中国转型期收入差距与刑事犯罪的动态变化研究

发布时间：2018-06-11 13:39

本文选题：收入差距 + 刑事犯罪　；参考：《浙江大学》2010年博士论文

【摘要】： 30年的改革开放,我国的经济建设取得了举世瞩目的成就,人民生活水平持续提高。经济快速增长的同时带动了远比经济增长本身复杂的社会转型。本论文研究关注我国经济社会转型过程中的犯罪率上升,并将其置于经典犯罪经济学分析框架内加以考察。在详细梳理相关文献和典型事实描述的基础上,本文推导了收入差距、城市化以及执法威慑等变量对犯罪参与的比较静态；并利用我国转型期分省面板数据和国家层级的时间序列数据,对相关假说做了详细的实证检验。综合起来,我们实证研究的发现包括以下几个方面。首先,相对收入差距扩大和绝对收入差距扩大,共同导致我国转型期刑事犯罪率的快速攀升。基于1988年到2007年的分省面板数据,固定效应模型的估计表明,相对收入差距对刑事犯罪率的弹性为0.34到0.45；而绝对收入差距每扩大1%将至少导致刑事犯罪率上升0.37%到0.52%。由于我国分省层面广泛存在不可观测的经济文化差异,可能与省级层面的犯罪率相关,同时时间维度历次“严打”运动和刑事政策调整也对犯罪供给构成异常冲击,因而无论单独控制省份固定效应还是时间固定效应,都将导致收入差距对犯罪率的影响被低估。其次,城市化进程加快也是我国转型期犯罪率上升的一个重要原因。和基于发达国家样本的研究发现不同,我们发现城市化进程不仅通过潜在犯罪获利机会的集聚增加犯罪参与激励,同时城市化进程中低收入群体收入提高和劳动力市场状况改善也将有利于犯罪率下降。同样基于1988年到2007年的分省面板数据,我们的实证分析显示,城市化水平在50%及以下时,城市化的犯罪率增加效应占优,而城市化水平超过这个分界点之后,犯罪率的减少效应开始发挥更大作用。以上估计在不同城市化指标设定中保持稳健,这意味着基于发展中国家样本的犯罪供给方程的估计,如果未加入城市化变量的二次项,很可能导致城市化效应的估计结果有偏误。再次,收入差距和执法变量对不同类型犯罪的影响存在长短期差异。基于1981年到2007年国家层级的时间序列的协整分析,我们发现盗窃犯罪、侵财犯罪和总犯罪率与收入差距及犯罪威慑变量存在长期协整关系,相反抢劫、伤害以及暴力犯罪序列与以上两个变量则不存在长期协整关系。收入差距与刑事犯罪破案率对盗窃犯罪、侵财犯罪及总刑事犯罪率的长期弹性,大约分布在1.88到2.13和-0.93到-1.14之间。基于ECM的短期分析表明,虽然短期内收入差距扩大犯罪率可能没有立即上升,但是由于二者之间存在长期均衡关系,所以短期内偏低的犯罪率将逐渐被修正到长期均衡水平。从犯罪治理看,以上估计意味着犯罪率的长期下降有赖于低收入群体生存条件的持续改善。最后,刑罚威慑是我国转型期犯罪率动态变化的另一个重要影响因素。由于执法变量和犯罪率之间存在典型的联立内生问题,直接估计二者关系将导致威慑效应被低估。本文利用1983年和1996年的两次“严打”运动生成的执法变量的外生变动,在动态面板数据的广义矩估计(GMM)框架内,对我国转型期刑罚威慑与犯罪率关系展开了详细的定量分析,我们的估计结果表明,犯罪威慑弹性大致分布在-0.45到-0.70之间；另外,由于我们的估计控制了惩罚概率,所以“严打”年份虚拟变量显著为负意味着惩罚强度增加同样存在犯罪威慑效应。相比国内已有的时间序列研究,本文基于分省面板数据的实证研究不仅样本量大大扩展,而同时控制省份固定效应和时间虚拟变量,也使得犯罪供给方程的估计较少受到不可观测的异质性和度量误差的影响,因而估计结果的稳健性显著改善。另外,和国际比较研究相比,中国分省经济社会文化差异较大,但又共享相同的司法体系,因而我们的估计也为犯罪经济学提供了来自发展中国家的新证据,对我国转型期刑事政策的制定也有相应的参考价值。
[Abstract]:After 30 years of reform and opening - up , our country ' s economic construction has achieved remarkable achievements , and the living standard of the people has improved continuously . At the same time , the rapid growth of the economy drives the social transformation which is far more complicated than the economic growth itself .
Based on the data of the provincial panel data and the time series data of the country level in China , a detailed empirical test is made on the relevant hypothesis . The findings of our empirical study include the following aspects .

Firstly , the widening of the relative income gap and the widening of the absolute income gap leads to a rapid increase in the rate of criminal crime in China ' s transitional period . Based on the data of the provincial panel data from 1988 to 2007 , the estimation of the fixed - effect model shows that the relative income gap has an elasticity of 0.34 to 0.45 with respect to the criminal offence rate ;
The absolute income gap will at least result in a rise of 0.37 % to 0.52 % of the criminal offence rate by 1 % . Due to the widespread non - observable economic and cultural differences at the provincial level , it may be related to the crime rate at the provincial level . In the meantime , the time dimension and the criminal policy adjustment also make an abnormal impact on the supply of the crime , so that the influence of the income gap on the crime rate will be underestimated whether separately controlled by the fixed effect or the time fixed effect .

Secondly , the acceleration of urbanization is also an important reason for the increase of crime rate in China ' s transitional period . Based on the research of developed country samples , we find that urbanization not only increases the crime participation incentive through the agglomeration of potential crime profit opportunities , but also the lower income group ' s income and the improvement of labor market situation in the urbanization process will benefit the crime rate .

Based on the co - integration analysis of the time series between 1981 and 2007 , we find that the long - term relationship between the crime of theft , the crime of invasion and total crime , the gap between the total crime rate and the income gap and the crime deterrence variable , and the long - term equilibrium relationship between the income gap and the crime rate of the crime of the crime of the crime of theft , the crime rate of crime and the crime rate of the crime of the crime of crime can be corrected to the long - term equilibrium level . From the perspective of crime control , the above estimation means that the long - term decline of crime rate depends on the continuous improvement of the living conditions of low - income groups .

Finally , the deterrent effect of penalty is another important factor in the dynamic change of crime rate in China ' s transitional period .
In addition , since our estimates control the probability of punishment , the significant negative means of the " severe " year virtual variable means that the increase in the penalty intensity also has a crime deterrent effect .

Compared with the existing time series research in China , the empirical research based on the data of the provincial panel not only greatly expands the sample size , but also controls the provincial fixed effect and the time virtual variable , so that the estimation of the crime supply equation is less affected by the non - observable heterogeneity and the measurement error , so that the estimation result is obviously improved . In addition , compared with the international comparative study , the economic and social culture of China is relatively large , but the same judicial system is shared . Therefore , our estimation also provides new evidence from developing countries for the crime economics , and has a corresponding reference value for the development of the criminal policy in China ' s transition period .
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2010
【分类号】：D631.2;F124.7

【参考文献】