Losses, hopes, and expectations for sustainable futures after COVID

The paper investigates public attitudes in the United States and the United Kingdom toward potential post‑COVID futures amid widespread societal disruption. Public discourse has featured competing narratives: a rapid "return to normal" versus a progressive "build back better" path emphasizing sustainability, fairness, and climate action. A second cross‑cutting debate concerns the distribution of power after the pandemic: strong government versus individual and community autonomy. The authors formalize these as two orthogonal dimensions—orientation (back to normal vs. progressive) and distribution of power (strong government vs. individual autonomy)—and operationalize them in four scenarios. The study aims to answer three questions: (1) Do people prefer a return to normal or a more progressive future? (2) Do they prefer futures characterized by strong government or individual autonomy? (3) Do their preferences align with their expectations for what is likely to happen and with what they think others want? The work is motivated by concerns about democratic backsliding during the pandemic, the politics of climate action, and the potential role of pluralistic ignorance in masking true public preferences.

The introduction situates the work within several literatures. First, it reviews pandemic impacts on livelihoods and inequalities, and the emergence of competing recovery narratives, including calls for a Green New Deal and resistance to linking climate policy to immediate economic relief. Second, it discusses civil liberties during COVID‑19, democracy under strain, and heterogeneous governance approaches, noting evidence that trust‑based, accountable policies yielded better outcomes. Third, it links to social perception research: the false consensus effect and pluralistic ignorance, especially regarding climate change, where people often misperceive social consensus, affecting willingness to express views and support policies. The study also considers moral values (Moral Foundations) and demographics known to correlate with climate attitudes, though ultimately finds moral foundations play limited roles. This background motivates measuring own preferences, perceived likelihoods, and perceived preferences of others across structured futures that reflect salient dimensions of public debate.

Overview: Two experiments (2020 and 2021) with representative U.S. and U.K. samples recruited via Prolific examined preferences for four future scenarios varying along two orthogonal factors: orientation (back to normal vs. progressive) and distribution of power (strong government vs. individual autonomy). Scenarios: (1) For freedom (back to normal + individual autonomy); (2) Grassroots leadership (progressive + individual autonomy); (3) Collective safety (back to normal + strong government); (4) Fairer future (progressive + strong government). All scenarios were designed to be positively intentioned and to differ principally on the manipulated dimensions.

Scenario calibration (Exp. 1 ancillary; Exp. 2 embedded): A separate U.S. calibration study (N=404; CloudResearch MTurk; mean age 40.7) presented a single randomly chosen scenario and asked participants to rate it on 10‑point scales for orientation (return to normal–build back better) and distribution of power (people power–government power). Results confirmed a clear 2×2 factorial perception as intended. In Experiment 2, all participants provided the same calibration ratings post‑scenario; combined and country‑specific results replicated the intended structure.

Experiment 1 (May–July 2020): Design: Within‑subjects 2×2; each participant evaluated all four scenarios in random order. Samples: U.S. N=277 (tested early May 2020, most states in lockdown); U.K. N=300 (tested early July 2020, emerging from first lockdown). Measures: After each scenario, participants rated (1) own preference ("own want"), (2) perceived likelihood of the scenario, and (3) presumed preference of others in their country ("others want"), all on 1–11 scales. U.S. participants also completed the 20‑item Moral Foundations Questionnaire (harm, fairness, ingroup, authority, purity; 6‑point scales). Demographics included age, gender, English proficiency, employment, education, and income. Open‑ended questions asked about what was missed, welcomed as lost, and what should stay different or change after the pandemic. Procedure included an attention/effort check. Analysis: Linear mixed‑effects models (lme4::lmer) with orientation and distribution of power as fixed effects (and random slopes for main effects), fit separately for each question and country. Simple (fixed factors only) vs. complex (with covariates) models were compared using BIC; simple models generally preferred. Power analyses used simR to estimate detectable effect sizes. Qualitative data were analyzed inductively using NVivo 12 and grounded theory coding; word clouds and word trees summarized themes; no demographic splits were applied to qualitative data.

Experiment 2 (late July–early August 2021): Design: Between‑subjects 2×2; each participant saw a single scenario constructed from a common syntactic template to maximize experimental control. Samples: U.S. N=401 (mean age 44.8), U.K. N=398 (mean age 45.1). Measures: Same three primary ratings (own want, likelihood, others want). Political leaning (1=very left/liberal to 11=very right/conservative) was collected as a potential moderator. Calibration ratings were obtained post‑scenario. Procedure included effort checks. Analysis: Omnibus 2×2×2 ANOVAs (orientation × distribution of power × country) via lm with Helmert contrasts and Type III sums of squares (car::Anova). Moderated regressions added political leaning (zero‑centered) and its interactions. Because Prolific does not ensure political representativeness, the authors also performed a reweighting (raking) analysis to match national election study distributions; results mirrored unweighted analyses (reported in supplement), so main text reports standard analyses.

Experiment 1 (within‑subjects, U.S. N=277; U.K. N=300):

• Own preferences: Strong preference for progressive futures over return to normal in both countries. U.S.: main effect of orientation β = −1.27, 95% CI [−1.45, −1.10], t=−14.22, p<0.001 (standardized −0.39); modest preference for individual autonomy β = 0.37, 95% CI [0.20, 0.55], t=4.17, p<0.001 (standardized 0.11); no interaction. U.K.: main effect of orientation β = −1.16, 95% CI [−1.33, −0.99], t=−13.31, p<0.001 (standardized −0.36); interaction β = 0.30, 95% CI [0.17, 0.44], t=4.52, p<0.001 (standardized 0.10) indicating preference for individual autonomy under back‑to‑normal but preference for stronger government when pursuing progressive change.
• Most‑preferred scenario (winner) distribution: U.S.: Fairer Future 39% (N=108) and Grassroots Leadership 40.1% (N=111) combined 79.1% (N=219) vs. return‑to‑normal For Freedom 14.1% (N=39) and Collective Safety 6.9% (N=19). U.K.: Fairer Future 38% (N=114) and Grassroots Leadership 36.3% (N=109) combined 74.3% (N=223) vs. For Freedom 15.7% (N=47) and Collective Safety 10.0% (N=30).
• Likelihood: Participants judged a return to normal more likely than progressive futures. U.S.: orientation main effect B = 0.75, 95% CI [0.58, 0.92], t=8.70, p<0.001 (standardized 0.28); interaction B = 0.41, 95% CI [0.30, 0.53], t=6.96, p<0.001—progressive + individual autonomy seen as particularly unlikely. U.K.: orientation main effect B = 0.69, 95% CI [0.54, 0.84], t=9.04, p<0.001 (standardized 0.26); distribution of power main effect negative (strong government seen as less likely overall), and interaction similar to U.S.
• Perceptions of others’ preferences: U.S.: believed others preferred return to normal more than they themselves did (distribution of power main effect β = 0.27, p<0.001; interaction β = 0.30, p<0.001); no orientation main effect. U.K.: participants correctly sensed overall preference for progressive futures but still overestimated others’ desire for a return to normal (orientation β = −0.38, p<0.001; distribution of power β = −0.19, p=0.006; interaction β = 0.35, p<0.001).
• Anticipated disappointment (own want minus likelihood): Older age (and U.S. stronger ingroup allegiance; U.K. higher income) associated with lower anticipated disappointment.
• Qualitative themes: Most missed pre‑pandemic: social interaction with family/friends, eating out, anxiety‑free public life, cultural activities, shopping, touch. U.K. respondents more often missed freedom of movement and freedom from mask‑wearing (timing effects). Welcome losses: many said “nothing”; others valued less commuting/face‑to‑face requirements, a slower pace, reduced stress, mindset shifts, less pollution/traffic, fewer crowds. Desired to retain/change: working from home and reduced commuting; slower life; more collective friendliness and care for others/environment; some desire to maintain distancing and limit crowds; higher U.K. emphasis on appreciating essential workers; U.S. respondents more often sought better healthcare and evidence‑based, less partisan politics. The most frequent word was “people,” often used in calls for prosocial behavior change.

Experiment 2 (between‑subjects, U.S. N=401; U.K. N=398):

• Own preferences: Main effects—orientation F(1,791)=21.55, MSE=8.85, p<0.001, η²=0.027 (progressive > back to normal); distribution of power F(1,791)=68.59, p<0.001, η²=0.080 (individual autonomy > strong government). Political leaning moderated preferences: adding political leaning significantly improved fit; leaning had a main effect F(1,783)=16.02, p<0.001, and interacted with orientation F(1,783)=31.90, p<0.001, η²=0.039 and distribution of power F(1,783)=9.99, p=0.002, η²=0.013. Right‑leaning participants were more favorable to back‑to‑normal and preferred individual autonomy more strongly; effects were attenuated for progressive scenarios.
• Likelihood: Return to normal judged more likely overall; progressive + individual autonomy seen as particularly unlikely. Country × orientation interaction was significant; political right respondents expressed greater confidence in their expectations (political leaning covariate significant).
• Perceptions of others’ preferences: Main effects—orientation F(1,791)=10.20, p=0.001, η²=0.013; distribution of power F(1,791)=88.50, p<0.001, η²=0.101. Participants presumed others preferred a return to normal more than progressive futures and correctly inferred others preferred individual autonomy over strong government. Political leaning did not improve explanatory power for this outcome.

Overall: Across both experiments and countries, people personally preferred progressive, sustainability‑oriented futures but believed a return to normal was more likely and that others preferred return to normal more than they do—an instance of pluralistic ignorance. Opposition to progressive futures was limited and concentrated on the extreme right; center‑right respondents were often ambivalent rather than opposed.

The findings directly address the research questions. First, respondents in both the U.S. and U.K. preferred progressive, sustainability‑ and fairness‑oriented futures over a simple return to pre‑pandemic normality. Second, preferences modestly favored futures with greater individual and community autonomy, though in the U.K. a progressive path was acceptable or even preferred with stronger government action. Third, preferences diverged from expectations: people expected a return to normal to be more likely and underestimated societal support for progressive futures, revealing pluralistic ignorance. This misperception is consequential because perceived social consensus shapes attitudes and willingness to advocate for policies, particularly around climate change. The temporal replication (one year apart) and robust results across methodological variations (within‑ vs. between‑subjects; revised scenarios) underscore the stability of these patterns. Politically, although left‑leaning respondents were more supportive of progressive futures, overall opposition was not widespread, suggesting scope for broad coalitions. The qualitative data enrich the interpretation, showing concrete aspects of daily life respondents wish to retain (e.g., remote work, reduced commuting) and a desire for more caring, less individualistic norms, aligning with the quantitative preference for progressive futures. The gap between wants, expectations, and perceived social norms highlights the need for public communication that accurately reflects majority preferences to avoid self‑fulfilling returns to the status quo.

The study contributes evidence that majorities in two Anglophone countries personally favor progressive, sustainability‑oriented post‑COVID futures, yet expect a reversion to pre‑pandemic norms and underestimate others’ support for progressive change. This combination signals pluralistic ignorance that could impede policy ambition. Policy‑relevant implications include: making visible the true social consensus for sustainable and equitable futures; leveraging consensus messaging (social and scientific) to increase motivation and support for climate and fairness policies; and creating deliberative and digital fora that reduce perceived polarization and foster constructive dialogue. Given that participants saw progressive + individual autonomy as unlikely, policymakers may need to articulate credible pathways and governance arrangements to achieve progressive goals. Future research should extend beyond the Anglosphere, explore a wider set of nonlinear and potentially disruptive futures (including inequality and tipping points), and further examine how correcting misperceived norms influences public support and behavior over time.

• Geographic scope: Samples were limited to the U.S. and U.K.; findings may not generalize to other cultural or political contexts.
• Political representativeness: Prolific’s representative panels match demographics but not political leanings; left‑leaning respondents were overrepresented (especially in the U.S.), though reweighting analyses yielded similar conclusions.
• Scenario space: The four scenarios, while grounded in public debate, do not span all plausible futures; they reflect relatively linear trajectories and exclude extreme tipping points or additional crises; limited treatment of inequality.
• Measures/design: Open‑ended qualitative data were collected only in Experiment 1; MFQ measured only in the U.S. sample of Experiment 1 and had limited explanatory power.
• Expectation vs. causality: Associations with political leaning are correlational; direction of causality cannot be established.