A person high on the neuroticism dimension of the widely accepted nonpathological Big Five personality model (e.g., Costa & McCrae, 1995; Goldberg, 1990; John & Srivastava, 1999) has characteristics that may make FV less likely. For example, such a person tends to be anxious, nervous, self-conscious, easily agitated in tense situations, and generally unable to manage stress effectively. As well, there is evidence with individuals and US states as analytic units showing that higher neuroticism is associated with lower FV levels (e.g., Claxton-Oldfield & Banzen, 2010; Handy & Cnaan, 2007; King et al., 2015; McCann, 2017).

Lower SES also is associated with lower levels of FV with individuals as the analytic units (e.g., Macchia & Whillans, 2021; Steptoe & Zaninotto, 2020; Wilson, 2000). This relation probably occurs because lower SES persons face more barriers to serving as volunteers such as limitations in income and other resources, including suitable knowledge and requisite skills (Musick & Wilson, 2008). Lower SES also has been related to lower US state FV rates (e.g., McCann, 2017, 2023).

Generalized trust refers to trust of persons external to family and friends. Where there is less trust, there is less community social interaction (Putnam, 2007) and less cooperation (Gächter et al., 2004). Where there is less opportunity for common purpose and cooperation, there is a weaker desire to participate in FV (Dinesen et al., 2020). Of course, trust and social interaction and cooperation are integral in most FV commitments. Researchers have found that generalized trust is positively associated with FV at the individual and state analytic levels (e.g., Dinesen et al., 2020; McCann, 2023; Miranti & Evans, 2019).

Nesbit et al. (2020) have noted that RED “remains an underexplored topic in volunteering studies” (p. 919). RED relates negatively to social trust (e.g., Dinesen et al., 2020; Nesbit et al., 2020; Putnam, 2000, 2007). Higher RED in an area heightens the probability of exposure to those who are racially/ethnically different. Trust thrives on the perception of similarity of others to oneself because such perceived similarity “is an indicator of shared norms and other behaviour-regulating features relevant for trust” (Dinesen et al., 2020, p. 444).

The meta-analysis by Dinesen et al. (2020) found that trust in neighbors was strongest as a factor when the assessment of RED centered on “geographically bounded … residential contexts” broadly defined as “neighborhoods … municipalities, metropolitan areas, regions, and countries” (p. 443). Dinesen et al. (2020) suggest that “a plausible interpretation of the stronger relationship between ethnic diversity and trust in neighbors than for generalized social trust is that exposure to racially/ethnically dissimilar others is a stronger and more directly relevant cue for trust in neighbors than for trust in other people in general” (p. 452). Public and social media accounts also tend to make RED “much more visible, seemingly more local, and of more immediate concern” (McCann, 2023, p. 112). Such trust positively related to FV in the state-level research of McCann (2023) and in the later unpublished analysis (see McCann, 2026a).

McCann (2023) also speculated that principles based on mere exposure and homophily are fundamental to relations between RED, trust, and FV. The homophily principle recognizes that people have strong desires and tendencies to associate and interact with similar others (e.g., Alesina & La Ferrara, 2000; Haun & Over, 2015; Zhao, 2023). Homophily and trust have been positively related in various contexts (e.g., Ahlf et al., 2019; Evans & Wensley, 2009; Kim, 2015). Higher exposure to RED “implies a larger dose of negative cues regarding the trustworthiness of others in general” (Dinesen et al., 2020, p. 445). Lower trust is associated with less interpersonal contact and cooperation (Gächter et al., 2004; Putnam, 2007). Most FV activities depend upon trust, interpersonal contact, and cooperation. Consequently, lower trust is likely to result in less FV.

The focus of the current research is on the relation of RED to FV. The general picture painted here is that higher levels of RED in an area are associated with a lower likelihood of engaging in FV in that area. Although this account tends to attach negative connotations to RED, it is the widespread popular view of RED that it is beneficial in other circumstances. For example, many believe that it may enhance cognitive skills, critical thinking, and creativity in the classroom (e.g., American University, 2019), may improve problem solving, innovation, and productivity in the workplace (e.g., Mattina, 2025), and may reduce prejudice, increase empathy, provide opportunities for stronger social cohesion, and foster equity at the societal level (e.g., Ricee, 2023). It should be stressed that the current project is solely concerned with the empirical relation of RED to FV, and that any potentially negative connotations regarding RED engendered by the research should not influence perceptions of other relational contexts.

The primary objective of the current study was to provide an initial cross-cultural test of the major findings of the McCann (2023) research and the unpublished analysis (see McCann, 2026a). The main hypothesis was that, although LAD FV rates may be negatively correlated with neuroticism and RED and positively correlated with socioeconomic grade (SEG) and trust, statistically controlling for trust of neighbors may eliminate any association between higher RED and FV rates. Such results would support the findings of McCann (2023) and the unpublished analysis and strongly suggest that trust mediates the relation between RED and FV in UK LADs as well as in US states.

The association between higher RED and lower trust has been found in both the US and the UK (e.g., Dinesen et al., 2020). However, Helbling et al. (2015) found that the relation is more likely to exist in countries demonstrating a stronger preoccupation with immigration issues. For example, Dinesen and Sonderskov (2018) noted that in the US “high levels of ethnic diversity paired with an idiosyncratic pattern of adverse race and ethnic relations has formed a national backdrop particularly conducive to conflictual local interactions eroding social trust” (ResearchGate version, p. 7), and “that political elites and their rhetoric play a role in connecting citizens’ experiences with diversity to their perception of the generalized other” (ResearchGate version, p. 8). These same factors also may be operative in England.

There also is evidence of adequate variability in FV rates (Sutcliffe & Holt, 2011) as well as sufficient differences in cultural, economic, and historical characteristics in the LADs of England to facilitate a replication of McCann (2023). For example, the LADs differ on the Big Five personality dimension of neuroticism (Rentfrow et al., 2015). As well, “the concept of a North–South divide, a line that dissects England, rendering the two halves economically, socially and culturally different, permeates English history” and shows clear differences in educational opportunities, poverty, wealth, employment, infrastructure, investment, transportation options, health, mortality rates, and political preferences (Burton, 2021). The seven regions of England (i.e., South West, South, South East, London, Midlands, North West, and North East) also differ markedly on social trust, including trust of others in general, of those met for the first time, and of those of another nationality (The Policy Institute, 2023). Furthermore, research shows that greater FV is associated with stronger “horizontal” social cohesion based on neighborliness, trust in others, and tolerance of people from other social groups, and that levels of such cohesion differ across the regions and LADs of England (Abrams et al., 2023). The LADs also have vastly different proportions of ethnic groups represented in their populations (Census, 2021). All of these categories of factors appear to be related to FV, thereby enabling a replication of the US findings. More specifically, it would seem possible to determine whether higher FV rates occur in LADS lower on neuroticism, higher on SEG, higher on trust, and lower in RED, and, whether the multivariable pattern of relations found with US state data also pertains to English LADs.

Cultural similarities between the US and the UK also suggest the possibility of successfully replicating the US results of McCann (2023) and the unpublished analysis (see McCann, 2026a) with English data. Although FV rates vary considerably across European nations (Gil-Lacruz et al., 2017), the rate was 24% in England in 2008 (Statista, 2024) compared to the rate of 26.4% in the US in 2008 (BLS, 2008). Therefore, the overall levels of FV in the US and England are relatively close. Researchers also have found that the association between higher ethnic diversity and lower trust may occur in both the US and the UK. For example, McCann (2023) found such a link in the US and Lymperopoulou et al. (2022) found that there was a negative association between ethnic diversity and trust in the UK. To the degree that there are cultural similarities between the US and the UK, it is reasonable that we should be somewhat more confident that the main findings of McCann (2023) will replicate in England.

However, although it is apparent that there are cultural similarities between the US and the UK, differences also exist that could be relevant to formal volunteering, ethnic diversity, and trust, as the following examples indicate. Americans are more likely to value and foster individualistic lifestyles while Britons tend to have a somewhat stronger focus on community and collectivistic responsibility (Harvey, 2023). Ideological polarization has been more pronounced in the US than the UK on cultural issues regarding the prioritizing of national customs and the importance of assimilating immigrants into the nation’s way of life (Silver, 2021). In a study of 13 countries including those as different as Canada, China, Netherlands, and India, the US was highest on social motivation for volunteering and England was the third lowest — only higher than Japan and Croatia (Grönlund et al., 2011). The US has a more diverse population speaking several different languages because it has a much longer and stronger history of immigration from many different countries, whereas English has long been the overwhelmingly predominant language commonly spoken in the UK (Harvey, 2023). The UK also has been demonstrated to be markedly higher than the US on social trust, including trust of others in general, those of a different social group, those in the neighborhood, those known personally, those met for the first time, and those of another nationality (The Policy Institute, 2023). Despite the many similarities between the two countries, these cultural differences between the US and the UK suggest that a successful cross-cultural replication of the general results pattern found by McCann (2023) and the unpublished analysis (see McCann, 2026a) was not a foregone conclusion.

It is important to know whether the US state-level relations also are found with the LADs of England as the analytic units. There are some studies using UK data that report relations of FV to neuroticism (e.g., Claxton-Oldfield et al., 2012), SEG (e.g., Sharp et al., 2023), trust (e.g., Storm, 2015), and RED (e.g., Hylton et al., 2019). Others report relations of neuroticism to SEG (e.g., Rentfrow et al., 2015), relations of SEG to trust (e.g., Janmaat, 2019), and RED (e.g., Pesta et al., 2023), and relations of trust to RED (e.g., Lymperopoulou et al., 2022). However, no existing research examines the interrelations of FV, neuroticism, SEG, trust, and RED in a multivariable framework with either individuals or LADs of England as the analytic units. The present study seeks to fill this void, at least at the LAD level of analysis.

The present study constitutes a cross-cultural test of the major thrust of the original study by McCann (2023). The study includes overall FV rates, resident neuroticism, SEG, trust of neighbors, and RED measures for each of 316 LADs in England. Variables are based on data collected between 2008 and 2015. Specifically, the following hypotheses are based directly on the results of McCann (2023) and the unpublished analysis (see McCann, 2026a):

Table 1 displays the means, standard deviations, and Pearson correlations for the 316 LADs. Notably, FV rates correlated significantly with neuroticism (-.42), SEG (.41), trust (.64), and RED (-.35). Trust also correlated significantly with RED (-.75).

Table 2 shows the results of two multiple regression equations demonstrating the capacity of neuroticism, SEG, RED, and trust to predict LAD FV rates. The assumptions of multiple regression were generally met for each equation. For example, scatterplots indicated appropriate linearity, there was no autocorrelation in the residuals of the two equations as indicated by a Durbin-Watson value of 2.01 in the first and 2.03 in the second, there was no evidence of multicollinearity in either equation as shown by VIFs ranging from 1.04 to 3.35 which are well below the common threshold of 10 and even the more conservative threshold of 5, and Mahalanbois distance analysis showed no multivariate outliers. Also, normal predicted probability (P-P) plots indicated no significant departure from normality for residuals, and the scatterplot of the standardized residuals and the standardized predicted values showed acceptably even spread, indicating support for homoscedasticity. In addition, histograms showed normality of residuals and Q-Q plots of the standardized residuals also appeared normal, although the Kolmogorov-Smirnov test suggested some departure from normality (p = .040) for Equation 1 while the Shapiro-Wilk test did not. In contrast, both tests indicated normality for Equation 2.

In Equation 1, neuroticism entered on Step 1 accounted for 17.3% of the FV rate variance, F(1, 314) = 65.61, p < .001. SEG entered on Step 2 accounted for another 5.2%, F(1, 313) = 20.84, p < .001. RED entered on Step 3 accounted for a final 13.3%, F(1, 312) = 64.45, p < .001. The significant βs were -.20 for neuroticism, .34 for SEG, and -.37 for RED.

Equation 2 had the same form as Equation 1, but trust entered on Step 3 and RED entered on Step 4. Trust accounted for a further 25.0% of the FV variance, F(1, 312) = 148.57, p < .001. However, RED now could account for only a nonsignificant .5% of the FV rate variance, F(1, 311) = 3.02, p < .001. The significant βs were -.15 for neuroticism, .13 for SEG, and .64 for trust.

The findings of McCann (2023) and the later previously unpublished analysis (McCann, 2026a) were successfully replicated with data for the 316 English LADs. The first 10 hypotheses were supported by the Pearson correlations. As expected, FV rates correlated positively with SEG and trust of neighbors, and negatively with neuroticism and RED. Neuroticism correlated negatively with SEG and trust. SEG correlated positively with trust. Trust correlated negatively with RED. Also as expected, RED did not correlate with neuroticism or SEG. These correlations substantively paralleled the US results found in McCann (2023) and in the unpublished analysis.

The 11^th hypothesis that “controlling for trust eliminates the association between RED and FV” also received support from the planned analysis using the 3-and-4-variable approach. Controlling for neuroticism and SEG did not eliminate RED as a predictor of FV rates, but adding trust as a third control did so. This pattern of results also emerged in the unpublished analysis. However, such an analysis applied to the McCann (2023) data did not wholly support this outcome. In the 4-variable equation, SES did not have a significant β weight, and trust as a third control substantially reduced but did not eliminate the predictive capacity of RED.

However, the UK support for the 11^th hypothesis was somewhat qualified by the exploratory 1-and-2-variable analysis in which controlling only for trust greatly reduced but did not entirely eliminate the capacity of RED to predict LAD FV rates. In other words, controlling for neuroticism, SEG, and trust eliminated RED as a predictor of LAD FV rates, but controlling for neuroticism alone only reduced the predictive capacity of RED in this UK context. In contrast, the US results based on McCann (2023) and the unpublished analysis showed that RED was eliminated with trust as the sole control variable.

The results strongly suggest that trust serves as a mediator between RED and FV. However, formal mediation analysis generally is not recommended when the researcher possesses only cross-sectional data without clear temporal sequencing (e.g., Fairchild & McDaniel, 2017; O’Laughlin et al., 2018; Shrout, 2011). Nevertheless, Shrout (2011) does suggest that mediation analysis with cross-sectional data might be somewhat informative if the research context is one “in which there are well-founded theories that describe the causal direction of the processes, and for which the interpretation of the cross-sectional measures is informative about the temporal process” (p. 857).

The present study assumed that the temporal sequence in the indirect path in the present study is from RED to trust to FV, but there has not been sufficient argument for this chain of events to justify mediation analysis in the current analytic strategy according to the stated qualifications of Shrout (2011). Nevertheless, in an exploratory fashion, hypothetically assuming that one could argue successfully for such a clear temporal sequence, an application of the Sobel Test of Mediation (Preacher & Leonardelli, 2025) did suggest that trust may mediate the relation between RED and FV. Equations including neuroticism and SEG as statistical controls yielded a Sobel Test statistic of -8.12, a standard error of .28, and a probability value less than .001. Equations without these two statistical controls produced a Sobel Test statistic of -11.10, a standard error of .26, and a probability value less than .001.

Geographical psychology researchers (e.g., McCann, 2023; Rentfrow et al., 2008; Rentfrow & Jokela, 2016) are aware that aggregate level results may or may not successfully extrapolate from one level to the other. There are dangers of committing the compositional fallacy (e.g., Pettigrew, 1997) or the ecological fallacy (e.g., Robinson, 1950) if such extrapolations are based purely on assumptions rather than empirical evidence at both levels. Nevertheless, despite cautions, it is often useful to put forth judicious aggregate-level speculations based on cross-level extrapolative rationale grounded in individual-level dynamics and vice versa (Rentfrow & Gosling, 2021). The speculation here is that the LAD-level results stem from the aggregation of corresponding individual-level relations.

The major thrust of the findings with English LADs is the same as in McCann (2023) and the unpublished analysis (McCann, 2026a) for US states: Statistically controlling trust eliminates or drastically reduces the relation between RED and FV. The speculation here is that the principle of homophily is at the root of the link between RED and trust. Perceived racial/ethnic differences tend to signal dissimilarity of values, attitudes, beliefs, and behaviors. High RED fosters distrust. In turn, lower trust is not conducive to interpersonal contact and cooperation. FV then suffers because it depends to a large extent on personal interaction and cooperative efforts with others — often initially strangers. Individual-level studies also support the plausibility of these speculations. For example, researchers have found that trust relates positively to homophily (Ahlf et al., 2019) and to FV (Miranti & Evans, 2019); that homophily relates positively to FV (Wiertz, 2016); and that RED relates negatively to trust (Putnam, 2007), homophily (Bacharach et al., 2005), and FV (Putnam, 2007).

Although relations between trust, homophily, and trust may indeed provide a plausible explanation for the current results, it remains possible that the interrelations of other variables might also furnish viable alternative interpretations. Variables such as the individual FV rates of different racial/ethnic groups, the differences in racial/ethnic population percents, the homeland cultural norms regarding FV of those not born in the UK, the specific barriers to FV faced by particular racial/ethnic groups, and the relative availability of FV opportunities for different racial/ethnic groups may be implicated to some degree in the relation between RED and FV rates. Of course, theoretically articulating and testing how such variables might afford an explanation of the association between higher RED and lower FV rates is well beyond the scope of the current study. At present, the explanation focusing on trust, homophily, and RED seems most viable, elegant, and parsimonious.

Regarding spatial autocorrelation, perhaps one could have somewhat anticipated that spatial clustering would be likely for the UK data. Relatively short distances even between more distant LADs fairly readily facilitates the geographic dispersion of ethnic groups in England. More closely adjacent LADs also make it more likely that clusters of LADs may share beliefs and attitudes on a multitude of issues and events, including FV and RED. Nonetheless, as previously noted, incorporating spatial autocorrelation adjustments into the UK analysis produced no substantive difference compared to the original OLS multiple regression coefficients. Similarly, when spatial autocorrelation was detected in the US data, there was no substantive difference between spatially adjusted and original OLS results.

Should we expect that the results patterns for the US (McCann, 2023; McCann, 2026a) and England regarding trust, RED, and FV demonstrated in the present study also applies to other nations? Both the US and England have a predominantly White population steeped in individualism where FV is an established part of the social fabric (e.g., Kemmelmeier et al., 2006). As well, Dinesen and Sonderskov (2018) noted that the US has a history of “high levels of ethnic diversity paired with an idiosyncratic pattern of adverse race and ethnic relations” (p. 7) and thereby may be particularly susceptible to the erosion of trust. They also assert that the rhetoric of elite politicians and their followers influences citizens to associate their experiences with RED to their perception of others in general and this subsequently leads to lower trust. England may provide a somewhat similar context. Dinesen and Sonderskov (2018) also pointed out that the relation between RED and trust may be weaker in other Western countries than in the US. However, England likely is among the relative exceptions.

Helbling et al. (2015) suggest that a strong negative relation between RED and trust is only likely in nations having a relatively high preoccupation with issues of immigration. However, there has been a tremendous increase recently in immigration in many countries, including England. Diversity, equity, and inclusion (DEI) movements also have proliferated, generally elevating political turbulence and polarization. Increases in RED undoubtedly have prompted commensurate concern with immigration issues making it more probable that many other countries will likely show the pattern of relation between trust, RED, and FV found in the present research and that of McCann (2023) and the previously unpublished analysis (see McCann, 2026a).