Patrick L. Hill1, Daniel K. Mroczek2 and Robin K. Young1, 1Carleton University, Ottawa, ON, Canada, 2Northwestern University, Evanston, IL, USA
Research on subjective well-being often focuses on identifying mean-level trends, and less on the individual differences that might influence patterns of change across the life span. The current chapter focuses on this latter topic with respect to examining personality traits as predictors and moderators of well-being change. Personality traits are known correlates of well-being across multiple indicators, and traits by definition include affective components. Our literature review touches on each of these points to underscore why traits appear likely candidates for influencing well-being trajectories, with a focus on generating hypotheses regarding why traits could play a role in influencing well-being changes. In so doing, we promote a view that research should consider both broad (e.g., Big Five) and more specific traits as moderators, as well as to incorporate experimental paradigms into the investigation. Ultimately, the goal is to move beyond investigating whether well-being changes over time and instead consider the question of “for whom?”
personality; Big Five; personality development; well-being changes; gratitude; forgivingness; mindfulness; life span development
When considering who may become happier or more satisfied over time, laypeople and researchers alike often focus on candidates such as social role changes (including getting married and getting a job), personal wealth, and even physical health and well-being. A less obvious candidate, though, may be one’s personality. Although we do not typically think about how traits and dispositions may predict fluctuations in well-being, their potential relevance becomes more apparent with the realization that they tend to predict all the potential catalysts for well-being change discussed above. Indeed, personality traits predict occupational attainment, divorce, income levels, as well as physical health and mortality risk (e.g., Hampson, 2012; Ozer & Benet-Martinez, 2006; Roberts, Kuncel, Shiner, Caspi, & Goldberg, 2007). Therefore, if any of these variables increases or decreases levels of well-being, personality traits may be an underlying catalyst for such changes.
The value of discussing personality traits in the context of changes in well-being is supported by considering four literatures, which serve as the four sections for review in the current chapter. First, personality traits by definition (e.g., Roberts, 2009; Watson, 2000) include affective components, some more so than others, such as extraversion and neuroticism. Therefore, we discuss the evidence for longitudinal changes on these two traits, as these fluctuations provide potential insights into whether subjective well-being also may be capable of changes. Second, we briefly review the wealth of evidence suggesting that personality traits do predict levels of subjective well-being, at least cross-sectionally, demonstrating that certain personality profiles appear more likely to promote well-being. Third, we consider the potential for each of the Big Five traits to moderate trajectories of well-being, noting the extant empirical evidence when relevant. Finally, we discuss the relatively recent work suggesting that “manipulating” personality could lead to changes in well-being, focusing on three specific traits: gratitude, forgivingness, and mindfulness. Our intent is to demonstrate that considering personality traits as moderators of well-being change is an important focus for future research, even though it is one that has been largely neglected to this point. Therefore, we have structured our chapter to guide this work, by producing testable hypotheses for future work using Big Five traits and more specific dispositional characteristics.
We begin by adding another perspective for considering the central question of this volume: namely that evidence for personality change may itself constitute evidence that happiness and well-being can change. Toward this point, we focus solely on the broad dispositions of extraversion and neuroticism. When we consider the lower-order facets that comprise these traits, their relevance becomes immediately clear. Neuroticism has been defined as higher levels of affective traits such as anger, anxiety, and depression (NEO; Costa & McCrae, 1995). Alternatively, extraversion is described with respect to greater positive emotions, warmth, and activity. Although neuroticism and extraversion clearly are not reducible to negative and positive affect, understanding whether these dispositions and their facets change over time provides insight into the broader question of whether well-being itself can change.
Studies have consistently demonstrated that extraversion and neuroticism have the capacity to change across the life course. With regard to mean-level trends, neuroticism tends to decline during young and middle adulthood (Roberts, Walton, & Viechtbauer, 2006), a pattern also mirrored by several of its facets (Soto, John, Gosling, & Potter, 2011). Extraversion often shows a more differentiated age trajectory, depending on the facet of interest (Roberts et al., 2006). However, studies consistently demonstrate that individuals vary in their patterns of change on these traits (e.g., Mroczek & Spiro, 2003; Small, Hertzog, Hultsch, & Dixon, 2003), underscoring the potential for both to fluctuate across time. Perhaps more importantly, these changes do, in fact, matter for aging and life span development. For instance, increases in neuroticism predict greater mortality risk, even when controlling for initial levels (Mroczek & Spiro, 2007).
Similar to the work on well-being, studies have examined the role of life events on the stability of extraversion, neuroticism, and other traits over time (Neyer & Lehnart, 2007). For instance, one study investigated the role of 12 different major life events (e.g., marriage, divorce, unemployment, retirement, etc.) and found some evidence that they may influence trait changes (Specht, Egloff, & Schmukle, 2011). However, on the whole, relatively few events significantly influenced changes on neuroticism and extraversion, or their patterns of rank-order stability. Such evidence provides another parallel to the well-being literature, which similarly has reported that major life events have less of a long-term influence on well-being than we may expect (see, e.g., Diener & Lucas, 1999). Next, we consider the relationship between personality and well-being more thoroughly, by reviewing the literature linking trait levels to reports of life satisfaction and well-being.
Our subheading alludes to perhaps the most substantial and well-known examination of personality as a predictor of well-being, namely, the classic meta-analysis conducted by DeNeve and Cooper (1998). Their thorough review of the literature pointed to three particularly important findings. First, all Big Five personality traits were linked to markers of life satisfaction, happiness, positive affect, and negative affect, in the expected ways. Extraverted, agreeable, conscientious, emotionally stable, and open individuals tended to fare better than their peers with respect to reporting greater well-being, with the exception that openness to experience was modestly related to negative affect. Second, although extraversion or neuroticism was the strongest predictor for each well-being outcome, the other three traits were not much weaker in their predictive value. Third, overall, these personality traits were shown to evidence, at best, moderate relationships with the well-being indicators of interest. Indeed, no meta-analytic estimate was greater in magnitude than .27, the effect of extraversion on happiness. As an effect size comparison, this is comparable to the correlation between gender and weight among adults, which is .26 (Meyer et al., 2001), but lower than the association between height and weight, which is .44. However, the extraversion-happiness association is roughly double the size of the association between ibuprofen and pain relief (.14) or college grades on job performance (.16) (Meyer et al., 2001). From this perspective, we may conclude that the effect of traits on happiness is substantial. It also may be the case that individual changes in happiness track, to some extent, individual changes in traits.
However, more recent meta-analytic work (Steel, Schmidt, & Shultz, 2008) suggests that the estimates reported by DeNeve and Cooper (1998) may have underestimated the relationships between personality and well-being, particularly for extraversion and neuroticism. In some cases, these relationships were closer to .4 or .5 in magnitude, although it differed greatly based on the personality measure. If these estimates could be considered more accurate, we might say that the personality-happiness relationship is comparable in size to the association between height and weight—in other words, an effect of considerable magnitude. In addition, this meta-analysis demonstrated the potential for specific traits to evidence stronger relationships with well-being than the broader Big Five composites. For instance, anxiety and depression were stronger (negative) correlates with life satisfaction and positive affect than impulsivity, another facet of neuroticism. Similarly, activity and positive emotions proved stronger correlates of positive affect than other extraversion facets, such as gregariousness and excitement seeking.
While the facet results were based on relatively few studies, they point to the fact that the happy personality lies beyond merely the Big Five traits. For instance, dispositional gratitude has demonstrated consistently moderate-to-strong relations with well-being (see Wood, Froh, & Geraghty, 2010, for a review), even when considering observer-reported personality (McCullough, Emmons, & Tsang, 2002). Although gratitude may appear an obvious correlate, given its natural connection to positive affect in general, this trait provides an interesting discussion point for at least two reasons. First, its influence on subjective well-being appears particularly strong, given that it holds across the adult years (Hill & Allemand, 2011), as well as when controlling for the NEO facets (Wood, Joseph, & Maltby, 2008). Second, the findings for gratitude present an interesting argument for how personality could influence well-being. Namely, when studying how and why well-being changes, researchers may wish to focus less on nominating specific life events that could alter these trajectories, and instead turn to understanding the individual differences that influence our interpretation of these events.
Along this front, another disposition worthy of consideration is forgivingness, defined as a dispositional tendency to forgive across different transgressions and transgressors (Roberts, 1995). As such, forgiving individuals differ in their interpretation and reaction to social stressors and exchanges (e.g., Berry, Worthington, Parrott, O’Connor, & Wade, 2005; Burrow & Hill, 2012). Forgivingness has been consistently connected to subjective well-being, potentially because it may foster relationship success and adaptive self-development (see, for a review, Hill, Heffernan, & Allemand, in press). While it is clearly connected to Big Five traits such as agreeableness and emotional stability (Mullet, Neto, & Rivière, 2005), research has demonstrated that forgivingness cannot be fully characterized by these broader traits (Steiner, Allemand, & McCullough, 2012). In sum, the happy personality appears to comprise both the ability to be thankful for the positive influences that others may bring, as well as to excuse them for the negative ones.
A final trait worth discussing is mindfulness. Mindful individuals are those better capable of observing and describing their internal and external environments, being less judgmental and impulsively reactive to a given event, and acting with awareness of the current and future consequences (Baer, Smith, Hopkins, Krietemeyer, & Toney, 2006; Brown & Ryan, 2003). Therefore, mindfulness reflects a trait similar to gratitude and forgivingness, in its potential for influencing one’s daily reactions and social exchanges, which in turn can have cumulative influences on well-being over time. As such, this trait has been nominated by several religious traditions as a catalyst for greater personal well-being (Brown & Ryan, 2003; Kabat-Zinn, 2003), and empirical research has consistently supported this suggestion. Indeed, mindful individuals tend to report greater happiness and positive affect, less negative emotions, and more satisfaction with life (e.g., Brown & Ryan, 2003; Schutte & Malouff, 2011; Weinstein, Brown, & Ryan, 2009). Moreover, a key goal of mindfulness-based stress reduction (MBSR) programs (Kabat-Zinn, 1998; Speca, Carlson, Goodey, & Angen, 2000) is to lessen both the perception and emotional reactivity to stressors as well as increase positive emotion. Indeed, techniques such as MBSR have been linked to improved immune function (Davidson et al., 2003; Witek-Janusek et al., 2008) as well as increased activation in brain regions that promote positive affect (Davidson et al., 2003). Indeed, people with lessened reactivity to stressors have lower mortality risk (Mroczek, Stawski, Turiano, Chan, Almeida, Neupert, & Spiro, 2013). Like gratitude and forgivingness, these effects appear to hold even when controlling for the broader traits like extraversion and neuroticism (Brown & Ryan, 2003; Mroczek et al., in press). Therefore, the happy personality is best described with respect to a collection of both broader and more specific traits.
Given the theoretical, conceptual, and empirical linkages between personality traits and well-being, it appears likely that our dispositions may influence our tendency to become more or less happy over time. Therefore, it is surprising that relatively little research has formally tested whether personality traits can moderate well-being trajectories, although the extant literature does suggest a potential role for personality. To organize this review, and chart the course for future research, we consider the potential for each of the Big Five traits to moderate well-being trajectories, starting again with the two most affective dispositions.
Given their links to positive and negative affectivity, most extant research has focused on the potential roles for extraversion and neuroticism. For instance, one study examined age and personality traits as predictors of positive and negative affect across a large adult sample (Mroczek & Kolarz, 1998). While cross-sectional in nature, this work provided several important insights regarding the potential for well-being to change over time, as well as predictors of these fluctuations. First, across participants, positive affect tended to increase with age, whereas negative affect tended to decline. However, these trends differed for males and females, insofar that age only seemed to predict negative affect for males. Second, the influence of age on positive affect differed by extraversion level for males. Specifically, highly extraverted males tended to always have high levels of positive affect, regardless of age, whereas age had a stronger influence on development trajectories for highly introverted males. In other words, the aging process appears to only influence positive affect for those men who were not already very extraverted in nature.
Charles, Reynolds, and Gatz (2001) went even further and estimated long-term (23 years) longitudinal positive and negative affect trajectories in more than 2,800 participants in the Longitudinal Study of Generations (LSOG). They used growth modeling to capture intra-individual trajectories of PA and NA. They found that PA trajectories were stable over time for young and midlife adults but declined for older adults. More relevant for the issue of personality moderation of well-being trajectories, higher extraversion was associated with greater stability in PA trajectories among all age cohorts. Consistent with the finding of decreasing NA with age (e.g., Watson & Walker, 1996), Charles et al. found decreasing trajectories of NA for all age cohorts. With respect to personality moderation, higher neuroticism was associated with less decline in NA for all age groups. In essence, Charles et al. (2001) documented that higher extraversion promotes maintenance or stability of PA, and that higher neuroticism attenuates the natural decline in NA seen in most adults. This was the first study using long-term longitudinal data to highlight the crucial role of personality in predicting well-being trajectories.
Building upon Charles et al. (2001), Griffin, Mroczek, and Spiro (2006) estimated decade-long trajectories of PA and NA in more than 1,500 mostly older adults (mean age 69). They found that PA declined, and replicated the finding that NA decreased with age (but increased slightly after age 70). Like Charles et al., they found a critical role of personality as a moderator of NA trajectories, with higher neuroticism associated with fewer declines in NA during midlife and steeper increases in older adulthood. Extraversion was associated with higher levels (intercepts) of PA but not with rate of change.
However, PA and NA are not the only aspects of happiness that personality can modify. Life satisfaction, a more “cognitive” or evaluative indicator of well-being, appears to change systematically over the life course, as documented by Mroczek and Spiro (2005) and Blanchflower and Oswald (2008). The latter found evidence for a U-shaped curve of well-being, with a nadir at around age 40 to 45. Mroczek and Spiro (2005) described the opposite: an inverted U function. However, they used a much older sample: their youngest participants were age 40 to 45. Placing the Blanchflower and Oswald (2008) curve side by side with the Mroczek and Spiro (2005) curve yields a function that has two nadirs. One nadir is at age 40–45 and the other at a much older age, at around 85–90. With respect to personality moderation of these curves, Mroczek and Spiro found that extraversion modified all three aspects of well-being trajectories: level, slope and curvature. In other words, persons higher in extraversion had higher life satisfaction trajectories that were more stable and less peaked than persons who were more introverted.
The aforementioned work presents one of the ways by which personality traits can contour changes in well-being, which is similar to either a “ceiling” or “floor” effect. Namely, individuals initially higher on those dispositions characteristic of the happy personality should be less likely to gain on well-being with time or show less pronounced natural declines in negative indicators of happiness (e.g., NA). Similar examples have been shown by research investigating changes in social well-being over a 9-year span of adulthood (Hill, Turiano, Mroczek, & Roberts, 2012). In that study, initially extraverted adults were less likely to gain on social integration, one aspect of social well-being (Keyes, 1998), and were no more likely to gain on the other facets of social well-being. As would be expected, extraverted individuals were much more likely to report higher social well-being at the start of the study, making it less likely for them to increase on these facets with time.
An alternative moderation effect can occur for individuals initially low on “happy traits.” For instance, that same study (Hill et al., 2012) found that initially neurotic individuals were more likely to experience gains in social well-being across three of its four facets. In this respect, personality traits also can act as “floor” effects in their influence on well-being change. Both of these moderation effects reflect ways by which personality traits can identify those individuals most and least likely to become happier with time. However, describing and identifying moderation effects for personality traits becomes more complicated when moving beyond the two Big Five traits most characterized by well-being. Indeed, extraversion and neuroticism accounted for the vast majority of the personality moderation effects shown in that study of social well-being change.
When we consider the other Big Five traits, deeper consideration is needed regarding why these dispositions could serve as moderators of well-being trajectories. Although personality traits do not appear to moderate the effect of life events on well-being (Yap, Anusic, & Lucas, 2012), these traits could still serve moderating roles by virtue of motivating the events that influence well-being. For instance, one potential route for personality is through predicting which individuals are more likely to attain those accomplishments that might coincide with greater well-being, such as marital or job success. Or conversely, personality traits may predict an attenuated likelihood to experience the events that can dampen well-being, such as poor health or unemployment.
In both cases, conscientiousness should prove valuable to maintaining or promoting adaptive patterns of well-being over time. Conscientious individuals are likely to achieve higher levels of education (Poropat, 2009), job success (Barrick & Mount, 1991; Judge, Higgins, Thoresen, & Barrick, 2006), and in turn income (e.g., Sutin, Costa, Miech, & Eaton, 2009). In addition, conscientiousness predicts a diminished likelihood of divorce, unemployment, and poor health (Roberts et al., 2007). Given the strong stability of well-being over time (Diener & Lucas, 1999), it is worth noting that none of these effects in and of themselves is likely to change well-being trajectories much. However, it is the cumulative benefit of being conscientious that makes it a likely candidate as a well-being moderator, likely acting more as a stabilizer that allows individuals to immunize themselves against the risks for ill-being that befall others (see also Hill, Nickel, & Roberts, 2013). This largely remains a topic for future inquiry, though, because little research thus far has considered the role of conscientiousness as a well-being moderator.
With respect to openness to experience, it is more difficult to make clear and consistent predictions regarding its influence on those events associated with well-being changes. Open individuals may live longer lives than their peers, but this effect appears dependent on which facet is examined (Jonassaint et al., 2007; Turiano, Spiro, & Mroczek, 2012). Mixed results also present with respect to the influence of openness on health (Goodwin & Friedman, 2006) and relationships (e.g., Noftle & Shaver, 2006), with results differing based on the specific outcome of interest. Accordingly, one would likely anticipate the “mixed bag” of results regarding the cross-sectional relations between this trait and well-being.
As such, openness to experience may be less consistent in the direction that it influences well-being change. Instead, one could predict this trait to hold an “opposite” role to conscientiousness. Namely, openness to experience may influence well-being trajectories less in their direction and more by virtue of creating greater fluctuations and variability. Similar to conscientiousness, though, this destabilizing effect may be difficult to identify in typical moderation tests. Therefore, future research needs to examine personality traits as predictors not only of change patterns, but also of well-being stability over time.
Agreeableness also has shown both positive and negative influences on physical health (e.g., Goodwin & Friedman, 2006) and appears to negatively predict income levels (Judge, Livingston, & Hurst, 2012). Moreover, it is less clear whether agreeableness, unlike openness to experience, would play a role on the variance around one’s well-being trajectory. One potential reason for this lack of clarity is that the facets of agreeableness are differentially related to life satisfaction (e.g., Wood et al., 2008), which may obscure relationships at the domain level. Indeed, similar claims could be made for most of these larger composite traits. Accordingly, we turn our attention next to how more specific traits can inform us with respect to well-being moderators.
The study of specific traits is valuable for three primary reasons. First, one often can make clearer predictions regarding the influence of specific traits on well-being, or the potential catalysts thereof, than when discussing the Big Five domains. Second, while initial research focused on the broader Big Five traits, empirical work has accrued to demonstrate consistent relationships between specific traits and well-being (e.g., Wood et al., 2008). Moreover, researchers have suggested the capacity for lower-order, specific trait facets to change over time (e.g., Hill, Payne, Jackson, Stine-Morrow, & Roberts, 2013; Jackson et al., 2009; Soto et al., 2011). Third, building from this work, researchers have demonstrated some efficacy in intervening upon these specific traits, and the outcomes of which are promising for researchers trying to influence well-being. Toward this end, we return to the three case examples we briefly described earlier: gratitude, forgivingness, and mindfulness.
The construct of gratitude provides one of the most frequent candidates for intervention in the field of positive psychology. Such interventions often entail either asking participants to describe something for which they are thankful on each day or sending longer “thank you notes” to individuals who have provided them frequent assistance in their lives. These interventions have shown some potential to increase levels of gratitude in both adolescent and adult samples (see, e.g., Lambert & Veldorale-Brogan, 2013; Wood et al., 2010), although there are still methodological concerns left to be addressed (Froh, Kashdan, Ozimkowski, & Miller, 2009). Perhaps more important to our aims, these interventions often find that participants report greater life satisfaction or positive affect by the end of the study (e.g., Emmons & McCullough, 2003). In other words, it appears that motivating individuals toward greater gratitude may influence changes in well-being. As such, this specific trait may prove an intriguing candidate for influencing long-term trajectories in well-being, because grateful individuals are likely to experience more positive changes over time, as they are oriented toward focusing on the “good” in the world (Wood et al., 2010). In other words, grateful individuals may be more susceptible to increases in well-being over time because they are better at identifying those events and circumstances that have improved their lives.
Another interesting candidate is forgivingness, which reflects a trait that not only changes as we age (Allemand & Steiner, 2012), but also may be susceptible to interventions. Indeed, a wealth of literature has demonstrated that samples ranging from youth to older adults can become more willing to forgive others following interventions (for a review, see, e.g., Baskin & Enright, 2005; Wade, Worthington, & Meyer, 2005). Moreover, although the literature on forgivingness development is brief, researchers have nominated a number of pathways by which these single sessions of forgiveness are likely to enact broader dispositional changes (Hill, Heffernan, & Allemand, in press). Again, important to our aims, these forgiveness interventions appear also to motivate changes in well-being, particularly with respect to attenuating negative affect (e.g., Allemand, Steiner, & Hill, 2013; Lin, Mack, Enright, Krahn, & Baskin, 2004). As such, forgiving individuals may be less likely to experience upward shifts on this component of well-being over time. Forgivingness may serve as a trait that buffers one against daily increases in hostility or revenge seeking, which in turn has a cumulative effect with respect to influencing longer-term changes in well-being.
A final candidate trait is mindfulness, which also has served as the target for a wide array of intervention studies (Brown, Ryan, & Creswell, 2007; Grossman, Niemann, Schmidt, & Walach, 2004; Shapiro, Carlso, Astin, & Freedman, 2006), mostly given the large interest it has received from the clinical psychology community. These studies consistently demonstrate that mindfulness training can have important effects on well-being over the course of the study (e.g., Grossman et al., 2004). Mindfulness may be a particularly interesting candidate moderator, given its influence on both positive and negative affect. Being aware of the current and future consequences of one’s actions could increase positive affect over time, by allowing individuals to recognize and capitalize those experiences that can produce happiness. In addition, increased awareness could reduce negative affect over time, by virtue of responding to daily stressors more flexibly, and in turn reducing allostatic load and the known health issues that it presents (McEwen & Seeman, 1999).
The evidence is mixed on whether well-being is stable or changing over the life span. Some studies have indicated stability (Diener & Lucas, 1999), whereas others have documented long-term changes at both the population and individual levels (Blanchflower & Oswald, 2008; Charles et al., 2001; Mroczek & Spiro, 2005). Regardless of the overall trends, personality traits present as excellent candidates for potentially influencing these trajectories over time. First, personality traits include affective components, particularly with respect to extraversion and emotional stability, and thus share conceptual connections with well-being. Second, personality traits, both broad and specific, have demonstrated consistent linkages with happiness and well-being. Third, research has begun to demonstrate long-term effects on this front, with personality traits predicting age trends and longitudinal changes on well-being (Hill et al., 2012; Mroczek & Kolarz, 1998). Fourth, intervening on personality, either in its state or trait form, appears to induce changes in well-being. As such, we encourage future researchers to consider the points made here and begin to examine the potential for traits to moderate changes in well-being, which would point again to the vast predictive power of personality (Roberts et al., 2007).
That said, any well-being changes are likely to be quite modest in magnitude, and may differ for males and females (Mroczek & Kolarz, 1998), which further exacerbates the known power issues with studying moderator effects (e.g., Aguinis, Beaty, Boik, & Pierce, 2005). In addition, as noted earlier, personality traits may moderate not only the direction of well-being changes, but also the variability one has around his or her set point. Therefore, researchers must consider multiple approaches and perspectives regarding what “moderation” means. However, given that other aspects of personality, such as life goals, appear to predict well-being changes (e.g., Headey, 2008; Hill, Burrow, Brandenberger, Lapsley, & Quaranto, 2010), we are encouraged by the ultimate prospects for research with personality traits.