Associations between the Importance of Well-Being Domains and the Subsequent Experience of Well-Being

Abstract

Prior cross-sectional research suggests that the importance assigned to well-being domains may be associated with actual self-reported well-being in these same domains. However, cross-sectional data cannot discern directionality, leaving an open question as to whether valuing well-being leads to higher actual well-being or the other way around—higher levels of well-being lead to valuing well-being more. In the present study, we used longitudinal data from 1209 employees to examine the associations between the perceived importance of six well-being domains (emotional health, physical health, meaning and purpose, social connectedness, character strengths, and financial stability) and subsequent well-being in these domains reported approximately 1 year later. Lagged linear regression models demonstrated that valuing character strengths and valuing social relationships were most strongly associated with subsequent well-being. None of the valuations were associated with higher subsequent emotional well-being and only one (importance of physical health) predicted physical health. We also found that higher valuations of character strengths and physical health were associated with lower ratings of subsequent financial stability. A stronger sense of the importance of each well-being domain was predictive of subsequent character strengths. Our findings suggest that living well appears to be achieved by valuing immaterial goods, especially social connectedness and character strengths, as opposed to domains such as financial stability or physical health.

1. Introduction

Research on subjective well-being is increasingly seeking to integrate multiple domains, reflecting an emerging interest in “complete well-being” [1,2,3,4]. The notion of complete well-being minimally requires “doing or being well” in all well-being domains [3]. These domains comprise emotional health, purpose in life, social connectedness, character strengths, physical health, and financial security [3,5], and have been shown to be highly valued by people [3,6]. Most people throughout the world would also insist on including additional well-being domains, such as spirituality [7] or inner peace [8]. Consideration of community well-being, which involves more than a simple aggregation of individuals’ self-reported well-being, is also desirable [9].

Each individual domain of well-being has been thoroughly studied over the years, and some have been included as part of more comprehensive measures. Research on character strengths as an integral facet of well-being has emerged only recently [see for example 6]. Thus, this domain is somewhat underdeveloped relative to the others and is rarely included in multidimensional measures of well-being [3,4,10]. However, it has been a perennial focus in disciplines of the humanities, such as philosophy and theology. It has also received increased attention in the social sciences in recent decades [11,12]. Although the character strengths domain has largely been avoided in some disciplines, it implicitly animates many contemporary scholarly debates [13]. The role of character strengths seems especially relevant for other domains of well-being. In particular, recent research found that it predicts future emotional health, physical health, social connectedness, and meaning and purpose [14,15,16,17]. What remains uncertain is whether valuing character strengths serves as a determinant of these other well-being domains.

Similarly, despite the fact that there is a large body of research on physical health, this domain is not usually included in measures of well-being [3]. The same is true for the financial stability domain [3,18]. However, nearly all well-being measures include items related to emotional health. A single dimension of this domain (life satisfaction) has itself been divided into 173 distinct sub-domains [19]. Unlike most other frameworks, the framework proposed by VanderWeele [3], which informed the design of the measure that is used in this study, incorporates both the neglected and the frequently included domains of well-being.

Life circumstances may require people to make trade-offs among various well-being domains. There is evidence that some people value health over happiness and life satisfaction [20,21,22] or other way around—life satisfaction over health [23]; others place health above their financial situation [24]. Nevertheless, it is generally preferable to do or be well in all well-being domains when that is possible.

Prior cross-sectional research has shown that the six abovementioned domains of well-being are indeed all highly valued and that this valuation is correlated with the experience of the same set of well-being domains [6]. It remains unknown whether valuing certain well-being domains may lead to a better subsequent experience of well-being, a relationship that has not been assessed in previous research. Therefore, the current study poses two research questions:

(1)

How is the importance of a particular well-being domain associated with the subsequent experience of well-being in the same domain? and

(2)

How is the importance of a particular well-being domain associated with the subsequent experience of well-being in other domains?

Since prior research has found that valuing domains of well-being is positively correlated with domains of actual self-reported well-being [6], in this study we tested the following research hypotheses:

H1. 

The importance of a particular well-being domain will be positively associated with the subsequent experience of well-being in the same domain.

H2. 

The importance of a particular well-being domain will be positively associated with the subsequent experience of well-being in other well-being domains.

To test these hypotheses, we used longitudinal data and applied a lagged regression analytic design. Our methodological approach includes adjustment for baseline values of the outcomes, which decreases the risk of reverse causality and provides insight into time-ordered associations among variables [25]. This analytic design has the potential to offer more robust support for directionality of the associations, as it controls for the possibility that higher well-being in certain domains might also lead to valuing those domains more.

2. Materials and Methods

2.1. Participants

Randomly sampled employees of a large, US company were invited to participate in the first wave of the Well-Being Survey in June 2018. Out of 15,000 employees of at least 18 years of age that were invited, 2370 agreed to participate and provided responses (response rate 15.8%). The following year, participants were invited to complete the second wave of the study (July 2019). A total of 1209 employees participated in both waves. This group constituted the analytic sample for this study.

Table 1. Participant characteristics at study baseline (N = 1209).

Baseline Characteristic	Statistic
Gender (females), %	84.5
Age (years), mean (SD)	43.5 (10.4)
Age, %
≤30 years	11.8
31–40 years	29.9
41–50 years	29.0
>50 years	29.3
Race, %
White	74.3
Black or African American	12.2
Hispanic/Latino	6.7
Asian	5.1
Other	1.8
Marital status, %
Single/never married	16.2
Married	62.4
Divorced	10.1
Widowed	1.3
Separated	1.3
Non-married partner	8.7
Education, %
High school	7.8
Some college but no degree	22.6
Associate degree	14.0
Bachelor’s degree	35.0
Graduate school or higher	20.7
Has child dependents, %	48.1
Has older adult dependents, %	27.2
Owns a home, %	72.3
Voted in previous elections, %	82.4
Religious service attendance, %
More than once a week	5.9
Once a week	14.6
1–3 times a month	11.6
Once every few months or once a year	39.8
Never	28.1
Spiritual practices, %
Daily	28.2
Not daily but more than once a week	24.8
Once a week	8.3
1–3 times a month	12.8
Once every few months or once a year	17.8
Never	8.2
Volunteering, %
More than once a week	5.03
Once a week	4.7
1–3 times a month	12.8
Once every few months or once a year	50.5
Never	26.9
Participation in community groups, %
More than once a week	10.1
Once a week	8.4
1–3 times a month	15.8
Once every few months or once a year	33.9
Never	31.9
Salary (USD), mean (SD)	73,117 (34,259)
Salary (USD), %
<40,000	12.3
40,000–49,999	16.3
50,000–59,999	13.5
60,000–69,999	22.5
70,000–79,999	3.6
80,000–99,999	13.8
100,000–119,999	11.8
120,000–139,999	0.0
140,000+	6.2 1

¹ This table was adapted from [26]. The statistics reported in this table are based on non-imputed data.

presents the sociodemographic characteristics of the sample at baseline. Females accounted for 84.5% of the sample (74.5% of employees in the organization were female). The mean age of the sample was 43.5 years (mean age of 45.6 years in the organization). Participants were mostly White and relatively well-educated office employees, which was also consistent with the structure of employees in the target population. Approximately 72% of participants reported owning a home, 62% were married, and 48% indicated that they were caring for at least one minor child. Most participants voted in prior elections (82%). Roughly 28% of the sample reported engaging in daily spiritual practices, and about 20% attended religious services at least once a week. Nearly 10% volunteered at least once a week.

The retention rate from the first wave was 51.2%. Participants who did not complete the second wave were significantly more likely to be young, male, non-White, not married, and not homeowners [14]. For both surveys, an email system within the organization was used to conduct a communication campaign (one week prior to survey administration), distribute letters of invitation, and send participation reminders. The surveys were also administered online, allowing participants to provide responses in a secure and anonymous space of their choice.

2.2. Measures

2.2.1. Predictor Variables

We measured the self-reported importance of the following six domains of well-being with single items: emotional health (“How important is being happy and satisfied with life, having good mental health, and being in control and able to deal with difficult emotions?”); meaning and purpose (“How important is having a sense of meaning in life, a direction to one’s activities, and pursuing what is most important?”); social connectedness (“How important is having close, meaningful, and supportive relationships and being respected by and connected to community?”); character strengths (“How important is having consistent thoughts and actions that contribute to the good of oneself and others?”); physical health (“How important is being sufficiently healthy to be able to carry out the important tasks in life now and into the future?”); and financial security (“How important is having sufficient financial resources and planning so as to be able to pursue one’s life goals and not overly worry about making ends meet?”). Scores ranged from 0 “not important at all” to 10 “extremely important.” The correlation matrix for the self-reported importance variables at baseline is presented in Table S1 in the Supplementary Material. Table S2 in the Supplementary Material presents the cross-lagged correlations between the importance variables after adjusting for the control variables.

2.2.2. Outcome Variables

Our analysis builds upon research that has validated a 40-item comprehensive well-being assessment [5,6], which assesses actual self-reported well-being in six domains: (1) emotional health; (2) physical health; (3) meaning and purpose; (4) character strengths; (5) social connectedness; and (6) financial security. A sample item from the emotional health domain is, “Overall, how satisfied are you with life as a whole these days?” Response categories ranged from 0 (not satisfied at all) to 10 (completely satisfied). A sample item from the character strengths domain is, “I get to use my strengths to help others” (0 = not true of me, 10 = completely true of me). Exact wording of all items can be found in [5].

The measure has good psychometric properties, including evidence of construct validity, convergent and discriminant validity, and test–retest reliability, as well as satisfactory internal consistency for domain-specific and overall scores [5]. The utility of the overall measure and separate sub-scales has also been demonstrated in several recent studies of well-being [14,16,26].

In this study, we calculated scores for each domain by averaging the responses across all items on a given domain. Since some items are negatively oriented, relevant items were reverse coded to ensure that a higher score implies greater well-being.

2.2.3. Control Variables

We included a rich set of control variables that are known to affect well-being. First, we controlled for demographics including: (1) gender (male vs. female), (2) age (≤30, 31–40, 41–50, >50), (3) race (White, Black/African American, Hispanic/Latino, Asian, other), (4) educational attainment (high school, some college, associate degree, bachelor’s degree, graduate degree), (5) marital status (single/never married, married, divorced, widowed, separated, non-married partner), (6) number of child dependents under the age of 18 living in the home, (7) caring for one or more older adult dependent living in the home (yes, no), and (8) wealth [owning a house (yes, no)], and (9) income (salary based on mid-point salary bands obtained from the human resource department of the employer). A number of studies have shown that these variables are predictive of various well-being domains [27,28,29,30].

We also controlled for religious service attendance (more than once a week, once a week, 1–3 times a month, once every few months or once a year, never), spiritual practices (daily, not daily but more than once a week, once a week, 1–3 times a month, once every few months or once a year, never), volunteering (more than once a week, once a week, 1–3 times a month, once every few months or once a year, never), participation in community groups (more than once a week, once a week, 1–3 times a month, once every few months or once a year, never), and voting in the last election (yes vs. no/not sure/not registered voter). Prior research indicated that these factors can influence various domains of well-being, including both emotional and physical health [31,32,33,34,35,36].

In addition to adjusting for all of the abovementioned control variables assessed in the first wave (the same wave as the predictor variables), we attempted to reduce the risk of reverse causality by adjusting for complete well-being assessed in the first wave (a composite of all six well-being domains computed as an arithmetic average).

2.3. Analytic Strategy

The prospective associations were examined using a series of lagged linear regression models with control for the prior outcome and extensive set of covariates. Standardized regression estimates (betas) were presented. Consequently, a set of 36 regression models was used to regress each of the six well-being outcomes on each of the six self-reported importance measures. In particular, the association between a self-reported importance of well-being domain j and a well-being outcome k was modelled as follows:

$$W{B}_{k,j,i}\left(T=2\right)={\alpha }_{0,k,j}+{\alpha }_{1,k,j}IW{B}_{j,i}\left(T=1\right)+{\alpha }_{2,k,j}{X}_i\left(T=1\right)+{\eta }_{k,j,i}$$

(1)

where i = 1, …, N; k = 1, …, 6; j = 1, …, 6.

Subscript i represents an individual, the variable WB indicates one out of six (k = 1, …, 6) well-being outcomes, IWB is one out of six importance of well-being variables (j = 1, …, 6). X is a vector of control variables including the first wave (T = 1) general well-being measure. α_1,k,j reflects an association between an importance of well-being predictor (j) and a subsequent well-being outcome (k). α_2,k,j is a vector that shows the association between control variables and the well-being outcome (k), and η_k,j,i is a disturbance term.

All missing values on the predictor, covariate, and outcome variables were imputed using chained equations (10 datasets were generated) [37,38]. Data were arranged in a wide format as suggested by Allison [39] and all outcome, predictor, and control variables were used in the procedure. We used Rubin’s formula [40] to pool estimates derived from regression models that were performed with each of the imputed datasets.

To examine the robustness of the results to missing data patterns, we reanalyzed the primary sets of models using complete cases. For both the primary and complete-case analyses, Bonferroni correction was used to correct for multiple testing. The interpretation of results focuses on unadjusted results. All statistical computations were performed using Stata 15.

3. Results

Table 2. The prospective associations between the importance of well-being domains and subsequent experience of well-being (standardized estimates/betas and 95% confidence intervals, N = 1209).

	Emotional Health	Meaning and Purpose	Social Connectedness	Character Strengths	Financial Stability	Physical Health
Importance of emotional health	0.022 (−0.022; 0.067)	0.071 ** (0.025; 0.119)	0.041 (−0.005; 0.087)	0.079 ** (0.030; 0.129)	−0.021 (−0.066; 0.024)	0.034 (−0.015; 0.083)
Importance of meaning and purpose	0.009 (−0.038; 0.056)	0.107 *** (0.059; 0.155)	0.037 (−0.011; 0.085)	0.164 *** (0.113; 0.215)	−0.035 (−0.081; 0.011)	0.013 (−0.039; 0.064)
Importance of social connectedness	0.034 (−0.011; 0.078)	0.100 *** (0.053; 0.147)	0.114 *** (0.068; 0.160)	0.120 *** (0.070; 0.170)	−0.008 (−0.053; 0.038)	0.043 (−0.010; 0.095)
Importance of character strengths	0.006 (−0.044; 0.056)	0.092 *** (0.041; 0.144)	0.068 ** (0.020; 0.116)	0.215 *** (0.164; 0.265)	−0.050 * (−0.097; −0.003)	−0.009 (−0.062; 0.043)
Importance of financial stability	−0.011 (−0.056; 0.033)	0.016 (−0.030; 0.062)	−0.011 (−0.058; 0.035)	0.147 *** (0.100; 0.195)	−0.034 (−0.078; 0.011)	0.034 (−0.015; 0.084)
Importance of physical health	0.035 (−0.010; 0.080)	0.049 * (0.003; 0.096)	0.023 (−0.022; 0.069)	0.135 *** (0.086; 0.184)	−0.059 ** (−0.103; −0.015)	0.060 * (0.010; 0.109)

* p < 0.05, ** p < 0.01, *** p < 0.001; The p-value cut-off for Bonferroni correction = 0.05/36 = 0.0014; Estimates significant after correcting for multiple testing are underscored. Importance variables were measured in June 2018. Well-being outcomes were measured in July 2019. We controlled for 2018 well-being levels, race, age, gender, marital status, voting, education, home ownership, salary, child dependents, older adult dependents, religious service attendance, spiritual practices, community participation, and volunteering.

displays the results of our primary analysis. Focusing on the outcomes by examining the table columns, we find that none of the valuations (including importance of emotional health) predicted subsequent emotional health, all of the valuations predicted subsequent character strengths, and all but one valuation (the importance of financial stability) predicted subsequent meaning and purpose. Additionally, valuing social connectedness and character strengths predicted the subsequent experience of social connectedness. Only the importance of physical health was associated with the subsequent experience of physical health, and only two valuations were associated—both inversely—with subsequent financial stability (the importance of character strengths and physical health).

Looking at the rows, the importance of character strengths and importance of physical health were both associated with four outcomes (including a negative association with financial stability). Valuing social connectedness was the next most consistent predictor of the outcomes, evidencing associations with subsequent meaning and purpose, social connectedness, and character strengths. The importance of financial stability predicted only one outcome (character strengths).

The examination of beta estimates measured across importance domains indicated the strongest link between importance and character strengths. The beta coefficients for this outcome were the highest compared to other well-being domains. This might indicate an especially substantial role of importance in cultivating character strengths. The weakest associations found for the set of importance indicators were with the emotional health domain, suggesting that emotional health might not be easily shaped by valuing well-being domains.

The importance of specific domains can be linked to positive changes in the corresponding domain of well-being, but these valuations also affect other domains. In order to understand the strength of these two types of associations, we examined the summary associations between valuations and both corresponding domains and other domains. The effects on the diagonal, which reflect the prospective association between a specific importance domain and its corresponding subsequent well-being domain, yielded an average beta of 0.092 (absolute values were taken into account as the aim was to compare the strength of associations). The off-diagonal betas, which explain cross-relationships between importance and well-being domains, were weaker and averaged 0.052.

Results of the complete case analysis were very similar to those from the primary analysis based on multiply imputed data (see Table S3 in Supplementary Material). Directionality of all associations was preserved and standardized regression coefficients were similar, with the confidence intervals of comparable width. However, the association between the importance of character strengths and the subsequent experience of financial stability no longer excluded the null in the complete-case scenario. Overall, this analysis provided evidence for the robustness of our results to missing data pattern. We also examined the correlations between the importance of domain ratings (see Table S1 in Supplementary Material). All were positive, with the financial stability and emotional health indicators of importance evidencing the weakest correlation and character strengths importance ratings and meaning and purpose importance ratings being correlated the most. As a supplementary analysis, we also examined the longitudinal associations of the 2018 domain importance ratings with subsequent 2019 domain importance ratings, adjusting for control variables (see Table S2 in Supplementary Material). All associations were positive, with the weakest average association obtained for financial stability importance.

4. Discussion

Building on a foundation established by previous cross-sectional research [6], our longitudinal analysis explored the prospective associations between valuing core domains of well-being and the subsequent experience of well-being in these same domains. Providing some support for our hypotheses H1 and H2, we found that most associations were positive. However, we also found that none of the valuations predicted subsequent emotional health, and two valuations were inversely associated with subsequent financial stability (the importance of character strengths and physical health), which was at adds with both tested research hypotheses. Regarding our first research question, the results indicated that the importance of a particular well-being domain is associated with subsequent experience of well-being in the same domain for all domains except emotional health and financial stability. Regarding our second research question concerning the cross-domain associations, we found that valuing two domains—character strengths and social connectedness—yielded the strongest associations with other well-being domains, while valuing character strengths and physical health were prospectively associated with the highest number of well-being domains. Our results also showed that the strongest association was observed between the importance of character strengths and the prospective self-reports of character strengths. Additionally, self-reports of subsequent emotional health (measured in the second wave) were found to be independent of the importance attached to well-being domains in the first wave (which provided no support for hypothesis H1), while each of the valuations predicted subsequent character strengths and five out of six valuations predicted subsequent meaning and purpose (supporting research hypotheses H1 and H2).

Our results indicating that valuing character strengths, emotional health, social connectedness, and sense of meaning and purpose predict subsequent sense of meaning and purpose in life add to evidence on the predictors of purpose and meaning in life that has previously been reported, such as mental well-being, social connections, and having a sense of purpose while at work [41,42,43]. A particularly important contribution of the present study is the finding that importance attached to well-being domains is associated with subsequent self-reported character strengths, which has yet to receive empirical attention. There is an abundance of research on the centrality of social connections and character strengths for the flourishing life [3,16,21,44], but until now the only evidence for an association between valuing these two domains and self-reported well-being in these domains was based on cross-sectional data [6]. Our longitudinal findings align with these previous results, which also found the strongest correlations between valuations and subsequent well-being for these two domains.

It is not surprising that valuing social connections is associated with subsequent well-being. Although valuations were not directly tested, this is one of the primary lessons from the Harvard Grant Study, a longitudinal investigation of human development of a single cohort of men that is now in its 8th decade [45]. After spending more than $20 million to follow these men throughout the course of their lives, one of the study directors famously declared, “Happiness is love. Full stop” (Vaillant, quoted in [45]). The quality of close relationships decades earlier was the most important predictor of a range of well-being outcomes, including happiness, physical health, and longevity, leading study’s current director to declare, “Taking care of your body is important, but tending to your relationships is a form of self-care too” (Waldinger, quoted in [46]).

Beyond tending to relationships, our findings suggest that simply valuing connectedness is associated with an enhanced sense of purpose, the development of character strengths, and the experience of more satisfying social relationships. People who are socialized to appreciate the importance of connectedness may be more likely, as Waldinger put it, to “tend” to their relationships in more effective ways, which could, in turn, increase the experience of well-being in a number of domains. It is interesting that the importance of connections was not associated with emotional health, but in fact none of our importance-related predictors showed evidence of association with this domain of well-being. These results add to prior evidence on the associations between mental health and temporary prior social connectedness [47], character strengths [15,16], meaning and purpose [41,48], and financial well-being [26]. However, they also highlight a crucial difference. Although certain well-being domains may be prospectively associated with one another, the same may not be true for valuations of well-being domains. It may be that valuing various aspects of well-being, while possibly leading to actions that enhance well-being, also perhaps leads to higher standards, thus making it more difficult to be emotionally satisfied. In this vein, our results are somewhat at odds with prior research reporting that individuals are inclined to trade-off levels of happiness with levels of physical health [20], and that these trade-offs substantially depend on the individual’s own levels of happiness and health [22]. Specifically, it has been reported that unhappy people are more likely to choose unhappy lives and unhealthy people are more inclined to prefer unhealthy life [22].

Principles derived from Acceptance and Commitment Therapy [49] might partially account for the lack of associations between valuing emotional health and subsequent self-reported domains of well-being. From this perspective, feeling happy and having good mental health might result from initially prioritizing the ability to embrace negative emotions, seek deeper meaning, and especially strive to live according to deeper values. It must be noted, however, that these emotional health effects may materialize over a longer time frame than the one-year period covered by our study. These principles might, however, help to explain why valuing character strengths in our study was associated with the subsequent experience of greater meaning and purpose, enhanced social connectedness, and higher reported character strengths.

As we have noted, valuing emotional health was the least influential importance domain for subsequent well-being, whereas individuals assigning value to character strengths were most likely to emerge with higher well-being. Similarly, valuing financial stability was only associated with one subsequent well-being domain—character strengths. Long-standing philosophical and religious traditions offer abundant reasons why placing a higher value on character development, as opposed to emotional health or financial stability might result in benefits in other domains of well-being [3,12,14]. Empirical research has been able to confirm some of the pathways that we would expect in light of the wisdom of the humanities. To take just one example, it appears that it is indeed “better to give than to receive”, as a preoccupation with hedonistic enjoyment may adversely affect a sense of inner peace or personal growth, while participation in benevolent service to others is associated with a variety of positive well-being outcomes [8,35]. In the same vein, we found that the increased importance of character strengths showed little evidence of association with the domains of physical health and emotional health, and was negatively associated with the financial stability domain. We speculate that people who value “doing the right thing” might make sacrifices to help others. These actions, in turn, might adversely affect their well-being in one of these domains, including compromising their short-term financial situation or sense of positive affect. Future research is needed to more fully explain these findings. In the meantime, such patterns suggest that the connection between human well-being and sustainable growth requires more direct attention, as hedonistic and material well-being might be at odds with other domains of well-being [6].

Previous research has found that a single survey item related to character strengths (“I always act to promote good in all circumstances, even in difficult and challenging situations”) was associated with a variety of well-being outcomes, including mental health, physical health, social connectedness, and purpose [14,15]. This suggests that actively engaging in behaviors that reflect good character might be more influential on other well-being domains than simply valuing character strengths. However, we also note that our study found that valuing all six of the well-being domains predicted the subsequent experience of character strengths, which was measured by seven survey items, including an item that is similarly worded to the item used in this previous study (“I am willing to face difficulties in order to do what is right”). It is possible that valuing these other well-being domains may affect subsequent well-being in a variety of domains through the pathway of improved character strengths. For example, if an individual values physical health, that person might develop a disciplined commitment to exercise (e.g., running outside even when the weather is cold), which might then affect several domains of well-being (e.g., emotional or physical health) [50]. In this example, the character trait of discipline is shaped by valuing physical health. We are not able to test this hypothesis with our data, but future empirical studies could incorporate a research design that would permit the investigation of this possibility.

4.1. Implications

Our findings also suggest that learning to place more value on some domains of well-being, such as character strengths or social connections, might lead to the greatest increases in subsequent self-reported well-being. This provides guidance on targeted interventions. For example, if resources are limited, interventions might be designed to encourage increased valuation of one or both of these domains. Effects could then be assessed in terms of improved well-being across the full range of domains, ideally in a randomized controlled trial. Future research could also seek to disentangle the reasons why valuations are related to subsequently reported well-being. For example, people might value the domains that they are skillful at achieving or experiencing more (such as in [22] for health and happiness), leading to increases in the experience of these domains, and in turn leading to higher valuations—a virtuous cycle. The failure to achieve a high level of a specific domain might be associated with devaluing that domain, or perhaps the opposite, as an individual might more strongly desire what has been difficult to obtain. Developmental stages in life might also shape these patterns. Longitudinal data with more than two waves would be required to explore how these associations might unfold across the life course.

Findings from this research can be also informative for social policy. High valuations of character strengths proved to be pivotal for improved meaning and purpose, and social connectedness. They were also instrumental in building character strengths, which might be indicative of a possible virtuous circle in this domain. The role of character strengths could be strengthened in the educational process with emphasis on such character strengths as moral compass, self-regulation, perseverance, and zest. The other domain with possible scope for policy actions is meaning and purpose. People striving for meaningful and purposeful life not only tend to lead such lives more often but are also able to build their character strengths. The importance of meaningful and purposeful life can be conveyed in the educational process by showing possible alleys for personal growth and presenting their well-recognized impacts for longevity [51,52] and lower risk of cardiovascular diseases [53,54].

4.2. Limitations

There are several limitations to our study. First, our self-reported data may be subject to unmeasured confounding or social desirability bias [55]. Further research might be conducted with other types of data (e.g., objective markers of health and well-being, such as health care records and clinical diagnoses). Second, we relied on single items for our measures of the importance of domains, and we hope that our findings encourage others to strengthen measurement in this area to further advance research along these lines. Third, our sample was drawn from a single workplace, whereas a randomized sample drawn from a broader community would expand the potential for generalizability. Fourth, our baseline response rate was low, and the rate of attrition at follow-up was around 50%. It is worth noting, however, that although the response rate and follow-up rate were suboptimal, participants retained in the analytic sample were similar to the employee population with respect to most major sociodemographic characteristics. Fifth, a longitudinal analysis with three waves of data would be preferable to two waves, as the prior values of outcomes could then be measured prior to the independent variables (i.e., at the pre-baseline wave rather than at the baseline wave). Adjusting for well-being levels measured simultaneously at baseline with independent variables may partially block some of the effects that the importance of the domains might have on subsequent well-being [25].

5. Conclusions

This research showed that the importance of character strengths and the importance of physical health, followed by the importance of social connectedness, were the most consistent predictors of well-being outcomes. However, we also found that none of the valuations predicted subsequent emotional health, the importance of physical health predicted only physical health, and only two valuations were associated—both inversely—with subsequent financial stability (the importance of character strengths and physical health).

Our research provided further evidence that living well appears best achieved by valuing immaterial goods, especially social connectedness and character strengths, as opposed to domains such as financial stability or physical health. This has important implications for sustainable growth, as our main findings are consistent with a “politics of being” rather than a cultural emphasis on “having” that is associated with resource-intensive economic systems rooted in environmentally degrading forms of consumerism [56]. Character strengths emerged as the only well-being domain that was predicted by the valuation of all other well-being domains. As valuing the domains of well-being is concerned with some of the most important ends of human life, evidence of associations between these valuations and the subsequent experience of the domain of character strengths suggests that character is indeed a pathway to attaining a life of flourishing.