基于Panel Data Model的卫生费用总量与构成模拟与预测
|
摘要
一、研究目的和意义
医药卫生事业关系亿万人民群众的健康,关系千家万户的幸福,是重大民生问题。然而,当前看病贵、乱开药、拉大网检查等百姓抱怨问题愈演愈烈,药品市场的“高价药易销”问题仍然突出。课题组前期研究发现,医疗费用负担重、看病贵问题是医疗改革的需优先解决的问题。
“看病贵”问题对社会造成了极其严重的后果,一方面造成卫生费用过快增长,直接加重了各类医保制度收支平衡,影响各类医保制度的平稳运行。另一方面,卫生费用的过快增长和医疗保障水平较低则进一步加重了居民医药负担,使得“看病贵”问题日益严重,威胁居民卫生服务可及性,因病致贫问题突出。围绕着“看病贵”问题,即医疗费用负担问题,医疗卫生相关利益团体相互埋怨,陷入失衡状态：医疗机构埋怨政府财政投入不足、收费标准过低等补偿机制问题,医疗保障部门担心医疗费用快速上涨、基金收支失衡,当看病贵等问题与医改挂钩,政府则担忧看病贵问题危及社会和谐稳定。
政府对此给予了前所未有的重视,《中共中央国务院关于深化医药卫生体制改革的意见》和《医药卫生体制改革近期重点实施方案(2009—2011年)的通知》相继出台,从宏观上确定了我国医改的基本方向、思路和政策框架,对于保障我国医药卫生事业持续健康发展,维护广大人民群众健康具有重大意义。
但是,如何进行改革,如何真正解决社会反响激烈的问题,实践关注民生的价值观?本文认为,透视卫生费用过快增长和控制卫生总费用这一社会热点,3个科学问题值得进一步思考：(1)卫生费用总量及构成的影响因素是什么?(2)将来中国医疗费用的走势如何?如何进行科学预测?(3)探讨适宜中国国情的卫生费用总量及构成的预测与控制模型,据此为研制实践中从根本上一揽子解决中国医疗卫生领域现有突出问题的策略思路提供依据。
因此,本研究的目的在于明确卫生费用总量与构成影响因素的维度及其指标体系,利用10省的Panel Data数据进行模拟确定卫生费用总量、构成及其影响因素的定量关系,在此基础上,对我国2015年卫生费用总量与构成进行预测,并比较其与国际普遍经验之间的差距,为我国医改提供参考依据。
二、材料与研究方法
本研究以卫生系统宏观模型为指导性研究方法,以Panel Data模型、规范差距分析等为应用性研究方法,在总研究项目前期研究的基础上,形成我国卫生费用总量与构成定量研究的框架和思路。
本部分定量研究搜集的资料主要来源于自于面上统计资料,包括2002-2010年的中国统计年鉴、中国卫生统计年鉴、中国人口和就业统计年鉴、卫生事业经费决算资料以及2009、2010年卫生部卫生经济研究发行的中国卫生总费用研究报告。将搜集数据进行整理后形成分析库,对数据进行单位根检验、数据转换、滞后期数确定,用Panel Data模型对卫生费用总量、构成与其影响因素之间的关系进行定量模拟与预测。
三、主要研究结果
(一)确定卫生费用总量及构成影响因素
运用“卫生总费用”、“影响因素”、“个人卫生支出”等关键词及其组合,在数据库“中国期刊全文数据库”搜索1979年至今文献1257篇,查阅的结果显示影响因素众多,有GDP、65岁以上人口比例、0-14岁人口比例、政府预算卫生支出、社会卫生支出、人口数、财政收入、财政支出、医生数、婴儿死亡率、城镇化集中程度、卫生事业费、城镇居民基本医疗保障计划、居民消费价格指数、中西药品及保健用品价格指数、每千人口医疗机构床位数、卫生服务技术密集程度、住院病人人均医疗费、农村居民人均医疗消费水平、城镇居民人均医疗消费水平、居民家庭人均可支配收入、居民消费水平、农村居民消费水平、城镇居民消费水平、城乡居民人民币储蓄存款年底余额、期望寿命、慢性病、社会保险覆盖率。影响卫生费用总量及构成的因素遍及结构、过程、结果以及包括人口需要、社会经济、政治等在内的外环境,且主要集中在外部子模,尤其是社会经济与人口需要维度。
依据专家咨询与论证确定的全面性、科学性、可评价性、系统性和代表性原则,最终确定影响总量和构成的主要变量维度及指标为：社会经济(人均GDP、城镇居民家庭人均可支配收入绝对值、农村居民人均纯收入绝对值)、人口(总人口、65岁以上人口比例、0-14岁人口比例)、结构(政府预算卫生支出占财政支出的比例)、过程(每门诊人次收费水平、每床日平均收费水平)、结果(孕产妇死亡率、围产儿死亡率)
(二)模型构建过程与结果
1、模型构建过程
在确定卫生费用总量及构成影响因素的基础上,对变量进行单位根检验、数据变换和滞后期数确定,为模型构建做准备。单位根检验结果发现,变量“个人现金卫生支出占卫生总费用的比例”、“总人口”、“孕产妇死亡率”、“围产儿死亡率”、“0-14岁人口比例”、“65岁以上人口比例”、“每门诊人次收费水平”、“每床日平均收费水平”等变量为平稳的序列,可直接纳入回归方程；而“卫生总费用占GDP的比例”、“人均GDP”、“城镇居民可支配收入”“农村居民人均纯收入”、“政府卫生支出占财政支出的比例”为非平稳序列,不能直接纳入回归方程,对其进行数据变换后纳入回归方程。
人均GDP对卫生费用总量与构成存在滞后效应。根据Akaike与Schwarz信息准则,滞后一期的人均GDP的Akaike与Schwarz函数值最小,因此确定人均GDP对卫生总费用占GDP的比例存在滞后一期的影响。
在此基础上,根据F统计量,确定模型形式是变系数模型、变截距模型、还是混合回归模型；根据Hausman检验结果,确定选择固定效应模型还是随机效应模型；继而运用panel data数据模型方法,构建卫生费用总量与构成及其影响因素的模拟回归方程。
2、模型模拟结果
使用固定效应模型模拟的卫生总费用占GDP的比例及其影响因素的关系如下：
其中,αi*为个体截面基准值之间的区别,LNH为对数变换后的卫生总费用占GDP的比例,LNGDP为对数变换后的人均GDP,LNGDP (-1)为滞后一期的对数变换后的人均GDP,LNO为对数变换后的政府预算卫生支出占财政支出的比例,MM为孕产妇死亡率,OL为65岁及以上人口所占比例,FF为每门诊人次收费水平。
模拟结果显示：卫生总费用占GDP的比例与滞后一期的人均GDP、政府预算卫生支出占财政支出的比例、65岁以上人口所占比例、每门诊人次收费水平呈正向关系,与当期人均GDP、孕产妇死亡率呈负向关系(R2为0.95)。
使用随机效应模型对个人现金卫生支出占卫生总费用的比例及其影响因素的关系模拟,方程形式为：
其中, OOP为对数变换后的个人现金卫生支出占卫生总费用的比例,LNGDP为对数变换后的人均GDP,LNGDP (-1)为滞后一期的对数变换后的人均GDP,LNO为对数变换后的政府预算卫生支出占财政支出的比例,MM为孕产妇死亡率,PE为总人数,FB为每床日平均收费水平。
模拟结果显示：个人现金卫生支出占卫生总费用的比例与政府预算卫生支出占财政支出的比例、每床日平均收费水平呈正向关系,与其它变量的关系无统计学意义(R2为0.81)
(三)卫生费用总量与构成预测及与国际普遍经验间差距
根据构建的Panel data模型,结合预测的2015年的相关数据,对2015年中国卫生费用总量和构成进行预测。根据模型预测,2015年卫生总费用占GDP的比例为6.5%。结合课题组前期研究,筹资公平性优于中国的113个国家样本的“卫生总费用占GDP比例”模型预测结果,按照国际普遍经验2015年中国卫生总费用应达到GDP的6.0%。两者相比,说明基于现实的模拟预测值高于国际较优国家普遍经验值,达到较为适宜的比例。研究得出的基于现实模型的结果提示,2015年个人现金卫生支出占卫生总费用的比例为35.7%,低于2015年根据国际普遍经验值38.5%,个人负担将低于国际普遍经验,出现较好的势头。但是需要注意的是,这是在当前政府加大预算卫生支出,政府预算卫生支出增长速度大大超过财政支出增长的情况下必然发生的结果,若政府预算卫生支出增长幅度不能确保最近年份的增速势头,则要实现国际普遍经验的目标估计还是有待观察的。
四、结论
(一)我国卫生总费用占GDP的比例较低,不同地区之间存在较大差距
中国卫生统计年鉴中数据显示,2001-2009年,卫生总费用的年均增速为16.6%,高于GDP的增长速度(15.2%)。2007年我国卫生总费用占GDP的比例仅为4.3%,较2000年的4.6%有所下降,在191个国家和地区中仅列第149位。即便是财政增加投入的2009年,也仅为5.1%。在世界各国卫生总费用占GDP的较高的比例和逐年增加的趋势下,中国卫生费用总量保持在相对低的水平,一直低于5.5%,这表明中国卫生费用并没有突破社会的支付能力,仍然有较大的增长空间。
尽管按照当前发展趋势,2015年卫生总费用占GDP的比例将达到6.5%,高于世界普遍经验确定的6.0%,但是需要注意的是,这是在当前政府加大预算卫生支出,政府预算卫生支出增长速度大大超过财政支出增长的情况下发生的结果,若政府预算卫生支出增长幅度不能确保最近年份的增速势头,则要实现国际普遍经验的目标估计还是有待观察的。因此,保持卫生费用的可持续增长是医改的关键。
与此同时,全国各地卫生总费用占GDP的比例存在较大差别。甘肃、云南、新疆等经济欠发达地区卫生总费用占GDP的比例较高,且有逐年增加的趋势。而上海、浙江、天津、广东、福建等经济发达地区卫生总费用占GDP一直偏低,一直低于5.0%。
(二)个人现金卫生支出比例依然偏高,不同地区间存在较大差距
2007年中国个人卫生支出占卫生总费用的比例依然高达55.3%,成为个人负担比例比较高的国家之一,在191个国家和地区中列第46位。即便是幅度增加财政投入的2009年,个人负担比例也达到了38.2%。明显高于2007年古巴、捷克、荷兰和英国的个人现金卫生支出占卫生总费的比例(分别为4.5%、14.8%、18.0%和18.3%)
各地个人现金卫生支出占卫生总费用的比例存在较大差别。上海、新疆、福建、甘肃等省份个人现金卫生支出占卫生总费用的比例较低,尤其是上海个人支出比例一直在30.0%以下。而吉林、黑龙江、广东等地区个人现金卫生支出占卫生总费用的比例的比例较高,一直在40.0%以上
(三)卫生事业没有享受到经济发展的平均利益
卫生费用总费用占GDP的比例受到人均GDP、政府预算卫生支出占财政支出比例、65岁及以上人口所占比例、孕产妇死亡率、每门诊次均费用等因素的影响。其中,卫生总费用占GDP的比例与当期的人均GDP呈负向关系,而与滞后一期的人均GDP呈正向关系,而且负向关系的系数要大于正向关系的系数。由于受当期GDP增长(分母)的影响,卫生总费用(分子)占GDP的比例在当期会有所下降；在GDP增长的基础上,国家加大了对卫生事业的投入,使得下年卫生总费用占GDP的比例又有所上升。但是负向关系的系数要大于正向关系的系数,表明尽管国家加大了对卫生的投入,但卫生事业并没有享受到经济发展的平均利益。
(四)提高政府卫生支出占财政支出的比例成为医改的必然选择
如果按照当前发展趋势,2015年个人现金卫生支出占卫生总费用的比例将为35.7%,要完成卫生十二五规划中提及的个人现金卫生支出占卫生总费用的比例降低到30%的目标非常困难。因此,针对影响模型因变量程度较大的自变量—一政府加大卫生投入入手具有较好的现实意义,相对于人口比例、经济发展以及收费标准等指标,切实提高政府卫生支出占财政支出的比例是较为合适的选择。
如果要使2009年的个人现金卫生支出占卫生总费用的比例降低到30.0%,按照上述模型,在其他影响因素不变的情况下,政府预算卫生支出占财政支出的比例需要由6.4%增加到14.5%,即政府需要在当前3930.7亿元的基础上追加4921.0亿元,此时卫生总费用占GDP的比例达到6.4%。
如果使2015年的个人现金卫生支出占卫生总费用的比例降低到30%,按照上述模型,在其他影响因素不变的情况下,政府卫生支出占财政支出的比例需要由预测的8.8%增加到13.4%,即政府需要在预测的17006.5亿元的基础上追加8907.0亿元,此时卫生总费用占GDP的比例达到7.7%。
1. Background and Objectives
The health care sector is a major livelihood issue, as it is closely related to the health of billions of people and the happiness of every household. However, unaffordable and unnecessary medical care has been complained more and more frequently. The problem in the drug market that expensive drugs sold well was still serious. Our earlier research found that affordable medical care should be given priority.
The problem has caused serious consequences to the society. On one hand, it caused the abnormal speed of health expenditure growth and affected the balance of health insurance; on the other hand, the over-rapid growth of health expenditure increased the burden of medical costs of the general public and aggravated the poverty caused by illness, making the problem " seeing a doctor is expensive " more serious. Medical institutions complained that the money provided to health was inadequate and the fee standards were very low. Medical security departments worried about the quick rise of medical expenditure and the disequilibrium of the fund. The governments worried that the problem of " seeing a doctor is expensive " might endanger the harmony and stability of the society.
The governments have paid much attention to the problems and promulgate policy which has oriented the development of the health system in order to solve them. But how to reform? How to solve the problem of social repercussions entirely? There are still some questions to answer:What variables affect health care expenditure and its constitution? What is the future trend of the Chinese health care expenses? How to conduct scientific prediction? How to simulate the relationship between health care expenditure and its constitution and the factors?
The objective of the study is firstly to explore the variables that affect health care expenditure and its constitution. And then use the panel data model of 10 provinces to analyze the relationship between health care expenditure and its constitution and the factors. On this basis, we use the database to predict health care expenditure and its constitution of 2015 in China. Finally, we make a comparison between the predicted expenditure and constitution of 2015 and the expenditure and constitution of the international experience.
2. Materials and Methods
The macro model of health system was selected as the principle methods, and selected the panel data model as the applied research method. The framework and thinking of the quantitative study of health care expenditure and its institution in China formatted on the basis of preliminary research of the total project.
The data used in the study was collected mainly from statistics including China Statistical Yearbook, China Health Statistical Yearbook, China Population and Employment Statistics Yearbook and health accounts funding information of the year from 2002 to 2010, and China Total Expenditure on Health Report published by China Health Economics Institute from 2009-2010.Then we collated those data collected from these statistics to be a database which could be analyzed directly. After Unit Root test, data conversion and confirming the lag period, we analyzed the relationship between health care expenditure and its constitution and the factors, and then predicted health care expenditure and its constitution of the year 2015.
3. Main Results
3.1 The factors which affected health care expenditure and its constitution
After using the keywords "health care expenditure", "influence factors" and " out of pocket spending " and their combination to search literatures in the CNKI database since 1979,we found that there were a lot of factors that affected health care expenditure and its constitution, such as GDP, population proportion of people over 65, population proportion of people below 14, the government budget health spending, social health spending, population, fiscal revenue, fiscal expenditure, the number of doctors, infant mortality, urbanization concentration, health care expenditure, basic medical security plan of urban residents, consumer price index, Chinese and western medicines and health supplies price index, medical institutions per thousand population, the intensity of health service, per capita health cost of hospital patients, per capita health consumption level of each rural resident, per capita health consumption level of each urban resident, per capita disposable real income, consumption level, rural consumption level, urban consumption level, the end of urban and rural residents RMB savings deposit balance, life expectancy, chronic disease, social insurance coverage. The factors distributed in structure submodule, process submodule, result submodule, and the external environment submodule including population needs submodule, social economy submodule, politics submodule, etc.
After the comprehensive consideration of expert consultation and the principle of comprehensive, scientific, able to be evaluated, systematic and representative demonstration,we finally decided to classified the factors which were put in the equation into 5 dimensions:economy dimension which included per capita GDP, per capita disposable real income of urban residents, per capita disposable real income of rural residents; demography dimension which included population, population proportion of people over 65, population proportion of people below 14; structure dimension which included the proportion of the government budget health in financial expenditure; process dimension which included per outpatient service charge and per hospital service charge; result dimension which included maternal mortality and perinatal mortality.
3.2 The process and result of model building
3.2.1 The process of model building
Based on the unit root test,we found that there were several stable factors which included the proportion of out of pocket spending, maternal mortality,wai births mortality, population, population proportion of people over 65, population proportion of people below 14, per outpatient service charge,and per hospital service charge. Other variables were unstable and should be transformed before putting into equation.
Per capita GDP had lagging effect to health care expenditure and its constitution. We found the lag period was 1 according to the Akaike s Information Criterion and the Schwarz Criterion. On this basis, we determined the general format of the model according to the F statistics and the influence form according to the Hausman tests. Then we built the equation of the relationship between health care expenditure and its constitution and the factors.
3.2.2 The result of the model building
The equation of the relationship between the total health expenditure in GDP and its influencing factors was simulated by fixed effect model as:
αi* is the difference of each cross section. LNH is the logarithmic transformation of total health expenditure in GDP. LNGDP is the logarithmic transformation of Per capita GDP, while LNGDP(-1) is the logarithmic transformation Per capita GDP lagged 1 period. LNO is the logarithmic transformation of the proportion of the government budget health in financial expenditure. MM is the maternal mortality. OL is the population proportion of people over 65. FF is per hospital service charge.
The result suggested that per capita GDP lagged 1 period, the proportion of the government budget health in financial expenditure, population proportion of people over 65, per outpatient service charge had a positive effect on the total health expenditure in GDP, while per capita GDP and maternal mortality had a negative effect.
The equation of the relationship between out of pocket spending and its influencing factors was simulated by random effect model as:αi* is the difference of each cross section. OOP is the proportion of out of pocket spending. LNGDP is the logarithmic transformation of Per capita GDP, while LNGDP(-1) is the logarithmic transformation Per capita GDP lagged 1 period. LNO is the logarithmic transformation of the proportion of the government budget health in financial expenditure. MM is the maternal mortality. OL is the population proportion of people over 65. FB is per ouypatient service charge.
The result showed that the proportion of the government budget health in financial expenditure, and per outpatient service charge had a positive effect on the proportion of out of pocket spending.
3.3The prediction of health care expenditure and its constitution and the gap between the prediction and the international common experience
We can predict Chinese health care expenditure and its constitution of 2015 according to the building model. Total health expenditure in GDP will be 6.5% in 2015, higher than 6.0% according to the international common experience which was drawn from the 113 countries of which the performance of Health system was better than China. The proportion of out of pocket spending of China will be 35.7% in 2015, lower than 38.5% according to the international common experience. But it should be paid attention that the result is based on the increasing investment on health these years. If the investment is decreased, the goal be achieved is not certain.
4. Conclusion
4.1TotaI health expenditure in GDP of China was still low, and there was a large gap between different regions
The annual growth speed of total health expenditure is 16.6% from 2001 to 2009, quicker than the growth of GDP 15.2%. Total health expenditure in GDP of China in 2007 was only 4.3%, ranked 149 in the 191 world countries, which is lower than that of 2000. Although total health expenditure in GDP worldwide kept a higher proportion and rising trend, China's total health care expenditure remained in a comparatively low level, below 5.5 percent all the time. This showed that Chinese health expenditure had not break the social payment ability and still has larger growth space.
Total health expenditure in GDP of 2015 will be 6.5%, higher than that of the international experience. But it should be paid attention that the result is based on the increasing investment on health these years. If the investment is decreased, the goal be achieved is not certain. Therefore the key of the health reform is to make sustainable development of the health expenditure.
Meanwhile, there was a large gap between different regions. Total Health Expenditure in GDP of underdeveloped region, such as Gansu, Yunnan, Xinjiang, was higher than the average and had a rising trend. While Total Health Expenditure in GDP of developed regions, such as Shanghai, Zhejiang, Tianjin, Guangdong, Fujian, was lower than 5.0%.
4.2 The proportion of out of pocket spending was still higher than that of the government and the social spending, and there was a large gap between different regions
The proportion of out of pocket spending of China in 2007 was 55.3%, leading China to be one of the countries in the world with the highest individual proportions, ranked 46 in the 191 world countries. The proportion of out of pocket spending was 38.2%at the large financial investment condition in 2009, higher than that of Cuba, the Czech republic, the Netherlands and Britain in 2007.
Meanwhile, there was a large gap between different regions. Shanghai, Fujian, Gansu shared a lower proportion of out of pocket spending, and the proportion of Shanghai was even lower than 30%.While Jilin, Heilongjiang, Guangdong shared a higher proportion, higher than 40%.
4.3 Health sector did not benefit equally from the development of economy
Per capita GDP had a negative effect on the total health expenditure in GDP, while Per capita GDP lagged 1 period had a positive effect. The result that the negative coefficient is larger than the positive coefficient suggested that the development of health service lagged behind the economy although the government enlarged the investment on health.
4.4 Raising the proportion of government health spending in financial expenditure is the inevitable way to solve the problems of "seeing a doctor is expensive"
The proportion of out of pocket spending in 2015 will be 35.7% according to the trend of 2001-2009. It is hard to achieve the goal that the proportion of out of pocket spending decreasing to 30.0%. Therefore, the government should invest more money to health to achieve the goal.
If we are ready to decrease the individual proportion to 30.0% in 2009, what we should do is increasing 492.1 billion yuan to the health, and the total Health Expenditure in GDP of China will be 6.4%.
If we are ready to decrease the individual proportion to 30% in 2015, what we should do is increasing 890.7 billion yuan to the health on the basis of 170.7 billion yuan, and the total Health Expenditure in GDP of China will be 7.7%.
医药卫生事业关系亿万人民群众的健康,关系千家万户的幸福,是重大民生问题。然而,当前看病贵、乱开药、拉大网检查等百姓抱怨问题愈演愈烈,药品市场的“高价药易销”问题仍然突出。课题组前期研究发现,医疗费用负担重、看病贵问题是医疗改革的需优先解决的问题。
“看病贵”问题对社会造成了极其严重的后果,一方面造成卫生费用过快增长,直接加重了各类医保制度收支平衡,影响各类医保制度的平稳运行。另一方面,卫生费用的过快增长和医疗保障水平较低则进一步加重了居民医药负担,使得“看病贵”问题日益严重,威胁居民卫生服务可及性,因病致贫问题突出。围绕着“看病贵”问题,即医疗费用负担问题,医疗卫生相关利益团体相互埋怨,陷入失衡状态：医疗机构埋怨政府财政投入不足、收费标准过低等补偿机制问题,医疗保障部门担心医疗费用快速上涨、基金收支失衡,当看病贵等问题与医改挂钩,政府则担忧看病贵问题危及社会和谐稳定。
政府对此给予了前所未有的重视,《中共中央国务院关于深化医药卫生体制改革的意见》和《医药卫生体制改革近期重点实施方案(2009—2011年)的通知》相继出台,从宏观上确定了我国医改的基本方向、思路和政策框架,对于保障我国医药卫生事业持续健康发展,维护广大人民群众健康具有重大意义。
但是,如何进行改革,如何真正解决社会反响激烈的问题,实践关注民生的价值观?本文认为,透视卫生费用过快增长和控制卫生总费用这一社会热点,3个科学问题值得进一步思考：(1)卫生费用总量及构成的影响因素是什么?(2)将来中国医疗费用的走势如何?如何进行科学预测?(3)探讨适宜中国国情的卫生费用总量及构成的预测与控制模型,据此为研制实践中从根本上一揽子解决中国医疗卫生领域现有突出问题的策略思路提供依据。
因此,本研究的目的在于明确卫生费用总量与构成影响因素的维度及其指标体系,利用10省的Panel Data数据进行模拟确定卫生费用总量、构成及其影响因素的定量关系,在此基础上,对我国2015年卫生费用总量与构成进行预测,并比较其与国际普遍经验之间的差距,为我国医改提供参考依据。
二、材料与研究方法
本研究以卫生系统宏观模型为指导性研究方法,以Panel Data模型、规范差距分析等为应用性研究方法,在总研究项目前期研究的基础上,形成我国卫生费用总量与构成定量研究的框架和思路。
本部分定量研究搜集的资料主要来源于自于面上统计资料,包括2002-2010年的中国统计年鉴、中国卫生统计年鉴、中国人口和就业统计年鉴、卫生事业经费决算资料以及2009、2010年卫生部卫生经济研究发行的中国卫生总费用研究报告。将搜集数据进行整理后形成分析库,对数据进行单位根检验、数据转换、滞后期数确定,用Panel Data模型对卫生费用总量、构成与其影响因素之间的关系进行定量模拟与预测。
三、主要研究结果
(一)确定卫生费用总量及构成影响因素
运用“卫生总费用”、“影响因素”、“个人卫生支出”等关键词及其组合,在数据库“中国期刊全文数据库”搜索1979年至今文献1257篇,查阅的结果显示影响因素众多,有GDP、65岁以上人口比例、0-14岁人口比例、政府预算卫生支出、社会卫生支出、人口数、财政收入、财政支出、医生数、婴儿死亡率、城镇化集中程度、卫生事业费、城镇居民基本医疗保障计划、居民消费价格指数、中西药品及保健用品价格指数、每千人口医疗机构床位数、卫生服务技术密集程度、住院病人人均医疗费、农村居民人均医疗消费水平、城镇居民人均医疗消费水平、居民家庭人均可支配收入、居民消费水平、农村居民消费水平、城镇居民消费水平、城乡居民人民币储蓄存款年底余额、期望寿命、慢性病、社会保险覆盖率。影响卫生费用总量及构成的因素遍及结构、过程、结果以及包括人口需要、社会经济、政治等在内的外环境,且主要集中在外部子模,尤其是社会经济与人口需要维度。
依据专家咨询与论证确定的全面性、科学性、可评价性、系统性和代表性原则,最终确定影响总量和构成的主要变量维度及指标为：社会经济(人均GDP、城镇居民家庭人均可支配收入绝对值、农村居民人均纯收入绝对值)、人口(总人口、65岁以上人口比例、0-14岁人口比例)、结构(政府预算卫生支出占财政支出的比例)、过程(每门诊人次收费水平、每床日平均收费水平)、结果(孕产妇死亡率、围产儿死亡率)
(二)模型构建过程与结果
1、模型构建过程
在确定卫生费用总量及构成影响因素的基础上,对变量进行单位根检验、数据变换和滞后期数确定,为模型构建做准备。单位根检验结果发现,变量“个人现金卫生支出占卫生总费用的比例”、“总人口”、“孕产妇死亡率”、“围产儿死亡率”、“0-14岁人口比例”、“65岁以上人口比例”、“每门诊人次收费水平”、“每床日平均收费水平”等变量为平稳的序列,可直接纳入回归方程；而“卫生总费用占GDP的比例”、“人均GDP”、“城镇居民可支配收入”“农村居民人均纯收入”、“政府卫生支出占财政支出的比例”为非平稳序列,不能直接纳入回归方程,对其进行数据变换后纳入回归方程。
人均GDP对卫生费用总量与构成存在滞后效应。根据Akaike与Schwarz信息准则,滞后一期的人均GDP的Akaike与Schwarz函数值最小,因此确定人均GDP对卫生总费用占GDP的比例存在滞后一期的影响。
在此基础上,根据F统计量,确定模型形式是变系数模型、变截距模型、还是混合回归模型；根据Hausman检验结果,确定选择固定效应模型还是随机效应模型；继而运用panel data数据模型方法,构建卫生费用总量与构成及其影响因素的模拟回归方程。
2、模型模拟结果
使用固定效应模型模拟的卫生总费用占GDP的比例及其影响因素的关系如下：
其中,αi*为个体截面基准值之间的区别,LNH为对数变换后的卫生总费用占GDP的比例,LNGDP为对数变换后的人均GDP,LNGDP (-1)为滞后一期的对数变换后的人均GDP,LNO为对数变换后的政府预算卫生支出占财政支出的比例,MM为孕产妇死亡率,OL为65岁及以上人口所占比例,FF为每门诊人次收费水平。
模拟结果显示：卫生总费用占GDP的比例与滞后一期的人均GDP、政府预算卫生支出占财政支出的比例、65岁以上人口所占比例、每门诊人次收费水平呈正向关系,与当期人均GDP、孕产妇死亡率呈负向关系(R2为0.95)。
使用随机效应模型对个人现金卫生支出占卫生总费用的比例及其影响因素的关系模拟,方程形式为：
其中, OOP为对数变换后的个人现金卫生支出占卫生总费用的比例,LNGDP为对数变换后的人均GDP,LNGDP (-1)为滞后一期的对数变换后的人均GDP,LNO为对数变换后的政府预算卫生支出占财政支出的比例,MM为孕产妇死亡率,PE为总人数,FB为每床日平均收费水平。
模拟结果显示：个人现金卫生支出占卫生总费用的比例与政府预算卫生支出占财政支出的比例、每床日平均收费水平呈正向关系,与其它变量的关系无统计学意义(R2为0.81)
(三)卫生费用总量与构成预测及与国际普遍经验间差距
根据构建的Panel data模型,结合预测的2015年的相关数据,对2015年中国卫生费用总量和构成进行预测。根据模型预测,2015年卫生总费用占GDP的比例为6.5%。结合课题组前期研究,筹资公平性优于中国的113个国家样本的“卫生总费用占GDP比例”模型预测结果,按照国际普遍经验2015年中国卫生总费用应达到GDP的6.0%。两者相比,说明基于现实的模拟预测值高于国际较优国家普遍经验值,达到较为适宜的比例。研究得出的基于现实模型的结果提示,2015年个人现金卫生支出占卫生总费用的比例为35.7%,低于2015年根据国际普遍经验值38.5%,个人负担将低于国际普遍经验,出现较好的势头。但是需要注意的是,这是在当前政府加大预算卫生支出,政府预算卫生支出增长速度大大超过财政支出增长的情况下必然发生的结果,若政府预算卫生支出增长幅度不能确保最近年份的增速势头,则要实现国际普遍经验的目标估计还是有待观察的。
四、结论
(一)我国卫生总费用占GDP的比例较低,不同地区之间存在较大差距
中国卫生统计年鉴中数据显示,2001-2009年,卫生总费用的年均增速为16.6%,高于GDP的增长速度(15.2%)。2007年我国卫生总费用占GDP的比例仅为4.3%,较2000年的4.6%有所下降,在191个国家和地区中仅列第149位。即便是财政增加投入的2009年,也仅为5.1%。在世界各国卫生总费用占GDP的较高的比例和逐年增加的趋势下,中国卫生费用总量保持在相对低的水平,一直低于5.5%,这表明中国卫生费用并没有突破社会的支付能力,仍然有较大的增长空间。
尽管按照当前发展趋势,2015年卫生总费用占GDP的比例将达到6.5%,高于世界普遍经验确定的6.0%,但是需要注意的是,这是在当前政府加大预算卫生支出,政府预算卫生支出增长速度大大超过财政支出增长的情况下发生的结果,若政府预算卫生支出增长幅度不能确保最近年份的增速势头,则要实现国际普遍经验的目标估计还是有待观察的。因此,保持卫生费用的可持续增长是医改的关键。
与此同时,全国各地卫生总费用占GDP的比例存在较大差别。甘肃、云南、新疆等经济欠发达地区卫生总费用占GDP的比例较高,且有逐年增加的趋势。而上海、浙江、天津、广东、福建等经济发达地区卫生总费用占GDP一直偏低,一直低于5.0%。
(二)个人现金卫生支出比例依然偏高,不同地区间存在较大差距
2007年中国个人卫生支出占卫生总费用的比例依然高达55.3%,成为个人负担比例比较高的国家之一,在191个国家和地区中列第46位。即便是幅度增加财政投入的2009年,个人负担比例也达到了38.2%。明显高于2007年古巴、捷克、荷兰和英国的个人现金卫生支出占卫生总费的比例(分别为4.5%、14.8%、18.0%和18.3%)
各地个人现金卫生支出占卫生总费用的比例存在较大差别。上海、新疆、福建、甘肃等省份个人现金卫生支出占卫生总费用的比例较低,尤其是上海个人支出比例一直在30.0%以下。而吉林、黑龙江、广东等地区个人现金卫生支出占卫生总费用的比例的比例较高,一直在40.0%以上
(三)卫生事业没有享受到经济发展的平均利益
卫生费用总费用占GDP的比例受到人均GDP、政府预算卫生支出占财政支出比例、65岁及以上人口所占比例、孕产妇死亡率、每门诊次均费用等因素的影响。其中,卫生总费用占GDP的比例与当期的人均GDP呈负向关系,而与滞后一期的人均GDP呈正向关系,而且负向关系的系数要大于正向关系的系数。由于受当期GDP增长(分母)的影响,卫生总费用(分子)占GDP的比例在当期会有所下降；在GDP增长的基础上,国家加大了对卫生事业的投入,使得下年卫生总费用占GDP的比例又有所上升。但是负向关系的系数要大于正向关系的系数,表明尽管国家加大了对卫生的投入,但卫生事业并没有享受到经济发展的平均利益。
(四)提高政府卫生支出占财政支出的比例成为医改的必然选择
如果按照当前发展趋势,2015年个人现金卫生支出占卫生总费用的比例将为35.7%,要完成卫生十二五规划中提及的个人现金卫生支出占卫生总费用的比例降低到30%的目标非常困难。因此,针对影响模型因变量程度较大的自变量—一政府加大卫生投入入手具有较好的现实意义,相对于人口比例、经济发展以及收费标准等指标,切实提高政府卫生支出占财政支出的比例是较为合适的选择。
如果要使2009年的个人现金卫生支出占卫生总费用的比例降低到30.0%,按照上述模型,在其他影响因素不变的情况下,政府预算卫生支出占财政支出的比例需要由6.4%增加到14.5%,即政府需要在当前3930.7亿元的基础上追加4921.0亿元,此时卫生总费用占GDP的比例达到6.4%。
如果使2015年的个人现金卫生支出占卫生总费用的比例降低到30%,按照上述模型,在其他影响因素不变的情况下,政府卫生支出占财政支出的比例需要由预测的8.8%增加到13.4%,即政府需要在预测的17006.5亿元的基础上追加8907.0亿元,此时卫生总费用占GDP的比例达到7.7%。
1. Background and Objectives
The health care sector is a major livelihood issue, as it is closely related to the health of billions of people and the happiness of every household. However, unaffordable and unnecessary medical care has been complained more and more frequently. The problem in the drug market that expensive drugs sold well was still serious. Our earlier research found that affordable medical care should be given priority.
The problem has caused serious consequences to the society. On one hand, it caused the abnormal speed of health expenditure growth and affected the balance of health insurance; on the other hand, the over-rapid growth of health expenditure increased the burden of medical costs of the general public and aggravated the poverty caused by illness, making the problem " seeing a doctor is expensive " more serious. Medical institutions complained that the money provided to health was inadequate and the fee standards were very low. Medical security departments worried about the quick rise of medical expenditure and the disequilibrium of the fund. The governments worried that the problem of " seeing a doctor is expensive " might endanger the harmony and stability of the society.
The governments have paid much attention to the problems and promulgate policy which has oriented the development of the health system in order to solve them. But how to reform? How to solve the problem of social repercussions entirely? There are still some questions to answer:What variables affect health care expenditure and its constitution? What is the future trend of the Chinese health care expenses? How to conduct scientific prediction? How to simulate the relationship between health care expenditure and its constitution and the factors?
The objective of the study is firstly to explore the variables that affect health care expenditure and its constitution. And then use the panel data model of 10 provinces to analyze the relationship between health care expenditure and its constitution and the factors. On this basis, we use the database to predict health care expenditure and its constitution of 2015 in China. Finally, we make a comparison between the predicted expenditure and constitution of 2015 and the expenditure and constitution of the international experience.
2. Materials and Methods
The macro model of health system was selected as the principle methods, and selected the panel data model as the applied research method. The framework and thinking of the quantitative study of health care expenditure and its institution in China formatted on the basis of preliminary research of the total project.
The data used in the study was collected mainly from statistics including China Statistical Yearbook, China Health Statistical Yearbook, China Population and Employment Statistics Yearbook and health accounts funding information of the year from 2002 to 2010, and China Total Expenditure on Health Report published by China Health Economics Institute from 2009-2010.Then we collated those data collected from these statistics to be a database which could be analyzed directly. After Unit Root test, data conversion and confirming the lag period, we analyzed the relationship between health care expenditure and its constitution and the factors, and then predicted health care expenditure and its constitution of the year 2015.
3. Main Results
3.1 The factors which affected health care expenditure and its constitution
After using the keywords "health care expenditure", "influence factors" and " out of pocket spending " and their combination to search literatures in the CNKI database since 1979,we found that there were a lot of factors that affected health care expenditure and its constitution, such as GDP, population proportion of people over 65, population proportion of people below 14, the government budget health spending, social health spending, population, fiscal revenue, fiscal expenditure, the number of doctors, infant mortality, urbanization concentration, health care expenditure, basic medical security plan of urban residents, consumer price index, Chinese and western medicines and health supplies price index, medical institutions per thousand population, the intensity of health service, per capita health cost of hospital patients, per capita health consumption level of each rural resident, per capita health consumption level of each urban resident, per capita disposable real income, consumption level, rural consumption level, urban consumption level, the end of urban and rural residents RMB savings deposit balance, life expectancy, chronic disease, social insurance coverage. The factors distributed in structure submodule, process submodule, result submodule, and the external environment submodule including population needs submodule, social economy submodule, politics submodule, etc.
After the comprehensive consideration of expert consultation and the principle of comprehensive, scientific, able to be evaluated, systematic and representative demonstration,we finally decided to classified the factors which were put in the equation into 5 dimensions:economy dimension which included per capita GDP, per capita disposable real income of urban residents, per capita disposable real income of rural residents; demography dimension which included population, population proportion of people over 65, population proportion of people below 14; structure dimension which included the proportion of the government budget health in financial expenditure; process dimension which included per outpatient service charge and per hospital service charge; result dimension which included maternal mortality and perinatal mortality.
3.2 The process and result of model building
3.2.1 The process of model building
Based on the unit root test,we found that there were several stable factors which included the proportion of out of pocket spending, maternal mortality,wai births mortality, population, population proportion of people over 65, population proportion of people below 14, per outpatient service charge,and per hospital service charge. Other variables were unstable and should be transformed before putting into equation.
Per capita GDP had lagging effect to health care expenditure and its constitution. We found the lag period was 1 according to the Akaike s Information Criterion and the Schwarz Criterion. On this basis, we determined the general format of the model according to the F statistics and the influence form according to the Hausman tests. Then we built the equation of the relationship between health care expenditure and its constitution and the factors.
3.2.2 The result of the model building
The equation of the relationship between the total health expenditure in GDP and its influencing factors was simulated by fixed effect model as:
αi* is the difference of each cross section. LNH is the logarithmic transformation of total health expenditure in GDP. LNGDP is the logarithmic transformation of Per capita GDP, while LNGDP(-1) is the logarithmic transformation Per capita GDP lagged 1 period. LNO is the logarithmic transformation of the proportion of the government budget health in financial expenditure. MM is the maternal mortality. OL is the population proportion of people over 65. FF is per hospital service charge.
The result suggested that per capita GDP lagged 1 period, the proportion of the government budget health in financial expenditure, population proportion of people over 65, per outpatient service charge had a positive effect on the total health expenditure in GDP, while per capita GDP and maternal mortality had a negative effect.
The equation of the relationship between out of pocket spending and its influencing factors was simulated by random effect model as:αi* is the difference of each cross section. OOP is the proportion of out of pocket spending. LNGDP is the logarithmic transformation of Per capita GDP, while LNGDP(-1) is the logarithmic transformation Per capita GDP lagged 1 period. LNO is the logarithmic transformation of the proportion of the government budget health in financial expenditure. MM is the maternal mortality. OL is the population proportion of people over 65. FB is per ouypatient service charge.
The result showed that the proportion of the government budget health in financial expenditure, and per outpatient service charge had a positive effect on the proportion of out of pocket spending.
3.3The prediction of health care expenditure and its constitution and the gap between the prediction and the international common experience
We can predict Chinese health care expenditure and its constitution of 2015 according to the building model. Total health expenditure in GDP will be 6.5% in 2015, higher than 6.0% according to the international common experience which was drawn from the 113 countries of which the performance of Health system was better than China. The proportion of out of pocket spending of China will be 35.7% in 2015, lower than 38.5% according to the international common experience. But it should be paid attention that the result is based on the increasing investment on health these years. If the investment is decreased, the goal be achieved is not certain.
4. Conclusion
4.1TotaI health expenditure in GDP of China was still low, and there was a large gap between different regions
The annual growth speed of total health expenditure is 16.6% from 2001 to 2009, quicker than the growth of GDP 15.2%. Total health expenditure in GDP of China in 2007 was only 4.3%, ranked 149 in the 191 world countries, which is lower than that of 2000. Although total health expenditure in GDP worldwide kept a higher proportion and rising trend, China's total health care expenditure remained in a comparatively low level, below 5.5 percent all the time. This showed that Chinese health expenditure had not break the social payment ability and still has larger growth space.
Total health expenditure in GDP of 2015 will be 6.5%, higher than that of the international experience. But it should be paid attention that the result is based on the increasing investment on health these years. If the investment is decreased, the goal be achieved is not certain. Therefore the key of the health reform is to make sustainable development of the health expenditure.
Meanwhile, there was a large gap between different regions. Total Health Expenditure in GDP of underdeveloped region, such as Gansu, Yunnan, Xinjiang, was higher than the average and had a rising trend. While Total Health Expenditure in GDP of developed regions, such as Shanghai, Zhejiang, Tianjin, Guangdong, Fujian, was lower than 5.0%.
4.2 The proportion of out of pocket spending was still higher than that of the government and the social spending, and there was a large gap between different regions
The proportion of out of pocket spending of China in 2007 was 55.3%, leading China to be one of the countries in the world with the highest individual proportions, ranked 46 in the 191 world countries. The proportion of out of pocket spending was 38.2%at the large financial investment condition in 2009, higher than that of Cuba, the Czech republic, the Netherlands and Britain in 2007.
Meanwhile, there was a large gap between different regions. Shanghai, Fujian, Gansu shared a lower proportion of out of pocket spending, and the proportion of Shanghai was even lower than 30%.While Jilin, Heilongjiang, Guangdong shared a higher proportion, higher than 40%.
4.3 Health sector did not benefit equally from the development of economy
Per capita GDP had a negative effect on the total health expenditure in GDP, while Per capita GDP lagged 1 period had a positive effect. The result that the negative coefficient is larger than the positive coefficient suggested that the development of health service lagged behind the economy although the government enlarged the investment on health.
4.4 Raising the proportion of government health spending in financial expenditure is the inevitable way to solve the problems of "seeing a doctor is expensive"
The proportion of out of pocket spending in 2015 will be 35.7% according to the trend of 2001-2009. It is hard to achieve the goal that the proportion of out of pocket spending decreasing to 30.0%. Therefore, the government should invest more money to health to achieve the goal.
If we are ready to decrease the individual proportion to 30.0% in 2009, what we should do is increasing 492.1 billion yuan to the health, and the total Health Expenditure in GDP of China will be 6.4%.
If we are ready to decrease the individual proportion to 30% in 2015, what we should do is increasing 890.7 billion yuan to the health on the basis of 170.7 billion yuan, and the total Health Expenditure in GDP of China will be 7.7%.
引文
[1]费菲.看病贵从何时缘起[J].首都医药,2010,(5)：19
[2]胡善联,张仁伟,高建民.“乡村卫生机构药品购销体制研究”总报告[R].卫生经济研究,2003,(4)：3-5.
[3]孙兰香.调整收费结构减轻患者负担[J].中国卫生经济,2002,21(8)：42.
[4]田璇.浅谈新时期医德医风建设[J].医学理论与实践,2005,(1)：1367-1368.
[5]刘军民.过度市场化与分权化——中国医疗卫生改革的双重误区[J].卫生经济研究,200(12)：3-10.
[6]尹秀云.走出医患关系认知误区[J].家庭中医药,2004,(12)：10-11.
[7]郑雪倩主编.医疗纠纷防范与对策[J].广东汕头大学出版社,2002：1-7
[8]沈鸿伟,王钦清.正确认识和处理医疗纠纷促进医患关系正常化[J].前进论坛,2003,(10)：25-26.
[9]林晓蕾.医生是“白狼”还是“天使”——委员交锋医患之争[EB/OL]. http://heilongJiang.northeast.cn/system/2006/02/09/050253818.shtml.2006-02-09.
[10]王俊秀.国务院研究机构称我国医疗改革工作基本不成功[EB/OL]. http://news.xinhuanet.com/cic/2005-07/29/content 3282394.htm.2005-07-29.
[11]赵志疆.医改不成功背后的政府失灵[N].燕赵都市报,2006-05-24(8).
[12]杜乐勋,张文鸣,黄泽民.医疗卫生绿皮书[J].中国医疗卫生发展报告,2006.
[13]杜乐勋,张文鸣,黄泽民.中国医疗卫生发展报告[M].北京：社会科学文献出版社,2006.109-111.
[14]中华人民共和国中央人民政府.2005年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2006-02/16/content 201218.htm
[15]中华人民共和国中央人民政府.2006年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content 1260216.htm
[16]中华人民共和国中央人民政府.2006年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content 1260221.htm
[17]中华人民共和国中央人民政府.2007年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content_1260188.htm
[18]中华人民共和国中央人民政府.2008年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content 1260198.htm
[19]中华人民共和国中央人民政府.2009年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content_1260221.htm
[20]中华人民共和国中央人民政府.温家宝所作政府工作报告(十一届人大三次会议)[EB/OL]. http://www.gov.cn/20101h/content 1555767.htm
[21]中华人民共和国中央人民政府.温家宝在十一届人大四次会议上所作政府工作报告[EB/OL]. http://www.gov.cn/20111h/content_1825233.htm
[22]国家统计局.《中国统计年鉴》,北京：中国统计出版社,2005,第一版：36-37
[23]卫生部.中国卫生统计年鉴,北京：中国协和医科大学出版社,2006：37-39
[24]世界卫生组织(日内瓦).2000年人人健康进度检查指标的制订[R].《人人健康》丛刊第四期("HEALTH FOR ALL " SERIES, No.4):30
[25]中共中央、国务院.关于卫生改革与发展的决定(中发[1997]3号)[J].中国农村卫生事业管理,1997,17(2)：3-9.
[26]周雁翎,差异悬殊：中国卫生保健事业将面临严峻挑战[J].中国改革,2002(4)：6-8
[27]新华网.关于构建社会主义和谐社会若干重大问题的决定[EB/OL]. http://news.xinhuanet.com/politics/2006-10/18/content 5218639.htm
[28]新华网.加快推进以改善民生为重点的社会建设[EB/OL]. http://news.xinhuanet.com/newscenter/2007-10/15/content_6883752.htm
[29]中华人民共和国中央人民政府.中共中央国务院关于深化医药卫生体制改革的意见 [EB/OL]. http://www.gov.cn/Jrzg/2009-04/06/content_1278721.htm
[30]Levin,A.,C.F.Lin,and C.S.J. Chu(2002)Unit root tests in panel data[J],asymptotic and finite-sample properties,Journal of Econometrics,Vol.108,1-24
[31]Im, K. S., M. H. Pesaran and Shin(2003)Testing for Unit roots in heterogeneous panels[J], Journal of Econometrics, Vol.115,53-74
[32]Choi, I(2001) Unit root test for panel data[J] Journal of International Money and Banking, Vol. 20,249-272
[33]吕延方,陈磊.面板单位根检验方法及稳定性的探讨[J].数学的实践与认识,2010,40(21)：49-61
[34]吴巧生,陈亮,张炎涛,成金华.中国能源消费与GDP关系的再检验—基于省级面板数据的实证分析[J].数量经济技术经济研究,2008,(6).27-40
[35]李子奈,叶阿忠.高级计量经济学[M].北京：清华大学出版社,2000.
[36]曾小敏,刘树仁,彭晓明,等.湖南省1997年卫生总费用影响因素的主成分分析[J].湖南医科大学学报(社会科学版),2000,2(4)：183-88
[37]何平平.我国卫生总费用的弹性测算[J].统计与决策,2006,(5)：87-90
[38]Herwartz, H.,Theilen, B.THE DETERMINANTS OF HEALTH-CARE EXPENDITURE NEW RESULTS FROM SEMIPARAMETRIC ESTIMATION [J].Health economics,2010,19(8):964-978
[39]Carrion-i-Silvestre, JL.Health care expenditure and GDP:Are they broken stationary?[J]. JOURNAL OF HEALTH ECONOMICS,2005,24(5):839-854
[40]Hansen, P., King, A.Health care expenditure and GDP:panel data unit root test results-comment[J]. JOURNAL OF HEALTH ECONOMICS,1998,17(3):377-381
[41]李秀娟.吴洋.医疗费用增长的原因及对策思考[J].卫生软科学,2002,16(3)：25-27
[42]黄成礼.人口老龄化对乙生费用增长的影响[J].中国人口科学,2004,(4)：36-43
[43]黄成礼.人口因素与卫生费用的关系[J].人口研究,2004,28(3)：24-30
[44]何平平.我国医疗费用增长因素的计量分析[J].Pacific Journal,2005,(11)：25-31
[45]何平平.我国卫生总费用的弹性测算[J].统计与决策,2006,(5)：87-90
[46]刘柏,赵振全.卫生总费用长期均衡和短期波动的经济计量检验[J].中国卫生经济,2008,27(6)：5-7
[47]刘树仁,周欢,彭晓明.卫生总费用影响因素的岭回归分析[J].卫生经济研究,2000,(9)：12-14
[48]赵郁馨.卫生总费用运行特点及增长因素[J].中国卫生经济,2002,21(3)：32-35
[49]刘国祥,宋泽萍,张长青.中国贫困地区卫生费用影响因素分析[J].中国卫生经济,1995,14(7)：24-25
[50]何平平.中国卫生总费用增长因素研究[J].统计与信息论坛,2006,21(1)：37-41
[51]郑云萍,温小霓.我国卫生总费用增长研究[J].中国卫生经济,2006,25(12)：16-18
[52]李军山,江可申,李雪强.我国医疗费用增长因素实证分析[J].价格月刊,2008,(10)：56-58
㈨何平平.协整分析与误差修正模型—经济增长、人口老龄化与我国医疗费用增长的实证研究[J].工业技术经济,2006,25(1)：122-124,135
[54]赵郁馨,高广颖,杜乐勋.中国卫生总费用发展变化趋势及其影响因素[J].卫生经济研究特稿,2000,(1)：7-9
[55]何平平.我国卫生总费用的弹性测算[J].统计与决策,2006,(5)：87-90
[56]危小军,廖伟,钟一鸣,等.遏制医疗费用快速增长构建和谐医患关系[J].赣南医学院学报,2007,27(5)：776-777
[57]金新政,张辉.卫生总费用膨胀之灰关联成因分析[J].卫生经济研究,1995,(12)：11-12
[58]李军山,江可申,李雪强.我国医疗费用增长因素实证分析[J].价格月刊,2008,(10)：56-58
[59]蔡善荣,阮红芳,李鲁,等.以主成分回归分析方法探讨人均卫生费用影响因素的研究[J].中国卫生事业管理,2001,(7)：399-400
[60]刘明霞,仕泉.自组织数据挖掘在卫生总费用预测中的应用[J].生经济研究,2003,(12)：10-12
[61]何平平.我国医疗费用增长因素的计量分析[J].PACIFIC JOURNAL,2005,(11):25-31
[62]蔡善荣,阮红芳,李鲁,等.以主成分回归分析方法探讨人均卫生费用影响因素的研究[J].中国卫生事业管理,2001,(7)：399-400
[63]雷海潮,刘兴柱,卞鹰.慢性病增加对我国卫生总费用及宏观经济的影响研究[J].中国卫生经济,1996,15(10)：27-28
[64]中国卫生总费用课题组.中国卫生总费用影响因素的计量经济学分析[J].中国卫生经济,1996,15(12)：27-28
[65]张宜民,顾雪非,张金齐.1990—-2002年我国卫生总费用筹资水平、结构及变化趋势分析[J].中国初级卫生保健,2004(10)：15-18
[66]陈琳红,肖云.我国卫生总费用筹资水平与结构分析[J].西部论坛.2010,20(6)：72-77
[67]耿瑛,裴丽昆,David Legge.增加政府投入提高医药卫生服务的公平性和可及性[J].中国卫生经济.2009,28(3)：12-14
[68]刘波,周雅静.我国医疗费用增长与经济增长的相关性分析[J].黑龙江对外贸易.2010,(9)：86-88
[69]钱东福,尹爱田,王鸿勇等.卫生事业发展状况的国际比较分析[J].中国卫生经济.2007,26(6)：17-19
[40]孟庆跃.卫生总费用、健康产出和卫生改革的国际比较和分析[J].卫生经济研究.2003,22(4)：23-25
[71]胡瑞荣,马旭东.“十一五”期间我国卫生总费用变化趋势及其影响因素分析[J].中国医疗保险.2011,(2)：40-45
[72]何平平.我国医疗支出增长因素研究[D].硕士学位论文.北京邮电大学.2007
[73]汪金鹏.1996-2002年我国卫生总费用筹资水平和机构分析[J].2006,(1)15-16
[74]卫生工作会新闻稿之四.卫生部明确提出“十二五”卫生发展的总体目标[EB/OL]. http://www.moh.gov.cn/publicfiles/business/htmlfiles/mohbgt/s3582/201101/50268.htm
[75]王朝昕,王颖等.适宜的中国卫生筹资总量分析[J].中国卫生资源.2011,14(1)：65-66
[76]中国新闻网.央行报告：2009年来人民币对美元汇率保持稳定[EB/OL]. http://www.chinanews.com/cj/cj-gncj/news/2009/11-11/1959823.shtml
[77]人民网.20日人民币汇率中间价：1美元对人民币6.5294元[EB/OL]. http://www.chinanews.com/fortune/2011/05-24/3061671.shtml
[78]李松光,王颖等.适宜的中国卫生筹资构成探讨[J].中国卫生资源.2011,14(1)：67-68
[79]中华人民共和国中央人民政府.温家宝在十一届人大四次会议上所作政府工作报告[EB/OL]. http://www.gov.cn/2011lh/content_1825233.html
[1]Mundlak Y. Empirical Production Function Fee of Management Bias[J].Journal of Farm Economics,1961,(43):44-56
[2]Nerlove,M. Estimation and identification of Cobb-Douglas production function[M].Chicago:Rand Mcnally & Company,1965
[3]白仲林.面板数据的计量经济分析[M].天津：南开大学出版社,2008
[4]白仲林.面板数据模型的设定、统计检验和新进展[J].统计与信息论坛.2010,25(10)：3-12
[5]Allison P D. Causal inference with panel data [EB/OL].http: pauldallison.com/downloads/Causal%20Inference.pdf.
[6]Kasprzyk D., Duncan GJ. Panel surveys[M].New York:John Wiley,1989:16-18
[7]Fitzgerald J, Gottschalk P. An analysis of sample attrition in panel data[J]. Journal of Human Resources,1998,33(2):251-299
[8]Abauf N, Jorian P. Purchasing power parity in the long run[J].Journal of Finance,1990,45(1): 157-174
[9]白仲林.同期相关面板数据退势单位根检验的小样本性质[J].数量经济技术经济研究,2006(5)：146-152
[10]段鹏,白仲林,张晓峒.个体协整条件下面板单位根检验统计量性质研究[J].统计研究,2009(4)：91-100
[11]Baltagi B H. Econometric analysis of panel data[M]白仲林,译.北京：机械工业出版社,2010：275-284.
[12]武大勇.计量经济学中的面板数据模型分析[D].武汉：华中科技大学,2006
[13]Marc Nerlove. An essay on the history of Panel Data Econometrics. The Ninth International Conference on Panel Data,Geneva,2000:4-9
[14]Hsiao C. Analysis of panel data.2nd Edition, Cambridge University Press,Cambridge,2003,1-5
[15]长自武.平行数据的单位根检验及其模型应用.城镇居民收入和消费的实证分析[D].乌鲁木齐：新疆大学,2007
[16]Bhargava, A., L. Franzini and M. Narendranathan. Serial Correlation and Fixed Effects Models [J]. Review of Economic Studies 1982,49(4):533-549
[17]Abuaf,N. and P. Jorion. Purchasing Power Parity in the Long Run[J]. The Journal of Finance, 1990,45(1):157-174,.
[18]Levin A. and C.F.Lin.Unit Root Tests in Panel Data, Asymptotic and Finite-sample Properties, UC San Diego.Working Paper,1992:92-23
[19]Flores,R., Jorion, P..Multivariate Unit Root Tests of the PPP Hypothesis [J]. Journal of Empirical Finance,1999,6(4):335-353
[20]O'Connell.The Overvaluation of Purchasing Power Parity [J]. Journal of International Economics,1998,44(1):1-19
[21]Taylor, M. and L. Sarno. The Behavior of Real Exchange Rates during the Post Bretton Woods Period [J]. Journal of International Economics,1998,46:281-312
[22]Maddala G.S. and Wu Shaowen, A Comparative Study of Unit Root Tests with Panel Data and a New Simple Test, Oxford Bulletin of Economics and Statistics,1999,61,631-52
[23]Moon, H.R. and B. Perron.. Testing for A Unit Root in Panels with Dynamic Factors[J]. Journal of Econometrics,2004,122(1):81-126
[24]Smith, L.V.. More Powerful Panel Data Unit Tests with an Application to Mean Reversion in Real Exchange Rates [J].Econometrics,2004,19:147-170
[25]Pedroni, P. Panel Cointegration:Asymptotic and Finite Sample Properties of Pooled Time Series Tests With an Application to the PPP Hypothesis Indiana University [J].Working Papers In Economics,1995,20(3):95-113
[26]Kao, C. Spurious Regression & Residual-Based Tests for Cointegration in Panel data[J]. Journal of Econometrics,1999,90:1-44
[27McCoskey, S., and C. Kao.A Residual-Based Test of the Null of Cointegration in Panel Data [J].Econometric Reviews,1998,17(1):57-84
[28]Pedroni, P. On the Role of Cross Sectional Dependency in Panel Unit Root and Panel Cointegration Exchange Rate Studies.Working paper, Indiana University,1997
[29]Pedroni, P. Critical Values for Cointegration Tests in Heterogeneous Panels with Multiple Regressors [J]. Oxford Bulletin of Economics and Statistics,1999,61(1):653-670.
[30]Pedroni, P. Fully Modified OLS for Heterogeneous Cointegrated Panels [J].Advances in Econometrics,2000,15:93-130.
[31]Pedroni, P. Purchasing Power Parity Tests in Cointegrated Panels [J]. Review of Economics and Statistics,2001,83(4):727-31.
[32]Joakim Westerlund. A Panel CUSUM Test of the Null of Cointegration [J].OXFORD BULLETIN OF ECONOMICS AND STATISTICS.2005,67(2):231-256
[33]Westerlund,J..New Simple Tests for Panel Cointegration[J].Econometrics Reviews,2005,24(3): 297-316
[34]Balestra, P.,M. Nerlove.Pooling Cross Section and Time Series Detain the Estimation of a Dynamic Model:The Demand for Natural Gas[J],Econometriea,1966,34:585-612.
[35]Costas MEGHIR. Frank WINDMEIJER. Moment Conditions for Dynamic Panel Data Models with Multiplicative Individual Effects in the Conditional Variance. ANNALES D'ECONOMIE ET DE STATISTIQUE.1999:55-56
[36]Ping Zhang. GENERALIZED ACCUMULATED PREDICTION ERROR AND MODEL SELECTION FOR CATEGORICAL PANEL DATA[J].Statistica Sinica.1997,7:527-544
[37]OLAF HUBLER. Multilevel and nonlinear panel data models[J].Allgemeines Statistisches Archiv,2006,90(1):121-136
[38]ALBAN THOMAS.Consistent estimation of binary-choice panel data models with heterogeneous linear trends[J]. Econometrics Journal,2006,9(2):177-195.
[39]Yu,YH. Determinants of variations in state per capita personal income:a panel data approach[J]. APPLIED ECONOMICS LETTERS,2010,17(3):235-329
[40]De Bandt O, Bruneau C, E1 Amri W.Convergence in household credit demand across Euro area countries:evidence from panel data[J]. APPLIED ECONOMICS,2009,41(27):3447-3462
[41]Ogwang T, Cho DI. Economic determinants of the consumption of alcoholic beverages in Canada:a panel data analysis[J]. EMPIRICAL ECONOMICS,2009,37(7):599-613
[42]Pampel FC. Population aging, class context, and age inequality in public spending[J].American Journal of Sociology,1994,100:153-195.
[43]Kilbourne BS, Farkas G, Beron K, England P.1994.Returns to skill, compensating differentials, and gender bias:effects of occupational characteristics on the wages of white women and men[J]. American Journal of Sociology,1994,100:689-719
[44]Osgood DW, Wilson JK, O'Malley PM.. Routine activities and individual deviant behavior[J]. American Sociological Review,1996,61:635-655.
[45]Mazzanti, Massimiliano.The dynamics of landfill diversion:Economic drivers, policy factors and spatial issues Evidence from Italy using provincial panel data[J]. Resources Conservation and Recycling,2009,54(1):53-61
[46]Sousa TRV, Leite PAM. Panel data analysis of health status in Northeast Brazil[J]. REVISTA DE SAUDE PUBLICA,2008.42(5)
[47]McGuire TG, Ayanian JZ, Ford DE, Henke REM, Rost KM, Zaslavsky AM. Testing for statistical discrimination by race/ethnicity in panel data for depression treatment in primary care[J]. HEALTH SERVICES RESEARCH,2008,43(2):531-551
[48]Paul Hansen, Alan King. The determinants of health care expenditure:A cointegration approach[J]. Journal of Health Economic.1996,15(1):127-137
[49]Suzanne K. McCoskey, Thomas M. Selden. Health care expenditures and GDP:panel data unit root test results[J]. JOURNAL OF HEALTH ECONOMICS.1998,17(3):369-376
[50]Paul Hansen, Alan King. Health care expenditure and GDP:panel data unit root test results-comment[J]. JOURNAL OF HEALTH ECONOMICS,1998,17(3):377-381
[51]Baltagi,BH., Moscone, F. Health care expenditure and income in the OECD reconsidered: Evidence from panel data[J]. ECONOMIC MODELLING,2010,27(4):804-811
[52]Moscone,F.,Tosetti, E. HEALTH EXPENDITURE AND INCOME IN THE UNITED STATES[J]. HEALTH ECONOMICS,2010,19(12):1385-1403
[53]白仲林.面板单位根检验理论及其应用研究[D].天津：南开大学,2005
[54]任燕燕,朱孔来.平行数据模型中的异方差问题的处理[J],统计研究,2005(2),75-77.
[55杨继生,王少平,艾春荣.工具变量法综列单位根检验的有偏性及其修正.数量经济技术经济研究,2006(2)：138-147
[56]张晓峒,白仲林.退势单位根检验小样本性质的比较[J],数量经济技术经济研究,2005(5).
[57]苏明,徐利君.基于面板数据的地方税收与经济差异性研究[J].经济理论与经济管理,2008(3)：57-62
[58]李小平,朱钟棣.国际贸易的技术溢出门槛效应—基于中国各地区面板数据的分析[J].统计研究,2004(10)：27-32
[59]郭庆旺,吕冰洋.税收增长对经济增长的负面冲击[J].经济理论与经济管理,2004,(8)：18-23
[60]郑义,徐康宁.结构变迁、工资差异与经济增长—基于75个国家和地区1978—2007年面板数据的实证研究[J].财经研究,2011,37(3)：17-26
[61]奉琼.外商直接投资与中国的水污染—基于面板数据的分析[J].经济研究导刊,2009,(12)：158-159
[62]李兰.基于省际面板数据的我国EKC实证研究[J].学术交流,2009,(3)：35-38
[63]谢童伟.中国县级教育水平与县人口迁移相互影响分析—基于2004年-2008年31省(市)县级面板数据的实证研究[J].清华大学教育研究,2011,32(1)：84-88
[64]宋帮英,苏方林.我国东中西部碳排放量影响因素面板数据研究[J].地域研究与开发.2011,30(1)：19-24
[65]周先波,田凤平.中国城镇和农村居民医疗保健消费的差异性分析——基于面板数据恩格尔曲线模型的非参数估计[J].统计研究.2009,26(3)：51-58
[66]徐文.对影响我国居民人均就医次数的因素的实证研究——基于一组面板数据的分析[J].消费导刊,2009,(15)：201-202
[67]徐文.对农村医疗保健消费的地区差异及影响因素的分析——基于山东省农村一组面板数据的分析[J].消费导刊,2008,(22)：6-9
[68]于东子.卫生投资对经济增长贡献的定性和定量研究.[D].杭州：浙江大学,2006
[2]胡善联,张仁伟,高建民.“乡村卫生机构药品购销体制研究”总报告[R].卫生经济研究,2003,(4)：3-5.
[3]孙兰香.调整收费结构减轻患者负担[J].中国卫生经济,2002,21(8)：42.
[4]田璇.浅谈新时期医德医风建设[J].医学理论与实践,2005,(1)：1367-1368.
[5]刘军民.过度市场化与分权化——中国医疗卫生改革的双重误区[J].卫生经济研究,200(12)：3-10.
[6]尹秀云.走出医患关系认知误区[J].家庭中医药,2004,(12)：10-11.
[7]郑雪倩主编.医疗纠纷防范与对策[J].广东汕头大学出版社,2002：1-7
[8]沈鸿伟,王钦清.正确认识和处理医疗纠纷促进医患关系正常化[J].前进论坛,2003,(10)：25-26.
[9]林晓蕾.医生是“白狼”还是“天使”——委员交锋医患之争[EB/OL]. http://heilongJiang.northeast.cn/system/2006/02/09/050253818.shtml.2006-02-09.
[10]王俊秀.国务院研究机构称我国医疗改革工作基本不成功[EB/OL]. http://news.xinhuanet.com/cic/2005-07/29/content 3282394.htm.2005-07-29.
[11]赵志疆.医改不成功背后的政府失灵[N].燕赵都市报,2006-05-24(8).
[12]杜乐勋,张文鸣,黄泽民.医疗卫生绿皮书[J].中国医疗卫生发展报告,2006.
[13]杜乐勋,张文鸣,黄泽民.中国医疗卫生发展报告[M].北京：社会科学文献出版社,2006.109-111.
[14]中华人民共和国中央人民政府.2005年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2006-02/16/content 201218.htm
[15]中华人民共和国中央人民政府.2006年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content 1260216.htm
[16]中华人民共和国中央人民政府.2006年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content 1260221.htm
[17]中华人民共和国中央人民政府.2007年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content_1260188.htm
[18]中华人民共和国中央人民政府.2008年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content 1260198.htm
[19]中华人民共和国中央人民政府.2009年国务院政府工作报告[EB/OL]. http://www.gov.cn/test/2009-03/16/content_1260221.htm
[20]中华人民共和国中央人民政府.温家宝所作政府工作报告(十一届人大三次会议)[EB/OL]. http://www.gov.cn/20101h/content 1555767.htm
[21]中华人民共和国中央人民政府.温家宝在十一届人大四次会议上所作政府工作报告[EB/OL]. http://www.gov.cn/20111h/content_1825233.htm
[22]国家统计局.《中国统计年鉴》,北京：中国统计出版社,2005,第一版：36-37
[23]卫生部.中国卫生统计年鉴,北京：中国协和医科大学出版社,2006：37-39
[24]世界卫生组织(日内瓦).2000年人人健康进度检查指标的制订[R].《人人健康》丛刊第四期("HEALTH FOR ALL " SERIES, No.4):30
[25]中共中央、国务院.关于卫生改革与发展的决定(中发[1997]3号)[J].中国农村卫生事业管理,1997,17(2)：3-9.
[26]周雁翎,差异悬殊：中国卫生保健事业将面临严峻挑战[J].中国改革,2002(4)：6-8
[27]新华网.关于构建社会主义和谐社会若干重大问题的决定[EB/OL]. http://news.xinhuanet.com/politics/2006-10/18/content 5218639.htm
[28]新华网.加快推进以改善民生为重点的社会建设[EB/OL]. http://news.xinhuanet.com/newscenter/2007-10/15/content_6883752.htm
[29]中华人民共和国中央人民政府.中共中央国务院关于深化医药卫生体制改革的意见 [EB/OL]. http://www.gov.cn/Jrzg/2009-04/06/content_1278721.htm
[30]Levin,A.,C.F.Lin,and C.S.J. Chu(2002)Unit root tests in panel data[J],asymptotic and finite-sample properties,Journal of Econometrics,Vol.108,1-24
[31]Im, K. S., M. H. Pesaran and Shin(2003)Testing for Unit roots in heterogeneous panels[J], Journal of Econometrics, Vol.115,53-74
[32]Choi, I(2001) Unit root test for panel data[J] Journal of International Money and Banking, Vol. 20,249-272
[33]吕延方,陈磊.面板单位根检验方法及稳定性的探讨[J].数学的实践与认识,2010,40(21)：49-61
[34]吴巧生,陈亮,张炎涛,成金华.中国能源消费与GDP关系的再检验—基于省级面板数据的实证分析[J].数量经济技术经济研究,2008,(6).27-40
[35]李子奈,叶阿忠.高级计量经济学[M].北京：清华大学出版社,2000.
[36]曾小敏,刘树仁,彭晓明,等.湖南省1997年卫生总费用影响因素的主成分分析[J].湖南医科大学学报(社会科学版),2000,2(4)：183-88
[37]何平平.我国卫生总费用的弹性测算[J].统计与决策,2006,(5)：87-90
[38]Herwartz, H.,Theilen, B.THE DETERMINANTS OF HEALTH-CARE EXPENDITURE NEW RESULTS FROM SEMIPARAMETRIC ESTIMATION [J].Health economics,2010,19(8):964-978
[39]Carrion-i-Silvestre, JL.Health care expenditure and GDP:Are they broken stationary?[J]. JOURNAL OF HEALTH ECONOMICS,2005,24(5):839-854
[40]Hansen, P., King, A.Health care expenditure and GDP:panel data unit root test results-comment[J]. JOURNAL OF HEALTH ECONOMICS,1998,17(3):377-381
[41]李秀娟.吴洋.医疗费用增长的原因及对策思考[J].卫生软科学,2002,16(3)：25-27
[42]黄成礼.人口老龄化对乙生费用增长的影响[J].中国人口科学,2004,(4)：36-43
[43]黄成礼.人口因素与卫生费用的关系[J].人口研究,2004,28(3)：24-30
[44]何平平.我国医疗费用增长因素的计量分析[J].Pacific Journal,2005,(11)：25-31
[45]何平平.我国卫生总费用的弹性测算[J].统计与决策,2006,(5)：87-90
[46]刘柏,赵振全.卫生总费用长期均衡和短期波动的经济计量检验[J].中国卫生经济,2008,27(6)：5-7
[47]刘树仁,周欢,彭晓明.卫生总费用影响因素的岭回归分析[J].卫生经济研究,2000,(9)：12-14
[48]赵郁馨.卫生总费用运行特点及增长因素[J].中国卫生经济,2002,21(3)：32-35
[49]刘国祥,宋泽萍,张长青.中国贫困地区卫生费用影响因素分析[J].中国卫生经济,1995,14(7)：24-25
[50]何平平.中国卫生总费用增长因素研究[J].统计与信息论坛,2006,21(1)：37-41
[51]郑云萍,温小霓.我国卫生总费用增长研究[J].中国卫生经济,2006,25(12)：16-18
[52]李军山,江可申,李雪强.我国医疗费用增长因素实证分析[J].价格月刊,2008,(10)：56-58
㈨何平平.协整分析与误差修正模型—经济增长、人口老龄化与我国医疗费用增长的实证研究[J].工业技术经济,2006,25(1)：122-124,135
[54]赵郁馨,高广颖,杜乐勋.中国卫生总费用发展变化趋势及其影响因素[J].卫生经济研究特稿,2000,(1)：7-9
[55]何平平.我国卫生总费用的弹性测算[J].统计与决策,2006,(5)：87-90
[56]危小军,廖伟,钟一鸣,等.遏制医疗费用快速增长构建和谐医患关系[J].赣南医学院学报,2007,27(5)：776-777
[57]金新政,张辉.卫生总费用膨胀之灰关联成因分析[J].卫生经济研究,1995,(12)：11-12
[58]李军山,江可申,李雪强.我国医疗费用增长因素实证分析[J].价格月刊,2008,(10)：56-58
[59]蔡善荣,阮红芳,李鲁,等.以主成分回归分析方法探讨人均卫生费用影响因素的研究[J].中国卫生事业管理,2001,(7)：399-400
[60]刘明霞,仕泉.自组织数据挖掘在卫生总费用预测中的应用[J].生经济研究,2003,(12)：10-12
[61]何平平.我国医疗费用增长因素的计量分析[J].PACIFIC JOURNAL,2005,(11):25-31
[62]蔡善荣,阮红芳,李鲁,等.以主成分回归分析方法探讨人均卫生费用影响因素的研究[J].中国卫生事业管理,2001,(7)：399-400
[63]雷海潮,刘兴柱,卞鹰.慢性病增加对我国卫生总费用及宏观经济的影响研究[J].中国卫生经济,1996,15(10)：27-28
[64]中国卫生总费用课题组.中国卫生总费用影响因素的计量经济学分析[J].中国卫生经济,1996,15(12)：27-28
[65]张宜民,顾雪非,张金齐.1990—-2002年我国卫生总费用筹资水平、结构及变化趋势分析[J].中国初级卫生保健,2004(10)：15-18
[66]陈琳红,肖云.我国卫生总费用筹资水平与结构分析[J].西部论坛.2010,20(6)：72-77
[67]耿瑛,裴丽昆,David Legge.增加政府投入提高医药卫生服务的公平性和可及性[J].中国卫生经济.2009,28(3)：12-14
[68]刘波,周雅静.我国医疗费用增长与经济增长的相关性分析[J].黑龙江对外贸易.2010,(9)：86-88
[69]钱东福,尹爱田,王鸿勇等.卫生事业发展状况的国际比较分析[J].中国卫生经济.2007,26(6)：17-19
[40]孟庆跃.卫生总费用、健康产出和卫生改革的国际比较和分析[J].卫生经济研究.2003,22(4)：23-25
[71]胡瑞荣,马旭东.“十一五”期间我国卫生总费用变化趋势及其影响因素分析[J].中国医疗保险.2011,(2)：40-45
[72]何平平.我国医疗支出增长因素研究[D].硕士学位论文.北京邮电大学.2007
[73]汪金鹏.1996-2002年我国卫生总费用筹资水平和机构分析[J].2006,(1)15-16
[74]卫生工作会新闻稿之四.卫生部明确提出“十二五”卫生发展的总体目标[EB/OL]. http://www.moh.gov.cn/publicfiles/business/htmlfiles/mohbgt/s3582/201101/50268.htm
[75]王朝昕,王颖等.适宜的中国卫生筹资总量分析[J].中国卫生资源.2011,14(1)：65-66
[76]中国新闻网.央行报告：2009年来人民币对美元汇率保持稳定[EB/OL]. http://www.chinanews.com/cj/cj-gncj/news/2009/11-11/1959823.shtml
[77]人民网.20日人民币汇率中间价：1美元对人民币6.5294元[EB/OL]. http://www.chinanews.com/fortune/2011/05-24/3061671.shtml
[78]李松光,王颖等.适宜的中国卫生筹资构成探讨[J].中国卫生资源.2011,14(1)：67-68
[79]中华人民共和国中央人民政府.温家宝在十一届人大四次会议上所作政府工作报告[EB/OL]. http://www.gov.cn/2011lh/content_1825233.html
[1]Mundlak Y. Empirical Production Function Fee of Management Bias[J].Journal of Farm Economics,1961,(43):44-56
[2]Nerlove,M. Estimation and identification of Cobb-Douglas production function[M].Chicago:Rand Mcnally & Company,1965
[3]白仲林.面板数据的计量经济分析[M].天津：南开大学出版社,2008
[4]白仲林.面板数据模型的设定、统计检验和新进展[J].统计与信息论坛.2010,25(10)：3-12
[5]Allison P D. Causal inference with panel data [EB/OL].http: pauldallison.com/downloads/Causal%20Inference.pdf.
[6]Kasprzyk D., Duncan GJ. Panel surveys[M].New York:John Wiley,1989:16-18
[7]Fitzgerald J, Gottschalk P. An analysis of sample attrition in panel data[J]. Journal of Human Resources,1998,33(2):251-299
[8]Abauf N, Jorian P. Purchasing power parity in the long run[J].Journal of Finance,1990,45(1): 157-174
[9]白仲林.同期相关面板数据退势单位根检验的小样本性质[J].数量经济技术经济研究,2006(5)：146-152
[10]段鹏,白仲林,张晓峒.个体协整条件下面板单位根检验统计量性质研究[J].统计研究,2009(4)：91-100
[11]Baltagi B H. Econometric analysis of panel data[M]白仲林,译.北京：机械工业出版社,2010：275-284.
[12]武大勇.计量经济学中的面板数据模型分析[D].武汉：华中科技大学,2006
[13]Marc Nerlove. An essay on the history of Panel Data Econometrics. The Ninth International Conference on Panel Data,Geneva,2000:4-9
[14]Hsiao C. Analysis of panel data.2nd Edition, Cambridge University Press,Cambridge,2003,1-5
[15]长自武.平行数据的单位根检验及其模型应用.城镇居民收入和消费的实证分析[D].乌鲁木齐：新疆大学,2007
[16]Bhargava, A., L. Franzini and M. Narendranathan. Serial Correlation and Fixed Effects Models [J]. Review of Economic Studies 1982,49(4):533-549
[17]Abuaf,N. and P. Jorion. Purchasing Power Parity in the Long Run[J]. The Journal of Finance, 1990,45(1):157-174,.
[18]Levin A. and C.F.Lin.Unit Root Tests in Panel Data, Asymptotic and Finite-sample Properties, UC San Diego.Working Paper,1992:92-23
[19]Flores,R., Jorion, P..Multivariate Unit Root Tests of the PPP Hypothesis [J]. Journal of Empirical Finance,1999,6(4):335-353
[20]O'Connell.The Overvaluation of Purchasing Power Parity [J]. Journal of International Economics,1998,44(1):1-19
[21]Taylor, M. and L. Sarno. The Behavior of Real Exchange Rates during the Post Bretton Woods Period [J]. Journal of International Economics,1998,46:281-312
[22]Maddala G.S. and Wu Shaowen, A Comparative Study of Unit Root Tests with Panel Data and a New Simple Test, Oxford Bulletin of Economics and Statistics,1999,61,631-52
[23]Moon, H.R. and B. Perron.. Testing for A Unit Root in Panels with Dynamic Factors[J]. Journal of Econometrics,2004,122(1):81-126
[24]Smith, L.V.. More Powerful Panel Data Unit Tests with an Application to Mean Reversion in Real Exchange Rates [J].Econometrics,2004,19:147-170
[25]Pedroni, P. Panel Cointegration:Asymptotic and Finite Sample Properties of Pooled Time Series Tests With an Application to the PPP Hypothesis Indiana University [J].Working Papers In Economics,1995,20(3):95-113
[26]Kao, C. Spurious Regression & Residual-Based Tests for Cointegration in Panel data[J]. Journal of Econometrics,1999,90:1-44
[27McCoskey, S., and C. Kao.A Residual-Based Test of the Null of Cointegration in Panel Data [J].Econometric Reviews,1998,17(1):57-84
[28]Pedroni, P. On the Role of Cross Sectional Dependency in Panel Unit Root and Panel Cointegration Exchange Rate Studies.Working paper, Indiana University,1997
[29]Pedroni, P. Critical Values for Cointegration Tests in Heterogeneous Panels with Multiple Regressors [J]. Oxford Bulletin of Economics and Statistics,1999,61(1):653-670.
[30]Pedroni, P. Fully Modified OLS for Heterogeneous Cointegrated Panels [J].Advances in Econometrics,2000,15:93-130.
[31]Pedroni, P. Purchasing Power Parity Tests in Cointegrated Panels [J]. Review of Economics and Statistics,2001,83(4):727-31.
[32]Joakim Westerlund. A Panel CUSUM Test of the Null of Cointegration [J].OXFORD BULLETIN OF ECONOMICS AND STATISTICS.2005,67(2):231-256
[33]Westerlund,J..New Simple Tests for Panel Cointegration[J].Econometrics Reviews,2005,24(3): 297-316
[34]Balestra, P.,M. Nerlove.Pooling Cross Section and Time Series Detain the Estimation of a Dynamic Model:The Demand for Natural Gas[J],Econometriea,1966,34:585-612.
[35]Costas MEGHIR. Frank WINDMEIJER. Moment Conditions for Dynamic Panel Data Models with Multiplicative Individual Effects in the Conditional Variance. ANNALES D'ECONOMIE ET DE STATISTIQUE.1999:55-56
[36]Ping Zhang. GENERALIZED ACCUMULATED PREDICTION ERROR AND MODEL SELECTION FOR CATEGORICAL PANEL DATA[J].Statistica Sinica.1997,7:527-544
[37]OLAF HUBLER. Multilevel and nonlinear panel data models[J].Allgemeines Statistisches Archiv,2006,90(1):121-136
[38]ALBAN THOMAS.Consistent estimation of binary-choice panel data models with heterogeneous linear trends[J]. Econometrics Journal,2006,9(2):177-195.
[39]Yu,YH. Determinants of variations in state per capita personal income:a panel data approach[J]. APPLIED ECONOMICS LETTERS,2010,17(3):235-329
[40]De Bandt O, Bruneau C, E1 Amri W.Convergence in household credit demand across Euro area countries:evidence from panel data[J]. APPLIED ECONOMICS,2009,41(27):3447-3462
[41]Ogwang T, Cho DI. Economic determinants of the consumption of alcoholic beverages in Canada:a panel data analysis[J]. EMPIRICAL ECONOMICS,2009,37(7):599-613
[42]Pampel FC. Population aging, class context, and age inequality in public spending[J].American Journal of Sociology,1994,100:153-195.
[43]Kilbourne BS, Farkas G, Beron K, England P.1994.Returns to skill, compensating differentials, and gender bias:effects of occupational characteristics on the wages of white women and men[J]. American Journal of Sociology,1994,100:689-719
[44]Osgood DW, Wilson JK, O'Malley PM.. Routine activities and individual deviant behavior[J]. American Sociological Review,1996,61:635-655.
[45]Mazzanti, Massimiliano.The dynamics of landfill diversion:Economic drivers, policy factors and spatial issues Evidence from Italy using provincial panel data[J]. Resources Conservation and Recycling,2009,54(1):53-61
[46]Sousa TRV, Leite PAM. Panel data analysis of health status in Northeast Brazil[J]. REVISTA DE SAUDE PUBLICA,2008.42(5)
[47]McGuire TG, Ayanian JZ, Ford DE, Henke REM, Rost KM, Zaslavsky AM. Testing for statistical discrimination by race/ethnicity in panel data for depression treatment in primary care[J]. HEALTH SERVICES RESEARCH,2008,43(2):531-551
[48]Paul Hansen, Alan King. The determinants of health care expenditure:A cointegration approach[J]. Journal of Health Economic.1996,15(1):127-137
[49]Suzanne K. McCoskey, Thomas M. Selden. Health care expenditures and GDP:panel data unit root test results[J]. JOURNAL OF HEALTH ECONOMICS.1998,17(3):369-376
[50]Paul Hansen, Alan King. Health care expenditure and GDP:panel data unit root test results-comment[J]. JOURNAL OF HEALTH ECONOMICS,1998,17(3):377-381
[51]Baltagi,BH., Moscone, F. Health care expenditure and income in the OECD reconsidered: Evidence from panel data[J]. ECONOMIC MODELLING,2010,27(4):804-811
[52]Moscone,F.,Tosetti, E. HEALTH EXPENDITURE AND INCOME IN THE UNITED STATES[J]. HEALTH ECONOMICS,2010,19(12):1385-1403
[53]白仲林.面板单位根检验理论及其应用研究[D].天津：南开大学,2005
[54]任燕燕,朱孔来.平行数据模型中的异方差问题的处理[J],统计研究,2005(2),75-77.
[55杨继生,王少平,艾春荣.工具变量法综列单位根检验的有偏性及其修正.数量经济技术经济研究,2006(2)：138-147
[56]张晓峒,白仲林.退势单位根检验小样本性质的比较[J],数量经济技术经济研究,2005(5).
[57]苏明,徐利君.基于面板数据的地方税收与经济差异性研究[J].经济理论与经济管理,2008(3)：57-62
[58]李小平,朱钟棣.国际贸易的技术溢出门槛效应—基于中国各地区面板数据的分析[J].统计研究,2004(10)：27-32
[59]郭庆旺,吕冰洋.税收增长对经济增长的负面冲击[J].经济理论与经济管理,2004,(8)：18-23
[60]郑义,徐康宁.结构变迁、工资差异与经济增长—基于75个国家和地区1978—2007年面板数据的实证研究[J].财经研究,2011,37(3)：17-26
[61]奉琼.外商直接投资与中国的水污染—基于面板数据的分析[J].经济研究导刊,2009,(12)：158-159
[62]李兰.基于省际面板数据的我国EKC实证研究[J].学术交流,2009,(3)：35-38
[63]谢童伟.中国县级教育水平与县人口迁移相互影响分析—基于2004年-2008年31省(市)县级面板数据的实证研究[J].清华大学教育研究,2011,32(1)：84-88
[64]宋帮英,苏方林.我国东中西部碳排放量影响因素面板数据研究[J].地域研究与开发.2011,30(1)：19-24
[65]周先波,田凤平.中国城镇和农村居民医疗保健消费的差异性分析——基于面板数据恩格尔曲线模型的非参数估计[J].统计研究.2009,26(3)：51-58
[66]徐文.对影响我国居民人均就医次数的因素的实证研究——基于一组面板数据的分析[J].消费导刊,2009,(15)：201-202
[67]徐文.对农村医疗保健消费的地区差异及影响因素的分析——基于山东省农村一组面板数据的分析[J].消费导刊,2008,(22)：6-9
[68]于东子.卫生投资对经济增长贡献的定性和定量研究.[D].杭州：浙江大学,2006