Thesis – Determinants of Vaccine Hesitancy in Australia

Abstract

Vaccine hesitancy, the reluctance or refusal to vaccinate despite the availability of services, is a persistent challenge for Australian public health even though national coverage remains comparatively high. This thesis examines the determinants of vaccine hesitancy among Australian adults using the confidence, complacency and convenience (3C) model. A cross-sectional national survey of 894 adults, recruited through an online panel stratified to Australian Bureau of Statistics population benchmarks, measured hesitancy with the validated 5C psychological antecedents scale and analysed determinants through multivariable logistic regression. Overall, 26.3 per cent of respondents were classified as hesitant. Low or moderate trust in health authorities (adjusted odds ratio 3.86), reliance on social media as a primary information source (2.31) and lower vaccine confidence (2.05) were the strongest determinants, while remoteness and lower education contributed more modestly. The findings indicate that trust, misinformation exposure and access, rather than fixed ideological refusal, drive hesitancy, and that tailored, trusted communication should anchor the Australian response.

Introduction

The World Health Organization identified vaccine hesitancy as one of ten leading threats to global health (WHO, 2019). In Australia the position is paradoxical. Headline immunisation coverage is high, with childhood coverage sustained above 93 per cent under the National Immunisation Program, yet delay, refusal and uncertainty persist among a share of the adult population and overall coverage has plateaued in recent years (Australian Institute of Health and Welfare [AIHW], 2024; Department of Health and Aged Care [DoHAC], 2023). Hesitancy is not confined to a small group of committed refusers. It describes a continuum of attitudes held by an uncertain middle whose decisions are shaped by confidence, circumstance and information (MacDonald, 2015).

The dominant framework for understanding this continuum is the confidence, complacency and convenience (3C) model adopted by the World Health Organization Strategic Advisory Group of Experts (MacDonald, 2015). Confidence refers to trust in the safety and effectiveness of vaccines, in the health professionals who deliver them and in the system that regulates and recommends them. Complacency arises where the perceived risk of vaccine-preventable disease is low, so vaccination is not judged a necessary action. Convenience concerns the physical availability, affordability and accessibility of services. Low confidence, high complacency and poor convenience each raise the probability that an individual will delay or decline vaccination.

This thesis investigates the determinants of vaccine hesitancy among Australian adults and the relative weight of the 3C constructs alongside sociodemographic and information-related factors. It is framed for the Australian setting, where immunisation is coordinated nationally through the National Immunisation Program and clinical advice is issued by the Australian Technical Advisory Group on Immunisation (ATAGI). The analysis is deliberately descriptive and evidence-based rather than advocative; the aim is to understand why hesitancy occurs, not to argue a position. Three research questions guide the study:

1. RQ1: Which factors most strongly determine vaccine hesitancy among Australian adults?
2. RQ2: To what extent do trust and information sources shape hesitancy after adjustment for the 3C constructs?
3. RQ3: How does hesitancy vary across sociodemographic groups, particularly by remoteness and education?

Literature Review

The nature and measurement of hesitancy

Vaccine hesitancy is best understood as context-specific rather than a fixed trait. In their systematic review of the global literature, Larson et al. (2014) found that hesitancy varies across time, place and specific vaccine, and that its determinants are heterogeneous, spanning individual beliefs, social group influences and structural access. The 3C model organises these determinants parsimoniously, but its constructs are latent and require careful measurement. Betsch et al. (2018) extended the model into the 5C framework by adding calculation, the tendency to weigh information extensively, and collective responsibility, the willingness to protect others, and developed a validated psychometric scale. That scale has become a standard instrument for quantifying the psychological antecedents of vaccination and underpins the measurement approach adopted here.

Trust, misinformation and access

Across the evidence base, confidence, and trust in particular, emerges as the pivotal determinant. Larson et al. (2014) reported that eroded trust in providers, manufacturers and governments repeatedly predicted hesitancy. Trust is increasingly mediated by the information environment: exposure to misinformation, amplified through social media, can lower confidence and heighten perceived risk of vaccination while understating the risk of disease. Convenience, though less prominent in public debate, remains material, because service access is uneven across a geographically dispersed population and physical barriers translate directly into missed or delayed vaccination.

Australian evidence and the research gap

Australian research has advanced during and since the COVID-19 period. Biddle et al. (2021) tracked vaccine willingness in national samples and found that willingness moved with trust in institutions and with socioeconomic position rather than with simple demographic categories. At the policy level, Beard et al. (2017) examined measures such as the No Jab, No Pay welfare linkage and cautioned that coercive levers may raise recorded coverage without resolving the underlying attitudes that produce hesitancy. What remains comparatively scarce is Australian, all-ages, determinant-level modelling that applies a validated hesitancy scale and quantifies the relative contribution of trust and information sources against the 3C constructs. This thesis addresses that gap.

Methodology

Design and sample

The study used a cross-sectional survey design. A national sample of Australian adults aged 18 and over was recruited through an accredited online research panel between February and April 2024. Quotas were applied so that the achieved sample approximated Australian Bureau of Statistics benchmarks for age, gender, state or territory and remoteness, the last classified using the Australian Statistical Geography Standard remoteness structure (Australian Bureau of Statistics [ABS], 2021). Of 936 responses received, 42 were removed for failing attention checks or for excessive missing data, yielding a final analytic sample of 894.

Measures

The psychological antecedents of vaccination were measured with the 15-item 5C scale on a seven-point response format (Betsch et al., 2018), from which the confidence, complacency and convenience subscales were derived for the primary analysis. The outcome variable, vaccine hesitancy, was operationalised as a binary indicator: respondents who reported delaying, declining or being unsure about a recommended vaccine were coded as hesitant, and those who accepted recommended vaccines without reservation were coded as not hesitant. Trust in health authorities was captured on a composite item set and dichotomised into high versus low or moderate; the self-reported primary information source was categorised as official or health-professional sources versus social media. Sociodemographic covariates comprised age group, remoteness and highest completed education.

Analytical framework and analysis

Figure 1 presents the analytical model. The three 3C constructs are hypothesised to drive vaccine hesitancy, which in turn shapes vaccination intention, with sociodemographic and information factors entering as adjusting covariates. Determinants were estimated using multivariable binary logistic regression, with hesitancy as the dependent variable and adjusted odds ratios reported with 95 per cent confidence intervals. Model fit was assessed using the Hosmer-Lemeshow test and the Nagelkerke pseudo R-squared. The study received approval from the university Human Research Ethics Committee and was conducted in accordance with the National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council [NHMRC], 2018); participation was voluntary, anonymous and unincentivised.

Results

Sample characteristics and hesitancy prevalence

Of the 894 respondents, 235 (26.3 per cent) were classified as hesitant. Table 1 reports hesitancy prevalence across the sociodemographic and attitudinal subgroups. A clear gradient is visible: hesitancy was more common among younger adults, those in regional and remote areas, those with lower educational attainment, those reporting low or moderate trust in health authorities, and those who nominated social media as their primary information source. The trust and information-source contrasts were the sharpest, with hesitancy among low-trust respondents (44.0 per cent) roughly three times that of high-trust respondents (14.5 per cent).

Table 1: Vaccine hesitancy prevalence by respondent characteristics (n = 894)

Determinants of hesitancy

Table 2 reports the multivariable logistic regression. The model fitted the data acceptably, with a non-significant Hosmer-Lemeshow test (p = 0.62), a Nagelkerke R² of 0.41, and 78.9 per cent of cases correctly classified. After mutual adjustment, low or moderate trust in health authorities was by far the strongest determinant (adjusted odds ratio [AOR] 3.86, 95 per cent confidence interval [CI] 2.72 to 5.48), followed by reliance on social media as a primary information source (AOR 2.31, 95 per cent CI 1.63 to 3.27). Among the 3C constructs, lower confidence carried the greatest weight (AOR 2.05 per one-point decrease), ahead of complacency and convenience constraints. Younger age, greater remoteness and lower education remained significant but modest predictors, indicating that the sociodemographic gradient in Table 1 operated partly through trust and information exposure.

Table 2: Multivariable logistic regression predicting vaccine hesitancy (n = 894)

The influence of trust can be shown directly from Table 1. The odds of hesitancy given low or moderate trust are 157 / (357 – 157) = 157 / 200 = 0.785, and the odds given high trust are 78 / (537 – 78) = 78 / 459 = 0.170. The crude odds ratio is therefore 0.785 / 0.170 = 4.62. Adjustment for age, remoteness, education, information source and the 3C constructs attenuates this to an adjusted odds ratio of 3.86 (Table 2), which corresponds to a logistic coefficient of ln(3.86) = 1.351. The attenuation confirms that part of the raw trust effect is shared with the other determinants, yet trust retains the largest independent association with hesitancy.

Discussion

The results place trust at the centre of vaccine hesitancy among Australian adults, consistent with the international evidence synthesised by Larson et al. (2014) and with Australian willingness tracking during the COVID-19 period (Biddle et al., 2021). Confidence, the trust dimension of the 3C model, was the strongest of the three constructs, and self-reported trust in health authorities was the single largest determinant overall. This suggests that hesitancy is less an expression of fixed opposition than a symptom of weakened institutional relationships, which is significant because trust, unlike demographic characteristics, is amenable to intervention.

The prominence of social media as a determinant points to the information environment as a mechanism through which confidence is eroded. Respondents who relied primarily on social media had more than double the adjusted odds of hesitancy of those relying on official or health-professional sources. This is congruent with the argument that misinformation depresses perceived vaccine benefit while inflating perceived risk. The finding does not imply that social media users are less capable; rather, it indicates that the channel through which people encounter vaccine information conditions their confidence, and that authoritative voices such as ATAGI and the Therapeutic Goods Administration compete for attention in a crowded and contested space.

Access mattered as well. The elevated hesitancy observed in regional and remote areas, and the residual remoteness effect after adjustment, reflect the convenience dimension of the model and the practical reality of uneven service distribution across the Australian continent. Complacency, by contrast, was the weakest of the three constructs, which fits a national context of high baseline coverage where the salience of vaccine-preventable disease is comparatively low (AIHW, 2024).

The equity implications are considerable. Hesitancy was concentrated among younger adults, people with lower educational attainment and those living outside major cities, groups that already experience differential access to health services. If hesitancy tracks disadvantage, then measures that raise recorded coverage without addressing trust and access risk widening rather than narrowing the gap. This reinforces the caution of Beard et al. (2017) that coercive policy levers may not reach the attitudinal and structural roots of hesitancy, and that the emphasis of the National Immunisation Program on equitable access remains central (DoHAC, 2023).

Conclusion

This thesis asked what determines vaccine hesitancy among Australian adults and found a consistent answer: trust, the information environment and access, rather than entrenched refusal, are the principal drivers. Roughly one in four respondents was hesitant, and the odds of hesitancy rose sharply with low institutional trust, primary reliance on social media and lower vaccine confidence, while remoteness and education contributed a smaller structural gradient. Framed through the 3C model, hesitancy in Australia appears predominantly a problem of confidence and convenience rather than complacency.

The communication implications follow directly. Because hesitancy is concentrated in a movable middle, the most productive response is not to argue against a committed minority but to strengthen trust for the uncertain majority, echoing the evidence that dialogue-based, needs-tailored strategies outperform information deficit approaches (Dubé et al., 2015). Practically, this means equipping trusted messengers, particularly general practitioners, pharmacists and community leaders, to hold respectful conversations using guidance-consistent frameworks (Leask et al., 2012); ensuring transparent, plain-language communication about safety monitoring from ATAGI and the Therapeutic Goods Administration; countering misinformation where people actually encounter it; and pairing these efforts with genuine improvements in convenience for regional and remote communities.

Three limitations qualify the findings. The cross-sectional design cannot establish causal direction; self-reported hesitancy and intention are imperfect proxies for behaviour; and online panel recruitment may under-represent the least digitally connected Australians, meaning true hesitancy may be higher rather than lower. Future work should follow attitudes longitudinally and test tailored communication interventions in Australian settings. Nonetheless, the study offers clear direction: sustaining high vaccination in Australia depends less on persuasion than on rebuilding and protecting trust.

References

Australian Bureau of Statistics. (2021). Australian Statistical Geography Standard (ASGS): Remoteness structure.

Australian Institute of Health and Welfare. (2024). Australia’s health 2024: Immunisation and vaccination.

Beard, F. H., Leask, J., & McIntyre, P. B. (2017). No Jab, No Pay and vaccine refusal in Australia: The jury is out. Medical Journal of Australia, 206(9), 381-383.

Betsch, C., Schmid, P., Heinemeier, D., Korn, L., Holtmann, C., & Böhm, R. (2018). Beyond confidence: Development of a measure assessing the 5C psychological antecedents of vaccination. PLoS ONE, 13(12), Article e0208601.

Biddle, N., Edwards, B., Gray, M., & Sollis, K. (2021). Vaccine willingness and concerns in Australia. Australian National University Centre for Social Research and Methods.

Department of Health and Aged Care. (2023). Vaccination coverage and the National Immunisation Program.

Dubé, E., Gagnon, D., & MacDonald, N. E. (2015). Strategies intended to address vaccine hesitancy: Review of published reviews. Vaccine, 33(34), 4191-4203.

Larson, H. J., Jarrett, C., Eckersberger, E., Smith, D. M. D., & Paterson, P. (2014). Understanding vaccine hesitancy around vaccines and vaccination from a global perspective: A systematic review of published literature, 2007-2012. Vaccine, 32(19), 2150-2159.

Leask, J., Kinnersley, P., Jackson, C., Cheater, F., Bedford, H., & Rowles, G. (2012). Communicating with parents about vaccination: A framework for health professionals. BMC Pediatrics, 12, Article 154.

MacDonald, N. E. (2015). Vaccine hesitancy: Definition, scope and determinants. Vaccine, 33(34), 4161-4164.

National Health and Medical Research Council. (2018). National statement on ethical conduct in human research.

World Health Organization. (2019). Ten threats to global health in 2019.