What the Evidence Finds: Employment, Wages, and Well-Being
This chapter maps where measured social and emotional domains are associated with adult outcomes: getting and keeping work, earning wages, choosing occupations, experiencing job satisfaction, and living a healthy and engaged life. The point is not to produce a league table of traits. It is to learn how the sign and size of a relationship change when the outcome changes, so a reader can ask a better career question than “Which personality is most valuable?”
Labor-market associations are real, modest, and uneven
The OECD evidence gives the clearest starting point for employment. Across almost all participating countries, adults with higher emotional stability and extraversion were more likely to be employed than unemployed or inactive after the analysis accounted for education, literacy proficiency, and demographic characteristics. Conscientiousness was positively associated with employment in about half of the countries. Openness and agreeableness played more limited roles: openness was positive in some countries and negative in a few, while agreeableness was negative in a small set and not linked to employment in most.
The pattern is important because it is both broad and incomplete. Emotional stability and extraversion recur often enough to be more than an isolated result, but “almost all” is not “everywhere,” and a repeated association is not a universal premium. Country exceptions may reflect differences in how people express a domain, how employers interpret it, and whether education or other signals provide a more reliable way to assess readiness. A social behavior that helps in one hiring culture may be less visible or less valued in another.
The relationship with literacy adds another layer. On average across the OECD countries, a one-standard-deviation increase in literacy was associated with about a four-percentage-point higher likelihood of employment. Comparable increases in extraversion or emotional stability were associated with approximately three percentage points. The comparison does not put literacy and personality on a common moral scale. It shows that social and emotional characteristics can have associations of a similar order for access to employment while remaining distinct from cognitive skill.
The OECD analysis also found that the employment associations of extraversion, emotional stability, and conscientiousness tended to be stronger among adults with low literacy. On average, each association was about two percentage points higher for adults at the lowest literacy level than for adults at high levels. This can be read as evidence of a possible compensatory role: when an employer has less cognitive information about readiness, observable persistence, composure, or social ease may carry more weight. It should not be read as evidence that social skill replaces literacy. It may instead show that different signals matter more when one signal is weak.
Wage evidence is somewhat different. The World Bank meta-analysis pools 613 coefficients, including estimates for socio-emotional and cognitive skills, using log real monthly wages and a random-effects model. Across the Big Five-related evidence, a one-standard-deviation increase is associated with roughly a 1.4–3.1 percentage-point increase in wages. The highest pooled association is attached to lower agreeableness at 3.1%, followed by emotional stability at 2.5% and openness at 2.4%. Conscientiousness and extraversion have smaller but still meaningful pooled associations in that synthesis.
The size is easier to understand through comparison. The paper contrasts the socio-emotional estimates with commonly reported 10–15 percentage-point returns to a one-standard-deviation increase in cognitive skills. The OECD’s separate adult-skills analysis likewise finds that, after accounting for education and literacy, emotional stability, extraversion, and conscientiousness have positive wage associations of roughly 2.8%, 2.6%, and 1.9% on average. These are not the same estimates: one is a cross-country survey analysis and the other pools reported coefficients from multiple studies. Their common implication is narrower and more defensible than either headline alone: social and emotional domains can add to wage differences, but they do not explain most of them.
The structure of the World Bank synthesis matters here. The 613 coefficients are not 613 independent people observed under one common experiment. They come from 16 papers with different samples, measures, models, and settings, and the random-effects approach allows the underlying effects to vary. A pooled estimate is therefore an average across a field of studies, not a number that can be assigned to a new worker. It is useful for asking whether a pattern recurs and roughly how large it is, but less useful for predicting a particular salary without knowing the person’s occupation, country, education, bargaining position, and employer.
The OECD variance decomposition gives another way to keep the scale in view. Across countries, the five social and emotional domains together account for about 1.5% of observed wage variation on average, compared with about 4.3% for literacy and 8.8% for educational attainment. The social-and-emotional share ranges from 0.3% in Austria to 4.6% in Croatia. In some countries it is comparable to the literacy contribution, but that does not make it the dominant explanation everywhere. An explained-variance share describes what a model accounts for among observed differences; it is not the fraction of a person’s career controlled by personality.
The pooled World Bank returns also vary by setting. Average returns are higher in high-income countries than in developing countries, and the paper reports larger average estimates for women than men in its subsamples. The developing-country evidence is concentrated in Latin America, so it cannot simply be extended to every lower-income labor market. These differences may reflect labor-market structure, institutions, selection into occupations, measurement, or the kinds of studies that were available. They are a reminder that a global-sounding claim may be built from a narrower evidence base than its wording suggests.
Employment and wages should also be kept conceptually separate. Employment is an access and retention outcome; wages are shaped after entry by occupation, task assignment, productivity, negotiation, tenure, and institutional pay structures. A domain can help make competence visible during hiring without producing a large wage difference once a person is in the job. Conversely, a domain may be linked to higher-paying occupations through selection without directly raising performance in every task. Treating the employment result as a wage promise skips the transition where much of the uncertainty lives.
That “modest but meaningful” position matters. A trait association does not need to dominate wages to affect a career. Small differences can compound through access to work, task assignment, promotion, and occupation. But the reverse is also true: a positive average association can be too small, too uncertain, or too context-bound to justify changing a person’s identity or choosing a career by questionnaire score.
The sign of a relationship needs equal attention. Lower agreeableness is not simply “being difficult.” In the World Bank synthesis, the category is connected with assertiveness, independent judgment, and willingness to challenge consensus. Those behaviors may be rewarded in negotiation, leadership, or competitive promotion contests. Yet the same label cannot be turned into a general recommendation to be less cooperative. Agreeableness may support trust, care, teamwork, and other outcomes. The measured sign tells us how a composite category relates to one outcome in the included evidence; it does not settle which behavior is desirable in every role.
The evidence is therefore best read as a map with several dimensions: access to employment, wage association, relative size, direction, and population. A map can show recurring terrain without telling one traveler exactly where to go.
The broader payoff is not measured only in wages
Employment and wages describe important economic consequences, but they do not exhaust what people are trying to secure through work. The OECD analysis links social and emotional domains to job satisfaction, health, life satisfaction, volunteering, political efficacy, self-employment, and occupational profiles. These relationships make the evidence harder to summarise—and more useful if the reader refuses to reduce every outcome to salary.
Job satisfaction illustrates the difference between having a job and experiencing it well. Emotional stability and extraversion are positively related to job satisfaction across nearly all participating countries, while agreeableness, conscientiousness, and openness matter in only some contexts. The association of emotional stability and extraversion is comparable in scale to the association of literacy with job satisfaction in the reported analysis. Even after wages, employment sector, and occupation are considered, the relationships with emotional stability and extraversion persist in most countries, although they become weaker. That persistence is consistent with two possible explanations: these domains may help people regulate emotions and navigate relationships, or people whose working conditions fit them may both report greater stability and feel more satisfied.
Health and life satisfaction show why economic returns should not be treated as the sole measure of value. Emotional stability has a strong and consistent positive association with both outcomes. Across OECD countries, a one-standard-deviation increase in emotional stability corresponds to roughly a nine-percentage-point higher likelihood of reporting high life satisfaction and good health. Extraversion is also positively related to life satisfaction in all countries, with an average increase of about three percentage points, and to good health with an average increase of about five percentage points. Agreeableness and conscientiousness are positive in many countries, while openness has more variable relationships.
These results do not mean that a person can purchase health or happiness by raising a score. The outcomes and the domains are measured in the same broad adult population, and self-reported traits may share response tendencies with self-reported health or satisfaction. The findings are adjusted associations, not evidence that a short intervention will produce a particular number of percentage points. Their practical value is to show that the traits relevant to work may also be connected to how people experience life, and that the relevant trade-off is not always income versus nothing.
The civic outcomes point in a different direction again. Extraversion and openness are positively associated with volunteering in most countries, with average relationships comparable to those of literacy. Openness, agreeableness, and emotional stability are positively associated with political efficacy—the belief that people like oneself can influence political affairs—in many countries. Conscientiousness has more mixed relationships with volunteering. A domain that is rewarded in a competitive wage negotiation may not be the same domain that predicts participation in a community. That is not inconsistency; it reflects different tasks, incentives, and forms of engagement.
Occupational choice adds a bridge between economic and non-economic outcomes. Extraversion is positively related to self-employment in almost all participating countries, and openness is positively related in many. The OECD also finds different average social and emotional profiles across occupations. These profiles can arise because people choose roles that suit their tendencies, because employers select people who fit an occupational norm, or because the work itself reinforces particular behaviors. The profile describes a pattern among workers; it does not state a requirement for every successful worker.
This distinction matters for career decisions. If a reader sees that sales managers tend to score higher on extraversion, the defensible conclusion is that some sales-management environments reward or attract behaviors associated with that domain. The indefensible conclusion is that a less-extraverted person cannot learn the work or succeed through a different way of building trust. Likewise, if caring occupations attract more agreeable workers but pay differently from leadership roles, the finding may reveal how labor markets price tasks and channel people—not the inherent worth of the people who perform them.
The same evidence can therefore answer different questions depending on the reader’s goal. Someone deciding how to enter employment may ask which behaviors make competence visible. Someone choosing among occupations may ask which tasks and environments support their preferred way of working. Someone evaluating well-being may care more about stress, social support, autonomy, and life satisfaction than about a small wage association. An organisation may need to ask whether it is rewarding the capability it needs or merely reproducing a familiar style.
A table of average associations cannot make those decisions for a person because it removes the task and the outcome from view. It can show that emotional stability is repeatedly linked to employment, satisfaction, health, and life satisfaction; it cannot tell us whether the best response is training, a different role, a different manager, a better workload, or a change in how competence is assessed. Those possibilities require a mechanism. The next question is how visibility, productivity, selection, and occupational context can create the patterns that the data reveal.