Models of well-being and ill-being

The purpose of this paper is to present models of well-being and ill-being which integrate the effects of different types of independent variables. Using the more powerful techniques provided by the LISREL software, the paper replicates and extends analysis previously reported in SIR (Headey, Holmstrom and Wearing, 1984a). A preliminary issue is whether well-being and ill-being constitute different dimensions or whether they are polar opposites. Factor analytic evidence, derived from two waves of an Australian panel study (1981, 1983) with samples of 942 and 878 respectively, confirms that indicators of well-being and ill-being form distinct, although not orthogonal dimensions. In the models of well-being and ill-being estimated from the panel data, we attempt to integrate research dealing with the impact of social background, personality, social networks and satisfaction with particular domains of life. Two key personality traits which influence both well-being and ill-being are self-esteem and personal competence. Social background (SES) has greater influence on ill-being than well-being. Having a well-developed social network, on the other hand, contributes more to enhancement of well-being than relief of ill-being. This is largely because a rich social network is associated with satisfaction with leisure, friends and marriage, which themselves are the life domains most closely connected to feelings of well-being. By contrast, the domain of health is relatively closely associated with ill-being. The conclusion discusses public policy implications of the finding that well-being and ill-being have different correlates and causes. Conventional welfare policies are designed to relieve ill-being. Quite different policies (“positive welfare” policies) are required to enhance well-being.