COVID-19 as Natural Intervention: Guilt and Perceived Historical Privilege Contributes to Structural Reform Under Conditions of Crisis.

The COVID-19 pandemic has been described as an unmasking of persistent racialized inequalities linked to South Africa's oppressive past. However, such observations lack empirical support. Here we examined whether COVID-19 lockdown conditions encouraged greater perceptions of continuing structural racism together with motivational and behavioral support for social justice, and whether guilt or empathic concern undergirded such responses. A national sample of White South Africans' data suggests that the pandemic served as a natural intervention, fostering greater acknowledgement of structural racism and support for redress through increased awareness of historical privilege and guilt in response to Black hardship. Guilt furthermore predicted a social justice motivation in relief efforts, whereas empathic concern predicted only charity motivation. These results suggest that "White guilt" is more consequential than empathic concern in contributing to structural reform but would require longer-term processes to support the translation of its motivational push into sustainable contributions to social justice. Supplementary Information: The online version contains supplementary material available at 10.1007/s42761-021-00073-7. © The Society for Affective Science 2021.

Over the past 25 years, South Africa has increasingly featured White denial of the roots of present-day racialized inequalities (Lefko-Everett et al., 2017). Although Whiteness continues to yield significant (economic) power and privilege (Chatterjee, 2019; Ratele & Laubscher, 2010), right-wing politics enforce the notion that the White minority is systematically oppressed and marginalized by the post-apartheid government (Van Zyl-Hermann, 2018). Such perceptions of victimhood often manifest in reduced acknowledgement and dampened empathic concern for continuing structural racism (Fourie & Moore-Berg, 2021), which serves the added benefit of maintaining the status quo and access to tangible material advantages (Milazzo, 2016; Unzueta & Lowery, 2008).

The unequal health and economic impact of the COVID-19 pandemic has highlighted the racialized nature of divisions between rich and poor worldwide (Bottan et al., 2020; Turok & Visagie, 2021). In South Africa, it has been described as an unmasking of the extent to which White South Africans and their descendants continue to benefit from centuries of systemic racialized privileging, so that the materiality of racism can no longer be ignored (Chatterjee et al., 2020; Haffajee, 2020). For example, the structural barriers (including infrastructure, water, food, sanitation, and digital access) to adhere to and survive COVID-19 lockdown restrictions faced by Black communities in township settings have been forced into view unequivocally (Coetzee & Kagee, 2020; De Groot & Lemanski, 2021).

While there has been greater articulation of White privilege in the media, it remains to be explored whether the COVID-19 disruption in public life has shifted White South Africans’ acknowledgement of structural racism (Salter et al., 2017) and their motivational and behavioral support for social justice (Durrheim & Dixon, 2018) or whether the pandemic strengthened a culture of paternalistic charity (Jackman, 1994). The distinction between a charity versus social justice motivation for helping is critical in terms of their impact on both human dignity and social inequality: whereas charity is associated with little appreciation for the root cause of a situation and reinforces the dependency of marginalized communities, a justice motivation is borne out of a longer-term desire to empower marginalized communities because of the recognition of historical injustice and resulting personal gain (Halabi et al., 2008; Swartz, 2016). Finally, we were uniquely interested in which group-based emotions, particularly self-focused guilt and/or other-focused empathic concern, in this context would be associated most strongly with structural reform (Iyer & Leach, 2008; Vollhardt & Sinayobye Twali, 2016).

Much has been theorized about White guilt and shame, and whether these self-directed emotions are appropriate responses to legacies of violent racial subjugation and persistent White privilege in post-colonial and post-slavery contexts (Milazzo, 2016; Phyfer et al., 2020; Sullivan, 2014; Vice, 2010). Yet a large body of empirical research in intergroup settings point to collective guilt’s association with support for affirmative action policies and efforts at restitution when experienced by members of historically advantaged groups (Doosje et al., 1998; Imhoff et al., 2012; Iyer et al., 2003; Leach et al., 2006; Swim & Miller, 1999). Shame’s scrutiny of the entire self typically renders individuals less able to connect with others’ suffering (Eisenberg, 2000; Fourie, et al., 2017a, b; see also Gausel & Leach, 2011), whereas guilt’s association with a sense of responsibility for harmful acts typically motivates repair (Iyer et al., 2004). Because group-based emotions parallel individual-level emotions (Salice & Montes Sánchez, 2016), collective guilt has also been associated more strongly with tendencies to apologize and with outgroup reparation support over time than collective shame (Brown & Cehajic-Clancy, 2008; Brown et al., 2008).

Empathy reflects the natural capacity to (i) understand and (ii) resonate with another’s emotional state, coupled with the (iii) desire to contribute to their well-being (Decety et al., 2016). Because these empathy phenomena are dissociable in terms of subjective experience, neural substrates, and behavioral outcomes, they should be distinguished (Decety & Yoder, 2016). The present research involved the latter aspect, empathic concern, which includes other-oriented feelings of compassion or sympathy for someone suffering (Batson et al., 2015). Unlike personal distress, which is an egoistic motivation to reduce personal aversive feelings, empathic concern is associated with altruistic motivation and resource sharing to reduce others’ suffering (Batson et al., 1987).

While empathic concern thus motivates prosocial behavior, various social and psychological factors along with situational motivations influence the extent to which, and towards whom, empathic concern is deployed (Bruneau et al., 2017; Cameron et al., 2016; Fourie et al., 2017b; Zaki, 2014). Moreover, in contexts of asymmetric power dynamics, where advantaged group members benefit from current status arrangements, empathic concern might be insufficient to produce prosocial change (Dixon, Durrheim et al., 2010 and Dixon, Tropp et al., 2010; Saguy & Kteily, 2014). In the absence of equitable goals and norms, empathic concern in such contexts might serve to mask structural inequality and maintain the status quo (Zaki & Cikara, 2015). Indeed, the stereotype content model predicts that people respond to social outgroups high in warmth but low in perceived competence or status with paternalistic positivity/pity that legitimizes inequality (Fiske et al., 2007, 2016).

Taken together, here we examined whether the unprecedented COVID-19 hard lockdown conditions fostered greater acknowledgement of continuing structural racism while raising White South Africans’ recognition of our implication and shared ethical responsibility for redress as beneficiaries of apartheid injustices (Leach et al., 2013; Rothberg, 2019). We tested the following specific hypotheses.

First, we hypothesized that a nationally representative sample of White South Africans would express greater empathic concern than self-incriminating guilt/shame in response to Black hardship during the 2020 COVID-19 hard lockdown. We had no specific predictions for personal distress, anger, and fear, which we included for comparison purposes.

Second, we predicted that greater acknowledgement of socioeconomic inequality as ingroup privilege during the pandemic would be associated with greater guilt and awareness of historical White privilege (Leach et al., 2002; Powell et al., 2005). Third, we predicted that heightened guilt (vs. empathic concern) and greater perceived historical privilege would predict progressive perceptual (perceived anti-Black structural racism), behavioral (support for redress), and motivational (social justice) intergroup outcomes but would be unassociated with right-wing populist attitudes and unsustainable charity. While existing work investigating both guilt and empathic concern’s reparative tendencies are inconclusive (Harth et al., 2008; Iyer et al., 2003), we argued that guilt’s association with ingroup responsibility for illegitimate advantage would be more consequential in generating motivational and actual support for restitution than empathic concern (Cehajic-Clancy et al., 2016). Moreover, empathic concern in response to low-status outgroups might be confounded by pity (Fiske et al., 2007). We explored age as another predictor of interest, given that our sample spanned more than one generation.

Finally, in line with the hypothesis that the pandemic served as a natural intervention, we predicted that support for redress and perceptions of anti-Black structural racism would be greater in the direct aftermath of the 2020 COVID-19 hard lockdown compared to similar data collected in 2018. Because of a growing sense of victimhood amongst the South African White minority (Van Zyl-Hermann, 2018), we also examined participants’ beliefs about racism directed towards the ingroup (“anti-White racism”; Saguy et al., 2013). Utilizing measures of both perceived anti-Black and anti-White racism provides a more accurate reflection of perceived racialized realities and differs from most WEIRD research (Rad et al., 2018) examining only anti-Black racism.

Methods

Participants

2020 COVID-19 Pandemic

An a priori power analysis using G*Power 3.1 (Faul et al., 2007) indicated that 395 participants would be sufficient to detect a small effect (f2 = 0.02) at an alpha of 0.05 with 80% power. We collected data from a national sample of 461 White South Africans between 12 June and 12 July 2020, the direct aftermath of the South African COVID-19 hard lockdown (level 5).1 Of these participants, we excluded 61 who failed to respond correctly to an attention check question, leaving 400 participants (age range: 18 to 71 years, M = 34.90, SD = 11.57; 40% male). The median income bracket of this sample was ZAR11,000-ZAR17,999 per month, 86% identified as Christian, and 62% obtained some tertiary education.

2018 Pre-pandemic

To determine whether (i) perceptions of racism and (ii) support for redress were greater during the pandemic than before, we compared these variables to similar data collected from a national sample of 195 White South Africans during August 2018 (age range: 20 to 74 years, Mage = 39.62, SD = 12.12, 23% male). The median income bracket of this sample was ZAR11,000-ZAR17,999 per month, and 54% obtained some tertiary education. Data was collected as part of a larger study that included additional measures less relevant to the present research question. Part of this study is reported in Fourie and Moore-Berg (2021).

The same panel service was employed to collect both data sets, and 84 respondents (21%) completed both surveys. Additional details on sampling can be found in the online Supplementary Material. All participants provided informed consent prior to participation.

Procedure and Stimuli

Participants completed online surveys specific to the South African context. Each survey stipulated that we do not endorse the legitimacy of artificial, historically imposed, racial categories, but that we accept the resultant realities of racialized identities, which continue to provide a context for people’s lived experiences. We furthermore indicated that, unless specified otherwise, the word “Black” was used inclusively, referring to all persons of color who were disempowered during apartheid.

COVID-19 Pandemic Measures

Affective Responses to the Pandemic

These items captured affective responses to the plight of Black individuals during the COVID-19 pandemic. Participants were asked “When considering the hardships of Black South Africans, what kinds of feelings did the current pandemic evoke for you?” for each of 6 different emotions on slider scales from 1 (not at all) to 9 (very much). Emotion items included compassion (empathic concern), guilt, shame, anger, fear, and personal distress. Empathic concern was operationalized as compassion to ensure that participants understood its meaning (Fourie et al., 2019; Zaki & Cikara, 2015).

COVID-19 Observations of Inequality

Seven questions probed participants’ general observations regarding socioeconomic inequality and the experiences of a relatively privileged White minority compared to a largely disadvantaged Black majority during the COVID-19 pandemic (see Supplementary Material). For example, participants were asked whether they agree with statements such as “During the current COVID-19 hard lockdown, White people had better access to medical facilities and essential services than Black people” and “Black people were more dependent on the distribution of food parcels than White people” on a scale ranging from 1 (not at all) to 9 (very much) (α = 0.93). Higher scores indicated greater acknowledgement of socioeconomic inequality.

Perceptions of Historical Privilege

Five questions assessed participants’ current awareness of being White and privileged, thus having benefited historically from apartheid injustices (see Supplementary Material). In this way, we assessed perceptions of implication arising from gross human rights abuses, specifically that of being a beneficiary (Rothberg, 2019). For example, participants were asked to indicate whether they agreed with statements such as “I have benefitted economically from apartheid: The inheritance I receive (will receive) from my parents is based on pre-1994 benefits” and “My children will continue to benefit economically from the injustices of apartheid” on a scale ranging from 1 (strongly disagree) to 7 (strongly agree; α = 0.76).

Perceptions of Racism

To assess perceptions of interpersonal and structural racism, we asked participants “How much did/does each form of racism affect Black African2 people in South Africa” (anti-Black racism) at two time points: during apartheid and today, and “How much does each form of racism affect White people today” (anti-White racism), on continuous sliders ranging from 1 (not at all) to 100 (very much; adapted from Rucker et al., 2019). To ensure that participants understood the distinction between these terms, interpersonal racism was defined as “negative attitudes and discriminatory behaviors by individuals towards members of specific racial groups,” whereas structural racism was defined as “institutional practices and structural factors (e.g., laws, policies) that routinely disadvantage specific racial groups.”

While we acknowledge that the term racism should be reserved for race-based oppression along historically institutionalized power dimensions, we examined perceptions of anti-White racism to gain a better understanding of White participants’ perceptions of feeling oppressed (i.e., victimhood).

Support for Redress (Behavior)

To assess support for real-world outcomes that will have a positive impact on historically marginalized individuals and communities in South Africa, participants were given the opportunity to sign five purportedly real petitions urging government to implement progressive policies. These included support towards a wealth tax for the 10% wealthiest citizens, Black Economic Empowerment legislation, Affirmative Action legislation, free tertiary education for previously disadvantaged students, and land expropriation policies (Dixon, Durrheim et al., 2010 and Dixon, Tropp et al., 2010; see Supplementary Material). For each petition, participants indicated whether they wanted to add their signature to the petition (coded as 1), to a counter-petition (coded as − 1), or to neither (coded as 0; α = 0.84). Using these codes, we created an average index of support for redress for each participant.

Prosocial Motivation

Participants were asked to indicate whether they have contributed personally towards relief measures for Black South Africans during the COVID-19 lockdown in terms of time or money (yes or no). Those who responded affirmatively were then asked to indicate what motivated their prosocial behavior on continuous sliders ranging from 1 (not at all) to 100 (very much; see Supplementary Material). Three items assessed charity motivation, for example, “It is my calling as a good person to assist the most vulnerable in society” (α = 0.58), and three items assessed social justice motivation, or a sense of moral responsibility for redress, for example, “I have helped because the unequal impact of the pandemic is largely shaped by our apartheid history” (α = 0.84).

Right-Wing Attitudes

To assess right-wing attitudes, participants were asked to indicate their agreement with six questions tapping into populist narratives presented in the media during the 2020 pandemic (see Supplementary Material). For example, participants were asked to indicate their agreement with statements such as “The government has abused COVID-19 to boost their own racist agenda and control White people”, and “During the pandemic, it has once again been up to predominantly White institutions (e.g., White churches, AfriForum) to alleviate the distress of the poor,” on slider scales ranging from 1 (not at all) to 9 (very much; α = 0.85).

Pre-pandemic Measures

The 2018 instrument used to assess participants’ perceptions about the impact of interpersonal and structural forms of racism against Black people (during apartheid and today) and against White people (today) was the same as that used during the 2020 COVID-19 pandemic. These data are reported in Fourie and Moore-Berg (2021).

The 2018 petition instrument assessed support/opposition for the same five measures of redress used in the 2020 COVID-19 pandemic survey. Because the petitions were framed inversely, i.e., whether participants supported policies that will have a negative impact on Black people, data from these petitions were first reverse scored and then averaged to create an index of support for redress (see Supplementary Material).

Results

COVID-19 Pandemic Data

To test our first hypothesis, we conducted a repeated measures ANOVA on the emotion items, which indicated that there were significant differences in the strength of reported emotions, F(4.09, 1633.68) = 70.42, p < 0.001, = 0.15, ε = 0.82.3 Confirming our hypothesis, planned contrasts indicated that empathic concern (M = 6.34, SD = 2.33) was rated as significantly higher than all other emotions (ps < 0.001, rs > 0.37). Guilt (M = 4.35, SD = 2.54) was rated as significantly lower than all emotions (ps < 0.01, rs > 0.15), except shame (M = 4.08, SD = 2.52), which was rated the lowest (p = 0.001, r = 0.16). Interestingly, empathic concern was positively associated with age (r = 0.14, p = 0.004), whereas guilt was negatively associated with age (r =  − 0.16, p = 0.001). No other emotions were significantly associated with age.

Next, we examined the hypothesis that observations of socioeconomic inequality, as it unfolded during the COVID-19 lockdown, would be associated with a heightened awareness of historical privilege, as well as guilt in response to Black hardship (see Table 1 for zero-order correlations and descriptive statistics). As expected, greater acknowledgement of socioeconomic inequality during the pandemic was strongly associated with an awareness of historical privilege, as well as with guilt and shame (rs > 0.45, ps < 0.001). COVID-19 observations of inequality were also positively, but modestly, associated with empathic concern and anger (rs > 0.17, ps < 0.01), and marginally with fear and personal distress (ps > 0.02).4 The difference in strength of correlations between COVID-19 observations of inequality and guilt, and between COVID-19 observations of inequality and empathic concern, was significant, Steiger’s z = 4.99, p < 0.001.

Table 1

Zero-order correlations between 2020 COVID-19 observations of inequality and historical privilege, guilt, shame, empathic concern, anger, fear, and personal distress (N = 400)

	Historical privilege	Guilt	Shame	Empathic concern	Anger	Fear	Personal distress
COVID-19 inequality	.58***	.46***	.45***	.18**	.17**	.12*	.07
M	4.44	4.35	4.08	6.34	4.75	5.12	5.15
SD	1.34	2.54	2.52	2.33	2.58	2.58	2.58
Quartiles	3.4, 4.4, 5.6	2, 5, 6	2, 4, 6	5, 7, 8.8	2.3, 5, 7	3, 5, 7	3, 5, 7

*p < .05. **p < .01. ***p < .001

To gain insight into how well perceived historical privilege and affective responses (particularly guilt and empathic concern) during the pandemic predicted outcome measures, we performed regression analyses.5 Descriptive statistics and intercorrelations of these variables are presented in Table 2, with simultaneous regressions for each outcome measure presented in Table 3.

Table 2

Descriptive statistics and variable intercorrelations: historical privilege, affect, and outcome measures during the 2020 COVID-19 pandemic (N = 400)

Variable	1	2	3	4	5	6	7	8	9	10	11
1. Historical privilege	-
2. Guilt	.28***	-
3. Shame	.29***	.79***	-
4. Empathic concern	.04	.34***	.34***	-
5. Anger	.10	.49***	.49***	.33***	-
6. Anti-Black structural racism	.38***	.45***	.44***	.22***	.31***	-
7. Anti-White structural racism	− .26***	− .10*	− .08	.07	.15**	− .13**	-
8. Support for redress	.34***	.39***	.36***	.11*	.17**	.35***	− .34***	-
9. Social justice motivation	.62***	.47***	.45***	.05	.24**	.50***	− .13	.45***	-
10. Charity motivation	.26***	.21**	.21**	.19**	.14*	.32***	.08	.31***	.49***	-
11. Right-wing attitudes	− .18**	− .12*	− .11*	− .04	.14**	− .17**	.36***	− .31**	− .05	.003	-
M	4.44	4.35	4.08	6.34	4.74	53.48	57.44	.13	59.68	83.84	6.30
SD	1.34	2.53	2.52	2.32	2.58	31.59	32.25	.56	30.70	15.88	1.88
Quartiles						27, 55, 80	31, 63, 85	− .2, 0, .6	35.6, 65.3, 86	74, 87.3, 98.3	5.0, 6.3, 7.8

Note: Perceptions of anti-Black and anti-White structural racism concern ‘today’

*p < .05. **p < .01. ***p < .001

Table 3

Simultaneous regressions predicting outcome measures as a function of historical privilege and affect during the 2020 COVID-19 pandemic (N = 400)

	Anti-Black structural racism (R2 = .29)a		Anti-White structural racism (R2 = .12)a		Support for redress (R2 = .21)		Social justice motivation (R2 = .50)		Charity motivation (R2 = .12)		Right-wing attitudes (R2 = .09)
	β	p	β	p	β	p	β	p	β	p	β	p
Historical privilege	.29	< .001	− .23	< .001	.25	< .001	.54	< .001	.24	.001	− .16	.002
Guilt	.28b	< .001	− .18b	< .01	.33c	< .001	.32c	< .001	.10	.16	− .19	.001
Empathic concern	.08	.10	.06	.23	− .01	.87	− .03	.55	.16	.02	− .06	.30
Anger	.12	.02	.24	< .001	− .01	.86	.08	.16	.03	.66	.26	< .001

Note: Perceptions of anti-Black and anti-White structural racism concern ‘today’. Regression analyses for social justice and charity motivation were performed on a subset of participants (n = 199), who indicated that they have contributed personally towards relief measures during the COVID-19 pandemic

aSimilar regression results are obtained for perceptions of interpersonal racism

bIndicates an estimate that is significant at p < .05 controlling for shame

cIndicates an estimate that is significant at p < .01 controlling for shame

Results of these analyses consistently showed the impact of greater perceived historical privilege and greater guilt, which were also positively associated (r = 0.28, p < 0.001), in predicting higher responses on progressive outcome measures: perceptions of anti-Black structural (and interpersonal) racism today (βs > 0.28, ps < 0.001), support for redress (βs > 0.25, ps < 0.001), and social justice motivation (βs > 0.32, ps < 0.001). Perceived historical privilege and guilt furthermore predicted reduced right-wing attitudes (βs >  − 0.16, ps < 0.002) and reduced perceptions of anti-White structural (and interpersonal) racism (βs >  − 0.18, ps < 0.01). These results could not be explained by shame, as guilt contributed unique variance even after controlling for shame (see Table 3).

Anger was related most significantly to greater right-wing attitudes and increased perceptions of anti-White structural (and interpersonal) racism (βs > 0.24, ps < 0.001), which suggests that it stemmed from a reactionary focus on perceived ingroup disadvantage, rather than on systemic outgroup injustices during the pandemic (Leach et al., 2006). Adding age as an additional predictor of interest did not change existing results but showed that younger age was associated significantly with more progressive outcomes: greater perceptions of anti-Black structural racism and greater support for redress (βs >  − 0.12, ps < 0.01), as well as reduced right-wing attitudes and reduced perceptions of anti-White structural racism (βs > 0.12, ps = 0.01; see Supplementary Table S1).

Of significance is that empathic concern was neither associated with perceived historical privilege (r = 0.04, p = 0.46), nor predicted any progressive outcome measures significantly (ps > 0.10). Instead, empathic concern significantly predicted only charity motivation (β = 0.16, p = 0.02). It should be noted that a charity motivation was dominant in the 199 participants who indicated that they have contributed towards emergency relief: only 33% agreed strongly with a social justice motivation for giving, while 63% agreed strongly with a charity motivation for giving (Fig. 1).

Fig. 1

Social justice versus charity motivation to contribute towards emergency relief measures during the 2020 COVID-19 lockdown. Data are from 199 participants who indicated that they have contributed personally

Although guilt was thus not experienced most prominently (cf. Swim & Miller, 1999), together with historical privilege, it was most strongly associated with observations of socioeconomic inequality during the pandemic, and the only affective response significantly predicting progressive perceptual, behavioral, and motivational outcome measures.

Comparison with 2018 Pre-pandemic Data

To test our final hypothesis, we compared (i) perceptions of racism today and (ii) support for redress following the 2020 COVID-19 lockdown to similar data collected during 2018. While these samples were only partially matched, observing significant progressive changes would provide at least partial support for our hypothesis that the pandemic served as a natural intervention.

As in 2018, participants perceived interpersonal racism today (M = 56.49, SD = 30.29) to have a greater impact than structural racism today (M = 53.48, SD = 31.59) on Black people during the 2020 COVID-19 pandemic, t(397) = 2.84, p = 0.005, d = 0.10. In stark contrast to the 2018 sample, however, where perceptions of anti-White racism today were exceedingly high, participants did not perceive significant differences in either structural or interpersonal racism experienced by Black and White people today during the pandemic, ts < 1.8, ps > 0.15, ds < 0.10 (Fig. 2). Moreover, perceptions of structural and interpersonal racism experienced by Black people today were significantly higher during the pandemic than in 2018 (anti-Black structural racism M = 15.80, anti-Black interpersonal racism M = 9.75, ts > 3.67, ps < 0.001, ds > 0.30), whereas perceptions of anti-White structural and interpersonal racism during the pandemic were significantly lower than in 2018 (anti-White structural racism M =  − 18.02, anti-White interpersonal racism M =  − 17.37, ts > 6.60, ps < 0.001, ds > 0.54). White participants following the 2020 COVID-19 lockdown thus showed greater acknowledgement of structural and interpersonal racism experienced by Black people today, coupled with reduced perceptions of anti-White structural and interpersonal racism today, than those in 2018. These changes suggest that notions of White people as a racially oppressed minority (i.e., White victimhood) in South Africa were less salient immediately following the 2020 COVID-19 lockdown.

Fig. 2

Perceptions about the impact of interpersonal and structural forms of racism on Black and White people (anti-Black and anti-White racism, respectively) during apartheid and today. Panel a represents data collected from a comparable sample of White South Africans during August 2018 (Fourie & Moore-Berg, 2021), while panel b represents data collected following the 2020 COVID-19 hard lockdown. Error bars represent 95% confidence intervals (CI). ***p < .001

Regarding support for redress, we also compared petition data for the 2020 sample to that obtained in 2018. Chi-square analysis indicated that support for redress for each of the five petitions was much more robust following the 2020 COVID-19 lockdown than during 2018 (χ2 > 39.00, ps < 0.001; Table 4). Moreover, the average support for redress score was significantly greater during the pandemic than in 2018 (Mann–Whitney U = 57,665.50, p < 0.001; Fig. 3). These data suggest that White participants were much more inclined to support economic policies of redress in the immediate aftermath of the COVID-19 pandemic than towards the end of 2018.

Table 4

Support for redress: responses to petitions urging government to implement progressive economic policies in 2018 and during the 2020 COVID-19 pandemic

Petitions		2018 (N = 195)		COVID-19 (N = 400)
		Count	%	Count	%
Wealth tax	Oppose	97	49.7	45	11.3
	Neither	66	33.8	170	42.8
	Support	32	16.4	182	45.8
Black economic empowerment	Oppose	116	59.5	107	27.0
	Neither	53	27.2	174	43.8
	Support	26	13.3	116	29.2
Free tertiary education	Oppose	68	34.9	65	16.4
	Neither	86	44.1	159	40.1
	Support	41	21.0	173	43.6
Affirmative action	Oppose	122	62.6	82	20.7
	Neither	52	26.7	173	43.6
	Support	21	10.8	142	35.8
Land expropriation	Oppose	132	67.7	141	35.5
	Neither	38	19.5	170	42.8
	Support	25	12.8	86	21.7

Note: % Support highlighted in bold face

Fig. 3

Boxplots indicating support for policies of redress in South Africa during 2018 (N = 195) and following the 2020 COVID-19 hard lockdown (N = 400). Data were averaged for 5 petitions where participants indicated support (coded as 1), opposition (coded as − 1), or neither (coded as 0). ***p < .001

Discussion

The present research examined whether White South Africans’ perceptions of continuing structural injustice and support for redress were encouraged through greater awareness of historical privilege and guilt experienced during the COVID-19 pandemic. Our data suggest that the pandemic and its societal impact served as a natural intervention: not only were observations of socioeconomic inequality during the pandemic associated strongly with perceived historical privilege and with guilt, but White acknowledgement of structural racism and support for economic redress were also greater in the immediate aftermath of hard lockdown compared to 2018. Although empathic concern for Black hardship was evoked most saliently, guilt together with perceived historical privilege consistently predicted greater progressive outcomes (perceived anti-Black structural racism, support for policies of redress, and social justice motivation), and reduced ultra-conservative perspectives (right-wing attitudes and perceptions of anti-White racism). By contrast, empathic concern predicted only charity motivation. These findings raise several questions for discussion and future research.

First, what was it about the pandemic that potentially raised awareness of racialized privilege and support for redress? We believe the unprecedented conditions of lockdown, which highlighted stark material discrepancies, provided a visceral confrontation with continuing structural racism and a unique opportunity to reflect on unearned privileges based on generations of racial discrimination. The pandemic thus served to bring the past into the present, which likely contributed significantly towards eliciting collective self-critical feelings of guilt (Leach et al., 2013). The fact that the pandemic was a “natural” intervention is also significant. In contrast to political intergroup interventions, where a focus on gross human rights violations is explicit (Fourie et al., 2013), the pandemic’s exposure of unjust racialized inequalities was more indirect, less threatening, allowing more room for self-critical group-based emotions to surface. The South African Truth and Reconciliation Commission process, for example, tended to conflate guilt with a legalistic, individualistic understanding of “criminal guilt,” which arguably undermined more collective feelings of wrongdoing (Swart, 2017; Verwoerd, 1999).

Second, how can we understand younger age’s association with heightened guilt, more progressive outcomes, reduced right-wing attitudes, and reduced perceptions of anti-White racism? These results follow an international trend, where the self-critical engagement with past immoral actions emerge more strongly amongst 2nd and 3rd post-war generations (Branscombe & Doosje, 2004). Because younger White South Africans have no causal responsibility for apartheid atrocities and its legacies as members of “perpetrator” or “participant” generations (Govier & Verwoerd, 2011), the acceptance of being implicated beneficiaries is less threatening legally, morally, and emotionally (Rothberg, 2019). Hence, they are potentially more likely to admit feelings of guilt and provide restitution than older generations. Their group identification might also be less cemented than pre-1994 generations, who were conditioned by a militarized, sanctified “anti-Communist” Whiteness (Verwoerd & Edlmann, 2021). Some evidence suggests that a weaker group identification enables stronger guilt and motivation for repair (Doosje et al., 2006; Powell et al., 2005).

Third, why was empathic concern and charity motivation evoked most saliently and why is this so problematic in the South African context? Faced with vivid, distressing, media images of Black people queuing for food, it is much easier to interpret large-scale hunger as a humanitarian crisis and respond with compassion and potentially short-lived (remote) help (Hunt et al., 2006; Västfjäll et al., 2014), which do not imply responsibility (Meerholz et al., 2019), than to face the dysphoria of guilt and moral implication in apartheid as crime against humanity (Xu et al., 2011). Indeed, empathic concern in response to Black hardship was entirely unrelated to perceived historical privilege in our sample. Moreover, charity strengthens a positive self-image as a “good person” and, especially, a “good White” (Jones, 2004). While we accept that in a crisis charity is better than indifference, the racialized nature of this charity, especially as Christian missionary outreaches, encourages a paternalistic “savior mentality” protecting Whiteness (Bowers du Toit & Nkomo, 2014).

The above interpretation helps to explain empathic concern’s weak association with outcome measures that will contribute to structural reform. In addition, we are drawn to the possibility that participants’ empathic concern responses more accurately reflected pity towards impoverished Black South Africans, a related but conceptually distinct construct (Boleyn-Fitzgerald, 2003). People typically feel worse when misfortunes befall pitied targets (social outgroups high in social warmth but low in status) than ingroup targets (Cikara & Fiske, 2011). Moreover, pitied targets are often the recipients of active helping (e.g., charity), but also passive harm (e.g., neglect; Fiske et al., 2007). To the extent that charity-as-White-pity contributes to the perpetuation of systemic racism, more than passive harm is at stake, however.

Fourth, we need to examine critically the data’s suggestion that White guilt holds promise in contributing to structural reform. Criticisms against the expression of White guilt include its association with ‘virtue signaling’ (Phyfer et al., 2020) and selfish relief from incriminatory feelings (Batson, 1998; Iyer et al., 2003), rather than with genuine other-directed behaviors to transform social injustices. While public displays of guilt may certainly be unproductive, we believe the present anonymous, private conditions of data collection mitigated against the pitfalls of White guilt becoming a public performance, placing undue pressure on Black people to respond (typically through forgiveness). We also agree that an exclusive focus on White guilt as an individual emotional state is insufficient to disrupt the structural systems that sustain and perpetuate racial inequality (Milazzo, 2016; Moore, 2019).

Nevertheless, we are encouraged that our data is supported by substantial previous research observing that advantaged groups’ collective experiences of guilt provide the impetus for social redress, as it “marks what concerns us, still, in spite of all rationalization or denial” (Cehajic-Clancy et al., 2016; Prade-Weiss, 2020, p. 8; Vollhardt & Sinayobye Twali, 2016). Indeed, for those participants who contributed towards relief measures during the pandemic, feelings of guilt strongly predicted a social justice, rather than a charity motivation for helping. Guilt in this context might therefore be understood as emanating not only from the collective misdeeds of group members in the past, but also from the acceptance of and continuous enjoyment of unearned material benefits in the present, especially during lockdown.

Finally, we concur that by itself, guilt is either too fleeting and infrequent (Iyer et al., 2004; Leach et al., 2013), or too ruminative and maladaptive when chronic self-blame focuses attention inward (Harder, 1995), to bring about lasting change. With the easing of lockdown conditions, evidence emerged that neither increased empathic concern and charity nor changes in political awareness and historical responsibility associated with guilt are likely to be sustained (Hamann, 2020). We contend that for genuine empathic concern and guilt to become “unsettling” beyond the immediate crisis (Krondorfer, 2015), these other-oriented and self-critical feelings need to be combined with the deliberate cultivation of capacities through intrapersonal/intergroup relational processes within contexts supporting equitable goals. Only then will their motivational push translate into sustainable commitments to promote social justice.

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 43.7 kb)