Psychological characteristics associated with COVID-19 vaccine hesitancy and resistance in Ireland and the United Kingdom

The COVID-19 pandemic, declared on March 11, 2020, has caused extensive morbidity, mortality, and economic disruption worldwide. In early 2020 several vaccines entered clinical trials, highlighting the urgent need to ensure adequate public uptake. Prior European surveys indicated substantial uncertainty and resistance toward a prospective COVID-19 vaccine. Moving beyond explicit reasons for vaccine opposition, the study adopts an attitude-roots perspective to identify psychological processes distinguishing vaccine hesitant and resistant individuals from accepters, with the aim of informing targeted public health messaging. The study set four objectives: (1) estimate prevalence of vaccine acceptance, hesitancy, and resistance in nationally representative samples from Ireland and the UK; (2) profile hesitant/resistant individuals on sociodemographic, political, and health indicators; (3) identify salient psychological characteristics differentiating hesitant/resistant from accepting groups; and (4) assess sources of COVID-19 information and trust in those sources across acceptance groups to guide communication strategies.

Existing work links vaccine hesitancy to multiple psychological and socio-political factors. Prior studies implicate altruistic beliefs, personality traits (e.g., neuroticism, conscientiousness), locus of control, and cognitive reflection in vaccine decisions. Hesitancy and resistance are associated with conspiratorial, religious, and paranoid beliefs; mistrust in authorities (government, scientists, healthcare professionals); authoritarian political views; societal disaffection; and anti-immigration attitudes. This literature suggests that cognitive styles, emotions, beliefs, trust, and socio-political orientations likely differentiate hesitant/resistant from accepting individuals, underscoring the need to map these constructs comprehensively for COVID-19 vaccination attitudes.

Design and samples: Cross-sectional surveys of nationally representative adult samples in Ireland (N=1041) and the UK (N=2025) were conducted by Qualtrics using quota sampling (matching population parameters: Ireland—sex, age, geography; UK—age, sex, income). UK data were collected March 23–28, 2020; Ireland March 31–April 5, 2020, during the first week of strict national lockdowns. Inclusion: adults ≥18, residents, English-speaking. Participants completed online surveys (median completion time ≈38 min Ireland; ≈29 min UK) and were reimbursed. Ethics approvals were obtained from the University of Sheffield and Ulster University. Power analyses prioritized detection of mental health disorder prevalences for the broader C19PRC study; target sizes were ~1000 (Ireland) and ~2000 (UK). Measures: Vaccine acceptance was assessed by a single question classifying respondents as accepting (Yes), hesitant (Maybe), or resistant (No). Sociodemographics included sex, age (mid-decade categories), birthplace/ethnicity, region, urbanicity, education, income, employment, religion, household composition, children, voting behavior (Ireland: party voted; UK: general election vote, Brexit referendum), and religiosity. Health indicators included current/past mental health treatment; personal and immediate family chronic conditions (diabetes, lung, heart disease); pregnancy; confirmed/suspected COVID-19 infection in self/close other. Psychological indicators: Personality (BFI-10); Locus of Control (internal, chance, powerful others subscales); Cognitive Reflection Test (3-item MCQ); Altruism/Identification with all Humanity (identify with others, care, help subscales); Conspiracy Mentality Scale; Paranoia (PaDS persecution subscale); Religious beliefs (Monotheist and Atheist Beliefs Scale—atheist items reverse scored to index religiosity); Trust in institutions (state: political parties, parliament, government, police, legal system; plus separate trust in scientists; and in doctors/health professionals); Authoritarianism (Very Short Authoritarianism Scale); Social Dominance Orientation (SDO-7 items); Attitudes toward migrants (economy, cultural enrichment items). COVID-19 information consumption and trust: frequency of obtaining and trusting information from newspapers, TV, radio, internet sites, social media, doctor, other health professionals, government agencies, and family/friends (1–4 Likert scales). Analyses: (1) Proportions accepting/hesitant/resistant; (2) Multinomial logistic regression (reference: acceptance, then re-estimated with hesitant as reference) to identify adjusted associations (AORs with 95% CIs) between sociodemographic/political/health predictors and hesitancy/resistance; (3) One-way ANOVAs with Bonferroni post-hocs to compare psychological variables across acceptance groups, reporting η² and combining hesitant+resistant vs accepting with Cohen’s d; (4) One-way ANOVAs comparing information consumption and trust across groups. Analyses were conducted in SPSS v25.

Prevalence: Ireland—accepting 65% (95% CI 62.0–67.9), hesitant 26% (22.9–28.3), resistant 9% (7.7–11.3). UK—accepting 69% (66.8–70.9), hesitant 25% (23.1–26.9), resistant 6% (5.2–7.3). Within the UK, Northern Ireland had the lowest acceptance (51%). Sociodemographic, political, health correlates (multinomial models): Ireland—Compared to accepters, hesitant were more likely female (AOR 1.62, 1.18–2.22), aged 35–44 (AOR 2.00, 1.06–3.75), and less likely to have received mental health treatment (AOR 0.63, 0.45–0.88). Resistant were more likely aged 35–44 (AOR 3.33, 1.17–9.47), city residents (AOR 1.90, 1.02–3.54), non-Irish ethnicity (AOR 2.89, 1.17–7.09), voters for Sinn Féin (AOR 3.22, 1.14–9.08) or Independent (AOR 4.15, 1.19–14.49), have an underlying health condition (AOR 2.59, 1.38–4.85), and be in lower income brackets (AORs ~3.16–5.73 vs highest). Resistant vs hesitant: non-Irish ethnicity (AOR 2.76, 1.05–7.19), underlying condition (AOR 2.68, 1.33–5.38), and lower income (AORs 2.82–5.44). UK—Hesitant were more likely female (OR 1.43, 1.14–1.80) and younger than 65. Resistant were markedly younger (over 10× more likely in ages 18–44; >4× ages 45–64 vs 65+), suburban residents (OR 2.13, 1.01–4.49), in lower income brackets, and pregnant (OR 2.36, 1.03–5.40). Resistant vs hesitant: younger age only (e.g., ~7× ages 18–24 and ~4× ages 25–44 vs 65+). Psychological profiles: In both countries, hesitant/resistant vs accepting showed: lower trust in scientists (Ireland d=0.51; UK d=0.38) and healthcare professionals (Ireland d=0.45; UK d=0.39), and lower trust in the state (Ireland d=0.31; UK d=0.16); higher conspiracy beliefs (Ireland d=0.21; UK resistant vs hesitant η²≈0.01) and religious beliefs (Ireland d=0.20; UK d=0.21); higher paranoia (UK d=0.27); higher social dominance (Ireland d=0.22; UK d=0.21) and authoritarianism (Ireland d=0.14); lower agreeableness (Ireland d=0.15; UK d=0.22) and conscientiousness (UK d=0.17), higher neuroticism (UK d=0.11); lower cognitive reflection (Ireland d=0.25; UK d=0.14); higher internal locus of control (Ireland d=0.14; UK d=0.16) and belief in chance (UK d=0.17), and lower belief in powerful others (UK d=0.19); more negative attitudes towards migrants (Ireland d≈0.27–0.29; UK d=0.11). Resistant vs hesitant: lower trust in scientists and healthcare professionals (Ireland η²≈0.05–0.06; UK η²≈0.03–0.04) and higher conspiracy beliefs (Ireland η²=0.02; UK η²≈0.01). Information consumption and trust: Ireland—Resistant consumed less information from newspapers, TV, radio, and government agencies, and more from social media (p<0.05) vs accepters; they trusted newspapers, TV, radio, doctors, other health professionals, and government agencies less (p<0.05). No significant differences between accepters and hesitant for consumption/trust. Resistant vs hesitant: resistant consumed less TV/radio and trusted newspapers, TV, radio, doctors, other health professionals, and government agencies less. UK—Resistant consumed less from newspapers and TV (p<0.05) and trusted newspapers, TV, radio, doctors, other health professionals, and government agencies less (p<0.05) vs accepters. No significant differences between accepters and hesitant in consumption/trust; resistant vs hesitant showed lower trust in newspapers, radio, doctors, and other health professionals.

Both Ireland and the UK exhibited substantial vaccine hesitancy (≈25–26%) and resistance (≈6–9%) early in the pandemic, with only two-thirds reporting willingness to accept a COVID-19 vaccine. Sociodemographic predictors overlapped (female sex, younger age, lower income), while some were context-specific (Ireland: non-Irish ethnicity, city residence, anti-establishment voting, underlying health problems; UK: suburban residence, pregnancy). Despite such contextual differences, hesitant/resistant individuals shared a consistent psychological profile across countries: lower trust in scientists, healthcare professionals, and the state; higher conspiracy, religiosity, and paranoia; greater social dominance and authoritarian orientations; lower agreeableness/conscientiousness, higher neuroticism; lower analytical reasoning; stronger internal locus of control; and more negative attitudes toward migrants. Resistant individuals particularly distrusted mainstream information sources and were more inclined toward social media. These patterns suggest that traditional authority-based messaging may be less effective with hesitant/resistant audiences and that tailored communication via trusted non-traditional messengers and channels may be needed to increase vaccine uptake.

The study quantifies early-pandemic prevalence of COVID-19 vaccine acceptance, hesitancy, and resistance in Ireland and the UK and delineates sociodemographic and psychological profiles associated with hesitancy/resistance. Findings indicate the need for targeted public health strategies: engage non-traditional and community/religious leaders, emphasize clear and personally relevant benefits, and disseminate messages across diverse media including social platforms. Future research should replicate psychological profiling and communication preference assessments in other countries and over time as vaccine characteristics and public perceptions evolve, to inform adaptive, context-specific interventions that maximize vaccine uptake.

- Non-probability internet-based quota samples (opt-in panels) preclude calculation of response rates and may differ from the underlying population; however, quotas and adjustments aimed to enhance representativeness. - Data were collected during the first week of the strictest lockdowns, potentially influencing attitudes. - Attitudes were measured toward a hypothetical vaccine with unknown efficacy and risk profile at the time. - Findings are limited to two Western European countries with similar sociocultural contexts; generalizability to other settings is uncertain. - Samples excluded institutionalized and hard-to-reach populations (e.g., those not online), limiting generalizability.