. The places where convergence takes place may be red light districts, bar districts, entertainment districts or large shopping areas. The lifestyle exposure theory assumes a relationship between activity patterns and the exposure to dangerous places at dangerous times, which links to the risk of criminal victimisation (Hindelang et al., 1978). Western studies have provided ample support for the lifestyle exposure and routine activities theories in developed countries (e.g., Miethe and Meier, 1994), most of which focuses on the risk of victimisation of serious crimes, such as assault, burglary or robbery. On the neighbourhood scale, the main criminogenic conditions include various forms of social and economic inequality, social disorganisation among neighbours and group demographics. Economic inequality, represented by a high level of poverty (Shaw and McKay, 1942), will lead to relative privation (Merton, 1957) and further result in delinquent behaviours. Social disorganisation identifies the features of neighbourhoods that affect the risk of victimisation, which is usually caused by subcultural conflicts (Blau, 1982). The delinquent subcultures theory developed in the 1950s is in line with this viewpoint (Cohen, 1955, 1958; Cloward and Ohlin, 1960; Miller, 1958). It focuses on the impact of delinquent subcultures on working-class neighbourhoods, the relationship of delinquents to the world around them, and the absence of legitimate opportunities for them. In contrast to social disorganisation, collective efficacy (Taylor, 2001) describes strong social cohesion among neighbours in the form of willingness to intervene when public order is threatened (Sampson et al., 1997). It emphasises the level and density of social ties in the neighbourhood context, as well as the frequency of the intervening interaction among neighbours. It encompasses the degree of mutual trust, shared expectations for informal social control and social cohesion. Drawing on this, residential instability (or mobility) (Kornhauser, 1978) is often used to measure social status, and highly mobile communities are assumed to have higher crime rates, regardless of the socioeconomic status

Theoretical Background

27

of the residents (Bursik and Webb, 1982). As for demographic composition, ethnic heterogeneity (Sampson and Groves, 1989) and the proportion of the most crime-prone cohorts present in a neighbourhood are the most commonly considered factors. Geographical conditions should also be paid attention to at the neighbourhood level, with reference to crime pattern theory. This theory explains how and why crime is distributed across places (Eck and Weisburd, 1995), the actual patterns of crime, the spatial-temporal patterns of offenders’ movement, as well as the types of environments where crime occurs. The locations most likely to experience concentrations of crime are near the routine activity nodes of a city, and usually around crime opportunity sites and facilitating environments (Wortley and Mazerolle, 2008). For example, highway improvements are thought to decrease the costs of shipping stolen property and increase the opportunities for effective law enforcement measures (Friedman et al., 1989). Therefore local land uses and travel path networks should be taken into consideration at the neighbourhood level, because they contribute a lot to the spatialtemporal patterns in crime levels. The majority of the above theories attempt to interpret changes in crime rates in terms of changes in economic inequality, social instability, cultural conflict, and demographic composition. Meanwhile, other theories, like institutional anomie theory (Messner and Rosenfeld, 1997) and social capital theory (Coleman, 1988; Sampson, 1999), place emphasis on the mediating role of social cohesion and the strength of cultural values that do not equate “success” with “money”. These theories work together to provide a basis for examining changing crime rates during periods of social change on various scales.

2.1.3 Crime-mediating conditions during periods of socioeconomic change While some aspects of social change are responsible for elevated crime rates, this association may be mediated by the strength of non-economic social institutions (Bernburg, 2002), including family cohesion, education levels, trust in polity, civic norms and social control, and the use of security precautions (e.g. not leaving the home unoccupied, not participating in dangerous activities away from home, not carrying expensive goods, etc.) (Miethe and McDowall, 1993). Messner and Rosenfeld’s (1997) institutional anomie theory argues that crime flourishes in societies where the institutional balance is skewed toward the economy, but is reduced where non-economic organisations (e.g. family,

28

Chapter Two

education, and polity) are capable of offsetting criminogenic influences. Meanwhile, social capital (Putnam, 1993) can affect individual delinquency choices significantly through the presence of civic norms and associational networks. I. Institutional anomie theory (IAT) Institutional anomie theory (IAT) (Messner and Rosenfeld, 1997) was generated in the context of economy-dominated social change (e.g. the American Dream), where individualism and the fetishism of money are dominant cultural values. Capitalist culture promotes anomic pressures for economic success at the expense of non-economic institutions such as family, education, polity, and religion, weakening institutional control. It is argued that strong families and the accompanying social cohesion can inhibit crime (Sampson et al., 1997), so neighbourhoods with more “social ties” and greater “participation in organisations” experience less crime (Sampson and Groves, 1989). Families can mediate anomic pressures by providing emotional support and social bonds (Messner and Rosenfeld, 1997). Education can lessen the impacts of social change on crime by monitoring and supervising members’ behaviours, thereby exercising social control, while creating environments in which people are strongly committed to their education and aspirations (LaFree, 1998). It is claimed that education may reduce crime through its “civilising” effect (Fajnzylber et al., 2002). Trust in political institutions reflects the legitimacy of these institutions, and is closely related to social control efforts (LaFree, 1998). Among the studies testing IAT, conditions commonly used to measure these institutions are marriage and single parent households, rates of enrolment in college and voter turnout for presidential elections (Messner and Rosenfeld, 1997; Savolainen, 2000; Kim and Pridemore, 2005). Their findings provide some support for the hypothesis derived from IAT, i.e. that the strength of social institutions is negatively related to crimes like homicide and property crimes. II. Social capital theory Besides the social institutions mentioned above, social capital is also predicted to be effective in reducing crime (Rosenfeld et al., 2001), and includes the dimensions of civic norms, social trust and associational networks (Buonanno et al., 2009). It attaches guilt and shame to criminal behaviour through civic norms, emphasises formal and informal social control, provides resources for individual goal attainment, and raises the

Theoretical Background

29

probability of detection by associational networks. Empirical studies have directly investigated the links between levels of social capital and crime rates (Rosenfeld et al., 2001; Messner et al., 2004). When it comes to the indicators of a community’s social capital, voter turnout and voluntary donations to charity are usually used as a reflection of participation in civic life (Putnam, 1993; Rosenfeld et al., 2001). Another indicator of social capital is voluntary blood donation (a so-called altruistic behaviour), which indicates how much people care for each other (Guiso et al., 2004). Trust indicators, population heterogeneity and family structure are also proxies for social capital, as they can measure general social relations and acquaintances. In this sense, social capital is more applicable on neighbourhood scales, whilst the aforementioned social institutions are relevant on macro scales. This is also consistent with collective efficacy (Taylor, 2001), which is used to describe the strong social cohesion among neighbours with a willingness to intervene when public order is threatened (Sampson et al., 1997). It encompasses the degree of mutual trust, shared expectations for informal social control and social cohesion. III. Collective efficacy Bellair (1997) found that neighbourhoods with more “social interaction” (e.g. visiting one’s neighbours) have lower levels of disorder. Warner and Rountree (1997) found that greater levels of “neighbouring activities” are related to lower assault rates, but they found no evidence that these mediate the relationship between neighbourhood structural characteristics and assault rates. Sampson et al. (1997) found that a combined measure of cohesion, mutual trust, and the expectation of intervention by others (which they labelled “collective efficacy”) reduces violent crime rates, and this mediates in part the effect of neighbourhood structural characteristics. Lastly, Markowitz et al. (2001) found significant relationships between neighbourhood structural characteristics and disorder, which are mediated by “cohesion” and “social control”, which could be measured by respondents’ intervention behaviour on behalf of others, their participation in neighbourhood organisations, their attachments or social ties, their trust in others, or “collective efficacy” (Sampson et al., 1997). Certain positive economic conditions, such as improvements in living standards and the welfare system, should not be neglected at this point either, because negative relationships can be expected between them and

30

Chapter Two

crime. The Engel coefficient 2 , which is defined as the proportion of income spent on food, is normally used as an indicator for living standards, and a high Engel coefficient indicates a relatively low living standard. As income per capita increases or general welfare improves, we can expect wealth, especially in urban areas, to increase, thus the incentive to offend will be reduced. However, the data for these economic indicators (Klein and Moore, 1983) at national and regional levels (Deadman, 2003) changes little over time, and at micro-geographic levels, they are unavailable from official statistical sources. To sum up, the criminogenic and mediating conditions affecting crime rates during social change comprise a complicated set of indicators; they interact with each other in complex ways to influence illegal behaviours, and vary according to the real situations in target areas. Certain explanatory factors, such as the impacts of cultural transition and social cohesion, are difficult to quantify and can only be fully explored by other methods. Based on the above information, a conceptual framework has been proposed (see Fig. 3-6) as a summary of the position taken in this research, and is also instructive for the latter part of this chapter and Chapter 3.

2.2 Practical differences in Chinese society It should be noted that theories in this field were mostly developed in western industrialised countries (Farrington, 1999), using data geographically biased towards western developed countries (Neapolitan, 1997). For example, comparative criminological studies (Bennett, 2004) have mostly been based on cross-national data (Bennett, 1991), and analysed the relationships between crime rates and social change (Lafree, 1999; Neapolitan, 1997) among nations, or provinces (states) in developed countries. China, in this sense, has been largely neglected, which renders the theories incomplete. So this research provides an opportunity to test these theories in a different society.

  2

The Engel coefficient is the ratio of annual food expenditure to total expenditure for a family or a country, following Engel’s Law (Engel, E., 1896), and is used to reflect the living standard of the target. It is expressed as:

E=

Food expenditures *100% Total expenditures

with a value between 0 to 100%. A high Engel coefficient indicates a poor family or country, while a low Engel coefficient is an indicator of a higher living standard.

Theoretical Background

31

2.2.1 The reality of Chinese society In the last 30 years, during the transition from a planned economy towards a market economy, China has experienced tremendous social, political, and economic changes, such as the breakdown of traditional cultural values, high levels of income inequality and migration, and increasing unemployment. At the same time, China’s crime levels have also been increasing dramatically. This increase in crime, combined with the social changes taking place in China since the beginning of the economic reform, has attracted the attention of many scholars. It provides a great opportunity for researchers to advance criminological knowledge and understanding in the context of China. Academic research into crime in China began in the late 1920s, after the revolution in 1911 (Zhang, 2002), but it was then interrupted by the civil war and the Japanese invasion in the 1940s. After a short period of recovery after the liberation in 1949, it began to disappear gradually in 1958 due to the spread of political struggle which continued until 1978. Since Reform and Opening-up in 1979, it has been revitalised in response to the rise in criminal behaviours associated with social and economic change. Government administered research was rare, due to the difficulty in interpreting the large volume of underreported data. The most famous national scale crime investigation was launched in 1985, entitled “Research on the contemporary criminal issues in China”, and had three phases. In the first phase, it sampled 123,400,000 people from eight provinces; the second phase, in 1987, covered 178,500,000 people in 13 provinces, and the last phase, in 1988, was conducted in eight provinces, covering 151,100,000 people. Findings from these surveys exposed significant underreporting of cases in the official record. It was estimated that recorded cases accounted for 32.6% of all crime in 1985, 19.42% in 1987 and 30.64% in 1988, which meant that 70% to 80% of criminal events went unrecorded. Studies exploring the relationships between crime rates and possible criminogenic factors have been mainly descriptive. Many scholars of sociology and criminology have done research into Chinese criminology, both in terms of theory and crime prevention strategies, such as work on the causes and prevention of crime (Zhou, 1995; Zhou, 1991), as well as the typology of crime. The majority of this research has focused on juvenile delinquency, drug-related crime, violent crime, and the relationships between crime rates and inequality, changing cultural beliefs and norms, the disruption of traditional social control, decreased social integration, massive migration into urban areas, and the altered age structure of the population (Deng and Cordilia, 1999). Taking the age and

32

Chapter Two

gender factors as examples, in China, it is argued that people aged between 14 and 25 are the most crime-prone cohort, and the majority among them are male.3 Females make up only 4% of China’s criminals, and the crimeprone age group for women is between 26 and 50. The type of crime committed also varies by gender. For instance, males tend to commit theft, robbery, assault and homicide, while females are inclined to commit theft, robbery and fraud (Cong, 2008). Some sociologists have also drawn attention to the impact of urbanisation on crime, such as Kong (2005a) who carried out a case study in Zhejiang Province, using local statistical data to illustrate the relationship between urbanisation and crime posited by Shelley (1981a). In his work, the annual proportion of the urbanised population, GDP per capita, and government expenditure per capita were indicators of urbanisation. In order to better understand the local urbanisation level, he also used Engel’s Coefficient, the unemployment rate, revenue disparity, the expenditure gap, the proportion of educated people, number of college staff, number of doctors, and number of public libraries and cultural centres in his study. Their clear correlations with crime were presented. Among these empirical studies, the identification of factors associated with crime has focused on the incongruity between market economy developments and government regulation, which results in urbanised communities becoming important breeding grounds for crime (Popenoe, 1983), while the most prominent crime-producing problems in urban areas are migration and inequality. I. Migration and crime Resulting from reduced migration controls and increasing economic liberalisation in China , millions of peasants have migrated to cities (Liu et al., 2001). Crimes committed by migrants (the so-called “floating population”) have become a major concern for social stability. It is estimated that China’s migrant population is more than 140 million, and they are a major source of criminals due to weakened social control. It is said that more than 50% of criminal cases in cities, 70% to 80% of which are theft cases, are committed by migrants (Chen and Yu, 1993). The proportion of criminal suspects who are members of the floating population in Beijing has increased from 18.5% in 1986 to 22.5% in 1990,   3

14 is the age of criminal responsibility (Chinese Criminal Law), and 25 is the upper age for “juvenile crime”. People aged between 14 and 25 are the most crimeprone cohort, in that they have accounted for more than 70% of offenders since the mid-1980s.

Theoretical Background

33

37.6% in 1992, 43% in 1993, and 50% in 1994; this number has even reached 70% in the areas on the fringes of the city (Jiao, 1994). These increases were not only seen in the seriousness of offences, but also in the diversity of crime types, especially in more developed urban areas. The Legal Daily (1995) once did a comparison of the proportions of crimes committed by migrants in 16 big cities in China in 1994 (Table 2-1). It found that migrants (who are not registered as permanent residents) in more developed cities, such as in the cities in Guangdong Province (Shenzhen (97%), Dongguan (85.4%) and Guangzhou (70%)), are much more likely to commit crimes. Table 2-1 Crimes committed by the migrant population (%) in selected cities in 1994 City Beijing Shanghai Tianjin Guangzhou Ningbo Xi’an Chongqing Urumqi

Percentage (%) 46.2 53.6 30 69.2-70 41.2 53 53.9 43.2

City Shenzhen Dongguan Hangzhou Wenzhou Guiyang Jinan Harbin Nanjing

Percentage (%) 97 85.4 50 48.6 30 28 19.6 47

Source: Legal Daily 1995.

II. Inequality and crime Another important factor is the greater inequality brought about by economic reform. Comparative research linked western ideology or capitalist economics to the sense of relative deprivation (Ma, 1999), the normlessness and the lack of social support in cities, through rapidly widening income inequality. Income inequality is usually measured by the Gini coefficient. 4 Sociologists (e.g. Ravallion and Chen, 2007) have

  4

The Gini coefficient (Gini, C., 1912) is the most widely used measure of inequality of income, and determines the extent to which the distribution of income among groups deviates from a perfectly equal distribution. Statistically, when the population is grouped by income in ascending order, there are several functions used in Chinese studies to calculate the Gini coefficient, and in this study, we apply Zhang’s (2007) method as below:

34

Chapter Two

compared the crime rate and the Gini index from 1981 to 2001 in China (Fig. 2-1). They found similar periodic features in both series, as well as a correlation between them. However, further in-depth research still needs to be undertaken in which the influence of other factors is controlled.

Fig. 2-1 The Gini Index and Crime rates in China from 1981 to 2001

At the neighbourhood level, along with economic development, Chinese households’ economic conditions and living standards have improved, thus there are more valuables attractive to offenders. At the same time, housing growth has stimulated residential mobility; the population is more heterogeneous, as large numbers of rural migrants settle in the cities; poorer neighbourhoods are juxtaposed with affluent neighbourhoods (Liu, 2008). All these are taken as criminogenic factors. However, empirical analyses on a neighbourhood scale in China are for the most part studied in law schools, and the focus of inquiry has been

    n

G=1-¦ Pi u  2Qi  Wi i=1

where Pi is the population ratio of group i, and

is group i’s ratio of income to the i

Qi

¦W

k

k 1 whole population income, can be measured by the equation , in that it is the accumulated ratio of income for group i. A low Gini coefficient indicates a more equal distribution, with 0 corresponding to perfect equality, while a higher Gini coefficient indicates a more unequal distribution, with 1 corresponding to maximal inequality. According to the UN, 0.4 is the international alert line, and values above 0.4 indicate great income inequality between groups.

Theoretical Background

35

theoretical and philosophical rather than based on empirical analyses of crime (Zhou and Cong, 2001). Based upon the above discussion of literature on social change and crime, especially that carried out in China, we find that there are still some issues to be explored further: 1) The relationships between social change and crime rates at different geographic scales. 2) Spatial cross-sectional comparison among provinces and cities, to indicate geographical differences in these relationships. 3) The extent to which different factors are significantly criminogenic or mediating at the ecological level. 4) Utilising relevant theories developed in the West in the context of China and, if necessary, adapting them to China’s particular situation.

2.2.2 Variable selection recognising the realities of Chinese society Drawing upon the above illustrations of social change and crime, it is easy to arrive at the conclusion that crime is a barometer of broader social conditions, in that a society that has experienced profound social change over a short period of time, such as rapid industrialisation or urbanisation, may experience particularly high rates of crime, above the “normal” (pre change) level. In other words, crime change is a reflection of rapidly changing contextual conditions and is inextricably connected with the society in which it exists. In fact, there are various socioeconomic conditions at various scales conducive to the precipitous increase of crime during periods of social change, and crime change in a particular place can be ascribed to a complex web of interacting factors. Although no aspects of society can be ignored in probing the phenomenon of crime (Shelley, 1981a), it is impossible to pinpoint accurately all the reasons for crime. For example, Quetelet (1842) and Guetty (1833) studied crime in terms of a variety of social and demographic factors, including age, sex, race, education, occupation, economics, as well as meteorological conditions (Shelley, 1981b). Other studies by criminologists have correlated crime rates with variables covering a comprehensive range of social aspects, e.g. population density, divorce rate, and urbanisation. Some of those factors widely used in western studies can be transferred directly into Chinese studies, but there are still some other factors, such as ethnic heterogeneity, voter turnout, blood donation rate, etc., that do not fit into the reality of Chinese society, because they are either not measured or not relevant. Considering China’s social reality and data availability, and based on my earlier review of the literature, the variables utilised in this research

36

Chapter Two

and their expected associations with the crime rate are categorised into several different dimensions and listed in Table 2-2. (1) The extensive economic development which was the initial goal of economic reform has been reflected by the variables in the “economy” dimension in Table 2-2. These variables are mostly assumed to be criminogenic because they provide more valuable targets. For example, the “GDP per capita” variable is normally used to indicate local economic strength in China, at either a national, provincial or city level. The higher value it has, the more developed the area is. (2) Demographic change is another prominent dimension of the social change brought by economic reform, and variables such as population density, urbanisation level (officially measured by the proportion of urbanised population in the China Statistical Yearbooks), divorce rate etc. are commonly used variables in related research. These variables are seen as a sign of eroding social bonds and social stability, so they are also supposed to have criminogenic effects on the change in China’s crime rates. For instance, a high divorce rate is an indicator of disruption to family cohesion, hence lessening the crime deterring effect arising from family bonds and relevant institutions. (3) Social disorganisation theory has emphasised the criminogenic influences stemming from conditions eroding social stability, such as the Gini coefficient, which measures social inequality, and unemployment levels. Related variables accessible in Chinese society are included in the table in the “social stability” dimension. (4) The “social harmony” dimension refers to the widely used Chinese term “harmony”; it is a mix of variables related to official crime prevention efficiency and levels of poverty. The latter is thought to be a factor that erodes social harmony by the public in China. (5) The “mobility” dimension focuses on the development of transportation, which is criminogenic because it can make escaping the scene of a crime easier, and the migration rate, which can increase social “disturbance”. (6) The “social improvement and citizens’ welfare” dimension is comprised of variables measuring improvements in residents’ daily life, such as the Engel index and their access to healthcare services (doctors per 10,000 people). (7) Cultural institutions, such as educational attainment levels, and cultural conflicts since economic reform (foreign capital utilised) are included in the “culture” dimension. (8) The “geography” dimension takes into account the target area’s climate and its influence on crime, as well as neighbouring areas’ crime spillover effects.

Theoretical Background

37

Table 2-2 Crime-related independent variables on macro scales Dimension

Variable

Demography

Population density Urbanisation level Natural population growth rate Divorce rate Proportion of GDP from industry Proportion of GDP from services GDP per capita Value of exports per capita (Communication with foreign cultures) Unemployment rate Gini coefficients Ratio of urban rural income Average income per capita Expenditure on civil affairs per capita Prosecution ratio Proportion of people in poverty Investment in public fixed assets per capita Ratio of administrative expenditure to public expenditure Length of railways per capita Length of highways per capita Passenger traffic per capita Net migration rate General consumer price index Engel urban index Engel rural index Number of doctors per 10,000 persons

Economy

Social stability Social harmony

Mobility

Social Improvement & Citizen Welfare

Expected association with crime rate (+) (+) (+) (+) (+) (-) (+) (+)

(+) (+) (+) (+) (-) (-) (+) (-) (-)

(+) (+) (+) (+) (+) (+) (+) (-)

38

Culture

Geography

Chapter Two

Number of public libraries per capita Proportion of people whose highest education attainment level is junior middle school Proportion of students enrolled in institutions for higher education Number of municipal cultural centres per capita Value of foreign capital utilised (foreign cultural conflicts) Annual average temperature Annual average precipitation Crime “spillover” from surrounding areas

(-) (-)

(-)

(-) (+)

(+) (+) (+)

Note: (+) indicates that an increase (decrease) in the level of the variable leads to an increase (decrease) in the crime rate, and (-) indicates that an increase (decrease) in the level of the variable leads to a decrease (increase) in the crime rate.

These variables are selected as the demographic and socioeconomic indicators of change and tested in the macro scale crime models. By contrast, analysis on a micro scale focuses on more detailed conditions instead of those mentioned above. At the neighbourhood scale, the measure for crime is residents’ perceptions of neighbourhood safety (fear of crime), due to the unavailability of official crime data, and indicators for the neighbourhood environment pay more attention to measures of neighbourliness and individual lifestyle information. For example, it has been found that crime messages can spread fast in tight-knit communities, so knowing local victims appears to affect residents’ fear of crime and their views about the environment (Tyler, 1984; Skogan and Maxfield, 1981; Greenberg et al., 1982). It has also been found that crime is encouraged by low levels of surveillance in public places and reduced by people’s willingness to challenge strangers (Shotland and Goodstein, 1984), and by the availability of social support (Sundeen and Mathieu, 1976). Measures of residents’ demographic characteristics derived from empirical studies are also included in neighbourhood scale analysis. For instance, women and the elderly are thought to have higher levels of fear of crime because of their selfperceived vulnerability to victimisation. Marriage may establish informal

Theoretical Background

39

social control (Sampson and Laub, 1990) and limit the opportunities for becoming a target for crime (Laub and Sampson, 2003; Giordano et al., 2002). Lifestyle conflicts between different groups can also be translated into concerns about crime (Newman, 1972; Newman and Franck, 1980). Detailed explanations on these measures are dealt with in Chapters 3 and 7.

2.3 Research aims and hypotheses This study is undertaken against the background of rapid social change in China since 1978 in terms of its demography, economy and culture. These changes are believed to have had significant impacts on crime levels and types. The whole project aims to examine crime in China since 1978 from a geographical, rather than sociological or psychological, perspective, and it will apply GIS and spatial data analysis methods to address crime issues. This involves investigating the changes in crime levels and types in China since 1978 on different geographical scales (national, provincial, city and neighbourhood); developing an understanding of the relationships between crime rates and various demographic, social, economic and other independent variables in different geographical contexts. Analyses of the relationship between crime and demographic and socioeconomic variables are actually about identifying crime inducing and deterrent factors during a period of rapid economic development, and finally bringing forward pertinent suggestions on crime control to policy makers and government agencies based on well-founded analysis. In view of the previously discussed literature, it is hypothesised that the effects of changing demographic and socioeconomic conditions on crime will vary with their characteristics and locations, as well as with geographical scale. This dissertation addresses the following hypotheses relevant to the literature on crime during the period of social change in China since the start of economic reform in 1978: 1) During the economic reform, the increase in acquisitive crimes (e.g. larceny, theft, robbery and fraud) has been much more dramatic than that in expressive crimes (e.g. homicide and assault); 2) The widening income inequality gaps during the economic reforms have a positive relationship with acquisitive crimes; 3) Increasing migration and urbanisation across the whole country has a positive relationship with crime; 4) Non-economic institutions, such as family support, education and polity institutions have negative relationships with crime, and they also condition the effects of other criminogenic factors; 5) Along with economic development, improvements in people’s living

40

Chapter Two

standards and the welfare system are thought to be deterrents to crime; 6) Relationships between the factors listed and crimes might differ according to geographical location and target characteristics; 7) There is a scale effect in the relationships between the proposed factors and crimes; in other words, the relationships between specific conditions and crimes will vary according to spatial and temporal scales. Besides the above hypotheses, detailed hypotheses for each scale will be proposed in the following chapters.

CHAPTER THREE DATA AND METHODOLOGY

This research is designed around four geographical scales, and involves the analysis of crime data in terms of a range of other social as well as economic, cultural, political and demographic data (Fig. 3-1). The scales of the analyses and the methods used are as follows: 1) National scale: time-series regression analysis using secondary data; 2) Provincial scale: time-series regression analysis, spatial cross-sectional and spacetime series regression analysis using secondary data; 3) City scale (in Guangdong Province): time-series regression analysis, spatial crosssectional and space-time series regression analysis using secondary data, and 4) Neighbourhood scale (in the city of Shenzhen, Guangdong Province): qualitative analysis of primary data from interviews and focus groups, and regression analysis of primary individual level data from a questionnaire survey.

Fig. 3-1 Scalar structure for research design

42

Chapter Three

3.1 Study area description China is located in eastern Asia between latitudes 4° and 53°30ƍN and longitudes 73°40ƍ and 135°05ƍE. It covers an area of 9.6 million km2 and has a population of around 1.4 billion. There are 23 provinces, five autonomous districts, four municipalities and two Special Administrative Regions, with varying crime levels and patterns. Guangdong Province was selected for city scale analysis because it has been at the forefront of China’s post-1979 economic reform and development, and has experienced high levels of migration and relatively high crime rates compared to other provinces. Guangdong is in the southern part of China, neighbouring Hong Kong and Macau (Fig. 3-2), with an area of 178,000 km2 and more than 95 million people. There are 21 cities in Guangdong Province, including Guangzhou, Shenzhen, Foshan and Dongguan.

Fig. 3-2 Map of Guangdong Province in China

Data and Methodology

43

During the past 30 years, millions of migrants have gathered in Guangdong from rural or other undeveloped areas. According to the national population census in 2005, the migrant (floating) population in Guangdong Province is more than 20 million, mainly concentrated in Guangzhou, Shenzhen and Dongguan. The economic and psychological contrasts in this region are thought to have contributed a lot to the high crime rates in this province, especially in cities within the Pearl River Delta. At present, the floating population in Guangdong Province accounts for approximately 25% of the total population, while the crimes committed by them account for more than 60% of the total, especially in the area of the Pearl River Delta, where 80% of crimes are committed by the floating population (21st Century Business Herald, 2005). The city scale study will focus on the major cities within the Pearl River Delta, which are Guangzhou, Shenzhen, Dongguan and Foshan, because official statistical data for Guangdong Province are only available for these four cities. An intensive study on a neighbourhood scale will be carried out in the city of Shenzhen. Shenzhen makes an interesting case study, as it was the first city to implement the economic Reform and Opening-up policy. It epitomises the social upheavals associated with China’s post-1979 transformation: most of the policies associated with economic reform were trialled in Shenzhen; the creative measures developed in Shenzhen were followed by other cities, and the extent and pace of Shenzhen’s development has become an example for other cities to learn from. The social problems that emerged are particularly prominent, like the large influx of migrants (it has been dubbed the “migrants’ city” (Tian, 2003)) and the extreme inequality. As for its crime rates, Shenzhen is one of the most crime-ridden cities in China. Another consideration in selecting Shenzhen as a study area is accessibility of data. Both qualitative and quantitative data are relatively easy to obtain through contact with officials who are familiar with the local situation.

44

Chapter Three

Fig. 3-3 An administrative map of Shenzhen

Geographically, Shenzhen is in the southern part of Guangdong Province, next to Hong Kong, with an area of 2,020 km2. There are six districts in Shenzhen (Fig. 3-3), 51 towns (the so-called city-street offices) and 612 neighbourhoods. It is reported that in 2008 the floating population in Shenzhen was more than 12 million, in contrast to the two million registered population. With a population density of 4,239 people per km2, Shenzhen is thought to be one of the most densely populated cities. This exerts great pressure on, for example, local land-use, traffic capacity, competition for employment, and income inequality (which is measured by the Gini index, and is reported to be the highest in China at 0.56 in 2006 (Xue et al., 2008)). Moreover, in part because of Shenzhen’s coastal location, it also experiences intense cultural conflicts, which is seen as a contributing factor in explaining the crime rates and neighbourhood dangers discussed in Chapter 2. Take robbery as an example; according to reports in the Southern Metropolitan News, there were 18,000 robberies in 2004 in the Bao’an district of Shenzhen, which has a population of 1.99 million. By contrast, there were only 2,182 reported robberies in Shanghai, which has a population of 18 million, in the same year.

Data and Methodology

45

Fig. 3-4 Locations of representative neighbourhoods

On a neighbourhood scale, pilot studies were carried out in 16 neighbourhoods (the red and yellow dots in Fig. 3-4). Further in-depth study was carried out in three representative neighbourhoods, namely TL, HL and HB (within the red circles in Fig. 3-4)5. The neighbourhoods TL and HB are in two of the districts within the Shenzhen Special Economic Zone (SEZ), while HL is outside the SEZ. All of these three neighbourhoods are comprised of two parts, the newer part and the older part, and the population in each neighbourhood is almost equally divided between these two parts. 1) TL neighbourhood, with an area of 7.5km2, is a traditional industrial neighbourhood with a majority of hired labourers in their 20s coming from other cities. It has a population of approximately 40,000, with the gender ratio at 54:46 (male: female) (neighbourhood workstation report, 2009); 2) HB neighbourhood (0.88km2) is marked by prosperous business activities and it is close to the city’s core business centre. It has a population of

 

5 In response to requests for confidentiality from neighbourhood officers, the three target neighbourhoods have been symbolised by the abbreviated names TL, HL and HB. The street maps for these neighbourhoods are used only for data analysis, without being presented; instead, land-use sketch maps are presented for readers’ information.

46

Chapter Three

approximately 38,000 (neighbourhood workstation report, 2010) with 32,000 migrants, and most of them are associated with the retail business (over 600 shops and restaurants in 2009 (neighbourhood workstation report, 2010)); 3) In the HL neighbourhood (2.2km2), most of the 42,000 residents live outside the SEZ but work within it. Compared to TL and HB, there is only a small amount of local industry and business activity in HL, and it is mainly residential. In all these three neighbourhoods, the majority of residents (about 90%) are so-called temporary residents, i.e. migrants from other cities rather than registered Shenzhen citizens.

3.2 Justification of the Methodology 3.2.1 Quantitative methods at macro scales Research on different geographical scales is designed to test specific hypotheses. On the national scale, the main aims are to answer the following questions: (1) What has the trend in the crime rate been like since the economic reform? (2) What are the most influential variables for each type of crime? (3) Is there any tendency for the most influential variables to change in importance over time? On the provincial scale, the research focuses on two elemts: (1) spatial cross-sectional and space-time series regression analysis among provinces at selected time points to identify the spatial distribution patterns over time and compare the crime trends between them; (2) similar analyses to those carried out on the national scale in selected provinces (or autonomous district/municipality). The above analyses can also illustrate scaling effects on the relationships between crime levels and the socioeconomic variables, which may vary according to crime type and geographical location. The research design on the city scale is a replica of that employed on the provincial scale. It will also be divided into two parts based on secondary data, and follow similar procedures. To sum up, research on macro scales (national, provincial and city) includes: (1) temporal change pattern analysis; (2) relationship analysis using demographic, socioeconomic and other variables as previously mentioned; (3) spatial cross-sectional and space-time series regression analysis, and (4) comparative analysis based on spatial cross-sectional data. 3.2.1.1 Temporal change pattern analysis The first crucial task is to examine trends in crime rate changes over time. The statistical equation is:

Data and Methodology

y(t)=b0 +b1t+e(t), t=1,2,...,31

47

(3.1)

where y(t ) is the annual crime rate for any specific crime type at time t, where t is calculated based on the difference between the value of the year and 1978 (the starting point). For example, the t value for 1979 is 1, for 1980 is 2 and so on. b1 is the slope coefficient reflecting the relationship between time t and y. b0 is the intercept, and e (t ) is the error term. The standard hypothesis testing procedure using the t-test is then used to test  whether the value of b is significantly different from the null hypothesis H0: b1=0. If the probability of the null hypothesis is p>0.05, the hypothesis H0: b1=0 should be accepted, which implies there is no significant linear  change over time; if p0, then these variables are called co-integrated at (d, b), denoted as CI(d, b). For the co-integration method utilised in this work, the I(1) variables are combined to generate cointegrating vectors I(0), and the long-run equilibrium relationship among them can be used to indicate their valid regression. The quantity of co-integrating vectors they could generate needs to be tested through Johansen and Juselius’s (1990) multivariate cointegration test using the vector autoregressive (VAR) model (Sims, 1980). The VAR model is a multiple time series generalisation of the AR model (Maddala and Kim, 1998, p34), and the model is also used to determine the long-run equilibrium relationship. Let

Yt =D +31Yt-1 +32Yt-2 +...+3k Yt-k +Q t (t 1,..., T )

(A-17)

where Y is a p u 1 vector of non-stationary variables integrated of order one (I(1)), D is a p u 1 vector of constant terms, 31 ,..., 3 k are p u p coefficient matrices to be estimated. Q t is a p u 1 Gaussian white noise vector with mean zero and finite variance. p depends on the number of variables included in the vector. For example for the case of two variables, Y = (crime rate and urbanisation level), p=2. The coefficient matrix 3 k can provide information on co-integrating or long-run equilibrium relationships among the variables. Its rank r indicates the number of cointegrating relationships existing among the variables in the vector Y t

t

t

(Johansen, 1988). It can be tested by two statistics, namely the trace ( Otrace ) and maximal eigenvalue ( k max ) statistics, and the test results will affirm the existence of long-run equilibrium relationships between the dependent variable (i.e. crime rate) and its determinants. If r is 0, there are no cointegrating vectors and p stochastic trends; if 01:

1 I

m xi )

exp( xi E m ) I

1  ¦ exp( xi E j ) j 2

(E-4) The vector ȕm= (ȕ0m … ȕkm … ȕKm ) includes the intercept ȕ0m and coefficients ȕkmfor the effect of xkon outcome m. Residents’ responses to crimes are analysed using multinominal logit regression, and data came from the questionnaire survey results. As noted, there are three possible responses by a resident (report to the police, report to a safety guardian, do nothing) and we build a model for each of the three possible responses.

Appendix E

400

Yi ~ a 0  a1i PSI i  a 2 i PNSM i  a3 i DNI i  a 4 i genderi  a5 i living i  a 6 i household i  a 7 i maritali  a8 i agei  a9 i educationi  a10 i employmenti  a11i incomei  a12 i workingtimei  a13 i hom eattendancei

(E-5) where i indicates neighbourhood i, and the dependent variable refers to respondents’ choices in neighbourhood i for a specific crime scenario. The multinomial logit regression models conducted the log likelihood tests for all independent variables’ partial regression coefficients, as listed in Table E-8, and if the p value is less than 0.05, indicating that at least one of the coefficients is not zero, in other words, this model fits better than the model with only the intercept term. It seems that the model for medium and minor crimes fits better than that for serious crimes. Table E-8 Model fitting information of the multinomial logit regression models

Serious crimes TL

Medium crimes Minor Crimes

Intercept Only Final Intercept Only Final Intercept Only Final

-2 Log Likelihood 244.678

ChiSquare

Df

Sig.

186.795 437.878

57.883

114

1.000

282.171 505.413

155.707

114

.006

367.461

137.951

114

.063

The Pearson test of goodness-of-fit results in Table E-9 showed that the p values for serious and minor crimes are lower than 0.05, indicating a significant F distribution; while the Deviance test p-values for all cases, larger than 0.05, because it is highly related with the variables presenting many zero observations, so the results are not reliable enough at this point. 2

Geography of Crime in China since the Economic Reform of 1978

401

Table E-9 Models Goodness-of-fit

Serious crimes TL

Medium crimes Minor crimes

Pearson Deviance Pearson Deviance Pearson Deviance

ChiSquare 16241.823 186.795 366.892 282.171 571.085 367.461

Df

Sig.

483 483 483 483 483 483

.000 1.000 1.000 1.000 .003 1.000

The multinominal logit regression results on the coefficients of the residents’ responses to crimes model in TL are presented below, with * meaning significant at 90% level, ** meaning significant at 95% level, and *** meaning significant at 99% level. 1. For serious crimes Residents’ responses to serious crimes take category seven, “leave it alone”, as the reference, and models for category zero, “other solutions”, category one, “report to police” and category two, “report to neighbourhood guardians” will be built upon this reference.

(1) Logit (

response0 ) response7

15.637  3.606( PSI 1)  2.724( PSI 2)  3.039( PSI 3)

6.02( PNSMI 1)  0.616( PNSMI 2)  2.456( PNSMI 3)  3.032(inc 1)  0.692(inc 2) 0.602(inc 3)  0.215(inc 4)  1.171(inc 5)  2.548(empl 1)  0.242(empl 2) 5.126(empl 3)  0.107(empl 4)  6.924(edu 1)  1.348(edu 2)  3.444(edu 3) 0.777(edu 4)  2.732( age 1)  2.38(age 2)  2.298(age 3)  3.059(age 4) 14.6(mari 1)

 12.225(mari 2)

 13.954(mari 3)

 0.337( work 1)  1.471 ( work 2)  1.114(act 1)  0.414(hou 0)  3.742(liv 1)  5.064(liv 2)  3.548 (liv 3)  2.696(liv 4)  0.843( gen 0)  1.352( NCI 1)  0.692( NCI 2)  0.697( NCI 3)

Appendix E

402

(2) Logit (

response1 ) 16.955  3.49( PSI 2)  1.474( PSI 3)  1.554( PSI 4)  0.909 response7

( PNSMI 1)  0.906( PNSMI 2)  1.956( PNSMI 3)  3.017(inc 1)  1.318(inc 2) 0.022(inc 3)  0.945(inc 4)  0.335(inc 5)  2.225(empl 1)  0.42(empl 2) 3.662(empl 3)  0.056(empl 4)  9.683(edu 1)  3.994(edu 2)  4.427(edu 3) 3.388(edu 4)  2.25( age 1)  2.406(age 2)  1.695(age 3)  3.628( age 4) 8.429(mari 1)  9.682( mari 2)  9.233(mari 3)  0.637( work 1)  0.605 ( work 2)  0.794(act 1)  1.143(hou 0)  2.784(liv 1)  1.432(liv 2)  2.791 (liv 3)  1.658(liv 4)  0.706( gen 0)  1.718( NCI 1)  0.983( NCI 2)  0.696( NCI 3)

(3) Logit (

response2 ) response7

6.13  4.231( PSI 2)  1.263( PSI 3)  1.731( PSI 4)  1.491

( PNSMI 1)  1.484( PNSMI 2)  2.555( PNSMI 3)  1.424(inc 1)  0.563(inc 2) 0.69(inc 3)  2.098(inc 4)  1.622(inc 5)  0.951(empl 1)  0.858(empl 2) 3.46(empl 3)  4.524(empl 4)  10.396(edu 1)  2.71(edu 2)  2.957(edu 3) 3.111(edu 4)  4.723( age 1)  3.839( age 2)  2.965(age 3)  4.978(age 4) 15.456(mari 1)**  14.883( mari 2)**  14.574( mari 3)  0.13( work 1)  0.936 ( work 2)  0.685( act 1)  0.526( hou 0)  1.968(liv 1)  1.952(liv 2)  2.692 (liv 3)  1.039(liv 4)  ( gen 0)  1.548( NCI 1)  1.138( NCI 2)  0.435( NCI 3)

From the equations above, it turns out that in TL, the ratio of residents’ responses to serious crimes to the choice of “leave it alone” are only significantly positive with their marital status when they choose “other solutions” or “report to neighbourhood guardians”. On the other hand, residents’ choice of “report to police” to the choice of “leave it alone” when dealing with serious crimes is not affected by other variables, which partly explains the high proportion selecting category one for serious crimes. 2. For medium crimes

Residents’ responses to medium crimes also take category seven, “leave it alone”, as reference.

Geography of Crime in China since the Economic Reform of 1978

403

(1) Logit (

response0 ) response7

10.81  4.328( PSI 2)

 1.868( PSI 3)  3.243( PSI 4)

21.123( PNSMI 1)  22.077( PNSMI 2)  22.949( PNSMI 3)  6.808(inc 1)  9.304(inc 2) 10.623(inc 3)  22.671(inc 4)  7.039(inc 5)  12.419(empl 1)  28.6(empl 2) 8.29(empl 3)  29.492(empl 4)  10.264(edu 1)  21.971(edu 2)  17.95(edu 3) 6.656(edu 4)  5.546( age 1)  0.814(age 2)  1.393(age 3)  1.93(age 4) 35.389(mari 1)  35.435(mari 2)  33.202( mari 3)  4.04( work 1)  2.014 ( work 2)  0.493(act 1)  2.413( hou 0)  8.304(liv 1)  9.384(liv 2)  14.113 (liv 3)  13.082(liv 4)  1.296( gen 0)  10.615( NCI 1)  9.63( NCI 2)  8.866( NCI 3)

(2) Logit (

response1 ) response7

20.585  16.946( PSI 2)  0.479( PSI 3)  0.991( PSI 4)  7.015

( PNSMI 1)  15.402( PNSMI 2)  16.435( PNSMI 3)  10.688(inc 1)  8.325(inc 2) 9.244(inc 3)  10.12(inc 4)  10.576(inc 5)  7.467(empl 1)  16.603(empl 2) 9.068(empl 3)  8.872(empl 4)  5.786(edu 1)  4.459(edu 2)  4.816(edu 3) 4.964(edu 4)  1.93(age 1)  1.745(age 2)  1.341(age 3)  1.117(age 4) 19.11(mari 1)  20.415(mari 2)  19.922(mari 3)  0.875( work 1)  11.11 ( work 2)  0.223(act 1)  1.683(hou 0)  9.279(liv 1)  10.982(liv 2)  7.777 (liv 3)  9.933(liv 4)  0.416( gen 0)  9.374( NCI 1)  7.915( NCI 2)  10.625( NCI 3)

(3) Logit (

response2 ) response7

25.565  15.619( PSI 2)  1.553( PSI 3)  0.548( PSI 4)  8.798

( PNSMI 1)  16.323( PNSMI 2)  17.791( PNSMI 3)  10.515(inc 1)  9.613(inc 2) 9.17(inc 3)  10.597(inc 4)  11.773(inc 5)  2.785(empl 1)  6.315(empl 2) 1.36(empl 3)  8.484(empl 4)  0.675(edu 1)  1.269(edu 2)  1.614(edu 3) 1.569(edu 4)  10.169(age 1)  9.526(age 2)  8.75( age 3)  13.628(age 4) 32.608(mari 1)  33.202(mari 2)  31.797(mari 3)  0.313( work 1)  8.416 ( work 2)  0.415(act 1)  1.987(hou 0)  8.984(liv 1)  9.161(liv 2)  7.367 (liv 3)  9.319(liv 4)  0.061( gen 0)  10.12( NCI 1)  8.8( NCI 2)  10.15( NCI 3)

Appendix E

404

From the results for responses to medium crimes in TL, when divided by response seven, response one and response two do not have significant relationships with the selected variables. For other responses, only category two for the variable “perception of neighbourhood safety improvement”, which is “a little bit worse than before”, has a significant negative relationship with the dependent variable. 3. For minor crimes

Residents’ responses to minor crimes also take category seven, “leave it alone”, as reference. (1) Logit (

response0 ) 78.972  16.623( PSI 2)  1.164( PSI 3)  2.266( PSI 4)  0.208 response7

( PNSMI 1)  1.098( PNSMI 2)  0.575( PNSMI 3)  14.691(inc 1)  15.504(inc 2) 13.893(inc 3)  13.351(inc 4)  16.129(inc 5)  15.235(empl 1)  1.609(empl 2) 15.338(empl 3)  33.079(empl 4)  3.864(edu 1)  3.695(edu 2)  3.421(edu 3) 4.033(edu 4)  15.668( age 1)  15.027(age 2)  15.276( age 3)  15.674(age 4) 17.089(mari 1)  15.567( mari 2)  15.702(mari 3)  0.842( work 1)  0.114 ( work 2)  0.846(act 1)  0.93(hou 0)  11.77(liv 1)  12.699(liv 2)  13.125 (liv 3)  13.782(liv 4)  0.188( gen 0)  0.839( NCI 1)  0.562( NCI 2)  2.08( NCI 3)*

(2) Logit (

response1 ) 59.008  18.334( PSI 2)  2.421( PSI 3)  0.497( PSI 4)  13.155 response7

( PNSMI 1)  13.182( PNSMI 2)  13.084( PNSMI 3)  0.001(inc 1)  0.289(inc 2) 0.492(inc 3)  1.633(inc 4)  0.359(inc 5)  14.143(empl 1)  13.946(empl 2) 15.16(empl 3)  13.342(empl 4)  16.894(edu 1)  15.954(edu 2)  15.695(edu 3) 16.77(edu 4)  1.185(age 1)  0.71(age 2)  0.043(age 3)  1.545(age 4) 14.581(mari 1)  15.199(mari 2)  14.668(mari 3)  0.858( work 1)  0.425 ( work 2)  1.293(act 1)  0.183(hou 0)  0.84(liv 1)  1.234(liv 2)  0.324 (liv 3)  0.343(liv 4)  0.673( gen 0)  1.587( NCI 1)*  0.322( NCI 2)  0.208( NCI 3)

Geography of Crime in China since the Economic Reform of 1978

405

(3) Logit (

response2 ) 57.918  14.167( PSI 2)  0.437( PSI 3)  0.117( PSI 4)  16.449 response7

( PNSMI 1)  14.932( PNSMI 2)  14.199( PNSMI 3)  1.176(inc 1)  15.563(inc 2) 15.234(inc 3)  14.631(inc 4)  12.526(inc 5)  1.902(empl 1)  0.197(empl 2) 0.255(empl 3)  0.777(empl 4)  1.166(edu 1)  14.468(edu 2)  14.639(edu 3) 16.086(edu 4)  1.043( age 1)  0.06( age 2)  0.297(age 3)  0.364(age 4) 15.587(mari 1)  15.822( mari 2)  15.31(mari 3)  0.216( work 1)  1.572 ( work 2)  0.161(act 1)  0.494(hou 0)  0.22(liv 1)  0.806(liv 2)  0.332 (liv 3)  0.405(liv 4)  0.724( gen 0)  0.089( NCI 1)  0.827( NCI 2)  1.52( NCI 3)*

According to the above models of residents’ responses to minor crimes in TL, higher categories of PSI and NCI exert significant negative impacts on their responses.



APPENDIX F QUESTIONNAIRE FOR CHAPTER SEVEN

Dear Residents: In order to get better understanding of residents’ perceptions on neighbourhood safety, and further improve our work on neighbourhood management and crime prevention, our neighbourhood committee is in collaboration with the “liveable neighboyrhood” scheme executed by Shenzhen Planning and Land Resources Bureau to do investigation in our neighbourhood. We hope to get your support and cooperation, and we promise to keep the confidentiality of your information. *The respondent to this questionnaire should be above 16 years old. 1. What is your general feeling on safety when living in this neighbourhood (please tick before the option): Very safe safe not for sure whether it is safe or not unsafe very unsafe 2. What is your opinion on neighbourhood safety improvement in recent 2 years: improved a lot improved a little no change a little bit worse than before much worse than before 3. Have you or your friends been victim of the following crimes in this neighbourhood within last 5 years (and list the date/time and venue for the last victimization): e.g. Theft of Car Never Once One to three times Three to five times More than five times Early morning during the Spring Festival in 2008, Street corner near X restaurant Theft of Car

Geography of Crime in China since the Economic Reform of 1978

Never Once One to three times More than five times_________________ Theft from Car Never Once One to three times More than five times_________________ Theft of Electric bicycles Never Once One to three times More than five times_________________ Burglary Never Once One to three times More than five times_________________ Unsuccessful Burglary Never Once One to three times More than five times_________________ Robbery Never Once One to three times More than five times_________________ Picking and stealing Never Once One to three times More than five times_________________ Sexual harassment Never Once One to three times More than five times_________________ Personal threaten Never Once One to three times More than five times_________________ Fraud Never Once One to three times More than five times_________________

407

Three to five times

Three to five times

Three to five times

Three to five times

Three to five times

Three to five times

Three to five times

Three to five times

Three to five times

Three to five times

4. When you are walking alone in this neighbourhood, please scale the possibility of been a victim of crime in following time slots (no more than 2 options for each question): 1) The most dangerous time slot(s): 9:00-13:00 13:00-17:00 17:00-21:00 21:00-1:00 1:00-9:00 2) The safest time slot(s): 9:00-13:00 13:00-17:00 17:00-21:00 21:00-1:00 1:00-9:00

408

Appendix F

5. How is your feeling when walking alone in this neighbourhood at night? Very safe safe not for sure unsafe very unsafe 6. Where in this neighbourhood do you think are most likely to be the places occurring crimes in general (multi-options): Commercial business centers (e.g. shopping mall) bus stations home recreational places restaurants schools industrial places Food markets remote streets remote parks or groves others 7. When you happened to be victim of a crime in this neighbourhood, what measures will you take to fix it (multi-option)? 1) Serious crimes (e.g. robbery, burglary of more than 1000 Yuan, etc.): Report to the police report to neighbourhood guardians Contact with family members living in this neighbourhood Seek for help from friends or fellow-townsman living in this neighbourhood seek for help from other neighbours Call family back living far away leave it alone others ________ 2) Medium crimes (e.g. theft of cycles, burglary no more than 500 Yuan, etc.): Report to the police report to neighbourhood guardians Contact with family members living in this neighbourhood Seek for help from friends or fellow-townsman living in this neighbourhood seek for help from other neighbours Call family back living far away leave it alone others ________ 3) Minor crimes (e.g. picking and stealing less than 100 Yuan): Report to the police report to neighbourhood guardians Contact with family members living in this neighbourhood Seek for help from friends or fellow-townsman living in this neighbourhood seek for help from other neighbours Call family back living far away leave it alone others ________

Geography of Crime in China since the Economic Reform of 1978

409

Please mark the places that you think a crime might occur in the neighbourhood street map below less than 3 (ż for daytime; ѕ for night time): Figure of the neighbourhood street map

8. Whether do you think your neighbours will offer help when you happened to be a victim of crime? Of course yes maybe yes not for sure maybe not definitely not 9. What is your opinion on neighbourhood safety management: Very satisfied satisfied ok unsatisfied very unsatisfied 10. Which measures following do you think will improve your safety feelings in this neighbourhood? Improve the propaganda on safety improve the management of rented houses use high-tech measures (CCTV, entrance guard More efficient patrols increase the frequency system, etc.) be tougher on crime punishments improve of patrols the management on surrounding traffic improve the management on local commercial businesses others __________________________________________ 11. How long have you been living in this neighbourhood? less than 3 months 3 to 6 months 6 months to 1 year 1 to 3 years 3 to 5 years 5 to 10 years more than 10 years 12. What is your opinion on neighbours’ intimacy? Very intimate intimate neither intimate nor unfamiliar unfamiliar very unfamiliar 13. How do you think the degree of trust among neighbours in this neighbourhood? Very trust with each other trust neither trust nor distrust distrust very distrust 14. What is the frequency of your attendances to neighbourhood activities in recent 2 years (e.g. seminars on healthcare, societies, etc.)? Very frequent (more than 5 times) Frequent (3 to 5 times) Occasionally (1 to 3 times) Only once Never

410

Appendix F

15. What is your opinion on neighbours helping each other? Very often help others Occasionally help others Never help with each other, just care about his own business 16. How will you measure the familiarity with neighbours: Very often 1) Say hello to each other when meet rare never occasionally 2) Chat with each other when convenient Very often rare never occasionally Very often 3) Visit your neighbour’s family occasionally rare never 4) To be good friends with your neighbours Very often occasionally rare never

often often often often

17. What is your age: 16-20 years old 20-30 years old 30-40 years old 40-50 years old 50-60 years old 60 years old and above unmarried and without a relationship 18. What is your marital status: unmarried but have a relationship married divorced widowed 19. What is your educational level: primary school and below junior middle school undergraduate graduate and above middle school

higher

20. What is your working status now: Full-time job with stable income full-time job with unstable part-time job with stable income part-time job income with unstable income unemployed but don’t worry about living unemployed and worry about living student retired with stable pension Retired without stable pension others (please indicate) ___________ 21. What is the monthly average income in your family: Below 415 Yuan 415-1000 Yuan 1000-2000 Yuan 2000-3000 Yuan 3000-5000 Yuan above 5000 Yuan 22. What is your feeling about current income and life: very satisfied satisfied just ok unsatisfied

very unsatisfied

Geography of Crime in China since the Economic Reform of 1978

411

23. When do you usually leave home for work? daytime (e.g. from 8am to 7pm) night time (e.g. from 8pm to 7am) others ___________ 24. Is there any family member stay at home when you leave for work? Yes No 25. How do you usually spend your spare time after work? Stay at home visit neighbours walk outdoors within go to recreational places this neighbourhood go shopping others___________ 26. Your current living status: Living in self-owned house Live alone live with others live with parents mate without children live with mate and children children only Live with parents, mate and children Living in rented house Live alone live with others live with parents live with mate and children mate without children children only Live with parents, mate and children

live with live with others live with live with others

Please leave your personal contact information in case there is need to do telephone follow-up. Name ______________ Gender ____________ Telephone (or Mobile) No. _________________________ Current Address ____________________________________________

INDEX 5-point Likert item, 64, 175 5-point Likert scale questions, 197 abduction, 5-6, 70 accessibility, 15, 24, 41, 66-68, 96, 127-131, 156, 161, 245, 247 accessibility of healthcare services, 156, 161, 246-247 accommodation status, 62 accuracy, 52, 101, 256-257 acquaintances, 27 acquisitive crimes (AC), 6, 37, 55, 69, 75-77, 83, 89, 98-99, 101, 185-187, 192-193, 224, 235-236, 247, 253 acquisitive crime change, 89 acquisitive crime levels, 99 acquisitive crime rates, 18, 92, 98 acquisitive criminal activity, 23 activity patterns, 24 Ad hoc, 67 ADF test, 125, 154, 162, 282, 289 adjusted crime rates, 74-75 adjusted data, 98 administrative expenses (RAEP), 59, 88, 100 administrative unit, 48, 168 affluence, 21, 88 aggregate crime trends, 16 aggregation bias, 68 AIC and SBC criteria, 70, 290 Akaike’s Information Criterion (AIC),80, 124, 153 altruistic behaviour, 27 American Dream, 26 amplifying impact, 90, 129, 135, 156 Annual Statistical Yearbooks, 107, 144

Annual Yearbooks, 144 anomic conditions, 20, 244 anomie, 19-20, 237, 244 anomie theory, 236-237, 244 anonymity, 55 ArcGIS 9.3, 179, 183, 211 ARDL method (autoregressive distribution lag model), 80 Asia reporting rate, 74, 98 assault, 5-6, 18, 24, 27, 30, 37, 55, 74-78, 187 associational networks, 18, 26, 59 associations, 17, 33, 68, 144, 243 at night, 61, 171, 177, 180-184, 195, 198-200, 203, 211, 216-217, 225, 248, 250, 392 atomistic fallacy, 54, 69 attachment, 27 attendees, 68 attitudinal questions, 68 audiotape, 173 audio-video recorder, 52, 173 Augmented Dickey-Fuller (ADF) test, 46, 282 augmented F-test, 80 authoritative reports, 64 autonomous districts, 40, 103 average crime growth rate (GR), 76 average crime rate (CR), 78 average income, 35, 79, 108, 127, 145, 179, 409 average share of total crimes (SR), 76 awareness space, 15-16 bar districts, 24 barometer, 33 Beijing Statistical Yearbooks, 109 benchmarked, 61, 174

Geography of Crime in China since the Economic Reform of 1978 benefits, 15 bidirectional, 47, 83, 126, 225, 287 bidirectionally, 83, 126, 161 binary independent variables, 183 binary logit regression models, 196 binary variables, 182-183, 195, 200 bivariate correlations, 196 blackmailed, 229 blood donation rate, 33 bottom up, 1 box plots, 109, 119 broader, 16, 33, 53-54, 102, 147, 174, 187, 238 broader scale, 51, 251, 257 burglarproof, 233 burglary, 6, 24-25, 74-75, 174, 177, 187, 193-194, 217-218, 227-230, 233, 250, 387, 390-393, 407-408 business area, 170, 216 business areas, 216-217, 250 business neighbourhood, 224, 227 business neighbourhoods, 224, 250 business-oriented neighbourhood, 248 capable guardian, 15, 24 capable guardians, 24, 211, 224, 230-231, 248 capital city, 104, 143-144 capitalist culture, 23, 26 capitalist economics, 31 capitalist ideologies, 23 cartographic school, 17 case clearance rates, 66 case study, 8-9, 11, 30, 41, 146, 150151, 165-166, 168, 179, 184, 217, 227, 234, 250, 257 categories, 50, 52, 55, 75-76, 81-82, 182, 193, 203, 212-214, 224-225, 248, 323-328, 392-393, 398, 403 causal, 46-47 causal relationships, 45, 47, 98 census tracts, 256 change trends, 81, 110, 133, 145 changing trends, 105, 109, 144, 240 characteristics, 1-2, 9, 16, 19-20, 29, 38-40, 68, 106, 158, 169, 171,

413

180, 183-186, 189-191, 201, 214, 219, 226, 229, 236, 249-253 checklist, 54, 175 Chicago school, 19 child maltreatment, 20 China Law Yearbooks, 6, 8, 57, 73, 109 China Prosecution Yearbooks, 109 China Public Security Yearbooks, 57 China Statistical Yearbooks, 2, 36, 57, 75 China Yearbooks, 57 China’s population registration system, 189 Chinese criminal justice system, 68 Chinese criminology, 31 Chinese Minister of Public Security, 68 Chinese society, 4, 11-12, 15, 30-31, 35-36, 59, 256, 260 Chi-square value, 198 choropleth maps, 186, 225 cities, 1, 3-5, 8-13, 23, 32-35, 42-46, 52, 61, 67-70, 77, 106, 144-149, 152-162, 166-167, 171, 245-249, 255-256 citizen welfare, 37, 75, 84-87, 110, 131, 134, 137, 148, 157, 159, 165-166, 240-241, 247-248, 255, 303-310 citizens’ welfare, 36, 90, 94, 99, 102, 136, 142, 290, 313, 318, 323 Citizens’ welfare improvements, 102 city scale, 10, 13, 42-43, 47, 60-61, 67-68, 70, 143-144, 146-147, 167, 245-249, 253, 255-256, 259 city-street offices, 44 city-street officials, 65 citywide responsibilities, 53 civic norms, 20, 27-28 civil war, 31 closed questions, 55 close-knit, 65 clustering, 50, 120 coastal areas, 8

414 coastal provinces, 121, 123, 127, 135, 142-143, 242-244 coastal regions, 124, 242 co-facilitator, 54 co-integrated, 48-49, 88, 128, 285286 co-integrating vectors, 48-49, 88, 128,156, 286 Co-integration analysis, 48, 85, 128 co-integration and causality method, 48 co-integration equation, 88, 128, 154, 293-294 co-integration equilibrium equations, 157 co-integration models, 167 co-integration multivariate testing, 100, 139, 166 co-integration rank tests, 88, 128, 156 Co-integration relationship analysis, 47, 88, 125, 128, 156, 239, 247 co-integration test, 48-49, 75, 82, 88, 104, 107, 111-112, 126, 128, 134, 137, 142, 146, 149-150, 154, 156, 163-164, 238, 242, 245, 257, 259, 281, 286, 291-292, 299-300, 307309, 341-349 collection box, 55 collective efficacy, 18, 26, 29, 65, 236 comfortable, 2 common trends, 88, 126, 128, 154 communism, 69 Communist Revolution, 1 community officials, 62, 175 community police stations, 65 Comparative analysis, 46, 52, 111 comparative case studies, 52 comparative criminological studies, 11, 30 Comparative research, 33 comparative studies, 88, 126, 128, 154, 259 comparisons, 10-11, 13, 52, 55, 61, 109, 111-112, 128, 133, 148, 152,

Index 157, 164, 166, 170, 183, 185, 218-219, 243, 259-260 complex system,253-254, 257 Comprehensive Statistical Data and Materials on 55 Years of the New China, 57 concealed, 70 conceptual framework, 30, 59 condone, 24 confidence interval, 55, 190-191 confidentiality, 57, 70, 265, 407 Constant-match average, 149 Constant-match sum, 149 contextual conditions, 35 contextual legal institution, 61 contingency table, 214-217 contradictory, 47, 165 contributing factor, 44 converge, 17-18, 26, 92-93, 132, 137, 140, 158, 288 convergence velocity, 97 conviction, 105 correlation, 22, 25, 32, 34, 47-48, 68, 71-74, 81, 104, 112-114, 125-126, 149, 154-157, 169, 198, 203, 208 correlation analyses, 70 corruption, 8-9 cost-effective, 56 court, 106-107 Cox & Snell’s R Square, 198 CPTED (crime prevention through environmental design) theory, 18 credit cards, 232 crime attractors, 172, 240 crime categories,80, 195, 226 crime change analysis, 143, 146, 245 crime change pattern, 167 crime change trend, 135, 147 crime changes, 6, 12 crime clear up rate, 162 crime control, 6, 9, 11, 18, 39, 55, 181, 234, 243, 260 crime control and prevention, 18, 50, 236 crime data, 14, 21, 38, 41, 55, 57, 62,

Geography of Crime in China since the Economic Reform of 1978 66-69, 76, 78, 100, 103-104, 112114, 124, 146, 152, 167, 181, 183, 186, 219, 225, 229, 236, 252, 256, 258, 260 crime deterrent effects,101 crime incidents, 169 crime inducing, 39 crime investigation, 31 crime levels, 6-11, 23, 27, 31, 39, 42, 46, 53-56, 62-63, 77, 101, 103, 129, 170, 189, 242, 256 crime mediating, 138, 142, 158, 167, 236, 240, 243, 245-246, 248-249, 253, 255 crime opportunities, 16-17, 24, 140, 226, 244 crime pattern theory, 17, 27 crime patterns, 10, 16, 19, 48, 50, 52, 107, 139, 213, 219, 249 crime prevention, 9, 17-18, 31, 36, 140, 143, 234, 245 crime problem, 9, 12, 54, 62, 175 crime prone, 104, 184-185 crime rate change patterns, 47, 143, 244 crime rate trends, 24, 77, 153 crime rate variation, 73, 85, 97, 154, 157-160, 162, 244, 253 crime rates, 6-7, 9, 11, 20, 22-32, 34-39, 42, 44, 47-48, 50, 55, 57, 72-74, 76-78, 88, 91, 94-95, 9798, 100-102, 107, 113-114, 140145, 148, 151-153, 156-160, 165167, 180, 183, 186-189, 225-226, 237-238, 241-249, 253-256, 259260 crime reduction programs, 19 crime sites, 17 crime spillover effects, 36, 60 crime statistics, 67, 144, 257 crime surge, 76 crime target, 17, 26, 239 crime time-slots, 225 crime types, 7-8, 10, 33, 48, 62, 66, 78, 92, 96, 102, 112, 176, 186, 195, 219, 226, 388

415

crime variation, 242 crime-attracting, 229, 250 crime-inducing, 54, 102, 183, 233 crime-mediating, 12, 27, 139, 146, 158, 168, 179, 184, 194, 202, 208, 243, 245, 252, 255 crime-producing, 22, 32, 183, 201 crime-prone, 7, 27, 32, 60, 124, 185186, 193, 201, 213, 229, 239, 250 crime-related conditions, 142, 245 crime-related studies, 68 crime-reporting rate, 229 crime-ridden, 43 criminal behaviours, 6, 16, 21, 28, 31, 52, 103, 256 criminal cases, 9, 32, 68, 124 criminal choice, 21 criminal counts, 111, 121-123, 140 criminal events, 16, 25, 31 criminal inclinations, 17 criminal offenses, 7, 107 criminal opportunities, 22, 24, 246 criminal profiles, 233, 258 criminal suspects, 32 criminal targets, 16, 25 criminal victimisation, 26, 66, 180 criminal victimisation surveys, 66 criminality, 15, 21, 23, 25, 89, 106, 258 criminals, 32, 75, 106-107, 109, 111-112, 121-125, 139-140, 172, 226, 229, 231-235, 240 criminals prosecuted, 111, 124 criminogenic, 10, 12, 22, 25-27, 3031, 34-36, 39, 53, 58-59, 62, 72, 89-90, 94-95, 103, 136-142, 146, 167-168, 170, 184-185, 197, 202, 208, 236, 238-248, 253, 255-256, 260, 299 Criminogenic conditions, 10, 22, 25-26, 53, 62, 170, 236 criminogenic effects, 36, 72, 94, 158, 247-248 criminogenic factors, 31, 34, 39 criminogenic variables, 58-59, 136, 142, 241, 245

416 criminological theories, 15, 20, 101, 103, 140, 143, 167, 244, 246-247 criminological theory, 256 critical values, 84, 114, 326 cross-national, 11, 22, 30 cross-sectional, 10, 35, 41, 46-47, 61, 68, 107, 111, 242 cubic interpolated, 150, 163 cubic-match interpolated dataset, 246 Cubic-match last, 150-151 cultural affinities, 140 cultural beliefs and norms, 31 cultural centers, 85, 89, 91, 129 cultural conflicts, 22, 24-26, 36, 38, 44, 59, 73, 136, 245 cultural infusions, 143, 244 cultural institutions, 36, 60-61, 102, 239 cultural transition, 30 cultural values, 4, 23, 27-28, 31 culture conflict theory, 24 cyberbars, 230 daily activities, 17 daily living expenses, 90, 95 Daily newspapers, 109, 111 dampen, 101, 139 damping effect, 92, 131, 137, 158 dangerous, 11, 26-27, 173, 214, 219, 222, 225, 229-230, 252, 408 data availability, 15, 35, 59-60, 67, 73, 104, 109, 112, 147, 152, 260 data fluctuations, 74 data limitations, 149, 166 data property check, 81, 125-126, 154-155, 291 debit cards, 232 decision making, 17 decrease, 2, 27, 31, 38, 47, 60, 78, 101, 103, 120-123, 152-153, 188189, 226, 238, 241-242, 249 degrees of freedom, 114, 198 delinquent behaviours, 26 delinquent subcultures theory, 26 demarcation, 17 demographic, 4, 7, 9-10, 16, 18, 21,

Index 24-27, 35-41, 46, 50, 57, 60, 6263, 66, 72-73, 85, 89, 99-100, 103, 106, 109, 111, 127, 135, 138-139, 141, 146, 148-149, 166, 167, 175, 180, 183-186, 190-193, 201, 214, 219, 226, 229, 236, 238, 242, 245, 250, 252, 255 demographic and socioeconomic characteristics, 180, 183-186, 219, 226, 229, 250, 252 demographic composition, 25-27, 186, 193 demographic physical disadvantages,169 Deng Xiaoping, 1 density map, 185, 213 dependent variables, 74, 81, 83, 101, 143, 147, 184, 238, 325, 327, 329, 331, 388, 394-395, 400 descriptive, 31, 67, 181, 183, 388 deterrent effects, 90-91, 101, 140, 142, 240, 244 deterrent impacts, 136 developed cities, 5, 33 development priority, 143, 242, 244 deviant behaviour, 22 deviant behaviours, 23-24 dichotomously coded, 198 differences, 3, 10, 24, 30, 35, 49, 52, 54, 60-61, 63, 67, 71, 74, 78, 82, 92, 122, 126, 131, 156, 158, 176, 185, 193, 218-219, 225, 238, 245, 249, 252, 284, 310 different geographic scales, 20, 34 diffusion effects, 257-258 digressions, 54 disadvantaged groups, 236 disadvantaged people, 3 disequilibrium, 238, 254, 298 disorder, 29 disparities, 14, 59, 186, 229, 236, 252, 260 dispersion, 24, 120-121 displacement, 92, 131, 137, 158 disproportionately, 7 distance, 21, 110

Geography of Crime in China since the Economic Reform of 1978 distribution pattern, 46, 107, 111112, 121-123, 242 districts, 26, 42, 44-45, 67, 105, 170-171 disturbance, 36, 76, 138, 285, 289 divorce rate, 21, 35-37, 61, 65, 75, 85, 110, 128-132, 135, 147, 163, 241, 246 domestic living standards, 2 Dongguan, 11, 13, 33, 42-43, 144146, 148, 152-154, 156-166, 245249, 254-256, 362, 374, 382 dramatic, 2 drug abuse, 8 drug trafficking, 8 drug-related crime, 8, 31 dynamic changes, 158, 243, 287 dynamic interactions, 48, 100, 238 dysfunctional nature, 236 ecm, 49, 92, 94, 96, 131, 137-139, 158-159, 239-241, 247, 287-288, 297 ecological association, 16 ecological fallacy, 72 ecological level, 18-19, 35, 258 econometrics, 48 economic activity, 21, 147 economic and psychological contrasts, 43 economic change, 7, 20, 22, 25, 27, 31, 226 economic conditions, 7-8, 22, 29, 34-35, 39, 52, 111, 245 economic development, 1, 4, 6, 9, 34-35, 39, 60, 72, 90, 103, 130, 136, 140, 142-144, 158, 160, 162-163, 186, 244-246, 250, 256 economic fluctuations, 159 economic growth, 89-90,100, 238, 255 economic inequality, 22, 26-27, 140 economic liberalization, 4 economic performance, 59 Economic prosperity, 101 economic recovery, 1 economic reform, 1-4, 6, 9, 12, 15,

417

31, 33, 35-36, 39, 42-43, 46, 59, 72-73, 79, 101-103, 106, 120, 144, 146, 186, 237-238, 253-254, 256, 259 economic reform and opening-up policy, 43 economic self-reliance, 1 economic strengths, 2 economic success, 25, 28 economics of crime, 20 edges, 16 education level, 4, 27, 61, 66, 184, 214 educational attainment level, 36, 61, 103, 136, 139, 181, 200, 202, 240-241, 248, 250, 256, 394 educational background, 193 eigenvalue, 88, 128, 157, 286, 294, 374-381 elasticities, 74 elites, 2 embryo, 19 emotional support, 28 empirical studies, 22-23, 25, 28, 32, 38, 47, 60, 101-102, 140-143, 167, 169, 233, 241, 243-248, 256-258 employment stability, 66 employment status, 66, 89, 184, 193, 197, 202, 208, 214, 228 endogenous, 97, 100, 254 Engel coefficient, 29, 60 Engel index, 2, 36, 85, 89, 94, 110, 130, 133, 136, 148 Engel’s Coefficient, 32 entrepreneurs, 2 environmental associations, 19 environmental conditions, 17, 63, 175 environmental criminologists, 18 environmental criminology, 16-18 equilibrium state, 49, 92, 100, 131, 137, 142, 158, 238, 244, 254, 259 equilibrium status, 96, 138 equilibrium terms, 94 error correction term (ecm), 92, 131,

418 158, 287-288 ethics issues, 57 ethnic composition, 60 ethnic heterogeneity, 19, 27, 35, 60 ethnic minority, 64 ethnic or racial heterogeneity, 23 Eviews, 84, 149, 167, 326 excess zero values problem, 184 exogenous, 51, 91, 100-101, 142, 254 exogenous shocks, 101, 142, 254 exotic cultures, 59 Expected Fraction of Agreement, 216 expenditure gap, 32 explanatory variables, 10, 13, 49-51, 59, 73-76, 88, 97, 100, 102, 105, 109, 128, 143, 226, 241, 244, 254, 287, 290, 389 exploratory spatial data analysis (ESDA), 20 export income, 89 expressive crime rate, 74, 81, 85, 91, 95, 237-238 expressive crimes, 8, 39, 57, 72, 7779, 100, 103, 188-189, 194-196, 226, 238-239, 241, 249, 255 face-to-face, 54, 178, 183 factor analytic school, 19 familiarity, 176, 411 familism, 229 family activities, 180 family attendance, 213 family cohesion, 23, 27, 36, 61, 6465, 184 family disruption, 18, 23 family institutional anomie, 61 family structure, 29 family support, 39, 193, 196, 202, 227-228, 250 family ties, 24 family’s reputation, 257 fear of crime, 14, 38, 61, 67, 169, 183, 200, 213, 227, 229, 252-253 fellow townsmen, 230-231 fetishism, 28

Index field observations, 68, 104, 258 fieldwork, 259 finer, 67, 149, 253, 259 first differences, 49, 82, 126, 156, 284 first order differenced, 82-83, 91-92, 127, 285, 296 first-hand knowledge, 176 first-hand materials, 168 floating people, 64, 171-172, 190, 249 floating population, 4, 32, 43-44, 60, 190 fluctuations, 6, 47, 74, 92, 94, 98, 104, 131-132, 158-160, 282, 295, 298 focus group, 10, 41, 53-56, 61-63, 69-70, 169, 175-176, 182-183, 189-190, 205, 230-233 follow up, 63 follow-on question, 176 forcible seizure, 187-189, 249 forecast error variance, 49, 101, 289 forefront, 42 foreign capital, 2, 36, 38, 59, 75, 111, 129, 130, 148 foreign trade, 1 forest area, 218, 225 forged, 70 formal measures, 205 formal responses, 251 formal solutions, 205, 228, 251, 257 Foshan, 13, 42-43, 144, 146, 148, 153-154, 156-165, 245-248, 254, 376, 383 four-quadrant chart, 119 framework, 12, 16, 30, 48, 59, 147, 242, 258 fraud, 7-8, 32, 39, 57, 77, 80, 176, 187-189, 194-196, 219-220, 229, 232, 249, 252, 388, 408 frequency, 26, 149, 167, 176, 179180, 194-195, 200, 226, 410 F-statistic values, 84, 127, 162 gambling, 8 gang crime, 8

Geography of Crime in China since the Economic Reform of 1978 GDP per capita, 2, 32, 36-37, 75, 110, 129-130, 133, 163 Geary’s C, 51-52 gender ratio, 45, 171, 190 general consumer price index, 37, 60, 75, 85, 110, 133, 148 general educational accessibility, 98 general educational level (SJM), 102-103 GeoDA, 51, 121 geographic profiling theory, 18 geographical contexts, 9, 39 geographical differences, 35, 60 Geographical features, 59-60, 112, 128, 140 geographical location, 40, 46, 124, 143 geographical relationships, 51 geographical scales, 9-11, 39, 41, 46, 70, 76, 103-104, 168, 258 geographically weighted regression, 50 geography, 12, 15-16, 19-20, 36, 38, 111, 131-132, 134, 136-137, 143, 243-244, 255, 259, 326, 328, 330, 332, 334, 336, 338, 340, 342, 344, 346, 348, 350, 352, 354, 356, 358, 359 geography of crime, 12, 15-16, 20, 259 geopolitical advantages, 142-143, 244-245 geopolitical positions, 144 geospatial, 19 Gini coefficient, 33, 36-37, 59, 75, 85, 110, 112, 133, 147 GIS, 10-11, 21-22, 39 global engagement, 2 Global Moran’s I index, 120 global spatial autocorrelation, 51 goodness of fit, 198, 209 goods disposal channels, 237 government agencies, 39, 258 government expenditure per capita, 32 government regulation, 32

419

gradient colours, 186, 215 Granger causality, 49, 84-86, 104105, 125, 127, 141, 154, 162-163, 290, 300-302, 305, 334, 336, 338, 340, 383-386 Granger cause, 85, 128-129, 154, 163, 290, 300-301 Granger non-causality, 49 grassroots, 9, 65 great leap forward, 6 great natural disaster, 6 Gross Domestic Product (GDP), 59 group discussion, 54, 56, 190-191, 231, 254 group values, 24 Guangdong Province, 10-11, 13, 33, 41-44, 106, 121, 126, 136, 143146, 151-154, 166-168, 244, 246 Guangdong Public Security Department, 154 Guangdong Statistical Yearbooks, 111 Guangdong Yearbooks, 148, 151 Guangzhou, 155-166, 246-249, 255, 257, 369, 379, 385 guardianship, 6, 169, 184-185, 200, 237 guilty, 106, 108 hallmark, 1 handlers, 26 HB, 45, 172, 209, 219 Henan, 13, 109, 111-112, 115, 117, 119, 121, 124-128, 131, 134-142, 243-245, 255, 332, 340, 348, 356, 362 Henan Statistical Yearbooks, 111 heterogeneous, 34, 54 hierarchical political system, 3 higher educational attainment, 102103, 139, 203, 241-242, 249, 256, 395 high-rise buildings, 236 high-value, 234 HL, 13, 45-46, 171, 173-176, 190191, 193-196, 199-203, 205-207, 210-214, 216, 218-220, 227-229,

420 250-252 home attendance, 197, 214, 252 homicide, 58, 73, 77, 81, 179, 188189 homogeneous, 50, 64 hot spots, 19, 186, 190, 214, 226, 230, 236-237, 253 house attendance, 182, 191, 201, 209-211, 213 household registration system, 106 hurdle models, 169, 387, 389, 391 hypothesis, 28, 39, 47-48 72, 82, 102, 120, 126, 136, 138-140, 158, 203, 214, 238-239, 241-242, 244, 249, 256-257, 285-286, 289, 292, 308-311, 396-397 illegal behaviours, 30, 73 illegal removal, 169 immediate environment, 18 immigrants, 23, 144-145 import and export trade, 2 improvements in living standards, 29, 60, 100 in the daytime, 63, 173, 178, 185, 198, 200-203, 206, 214-215, 219220, 228, 230, 237, 251-253, 401 in the long run, 89, 98-99, 101-103, 132, 140-142, 157, 162, 239, 242, 244-245, 247, 249-250, 254, 256, 299-300 in the short run, 95, 97, 102-102, 132, 138-139, 141-142, 160, 166, 242, 244, 248, 250, 254, 256, 298, 300 income gap, 90, 101, 129-130, 132, 136, 147 income growth (AI), 240 income inequality, 21-22, 31, 33, 39, 44, 64, 72, 127, 140, 142-143, 158, 163, 244, 246-248 income level, income per capita, 30, 37, 75 incongruity, 32 inconsistent, 47, 52, 83, 101, 114, 142, 158, 237, 242, 244, 247, 249, 253

Index independent variables, 9, 37, 39, 7374, 81, 83, 85, 88, 91-92, 101, 104, 110, 127-128, 133-134, 142, 147, 149, 154, 156-157, 163, 184-185, 197-200, 203-204, 209, 214, 228, 239, 243, 245, 281, 288, 291, 296,, 298, 300, 326, 328, 330, 332, 395-397, 402 independents, 47 in-depth, 11, 14, 34, 45, 56, 231, 250, 390 individual level data, 41, 168 individual lifestyle information, 38 individualism, 25, 28 industrial areas, 172, 219, 226 industrial economic growth, 90 industrial neighbourhood, 45, 171, 253 industrial places, 219-220, 253, 410 industrial revolution, 23 industrialization, 278 industrialized neighbourhood, 230 inequality, 3, 6, 21-23, 31-33, 36, 39, 43-44, 59, 64, 72, 85, 90, 98, 102-103, 112, 121-123, 127-128, 140-143, 145, 158, 163, 239, 241, 245-247, 256 inequality gap, 6, 22, 39, 103, 145, 257 inertia, 17 influential variables, 46 informal conversations, 175 informal response, 206, 252 informal social control, 26, 28-29, 39, 237 informal solutions, 206, 258 informants, 70, 231 infrastructure, 3, 291 inhibitor, 102 inland regions, 124, 243 inner provinces, 123, 136, 140, 42143, 243, 245 insecurity level, 10, 219, 253 insignificant, 51, 91, 102, 134, 138, 141, 211, 228, 243, 251, 255, 297, 299-300

Geography of Crime in China since the Economic Reform of 1978 institutional anomie theory, 25, 2728, 60, 140, 142, 239, 241, 243, 246 institutional theory, 246 integrated, 22, 48-49, 82, 88, 126, 128, 155-156, 170, 259, 281, 284-286, 364 intelligence-led policing, 18 intercept, 47, 82, 126, 284, 391-395, 400-401 inter-map comparison, 214 internal reform and external opening-up, 1 International Crime Victimisation Survey, 76 international transport hub, 145 interpolated data, 150, 163, 166-167, 245-246, 249 interpolated dataset, 163, 166, 246 interpolation, 149-150, 164, 166167, 249 inter-provincial, 122 interrelationships, 48 intervention behaviour, 29 interview, interviewees, intimacy, 176, 180, 410 investigation reports, 57 isolation, 2 Japanese invasion, 31 judiciary, 106 juvenile delinquency, 31 juxtaposed, 34 Kappa, 170, 186, 214-216, 218 Khisto, 216-217 labour force, 59 lag effect, 74, 76, 102, 138, 241, 243 lag order, 73, 82, 292-293, 296, 374381 Lagrange Multiplier (LM) tests, 51 laid-off, 9, 23 land uses, 17, 27 Landlords, 172, 234 landmarks, 63, 181 land-use choropleth maps, 186

421

land-use data, 171, 185, 213 laptops, 233 larceny, 7-8, 39, 57, 72, 74, 76-78, 84-85, 88-89, 93-94, 98-103, 187-188, 219-220, 226, 227, 237241, 252, 255 latitudes, 42 law, 7, 16, 25, 27, 34, 59, 61, 69, 106, 108 law enforcement, 27, 69, 106-108 Law Yearbooks, 6, 8, 57, 66, 73, 109 legitimate opportunities, 26 length of railway, 37, 60, 75, 110, 129-130, 132, 142, 290 lifestyle, 24-26, 38-39, 231 Lifestyle conflicts, 39 lifestyle exposure theory, 25-26 Likert-type, 65, 180 Likert-type scale, 178, 180 linear correlation coefficients, 126, 154 linear regression, 91-92 linear time trend, 82, 126 linear-match last, 150 listwise deletion, 69 living pressures, 239, 248 living standards, 1-3, 6, 29-30, 34, 40, 60, 72, 89-90, 100-102, 129130, 132-133, 243, 256 living status, 64, 182, 185, 192-193, 195, 199, 203-204, 249, 412 local land-use, 44, 229 local safety issues, 175 local statistical data, 32 logarithm, 74, 91-92, 96, 293 logarithmic transformation, 125, 154 logistic regression, 169, 197, 202203, 207-208, 226-227, 250-251, 394 logit regression model, 183-185, 198, 396, 401 log-transformed, 74 loitering, 219 long run, 89, 98-99, 101-103, 132,

422 140-142, 157, 160, 162, 238, 241, 243-244, 246, 248-249, 253, 255, 287, 297-299 longitudes, 42 long-run equilibrium, 49, 88-91, 9597, 100, 128, 131, 137, 139, 142, 157-159, 238, 241, 243, 247, 254, 259, 281, 285-288, 294-295, 297, 299, 349, 351, 353, 355 long-run relationship, 92-94, 96, 101-102, 131, 133-135, 137, 154, 157-158, 164-165, 247, 254-255, 281, 286-287, 300 long-run, 49, 76, 88-97, 100-102, 109, 112, 129, 131-135, 137, 139, 141-142, 149, 154, 157-159, 164166, 238-239, 241, 243, 245, 247, 253-255, 259, 281, 285-288, 294295, 297, 299-300, 349, 351, 353, 355 loopholes, 9 lower trust, 205, 251 macro and micro scales, 237 macro scale, 10, 22, 26, 29, 37-38, 46, 57, 67, 168, 253-254 Mainland China, 105, 146, 170 mainstream society, 24 maladjustment, 1 management improvements, 178, 181 marital status, 182, 184-185, 192193, 199, 211, 250, 394, 403, 411 market area, 189, 225, 229 market economy, 2-4, 31-32 market mechanisms, 1-2 marriage status, 65 Maximum Fraction of Agreement, 216 maximum lag, 73, 82, 84, 326 maximum likelihood estimation, 51, 69 mediating, 12, 27, 30, 35, 58, 60, 72, 91, 131, 136, 138-139, 142, 146, 158, 167-168, 179, 183-184, 194, 197, 202, 208, 236, 240, 243, 245-246, 248-249, 252-253, 255-

Index 256, 300 mediators, medium crime, 64, 206, 211, 228, 251, 401, 403, 405, 409 medium level crime, 179, 205, 210211 memory lag, 70 mental map, 14, 55, 63, 170, 176, 181, 183, 185, 213, 219, 229, 236, 251 micro scale, 25, 38, 52-53, 57, 168, 237, 253, 258-259 micro-geographic levels, 30 migrant (floating) population, 43 migrants, 3-4, 9, 23, 25, 32-34, 43, 46, 60, 144, 146 migration, 3-6, 8, 23-24, 31-32, 3637, 39, 42, 72, 74, 103, 106, 110, 132-133, 142, 148, 163, 245, 256 Ministry of Public Security, 68, 107 minor crime, 64, 68, 179, 205-206, 212-213, 228, 236, 251-252, 401402, 405-406, 409 minor level crimes, 211, 228 misleading, 18, 68 missing at random (MAR), 69 missing data, 69, 109 missing not at random (MNAR), 69 mixed degrees of integration, 82 mobile phones, 233 mobility, 4, 23, 25-26, 34, 36-37, 52, 64-65, 98, 103, 110, 129-137, 140-141, 148, 157, 159, 161-163, 165, 240, 243, 246-248, 255-256, 326, 328-329, 331, 349, 351, 353, 355, 357, 359-361, 377, 379, 381, 383-385 model fit, 197-199, 290, 394-395, 401 model summary, 203-204, 209, 211212 modernization, 227 Modernization theory, 227 modifiable areal units problem, 18, 258 momentous shifts, 2

Geography of Crime in China since the Economic Reform of 1978 Moran’s I, 51-52, 111, 120 motivated offenders, 16-17, 55, 64, 140 motives, 25 multicollinearity, 74, 198, 203, 208, 257, 288 multinomial logit regression model, 185, 401 multinominal regression models, 169 multiple equilibriums, 68 multiple regression, 198 multiple victimisations, 194-195 multi-scale, 66, 168, 259 multivariate, 48, 74, 76, 88, 100, 128, 139, 156, 166, 169, 183-185, 226, 286 multivariate co-integration test, 88, 128, 156, 286 multivariate framework, 48 multivariate logistic regression models, 169, 226 multivariate logit regression model analysis, 183-185 municipalities, 42 mutual support, 53, 61-64, 170, 175 mutual surveillance, 25 mutual trust, 26, 29 Nagelkerke’s R Square, 198 national, 1-2, 5, 10-13, 19, 21, 30-31, 36, 39, 41, 43, 46, 57, 59, 66-67, 72-76, 100-101, 103-105, 107, 109, 111-113, 124-126, 140, 143145, 149, 151, 237, 241-242, 244-245, 254-256, 258-259 national development priority policies, 143, 244 national enterprises, 2 National Government, 105 national policy, 140, 242 national population census, 43, 145 national scale, 12, 31, 41, 46, 57, 72-74, 76, 100-101, 103, 107, 109, 111-113, 124-126, 143, 151, 237, 241-242, 245, 255-256, 290 National Science Foundation, 66

423

natural growth rate, 60, 85, 110, 128-130, 147 natural population growth, 37, 74, 89, 135 natural population increases, 90 NCI score, 180, 184-185, 192, 197, 202, 208, 400 negative, 20, 51, 78, 80, 96, 120, 129, 131, 133-134, 199, 205, 238, 241, 288, 386-387, 389, 400, 406 negative coefficient, 49, 92, 94, 131, 136-138, 158 negative influence, 166 negative relationships, 29, 39, 92, 94-96, 101, 133, 139-142, 157160, 243, 255, 299, 394, 405 neighbourhood safety issues, 63, 175 neighbourhood, 4, 10-11, 13, 18-21, 25-29, 34, 38-39, 41, 43-46, 5356, 60-65, 67, 130, 132, 136, 140, 143, 149, 168-176, 178-180, 182186, 189-191, 193, 195-197, 199203, 205-208, 211, 213-214, 218220, 226-229, 231-236, 245, 249253, 259-260, 389, 394, 401, 405, 407-410 neighbourhood cohesion, 179, 182, 193-194, 196 neighbourhood committee office, 55 neighbourhood conditions, 200 neighbourhood crime level changes, 54 neighbourhood environment, 38, 64, 171, 176, 236 neighbourhood environmental conditions, 175 neighbourhood features, 53, 63, 176, 214, 219, 228, 236, 250, 252, 394 neighbourhood guardians, 205-206, 211, 213, 228, 235, 251, 402-403, 409 neighbourhood insecurity, 213-215, 229 neighbourhood level, 10, 13, 20, 27, 34, 60, 64, 189

424 neighbourhood management, 64, 199, 227, 250, 407 neighbourhood management improvement, 199, 227, 250 neighbourhood organisations, 29 neighbourhood relationships, 257 Neighbourhood residents’ committees, 65 neighbourhood safety, 13, 38, 55-56, 61-64, 169, 175, 178-179, 181186, 196, 200-202, 205, 209, 218, 226-227, 234, 236, 250, 252-253, 394, 400, 405, 407, 410 neighbourhood safety improvement measures, 186 neighbourhood safety management, 181, 184-185, 197, 208, 211, 213, 228, 251, 410 neighbourhood scale, 19, 26, 29, 34, 38, 41, 43, 45, 52, 61, 143, 149, 167, 170-171, 175, 186, 226, 229, 236, 245, 249, 258-260 neighbourhood self defence, 61 neighbourhood structural characteristics, 29 neighbourhood surveillance, 201, 226, 250 neighbourhood workstation report, 45-46 neighbourhood’s street map, 63 neighbouring provinces, 50, 52, 109, 120, 242 neighbourliness, 38, 55, 61, 170, 178-179, 183, 185, 193, 200, 208, 211, 227, 236, 251-252, 394, 400 neo-positivist, 56 neutral detachment, 56 new opportunity theories, 17 new village, 218-219, 225, 229, 236, 252 newer area, 173, 194 newer part, 45, 171-172 newer residential area, 172 NICR, 109, 111, 129-130, 132, 134, 137, 326, 328, 330, 332, 350, 352, 354, 356

Index nodes, 16, 27, 250 Non-economic institutional variables, 142 non-economic organisations, 27 non-linear, 74, 254 non-significance, 97 non-stationarity, 50, 125-126, 281283 non-stationary, 47-48, 81-82, 112, 126, 149, 154, 156, 164, 166, 280-287, 291 normalisation, 133 normalised coefficients, 133 normlessness, 33 northern inner provinces, 123, 242 notes, 54, 84, 89-91, 96, 175, 310, 326, 350, 358 NPS, 182, 185, 192, 199, 201-205, 227, 250, 387-388, 391-397, 400 number of doctors per 10,000 people, 60 objectivity, 57 observations, 51, 68-69, 73, 82, 104, 113, 149, 258, 282, 287, 395, 401 observer, 56 offence sites, 181 offenders, 9, 16-18, 25, 34, 55, 64, 140, 169, 196, 228, 230, 232-234, 236, 239-240, 250, 257 offenders’ motivation, 27, 64, 257 official agencies, 53 official census record, 190 official crime data, 14, 38, 55, 68, 181, 183, 186, 219, 225, 229, 236, 252, 256, 260 official crime records, 62 official criminal records, 53 official data, 10, 63, 67, 176, 219225, 229 official records, 69, 225, 229-230, 259 official statistics, 66-68, 104, 175, 258 old village, 218-219, 225, 229, 236, 252 older areas, 173

Geography of Crime in China since the Economic Reform of 1978 older part, 45, 171-172 older residential area, 172 OLS linear regression, 91 on-site investigations, 175-176 open-ended questions, 63 opening-up policy, 1, 9, 43, 76, 144 opportunities, 3-4, 8, 16-17, 22-24, 26-27, 39, 55, 59, 101, 138, 140, 145, 169, 196, 226, 229, 232, 239-241, 244, 246, 250, 256, 260 opportunity theory, 226, 238-239, 246 optimal lag length, 82, 126, 155 Orchard and Woodbury’s approach, 69 order of integration, 76, 285 ordinal logistic regression models, 169 ordinal regression model, 185 ordinal variable, 66 Organised crime, 8 original data, 100, 150, 165-166, 249, 362, 365, 368, 371 original forms, 126, 156 original interpolated data, 166-167, 249 original recorded crime data, 100 OSEZ, 170-171 outlier, 111, 121-123 parallel studies, 236, 259 participants, 54, 56, 63, 175, 183, 191 participation in civic life, 29 participation in organisations, 28 passenger traffic, 37, 110, 129, 132133, 136, 148, 163 paths, 17 PCT (port cargo throughput), 147 peak, 7, 77, 114, 124-125, 237-239 Pearl River Delta, 43, 144-145, 148 Pearson correlation coefficients, 114 peer group activities, 180 per capita income, 2, 89 percentage correctly predicted, 198 perceptions on neighbourhood safety, 13, 169, 196, 200, 227,

425

407 periodic features, 34 permanent population, 146, 149 permanent resident card, 189 permanent residents, 33, 64, 106, 145, 152, 189-190 personal information, 70, 178 personality, 56 person-situation interaction, 17 perturbation, 92, 131, 137, 158 piecewise, 47 pilot study, 171, 259 place, 1, 4, 11, 14, 16, 18-20, 22, 25-27, 35, 38, 50, 56, 63-64, 72, 104, 106, 124, 146, 150, 183, 185-186, 201, 205, 213-214, 218219, 225-226, 229-231, 234-235, 237, 242, 244, 250-252, 254, 258, 409 place-based analyses, 19 planned economy, 1-2, 30 point-based mental maps, 219 police agency, 53, 107, 182 police department, 106-107 police patrols, 201, 250 police records, 181, 219 police services, 257 police station, 4, 65, 225, 233, 257 policy makers, 39, 143, 245 political and legal development, 60 political and legal systems, 6, 103, 256 political convulsion, 76, 103, 237 political disturbance, 76 political institutions, 28 political manipulation, 68 political struggle, 31 polity, 25-28, 39, 60-61 polity institutions, 39 population characteristics, 183, 189, 191 population composition, 23 population density, 23, 35-37, 44, 60, 74, 85, 99, 102, 110, 128-130, 133, 135, 141, 147, 160, 163, 240 population heterogeneity, 18, 25, 29,

426 64 population increase (NGR), 90, 238239 Population mobility, 4, 98, 129-130, 132-133, 136, 240, 246 population turnover, 19 population urbanisation, 125, 129, 246 portable electrical goods, 232 portable valuable goods, 101 POS machine, 232 positioned subject, 56 positive, 22, 29, 39, 51, 60, 72, 92, 94, 96, 101-103, 120, 129, 131, 133-141, 157-160, 165, 200, 205, 211, 213, 238, 241, 243, 255, 290, 300, 386-387, 394, 400, 403 potential targets, 140 poverty, 2-3, 18-19, 21-23, 26, 3637, 64, 110, 148, 163 power control, 106 precipitation, 38, 60, 109, 111, 129130, 132, 242 predatory criminal victimisation, 180 predictive ability, 211, 227-228 pre-reform, 3-4, 6 primary data, 10, 12-13, 41, 57, 61, 67-68, 104, 175, 178, 256, 258 priority, 1 private enterprise, 2 problem crime area, 18 problem-oriented policing, 18 processing flow, 106 procuratorate, 68, 106-107, 112-113 professionalism, 69 property crime, 20, 23, 28, 101, 142, 196, 219, 226, 238-239 property surveillance, 196 prosecuted criminals, 107, 109, 111112, 114, 121, 124-125, 139 prosecution, 37, 106-107, 109-110, 112-114, 129-133, 136, 140-142, 147, 243-244 Prosecutorial Yearbooks of China, 109, 111, 113

Index pro-social, 25 prosperous, 2, 45, 172, 196, 201, 232, 250 prostitution, 8 provinces, 3, 8, 10-11, 13, 20, 30-31, 35, 42, 46, 50, 52, 59, 61, 67-68, 70, 105-107, 109-112, 114-115, 119-128, 133-143, 146, 171, 231, 242-244, 256 provincial scale, 13, 41, 46, 105, 107, 109-113, 139-140, 142-143, 146-148, 152, 154, 166, 242-245, 253-256, 259 proximity, 17, 172 proxy, 60, 107, 114, 200 PS, 182, 185, 192, 199, 201-202, 204-205, 227, 250, 387-388, 391396, 400 public and private space, 17 public control, 65 public facilities, 235-236 public fixed asset investment (PFAI), 130, 136 public libraries, 32, 38, 61, 75, 111, 129-130, 132-133, 141, 148 public monitoring, 233 public order, 26, 29, 65 public security, 7, 57, 61, 68, 75, 91, 98, 101, 107, 154 public transportation, 145 punishment, 4, 9, 106 purse-stealing, 219 Quadratic-match average, 149 Quadratic-match sum, 149 qualitative, 10, 12, 41, 43, 52-53, 56, 61, 69, 169, 182, 233, 236, 259 qualitative and quantitative, 10, 12, 43, 52 qualities of place, 19 quantitative, 10, 12, 43, 46, 52-53, 55, 59, 61, 63, 169, 182, 200, 216, 236, 259 quartile maps, 121 questionnaire survey, 10, 13, 41, 5355, 63, 65, 71, 169, 175-176, 178, 182-183, 194, 202, 207, 213, 226,

Geography of Crime in China since the Economic Reform of 1978 400 questionnaire surveys, 61, 71, 169, 175, 205, 219, 256 questionnaires, 56, 62, 178, 389 R square value, 199 racial and ethnic, 19, 170 random error, 68, 150, 163-164, 166-167, 245, 249 random sample, 55 random shocks, 97, 132 rape, 7-8, 57, 77-78, 80, 188-189, 257 rapid social change, 9, 15, 23, 39 Ratio of administrative expenses to public expenditure, 61 Ratio of Public security expenses to public expenditure, 61 rational choice theory, 17, 25, 226, 238 recorded cases, 31 recorded crime, 14, 23, 62, 73, 7678, 100, 103, 107, 112-114, 146, 148, 152, 166, 181, 186, 220, 236, 258, 260 recorded criminal cases, 124 recreational areas, 219, 229, 252 recreational facilities, 235 red light districts, 26 Reform and Opening-up, 1, 3, 9, 31, 43 registered population, 44, 106, 190 registered residents, 189-190 registration system, 4, 106, 189, 234 regression model, 20, 50, 169, 183185, 198, 208, 226-227, 257-259, 280, 290-291, 300, 386, 394-396, 400-401 regulations, 144, 233-234 relationships, 9-11, 19, 22, 29-31, 34-35, 39-40, 46-48, 50-52, 55, 69, 71, 73-76, 91-94, 96-97, 100102, 111-112, 125, 128, 131, 133143, 154, 156-159, 164-167, 183, 193, 196, 200, 236, 238, 240-245, 247-248, 252, 254-260, 286-287, 290, 297, 299-300, 400, 405

427

relative deprivation, 33 relax, 1 reliability, 57, 67, 103, 236, 258 religion, 25, 28 rental bill, 234 repeat crimes, 226 repeat victimisations, 226 repetitions, 54 replica, 23, 46, 146, 245 reporting rate, 79, 100, 229-230, 252 Reports on the Work of Government, 109, 111 representative, 1, 45, 53, 55-56, 61, 70, 111, 124, 146, 175, 191, 242, 249 Research on the contemporary criminal issues in China, 31 resident’s perception, 197 residential area, 172, 214, 218-219, 229, 232, 235, 252 residential instability, 18, 21, 26 residential mobility, 23, 25, 34, 6465 residential neighbourhood, 226, 229, 250, 252 residential stability, 193 residential turnover, 65 residents’ demographic composition, 186 residents’ fear of crime, 14, 38, 61, 200, 213, 253 residents’ mental maps of unsafe places, 14 residents’ victimisation experiences, 13, 179, 183-184, 194, 196, 200, 219, 226-227, 250, 386 residents’ victimisation logistic regression model, 227 respondents, 29, 55-57, 63, 65-66, 70, 169-170, 173, 176, 178, 181, 184-185, 193, 201, 205, 213, 219, 235-236, 250, 252, 400-401 response patterns, 206 responses to crime, 13, 64, 169, 183, 185, 205, 207-208, 226, 236, 251,

428 400-401 responses to offences, 178, 183 responses to victimisation, 179, 205, 207, 227 resurfaced, 19 resurgence, 19 retrospective surveys, 70 revenue disparity, 32 review-board protocol, 57 risk of crime, 63, 176 risks, 17, 150, 185-186, 196, 233 RMC, 208, 210 RMNC, 208, 212 robbery, 7-8, 22, 26, 32, 39, 44, 57, 76-77, 80, 148, 152-154, 166, 172-173, 179, 188-189, 194-196, 205, 219, 230, 235, 249, 388, 408 robustness, 151, 163, 245 routine activity, 17, 24-27, 63, 169, 178, 180, 226, 230, 235-236, 257 RSC (crime clear-up rate), 147 safety management, 182, 184-185, 196-197, 208, 211, 213, 228, 251, 410 sample period, 49, 97, 132, 160, 289 sample selection, 190-191 sample size, 55, 190 satisfaction with current neighbourhood safety management, 184-185 scale dependent, 50 scale effect, 40, 103-104, 168, 237, 245, 253, 255, 258 scale problem, 70 scale-down, 168, 258-259 scaling up, 168 scatter plot, 52 scenario, 179, 207, 228, 251, 401 school area, 218-219, 225, 230, 252 Schwarz Information Criterion (SIC), 82, 126, 155 Schwarz-Bayesian (SBC) criteria, 49 scrap collection depots, 235 seasonal feature, 225 secondary data, 10, 12, 41, 46, 57,

Index 59, 61, 66, 68, 104, 175, 256, 258 secondary statistical data, 168 second-hand, 196, 235, 250 security measures, 232 security precautions, 27 security protection measures, 231 self-defense consciousness, 231 self-perceived vulnerability, 38, 252 self-protection, 196, 228 self-reporting surveys, 66-67 semi-official guardians, 234 semi-public control, 65 sensitivity, 150, 163, 166-167, 246, 249 sentencing, 4, 106 serious crime, 8, 26, 64, 179, 205209, 227-228, 236, 251-252, 401403, 409 seriousness, 17, 150, 185-186, 196, 207, 227, 251 Shanghai, 1, 5, 13, 33, 44, 111-112, 114, 116, 118-120, 124-143, 243245, 54-256, 327, 335, 343, 351, 359 Shanghai Statistical Yearbooks, 111 Shenzhen, 1, 5, 8, 10-11, 14, 33, 4145, 52-53, 62, 64, 67, 144-149, 152-166, 168-172, 186-191, 231, 233, 245-249, 254-256, 371, 380, 385, 407 Shenzhen Politics and Law Yearbook, 147, 186-187 shopkeepers, 201 short-run, 49, 76, 88, 92-102, 112, 125, 131-133, 136-140, 142, 149, 154, 158-166, 238-239, 241, 244245, 247-248, 253, 255, 259, 287-288, 295, 297-300, 357, 359361 significance testing, 82 significant, 1, 5, 9-10, 29, 31, 39, 47, 50-51, 72, 76, 84, 89-92, 94-96, 100-102, 113, 120, 127-128, 133134, 136, 138-141, 154, 157-160, 162-166, 193, 195-200, 203, 205, 211, 213, 227-228, 238, 240-241,

Geography of Crime in China since the Economic Reform of 1978 243-244, 247, 249, 256-257, 280281, 288-290, 297, 299-300, 350, 358, 364, 391, 394-396, 400-405 similarity, 165, 214, 216, 218 single victimisation, 194-195 situational dynamics, 17 small and medium sized shops, 196, 250 smuggling, 7-8 snapshot, 153, 178 social and economic change, 7, 20, 25, 31, 226 social and economic inequality, 26 social and physical environmental characteristics, 20 social aspects, 35 social behaviours, 231 social bonds, 23, 28, 36 social capital, 27-29, 60, 140 social change, 2, 9, 11-12, 15, 20-31, 34-36, 39, 57, 59, 72, 102-103, 238-239, 253, 260 social cohesion, 10, 21, 26-30, 53, 55, 61-64, 73, 170, 202 social cohesiveness, 54, 63, 175, 183 social conditions, 24, 35, 52, 194 social construct, 170 social context, 73 social control, 4, 21-23, 26-29, 3132, 39, 63-65, 178, 236 social determinants, 50 social disorganisation, 19, 21-23, 26, 36, 63, 102, 143, 178, 235, 243, 245-246, 257 social efficacy, 170 social events, 100, 237 social factors, 47 social harmony, 35, 37, 75, 84-86, 89-96, 98-100, 110, 131, 134, 136-137, 139, 148, 157, 159, 161-163, 165-166, 241, 243, 247248, 254-255, 302, 304, 306, 308, 309, 312, 317, 375, 377, 379, 381-385 social hierarchy, 3

429

social improvement, 35, 74, 84-85, 87, 89, 91, 94-95, 99-100, 110, 136, 142, 148, 159, 161-163, 240-241, 290, 299, 305, 307-309, 313, 318, 323, 326, 328, 332, 375, 377, 379, 381-385 social inequality, 3, 23, 36 social instability, 27, 59, 72, 77, 90, 101, 130, 140, 241 social institutions, 25, 27-29, 60-61, 72, 103, 235, 239, 241 social integration, 31 social interaction, 29 social networks, 231 social organisation, 22, 100, 140 social problems, 43 social reality, 35 social relationships, 196 social responsibility, 61 social stability, 9, 32, 36-37, 110, 131, 134, 136-137, 139, 142, 147, 157, 159-165, 247-248, 254-255, 374, 376, 378, 380, 382, 385 social status, 26, 240 social structure, 3, 23 social support, 33, 38, 140, 170, 230 social ties, 26, 28-29, 65, 180 social transition, 21 social trust, 28 social upheavals, 43 social welfare, 140, 243 socialism, 6, 69 socialistic market economy reform, 1 society, 1 socio-demographic characteristics, 169 socioeconomic, 3, 7, 16, 18-22, 2627, 35, 38-39, 46, 48, 50, 57, 7374, 100, 102-103, 109, 111, 124, 139, 141, 146, 148-149, 166, 170, 175, 180, 183-186, 219, 226, 229, 236-238, 242, 245, 250, 252, 259 sociologists, 9, 11, 21, 32-33 southern coastal provinces, 123, 242 space-time crime patterns, 48

430 space-time series, 10, 41, 46, 111 spatial analysis, 9, 20, 111 spatial and ecological distribution of crime, 47 spatial and temporal scales, 40, 237, 259-260 spatial autocorrelation, 50-52, 120 spatial autoregressive, 51 spatial clustering pattern, 120 Spatial cross-sectional, 10, 35, 41, 46, 61, 68, 107, 111, 242 spatial data analysis, 11, 20, 39, 70 spatial dependence, 111 spatial dimension, 19, 21 spatial distribution, 19-20, 22, 46-47, 107, 111-112, 121-123, 183, 186, 242 spatial entities, 70 spatial error model, 50, 52 spatial heterogeneity, 50, 258 spatial lag, 50-52 spatial measurements, 20 spatial pattern, 14-15, 55, 111, 120, 170, 181, 219, 222, 225 spatial perspective, 19 spatial regression, 20, 50 spatial relationships, 50-51 spatial scale, 143, 245, 253-255, 258 spatial statistics, 20 spatial structure, 20, 22 spatial studies of crime, 19, 21 spatial units, 53, 73, 264-265 spatial variation, 22 spatial weights, 53 spatially distributed inequality, 143 spatial-temporal, 9, 18, 27, 107, 111-112, 114, 121-123, 139, 146, 152, 170, 185, 219, 230, 242, 245 Special Administrative Regions, 44 special economic zone, 1, 45, 143144, 146, 170, 245 spillover, 39, 112, 144 SPSS, 201 spurious correlation, 81, 104, 112, 126, 149, 154 spurious regression, 49-50, 77-78,

Index 287, 298 standard deviation, 122 standardized coefficients, 95 state-owned, 3-4 stationarity, 48, 50, 75, 154, 166, 281-283, 285, 287, 291 stationary, 47-48, 81-82, 126, 155156, 164, 280-287, 291 statistical representativeness, 55 stochastic, 48-49, 51, 281-288, 298 strain theory, 22, 25, 239 street map, 63, 181, 213, 410 stretching out, 74 study area, 12-13, 42-43, 50, 171173 subcultural conflicts, 26 sub-districts, 170 subjective bias, 56, 70 subjectivity, 56 subsistence, 22 suitable target, 17-18, 26, 226, 229, 231 supplicant, 56 Supreme People’s Court, 68, 107 Supreme People’s Procuratorate, 68, 107 surveillance, 17, 25, 38, 196, 201, 226, 250 suspects, 32, 106, 189, 234 systematic deletion, 69 target, 18-21, 24-26, 30, 36, 39-40, 50, 53, 59-60, 64, 105-106, 109111, 139-140, 142-143, 167, 169, 172, 175, 178, 183, 190, 226, 229-231, 235-236, 239, 243-245, 255, 259-260, 291, 394 t-distribution, 114 telecommunication, 257 Telescoping, 70 temperature, 38, 60, 109, 111, 129133, 140, 142, 242-243 temporal analyses, 61, 107 temporal change, 46, 85, 111, 114, 119, 189 temporal comparisons, 111, 152, 170, 185

Geography of Crime in China since the Economic Reform of 1978 temporal lag, 102, 138, 237-238, 241, 243, 253, 255 temporal scale, 40, 72, 141, 237, 248, 255-256, 259-260 temporal similarities, 186 temporal trends, 166 temporal variation, 10, 63, 72, 76, 100, 139 temporary resident cards, 189-190 temporary residents, 46 territoriality, 17 territory, 16 theft from car, 179, 195, 388, 408 theft of car, 179, 195, 388, 408 theft of electronic-cycle, 179, 195, 388, 408 threshold levels, 184 Tianjin Academy, 66 tight-knit communities, 38 time control, 230 time lags, 104 time scale, 101-102, 104, 141-142, 167, 241, 244-245, 253 time slot, 63, 109, 182, 186, 219, 222, 225, 229, 252, 408 time-consuming, 56 time-series, 10, 41, 47-50, 73-75, 81-82, 104, 111-112, 124-126, 149, 154, 259, 280-285, 289, 291, 300 towns, 44, 189, 230-231 traditional correlation analysis, 48 traditional cultural values, 31 traditional OLS regression, 82-83, 297 traditional social control, 31 traffic capacity, 44 transcribed, 54, 175 transform, 2, 43, 74, 96, 125, 143, 154, 244 transit crime, 20 transition, 2-3, 20-21, 30-31 transport interchanges, 219, 252 transportation, 145, 244 travel costs, 21 travel path networks, 27

431

trust in others, 29, 55, 183 trust in the polity, 60-61 t-test, 47, 49, 55, 92, 113, 289, 294, 364, 396 turnover, 19, 65, 201, 234, 250 unattainable goods, 23 unavailability, 38, 147 underdeveloped, 2 underestimation problem, 68-69 underreporting, 31, 67-68, 76, 100, 103, 167, 220, 256, 258 undeveloped areas, 43, 145 unemployment, 3, 23, 31-32, 36-37, 59, 66, 71-72, 75, 90, 101, 110, 129-130, 132-133, 136, 140, 147, 158, 160, 163, 180, 241, 243, 246 unidirectional, 49, 84-85, 127, 162163, 289 uniform pattern, 119 unit root test, 48, 82-84, 88, 125-128, 154-156, 281-285, 291-292, 325328, 331, 362, 365, 368, 371 United Nations Interregional Crime and Justice Research Institute (UNICRI), 66 universal model, 142 unofficial crime data, 55, 181, 183, 186, 219, 229-230, 252, 256, 260 unofficial data, 67, 170, 181, 220225, 229 unrecorded, 31, 190 unreported, 53, 62, 64, 170 unsafe feelings, unsafe places, 14, 63, 183, 213-214, 218-219 upper limit, 180, 191 urban areas, 1, 3, 5, 23-24, 30-33, 74, 90, 132, 136, 239 urban inequality, 98, 102, 128 urban instability, 100 urban population, 4, 59, 89 urbanisation, 4, 8, 21, 23, 32, 35-37, 39, 59, 74, 90, 101, 125, 129, 142-143, 160, 163, 186, 235, 243-246, 286 vandalism, 219

432 VAR model, variance decomposition, 49, 76, 9799, 101, 125, 132, 142, 154, 160, 289, 298, 300, 310-324 variance-covariance matrix, 71 VDC, 49, 97, 100, 132, 160, 289, 298-299 VECM model, 93, 95-96, 154, 287, 295-296, 298, 357, 359-361 vector autoregressive (VAR) model, 49, 158, 286 vector error correction model (VECM), 49, 76, 92, 131, 158, 287, 299 vice versa, 49, 51 victimisation, 13, 16, 19, 26, 38, 63, 66-67, 76, 103, 169, 176, 178181, 183-184, 186, 194-196, 198, 200, 205, 207, 219, 226-227, 236, 249-250, 252-253, 257-258, 386389, 394 Victimisation Survey Data, 103, 258 victims, 16, 25, 38, 64-65, 189, 195196, 200, 226-227, 229, 231-232, 236, 239-240, 249 VIF (variance inflation factor), 198 violent crime, 8, 20, 23, 29, 31, 101, 142, 238 visibility, 17

Index voluntary donations, 29 voter turnout, 28-29, 35, 60 vulnerability, 38, 180, 189, 236, 252 vulnerable, 169 weakened social organisation, 22, 140 welfare system, 2, 29, 40, 60, 72 well-founded analysis, 9, 39 western developed countries, 11, 30 western ideology, 33 western industrialised countries, 11, 30 western literature, 55, 252 western work, 169 white-collar crime, 8-9 widening income gap, 90, 101 widening income inequality, 33, 39, 72, 142-143, 245 widening inequality, 239, 256 willingness to report incidents, 63, 65, 178 worker lay-offs, 77, 103, 237 working time, 184-185, 195, 197, 199-200, 208, 213 workload, 68 zero inflation, 184, 386 zoning problem, 70 Z-value, 55