| BackgroundHappiness is an eternal topic of human beings.During the past decades,not only academic researchers have increased interest in the topic,more and more discussions are being held among policy makers,because its potential could shed light on the political,economic,social,and psychological conditions of populations.Now happiness is being used widely as an indicator for the evaluation of policies in many sections,including quality of life,health care,public health,social services,etc.With the aging of population in China,health conditions,elderly care and quality of life among the seniors have become a challenge for society,which needs to be solved urgently.According to the 2018 Statistical Bulletin of the National Bureau of Statistics,population aged 60 and over has reached about 250 million,accounting for 17.9%of the total population.According to the Blue Book on Aging:Survey Report on the Living Conditions of China’s Urban and Rural Older Persons(2018),only about 30%of the elderly are in good health,and the main problems the elderly population facing are low income,functional limitations,and underdeveloped constructions.Moreover,according to the China Family Development Report(2015)issued by the National Health and Planning Commission,64.3%of families in China are dominated by core family consisting of husband and wife,and children;Rural left-behind children accounted for 35.1%of all rural children,and left-behind seniors accounted for 23.3%of all the rural seniors.On the one hand,rapid aging of population structure,domestic migration and shrinkage of family size are bringing pressures on the traditional elderly care of family.On the other hand,the society is not yet prepared for the coming of elderly service.Especially in the context of modernization,the elderly group is easy to be marginalized due to low social participation,all of the listed situations may impact senior happiness.Therefore,whether the elderly in China can live happily is an urgent issue for individuals,families and society.At the time of data collecting,although the basic old-age security system,the basic medical security system,and the minimum living security system had been established in urban and rural areas,and covered most of the seniors.However,level of protection is still low,most of the elderly have to rely on their children economically,and the living standard is low.In addition,due to the divided society between urban and rural areas long time in the past,migrant population and regional unbalanced development,urban,rural and migrant seniors have different social security,economic conditions,social participation and availability of public resources,could these social conditions lead to different levels of happiness among senior groups?This is one of the objectives of this study.Social integration refers to the acceptance and recognition by society,and gets support by social groups in need.As to the research of social integration,migration is mainly the subject in most studies.The important dimensions of social integration include economic integration,social communication,political integration and cultural integration.The World Health Organization has been promoting the concept and actions of social integration in disease prevention,health promotion and health care.In China,some studies have started with the constructions of conceptual dimensions and contents,anyhow,empirical studies are still in the primary stage without a universal concept of social integration.Based on the social correlates of happiness,this study started with constructing dimensions of social integration and further exploring the possible factors related to senior happiness from the perspective of social integration.Cultural integration was not taken into consideration in this study because of minor cultural difference from the domestic migration.It is hypothesized that the higher level of social integration and the greater happiness of the elderly group is supposed to present.On the basis of the CHARLS questionnaire and data,variables related to the dimensions of social integration were identified at the individual and community levels,which were analyzed for the associations with happiness and were detected for the difference among different senior groups from the perspective of social integration.ObjectivesThis study began with the meaning of happiness study,situation of population aging in China,the problems faced by different senior groups,and then the measurement of happiness,social correlates of happiness,social integration theory and conceptual dimensions were discussed.On this basis,variables related to social integration were extracted from CHARLS questionnaire to construct the research framework in the perspective of social integration.Descriptive analysis of demographic and physical conditions,family characteristics,and features of individual level social integration were presented first;Variables were analyzed to understand the association with happiness among different seniors groups by using multiple logistic regression models,and further to be analyzed from the perspective of individual level social integration;Variables related to happiness were further analyzed by using multi-level regression model so as to understand the relation of community,community characteristics and community-level social integration with happiness;Finally,based on the results,policy implications were presented.The concrete research objectives are listed as followings:(1)To understand the situation of population aging,the problems faced by different senior groups;to understand happiness measurement,happiness related variables,and the relation between happiness study and social integration.(2)Based on the conceptual dimensions of social integration and the CHARLS questionnaire,the variables related to social integration were extracted to construct the framework for the study.(3)To describe and analyze the associations of characteristics of demographic and physical conditions,family characteristics,and features of individual level social integration with happiness,by use of chi-square test.(4)Happiness related variables were further analyzed among seniors groups by using multiple logistic regression models.Meanwhile,the correlations of individual level social integration and family with happiness were compared,based on the deviance of likelihood ratio and ROC contrast estimations.(5)Based on the multi-level regression model,the correlations of community,community characteristics and community level social integration with happiness were analyzed among different elderly groups.(6)Based on the results and discussion,policy implications in perspective of social integration were presented,targeting different groups of the elderly.Data and methods1.DatasetsThe dataset used in this study were extracted from the 2013 China Health and Retirement Longitudinal Study(CHARLS)where there was the happiness scale and the 2011 CHARLS community baseline data.In this paper,the elderly aged 60 and over were used as research samples,and the final sample size was 8911.Among them,men accounted for 48.77%,and only 20%of the sample size aged 75 years old or over.2.Variable extraction and dimension construction of social integrationThe definition of happiness depended on the measurement of HWB-12 scale.According to the study framework,three domains:demographic and physical conditions,family characteristics,and domain of individual level social integration were first constructed with groups of happiness-related variables extracted from CHARLS questionnaires.Social integration is studied from individual level and community level.Individual level social integration includes economic integration,participation in social life and political integration;and community level social integration includes the region where the community is located(the Eastern,Central and Western regions categorized by the National Bureau of Statistics),and number of community facilities and organizations related to elderly services.3.Statistical methods(1)Correlations of variables from the three domains with happiness were presented by use of the univariate analysis,and some variables were selected for multiple regression analysis.(2)The selected variables were further analyzed based on multiple logistic regression models among the senior groups.The correlation strength was showed by odds ratio OR and 95%CI.Meanwhile,the regression models were evaluated using ROC(Receiver Operating Characteristic Curve)and AUC(Area under the ROC Curve).(3)Slight collinearity of the independent variables may extrude the variables with significant contributions,due to the traditional multiple regression analysis.VIP(variable important for projection VIP)score could visually show the contributions of all independent variables to happiness by use of partial least squares(PLS)regression models.(4)The correlations of individual level social integration and family with happiness were compared and tested by using the Deviance of Likelihood Ratio(LR)and the ROC contrast estimation(Mann-Whitney C statistic)from nested logistic regression models.(5)According to the hierarchical characteristics of CHARLS data structure,multi-level regression models were adopted to analyze the correlations of community,community characteristics and community level social integration with happiness among elderly groups.Two-level regression models were used while individuals as low level and community as high level.First,based on the random effects of the empty model:the residual variance σ2μ0 and the intra-class correlation coefficient(ICC)from community level,whether the community had an impact on happiness and the magnitude of the effect could be determined;then,based on the fixed effects of multi-level regression models,the correlations of community characteristics and community level social integration with happiness were presented.Statistical analysis was performed using Statistical Analysis System(SAS University Edition)software.Results1.Sample characteristicsThere were 5,054 rural seniors,1,737 migrant seniors and 2,120 urban seniors in the sample.According to the definition of happiness,55.72%of rural seniors were happy,while 67.07%of migrant seniors and 70.90%of urban seniors were happy.There was significant difference in happiness percentage among the senior groups;the happiness percentages of the elderly in the western,central and eastern regions were:54.49%,63.02%,67.08%,there was significant regional difference.2.Variables related to happiness by univariate analysisA total of 21 independent variables were selected from three domains.Among them,8 variables from domain of demographic and physical conditions:age,gender,marriage,education,self-rated health,chronic diseases,limitations in activity,and self-reported memory;6 variables from family characteristics:number of children,living with children,weekly contact with children,grandchildren care,any transfer from children,and perceived future help from children;7 variables from domain of individual level social integration:medical insurance(employee medical insurance or public medical insurance),pensions,living standard compared with neighbors,monthly consumption,housing conditions,communist party membership,and participation in social life.3.Variables related to happiness by multiple logistic regression models among the senior groups,the common findings were:(1)Chronic diseases,self-rated health,self-reported memory,limited activity,and the older were significantly associated with happiness.(2)More children,living with children,and weekly contact with children were found no significant relation with senior happiness.But perception of future help from children was a significant positive variable correlated with happiness.(3)There were significant positive correlations between pensions,higher living standards and better housing conditions with happiness,while party membership and medical insurance were not significant variables.4.Variables related to happiness by multiple logistic regression models among the senior groups,the different findings were:(1)For rural elderly,grandchildren care(OR:0.86,95%CI:0.75-0.99)and more monthly consumption(OR:0.86,95%CI:0.75-0.99)were significantly negative correlated with happiness.(2)For migrant elderly,there was significant positive correlation between the participation of social life and happiness(OR:1.44,95%CI:1.13-1.84).(3)For urban elderly,any transfer from children had significant positive correlation with happiness(OR:1.41,95%CI:1.12-1.76).Although the weekly contact with children,medical insurance,and party membership did not reach statistical significance based on regression analysis,the VIP score graph showed that the contribution to happiness was similar to the significant variable such as any transfer from children.5.Correlation comparison of individual level social integration and family with happiness based on the Deviance of Likelihood Ratio(LR)from nested logistic regression models(1)For rural elderly,the Deviance of family and social integration were:224.42(DF=6),126.51(DF=7),and the chi-square test had significant significance.(2)For migrant elderly,the Deviance of family and social integration were:40.35(DF=6),39.98(DF=7),and the chi-square test was significant,but the roles of individual level social integration was similar to that of the family.(3)For urban elderly,the Deviance of family and social integration were:43.16(DF=6),37.66(DF=7),and the chi-square test was significant as well,but the family had stronger roles than individual level social integration did.6.Correlation comparison of individual level social integration and family with happiness based on the ROC contrast estimation from nested logistic regression models(1)For rural elderly,the ROC contrast estimations of family and social integration were:0.014,0.0157,and the chi-square test of both was P<0.001.(2)For migrant elderly,the ROC contrast estimations were:family 0.00696(chi-square test P>0.1),social integration 0.0168(chi-square test P<0.01),and role of individual level social integration in happiness was greater than that of family.(3)For urban elderly,the ROC contrast estimations were:family 0.015(chi-square test P<0.01),social integration 0.00854(chi-square test P<0.05),and family was more important for happiness than individual level social integration.7.Whether the community was related to happiness based on the empty-model of the multi-level regression models(1)For rural elderly,σ2μ0 and the corresponding standard error were 0.2227(0.042),P<0.001.The intra-class correlation coefficient ICC was 0.06342,which can be explained as 6.34%of happiness attributed to the community.(2)For migrant elderly,σ2μ0 and the corresponding standard error were 0.2488(0.078),P<0.001.The intra-class correlation coefficient ICC was 0.07033,which can be explained as 7.03%of the happiness attributed to the community.(3)For urban elderly,σ2μ0 and the corresponding standard error were 0.0931(0.0558),P<0.05.The correlation coefficient between groups was ICC:0.02753;it can be explained that 2.75%of happiness is attributed to the community.8.Correlations of community characteristics and community level social integration with happiness based on the fixed effects of multi-level regression models(1)The number of community facilities and organizations was an importantvariable for senior happiness.The regression coefficients and corresponding standard errors were:for rural elderly 0.03397(0.01819),P<0.10;migrant elderly 0.04895(0.02101),P<0.05;urban elderly 0.0395(0.01849),P<0.05.(2)In comparison with a community in the western region,the regression coefficient and corresponding standard error of a community located in the central region:0.4199(0.1029),P<0.001;the regression coefficient of a community in the eastern region and the corresponding standard error:0.5249(0.09869),P<0.001.Both are significantly positively correlated with the happiness of rural elderly,but there was no statistical difference between the community locations and the happiness of migrant or urban elderly,indicating little regional difference among urban communities.9.Correlation comparison of family,individual level social integration and community level social integration with happiness based on the Deviance of Likelihood Ratio(LR)from multi-level regression models(1)For rural elderly,the Deviance of family,individual level social integration and community level social integration were:80.45(DF=6),138.54(DF=7),37.16(DF)=3),all the chi-square tests were significant.(2)For migrant elderly,the Deviance of family,individual level social integration and community level social integration were:17.14(DF=6),35.65(DF=7),5.76(DF=3),statistical significance were presented for family and individual level social integration by chi-square test,while the community level social integration was not significant.(3)For the urban elderly,the Deviance of family,individual level social integration and community level social integration were:51.86(DF=6),25.51(DF=7),6.26(DF=3)and the chi-square test had statistical significance for family and individual level social integration,while chi-square test for community level social integration at P<0.10.ConclusionFollowing points could be concluded based on the above results:(1)Different senior groups have different happiness level:urban seniors are the happiest while rural seniors are the unhappiest.(2)In the perspective of individual level social integration:economic integration is most significant for happiness of rural seniors in comparison with political integration and social communication;both economic integration and social communication are important for happiness of migrant seniors;role of economic integration is important for happiness of urban seniors,but not so strong as for rural or migrant happiness(3)In the perspective of community level social integration:community locations(East or middle versus west)are significant for happiness of rural seniors;number of community facilities and organizations are significant for all the senior groups(rural senior group:P<0.10).(4)Social integration versus family roles:both are important for senior happiness However,social integration is more important for happiness of migrant seniors than family;while family is more important for happiness of urban seniors than social integration.(5)Different senior groups presented different happiness related variables:for rural seniors,caring grandchildren is negatively correlated with happiness(P<0.10);for migrant seniors,participation in social life is significant for happiness;for urban seniors,transfer from children to parents was positively associated with happiness,which may reflect parental satisfaction with children.(6)Elderly happiness from traditional family such as many children,co-residence,and frequent contacts no longer exist,but perceived future help from children is important for all seniors.(7)Other:health conditions are most relevant for senior happiness;but medical insurance is not significant for happiness,which needs more evidence.Policy Implication(1)Happiness gap:different happiness levels among senior groups reflect the difference of social integration among senior groups and unbalanced development across geographical areas.The central and western regions,especially rural areas should be helped and invested more in order to narrow the difference gap and improve happiness of rural seniors in the central and western regions.(2)In the perspective of individual level social integration:happiness of rural seniors are more dependent on economic integration,which imply the importance of improving level of social security,living standard,and housing conditions.Besides economic integration,social communication are important for happiness of migrant seniors,which imply the importance of social inclusion in reality.Besides economic integration,happiness of urban seniors is more dependent on family.(3)In the perspective of community level social integration:community location is correlated to happiness of rural seniors instead of migrant or urban seniors,which may imply the underdevelopment of rural areas.Number of community facilities and organizations are significant for senior happiness,which confirms that the policy of community construction is right and more senior-friendly communities are needed.(4)Family roles:family is still important for happiness of all the seniors.Family policy should focus on relieving the burden of grandchildren care and elderly care instead of emphasizing the form of filial culture.(5)Different strategies should be made for happiness of different senior groups:improvement of living standard and helping with caring grandchildren may raise happiness levels of rural seniors;more inclusions of migrant seniors in various activities could create chances for social communication;visits from children are supposed to bring urban seniors happiness.(6)Others:because health conditions are most relevant for senior happiness.Health policies should promote healthy life styles,prevent chronic diseases,provide health care for all the seniors,and elderly care for disabled seniors.Strength and Weakness1.Strengths(1)Topic selection:China’s long-term dualistic society,leading to different populations.Rural,migrant,and urban seniors have different levels of social security,living environment and social backgrounds.The study about social integration and happiness have practical meaning in reality.(2)Content:It was the first time to extract relevant variables from CHARLS questionnaire to construct the dimensions of individual level and community level social integration.Based on conceptual and measurement frameworks of social integration,results could be explained in the perspective of social integration.The hierarchical nature of the data provides an opportunity to analyze social integration at the community level.(3)Happiness Scale HWB-12 was used for the first time in the measurement of elderly happiness in China.According to the literatures,this happiness scale is more sensitive to reflect the social factors than the life satisfaction.The evaluation of the scale could be used as a reference for future cross-country comparison studies of senior happiness.2.WeaknessesThis study is mainly based on the 2013 cross-sectional data,and the correlation cannot be inferred to be causal.Dimensions of social integration was constructed according to the CHARLS questionnaire,which has certain objective limitations.As a remedy,it is better to have some interviews or field surveys.Moreover,HWB-12 is an imported scale,whether it is suitable for surveys under Chinese culture and social background should be taken into consideration.Language validation is a necessary procedure in application.The results are based mainly on analysis of CHARLS data from 2013,because China changes rapidly in social and economic development,the policy implication needs careful interpretation in terms of "outdated". |
PDF Full Text Request