Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Recent research suggests the validity of the construct of Left-wing Authoritarianism (LWA). Like its well-studied parallel construct Right-wing Authoritarianism, LWA is characterized by dogmatism, punitive attitudes toward dissenters, and desire for strong authority figures. In contrast to RWA, LWA mobilizes these traits on behalf of left-wing values (e.g. anti-racism, anti-sexism, and wealth redistribution). I inductively examined the extent to which RWA and LWA predicted, in April 2020, Americans' endorsement of 19 authoritarian policies and practices intended to mitigate the effects of the COVID-19 pandemic. For 11 of these policies (e.g. abrogating the right to trial by jury for pandemic-related crimes), both RWA and LWA independently positively predicted endorsement. These findings are consistent with recent work showing psychological similarities between the two constructs.
Keywords: Right-wing Authoritarianism, Left-wing Authoritarianism, COVID-19In March 2020, the COVID-19 pandemic generated concern that authoritarian leaders around the world would seize even more power, and that leaders who had hitherto adhered (mostly) to liberal democratic norms would take advantage of the crisis to usurp power and infringe upon human rights (Funk & Linzer, 2020; Gebrekidan, 2020). Although most commentators and national publics agreed that the emergency justified certain temporary authoritarian measures (e.g. restrictions on travel and assembly), many feared that some leaders would hold on to these powers after the emergency had passed, and that they would use them as pretexts to harass or neutralize their political opponents. Putatively pandemic-mitigating authoritarian measures included rule by decree, intrusive surveillance, and expanded powers to detain criminal suspects.
Straightforwardly, within-country variation in support for authoritarian policies is partly driven by variation in authoritarianism as an individual difference dimension. Right-wing Authoritarianism (RWA) was long considered a personality trait (Adorno, Frenkel-Brunswick, Levinson, & Sanford, 1950; Altemeyer, 1981) characterized by conventionalism (adherence to conventional values), authoritarian submission (placing high value on obedience and respect for authority), and authoritarian aggression (punitive attitudes toward those who deviate from conventional values). More recent work has modified or enlarged this conceptualization. For example, although RWA is related to personality, it is probably better regarded as a set of related beliefs or attitudes than as a personality trait (Duckitt, 1989). Most relevant to the present study, it remains a matter of controversy whether a parallel construct, Left-wing Authoritarianism (LWA), is valid and useful (Altemeyer, 1996; Conway III, Houck, Gornick, & Repke, 2017; Costello, Bowes, Stevens, Waldman, & Lilienfeld, 2020; Costello & Lilienfeld, 2019; Costello, Stevens, & Lilienfeld, 2020; Mullen, Bauman, & Skitka, 2003). This debate is part of a broader discussion regarding whether psychologists have exaggerated the cognitive and personality differences between right- and left-wingers (Clark & Winegard, 2020).
Some psychologists have dismissed LWA as a “myth” (Stone, 1980), comparable to the Loch Ness Monster (Altemeyer, 1996). Conway III et al. (2017) postulated that LWA resembles RWA in being characterized by high levels of dogmatism, punitive attitudes toward dissenters, and desire for strong authority figures. What distinguishes LWA from RWA is the content of the dogmatically defended values. For individuals high in LWA, these values include pro-environment and anti-racist and anti-sexist beliefs, and rejection of traditional Christian moral restrictions. This model was supported with data showing associations of a new LWA instrument with measures of dogmatism, prejudice, and attitude strength. More recently, Costello and colleagues (Costello, Bowes, et al., 2020; Costello & Lilienfeld, 2019; Costello, Stevens, & Lilienfeld, 2020) created and validated a longer and more wide-ranging LWA instrument and explored its factor structure, nomological network (compared to RWA), and external validity. A three-factor structure (Anti-hierarchical Aggression, Left-wing Conventionalism, and Prohibition of Protected Speech) yielded the best fit. LWA's nomological network shared numerous features with RWA's nomological network, including high levels of dogmatism, affective polarization, and moral disengagement. Results from a modified Tangram Help/Hurt task showed that LWA, over and above political ideology, predicted behavioral aggression toward political opponents (Costello, Bowes, et al., 2020).
Thus, a growing body of evidence supports the validity and predictive power of the construct of LWA. The current study addresses the question: To what extent do RWA and LWA predict Americans' endorsement of authoritarian policies intended to mitigate the effects of the COVID-19 pandemic? Finding that RWA predicts support for authoritarian policies would not be very newsworthy. However, given the ongoing controversy regarding the LWA construct (Costello, Bowes, et al., 2020), the current study serves as a timely test of its predictive usefulness. Importantly, most proposed authoritarian policies to mitigate the COVID-19 pandemic fall outside the domains directly covered by LWA-associated values (race, gender, wealth redistribution). Thus, this study affords the opportunity to extend the nomological network of LWA to incorporate issues of great public concern at a particular historical moment.
Authoritarianism as a general set of attitudes includes dogmatism, a preference for conformity, willingness to coercively enforce behavioral standards, punitiveness toward perceived enemies, and a strong concern with hierarchy (Costello, Bowes, et al., 2020). However, this characterization does not in itself provide a complete means of identifying what constitutes an authoritarian policy. Such judgments are colored by political ideology and other factors. For example, a recent proposal (Bartholet, cited in O'Donnell, 2020) to ban homeschooling in the U.S. described the practice as authoritarian, whereas proponents of homeschooling regard banning it as authoritarian. I asked participants to indicate their level of endorsement of a wide range of putatively pandemic-mitigating policies, many of which would undoubtedly be judged as “authoritarian” by only a minority of Americans. These policies were mostly drawn from COVID-related news and commentary of April 2020. I also considered whether a proposed or implemented policy could be viewed as violating components of the U.S. Constitution, particularly Article I and the First, Second, Fourth, Fifth, Sixth and Fourteenth Amendments.
The range of policies selected means that some expectations of the current study's findings are obvious. For example, RWA will be positively associated with endorsement of closing abortion clinics. During April 2020, conservatives were less likely than liberals to abide by social distancing restrictions (Rothgerber et al., 2020). Therefore, to the extent that RWA is associated with conservatism and LWA is associated with liberalism, the former will be negatively associated, whereas the latter will be positively associated, with endorsement of these restrictions. Importantly, however, the current study concerns the predictive power of RWA and LWA, not of mainstream liberalism or conservatism. Most interestingly, it is not at all obvious how RWA and LWA will differ (if at all) with respect to their associations with endorsement of rule-making by executive decree, legal restrictions on pandemic-related speech and on the right to protest government actions, or encroachments on (Anglo-American) due process rights for individuals accused of crimes. My analyses of these associations are exploratory; no a priori hypotheses were tested.
Furthermore, the current study does not address the relationship between political beliefs and sensitivity to threat, nor does it explore other potential motivational links between authoritarianism and reactions to the pandemic. A considerable body of evidence indicates that conservatives are more sensitive than liberals to threat, particularly disease threat (see Conway III, Chan, & Woodard, 2019 for review), raising the puzzle of why, in the current pandemic, conservatives have been more sanguine than liberals about the danger it poses (Conway III, Woodard, Zubrod, & Chan, 2020). Variation in COVID-19 threat appraisal might be driven by variation in RWA or LWA (as found for mainstream conservatism and liberalism by Conway III et al., 2020), but that question is beyond the scope of the current study.
To summarize, this study examined associations between RWA and LWA (predictors) and endorsement of 19 putatively authoritarian pandemic-mitigation policies (outcomes). I incorporated five covariates into my analyses. With respect to policies mandating and enforcing social distancing, individuals whose incomes are being more strongly reduced by these restrictions might disapprove of them more than individuals whose incomes are less affected, irrespective of political beliefs (but see Conway III et al., 2020). Four variables might confound associations between RWA and/or LWA and endorsement of putatively authoritarian responses to COVID-19: gender, age, local COVID-19 prevalence, and African-American identity (because African-Americans have suffered higher rates of COVID-19 mortality than other Americans [APM Research Lab, 2020]).
I recruited 550 U.S. resident participants using Prolific.co, an online survey service with higher minimum payments to participants than is typical of the more widely used Amazon Mechanical Turk. All participants were paid $1.90. I excluded one participant for showing signs of bogus responding, one for indicating an age of 2, 10 for declining to provide a ZIP code, and 10 for indicating a gender other than “male” or “female” in answer to the free-response question.
Among the 528 participants comprising the final sample, 313 (59.3%) were female. Participants' ages ranged from 18 to 79 (M = 31.8; SD = 10.4; Med = 30). The sample's distribution of ethnic identities, household income levels, and educational attainment levels are shown in Supplementary Table S1.
Informed consent was obtained from all participants in accordance with the terms of UCLA IRB Approval #20-000738.
All participant responses were collected on April 22, 2020. After indicating consent, participants were instructed that they would first be asked 40 questions about their general political opinions, after which they would be asked questions about their views on how best to respond to the COVID-19 crisis specifically. Participants then completed the 18-item short form of the ACT (Authoritarianism-Conservatism-Traditionalism) Scales (Duckitt, Bizumic, Kruauss, & Heled, 2010) and the 22-item short form of the Left-Wing Authoritarianism Index (Costello & Lilienfeld, 2019). Which of these two instruments was presented first was randomized. Each of the three ACT scales comprises six items. Half the items are reverse-keyed. Examples of items include Our prisons are a shocking disgrace. Criminals are unfortunate people who deserve much better care, instead of so much punishment (Authoritarianism, reverse-keyed), Our country will be great if we show respect for authority and obey our leaders (Conservatism), and God's laws about abortion, pornography, and marriage must be strictly followed before it is too late (Traditionalism). Because the LWA Index is not yet published in a peer-reviewed journal, I include a complete list of its short-form items as Table S2A. An example is Bigots must be taught to shut-up and stay in their place. Items within each instrument were presented in randomized order. Participants responded on 7-point Likert scales anchored by “Strongly Disagree” and “Strongly Agree.”
Participants were then presented (in randomized order) with 19 statements regarding pandemic-related policies, or policy-relevant actions. They were asked to indicate their endorsement of each item on a 7-point Likert scale anchored by “Strongly Disagree” and “Strongly Agree.” These items, and their descriptive statistics, are listed in Table 1 . For every item, responses ranged from 1 to 7. One policy item, Government policies to reduce the spread of COVID-19 should rely more on persuasion than on enforcing rules about social distancing, can be regarded as contra-authoritarian.
Policy endorsement descriptive statistics (7-point Likert scales, 1 = strongly disagree, 7 = strongly agree).
| Policy item | N | M ± SD | Median | Skew | Kurtosis |
|---|---|---|---|---|---|
| I would report it to the police if I saw someone violating social distancing rules. | 528 | 3.2 ± 1.9 | 3 | 0.39 | 2.02 |
| Government policies to reduce the spread of COVID-19 should rely more on persuasion than on enforcing rules about social distancing. | 528 | 3.9 ± 1.7 | 4 | −0.00 | 2.06 |
| People can't be trusted to follow social distancing guidelines unless they are threatened with punishment. | 527 | 4.1 ± 1.9 | 5 | −0.27 | 1.86 |
| I support the idea of an official federal government certificate that would be issued to people with proven immunity to COVID-19. | 527 | 4.2 ± 1.8 | 4 | −0.36 | 2.28 |
| I support the idea of requiring all smartphones to run an app that tracks a person's movements and notifies them when they are near a person infected with COVID-19. | 528 | 3.1 ± 2.0 | 3 | 0.52 | 1.94 |
| For as long as the COVID-19 pandemic continues. | |||||
| . Americans need to follow government orders about social distancing, even if they disagree with them. | 526 | 6.1 ± 1.4 | 7 | −1.81 | 6.09 |
| …governments should have the power to prohibit the spread of misinformation about COVID-19 that could endanger people if it were widely believed. | 527 | 5.5 ± 1.7 | 6 | −1.19 | 3.62 |
| …government officials should put churches under surveillance to make sure that they're not holding services that violate social distancing rules. | 527 | 4.4 ± 2.1 | 5 | −0.34 | 1.78 |
| …the constitutional right to protest against government actions should be restricted. | 527 | 3.3 + 1.8 | 3 | 0.29 | 2.03 |
| …sales of firearms should be banned. | 527 | 3.8 ± 2.1 | 4 | 0.12 | 1.76 |
| …sales of non-essential goods should be banned. | 528 | 2.8 ± 1.8 | 2 | 0.75 | 2.47 |
| …abortion clinics should be closed. | 527 | 2.6 ± 2.0 | 2 | 1.00 | 2.71 |
| …governments will need to run the economy by deciding what goods are to be produced, and in what quantities. | 526 | 4.0 ± 1.7 | 4 | −0.25 | 2.20 |
| …illegal activities that increase the spread of the virus should be punished directly by government officials, without the right to trial by jury. | 527 | 2.9 ± 2.0 | 2 | 0.71 | 2.22 |
| …heads of national, state, and local governments should be able to order new restrictions on activities that could spread the virus, without needing to consult legislative bodies (such as Congress or state legislatures). | 527 | 4.4 ± 1.9 | 5 | −0.44 | 2.10 |
| …crimes that take advantage of the pandemic (like stealing masks) should be punished much more severely than the same crimes when committed under normal conditions. | 528 | 4.6 ± 1.8 | 5 | −0.48 | 2.20 |
| …public health experts should be given more authority than elected politicians. | 527 | 5.1 ± 1.8 | 6 | −0.85 | 2.74 |
| …public health authorities should test people for the virus, even if they don't want to be tested, to obtain data that the authorities have decided they need to get the pandemic under control. | 527 | 4.2 ± 1.9 | 4 | −0.25 | 1.99 |
| …citizens of foreign countries should be banned from entering the United States. | 526 | 4.5 ± 2.0 | 5 | −0.28 | 1.89 |
Next, participants were asked a set of demographic questions: age, gender, ethnicity, education level, pre-pandemic household income (specified as Feb. 15, 2020), and current ZIP Code. Finally, they were presented with the statement “My income is being reduced by the social distancing restrictions being imposed to reduce the spread of COVID-19” and asked to endorse it on a 7-point Likert scale anchored by “Strongly Disagree” and “Strongly Agree.” I will refer to this item as income reduction.
Data and Stata commands used in this paper are available at http://doi.org/10.5068/D1RH4K.
For the LWA Index and the three ACT scales, missing responses (44 of 21,120 = 0.21%) were imputed using chained regression equations, separately for each of the four scales. For each missing value, the mean value from 20 imputations was substituted. Because the three ACT scales were highly intercorrelated (Authoritarianism-Conservatism, r = 0.70; Authoritarianism-Traditionalism, r = 0.59, Conservatism-Traditionalism, r = 0.69, all p < .001 at N = 528), overall RWA scores were calculated as the mean across the three scale means.
Participants' state and county of residence were recovered from their ZIP Codes, using the USPS Zip Code Lookup and Google Maps. County COVID-19 rate was scored as cases per 100,000 residents on April 22, 2020, as reported by USA Facts: https://usafacts.org/visualizations/coronavirus-covid-19-spread-map/. Because this variable was strongly right-skewed ( Table 2 ), I used its zero-skewness log transform in analyses.
Descriptive statistics of continuous predictor variables.
| Cronbach's α | Mean ± SD | Median | Range | Skew | Kurtosis | ||
|---|---|---|---|---|---|---|---|
| LWA | 0.93 | 3.67 ± 1.19 | 3.77 | 1.00–6.91 | −0.12 | 2.55 | |
| RWA | 0.91 | 3.12 ± 1.15 | 3.06 | 1.11–6.11 | 0.33 | 2.40 | |
| Authoritarianism | 0.76 | 3.55 ± 1.20 | 3.50 | 1.00–7.00 | 0.11 | 2.63 | |
| Conservatism | 0.85 | 3.11 ± 1.35 | 3.00 | 1.00–7.00 | 0.35 | 2.44 | |
| Traditionalism | 0.84 | 2.73 ± 1.37 | 2.33 | 1.00–7.00 | 0.84 | 2.91 | |
| County COVID-19 cases/100,000 | N.A. | 300.9 ± 503.2 | 111.6 | 2.6–2977.1 | 2.96 | 11.69 | |
| Income reduced by social distancing | N.A. | 4.08 ± 2.29 | 4.00 | 1.00–7.00 | −0.12 | 1.50 | |
To test associations between LWA, RWA, and covariates (the predictor variables), and endorsement of putatively authoritarian pandemic-related policies (the outcome variables), I used hierarchical multiple regression, as instantiated in the hireg module (Bern, 2005) of Stata 16.1. Each outcome variable was modeled separately. In the first step of each model, age, gender, African-American identity, and county COVID-19 rate were entered as predictors. For the four policies pertaining to social distance restrictions, income reduction was also entered in the first step. In the second step of each model, LWA and RWA were entered as predictors.
Table 2 shows reliabilities (Cronbach's alpha) and descriptive statistics for the LWA Index, the full set of ACT Scale items, and each of the three separate ACT scales. This table also shows descriptive statistics for county COVID-19 rate and for the income reduction item.
Tables S3 and S4 show associations among predictor variables. Women scored higher in LWA than men. County COVID-19 rate was negatively correlated with age and positively correlated with LWA. Income reduction was negatively correlated with age and positively correlated with LWA. Age was negatively correlated with LWA and positively correlated with RWA. RWA and LWA were negatively correlated.
Tables 3 and S5 show the results of the hierarchical regressions. To reiterate, first step models included the predictors age, gender, African-American identity, county COVID-19 rate, and (only for the policies pertaining to social distancing) income reduction. In general, these models performed poorly, explaining >4% of the variance in policy endorsement for only two policies. For eight of the 19 policies, first step models did not even attain statistical significance, in spite of the large sample size. Second step models, entering RWA and LWA, increased the variance explained significantly (p < .001) for every policy item except persuasion better than rules for promoting social distancing. For the other 18 policies, second step models were significant at p < .001, and they explained between 0.04 and 0.37 of the variance in policy endorsement.
Results of hierarchical regression analyses.
| Policy endorsement (outcome) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Model | Predictors: standardized βs | R 2 | R 2 change | |||||||
| Age | Female | Afr-Am | Log COVID | Income reduction | RWA | LWA | ||||
| Would report to police | 1 | −0.03 | −0.11* | −0.02 | 0.06 | 0.11* | – | – | 0.03** | – |
| 2 | −0.00 | −0.16*** | −0.05 | 0.05 | 0.05 | 0.21*** | 0.36*** | 0.14*** | 0.11*** | |
| Persuasion better than rules | 1 | 0.04 | −0.05 | −0.06 | −0.05 | 0.06 | – | – | 0.01 | – |
| 2 | 0.03 | −0.03 | −0.06 | −0.04 | 0.08 | −0.01 | −0.09 | 0.02 | 0.01 | |
| Need threat of punishment | 1 | −0.13** | −0.04 | −0.04 | 0.06 | 0.02 | – | – | 0.03* | – |
| 2 | −0.10* | −0.09* | −0.07 | 0.04 | −0.04 | 0.19*** | 0.36*** | 0.15*** | 0.12*** | |
| Must follow distancing orders | 1 | −0.02 | 0.11** | −0.03 | 0.02 | −0.01 | – | – | 0.01 | – |
| 2 | 0.04 | 0.08 | −0.03 | −0.01 | −0.07 | −0.13** | 0.27*** | 0.12*** | 0.11*** | |
| Certificate of immunity | 1 | −0.01 | −0.09* | 0.00 | 0.08 | – | – | – | 0.02 | – |
| 2 | 0.00 | −0.12** | −0.02 | 0.07 | – | 0.18*** | 0.23*** | 0.08 | 0.06*** | |
| Mandatory tracking app | 1 | 0.00 | 0.01 | 0.03 | 0.03 | – | – | – | 0.00 | – |
| 2 | 0.02 | −0.04 | −0.01 | 0.03 | – | 0.27*** | 0.32*** | 0.12*** | 0.12*** | |
| Prohibit misinformation | 1 | −0.11** | 0.09* | 0.02 | 0.08 | – | – | – | 0.03** | – |
| 2 | −0.04 | 0.04 | 0.00 | 0.04 | – | −0.08* | 0.37*** | 0.18*** | 0.15*** | |
| Surveillance of churches | 1 | −0.12** | 0.04 | −0.01 | 0.05 | – | – | – | 0.02* | – |
| 2 | −0.04 | −0.02 | −0.04 | 0.01 | – | 0.06 | 0.49*** | 0.23*** | 0.21*** | |
| Restrict right to protest | 1 | −0.17*** | 0.06 | −0.01 | 0.02 | – | – | – | 0.03** | – |
| 2 | −0.14*** | 0.00 | −0.05 | 0.01 | – | 0.33*** | 0.40*** | 0.22*** | 0.19*** | |
| Ban firearms sales | 1 | −0.15*** | 0.21*** | 0.01 | 0.12** | – | – | – | 0.08*** | – |
| 2 | −0.07 | 0.14*** | −0.02 | 0.09* | – | 0.06 | 0.49*** | 0.29*** | 0.21*** | |
| Ban nonessential items | 1 | −0.08 | −0.07 | −0.07 | −0.03 | – | – | – | 0.02 | – |
| 2 | −0.03 | −0.13** | −0.11** | −0.05 | – | 0.17*** | 0.41*** | 0.17*** | 0.15*** | |
| Close abortion clinics | 1 | 0.14** | −0.10* | 0.10* | −0.10* | – | – | – | 0.05*** | – |
| 2 | 0.04 | −0.09** | 0.06 | −0.06 | – | 0.60*** | −0.07 | 0.42*** | 0.37*** | |
| Government-run economy | 1 | −0.13** | −0.04 | 0.01 | 0.04 | – | – | – | 0.02* | – |
| 2 | −0.08* | −0.10* | −0.02 | 0.02 | – | 0.20*** | 0.41*** | 0.18*** | 0.16*** | |
| Restrict right to trial by jury | 1 | −0.17*** | 0.02 | 0.08 | 0.03 | – | – | – | 0.04*** | – |
| 2 | −0.16*** | −0.04 | 0.02 | 0.03 | – | 0.45*** | 0.41*** | 0.30*** | 0.26*** | |
| Restrictions by executive decree | 1 | −0.04 | 0.04 | −0.00 | 0.06 | – | – | – | 0.01 | – |
| 2 | 0.01 | −0.01 | −0.04 | 0.04 | – | 0.14** | 0.38*** | 0.13*** | 0.13*** | |
| Emergency-enhanced punishment | 1 | −0.10* | −0.12** | −0.08 | 0.00 | – | – | – | 0.03** | – |
| 2 | −0.07 | −0.15*** | −0.10* | −0.01 | – | 0.09* | 0.22*** | 0.07*** | 0.04*** | |
| Authority to public health experts | 1 | −0.16*** | 0.02 | −0.01 | 0.10* | – | – | – | 0.04*** | – |
| 2 | −0.07 | −0.04 | −0.03 | 0.06 | – | −0.07 | 0.47*** | 0.27*** | 0.23*** | |
| Mandatory COVID-19 testing | 1 | −0.06 | −0.03 | −0.09* | 0.04 | – | – | – | 0.02 | – |
| 2 | −0.01 | −0.09* | −0.12** | 0.02 | – | 0.12** | 0.40*** | 0.15*** | 0.13*** | |
| Ban foreigners from entering | 1 | −0.05 | 0.04 | 0.01 | −0.04 | – | – | – | 0.01 | – |
| 2 | −0.10* | 0.03 | −0.02 | −0.01 | – | 0.43*** | 0.00 | 0.18*** | 0.17*** | |
After controlling for all other predictors, both RWA and LWA were significantly positively associated with endorsement of would report to police, need threat of punishment, certificate of immunity, mandatory tracking app, restrict right to protest, ban nonessential items, government-run economy, restrict right to trial by jury, restrictions by executive decree, emergency-enhanced punishment, and mandatory COVID-19 testing. RWA was positively associated, whereas LWA was not associated, with endorsement of close abortion clinics and ban foreigners from entering. LWA was positively associated, whereas RWA was not associated, with endorsement of surveillance of churches, ban firearm sales, and authority to public health experts. LWA was positively associated, whereas RWA was negatively associated, with must follow distancing orders and prohibit misinformation. Some associations with age, gender, and county COVID rate remained significant in second step models. Age was negatively associated with endorsement of need threat of punishment, restrict right to protest, government-run economy, restrict right to trial by jury, and ban foreigners from entering. Women, compared to men, showed lower levels of endorsement of would report to police, need threat of punishment, certificate of immunity, ban nonessential items, close abortion clinics, government-run economy, emergency-enhanced punishment, and mandatory COVID-19 testing. Women showed a higher level of endorsement than men of ban firearm sales. County COVID rate was positively associated with endorsement of ban firearm sales.
Table S6 shows, for each policy item, whether the RWA and LWA unstandardized regression coefficients differed significantly from each other, based on confidence intervals. LWA coefficients were significantly greater than RWA coefficients for follow distancing orders, prohibit misinformation, surveillance of churches, ban firearm sales, ban nonessential items, government-run economy, restrictions by executive decree, authority to public health experts, and mandatory COVID-19 testing. RWA coefficients were significantly greater than LWA coefficients for close abortion clinics and ban foreigners from entering.
Reacting to the severe public health and economic impacts of the COVID-19 pandemic in Spring 2020, many citizens of liberal democratic nations have tolerated, or even demanded, actions from their governments that they would view as unacceptably heavy-handed under normal conditions. The current study used this situation to extend the nomological networks of the well-established construct of Right-wing Authoritarianism (Altemeyer, 1981; Duckitt et al., 2010), and the promising, but so far little-explored, construct of Left-wing Authoritarianism (Conway III et al., 2017; Costello, Bowes, et al., 2020). Both traits predicted endorsement of a range of putatively pandemic-mitigating policies and practices that many would regard as authoritarian. A few of these results were unremarkable, or even slightly circular, e.g. one of the ACT items condemns abortion, and one of the policies endorsed more by people higher in RWA was closing abortion clinics. However, most of the results illuminated, some of them rather surprisingly, how people with authoritarian attitudes respond when, due to a crisis, a menu of normally taboo authoritarian policies appears on the table of mainstream public debate in a liberal democracy. For example, RWA is positively related to conservatism (Crowson, Thoma, & Hestevold, 2005), and one component of American conservatism is advocacy of free markets, and yet RWA was positively associated with endorsement of the government running the economy.
Although I did not measure perception of the danger posed by COVID-19, it is unlikely that variation in objective risk of death or serious illness drove the observed associations between RWA or LWA and increased endorsement of putatively authoritarian policies. None of the three covariates chosen because of their association with COVID-19 mortality risk (county COVID-19 rate, age, or African-American identity) was consistently positively related to endorsement of these policies.
This study builds on recent work (Costello, Bowes, et al., 2020) that has demonstrated the existence of authoritarian attitudes on both ends of the political spectrum, and has documented numerous psychological similarities between RWA and LWA. Two such similarities are belief in a dangerous world, and preference for state control. Consistent with these findings, I found that in response to the danger posed by a deadly pandemic, people high in RWA and people high in LWA agreed on the need for enhanced state control in several domains, including restrictions on the right to protest, punishment without the right to trial by jury, and surveillance via a mandatory tracking app. The policies on which people high in RWA and people high in LWA disagreed tended to be those most directly tied to American right-wing vs. left-wing values, e.g. religion, abortion, and immigration.
This study had three major limitations. First, no other political attitudes, besides RWA and LWA, were measured. Controlling for other attitudes would have changed the observed statistical associations between authoritarianism and policy endorsements. As just one example, RWA is distinct from, but positively correlated with, Social Dominance Orientation (SDO: Pratto, Sidanius, Stallworth, & Malle, 1994; Roccato & Ricolfi, 2005), defined as degree of preference for inequality among social groups. Controlling for SDO might have reduced the association between RWA and endorsement of ban foreigners from entering. Second, as a cross-sectional study, this work cannot address the possibility of reversed causality, i.e. that the COVID-19 pandemic, as a perceived threat, has increased levels of authoritarianism (see Duckitt, 2001). Finally, my study population consisted exclusively of U.S. residents, so its findings do not generalize to other countries. Exploring this general topic internationally would require compiling country-specific lists of putatively authoritarian pandemic-mitigation policies. Furthermore, the validity of the LWA construct outside the U.S. has not been demonstrated.
LWA holds considerable promise as an explanatory construct in political psychology (Conway III et al., 2017; Costello, Bowes, et al., 2020). Both the course of the COVID-19 pandemic, and the general U.S. political scene, have changed considerably since the data for this study were collected. Therefore, replicating it exactly might be impossible. In general, however, the relationship of LWA to policy preferences comprises a fruitful topic for future research.
Joseph H. Manson:Conceptualization, Data curation, Formal analysis, Writing - original draft.