【作者】 李丹；

【导师】 肖智；

【作者基本信息】 重庆大学 ， 数量经济学， 2013， 博士

【摘要】 自1978年改革开放以来，中国经济经历了30多年的高速增长，被国际社会称为“中国增长奇迹”，引起了广泛关注。与此同时，一些与快速经济增长冲突的不和谐社会现象也随之产生，如社会保障、劳动就业、收入差距，贪污腐化、治安恶化等，对经济增长产生了不利影响。本文将上述经济增长过程中并发的社会问题视为社会代价，以此描述经济增长产生的负面效应，并将其作为衡量经济增长质量的重要指标，以中国经济增长为背景，基于最小二乘支持向量机方法，从定量的角度测算社会代价对经济增长的影响。首先，从社会代价视角探讨了贫富分化的表现及特点，形成的原因以及测量的方法和指标，在此基础上对贫富分化现象作出经济学分析。选择城乡居民恩格尔系数差值、基尼系数、城市居民可支配收入/农村居民纯收入指标来描述贫富分化程度，人均GDP及GDP不变价增长率指标描述经济增长状况。通过本文提出的最优参数支持向量回归方法，构造贫富分化对经济增长影响关系模型。实证结果表明：贫富分化初期，收入差距的增加产生激励作用，促进经济增长；贫富分化中期，激励作用与抑制作用并存；贫富分化发展到一定程度，激励作用让位于抑制作用；贫富分化的不稳定对经济有负向作用。其次，从社会代价视角探讨了腐败的表现及特点，对其形成的原因以及理论根源进行介绍，在此基础上对腐败现象作出经济学分析。选择腐败感知指数、贪污渎职人数/公职人员数、贪污渎职案件/公职人员数指标来描述腐败蔓延程度，人均GDP及GDP不变价增长率指标描述经济增长状况。通过最优参数支持向量回归方法，构造腐败蔓延对经济增长影响关系模型。实证结果表明：腐败蔓延对经济增长的抑制作用较为显著。政治腐败对经济增长的作用大于社会腐败整体水平；腐败蔓延对经济增长的负向作用具有非线性特征；腐败发展速度与经济增长速度早期呈负相关，且接近线性关系，但随着腐败的深入发展，其负效用开始呈现非线性特征，且边际递减。再次，从社会代价视角探讨了犯罪的表现及特点，给出了犯罪的成因理论，探讨了犯罪对于经济增长的抑制或推动作用。选择人均公共安全支出、公安机关刑事案件立案数、每万人中犯罪嫌犯人数描述犯罪状况，人均GDP及GDP不变价增长率指标描述经济增长状况。通过本文提出的最优参数支持向量回归方法，构造犯罪增长对经济增长影响关系模型。实证结果表明：对中国当前经济社会状况而言，犯罪增长不利于经济增长。犯罪对经济增长的负作用表现出非线性特点；加大人均公共安全支出有利于削弱犯罪对经济增长的不良影响；犯罪状况起伏越大，对经济的影响越大，且具有乘数效应。最后，综合测算贫富分化、腐败蔓延、犯罪增长三类因素对经济增长的共同作用。通过最优参数支持向量回归方法，构造社会代价对经济增长共同影响关系模型。认为贫富分化对经济增长具有阶段性特点，适度的放宽对贫富分化问题的约束，将有利于经济增长；腐败及犯罪则对经济具有明显的抑制作用，一旦放任腐败和犯罪继续发展，将产生更为严重的负面效果。更多还原

【Abstract】 Since the reform and opening from1978, in the process of economic growth Chinahas obtained gratifying achievements of high-speed economic growth in more than30years, which is known as “China’s growth miracle” by the international society, andarouses great attention of scholars. At the same time, we can see that somedisharmonious social phenomenon that are accompany with and conflict to rapideconomic growth phenomenon has also been arising, such as social security,employment, income gap, embezzlement and corruption, deterioration of public security,etc. These had a negative effect on the economic growth. Against the background ofChina’s economic growth, we try to measure social cost of economic growth based onleast squares support vector machine (SVM) method, from the quantitative perspective.First of all, from the perspectives of social costs we discussed the performance andcharacteristics, its cause of formation and theoretical source of the gap between rich andpoor. Based on this economics analysis of the gap between rich and poor phenomenonwas made. Selected Engel’s number difference between urban and rural residentsdepartment, the Gini Coefficient, disposable income of the urban residents/net incomeof rural resident index to describe the gap between rich and poor, and took growth rateof GDP and GDP per capita constant indicators to describe the economic growth.Structure the relationship model between the gap between rich and poor and economicgrowth through the optimal parameter support vector regression model. The resultsshow that: in the early period of the gap between rich and poor, the increase of incomeinequality will promote economic growth; in the middle the incentive effect andinhibition coexisted; to a certain extent, the gap between rich and poor developmentencouraging role in inhibition poor has a negative effect to economic instability.Secondly, from the perspectives of social costs we discussed the performance andcharacteristics of corruption, probed into the its cause of formation and introduces thetheoretical origin, on the basis of that economic analysis is made about corruption.Choose corruption perceptions index, the number of graft/office staff number, graftcase/office staff number index to describe the degree of inequality, and selectedconstant rate of growth on GDP and per capita GDP indicators describing the economicgrowth. Inhibition effect of spread of corruption on economic growth is significant.Structure the relationship model between political corruption and economic growth through the optimal parameter support vector regression model. The results show that:political corruption effect on economic growth is greater than the corruption of societyas a whole; corruption spread the negative effect on economic growth with nonlinearcharacteristic; corruption was found to be inversely associated with the earlydevelopment speed and economic development and showed the nearly linearrelationship. But with the further development of corruption, its negative effects beginto present nonlinear characteristics with diminishing marginal.Again, from the perspectives of social costs we discussed the performance andcharacteristics of crime, present the causes of crime theory and explained how the crimeinhibit or promote economic growth. Selected spending of public security per capita,criminal case of registered by public security organs, suspect and crime number index inevery ten thousand people to present the crime state and growth on GDP and per capitaGDP indicators describing the economic growth. Structure the relationship modelbetween crime growth and economic growth through the optimal parameter supportvector regression model. The results show that: as to the current economic and socialsituation in China, crime growth is not conducive to economic growth. The negativeimpacts of crime on economic growth showed nonlinear characteristics; Increasingpublic safety spending per capita is helpful to weaken the adverse effect on economicgrowth; The bigger ups and downs of criminal status was, the greater the impact on theeconomy would be, and it had a multiplier effect.Finally, there was a comprehensive measure about the interaction of three kinds offactors including the gap between rich and poor, rampant corruption, crime growth oneconomic growth. Structure the relationship model between social cost and economicgrowth through the optimal parameter support vector regression model. The gapbetween rich and poor had phased characteristics on economic growth, the widening ofmodest constraint on the gap between rich and poor would be conducive to economicgrowth; Corruption and crime, with a clear inhibitory effect on economic growth andonce allowed to be developed further, would have more serious negative effect.更多还原