Social Needs and Happiness: A Life Course Perspective

Social needs have been conceptualized and investigated in various ways (e.g., Deci and Ryan 2008; Lindenberg 1996, 2013; Maslow 1943; Max-Neef 1992; McClelland 1985). Although there is some overlap between these theoretical approaches on social needs (e.g., Ormel et al. 1999; Steverink et al. 2019), the possibility of substitution between different social needs and a life course perspective on such processes is rather unique to the Social Production Function (SPF) theory (Steverink and Lindenberg 2006). For example, both substitution between social needs and concrete predictions on divergent life course trajectories do not directly follow from the self-determination theory (e.g., Deci and Ryan 2008; Wehmeyer et al. 2017). SPF theory is therefore particularly useful for the purposes of our study.

SPF theory assumes that humans are motivated to improve their happiness via the fulfilment of basic physical and social needs which can be ordered hierarchically (Fig. 1): subjective well-being as the superordinate level (level 1) emerging from social and physical well-being (level 2) which can be derived via the fulfilment of three social and two physical needs (level 3) (see Ormel et al. 1999). SPF theory assumes that these three levels of subjective well-being¹ are universal thus all human beings share these needs. However, the fulfilment of the social and physical needs requires resources like money and social relations, which may differ between individuals, and vary according to time, place, and circumstances (see for comparable theories Max-Neef 1992, p. 199). Consequently, the needs themselves are universal, but the resources and restrictions people have and encounter to fulfil their needs may differ, as illustrated in Fig. 1.

Although SPF theory is about both social and physical needs, the present study has been specifically focused on the three social needs of SPF theory (see also Steverink and Lindenberg 2006; Steverink et al. 2019). Although the fulfilment of physical needs may also be expected to differ over the life course, these physical needs are less dependent on the social context and, therefore, considered to fall outside the scope of the present study. First, affection is the sense that people care about you, regardless of your status and behavior. The need for affection can be fulfilled by relationships that provide, among others, love, trust, and understanding (cf. Maslow 1943). This need is comparable to the need for relatedness (Deci and Ryan 2008), need for intimacy (McClelland 1985), and need to belong (attachment; Baumeister and Leary 1995). Second, behavioral confirmation is the sense that you do the right things according to the people in your social environment. Relationships that provide approval, appreciation and encouragement fulfill the need for behavioral confirmation, which consists of, among others, doing good, being useful, contributing to a common goal, and being part of a group. The need for behavioral confirmation is comparable to the need to belong (feeling accepted by the group; Anant 1967). Third, status is the sense that you achieve more than others, and are known for your achievements, skills or assets. Status needs can be fulfilled by relationships that provide respect, praising, and influence (cf. Maslow 1943) and compare to the need for competence (Deci and Ryan 2008) or need for achievement (McClelland 1985). Taken together, according to SPF theory, the fulfilment of social needs is dependent on the reaction from others on who a person is (affection), what a person does (behavioral confirmation) and what a person has (status) (see Steverink and Lindenberg 2006).

Importantly, SPF theory further states that the resources used to fulfil these social needs can be substituted. For instance, after a love relationship breaks down, someone can shift focus to family relations to fulfil the need for affection. Moreover, SPF theory indicates that people might even be able to substitute one social need for another when the latter is easier for them to fulfil. For example, when someone loses her job and thereby loses status, this person might start to focus more on fulfilling the needs of affection and behavioral confirmation instead, in order to maintain similar levels of happiness. The increased fulfilment of these other needs might yield a relatively stable level of happiness, despite the loss of status. For example, affection was relatively more important for the happiness of people with a lack of status than for their high status peers (Nieboer and Lindenberg 2002). In line with SPF theory and substitution processes, we therefore expect that people can have different strategies and resources to fulfil social needs, which may result in different profiles of need fulfilment.

To examine differences in social need fulfilment between age groups, we distinguish between six age groups in the Dutch population characterized by specific goals and life events (Table 1). Generally, the adult life course is divided into three main developmental periods: young adulthood, middle adulthood, and late adulthood (Sigelman and Rider 2018). In young adulthood we distinguish between early and young adults (Arnett 2000), in middle adulthood, between middle and late adults (Lachman 2004) and in late adulthood we distinguish between young and oldest-old (Wink and James 2007). Although cutoffs are always arbitrary and there is no consensus on these categorizations in the literature, the age groups we used do represent commonly distinguished developmental stages in adulthood (e.g., Ebner et al. 2006). Based on the ‘typical’ life course in the Netherlands, we can expect differences in need fulfilment between these six age groups. Combined with the variation in social contexts along the life-course and availability of resources such as money, friends and time for family we expect different social need fulfilment profiles, as outlined below.

This paper was aimed to examine whether levels of specific social need fulfilment and their association with happiness differ between age groups. According to SPF theory, the fulfillment of different social needs along the life course can be substituted to derive the same level of happiness. Substitution implies that disharmonious need fulfilment profiles might exist, such as low affection with high confirmation and medium status, and that both the prevalence and yields of these profiles may differ across the life course. For instance, middle adulthood is characterized by a career peak and therefore profiles with high status fulfilment may be expected in this age group. Our first aim was to identify such disharmonious social need fulfilment profiles, which we expected to differ between age groups. Subsequently, we compared average levels of social need fulfilment between different age groups, as changing social contexts were expected to influence the available resources to fulfil the social needs. Third, we aimed to examine whether the relation between the fulfilment of specific social needs and happiness differs between age groups. A difference in the importance of social needs for happiness between age groups would support the tenets of socioemotional selectivity theory, whereas stable associations between social needs and happiness would support SPF theory.

This study included 11,487 adult respondents (age 18⁺) of the ongoing Dutch crowdsourcing study HowNutsAreTheDutch (HND; Dutch: HoeGekIsNL), who reported on the fulfilment of their social needs.² HND allows respondents to investigate their mental health dimensions and compare themselves to others by visiting the internet platform hoegekis.​nl. All procedural details are provided elsewhere (Van der Krieke et al. 2016). Respondents first completed a sociodemographic questionnaire (n = 14,489), after which they completed one or more topical modules (e.g., living situation, mood, or well-being). Respondents were in majority women (67%), on average 45 years old (SD = 15), and higher educated (76%). Sampling biases were intended to be accounted for by weighting our results to match demographic strata in the general Dutch population.

The HND study protocol was assessed by the Medical Ethical Committee of the University Medical Centre Groningen. The committee judged the protocol to be exempted from review by the Medical Research Involving Human Subjects Act (in Dutch: WMO) because it concerned a non-randomized open study targeted at anonymous volunteers in the general public (registration number M13.147422 and M14.160855).

All analyses were performed in R (R Core Team 2018), unless specified otherwise. R scripts are provided at https://​osf.​io/​njwuy/​. Descriptive statistics of the sample and age groups were derived first, including group means and their standard deviation (SD), and standardized mean difference d between the age groups, heuristically interpreted as small from .20 to .40, medium from .41 to .79, and large after .80 (Borenstein 2009). The association between the social needs was examined using correlations and scatterplots. We classified correlations (r) as very weak if between .00 and .29, weak between .30 and .49, moderate between .50 and .69, strong between .70 and .89, and very strong from .90 onwards (Mukaka 2012). In this study, only estimates with alpha levels below .001 will be interpreted to avoid type I errors.

Distinct social need fulfilment profiles were explored with Latent Class Analyses (LCA) in LatentGOLD (Vermunt and Magidson 2005). LCA is particularly useful for cluster analyses with a small number of items and a relatively large sample size (Eid et al. 2003). As only 16 respondents fell in the oldest-old group, this group was excluded from all cluster analyses. We fitted various types of LCA models, each with 1 up to 10 clusters. For each model, the number of clusters was identified by selecting the solution with the lowest BIC value. First, to examine cluster solutions for the general population, the three ordinal social need variables were clustered in the total sample and the weighted sample. Second, the same cluster analyses were performed for the five age groups separately. Third, three multi-group LCA’s were performed to examine the differences between age groups in more detail.

The first multi-group LCA has an equal number of clusters across all age groups, but the model parameters may differ across the age groups. This analysis gives insight into differences in cluster solutions between the different age groups. The second multi-group LCA is constrained insofar that the association between the latent classes and the social need variables is equal for all age groups, while the cluster sizes may differ across the age groups. In this way, the interpretation of the clusters is equal across the age groups, and we could examine if certain clusters are only present in specific age groups. The last multi-group LCA was most constrained because the equality constraints of the second model were applied extended with the constraint that the cluster sizes are equal across age groups. For this model, we took the same number of clusters as we had identified in the second model. A deviance test was applied to test whether the cluster sizes were equal across age groups. The multi-group LCA’s were performed using both the weighted and unweighted sample.

The differences in social need fulfilment between the age groups were examined using descriptive statistics, ANOVA analyses and pairwise t-tests. To examine whether comparisons of average social need fulfilment levels between age groups are valid, we fit measurement invariance tests (Supplement 5), which supported configural, metric, scalar, and strict invariance (CFI > .96, RMSEA < .06; Putnick and Bornstein 2016), in line with previous reports in five independent samples (Steverink et al. 2019).

Finally, age group differences in the associations between the social needs and happiness were examined in a series of regression analyses. As the relative benefit of a social need for happiness might depend on the level of need fulfilment, it is important to control in these analyses for the average levels of need fulfilment in each age group. For example, based on social emotional selectivity theory, we expect that status is stronger associated with happiness in younger adults compared to older adults. Simultaneously, younger adults are expected to have lower levels of status fulfilment than older adults as they still have to develop skills and a career. To ensure that our interaction effects do not reflect diminishing marginal returns (the higher the need fulfilment, the lower the relative benefit), our regression models are adjusted with quadratic terms of the social needs (see Nieboer and Lindenberg 2002).

In total, four regression analyses were performed. First, the association between social need fulfilment and happiness is examined (model 1), adjusted for gender, education level, and diminishing marginal returns (model 2). Second, the association between social need fulfilment and happiness is compared between age groups via their interaction terms (age group*social need, see model 3), which was also estimated adjusted for gender, education level and diminishing marginal returns (in model 4). All regression models were repeated using sample weights to derive more generalizable conclusions.

In general, conclusions from weighted and unweighted analyses were very similar. Therefore, we mainly discuss the results from the weighted analyses. Only when the results differ, we will also discuss the results from the unweighted analyses.

To explore profiles of social need fulfilment, disharmonious combinations were examined using scatterplots and cluster analyses. Respondents seldom reported low levels of affection combined with high levels of behavioral confirmation or vice versa (Fig. 2), and no respondent experienced high levels of status and low levels of affection or confirmation. A few respondents reported low levels of status and high levels of affection or behavioral confirmation. Overall, patterns of disharmonious social need fulfilment were not found.

LCA clustering of the social needs in the total sample in general did not reveal clusters with disharmonious need fulfilment profiles. The 6-cluster model fitted the data best, given the lowest BIC value using the total sample (Supplement 4.2). People in different clusters scored generally lower or higher on all needs simultaneously, which indicates that there were no clusters of respondents with disharmonious need fulfilment profiles. These results were similar when all age groups were studied separately, which still showed clusters with lower or higher need fulfilment in general, although the number of optimal clusters differed between 4 and 6 (Supplement 4.6). Thus, in these analyses the clusters differed only in their level of the fulfillment of all needs (low to high), and not in terms of different fulfilment levels of the different needs (one low, one high). Multi-group LCA clustering of the social needs resulted mainly in clusters of generally low and high profiles of need fulfilment as well. The first multi-group LCA (equal number of clusters, with possibly different model parameters across age groups) revealed that the 4-cluster model fitted best. These 4-cluster solutions were found to be very similar for all age groups, with again only generally lower or higher scores on all social need fulfilments. The second multi-group LCA (with equal associations between latent classes and variables, with possibly varying cluster sizes across age groups) revealed that the 6-cluster model provided the best fit. Comparing these models with and without equality constraints on the cluster sizes across the age groups (using the deviance test), revealed that the models with varying cluster sizes showed a significantly better fit to the data (χ² = 142; df = 20, p < 0.0001). The models with varying cluster sizes are therefore interpreted. Next to low and high cluster profiles, we also found clusters with relatively higher levels of affection need fulfilment (see Fig. 3). These clusters were most common for emerging and young adults.

To examine age group differences in average social need fulfilment levels, ANOVA analyses were performed. The analyses showed significant, but small age group differences in affection (F_(5,11309) = 12.60, p < .001), behavioral confirmation (F_(5,11305) = 7.67, p < .001), and status (F_(5,11304) = 4.45, p < .001). The maximum difference between age groups was 0.48 (on a 9-points scale; see Supplement 2 for details). All effect sizes for differences in social need fulfilment between age groups were below d = 0.21, which is typically regarded to indicate a small effect (Borenstein 2009). These results thus indicate that need fulfilment levels are highly similar between age groups (Fig. 4). Affection fulfillment levels were highest in young adulthood, and slightly higher than behavioral confirmation fulfillment levels. In middle and late adulthood, the fulfilment of behavioral confirmation was slightly higher than affection. Status fulfilment levels were slightly higher in older age groups. In sum, some age differences were observed, but the effect sizes were all very small in magnitude.

In the unweighted analyses, social need fulfilment scores were slightly higher and showed peaks for affection (about age 30) and behavioral confirmation (about age 60) (Fig. 4b). These peaks seemed to reflect the higher proportion of women in the unweighted analyses. Post-hoc tests indeed showed significant differences between women and men for affection (two-sample t_(7613.10) = 8.28, p < .001) and status (two-sample t_(7212.70) = − 16.96, p < .001), but similar levels of behavioral confirmation (two-sample t_(7115.40) = − 1,08, p < .28). Women reported 0.28 points higher average levels of affection fulfilment (d = 0.15) but 0.57 points lower average levels of status fulfilment than men (d = − 0.31, see Fig. 4c).

Figure 5 shows that the variance in need fulfilment looked similar between all age groups. Although the sample scores were highly heterogeneous, in all age groups about half of the people scored between 5 and 7 for fulfilment of affection and behavioral confirmation, and between 3 and 5 for status fulfilment. Taken together, our results suggest that the difference in need-fulfilment between age-groups were marginal at best, and that throughout the entire life course, levels of affection and confirmation fulfilment were higher than status fulfilment levels.

Age group differences in the relationship between social needs fulfilment and happiness were examined by using regression analyses. The regression models are presented in Table 2. The fulfilment of each social need was associated with happiness and combined they explained up to 31% of the total variance in happiness (see Model 1). The effect sizes of social needs were predominantly independent from gender and educational level (see Model 2), even though lower education was associated with less happiness. Controls for diminishing marginal returns reached significance (Model 2), which means that at a higher level of social need fulfilment, the relative effect of additional need fulfillment on happiness decreased. However, the effect size of these diminishing marginal returns was of little importance in the effect of social needs on happiness. The older age groups were on average happier than the younger age groups (Model 3).

The interaction effects did not show a clear picture (Model 4, Fig. 6). With all 15 interaction terms included (3 needs*5 age groups), the model explained only 1 percentage point more variance in happiness. Thus, the combined effect of all interactions was very small. The separate significant interaction effects were small as well, and lost statistical significance in the unweighted analyses (see Supplement 6). Overall, we did not find consistent support for any of our hypotheses on age group differences in the relation between need fulfilment and happiness.

Substitution does seem to happen in some people, but only in one direction, as some respondents reported low status fulfillment but high affection and behavioral confirmation. This observation is in line with SPF theory which predicts that status is more difficult to obtain than affection and therefore status is more likely to be substituted for affection than vice versa (e.g., Steverink 2001).

Overall, few differences in the fulfilment of social needs were observed between age groups, contrary to our expectations. One possible explanation is that when people report on the fulfilment of social needs, they use their age group as frame of reference when making comparisons (Pérez-Asenjo 2011). For example, young adults might feel that, compared to other people of the same age, they have relatively high levels of status, while they might have answered differently if they would have thought of a comparison with older adults (cf. Suls et al. 2002).

The observation that the fulfilment of status needs did not decline after the retirement age does not align with previous research (e.g., Steverink and Lindenberg 2006). This stability of status after retirement might reflect the overrepresentation of higher educated people in our sample, despite our weighted analyses. Possibly, people with high education maintain their status (e.g., as a doctor or professor), while lower educated people lose their status (e.g., as a construction worker or nurse). In line with this idea, a previous study of retiring people showed little difference in happiness (operationalized as life satisfaction) in higher educated retirees but a decline after retirement in retirees with lower education (Wetzel et al. 2016). Future studies could explore this status effect in more detail, and examine the frame-of-reference hypothesis.

Although we did not find major differences in social need fulfilment between age groups, we did find differences between men and women. The peaks for affection (about age 30) and behavioral confirmation (about age 60) disappeared in the weighted estimates, seemingly due to the high proportion of women in the unweighted analyses, with diluted effects in men. Women reported slightly higher levels of affection than men, but significantly lower levels of status, which has previously been observed by Steverink et al. (2019). Future studies might expand upon our results and investigate whether gender differences also influence the ways in which people fulfil their needs.

The small age differences in need fulfilment over the life course point to individual rather than age-related effects on social need fulfilment. Individuals are likely to differ considerably in the resources they have available to fulfill needs, which can be corroborated by the large range in need fulfilment within age groups. Importantly, the observed stability in need fulfilment levels across the life course could indicate that people primarily differ in the ways in which they fulfill their needs, leaving overall need fulfilment levels unchanged. Consequently, people first compensate for losses in affection by seeking alternative means to fulfil this same need (in line with SPF theory) and only after having lost many resources, people start to substitute between needs, such as compensating status loss by increasing the production of affection. It may be that our sample, despite age differences, is too homogeneous in health and other resources to observe consistent need substitution.

To summarize, based on our models, it is likely that age groups do not differ substantively in their average need fulfilment, but only in the ways they derive the same degree of need fulfilment, which we could not test. Future studies can examine how different people fulfil their needs for affection, behavioral confirmation, and status in more detail. Ideally, these studies also sample different populations, as the Dutch are among the happiest populations on earth (Helliwell et al. 2019).

Our third aim was to explore whether and how age groups differed in the relative importance of the fulfilment of the three different social needs for their happiness. Our key observation is that the fulfillment of all three social needs seems to be equally important for happiness across the adult life course. Some significant interactions were found for affection and behavioral confirmation (∆R² < .01), but these effects were only found in the weighted analysis and are potentially due to the truncated values for the very high weights. Indeed, when examining the same question with merged age-groups (young [age 18–40], middle [age 40–65], and old [age 65⁺] adults), no significant interactions occurred (Supplement 7).

SPF theory asserts that all three social needs are universal, and thus produce well-being, regardless of age. This was supported by our regression models. These observations are not entirely in line with socioemotional selectivity theory insofar that older adults are assumed to derive more well-being from close relations (i.e., affection) than younger adults, while the latter are assumed to get more well-being from investing in long term goals (including status). However, according to both SPF and socioemotional selectivity theory, the strongest effects may be expected at the highest ages, when resources seriously decline (SPF) and time horizon shortens (SST). Our pool of respondents, however, comprised only sixteen people over 80 years old. Perhaps the effects of the shifting importance of specific needs on happiness may not have been visible in the present sample, because the decline of fulfilment of social needs is better observed in the oldest-old. Furthermore, we cannot rule out that some people answered the questions on need fulfilment relative to their peers (e.g., “My status is higher than the status of my peers”), while others may have kept an ideal situation in mind, regardless of age (e.g., “My status is lower than I wish it were”). Future studies should explore possible changes in social need fulfilment in the oldest-old and frame-of-reference effects in more detail.

As we compared people from different age groups, cohort effects might have played a role in these results. Our society, social roles, and resources changed significantly for different generations over the past century (e.g., Mitchell 2006; Pilcher 1994). Our measures of measurement invariance, however, indicated that people from different age groups answered the items of the scales on social need fulfilment in a similar way, and this has recently also been confirmed in five independent samples with various age groups (Steverink et al. 2019). Therefore, we can assume that the age groups interpreted the questions about need fulfilment similarly. Although developmental questions in principle require a within-individual longitudinal design, such long-span birth-cohort studies are scarce. Moreover, changes in social and economic circumstances at the moment of measurement renders conclusions about young adults obsolete once this sample became old, which also impedes inter-generational comparisons. Although we did establish measurement invariance, some cohort effects cannot be ruled out. Due to the cross-sectional design, reversed causality (or third factors) might also play a role, as happier people might be more able to fulfil their social needs.

Furthermore, the sample used here is not completely representative for the Dutch population. Even though our selected sample showed a wide range of scores and we weighted our results based on population strata (see method section), selection effects may have played a role. For instance, only people willing to spend their free time and who could find the website of hoegekis.​nl filled out the online questionnaires. About 2.5 million people in the Netherlands (15%) lack the reading and writing skills to participate (Algemene Rekenkamer 2016). Still, such selection effects may attenuate the estimated effect sizes at worst, but not alter the underlying processes and our main conclusions. Replicating these results in different and less advantaged samples could strengthen the claim that age groups are strikingly similar in need fulfilment.

Finally, happiness can be measured in various ways (Linton et al. 2016). Although the single-question happiness scale has been well-validated (Abdel-Khalek 2006), different results could be obtained with different measurements of happiness. For example, status fulfilment previously showed stronger associations with positive affect, and behavioral confirmation with negative affect, while affection is more associated with life satisfaction (e.g., Nieboer et al. 2005; Steverink and Lindenberg 2006; Steverink et al. 2019). Different happiness measures might also lead to different outcomes for different countries, as for example the fulfilment of social needs is higher correlated with positive emotions than life evaluation in Northern Europe, but not in Africa (Tay and Diener 2011). Using a different measure of happiness might thus result in slightly different results.