Abstract: Altruism is a universal human trait, but little is known about its within-population variation. Socio-economic status (SES) has been found to positively impact altruism, but the specific socio-economic variables behind this relationship have remained elusive. This study aimed to determine which facets of SES predict altruism using a lost letter paradigm and a novel lost letter method. Six hundred letters (half dropped on the pavement, half sent to residential addresses) were distributed in 20 suburbs of Perth (Australia) differing in socio-economic variables. Letters distributed in high-SES neighbourhoods were more likely to be returned than letters distributed in low-SES neighbourhoods. Educational attainment and occupation status were the specific socio-economic variables underlying this association, while economic resources and crime rate were not associated with the likelihood of a letter being returned. These results suggest that altruism blossoms in neighbourhoods that are populated with highly educated individuals working in high-status jobs. The relationship between education and prosocial inclinations may be mediated by cognitive ability, self-control and high levels of socialization. Having experienced sustained exposure to norm-abiding models, more educated people may also be better at internalizing cultural norms of helping behaviour, thus creating a more altruistic environment where they reside.
Discussion
The current study revealed substantial and systematic variation in altruistic tendencies across urban suburbs of different socio-economic characteristics. This variance appears to be conditioned by the education and occupation level of residents in the suburb, and was not consistently influenced by economic resources or crime rate.
Socio-economic status
Both the original lost letter experiment and the novel, modified letterbox method provided support for the hypothesis that area-level SES (measured by IRSAD) was positively correlated with helping behaviour. This result is in consensus with previous research reporting a link between SES and letter return rates in the lost letter experiment (Brown and Reed 1982; Grueter et al. 2016; Holland et al. 2012; Nettle et al. 2011; Silva and Mace 2014). Multiple drivers underlying the lower levels of prosociality in low-SES areas are conceivable, for example time constraints resulting from the need to make ends meet (Holland et al. 2012; Lynam et al. 2000), lower sense of control over the environment (e.g. Gallo et al. 2005) or – more mundanely – higher tolerance of litter rates (see Khatib et al. 2007).
However, the above finding is in conflict with studies by Piff et al. (2010, 2012) who found upper-class individuals to be less prosocial and more unethical in measures such as willingness to cooperate with a game partner and attitudes on charitable donations. Côté et al. (2015) recently showed that higher-income individuals are not more selfish across the board but a tendency to be less generous emerges only under conditions of high economic inequality. Piff et al. (2012) suggested that – among others – this was because high-SES people have abundant resources to deal with the downstream costs of unethical behaviour (e.g. money for a speeding fine), while lower-SES individuals may need to be more careful as they incur greater relative consequences for social deviation. Piff et al. (2012) proposed that low-SES individuals have a greater interest in the wellbeing of others because it affects their ability to draw resources from them. Thus lower-class individuals’ willingness to engage in altruistic behaviour can be seen a function of economic interdependence. Relatedly, Amir et al. (2018) invoked an uncertainty management framework to account for the greater prosociality observed in economic games among economically deprived children. In this framework, cooperation with social partners and prosociality reflect the adaptive internalization of a risk-mitigating strategy in the face of uncertain returns associated with early life deprivation.
The difference between the results of Piff et al. (2010, 2012) and lost letter-based studies could stem from the fact that the former analysed variation at the individual level, whereas the latter examined neighbourhood-level differences (Holland et al. 2012). Perhaps high-SES neighbourhoods foster altruism, yet within any one neighbourhood, the poorer individuals are more altruistic than the wealthier ones (Holland et al. 2012).
Another reason for the difference between our findings and those of Piff et al. (2010, 2012) could be that their experiments measured altruistic tendencies towards people in general (no specific group) in a range of environments. In contrast, the lost letter experiment used in this study measured altruistic behaviours within one's own ‘home environment’ (their suburb or street) towards (presumably) members of their own group; individuals who encountered lost letters would have probably assumed that the letter was distributed by a resident when walking through the area.
Lastly, as suggested by Holland et al. (2012), the experiments used to analyse altruism by Piff et al. (2010, 2012) may be more competitive than the small, cooperative task of returning a lost letter, resulting in different behaviours. For example, upper-class individuals may be more likely than lower-class individuals to return a letter in a cooperative task, but they may also be more likely to deceive another player in a laboratory-based economic game. Future studies should incorporate multiple measures of altruistic behaviour (such as those used by Piff et al. 2010, 2012) to determine if the patterns seen in this study are unique to the lost letter experiment.
Socio-economic variables
The principal aim of this study was to disentangle the association of different socio-economic variables with altruistic behaviour. Crime was predicted to reduce altruism by lowering trust, but a suburb's crime rate was largely unrelated to the expression of prosocial behaviour. Only in the model where economic resources were included did crime rate become significant. Therefore, the variance explained by crime rate may be accounted for by other SES characteristics such as education, which was included in all other models. It may also be that crime has a threshold effect and needs to be at a certain rate before it begins to affect peoples’ altruistic tendencies. The suburbs analysed in this study may not have had sufficient crime rates to demonstrate this effect.
Economic resources, as a characteristic of SES, also did not have a significant effect on letter return rate. This suggests that demographic factors such as individuals’ assets, house prices and average household income are not related to suburb-level altruistic behaviours. Holland et al. (2012) suggested that low-SES individuals may be too preoccupied with meeting their individual needs to be willing to spend time helping others. This hypothesis suggests that individuals with more economic resources will be better equipped to meet their needs and will, therefore, have more time and energy to engage altruistically with others. The current dataset does not rule out the hypothesis that, when time itself is not a limited resource, people may be more willing to engage in prosocial behaviours. It should be noted that the location of this study does not experience widespread socio-economic deprivation where a great proportion of individuals do not have access to basic needs such as clean water, food and housing. Perhaps this hypothesis may be relevant in more economically deprived contexts where economic resources may influence altruism.
IEO was found to be significantly associated with whether a letter would be returned or not. The effect of IRSAD on letter return rate may largely be explained by the composite variables that it shares with IEO. This finding suggests that the component of SES that affects a neighbourhood's letter return rate is the education and occupation status of individuals in that suburb. To our knowledge this is a novel finding that has not been reported previously. However, along a similar vein, there is one recent study which documented a positive correlation between historical rates of primary education and civic honesty (Cohn et al. 2019). Because IEO incorporates both education and occupation variables, we cannot distinguish whether both, or just one or the other, of these variables influence altruistic behaviour within a suburb.
This study has isolated education and occupation as the likely leading socio-economic variables behind the often found relationship between SES and letter return rates. However, we still do not fully understand the mechanism behind this link. We do not know what aspects of education and occupation status may lead individuals within a suburb to behave more altruistically. Education and occupation may also be associated with a third variable that may be driving the patterns in the results. For example, individuals who have achieved a high education level or who are in high-status jobs are more likely to possess greater cognitive abilities (Schmidt and Hunter 2004; Strenze 2007). It may thus be possible that the significant effect of IEO on letter return rate reflects an underlying effect of cognitive ability. Previous studies in behavioural economics have shown a link between cognitive ability and altruistic behaviour (Jones 2008). Cognitive ability has been found to be negatively correlated with a preference for immediate rewards and impulsivity (Jensen 1998; de Wit et al. 2007). Cognition in more stressful and harsh environments associated with lower SES may be focused more on temporal discounting and lower levels of self-control (Coall et al. 2012; Frankenhuis et al. 2016; Mullainathan and Shafir 2013; Sheehy-Skeffington and Rea 2017), conditions that discourage altruistic behaviour (Osiński et al. 2017). It is important to note that extrapolating from SES at a relatively crude area-level analysis to individual differences in cognitive ability (and thus altruism) is problematic. The relationship between these factors and education is probably more complex, and dependent on many factors (e.g. opportunity, value placed on education, etc.).
An evolutionary mechanism underlying the finding that education and occupation are the primary drivers of prosocial behaviour may be that educated people have more opportunities to learn, to be taught and to receive feedback and thus are more likely to adopt or maintain cultural norms of prosociality. Individual behavioural decisions (as to whether to act proscocially) are influenced by expectations of the behaviour of others in the local social environment (Bichierri and Xiao 2009). In turn, these decisions also influence the local social environment, by conveying to others information about local norms of cooperative behaviour (cf. Schroeder et al. 2014).
Modified lost-letter experiment
The novel letterbox method incorporated in this study featured a significantly lower return rate compared with the original pavement method. Both methods, however, exhibited the same SES patterns in the data. One explanation for the differing return rates is that ignoring a letterbox letter ends all future possibilities for the letter to be returned, but pavement letters may be picked up by someone else (however, one could also argue that receiving a lost letter in someone's letterbox increases the recipient's pressure to do something about it). Another possible explanation is that returning a pavement letter may incur a smaller cost in terms of time and effort, because an individual could already be heading in the direction of a post box, compared who a letterbox recipient who would have to make a separate trip. Furthermore, individuals may behave more prosocially when encountering pavement letters because there is a chance that their actions are being observed by bystanders and influence their reputation (sensu Raihani and Bshary, 2015). Additionally, there remains the possibility that letterbox recipients may have uncertainty about what to do with the wrongly addressed letter.
The similar socio-economic patterns found in the results from both methods suggest that the letterbox method may be a useful alternative to the pavement method as it may not be as susceptible to some of the potentially confounding variables such as non-residents encountering the letters and differing rates of pedestrian foot traffic in different neighbourhoods. The letterbox method has additional advantages that should be considered for future experiments. The method enables the letter to be distributed at any time, unlike pavement letters, which must be distributed on wind- and rain-free evenings. The letterbox method also eliminates the lengthy process of distributing letters by hand and provides easy access to remote or rural areas. Additionally, the letterbox method allows for more letters to be distributed in any given area, owing to the elimination of the possibility of individuals encountering multiple letters while walking through a neighbourhood.
Cultural group selection and prosociality
Cultural group selection theory posits that groups whose members engage altruistically with each other are more successful in intergroup competition than groups whose members lack such locally stable cooperative cultural norms (Henrich 2004; Richerson et al. 2016). However, the great variation in altruistic tendencies exhibited by the different suburbs suggests that cultural group selection does not function at the scale of the city. Instead, we may see large populations splitting up into smaller sub-groups with their own set of altruistic norms which may be the result of cultural group selection operating on this smaller scale. However, since populations of city suburbs are not natural groups but administrative divisions, it is unclear if these are subject to cultural group selection. Alternatively, variation in altruism attributed to different cultural norms could in fact reflect individual adaptations to different environments with varying levels of socio-economic harshness (Mace and Silva 2016).
Posts
No comments:
Post a Comment