- Original authors / creators
- Martin Korndörfer; Boris Egloff; Stefan C. Schmukle
- Original publication
- PLoS ONE · PLOS · 2015-07-20
- License & retained notices
- CC BY 4.0
© 2015 Korndörfer et al
Article-specific license evidence ↗Exact license notice
https://creativecommons.org/licenses/by/4.0/This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Article-specific XML license URL: http://creativecommons.org/licenses/by/4.0/ - Edition changes
- Narrative text edition with normalized whitespace. Figures, tables, display equations and separately licensed material are omitted; consult the preserved XML and original publisher for definitive content.
- Evidence limits
- Adjacent research, not a directly sampled family-office study. Applicability to a family office requires separate assessment. We believe that this research embodies some potentially advantageous characteristics: First, we examined large and representative samples (1,421 < N < 37,136), whereas most previous studies have based their analyses on multiple small samples that sometimes even consisted entirely of students. This reduced not only the statistical power of each single study but also the power of the combination of these studies. That is, when combining multiple studies in one article, the persuasiveness and robustness seem to increase, but actually, the power diminishes as more statistical tests are executed [48–53,56–58]. This problem has recently been discussed extensively in the literature and has led to an increase in caution against the results of these small-sample multistudy articles [51] that are best avoided [48]. By contrast, our large samples allowed for an analysis of the proposed negative effect with extremely high statistical power. Second, regarding the measurement of social class, we used a composite measure that comprised the three objective socioeconomic indicators of income, education, and occupational prestige in our analyses (we were restricted to income and educational attainment only in Study 2 and for the year 2012 in the studies using the GSS; see [17,69–71]). By using the same state-of-the-art composite measure in all of our analyses, we minimized the possible dominating effect of any single indicator. Furthermore, we complemented our main analyses by also using a measure of subjective social class. Third, we were cautious when we choose and handled the covariates. We always used the same covariates—age and gender—across all studies (because Studies 1 and 2 analyzed donation behavior at the household level, age and gender were not included as covariates in these analyses), whereas previous research often used covariates (age, ethnicity, religiosity, and different combinations of them) that varied inconsistently across studies. Furthermore, following Simmons et al.’s [58] advice, we additionally reported the results of all analyses without the inclusion of covariates to provide evidence for the robustness of the results across different analytical methods. Despite these potential positive features of our research, there are clearly some limitations that should be noted. First, because we relied on representative international survey panels, seven of our eight studies were based on self-reports. In the studies using the ISSP and the GSS, respondents retrospectively reported how often they donated and volunteered for charity and how often they had engaged in some prosocial everyday behaviors in the last year. In the study using the CEX, participants documented their income and expenses (also those for charity) in a “Diary Survey”—and we were therefore able to at least minimize self-serving hindsight bias [112]. Yet, all these self-reports are necessarily subject to social desirability and self-presentation biases [113,114], and these biases might even be enhanced among higher social class individuals [115–117]. Importantly, these potential limitations do not apply to the results of our Study 8: Fortunately and rather uncommon for a large panel, the SOEP also assesses directly observed prosocial behavior: the allocation of points in an economic game. Specifically, participants played a variant of the “trust game” [101] across 3 consecutive years, and we simultaneously measured trust and trustworthiness [100,102]. Second, we were able to use only a limited selection of prosocial behaviors. Prosociality is a broad concept that covers various actions that are aimed at benefitting others [59–61]. In the present studies, we assessed four very different behaviors—from the allocation of money or money-like points to face-to-face helping in an organized or everyday setting. Future research, however, should expand the focus to other forms of prosociality (e.g., to observable helping behaviors in field experiments using non-student populations). A third limitation lies in the unclear personal motivation for prosociality. That is, individuals might act generously not just because they want to help others. It is also possible that they act benevolently because they expect some sort of direct or indirect reciprocity that may help them in the future [118–121] or because of a desire for “prosocial” prestige [34,122,123]. Moreover, the (expected) prosocial behavior of others may have established a social norm that prompted individuals to behave philanthropically when they would not have done so in private [124–126]. For example, in a field experiment, Martin and Randal [127] placed a transparent donation box in an art gallery with free admission and found varying frequencies and amounts of donations depending on whether or not, how much, and what kind of money (coins vs. bills) was visible in the box. Thus, even if the motivation for the prosocial act is not relevant in terms of the key question of the present paper, future research should try to uncover the underlying motivation for prosociality and examine whether individuals from different social classes vary in their motivation for prosociality. A recent study by Stephens et al. [128] at least pointed in the direction of different motives: Students whose parents did not have a college degree (and who therefore might be from a lower social class) endorsed more interdependent motives (e.g., helping their families, giving back to their communities) for attending a university than students who had at least one parent with a college degree. In sum, although our results certainly do not fit with those presented in the psychological literature [1–3], they are in line with other research on the effects of social class on prosociality. For example, in research outside the field of psychology involving large-scale investigations in Taiwan, Canada, and the United States, individuals of the higher social realms were reported to volunteer and donate more often and to donate more to charity than those in the lower realms [8–11,37–40]. In terms of the monetary amounts of the donations relative to household income, the literature in sociology and economics furthermore reported a u-shaped relation to a household’s social class [36,40,42–45]. However, this held only for donor households. When donor and nondonor households were combined in a joint analysis, as also corroborated in our Studies 1 and 2, the artifactual u curve (or decreasing tendency as we found in Study 2) changed into a positive effect ([43,44]; for some other methodological issues that may produce the u curve, see also [36]). All in all, cumulative evidence therefore strongly points toward a positive effect rather than a negative effect of social class on prosocial behavior.
- Retraction / correction check
- No linked notice identified in retrieved XML or Crossref relation/update-to fields. Article-specific repository XML related-article links, Crossref DOI relation/update-to fields, and a bounded publisher-page notice search where accessible; unavailable pages and unregistered/unlinked notices remain coverage limits. Verified 2026-11-05. Automated/source checks are not human peer review.
- Preserved record
- Rights and full text checked 2026-11-05. Included in this volume on 2026-11-05. Original source dates remain unchanged.
Download the preserved source
SHA-256 (retrieved artifact): 7a4b8303c85174cae089c5e681c9b67b01ad804c7e276e3b1817313ddc22771b
SHA-256 (preserved text edition): 8d192c33e3625f812297d90982cd0e21072ef87ee67a9752f02c34c7424b3f91