Background to the Present Interest in the South African Institute of Race Relations
The South African Institute of Race Relations (hereafter the IRR) has recently found itself in the crosshairs of academics and researchers over problems with its research practices and its media campaigns pertaining to issues in the social and climate sciences, and issues related to the COVID-19 pandemic. Their campaign against Critical Race Theory (CRT) and their association with climate change ‘scepticism’ (better characterised as climate change denial) has been likened to US conservative fear-mongering given its near identical talking-points and rhetoric. A recent open letter, formulated and signed by academics, former researchers and presidents of the IRR, and concerned stakeholders, details some of these issues.1 The concern of these stakeholders is that the IRR is distributing misinformation as a trusted source.
So far, the discussion about the social-scientific and climate-related issues that stakeholders take umbrage with has only taken place in the media. Limiting this discussion to the media could have the negative result of making the exchanges with the IRR appear to be mere differences of opinion rather than matters resting on evidence and expertise. In its media statements, the IRR references its reports and surveys which, they argue, back up their claims with relevant evidence.2 The structure of debate in the media, where claims of different evidentiary weight can be disseminated and promoted irrespective of how likely they are to be true, can suggest the misleading conclusion that the options presented are equal epistemic competitors even when they are not. This is called a false balance in reporting, and it can distort public perception about whether there is a consensus of expert opinion on an issue and how much evidence each option presented has (Dixon and Clarke 2013; Koehler 2016). Worse still, an opinion can be believed to be true merely because it is popular and dominates public discourse which, in turn, can make such a view all the more popular and popularly held (Schmitt Beck 2015).
The media, for better or worse, influences debates and interpersonal discussions about politics that are reflected in public opinion. This has a direct impact on democracy and democratic processes that rely on public opinion (e.g. Iftikhar et al. 2016). Media reports of the results of opinion polls can have self-reinforcing feedback effects on public opinion in which reports of policy preferences influence further policy conformity to those policy preferences (Toff 2018). If what is reported is inaccurate or misleading, public opinion can be shifted towards those inaccurately or misleadingly reported views (Watts et al. 2021). Moreover, since politicians are more likely to demand change in policy and strategy when party support is slipping in opinion polls (Schumacher and Öhberg 2020), we can see how misleading and inaccurate reports of public opinion disseminated in the media can undermine and compromise democratic processes, particularly in public deliberation about policy issues.
The focus of this article is my claim that the IRR uses survey data, which is then disseminated in the media,3 to mislead the public on matters pertaining to race relations and racism. I will not discuss the technical statistical details of the case or measure the extent to which the IRR misleads empirically.4 I restrict myself to arguing how the IRR misleads the public on race relations with the intention to influence public policy debates, using data from their surveys and opinion polls.
I make my case by looking at how conclusions not supported by data are motivated and misread off surveys presented in the reports of the IRR. In this, I argue that the IRR misrepresents the views and opinions of the South African public (§2) for politically motivated ends (§3). The reports I examine are authored by their Head of Research, Anthea Jeffery, and the surveys that they are based on are mostly carried out by the company MarkData (Pty) Ltd on behalf of the IRR (Jeffery 2021: 5). These reports come to conclusions and make recommendations that do not follow from the evidence or data that they present.
Making direct reference to their surveys, I show how the IRR is mistaken in their claim that they can conclude that racism is not an important problem from a question that asks respondents what the most important problem in the country is (§2-1). I then argue that believing that politicians ‘play the race card’ to deflect from their own failures is not evidence for a respondent's beliefs about whether racism is a serious problem; I argue that we also cannot tell from this data if respondents believe that politicians are exaggerating the seriousness of the problem of racism, despite the IRR's claim to the contrary (§2-2). I follow by making the point that the results of opinion surveys about race and economic advancement are evidence of what people believe, but that those beliefs can also be mistaken (§2-3). The point is that whatever beliefs are held by the public about how certain policies affect the advancement of members of different race groups, they are not a substitute for evidence-based investigations about the effects of policy on actual demographic outcomes. When asking about what policies the public support and their opinions on social policy, I caution against the IRR's approach of asking loaded and ambiguous questions, because such questions undermine the ability of respondents to answer accurately or honestly to response items. The form of the IRR's questions makes it logically impossible to interpret the meaning of their answers because of the ambiguities baked into their survey questions and their consequent responses (§2-4).
Beneath these criticisms lies a deeper conceptual issue with the IRR's research. I show that the narrow characterisation of racism as what their former COO, Sinethemba ‘Gwen’ Ngwenya, called ‘relational racism’5 limits the scope of their questions about racism to only partially addressing how people experience racism, thereby undermining the more general claims they make about how often people in South Africa experience racism or how pervasive a problem racism is thought to be, according to their research (§2-5). Their findings also cannot tell us how pervasive a problem racism is given the nature of opinion polls, beyond the problems inherent in the questions that they ask.
I bring together my criticisms of the faults of IRR surveys and their analyses through an assessment of their purpose (§3). I argue that the faults in the IRR's reports are not just mistakes but are a motivated pattern of misinterpretation and strategic misrepresentation of race relations in South Africa for political ends. I argue that this is what makes IRR reports constitute disinformation. The purpose of the IRR's survey data and their motivated analyses of their opinion polls is to give support to the policy positions and proposals that they favour. In this way, the IRR is not only factually incorrect, but also strategically misleads the public and legislators about race relations and policy issues by spreading disinformation.
I end by giving a summary of the case I have made against the IRR, emphasising that the IRR is not a reliable source of information about race relations – particularly racism – and policy issues in South Africa (§4).
The IRR's Attitudinal Surveys and What They Say about Racial Issues
A useful and up-to-date resource for getting an overview of the IRR's argumentative strategy in using survey data to support certain views and policy recommendations can be found in their May 2021 report, Critical Race Theory & Race-Based Policy: A Threat to Liberal Democracy by their Head of Research and Social & Ethics Committee member, Anthea Jeffery (Jeffery 2021). I will be using this report and its analysis of survey data as an exemplar of a deep set of issues in the IRR's research and reporting on race and policy. It is appropriate to use this report as an exemplar because it is their authoritative review of the IRR's survey work on race and policy since the turn of democracy until 2021. This report uses questions and methodology that are similar to all the IRR's other field surveys about the state of race relations since 2001. That it is written by their Head of Research makes it well-placed to be taken as an official view of the IRR on race relations and policy. I focus my assessment on the IRR's claim that ‘opinion polling over many years shows that black South Africans . . . regard racism as a relatively insignificant problem’ (Jeffery 2021: 13). From this claim, the IRR argues for numerous anti-race-based policy positions, such as those against affirmative action and their ideological stance against CRT (§3). I dispute the claim that ‘black South Africans . . . regard racism as a relatively insignificant problem’ is a conclusion that can be drawn from their data, despite their lobbying efforts against race-based policies being based on this claim. By undermining this claim and its purported relationship to evidence, I undermine their arguments for the positions that they lobby for that depend on the claim.
I reproduce and briefly discuss tables of data that Jeffery presents in her 2021 report to illustrate the IRR's pattern of deeply flawed, but politically motivated, surveying and data analysis. I show how they use their surveys to strategically mislead the public and legislators about the problem of racism in relation to policy in South Africa.
You Can't Determine All That Is Important by Asking People What Is Most Important
Table 1 below presents the results of the IRR's 2020 field survey, as shown in Jeffery's 2021 report. This survey asked respondents to identify what they thought were the two most serious unresolved problems since 1994. Jeffery informs us that no prompting was provided and that respondents were free to list any issue that was important to them. She notes that the proportions in the table exceed 100 percent ‘because more than one problem [i.e. two problems] could be mentioned’ (Jeffery 2021: 5).
IRR opinion poll, 2020 (#1 in Jeffery 2021)
| Most serious problem | Total | Black | Coloured | Asian | White |
|---|---|---|---|---|---|
| Unemployment | 53.4% | 56.0% | 52.5% | 35.0% | 38.1% |
| Crime, safety/security | 22.0% | 18.3% | 29.3% | 51.2% | 36.9% |
| Corruption | 18.2% | 15.5% | 22.1% | 27.7% | 34.6% |
| Housing | 16.6% | 17.0% | 21.7% | 13.0% | 9.9% |
| Service delivery | 10.9% | 11.9% | 6.4% | 4.5% | 9.0% |
| Water/Sanitation | 10.0% | 11.8% | 3.3% | 1.2% | 4.0% |
| Education (poor) | 8.6% | 9.2% | 9.1% | 5.9% | 4.2% |
| Poverty/social inequality | 8.0% | 8.4% | 9.9% | 8.8% | 3.1% |
| Infrastructure | 6.3% | 7.1% | 3.4% | 1.6% | 4.2% |
| Women and children abuse | 5.5% | 5.7% | 4.7% | 7.4% | 4.3% |
| Land reform | 4.0.% | 4.8% | 1.1% | 0.0% | 1.4% |
| Corrupt leadership | 3.6% | 2.6% | 7.0% | 15.7% | 4.8% |
| Inequality | 3.5% | 3.6% | 2.3% | 0.6% | 5.1% |
| Racism/discrimination | 3.3% | 3.1% | 1.7% | 2.9% | 6.8% |
The most serious problem reported by South Africans, by a wide margin, is that of unemployment. This finding has been consistent with unemployment or joblessness being ‘flagged as the main concern of most South Africans in all seven of the IRR's field surveys, going all the way back to 2001’ (Jeffery 2021: 9). A little more than half the respondents in 2020 named unemployment as the most serious unresolved problem facing the country since 1994 while only about 3.3 percent reported it to be ‘racism/discrimination’. In Jeffery's discussion of these results, she claims that the data shows that ‘only 3 percent of black respondents identified racism as a serious unresolved problem’ (Jeffery 2021: 6).6 But the field survey did not find that only 3 percent of Black people identified racism as a serious unresolved problem; it only found that 3 percent of Black people mentioned ‘racism/discrimination’ as the most serious unresolved problem since 1994. The only implication that can be drawn in this regard is perhaps that respondents view other issues as more important than ‘racism/discrimination’, not that ‘racism/discrimination’ is not an important or serious unresolved problem. Even such an implication cannot be certain, because no such comparison between the items listed by respondents was made by the survey. Furthermore, this listing of responses does not take into account whether respondents believe there is a link between racism and unemployment, crime, inequality, or other issues that respondents mentioned.
To see how seriously misleading Jeffery's misinterpretation of the data is, one only needs to imagine a scenario in which the rest of the people in the country (97 percent) ranked the item ‘racism/discrimination’ as the third or fourth most serious unresolved problem in the country since 1994. This would mean, in this example, that everyone in South Africa would believe that racism is a serious unresolved problem, but this would not register on these survey results. In fact, if this were the case (ceteris paribus), the survey results presented would remain exactly the same as those that we see in Table 1. If 100 percent of the country thought racism was a serious problem but only 3 percent thought it was the most serious problem facing the country, the results registered by this survey would remain identical. The survey cannot actually tell us if anyone thinks that racism is a serious problem. This is because the survey design is not equipped to answer questions about what people find to be a serious problem. This survey only shows us what people say is the most serious unresolved problem since 1994 in South Africa. It cannot address the question, ‘do you think racism is an important unresolved problem?’ or whether racism is a serious problem, as Jeffery purports it does, because that is not what the survey asks.
Suffering False Dichotomies and the Motivated Misinterpretation of Preferences
Given the common talk of the legacies of Apartheid and colonialism in the country, especially by politicians who link present failures to colonisation, Apartheid, and racism, the IRR's 2020 survey investigated whether South Africans agree with the statement: ‘All this talk about racism and colonialism is by politicians trying to find excuses for their own failures’ (Jeffery 2021: 7). Their results are reproduced in Table 2 below.
All this talk of racism and colonialism is by politicians who are trying to find excuses for their own failures (#4 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| Agree | 54.0% | 52.0% | 58.7% | 65.1% | 63.2% |
| Disagree | 16.0% | 17.5% | 8.8% | 8.6% | 12.6% |
| Neither agree nor disagree | 29.2% | 29.8% | 32.0% | 26.3% | 23.0% |
Using racism as an excuse for problems or failures that presumably have no direct causal connection to racism is colloquially called ‘playing the race card’. Playing the race card is to mention race or racism as a deflection tactic from some more pertinent issue under consideration. Jeffery's report shows that most South Africans believe that government uses racism and colonialism as an excuse for its own failings. In other words, South Africans believe that politicians bring up racism and colonialism to deflect attention away from their own failings – viz., they play the race card.
It is possible for respondents to believe that politicians play the race card to deflect from their own failures whilst simultaneously believing that racism is a serious problem which the government is not addressing. These two beliefs could be related: some respondents could believe that the government does not seriously address racism in order to be able to use racism to direct attention away from their own failures.
The interpretation of this data as showing that ‘more than half of all respondents (54 percent) agreed that politicians are exaggerating the problems posed by racism and colonialism to excuse their own shortcomings’ (Jeffery 2021: 7) is unfounded since it has no basis in the evidence presented. What the evidence shows is that 54 percent of people believe that politicians ‘play the race card’, but the evidence does not show that people believe that politicians exaggerate the problem of racism. A different set of questions would have been required to support such a claim.
Jeffery argues that the data show us that ‘politicians may seek to play up racial differences for their own purposes, but most South Africans are well aware of the importance of better skills and increased earnings in reducing racial inequalities and building inclusive prosperity’ (Jeffery 2021: 7). This conclusion is misleading for the reasons I have just given, but also for the reasons I give in relation to the data produced in Table 3.
Best way to improve lives (#7 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| More jobs and better education | 72.8% | 71.9% | 77.4% | 78.7% | 74.2% |
| Better service delivery | 17.8% | 18.1% | 16.5% | 16.2% | 17.8% |
| More BEE and AA in employment | 3.3% | 3.8% | 0.9% | 2.0% | 2.1% |
| More land reform | 4.2% | 4.2% | 3.5% | 2.3% | 4.7% |
The policy judged to be the ‘best way to improve lives’ by almost 73 percent of respondents is ‘more jobs and better education’. That this is chosen as the best way to improve lives does not exclude other endeavours also being considered as paths for improving lives. Supporting a ‘best way’ to improve lives does not prevent one from supporting other ways of improving lives to be implemented in tandem with the ‘best way’. So, although it may be true that ‘most people, including most black South Africans, have persistently identified “more jobs and better education” as the key to upward mobility’ (Jeffery 2021: 9), it does not mean that they have no desire to pursue other ways they may believe lives could be improved. We cannot conclude, as the IRR does, that ‘very few blacks . . . (around 5 percent on average) have wanted the government to pursue “more” BEE, EE, or land reform initiatives’ (Jeffery 2021: 9) just because only about 5 percent of Black people think that this would be the best way to improve their lives. This is a false dichotomy, an instance of the fallacy of false alternatives in the form of a false dilemma (cf. Tomić 2013).
Another way in which respondents are presented with a false dilemma is when they are asked to choose between receiving a tax-funded voucher and not receiving a tax funded voucher. For example, as shown in Table 4, the IRR asks, ‘would you like tax-funded education vouchers to send children to a school of your choice?’ rather than more open-ended questions, like ‘what school funding option/model do you prefer?’, in which they could present different funding models respondents could compare, rank, or express preferences about. This would allow respondents to choose what school funding model they would prefer rather than having to choose between receiving a tax-funded voucher or not receiving a tax-funded voucher. Of course, when framing the choice as receiving a voucher or not receiving a voucher, this formulation of the question biases responses. Moreover, reporting the results of such a survey as ‘popular support for the voucher option’ (Jeffery 2021: 11) is brazenly misleading because it suggests a comparison between options when, in fact, no other options were offered.
Would you like tax-funded education vouchers to send children to a school of your choice? (#9 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| Yes | 77.9% | 79.6% | 79.7% | 65.4% | 65.6% |
| No | 15.6% | 13.8% | 17.0% | 29.8% | 24.2% |
| Don't know | 6.6% | 6.6% | 3.3% | 4.8% | 10.2% |
The same problem faces each of the survey questions in the IRR's report where respondents have to choose between receiving any sort of tax voucher as opposed to not receiving one, such as vouchers for healthcare and housing (see Jeffery 2021: 11). This error is also found in previous survey reports, such as when the claim is made that in ‘2016 some 85 percent of black respondents expressed support for schooling vouchers, while 83 percent endorsed healthcare and housing vouchers’ (Jeffery 2021: 12) although there were no other options aside from tax funded vouchers presented to respondents. Despite the merits of tax funded vouchers for public services, the structure of this question biases responses. Moreover, the IRR reports these results in a disingenuously misleading fashion because the support reported for tax-funded vouchers is not a result that emerges out of a comparison with relevant alternatives although it is presented as if it is preferred over all other alternatives.
My final example of the falsely dichotomous formulation of questions in IRR's surveys and their interpretations of them in reports is found in Table 5, where two options are presented as exclusive when they can, in fact, be held in tandem. The question asks if the respondents prefer a political party that promises faster economic growth and more jobs compared to one that promises land expropriation without compensation. Respondents expressed preference for faster growth and more jobs. The trouble comes when this data is interpreted. The IRR claims that its surveys show that ‘Black support for increased land reform has . . . been consistently low over many years’ (Jeffery 2021: 12). But their surveys do not ask if Black people support increased land reform; they only ask if respondents would prefer a political party that promises various land reform policies over one promising faster economic growth and more jobs. Presenting the preference for more jobs and faster growth as an exclusive option is a dubious, unfounded misinterpretation of the data and a serious misrepresentation of people's opinions.
Do you prefer a political party which promises faster economic growth and more jobs, or one which promises land expropriation without compensation as redress for past wrongs? (Choose one) (#10 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| Faster growth and more jobs | 81.4% | 80.7% | 85.1% | 90.5% | 81.8% |
| Land expropriation without compensation | 14.6% | 15.2% | 12.3% | 6.3% | 14.2% |
| Don't know | 3.9% | 4.0% | 2.6% | 3.1% | 4.8% |
Evidence of What People Believe Isn't Evidence of What Is True
Opinion polls can tell us what ‘common-sense’ or public opinion is on an issue. They can tell us what people generally believe to be true. What we generally believe to be true can, and often does, turn out to be wrong because, as we know from empirical studies, common-sense intuition and public opinion are often false (cf. Peels and Ridder 2020). For this reason, we cannot appeal to popular opinion or intuition when we want to know what the correct answer is to a question that is independent of what common-sense or public opinion may dictate. That is the case with the statement in Table 6. Agreeing or disagreeing with the statement, ‘with better education and more jobs, the present inequality between the races will steadily disappear’ tells us about what we believe – it does not tell us if the statement is true.
With better education and more jobs, the present inequality between the races will steadily disappear (#5 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| Agree | 73.4% | 72.5% | 80.5% | 78.8% | 72.0% |
| Disagree | 11.3% | 11.2% | 5.2% | 12.8% | 16.9% |
| Neither agree nor disagree | 14.8% | 15.7% | 14.3% | 8.4% | 10.2% |
Let us take for granted, at least for the sake of my argument, that economic prosperity and universal quality education are essential features of long-term sustainable equity among races in South Africa. This claim itself may also be another common-sense or intuitive presumption about the workings of South African society that would need support from empirical evidence. Establishing what is public opinion or common-sense on economic progress, 73 percent of respondents agreed with the statement: ‘With better education and more jobs, the present inequality between the races will steadily disappear’ (Jeffery 2021: 8).7 In an appeal to popular opinion, the interpretation of this result (in combination with the misinterpretation of Table 2) gives support to the claim: ‘Politicians may seek to play up racial differences for their own purposes, but most South Africans are well aware of the importance of better skills and increased earnings in reducing racial inequalities and building inclusive prosperity’ (Jeffery 2021: 7). Yet saying this gives the impression that the statement in Table 6 is a well-established fact rather than an opinion. Note that Table 6 mentions ‘more jobs’, and the statement in Jeffery's analysis and conclusion mentions ‘increased earnings’, which are significantly different things. Nevertheless, the main point I am making here is that taking public opinion to be indicative of fact is a fallacious mistake. Evidence of what people believe is not necessarily evidence of what is true. It is for this reason that sometimes the best decision-making process for policy makers, although deliberative, may not always be to settle on exactly what survey data show that people believe, rather than what has, in fact, been found to be the case (Solomon and Abelson 2012).8
The belief that with ‘better education and more jobs, the present inequality between the races will steadily disappear’ is not a new presumption among Black people in South Africa. Blacks who were colonially assimilated in the nineteenth century believed this, too. Despite their relatively better education and having access to more jobs with higher pay, their advancement in South African society was stifled by various racist barriers to their full participation in the society that they were assimilating into. The Kholwa of Edendale are a tragic case study of the failure of this optimistic view, at least in colonial South Africa (see Meintjes 2020). One may think that things have changed so considerably, with the repeal of racist laws, that it may be true today that education and more jobs can close the gap in racial inequality. Contemporary evidence that I detail below suggests otherwise.
The expectation that education and employment will close the racial income inequality gap is also a common expectation among the contemporary public in the US (Emmons and Noeth 2015). Studies have shown that this expectation is unfounded, because educational attainment has not had the effect of closing the racial income gap despite the fact that education typically boosts income and wealth (see Emmons and Noeth 2015 and references therein). This has many complex reasons, one of which, specifically in the US context, is the racially differentiated burden of debt that students carry when they graduate. Data in the US context shows that ‘higher education alone cannot level the playing field’ (Emmons and Noeth 2015: 3). Given how different the US and South Africa are, we cannot simply conclude that the same applies to South Africa, although some of the reasons for continued inequality (e.g. debt at graduation) are suggestive.
There may be similarities in US studies that suggest that education and employment alone may not close South Africa's racial income inequality gap. An avenue to empirically investigate this problem may be to measure what difference the so-called ‘Black-tax’ and study debt has on racial inequality in comparison to the racial disparities in student debt in the US. In the US, part of ‘the black-white income gap is driven by differences in wages and employment rates between black and white men’ (Chetty et al. 2020). This is also true of South Africa: differences in the absorption rates of equivalent employees differs by race in South Africa (see Gradín 2019) above and beyond the large employment and wage gap between races in South Africa.
White people in South Africa have the highest absorption rate into the labour force at 61.6 percent in the first quarter of 2021, as compared to 48.3 percent of Indian people, 42.9 percent of Coloured people, and 35 percent of Black people (Statistics South Africa 2021a: 28–29). Whites enjoy the privilege of being both the most desired workers in the economy and those who land jobs that tend to pay them better than other demographic groups. How we know that Whites continue to be unfairly favoured by employers is that ‘segregation and stratification’ in employment ‘remain when blacks and whites with similar characteristics are compared’ (Gradín 2019: 553).
Although more jobs and better levels of education are much needed in South Africa, especially considering its chronic structural unemployment of 34.4 percent in Q2 of 2021 being the highest that has ever been recorded since reporting began in 2008 (Statistics South Africa 2021b: 2), the evidence suggests that more jobs and better education will not be enough to close the racial income inequality gap. Moreover, despite common sense and intuition, it does not seem to be the case that ‘with better education and more jobs, the present inequality between the races will steadily disappear’ (Jeffery 2021: 8). This is because of the discrimination in the workplace already discussed. We cannot, from public agreement with the statement about better education and more jobs, conclude that the statement is true, especially given the countervailing empirical evidence against such optimism.
An important realisation about improving the economic fortunes of people in South Africa who are not White is that favour and discrimination play a role in the continued unequal and inequitable outcomes we observe in contemporary South Africa. In a large empirical study of the Southern African Customs Union (comprising Botswana, Eswatini, Lesotho, Namibia, and South Africa), the World Bank found that race is the greatest single predictor of inequality in the region – especially in Namibia (which was for a time ruled by the South African Apartheid government) and South Africa (Sulla et al. 2022). The most recent up-to-date empirical evidence available shows that racism and racial discrimination, or at least racism's historical legacy and effects, are still a serious unresolved problem in South Africa today, despite the IRR's insistence otherwise.
Leading, Loaded, and Ambiguous Questions Obfuscate the Meaning of an Answer
The IRR field surveys have leading, loaded, and ambiguous questions which produce inaccurate data that are impossible to interpret with any accuracy because of the structure of their questions and how they confound variables that affect item responses. The question in Table 7 is an exemplar of these problems.
The different races need each other for progress and there should be full opportunity for all (#6 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| Agree | 71.6% | 70.5% | 81.4% | 81.5% | 68.2% |
| Disagree | 8.7% | 9.1% | 2.1% | 7.7% | 12.3% |
| Neither agree nor disagree | 19.2% | 19.9% | 16.5% | 10.8% | 18.2% |
Table 7 presents us with a leading, double-barrelled, loaded, and ambiguous question. The ‘question’ asked is if the different races in South Africa need each other for progress and if there should be opportunity of all. It is a leading question in respect to whether races need each other (it could be the case that races do not need each other in South Africa, although it could also be an ethical imperative that the races should co-operate and aid one another). It is a double-barrelled question because it asks, on the one hand, if the races need each other and also asks, on the other hand, if there should be full opportunity for all – in the same response item. A person who does not think that the races need each other but believes that there should be opportunity for all would have no way to accurately register their views to this item, and their response would be misleading no matter what response they gave to this item. The question is also ambiguous. In what way do the races need each other, if at all? Is it that the races need each other for moral progress or for economic progress, or perhaps some other kind of progress? Does ‘full opportunity for all’ mean endorsing representative quotas which extend opportunity to all or ‘Other’ segments of the South African population, or is ‘full opportunity for all’ meant in some more narrow sense which rejects all affirmative action policy and racial quotas as being limiting to ‘full opportunity’? Is support for ‘full opportunity for all’ a euphemistic endorsement of a race-blind meritocracy? The question posed by the IRR resists accurate, confident, or reliable interpretation and does not usefully measure any particular belief or attitude.
Characterising Racism as Explicit Racial Animus is Inadequate and Misguided
A problem exemplified by the IRR's reporting on and discussions of racism, particularly in relation to their policy suggestions relying on their surveys, is that they have a very narrow understanding of racism and racial discrimination. In their report Race Relations in South Africa – Reasons for Hope 2018: Holding the Line, Jeffery opens by recounting what were at the time recent ‘racial incidents’ from across the country in which people – almost exclusively White people – committed racially motivated acts of aggression, assault, and murder (see Jeffery 2018: 1–3). Racism is framed in terms of her examples of incidents of explicit racial animus and bigotry that at times are acted out to levels of almost cartoonish iniquity. For her, racism is embodied in these kinds of discrete events. This is a theme that shows up in other IRR reports, including Jeffery's more recent reports.
Nevertheless, a notable survey result from research in that report undertaken telephonically in December 2017 by Victory Research for the IRR was that ‘77 percent of black respondents have never personally experienced racism directed against them’ (Jeffery 2018: 3), while 72 percent of respondents overall said the same (Jeffery 2018: 6). In this research, White people reported the highest levels of personal experience of racism, with 53 percent of Whites saying that they had personally experienced racism directed against them as compared to only 23 percent of Black people (Jeffery 2018: 6). This is how the IRR has produced research supposedly claiming that White people have experienced the most racism in South Africa since the end of Apartheid (see Jeffery 2018: 3).
In Jeffery's subsequent Race Relations in South Africa – Reasons for Hope 2019: Unite the Middle, the claim that Whites are the most racially discriminated-against group in South Africa since 1994 is given more self-reported data to support it (see Jeffery 2019b: 8). Furthermore, the case is made in this report that there is no evidence for there being ‘ubiquitous anti-black racism’ in South Africa (Jeffery 2019b: 3). Jeffery quotes her colleague John Kane-Berman, a Policy Fellow at the IRR, to make the point. Kane-Berman argues that by using complaints received by the Human Rights Commission for race-related complaints, and assuming that only 10 percent of ‘offenses’ would be reported, we could calculate ‘an (inflated) figure of [only] 13.3 racist incidents a day’ in South Africa (Jeffery 2019b: 3). This is argued to show that ‘even a tenfold inflated number of racist incidents pales into insignificance when measured against crime’ and that racial incidents defined in this way ‘almost disappear’ when we compare them to the population size of South Africa (Jeffery 2019b: 3; emphasis mine).
There are conceptual and political issues with viewing racism as an event or thinking that racism only arises when ‘racist incidents’, like those investigated by the Human Rights Commission, occur (see Modiri 2016). The conceptual issue, which is what I expand upon, is related to the relational definition of race that the IRR gives. We need to ask: What is the relational aspect of ‘relational racism’ to which the IRR refers? If part of the definition of ‘relational racism’ assumes that, in a relationship between individuals in a certain social context, one is dominant and the other is subordinate (if that is the relation), then the conceptualisation of racism as various ‘incidents’ does not capture the phenomenon being investigated. Racial subordination exists beyond the event of bigotry or insult, and even beyond the ‘occasional’ acts of discrimination that may have established or perpetuate the racial hierarchy. Even without a hierarchical framing of racism, if racism stems from prejudice and/or discrimination on the basis of the racial groups to which people are thought to belong, the existence of racism depends on a broader social order out of which racial identities and classifications come to be used in this way. People have to be seen to belong to race groups to be discriminated against on the basis of race, and that discrimination could be motivated by differential beliefs about – or differential dispositions towards – members of different race groups. This would mean, even under this very thin view of racism, that racist incidents are merely symptomatic of broader racial social relations.
Thus, to frame racism as events between individuals misses the socially situated and process-oriented aspects of racism. To put it another way: unlike some other kinds of interpersonal animosity, racism implicates social groups with racialised identities and practices that would need to have been established beforehand for anyone to be able to be racist to another person in the first place. Racism can result in events, such as suffering racial insult, prejudice, or discrimination – but even in a relational understanding of racism, that must be part of a broader racialised social process for racism as an event to be able to occur.
If the conceptualisation of racism as incidents or events is inadequate – or, at the very most, incomplete – this would call into question the validity of the IRR's surveys that pertain to racism. This would be because their concept of racism would lack validity by failing to properly characterise their topic of study. Why such validity – construct validity – is important is because it determines whether or not the data collected on a topic can be informative and accurate. A lack of construct validity undermines the integrity of the data collected on a subject of investigation. The IRR's understanding of racism as relational and pertaining only to incidents or events fails to adequately capture the phenomenon of racism and thus creates inherent limitations for any investigation about racism that they may undertake using their definition of racism.
Although there are many competing definitions of racism (e.g. see Zack 2021) and perhaps multiple ‘racisms’, the manner in which racism is conceptualised by the IRR and the manner in which they formulate their questions about racism reinforce a presumption about the nature of racism that they already have in mind, rather than reveal what their respondents may believe. Beyond the problem of concept validity, the IRR asks about racism in a way that biases responses. Asking respondents about what racist experiences they have had biases their responses to think of racism in terms of incidents or events, such as those Jeffery cites in her discussion of Kane-Berman's argument.
In line with previous and subsequent reports by the IRR, particularly those authored by Jeffery, it is concluded that the ‘IRR's 2018 field survey once again cuts through the increasing political “noise” around race to provide vital insights into what ordinary South Africans think on race-related issues’ (Jeffery 2019b: 12). What the data is said to reflect and what the IRR claims to have found is that ‘contrary to mainstream media perceptions and the hostile accusations often aired on Twitter and other social media, racism is not a major issue for most South Africans’ (Jeffery 2019b: 12). Jeffery's 2021 report, which has been the exemplar of this argumentative strategy in my discussion, also reflects this view using the data presented in Table 8 and Table 9 below.
Race relations since 1994 (#2 in Jeffery 2021)
| Most serious problem | Total | Black | Coloured | Asian | White |
|---|---|---|---|---|---|
| Improved | 41.7% | 42.8% | 45.0% | 33.9% | 31.6% |
| Stayed the Same | 26.4% | 26.6% | 20.2% | 34.9% | 27.6% |
| Became worse | 25.8% | 24.1% | 29.5% | 27.2% | 35.2% |
Have you personally experienced any form of racism over the past five years? (#3 in Jeffery 2021)
| Total | Black | Coloured | Asian | White | |
|---|---|---|---|---|---|
| Yes | 16.6% | 16.0% | 8.5% | 20.7% | 28.4% |
| No | 80.6% | 81.3% | 87.8% | 75.8% | 68.9% |
| Not answered | 2.8% | 2.7% | 3.7% | 3.4% | 2.7% |
Despite the confident declarations of racism not being a problem for most South Africans across reports produced by the IRR, the data that they present do not support this inference (see §2-1). Their surveys are not designed to be able to answer the question of whether racism is a major or important issue to South Africans; yet consistently in their analyses of their surveys they claim that their data shows that racism is not a major issue in South Africa. But the IRR has no way to ascertain the fact of the matter either way given how they have chosen to formulate the questions of their surveys. In fact, on review of the IRR's argumentative strategy in their research, it seems as if the design of their questions and their conceptualisation of racism are structured towards artificially producing the result that they desire: namely, their surveys and definitions are geared towards producing the result that ordinary South Africans do not think that racism is a major or serious problem in contemporary society (and that Whites are the greatest victims of racism in South Africa since 1994). The slanted design of the IRR's research raises the question: To what ends is the construction of this data designed to be put, including the interpretation and presentation of this data, and are these ends the purpose of the research's slanted design?
With all the problems with the IRR's surveys in mind, and with the knowledge of the outcomes these surveys consistently produce across a diversity of error types in the collection, presentation, and interpretation of data, we now move to thinking about what the purpose of the IRR's surveys, and the extensive political commentary that accompanies them, could be. The argument that I make is that these mistakes are not mere coincidences. I argue that the consistency of these errors across reports, and the consistency in the results or policies that these errors support across error types, suggests that these mistakes are strategically and systematically contrived to produce conclusions that are convenient to – and favourable for – the lobbying efforts of the IRR.
The Purpose of IRR Surveys is to Lobby for Specific Policies, Not to Seek Truth
The surveys in the IRR's reports foreground their political commentary and policy proposals that make up the more substantive parts of their reports. This may be because a central function of the IRR is its role as a political pressure or lobby group and think tank. It frequently makes submissions to parliament and lobbies the government on issues of affirmative action, employment equity, land redistribution, education policy across all levels of formal education; most recently the IRR has been arguing for individual citizens’ rights to own firearms, among other issues.9 Even outside of direct state lobbying, the IRR aims to have influence on the public through its engagement with the media, running its own ‘news’ outlets, producing publications such as the reports discussed, influencing school communities by going as far as having its members act as consultants or representatives on issues such as the purported dangers of Critical Race Theory (CRT) in schools, and appearing as expert witnesses in court cases pertaining to farm murders and free speech.10 This may explain why the most substantive parts of their reports are increasingly not data but commentary that is often tangential to the data that they present.
In Jeffery (2021), more than half of the report – pages 13–29 of a thirty-page report with references – is dedicated to giving political commentary on problems with affirmative action and ‘the dangers of CRT’. This part of the report is suggested to have support from the data of their field surveys, which have purportedly found that racism is or race relations are not a serious problem in South Africa. Their discussion of affirmative action is foregrounded by the issue of racism being framed as a ploy politicians use to get ahead and divert attention away from their own failures, but which is not a serious problem in the country, as is meant to be evidenced by South Africans’ experiences reflected in their surveys. As I demonstrated in §2, these claims are not supported by the data in their surveys. Nevertheless, these unsubstantiated claims are the foundation upon which the IRR build their case and policy recommendations about race-based policy and CRT.
The reports of the IRR are geared towards and lead up to them giving their policy recommendations about how South Africa should move forward. The dismal failures of South African instantiations of corrective race-based policy such as Black Economic Empowerment (BEE) legislation and the IRR's rejection of Employment Equity (EE) policy are discussed in order to propose the IRR's ‘Economic Empowerment for the Disadvantaged or EED’ strategy (Jeffery 2021: 13–20). Jeffery points out that South Africa's affirmative action policies have failed to uplift the majority of the population that they are meant to benefit. These policies have rather enriched a small elite and/or politically connected proportion of that population. These failures are well-documented, and area-specific studies, such as on land reform (e.g. Mtero et al. 2019), reflect this general trend. These and other issues with South Africa's race-based policies are discussed in tandem with corruption, ‘cadre deployment’, and fraud as part of a string of the ANC government's developmental failures. The discussion of these distinct issues together could function to bring further doubt to the reader's mind about the legitimacy of race-based policy, a rhetorical technique that looks intentional given the IRR's association of affirmative action with unfairness and corruption (e.g. cf. Jeffery 2021: 17). Although the discussion is focused on EE and BEE legislation, the argument the IRR makes is against affirmative action and race-based policy more generally. Drawing on and quoting controversial conservative thinkers like the US economist Thomas Sowell, the case is made that affirmative action can never work in principle. Jeffery, explaining why she believes affirmative action is untenable, says:
The reason is simple: people are not ‘blank slates’ who are all inherently the same and can therefore be slotted into any role. Instead, individuals differ markedly from one another in terms of their ages, education levels, aptitudes, skills, experiences, and personality attributes (Jeffery 2021: 13).
For this to make affirmative action untenable, the differences between individuals need to reflect on the race-group level, with a racially differentiated distributions of these traits. This would mean that it is not just that individuals differ in these ways, but that race groups differ in these ways as well, in such a way as to make affirmative action untenable in principle. The implication of this quote – which is essentially the argument of The Bell Curve (Hernstein and Murray 1994; cf., Gould 1996) – is that we should not expect or enforce representative racial distributions because individuals of different races differ from one another on the basis of a number of traits that Jeffery enumerates. I do not assess whether this is a racist claim; rather, I assert that, for Jeffery to make such a bold claim, more substantive evidence about the distribution of traits among individuals and across race groups (and identifying the reasons for these distributions) is needed. For instance, those who argue that racism is still an issue provide empirical evidence of the role that racism continues to play in producing inequitable social outcomes in the economy (e.g. Gradín 2019; Sulla et al. 2022). To counter these kinds of claims, counter-evidence or a plausible (re-)interpretation of the available evidence that undermines the causal role that racism plays in creating racially inequitable outcomes would need to be presented. To say that individuals are different and then to imply that race groups are also different in such a way as to make affirmative action untenable is presumptuous, to say the least, and is not supported by any evidence presented by the IRR.
It is against this backdrop – the denial of racism as a serious problem and the purported untenability of affirmative action policy – that it is argued that CRT is an ideologically misled view of race relations and a perspective about the world that is dangerous and damaging. Jeffery says that CRT is a form of Marxist political ideology (Jeffery 2021: 27) that ‘declines to treat people as individuals’ and instead sees ‘them solely as representatives of their socially constructed racial groups’ (Jeffery 2021: 23). Jeffery claims that CRT ‘therefore has no basis on which to recognise people's differing strengths and weaknesses’ (Jeffery 2021: 23) – a failure to recognise this across individuals (and, by implication, race groups) that makes affirmative action policies untenable.
Jeffery brings together her case at the end of the report by arguing that the South African public do not want affirmative action policy, as is purportedly evidenced by IRR surveys. Comparing complimentary IRR polls on Employment Equity with those of the Helen Suzman Foundation, Jeffrey claims that they show that ‘public opinion has . . . remained remarkably consistent in rejecting the EE Act's key demand [for fair racial representation in the workplace] for more than 25 years’ (Jeffery 2021: 10). She claims that affirmative action policy only remains policy in South Africa because of the influence of CRT and CRT-like ideology.11 Jeffery argues that although ‘most black South Africans . . . have repeatedly and consistently identified unemployment, crime, corruption, and poor housing as far more pressing challenges, CRT ideology nevertheless lends some credence to the ANC's view’ that ‘racism is South Africa's most urgent and pervasive problem’ (Jeffery 2021: 28).
The idea that racism remains a serious problem in South Africa is attributed to politicians since it is claimed that only about 3 percent of the South African public believe that racism is a serious problem. Moreover, the view that racism remains a serious problem is presented as a way in which politicians use race to divert attention away from their own failings. These claims are drawn from IRR survey data and analyses (see §2). The surveys serve the function of supporting the IRR's case against race-based policies and the belief that racism remains a serious problem in the country (except against Whites, according to their research). The survey data is meant to support the view that the government plays the race card and, by so doing, directs attention away from more serious problems like the failure of the government. According to the IRR, racism is a problem which very little of the population (except Whites) is affected by or cares about.
The point of the IRR's surveys is to depict race-based policy as illegitimate and having very little public support. Their surveys are meant to give credence to the view that the IRR represents what the public actually want (despite what politicians say and argue for). But support for these claims by the IRR is not grounded in evidence. Their views are not supported by good empirical or social-scientific work, not least by survey data, despite what they claim, as I demonstrated in §2. This suggests that the sole purpose of the IRR's surveys, since they have no scientific value, is to serve as a useful lobbying tool for their proposed policies, like those against affirmative action and other race-based policies. The purpose of the IRR's research in this regard is not to present us with accurate information about race relations and policy issues in South Africa; rather, the function of their research is purely political and ideological, although it pretends to be presenting the public with ‘objective’ scientific findings. The IRR's surveys do not meet basic scientific and ethical research standards: they do not accurately represent the views of the South African public and they strategically mislead stake holders about the state of race relations in the country. This is not good.
Concluding Remarks: The IRR as False Witness
The IRR argues that their survey data shows that the general public do not believe that racism remains a serious problem in South Africa. They claim that the idea that racism is a serious social issue that we (still) have to tackle is a view promoted by politicians who use the issue of racism to deflect attention away from their own failures.
I showed how these conclusions are misleading interpretations of their surveys – conclusions that are not supported by and do not follow from their own data. Moreover, I discussed how the leading, loaded, and ambiguous questions that the IRR asks cannot be interpreted accurately or meaningfully because of their form and how they confound variables. As a consequence, the links made by the IRR between views about racism, the public's supposed lack of support for affirmative action or race-based policy, and the influence of CRT on South African politics are all conclusions not grounded in the evidence that they present us with. The strategic errors in the IRR's surveys and their accompanying commentaries demonstrate that the IRR is not a reliable source of information about contemporary race relations and policy in South Africa.
Moreover, I have argued that the manner in which the IRR misinforms the public and policy stakeholders about race relations, particularly about racism, is motivated and misleading in ways that make it damaging to democratic processes. I showed how the IRR's surveys are used to attribute specific views to demographic groups which the evidence does not show that they hold. In this regard, I highlighted the danger that the IRR poses to deliberative democratic processes in its reporting and dissemination of disinformation about public opinion. It is because disinformation undermines the purpose of public deliberation in democratic processes that I have argued that the IRR is a danger to South African democracy in the way that it strategically misleads us about race relations, racism, and public policy.
Acknowledgements
I thank Heather Brookes, Roger Southall, the editors of this journal, and two anonymous reviewers for their useful feedback on this manuscript, along with Zakhele Mthembu and Dave Martin for their probing sceptical questions about my criticisms of the IRR on social media where I alluded to some of the core ideas of this article. It was through these engagements and those with lobbyists of the IRR who challenged my public criticisms of their work that I was eventually motivated to write this article. I thank the FirstRand Foundation for their support as a recent FirstRand Scholar and Macquarie University as a recipient of their International Macquarie University Research Excellence Scholarship (iMQRES). The views expressed in my research do not necessarily represent the position of any funder, research team, organisation, or institution with which I am affiliated. The work in this article is a continuation of the ‘scholactivism’ I have been engaged in as an alumnus of the Canon Collins Educational & Legal Assistance Trust. I thank Patty A. Gray for proof-reading the text for this article. Any errors that remain are my own.
Notes
The open letter can be found on the Daily Maverick website (https://www.dailymaverick.co.za/article/2021-09-17-irrs-current-approach-betrays-the-legacy-of-its-founders-and-does-a-disservice-to-the-people-of-south-africa/, accessed 17 September 2021). Although I was involved in the committee that formulated this open letter, I could not sign it due to my reservations about ‘rescuing’ the IRR or pleading that it return to its liberal roots, given the oppressive history of liberalism in South Africa and the IRR's present-day networks. Although the IRR of today is significantly different from the IRR of the past, it does not mean that the historical IRR was not problematic on several fronts in ways which could have affected its data and research. For an introduction to the oppressive history of liberalism in South Africa, see Eddy Maloka's book Friends of the Natives: The Inconvenient Past of South African Liberalism (Maloka 2014; cf. Dladla 2015) and, particularly on the early IRR and its liberal roots, see Anjuli Webster's thesis Racial Liberalism and the South African Institute of Race Relations, 1929–1950 (Webster 2018).
Numerous defences of the IRR's surveys have been mounted publicly by the IRR and its associates. For instance, Gabriel Crouse, Head of Campaigns for the IRR who recently authored their report against #BLM (see Crouse 2020), claims that IRR's data ‘integrity is transparently robust’ (see: https://irr.org.za/media/letter-roger-southall-invited-to-2018throw-off-his-jejune-anti-irr-beret2019-businesslive, accessed 30 October 2021); and Terence Corrigan, the Project Manager of the IRR, makes a case for the trustworthiness of IRR data through comparisons of IRR data with data collected by other agencies (see: https://dailyfriend.co.za/2021/10/29/polling-and-politics/, accessed 30 October 2021). Using the IRR's surveys as an example, I show how the IRR's data is neither transparent nor robust but is, rather, seriously flawed in ways that make its data inherently misleading and inaccurate (see discussion in §2), thus undermining claims made by the likes of Crouse and Corrigan about the integrity of IRR data.
The IRR has a strong web presence: it hosts its publications online; it has official and affiliated websites for its various projects to influence the public and policy; it has IRR and IRR-affiliated audio-visual YouTube channels such as the Big Daddy Liberty Show, hosted by Sihle Ngobese (Head of Strategic Operations at the IRR); it has podcasts such as 2 Crickets in A Thorn Tree, hosted by Nicholas Lorimer and Gabriel Crouse; and it has its own ‘online newspaper’ (not registered with the Press Council Ombudsman), deputy-edited by Marius Roodt, called the Daily Friend (their website does not name their editor at present), whose purpose is to disseminate content advocating for ‘classical liberalism as an effective way to defeat poverty and tyranny through a system of limited government, a market economy, private enterprise, freedom of speech, individual liberty, property rights, and the rule of law’ (https://dailyfriend.co.za/about/, accessed 16 September 2021).
It would be of interest to measure the effect that IRR materials have on influencing policy position support and how much they sway opinion on claims about racism in SA. Such an experiment is beyond the scope of this article, but this article provides motivation for such research.
In an interview with Stephen Grootes for 702 Radio on 8 February 2017, Ngwenya admits that they, as the IRR, ‘do concede, however, that the survey is slightly biased towards relational racism, which is racism based on your daily interactions as compared to taking a look at institutional or systemic racism’. You can find this interview here: https://www.702.co.za/articles/244114/over-70-of-south-africans-do-not-experience-daily-racism-sairr-survey (accessed 16 September 2021).
This is a claim often repeated in the press as a finding of the IRR from their national surveys. This claim dominated stories in the press about IRR's 2017 report by Jeffery (e.g. https://businesstech.co.za/news/business/155443/only-3-of-south-africans-consider-racism-a-serious-problem-irr/, accessed 30 October 2021) and is a common motif in the use of their data in submissions to parliament or public ‘debates’ against race-based policy. Their reporting of Whites being the most affected victims of racism since 1994 (see §2-5 for a discussion) is also used to support the myth of White genocide in South Africa.
It may be of interest to note that more White people disagreed with this statement than any other group, even if marginally.
A reviewer worries that this – the ability of policy makers to make evidence-based decisions based on the facts of a matter rather than public opinion – ‘undermines the paper's larger critique of the IRR (as reflected in the title), that the IRR's misrepresentation of public opinion poses a threat to deliberative democracy’. This is not true, as at least two problems would remain despite the policy maker's decision: first, the policy maker would be misinformed about what it is that the people want or support, and second, the misinformation about public opinion would still have a negative epistemic effect on deliberative processes and their outcomes. The epistemic environment in which they exist would have been compromised. This could create further damage to the democratic good of the responsiveness of policy makers to public opinion and to the legitimacy of the policy making process. The misrepresentation of public opinion also has the negative effect of misleading us about what the people want from their leaders or government. A lot remains at stake even if misinformation is not taken up by decision makers, especially if this misinformation spreads in society and causes conflict with the policy decisions leaders and government representatives end up taking. An example of these stakes is the public fallout and conflict between government and citizens due to misinformation about evidence pertaining to the COVID-19 pandemic (Kim et al. 2020). Misinformation can be dangerous and damaging even if (or even when) it is not reflected in policy decisions, as it can be the basis of conflict and doubts about the legitimacy of policy and the process of the creation and implementation of policy.
John Endres, the IRR's CEO Elect, claims in a recent newspaper article: ‘Over the past seven years, the IRR has produced over 450 policy submissions and policy reports, covering topics ranging from property rights to freedom of speech and electoral democracy, from water policy to electricity provision, consumer spending, demographic trends, education and health policy, communication, crime, corruption, unemployment, empowerment, women's rights, gun policy, gay rights, economic reform, the minibus taxi industry and much else besides’ (published online on 21 September 2021 on News24: https://www.news24.com/news24/columnists/guestcolumn/right-of-reply-john-endres-the-irr-holds-the-liberal-line-against-the-left-20210921; accessed 24 September 2021). In the IRR's 91st Annual Report 2020, the IRR claims: ‘The Institute's approximately 30 staff members produced and placed no fewer than 85 reports, 6 policy submissions, 1 214 opinion pieces, 123 media releases, 540 media interviews, 188 briefings, 27 events, 556 audio and video clips, and 2 mini-documentaries in the public domain in 2020’ alone (South African Institute of Race Relations 2021: 12). The IRR is prolific and works with other organisations within national and international networks with similar ideological outlooks and goals on their campaigns, legal cases, submissions, and dissemination of ideas to the public. There is a ‘revolving door’ between these organisations, e.g.: the IRR's Centre for Risk Analysis consultancy established by former IRR CEO Frans Cronje, of which current IRR CEO Endres is Director; the Freedom Advocacy Network whose Director, Hermann Pretorius, is the head of Strategic Initiatives at the IRR; African Liberty – which describes itself as ‘a project of the Cato Institute and IMANI Africa, thereafter the Atlas Network’ – which has former IRR CEO Cronje on their advisory board; the leadership of the Democratic Alliance, whose Federal Chairperson Helen Zille was a Senior Policy Fellow at the IRR; the Free Market Foundation, which has IRR Council member Martin van Staden and IRR Head of Campaigns Chris Hattingh on their Board of Non-Executive Directors, etc. There is a substantial history of collaboration across these organisations (some of which are sub-organisations or ‘brands’ of the IRR) and their affiliates. This collaboration is apparent, for example, in both Zille and Ngwenya's various leadership roles at different times in the Democratic Alliance and the IRR (e.g. Ngwenya was a COO of the IRR and the Head of Policy for the DA), or in the IRR Council member Adv. Mark Oppenheimer representing AfriForum in court whilst AfriForum assists and supports the IRR in some of its campaigns, such as resistance against land expropriation, racial legislation, and gun control. There historically has been some controversy in the media about the relationship between the IRR and AfriForum related to the fact that ‘AfriForum has in the past paid the IRR to conduct research’ and that ‘AfriForum and AfriSake, another AfriForum offshoot, are listed as ‘sponsors and donors’ in the IRR's annual reports of 2015 and 2016’, which Cronje later claimed was done in error (see: https://www.news24.com/News24/verwoerd-documentary-must-retracted-urges-irr-afriforum-says-neewat-20190313, accessed 27 September 2021). This, at the very least, suggests financial ties, ties between personnel, and a history of continued collaboration between these organisations. These networks merit further investigation but are beyond the scope of this article.
See their anti-CRT website ‘Educate, don't indoctrinate’ https://edonti.org/ (accessed 16 September 2021). It is notable that their media tab about CRT at the time of writing comprises of mostly US conservative media content and carries very little content about the purported situation in South Africa. The US conservative media slant of their website is suggestive of their orientation and concerns. For an example of how members of the IRR are called as expert witnesses in court cases, see discussion by Eusebius McKaiser about the recent AfriForum vs EFF court case in which Crouse was called to the stand by Adv. Mark Oppenhimer who was council for AfriForum in this case (see: https://iono.fm/e/1219678, last accessed 1 September 2022) and see comments in the judgment passed against AfriForum in the same case (see: https://www.politicsweb.co.za/documents/afriforum-vs-eff-malema--ndlozi-the-high-court-jud, last accessed 1 September 2022).
Jeffery links CRT to socialism and the antiquated idea that the South African government, through the African National Congress (ANC) and its alliance partners such as the South African Communist Party (SACP), are engaged in a ‘National Democratic Revolution’ – a claim Jeffery makes in this report (Jeffery 2021, 19; 28) and has repeated in her books (Jeffery 2019a; 2009), but which political scientists and historians find no evidence for (Ngqulunga 2020; Gumede 2009). She presents a revisionist account of the transition to democracy that runs counter to the evidence uncovered by historians who have lived through, studied, and written about the period from primary sources (e.g. Aitchison 2015; Mchunu 2021). It is said of her account that it ‘is simply not true’ (Gumede 2009, 56) and that ‘her case is undone by her ideological passion and fever’ since ‘she overlooks, avoids and/or dismisses information and evidence that runs counter to her thesis’ (Ngqulunga 2020, 152). I find this a general feature of Jeffery's work: she promotes an ideological stance that is impervious to countervailing evidence. It is this same dogged ideological commitment that is apparent in her reports for the IRR, particularly in her analysis and interpretation of survey data. But, as I have shown in §2, the problems with the IRR's research extend beyond the person of Jeffery. Rather, these problems are inherent in the way the IRR conducts its research.
References
Aitchison, J. 2015. Numbering the Dead: The Course and Pattern of Political Violence in the Natal Midlands, 1987–1989. Pietermaritzburg: The Natal Society Foundation.
Chetty, R., N. Hendren, M.R. Jones, and S.R. Porter. 2020. ‘Race and Economic Opportunity in the United States: An Intergenerational Perspective’, The Quarterly Journal of Economics 135 (2): 711–783. https://doi.org/10.1093/qje/qjz042.
Crouse, G. 2020. ‘Because #BlackLivesMatter: What Institutions Must Know About the BLM Global Network’ (Report). Johannesburg: South African Institute of Race Relations. https://irr.org.za/reports/occasional-reports/files/because-blm-report-final.pdf.
Dixon, G.N., and C.E. Clarke. 2013. ‘Heightening Uncertainty Around Certain Science’, Science Communication 35 (3): 358–382. https://doi.org/10.1177/1075547012458290.
Dladla, N. 2015. ‘Friends of the Natives: The Inconvenient Past of South African Liberalism’, New Voices in Psychology 11 (1): 138–149. https://doi.org/10.25159/1812-6371/1810.
Emmons, W.R., and B.J. Noeth. 2015. ‘Why Didn't Higher Education Protect Hispanic and Black Wealth?’ In the Balance: Perspectives on Household Balance Sheets 12: 1–3. https://www.stlouisfed.org/-/media/project/frbstl/stlouisfed/publications/in-the-balance/images/issue_12/itb_august_2015.pdf
Gradín, C. 2019. ‘Occupational Segregation by Race in South Africa after Apartheid’, Review of Development Economics 23 (2): 553–576. https://doi.org/10.1111/rode.12551.
Gould, S.J. 1996. The Mismeasurement of Man. New York: WW Norton & Company, Second Edition.
Gumede, W. 2009. ‘Review: People's War: New Light on the Struggle for South Africa’, Focus 55: 56–57.
Hernstein, R.J. and Murray, C. 1994. The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.
Iftikhar, I., R. Ullah, N. Naureen, and H. Ali. 2016. ‘Deliberative Democracy: Effect of News Media and Interpersonal Conversation on Quality of Public Opinion’, South Asian Studies 31 (1): 43.
Jeffery, A. 2009. People's War: New Light on the Struggle for South Africa. Johannesburg: Jonathan Ball Publishers.
Jeffery, A. 2018. Race Relations in South Africa – Reasons for Hope 2018: Holding the Line. South African Institute of Race Relations.
Jeffery, A. 2019a. People's War: New Light on the Struggle for South Africa. Johannesburg: Jonathan Ball Publishers.
Jeffery, A. 2019b. Race Relations in South Africa – Reasons for Hope 2019: Unite the Middle. South African Institute of Race Relations.
Jeffery, A. 2021. ‘Critical Race Theory & Race-Based Policy: A Threat to Liberal Democracy’, @LIBERTY: The Policy Bulletin of the IRR 1/2021 (49): 1–30.
Kim, H.K., J. Ahn, L. Atkinson, and L. Kahlor. 2020. ‘Effects of COVID-19 Misinformation on Information Seeking, Avoidance, and Processing: A Multicountry Comparative Study’, Science Communication 42 (5): 586–615. https://doi.org/10.1177/1075547020959670.
Koehler, D.J. 2016. ‘Can Journalistic “False Balance” Distort Public Perception of Consensus in Expert Opinion?’ Journal of Experimental Psychology: Applied 22 (1): 24–38. https://doi.org/10.1037/xap0000073.
Maloka, E. 2014. Friends of the Natives: An Inconvenient History of Liberalism in South Africa. Durban: 3rd Millenium.
Mchunu, M. 2021. Violence and Solace: The Natal Civil War in Late-Apartheid South Africa. Charlottesville, VA: University of Virginia Press.
Meintjes, S. 2020. In the Shadow of the Great White Queen: The Edendale Kholwa of Colonial Natal, 1850–1906. Pietermaritburg: The Natal Society Foundation. http://natalia.org.za/NSF_books/In The Shadow of the Great White Queen.html.
Modiri, J. 2016. ‘The “Event” of Racism’, The Daily Maverick 17 November. https://www.dailymaverick.co.za/opinionista/2016-11-17-the-event-of-racism/.
Mtero, F., N. Gumede, and K. Ramantsima. 2019. ‘Elite Capture in Land Redistribution in South Africa’, PLAAS Research Report No. 55. Institute for Poverty, Land and Agrarian Studies. http://hdl.handle.net/10566/5089.
Ngqulunga, B. 2020. ‘People's War: New Light on the Struggle for South Africa by Anthea Jeffery’, Transformation: Critical Perspectives on Southern Africa 104 (1): 146–153. https://doi.org/10.1353/trn.2020.0040.
Peels, R., J. De Ridder, and R. van Woudenberg (eds). 2020. Scientific Challenges to Common Sense Philosophy. New York: Routledge.
South African Institute of Race Relations. 2021. 91st Annual Report, 1 January to 31 December 2020. Braamfontein: South African Institute of Race Relations.
Schmitt Beck, R. 2015. ‘Bandwagon Effect’, The International Encyclopedia of Political Communication, 1–5. https://doi.org/10.1002/9781118541555.wbiepc015.
Schumacher, G. and P. Öhberg. 2020. ‘How Do Politicians Respond to Opinion Polls? An Experiment with Swedish Politicians’, Research and Politics 7 (3). https://doi.org/10.1177/2053168020955106.
Solomon, S., and J. Abelson. 2012. ‘Why and When Should We Use Public Deliberation?’ Hastings Center Report 42 (2): 17–20. https://doi.org/10.1002/hast.27.
Statistics South Africa. 2021a. ‘Quarterly Labour Force Survey, Quarter 1: 2021’. June 2021. http://www.statssa.gov.za/publications/P0211/P02111stQuarter2021.pdf.
Statistics South Africa. 2021b. ‘Quarterly Labour Force Survey, Quarter 2: 2021’. August 2021. http://www.statssa.gov.za/publications/P0211/P02112ndQuarter2021.pdf.
Sulla,V., P. Zikhali, P.F. Cuevas, 2022. Inequality in Southern Africa: An Assessment of the Southern African Customs Union (English). Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/en/099125303072236903/P1649270c02a1f06b0a3ae02e57eadd7a82.
Toff, B. 2018. ‘Exploring the Effects of Polls on Public Opinion: How and When Media Reports of Policy Preferences Can Become Self-Fulfilling Prophesies’, Research & Politics 5 (4): 205316801881221. https://doi.org/10.1177/2053168018812215.
Tomić, T. 2013. ‘False Dilemma: A Systematic Exposition’, Argumentation 27 (4): 347–368. https://doi.org/10.1007/s10503-013-9292-0.
Watts, D.J., D.M. Rothschild, and M. Mobius. 2021. ‘Measuring the News and its Impact on Democracy’, Proceedings of the National Academy of Sciences of the United States of America 118 (15): 1–6. https://doi.org/10.1073/pnas.1912443118.
Webster, A. 2018. ‘Racial Liberalism and the South African Institute of Race Relations, 1929-1950’. MA thesis, University of Dar es Salaam.
Zack, N. 2021. ‘Alberto G. Urquidez, (Re-)Defining Racism: A Philosophical Analysis, Palgrave Macmillan, 2020, Viii +421 Pp. the Concept of Racism and the Adjective Racist’, Ethical Theory and Moral Practice 24 (3): 673–677. https://doi.org/10.1007/s10677-021-10163-x.