【作者】 黄金辉；

【导师】 胡善联；

【作者基本信息】 复旦大学 ， 社会医学与卫生事业管理， 2007， 博士

【摘要】 一、研究背景和研究目的城市化水平高低是一个国家经济和社会发展水平的重要标志。城市化是实现工业化的产物，各国在工业化中，必然会经历传统农业部门的剩余劳动力转移到现代工业部门、农村人口转化为城市人口的过程，这其实就是城市化的过程。与国外相比，我国的城市化在概念上更加广泛，城市不但包括传统意义上的城市，还包括城镇。根据1990年颁布的《中华人民共和国城市规划法》，这里的镇是指规模不及城市，但具有城市的性质和功能，具有一定城区面积和人口数量、人口密度，为大部分从事非农业生产的人群所集居的介于农村和城市之间的社区，它包括国家已经批准的建制镇和尚未设建制镇的相对发达的农村集镇。改革开放以来，中国发生了翻天覆地的变化，农村人口转移到城镇经历了一个渐进过程，由于户籍制度、社会保障制度、农村土地承包制度的约束，以及乡镇企业就地发展的客观现状，农村人口有相当部分是就近转移到小城镇，从事二、三产业。发展小城镇是我国推进城镇化的重要途径，因此，国内很多学者都称中国的城市化为“城镇化”。现在我国的城镇化，从地域上其实主要是小城镇的迅速发展。根据国家建设部颁布的城市建设年度统计公报和村镇建设年度统计公报，自2000年到2005年，城镇化水平上升了7个百分点，城镇人口增长了0.81亿，其中，同期设市城市人口增长约0.01亿，小城镇人口增长0.80亿。城镇人口增长的主要人口流入区是建制镇与集镇。对于建制镇与集镇，国内学者也称之为“小城镇”或者“新城镇”。由于当前小城镇的人口总数与人口密度、规模、面积、经济水平、模式等都在不断地变化当中，而其趋势是慢慢向标准的城镇转化，是正在形成中的新城镇，因此，在本研究中也称之为新城镇。新城镇是我国工业化、城市化过程中出现的具有时代特色的现象，是农村向城市的过渡地区，也是城市和农村的结合体。不同于欧美发达国家在推行工业化城市化的时候采取的是技术自然演进的道路，由于我国现在正处于产业结构调整、国有企业改革阶段，城市本身就有上千万的剩余劳动力，因此，农村居民目前无法进入城市的传统就业市场。近些年来，小城镇人口的就业主要集中在乡镇企业和私营企业中。因而，新城镇的居民不能自然而然地纳入企业、社会的各种福利和保障范畴。我国的医疗保障体制，主要由城镇的职工基本医疗保险制度和农村的新型合作医疗组成。其中，职工基本医疗保险覆盖对象以城市居民为主，农村合作医疗则以农村居民为主。新城镇中的大部分居民一方面不能享受城镇职工基本医疗保险，另一方面也不能享有合作医疗，对于这部分居民的医疗保障，我国的医疗保障体系出现了制度上的缺失。与此同时，城镇的医疗服务价格也高于农村。这些情况都会影响了新城镇居民的卫生服务利用，降低了疾病经济风险的抵御能力。随着小城镇的发展，新城镇中的居民会越来越多。本研究通过对新城镇居民的卫生服务需要、需求、医疗保险需求评估，结合其经济和社会学特点，旨在探索对新城镇居民的医疗保障模式。二、研究方法在新城镇居民的卫生服务利用研究中，主要通过构建卫生服务需求模型，来探索需求影响因素及其强度，所采用的方法是1998、2003年国家两次卫生服务调查数据分析时所采用的四步模型法；在研究新城镇居民的医疗保障影响因素时，为了同时考察多种不同类型医疗保障模式的相互关系和影响因素，选用了多分类logistic模型来进行分析；在进行新城镇居民的医疗保险意愿购买与意愿支付研究中，首先需要设计明确的保险方案，而后采用Willing-To-Pay(WTP)中的Contingent valuation method(CVM)方法来调查其意愿支付能力，在调查中应用“Take-it-or-leave-it”的方式来询问调查对象，获得意愿购买和意愿支付数据后建立医保的需求模型，并用二分类的随机效应logistic回归模型来进行拟合。各类方法和指标的含义及其计算详见论文相关部分。三、资料来源本研究的数据包括两部分，第一部分是卫生部的《2003年中国卫生服务调查研究》，选取了其中有关卫生服务利用、医疗保险等方面的数据。第二部分主要由调查员入户调查的方法来获取。本研究选取了山东省、四川省作为样本省，每个省选取2个城镇，采用了分层、整群、随机抽样相结合的抽样方法来选择样本。共调查了2，671户，9，478人。四、研究结论(一)新城镇居民的疾病经济负担比较重新城镇居民的次均门诊费用约163.2元，介于2003年城市和农村水平之间，新城镇的次均门诊费用为农村水平的1.79倍；新城镇居民的次均住院费用约3，800元，也介于城市和农村之间，是农村水平的1.43倍。新城镇居民的医疗费用主要由自己承担，门诊费用的报销比例不到3％，住院费用占住院总费用的报销比例不到22％。年个人支付的医药总费用约为316.8元，占年收入的6.2％。新城镇居民所承担的疾病经济负担较为严重。(二)新城镇居民的高疾病经济风险，其原因在于缺乏风险分担机制在全部15岁及以上的调查对象中，有近72％的人口没有覆盖任何保险。由于疾病的治疗一般是家庭而非纯粹的个人行为，因此本研究主要测算家庭的疾病经济风险。当家庭有成员患病时，家庭年医疗自费支出占其收入的比例达到三分之一左右，占家庭年均支出的比例更高达45％，已经达到了WHO所界定的灾难性支出家庭的标准。通过分析新城镇居民所支付的医疗费用可以发现，高疾病经济风险的原因在于医疗费用的分担作用太低：门诊医疗费自付比高达97％以上、住院医疗费用自付比超过78％。如果假定门诊和住院费用占医疗总费用的比例各约50％，则在新城镇居民的医疗风险分担机制下，居民自己需要承担的医药费占总医药费的比例在88％左右。(三)新城镇居民医疗保险的筹资可行性和意愿购买通过构建模型后，本研究预测了年人均保费的中位数：重病大病保险方案、住院医疗保险方案和门诊医疗保险方案的年人均预测保费分别为110.1元、100.2元和72.4元，分别占年人均收入的2.2％、2.0％和1.4％。因此，对于重病大病保险和住院保险的意愿支付额占个人收入的筹资比例达到了4.2％，刚好相当于我国城镇职工基本医疗保险中统筹基金的筹资比例。由于本研究所估算的是个人的意愿支付水平，如果考虑到雇主支付水平，则重病大病和住院保险的筹资潜力可能会大于个人收入水平的4.2％，即大于当前的城镇职工基本医疗保险统筹基金的筹资水平。新城镇居民对个人重病大病医疗保险、住院医疗保险和门诊医疗保险的真实意愿支付分别是年人均125元、141元和88元。而新城镇居民的年人均住院总费用和门诊总费用分别为140.8和185.5元，经偿付比例换算后为140.8和111.3元，这就是纯保费。对于门诊医疗保险而言，由于其意愿支付的价格低于纯保费，从基金平衡的角度还没有公司愿意来提供这种保险。对于住院医疗保险而言，意愿支付的保费已经大于纯保费，因此会有公司愿意提供这种保险。如果考虑用人单位的责任。从这个角度，至少对于住院保险而言，具有相当大的筹资潜力，再配以良好的风险管理措施，应该能足以保证保险建立后的基金平衡。(四)住院医疗服务是新城镇居民医疗保险所保障的优先覆盖范畴从理论上，医疗保险应该首先覆盖住院服务，而非门诊服务。在本研究中，则通过构建模型，测算出了不同医疗保险项目的价格弹性，从低到高依次为：重病大病保险、住院医疗保险和门诊医疗保险。在研究所测算从参保率和意愿支付水平上，都反映出调查对象对住院保险的偏好：其参保率比门诊医疗保险高了13个百分点，年人均保费则高出近一倍。因此，无论是从新城镇户主的保险偏好，还是其意愿支付水平，还是资源配置，新城镇居民的医疗保险都应该首先覆盖住院医疗服务，而非门诊医疗服务。(五)家庭参保是新城镇居民医疗保险保障对象的优先选择在推出新城镇居民医疗保险时，保障对象可能会有两种选择：个人，以及家庭。当前的城镇职工基本医疗保险，保障的对象就是个人。本次研究，探索了家庭医疗保险的意愿购买和意愿支付情况。结果发现，个人在各个医疗保险方案中的参保率都比相应的家庭参保率高出近2-4个百分点；同时，家庭医疗保险的意愿支付水平是相应个人医疗保险意愿支付的2.3-3倍。理论上，虽然家庭参保率比个人参保率低了几个百分点，但由于家庭平均人口数为3.55人，意味着以家庭为单位参保的投保人总数要远远高于以个人为单位的投保人数目，而且能降低逆向选择。然而，由于家庭参保的意愿支付水平不到个人意愿支付水平的3倍，所带来的问题就是：筹资水平是否足以维持基金平衡。本研究测算的是家庭的意愿自付水平，故在实际保险费的筹集中，还应该考虑雇主的责任。由于新城镇居民中有大量的儿童，因此也可以考虑由政府给予一定的资助，来获得按家庭参保所需的足够保险费。(六)政府组织的医保是新城镇居民医疗保险模式和组织方式的优先选择当前的城镇职工基本医疗保险体系，在短期内并不一定能适用于新城镇居民，需要建立新的适合其特点的医疗保险。新城镇医疗保险的组织，有几种选择：政府组织、社会团体组织或商业保险公司组织，三者各有优劣。政府组织和社会团体组织能保证公平，商业保险公司运作则能体现效率。国外也各有这三者的成熟经验。然而，在调查中发现，近一半的调查者都偏好政府组织，认为这种方式值得信赖。因此，新城镇居民的医疗保险，原则上可以考虑由政府来组织，以体现政府责任，校正保险市场的失灵。更多还原

【Abstract】 1. BackgroundUrbanization is an inevitable process along with China’s industrialization. Thesuperfluous rural labor force will swarm into urban areas when China is transformingfrom agricultural to industrial society. The urban areas include cities and towns inChina. In "Low of Chinese City Layout", town is defined as the area where people areliving between city and rural areas, and have the similar function to cities. It includestown permitted by government and areas with relatively high-density population.China is accelerating its urbanization through industrialization and economicdevelopment. As there were strict restriction to move to Chinese cities, more andmore superfluous rural labor force migrated from rural areas to townships. Most ofmigrants were employed by private sector in second industry or third industry.Compared to’ rapid development in urbanization in developed countries, Chineseurbanization is experienced a boomtown. Then, Chinese urbanization is called"urbanization and boomtown" by some researchers.There was about increased 81 million populations who lived in Chinese urban areasfrom 2000 to 2005. The level of urbanization is raised by 8%. However, populationmigrated to cities and towns are 1 and 80 million, respectively. It seems that towns aremain living areas for migrants from rural areas.Cities and towns are experiencing the changes in many factors in urbanization, such as population, population-density, scale, region, and economic level, etc. Township istransferring from small scale living area to city. It is a boomtown-type city.Chinese health security system is composed of urban employee’s basic healthinsurance system and rural new cooperative system (NCMS). The beneficiaries ofurban employee’s health insurance were employees and retirees. It is financed mainlyfrom employers. NCMS covers the population living in rural areas. As citizens intownships are employed by private sectors with small scale, most of them can not beinvolved in urban social health insurance. Moreover, NCMS can not cover migrantsfrom rural areas to townships. Then, most of the citizens in townships didn’t coveredby any health insurance. The low coverage of health insurance lows the risk-pooling,and decreases the utilization of health services,This study aims to: estimate the need and demand of health services and healthinsurance in Chinese townships; explore the determinants for health care and healthinsurance; make the health security model for citizens in townships.2. MethodologyFour-part model applied in "national household study" in 1998 and 2003 will be usedto estimating health care need and demand for health care. This model will also beused to explore the determinants. To identify the factors affecting health insurancecoverage, multinomial logistic model will be applied. The contingent valuationmethod (CVM) was used to estimate the willingness-to-pay (WTP) for healthinsurance program. Take-it-or-leave-it was used in questionnaire survey.3. Data sourceData used in this study came from field survey and second-hand data. Firstly, data abouthealth care need, health care utilization, and health insurance coverage from national household study in 2003 was collected. The study is also based on the data obtained from a household surveyconducted in four townships in two provinces-Sichuan in the southwest and Shandong in thenortheast. One district was sampled in each city. Within each district, about 550 households wererandomly selected to interview. Approximately 9,478 individuals from 2,671 households wereresponded.4. Results4.1 Economic burden of disease in township citizens.The average expenditure per clinical visit in township was about￥163.2, which is lowerthan that in urban, but as 1.79 times higher than that in rural. An inpatient expenditure in townshiplevel was about￥3,800, which is also lower than that in urban, but as 1.43 times higher thanthat in rural. The reimbursement ratio for ambulatory and inpatient services expenditure are lessthan 3%, and 22%, respectively. Each citizen should pay about￥316.8 for health care servicesper year, accounting for about 6.2% of annual income. It is proved that the economic burden ofdiseases for township citizen was rather heavy.4.2 Poor risk-pooling is one of the reasons on higher economic burden of diseaseAn estimated, 72% of sample individuals aged over 14 years old did not have any healthinsurance. The economic burden of diseases at household level is accounted for more than onethird of household income, if any member suffered from disease in the past year. When health careexpenditure contributed to 45% total household expenditure is regard as catastrophic expenditurehousehold by WHO. One possible reason for those catastrophic expenditure household was poorrisk-pooling and risk-sharing system. The co-payment rate for outpatient services and inpatientservices were over 97% and 78%, respectively. The out-of-pocket ratio is about 88% to totalhealth care expenditure.4.3 Demand for health insurance in townshipsThis study surveys willingness-to-pay for different health insurance programs: major catastrophicdisease insurance (MCDI), inpatient expense insurance (IEI), and outpatient expense insurance (OEI). The median annual premium for these three programs was estimated through modeling. Asestimated,￥110.1 will be paid for one year MCDI, compared to￥110.2 for IEI and￥72.4 forOEI. The expected premium of MCDI is accounted for about 2.2% annual income, which is higherthan that of IEI and OEI (2.0% vs. 1.4% respectively). The individual’s willingness-to-pay forinpatient expense insurance program and major catastrophic illness insurance program was about4.2% annual incomes. The financing potential for health insurance may be higher than 4.2% ofincome if the employee’s financing level is considered.People are willing to pay more money for MCDI (￥141 per year) then for IEl (￥125 per year)and OEI (￥88). As the average expenditure for inpatient services and outpatients services were￥140.8 and￥185.5, respectively. If insured provided IEI and OEI should pay about￥140.8and￥111.3 per year, respectively. So insurance companies will prefer IEI but deny OEI in orderto get net profit.4.4 Inpatient services are the priority setting in health insurance benefit package for townshipcitizen. From health insurance theory, the medical insurance benefit package should includeinpatient health services instead of outpatient services. In this study, the price elasticity to demandfor OEI program is higher than that for OEI. In addition, about 13% of more people want to buyIEI than OEI. The annual average premium for IEI is 2 times higher than that for OEI. It seemsthat health insurance program should be covered inpatient services at first.4.5 Family health insurance is preferred to individual health insuranceThere are two possible health insurance models for township citizens: individual health insurancemodel and household health insurance model. In this study, demand for individual health.insurance programs and family health insurance programs were surveyed respectively. Morepeople are willingness-to-buy individual program rather than household programs. Thewillingness-to-pay for family health insurance program is 2.3- 3.0 times higher than that forindividual health insurance program. Even if the willingness-to-buy ratio of household programs are 2%-4% higher than that of individual programs. Family insurance programs will cover moreinsured than individual programs. Compared to family health insurance program, individual healthinsurance programs will be more difficult to control the adverse-selection.As the family health insurance scheme will cover some vulnerable population, such as children,unemployed, etc. government should provide financial support for this program.4.6 Government-oriented model is preferred in health insurance administration in townshipsAs the different social-economic characteristics between urban and township, current urbanemployee’s basic health insurance system can not cover most population who are living in town inthe near future. Township needs to establish a new health insurance system based on its economiclevel and development level. There are three possible models for township health insurance: Thefirst is government oriented; the second is non-government organizations (NGOs) oriented; thethird is private health insurance. In this study, more than half of sample individuals would preferto government-oriented model.更多还原