The Value of a Purposeful Life: Sense of Purpose Predicts Greater Income and Net Worth

Abstract

Having a sense of purpose in life appears valuable across life domains, though it remains unclear whether purpose also provides financial value to individuals. The current study examined sense of purpose as a predictor of concurrent and longitudinal income and net worth levels, using two waves of the MIDUS sample of adults (N = 4660 across both assessments). Participants who reported a higher sense of purpose had higher levels of household income and net worth initially, and were more likely to increase on these financial outcomes over the nine years between assessments. Interaction tests suggested some evidence of age moderation, but gender did not appear to moderate the influence of purpose on economic outcomes.

Having a purpose in life entails having a clear long-term direction toward which to strive, that organizes one’s behaviors and sense of self (McKnight & Kashdan, 2009; Ryff, 1989). The value of having a purpose in life has been recognized for centuries as a variable integral to positive health and well-being. Adults with a greater sense of purpose tend to report better emotional well-being (e.g., Zika & Chamberlain, 1992) and physical health (Scheier et al., 2006), experience less risk for cognitive decline later in life (Boyle, Buchman, Barnes, & Bennett, 2010), and even enjoy greater longevity (Hill & Turiano, 2014). However, it remains unclear whether purposeful individuals fare better with respect to economic outcomes; in other words, what is the literal value of having a purpose in life?

Research has demonstrated that individual dispositions can predict individual-level economic outcomes such as personal net worth and income (Judge, Livingston, & Hurst, 2012; Author Citation, under review). For instance, conscientiousness, a proclivity toward being organized and industrious, tends to predict greater financial success concurrently and in the future (Judge et al., 2012; Author Citation, under review). Central to this argument is the notion that our dispositional characteristics influence how we make daily and long-term decisions in ways that either facilitate or hinder our ability to accrue wealth.

Sense of purpose may be one such characteristic, given its potential to influence both building assets and reducing liabilities. For one, if purposeful individuals tend to be physically and psychologically healthier, they will incur fewer health care costs and miss work less frequently. Purposeful individuals also may be more focused on their occupational objectives, as one study suggested that individuals who rated occupational goals as more important during the transition to adulthood reported a greater sense of purpose in adulthood (Hill, Jackson, Roberts, Brandenberger, & Lapsley, 2011). If so, purposeful individuals may strive toward occupational success, which would likely increase personal income.

The possibility that purposeful individuals benefit financially over the long haul is consistent with prevailing theoretical perspectives. McKnight and Kashdan (2009) suggested that purposeful individuals differ in their resource allocation, as they focus on facilitating their efforts to achieve long-term aims. Studies show that purpose correlates positively with more expansive future time perspectives (Rappaport, Fossler, Bross, & Gilden, 1993; Hicks, Trent, Davis, & King, 2012) and with a greater sense that their time is being used effectively to fulfill downstream goals (Bond & Feather, 1998). As such, purposeful people may be more likely to save money or make investments that support downstream goals, and not squander resources based on impulsive decisions. However, research is needed with respect to whether purpose prospectively promotes personal wealth, as well as whether these associations are specific to purpose.

The current study examined whether sense of purpose predicts greater financial success, using the MIDUS longitudinal sample of adults (Brim, Ryff, & Kessler, 2004). We considered both household income and net worth as financial outcomes of interest. When predicting net worth (total assets minus debts), we controlled for household income levels to test the unique association purpose held on net worth, which includes incoming income. To identify the unique value of purpose, we examined whether it proved a significant predictor of net worth above and beyond the role of the Big Five personality traits, as well as general well-being (measured as life satisfaction). Furthermore, we examined age and gender as moderators of the purpose associations with economic outcomes. These interactions were tested both with cross-sectional data, as well as for the prospective associations.

Method

Participants

Data came from participants of the MIDUS study (Brim et al., 2004), which is a sample of 7,108 participants first recruited in 1995–96 and followed up in 2004–06. At the first wave (MIDUS 1), participants were residents of the United States, aged 25 to 74 (M = 46, SD = 13). The average longitudinal follow-up interval was around 9 years (range: 7.8 to 10.4 years). At MIDUS 2, 4,963 were successfully contacted to participate in another phone interview of about 30 minutes in length (75% total response rate – adjusting for the 8% too ill to be interviewed or deceased; see Radler & Ryff, 2010). Attrition analysis indicated that participants in the longitudinal sample reported higher purpose in life t(6289) = 8.14 (d = .21), higher life satisfaction t(7067) = 4.53 (d = .11), higher household income t(6108) = 7.75 (d = .20), higher net worth t(5672) = 6.05 (d = .16), higher education t(7093) = 15.31 (d = .36), higher conscientiousness t(6268) = 6.17 (d = .16), lower neuroticism t(6262) = −2.43 (d = −.06), lower extraversion t(6269) = −2.75 (d = −.07), and lower agreeableness, t(6269) = −4.61 (d = −.12), all p’s < .01. In addition, longitudinal respondents were more likely to be white χ²(6176) = 87.15, and married χ²(7103) = 61.99, but less likely to be retired χ²(7058) = 22.79, all p’s < .01. There was substantially missing data on economic variables likely because participants refused to provide this information or they did not know. The sample sizes for each individual analysis are provided in Tables 1 and 2. Even with such reductions, as evident in previous work with the MIDUS (Judge et al., 2012; Author Citation, under review), the current sample size is sufficient for detecting significant results when predicting economic variables in multiple regression frameworks. The sex distribution of MIDUS participants was generally balanced (53% female), and most participants were white (approximately 93%). Over two-thirds of participants had more than a high school education, and most MIDUS participants were married at MIDUS 1 (70%).

Table 1.

Regression analyses predicting concurrent income and net worth at MIDUS 1. The middle column reflects initial analyses and the right column reflects testing for moderation by age.

Predictor	Initial			Testing Moderation by Age
	B (s.e.)	p	95% CI	B (s.e.)	p	95% CI
Income as Outcome (n = 5880)
Purpose in Life	2578 (795)	.001	1020 to 4136	1729 (1056)	.102	−340 to 3797
Age	1792 (897)	.029	32 to 3551	1769 (898)	.049	8 to 3530
Gender	7873 (1507)	<.001	4918 to 10828	7816 (1509)	<.001	4858 to 10775
Education	16305 (759)	<.001	14816 to 17794	16314 (760)	<.001	14825 to 17803
Race (0 – White; 1 – Other)	−8669 (2467)	<.001	−13506 to −3832	−8809 (2470)	<.001	−13650 to −3968
Marital Status	−35347 (1590)	<.001	−38464 to −32231	−35353 (1590)	<.001	−38470 to −32236
Retirement Status	−28749 (2453)	<.001	−33557 to −23940	−28823 (2460)	<.001	−33645 to −24000
Life Satisfaction	4278 (793)	<.001	2722 to 5834	4282 (795)	<.001	2724 to 5840
Extraversion	4116 (924)	<.001	2304 to 5928	4077 (925)	<.001	2263 to 5890
Agreeableness	−5020 (885)	<.001	−6754 to −3286	−5010 (885)	<.001	−6745 to −3275
Neuroticism	1254 (767)	.102	−249 to 2758	1246 (767)	.105	−258 to 2750
Conscientiousness	2819 (785)	<.001	1281 to 4357	2784 (785)	<.001	1245 to 4323
Openness	750 (880)	.394	−976 to 2476	786 (881)	.372	−940 to 2513
Age by Purpose	---			−246 (719)	.732	−1656 to 1164
Gender by Purpose	---			1871 (1435)	.192	−942 to 4684
Model R2	.23			.23
Net Worth as Outcome (n = 5472)
Purpose in Life	14680 (2789)	<.001	9252 to 20107	6318 (3684)	.086	−904 to 13539
Age	64617 (3120)	<.001	58503 to 70732	63433 (3103)	<.001	57349 to 69517
Gender	9993 (5217)	.056	−234 to 20219	8126 (5193)	.118	−2054 to 18307
Education	22678 (2727)	<.001	17332 to 28025	22887 (2711)	<.001	17572 to 28201
Race (0 – White; 1 – Other)	−35872 (8564)	<.001	−52661 to −3558	−36238 (8520)	<.001	−52940 to −19536
Marital Status	−14797 (5733)	.010	−26037 to −3558	−14281 (5698)	.012	−25453 to −3110
Retirement Status	53537 (8713)	<.001	36456 to 70618	58483 (8683)	<.001	41460 to 75506
Income	68862 (2757)	<.001	63457 to 74267	68874 (2741)	<.001	63501 to 74248
Life Satisfaction	69 (2802)	.980	−3585 to 8337	1148 (2788)	.681	−4318 to 6613
Extraversion	3646 (3197)	.254	−2620 to 9913	4347 (3180)	.172	−1887 to 10581
Agreeableness	−11614 (3197)	<.001	−17621 to −5608	−12172 (3046)	<.001	−18143 to −6201
Neuroticism	−1869 (2656)	.482	−7076 to 3337	−2121 (2640)	.422	−7296 to 3054
Conscientiousness	6695 (2720)	.014	1363 to 12028	6662 (2705)	.014	1360 to 11964
Openness	2376 (3040)	.435	−3585 to 8337	2717 (3024)	.369	−3212 to 8645
Age by Purpose	---			19991 (2504)	<.001	15082 to 24900
Gender by Purpose	---			10768 (4957)	.030	1050 to 20485
Model R2	.28			.29

Table 2.

Longitudinal regression analyses predicting income net worth at MIDUS 2, controlling for initial levels. The middle column reflects initial analyses and the right column reflects testing for moderation by age.

Predictor	Initial			Testing Moderation by Age
	B (s.e.)	p	95% CI	B (s.e.)	p	95% CI
Income as Outcome (n = 3594)
Purpose in Life	4461 (977)	<.001	2546 to 6376	5535 (1283)	<.001	3019 to 8051
Age	−14276 (1084)	<.001	−16401 to −12151	−13865 (1087)	<.001	−15997 to −11733
Gender	5609 (1801)	<.001	2079 to 9140	5957 (1810)	.001	2409 to 9506
Education	12982 (935)	<.001	11149 to 14815	12911 (934)	<.001	11080 to 14742
Race (0 – White; 1 – Other)	−7507 (3495)	.032	−14361 to −654	−7715 (3491)	.027	−14560 to −870
Marital Status	−7254 (2043)	<.001	−11260 to −3248	−7315 (2040)	<.001	−11315 to −3315
Retirement Status	−4155 (2976)	.163	−9991 to 1680	−5002 (2981)	.093	−10847 to 842
Life Satisfaction	2580 (990)	.010	639 to 4520	2388 (989)	.016	449 to 4328
Extraversion	2117 (1092)	.053	−24 to 4259	1947 (1091)	.075	−193 to 4087
Agreeableness	−2127 (1039)	.041	−4163 to −90	−2003 (1038)	.054	−4037 to 31
Neuroticism	1343 (917)	.143	−455 to 3141	1347 (915)	.141	−447 to 3142
Conscientiousness	16 (962)	.986	−1869 to 1902	−52 (961)	.957	−1934 to 1833
Openness	−482 (1059)	.649	−2558 to 1593	−571 (1058)	.590	−2645 to 1504
Age by Purpose	---	---		−3476 (937)	<.001	−5314 to −1639
Gender by Purpose	---	---		−796 (1781)	.655	−4287 to 2696
Model R2	.33			.33
Net Worth as Outcome (n = 2889)
Purpose in Life	20857 (4974)	<.001	11103 to 30610	14715 (6710)	.028	1558 to 27872
Age	22487 (4974)	<.001	10813 to 33652	22571 (5881)	<.001	11040 to 34102
Gender	18239 (9036)	.044	522 to 35956	16556 (9124)	.070	−1334 to 34445
Education	32765 (4754)	<.001	27945 to 46586	37493 (4757)	<.001	28165 to 46821
Race (0 – White; 1 – Other)	−56016 (18074)	.002	−91456 to −20577	−56683 (18083)	.002	−92140 to −21225
Marital Status	7237 (10262)	.481	−12885 to 27359	7190 (10262)	.484	−12932 to 27313
Retirement Status	−41661 (15766)	.008	−72576 to −10746	−42178 (15824)	.008	−73207 to −11150
Income	55384 (4982)	<.001	45617 to 65152	55324 (4736)	<.001	45552 to 65095
Life Satisfaction	11242 (5075)	.027	1291 to 21193	11509 (5086)	.024	1537 to 21481
Extraversion	−134 (5452)	.980	−10824 to 10557	−110 (5456)	.984	−10809 to 10588
Agreeableness	−18012 (5181)	<.001	−28172 to −7853	−18130 (5184)	<.001	−28295 to −7966
Neuroticism	4008 (4572)	.381	−4956 to 12973	4000 (4573)	.382	−4966 to 12966
Conscientiousness	7272 (4867)	.135	−2272 to 16816	6962 (4872)	.153	−2591 to 16516
Openness	−850 (5292)	.872	−11227 to 9526	−494 (5301)	.926	−10888 to 9900
Age by Purpose	---			−1254 (4985)	.801	−11028 to 8520
Gender by Purpose	---			12736 (9080)	.161	−5067 to 30540
Model R2	.50			.50

Note: Models included initial levels on the financial outcomes as predictors, and r² values reflect the inclusion of this predictor.

Measures

Sense of purpose

Participants completed a three-item reduced version of the purpose in life subscale from the psychological well-being scales (Ryff, 1989). Participants reported on a scale from 1 (Strongly Disagree) to 7 (Strongly Agree) to the items: “Some people wander aimlessly through life, but I am not one of them”; “I live life one day at a time and don’t really think about the future” (reversed); and “I sometimes feel as if I’ve done all there is to do in life” (rev.) (M = 5.50; SD = 1.21; range = 1–7; α = .36). Though the reliability was not ideal, this measure has demonstrated predictive validity in previous work with the MIDUS sample for important outcomes such as mortality risk (Hill & Turiano, 2014).¹

Personality traits

Participants completed short measures of the Big Five traits based on Goldberg’s (1992) markers (Lachman & Bertrand, 2001). Respondents were asked whether 25 adjectives described themselves from 1 (not at all) to 4 (a lot). The Big Five traits were assessed as follows: neuroticism (moody, worrying, nervous, calm (rev.), α = .74); extraversion (outgoing, friendly, lively, active, talkative, α = .76); openness (creative, imaginative, intelligent, curious, broad-minded, sophisticated, adventurous, α = .77); conscientiousness (organized, responsible, hardworking, careless (rev.), α = .58); agreeableness (helpful, warm, caring, softhearted, sympathetic, α = .80).

Life satisfaction

Life satisfaction was employed as our measure of general subjective well-being. Life satisfaction was assessed by asking participants, “How satisfied with life are you now?” on a scale from 1 (Not at all) to 4 (A lot) (M = 3.53, SD = 0.69).²

Household income

At both measurements, household income was computed (in dollars) from several different questions. Total household income included both the respondent and their spouse/partners income from wages over the past month, and over the past 12 months the total household social security, government assistance, and other sources of income. Participants responded to these questions using a rating scale that reflected different ranges of monetary values, and MIDUS used these responses to generate household income variables. Any respondent reporting over $300,000 was capped at 300,000 to reduce the effect of outliers. The MIDUS data files do not contain information on which percentage of respondents were top-truncated to $300,000. At MIDUS 1, mean household income was $71,700 (SD = $61,282), and the median was $55,000. At MIDUS 2, mean household income was $71,363 (SD = $60,463), and the median was $57,500.

Net worth

At both measurements, participants reported the value (in dollars) for six types of assets (stocks/bonds, savings/checking accounts, retirement funds, homes/other real estate, vehicles, businesses/farms), as well as for their debts and liabilities (loans, mortgages, credit card debt) using similar rating scales as for income. Net worth was calculated by subtracting debts from assets, by the participants themselves. Negative net worth values were reset to zero before data release, due to privacy and human subjects concerns. This bottom-truncation occurred for 13% of participants at MIDUS 1 and 11% at MIDUS 2. Additionally, at MIDUS 1, if a participant had an estimated net worth higher than $1 million, the value was set to that amount, which affected only 2% of the sample. At MIDUS 1, mean net worth was $120,720 (SD = $209,397), and the median was $32,500. At MIDUS 2, in 1995 dollars, mean net worth was $280,623 (SD = $614,337), and the median was $137,700.

Analytic Plan

Regression analyses were conducted using SAS statistical software (SAS Institute Inc., 2002–2004). First, using MIDUS 1 data, we predicted initial income and net worth values from purpose in life scores, other psychosocial predictors (the Big Five traits and life satisfaction) and demographic variables (age, gender, education, race / ethnicity, marital status, and retirement status). Second, the same regression analyses were employed for predicting income and net worth at MIDUS 2, controlling for initial levels in order to predict residualized change over time. Third, age and gender were examined as moderators of the purpose associations with concurrent and prospective income and net worth, including all other previous predictors in the models. To simplify interpretations, age, education, income, purpose, personality traits, and life satisfaction were entered as standardized variables. Given the large sample size and the number of analyses, we employed a more restrictive alpha threshold of .01 for discussion of significance, though exact p-values are reported in the tables.

Results

First, we examined the associations of purpose in life on concurrent household income and net worth levels. Table 1 presents the findings on this front. Sense of purpose had a significant, unique positive association with both financial outcomes. The regression coefficient for purpose in life can be interpreted with respect to a one standard deviation increase in purpose corresponding to $2578 in greater income, and $14,680 in greater net worth (accounting for the influence of income), controlling for the other variables in the model.

Second, we investigated the associations between purpose in life and prospective household income and net worth levels. Table 2 presents the regression analyses toward this end. Sense of purpose uniquely predicted prospective levels of income and net worth at MIDUS 2. The regression coefficient for purpose in life can be interpreted as a one standard deviation increase in purpose being associated with a unique increase of $4461 in income and $20,857 in net worth over time, even controlling for the other variables.

Third, we considered whether the associations between purpose on household income and net worth were moderated by age or gender. The rightmost column of Tables 1 and 2 present the analyses including both age-by-purpose and gender-by-purpose interaction terms in the regression models. Age was a significant moderator for initial net worth and the longitudinal analysis for household income (see Supplementary Figure 1); gender failed to reach significance across models. We utilized the Johnson–Neyman technique, a procedure allowing researchers to identify at which levels of the moderator the associations occur, to explore the associations between purpose and economic outcomes at different ages. With respect to initial net worth, for adults aged 25–33 (−0.95 SD on age and below), there was a weak negative association between purpose and net worth such that younger individuals with higher purpose scores had lower net worth. Between ages 34–42 (−0.35 to −0.94 SD on age), the association of purpose on net worth was non-significant. For adults older than 42 though, the association was always positive and increased in strength with age. The direction of age moderation was reversed though for longitudinal analyses with household income. Specifically, the association between purpose and prospective household income was statistically significant and positive for adults who started the study between the ages of 20–35, but was not significant for those who started the study older than 35.

Discussion

Research has consistently demonstrated the benefits of having a purpose in life for healthy development and aging (see Ryff, 2014 for a review). The current study extended this work into the economic domain by examining the role of purpose on household income and net worth concurrently and prospectively. Above and beyond known predictors of financial success (e.g., demographics, personality traits), as well as conceptual correlates of purpose (well-being), individuals who reported a higher sense of purpose in life tended both to have higher household income and net worth initially, as well as greater increases on these outcomes over the following decade. These analyses also rule out several alternative explanations, such as the role of Big Five personality traits or well-being.

Instead, the picture painted is that having a purpose in life may hold real economic consequences. Having a purpose appears integrally linked to greater agency and engagement in daily life (Scheier et al., 2006), as well as a number of positive health outcomes (Boyle et al., 2009; Hill & Turiano, 2014; Scheier et al., 2006). As such, the value of purpose for financial success may result from the greater capability and propensity for purposeful individuals to pursue their long-term goals, which in turn promotes the accrual of assets. If so, one might anticipate a relatively long time course for the financial benefits of a sense of purpose, which is reflected in the age interactions found for initial net worth and longitudinal income levels. Initially, the association between net worth and purpose was stronger for older adults, presumably because these individuals had experienced more years during which to accrue the benefits of purpose. However, the association between purpose and household income was stronger longitudinally for younger adults, suggesting that a sense of purpose benefited them more over time, compared to older adults. These findings suggest that purposefulness may be most beneficial during the transition between young and middle adulthood, during which individuals are more likely to attain greater occupational success. That said, as this interaction effect was not consistent across all models in Tables 1 and 2, future research is needed to better understand the specific factors (and timing of those factors) involved in why purposeful individuals appear better at financial accrual. Little support though was found that the benefits of purpose differ between males and females (though see Supplementary Table 1).

The current work though is not without its limitations, which should provide avenues for future research. First, economic data in MIDUS was assessed via self-report, and thus findings should be supplemented with studies that include objective markers of financial status. In addition, work is needed to examine whether these associations replicate during alternative economic climates. Second, for the sake of increasing participant privacy, MIDUS participants’ data were bottom- and top-truncated, and participants were asked only to report into which income ranges they fell instead of specific values. As such, more continuous data would be preferable. However, these truncations did not affect a large portion of the sample, and if anything, likely served to attenuate the effect sizes found in the current work. Third, MIDUS participants only were asked to report on their sense of purpose, and did not provide information on the content of their purposes. In particular, it would be of interest to see whether financially-oriented individuals may reap greater benefits to their net worth than those more prosocially- or artistically-oriented. Fourth, it is essential to replicate these findings in other large-scale population samples, to ensure these findings were not unique to MIDUS. Moreover, it is important to replicate these findings with longer measures of sense of purpose, to deal with the issues of reliability, though the consistent cross-sectional findings across measurement occasions suggest the replicability of the results (see Supplementary Table 1). These caveats aside, the current findings provide evidence that even when it comes to finances, finding a purpose in life appears to be well worth it.

Highlights.

• Having a sense of purpose appears adaptive for health and well-being.

• We examined its association with economic success as well in MIDUS study.

• Sense of purpose positively associated with initial household income and net worth.

• Sense of purpose prospectively predicted gains on these outcomes over time.