Veenhoven (1998) and Cummins and colleagues (1998) have proposed a systems-theory approach to understanding well-being. In an extensive review of quality-of-life measures, Hagerty and colleagues (2001) argued that none of the 22 quality-of-life scales that they evaluated were based on a well-established theory (that is, an empirically supported “nomological net” of concepts and causal paths that specify how quality-of-life is related to exogenous and endogenous variables), and they proposed a systems-theory approach as a potential solution.
This approach distinguished between inputs, throughputs and outputs. In Hagerty et al.’s (2001) account, input variables are environmental factors that influence quality-of-life, such as Gross Domestic Product (GDP), political freedom and health services. Throughput variables refer to an individual’s reactions and choices in this environment. Quality-of-life measures typically use objectives measures as throughput variables, such as achieved education and personal health. Finally, output variables measure the results of the input and throughput variables. Veenhoven (1998) cites domain-specific and overall SWB, personal survival, and “contribution to the human heritage” as examples of output variables. It is important to note that output variables can have causal feedback effects on both input and throughput variables, influencing them either positively or negatively.
My early theoretical work (Jayawickreme, Forgeard, & Seligman, 2012) adopts their framework as a direct antecedent, which distinguishes between inputs, processes and outcomes for individual well-being. What goes into the three classes of variables, however, distinguishes this framework from theirs. The relevant concepts are summarized under the appropriate heading below.
This model includes as input variables two kinds of influences: exogenous and endogenous predictors of well-being. First, exogenous predictors include environmental variables such as income, education, and genetics. These input variables, such as income, green space, and clean water, fulfill exogenous needs and predict outcomes associated with well-being. The objective-list approach (below) consists of a compilation of exogenous input variables. These variables including resources and income—afford the opportunity to engage in valuable activities—and so contribute towards an individual’s well-being. Second, we also add personality variables—these are traits that predict well-being (Ryan & Deci, 2000). These endogenous variables are traits that include optimism, neuroticism, curiosity, abiding values, strengths and talents, and the trait of positive affectivity, which are all related to well-being.
Process variables are internal states that influence the choices that individuals make; the outcomes of these choices are the behaviors that constitute the outcome variables. Following Carver and Scheier’s (1981) self-regulatory model, individuals respond to their environment by engaging in activities to achieve their goals. This is equivalent, utilizing Sen’s (1999) language below, to their choosing between different capabilities in order to achieve functionings. These choices can be affected by a number of variables, including specific beliefs or cognitions that they may have regarding their choices, the explanations they make, moods, emotional states that are consequences and correlates of the choices. Importantly the engine places the subjective variables, such as mood, positive emotion, and cognitive evaluations in the process part of the model. Note that while Hagerty et al.’s (2001) definition of a throughput focuses on objective measures of choice, the category of process defined here focuses on capabilities and subjective states.
Outcome Variables: Preferences, Behavior, and Goal-Driven Functionings
The outcomes of the engine approach are the voluntary behaviors that characterize well-being: positive relationships, positive accomplishment, engagement in work, love or play, authentic, autonomous behavior, and meaningful activity. Following Seligman (2011) and Sen (1999), the approach defines well-being outcomes in terms of what people, when free from coercion, would choose to do for their own sake. Although individuals may sometimes pursue these outcomes for other ends (e.g., they may for instance think that accomplishment will bring positive emotion), many choose to do so because these outcomes are intrinsically motivating by themselves. Such an outcome should satisfy three conditions:
- It contributes to well-being and a life well-lived
- Many people pursue it for its own sake, not merely to get any of the other elements
- It is defined and measured independently of the other outcomes
Such behaviors constitute what Sen (1992) and Nussbaum (2011) define as functionings, or valuable doings that grow out of inputs. Such goal-driven functionings in objective list theories are the activities that individuals engage in to fulfill important goals; such goal-motivated activity is indicative of well-being (Brunstein, Schultheiss, & Grassmann, 1998; Hofer, Busch, & Kiessling, 2008).
Implications of this work
This paper has been the foundation for further discussions on what the specific dimensions of well-being that should be the target of public policy (Forgeard, Jayawickreme, Kern, & Seligman, 2011; Jayawickreme & Pawelski, 2013; see also Brethel-Haurwitz & Marsh, 2014). Targeting and assessing variables at all levels will inter alia help determine the existence and effectiveness of specific levers of change – in other words, the degree to which specific aspects of well-being are malleable or changeable in response to public policy intervention, and the degree to which changes in one lead to changes in others. I have also been working on a large research initiative over the past three years aimed at applying the Engine model framework to higher education populations (Jayawickreme & Dahiil-Brown, 2016; Dahiil-Brown & Jayawickreme, 2016; Jayawickreme, Brocato, & Pryor, in preparation; see also https://www.insidehighered.com/news/2014/04/29/wake-forest-u-tries-measure-well-being). This project will identify the resources, conditions and skills needed for lifelong wellbeing among undergraduate student populations, as opposed to simply assessing wellbeing outcomes. Such information would allow university offices and staff to develop effective programming to support student wellbeing and development of the processes that promote long-term well-being. We have completed two pilot administrations of a multidimensional measure of student well-being, and hope that this measure become widely used among U.S institutions.