COVID-19 vaccine hesitancy and resistanceCorrelates in a nationally representative longitudinal survey of the Australian population

Fifty-nine percent of Australians said they would definitely get a COVID-19 vaccine. That sounds like good news — until you do the arithmetic. Herd immunity requires roughly seventy-five percent coverage, assuming a vaccine around eighty percent effective. The gap between those two numbers is sixteen percentage points. And sixteen percentage points, at population scale, is the difference between a pandemic that ends and one that keeps finding its next host. That gap is exactly what Ben Edwards, Nicholas Biddle, Matthew Gray, and Kate Sollis set out to map. Their paper, published from a nationally representative longitudinal survey of over three thousand Australian adults, asks a question that is deceptively simple: who fills that gap, and why? The answer turns out to hinge less on biology and more on belief. To understand the stakes, it helps to know that the seventy-five percent figure isn't a single fixed target. Edwards and colleagues walk through the math: with a perfectly effective vaccine, you'd need about sixty-seven percent coverage; with a vaccine that's only sixty percent effective, you'd theoretically need to vaccinate the entire population. The seventy-five percent estimate assumes a vaccine around eighty percent effective — a reasonable baseline for the vaccines then in development. So the shortfall isn't academic. It's structural. And the researchers found that even within that shortfall, attitudes aren't uniformly distributed. Six percent of respondents said they would definitely not get vaccinated — resistant, full stop. Seven percent said probably not, marking high hesitancy. Another twenty-nine percent said probably yes, the low-hesitant group. Only fifty-eight point five percent were in the "definitely" column. The World Health Organization's advisory group had already called on governments to respond proactively to hesitancy hotspots using social and behavioral insights. This study was built to produce exactly those insights. The data came from the August ANUpoll, drawn from the Life in Australia panel. The sample of three thousand sixty-one adults was weighted to match the Australian population on key demographic and geographic benchmarks, with ninety-four percent of interviews conducted online and the rest by telephone. Vaccine intention was measured with a single question: if a safe and effective vaccine for COVID-19 is developed, would you get it? Four ordered response options — from definitely not to definitely yes. The researchers analyzed these as an ordered spectrum using a statistical model called an ordered probit, which estimates how various characteristics shift the probability of a respondent falling into each category. They ran five nested models, progressively adding health variables, COVID-related behaviors, and political and social attitudes, ultimately reporting marginal effects for each outcome category — meaning the results are readable as percentage-point shifts in probability. Now to the findings. On the hesitant and resistant side, three clusters emerge: demographic, attitudinal, and political. Demographically, being female was one of the clearest predictors. In the final model, being female reduced the probability of being in the "definitely vaccinate" category by six point one percentage points and raised the probability of resistance by one point eight points. Living in the most disadvantaged neighborhood quintile had an even larger effect: it raised the probability of resistance by three point three points and reduced the probability of definite vaccination by eight point four points. Higher household income cut in the other direction, making wealthier respondents more likely to intend to vaccinate, though the effect per dollar was small. The attitudinal correlates were the most powerful. Respondents who believed too much fuss was being made about COVID-19 were thirteen point five percentage points less likely to say they'd definitely get vaccinated, five point nine points more likely to be outright resistant, and three point eight points more likely to be highly hesitant. That's the single largest effect in the model. Religiosity also predicted hesitancy and resistance, though the authors flag that this finding should be treated with caution — statistical significance was only at the ninety percent confidence level rather than the more standard ninety-five percent. Then there's populism. The paper included a measure of populist attitudes — broadly capturing anti-establishment sentiment — and found that higher populist sentiment was associated with greater resistance and hesitancy and a lower probability of being in the "definitely vaccinate" group. The effects weren't enormous in isolation, but they fit a pattern. Flip the lens and the pattern becomes even clearer. On the acceptance side, what predicts willingness to vaccinate? Higher household income, yes. But the more telling predictors are behavioral and institutional. Respondents who had downloaded the COVIDSafe app — Australia's contact-tracing application — were substantially more likely to intend to vaccinate, with the association amounting to an eleven percentage-point higher likelihood. Those who reported greater adherence to social distancing were seven percentage points more likely to be in the "definitely" column for every one standard deviation increase in that behavior. Confidence in state or territory government, confidence in hospitals, and more supportive attitudes toward migration all predicted greater vaccine acceptance. What unifies those acceptance-side predictors is institutional trust and prior civic engagement. The people who were already following public health guidance, who trusted the institutions delivering that guidance, and who had opted into a government app were the same people most likely to step forward for vaccination. The divide isn't primarily about age or geography. It's about orientation toward public institutions. That finding reshapes the policy question. If hesitancy were primarily demographic — a matter of reaching the young, the rural, the less-educated — the solution would be logistical: more clinics, more outreach, better access. But when hesitancy maps onto beliefs about COVID-19's severity, populist skepticism of government, and distrust of health systems, the solution becomes harder. You can move a clinic. You cannot easily move a worldview. Edwards and colleagues are clear about this distinction. For the hesitant — the thirty-six percent who sit in the "probably not" and "probably yes" categories — public health messaging has genuine purchase. Countering misinformation, building trust through community voices, using doctors and local leaders to deliver targeted messaging: these are established tools, and the behavioral correlates in the data point to where they should be deployed. The COVIDSafe app association is a useful diagnostic: communities with low app uptake and low social distancing compliance are communities where hesitancy is likely to cluster. For the resistant six percent, the paper is more cautious. It notes that "alternative policy measures may well be needed" to close the gap, and acknowledges that a systematic review found majority public support for compulsory vaccination in general — but immediately qualifies that none of those studies were conducted during a pandemic with already-restricted civil liberties. The tone is measured. That is itself a form of honesty about the limits of the evidence. There are real limits to acknowledge. This study measures intentions in August two thousand twenty, before any vaccine was approved, before efficacy data were public, and before the specific vaccines Australians would be offered were known. Intentions and behavior don't always align. The survey also didn't capture some potentially important variables — prior vaccination history, specific safety concerns — that might refine the picture. But the core finding holds regardless of those caveats. The bottleneck to ending the COVID-19 pandemic, in Australia as in many countries, was not a biological one. Vaccines were coming that were highly effective. The bottleneck was belief. And this study gave policymakers something concrete: a map of who holds what beliefs, and what attitudes sit underneath them. Knowing that the resistant and hesitant cluster around distrust of institutions, around the conviction that the pandemic was overstated, and around populist skepticism of government isn't just academically interesting. It tells you that a public health campaign built around government authority is likely to backfire with exactly the audience it most needs to reach. Closing a sixteen percentage-point gap before a virus finds the people it would otherwise kill is, at its core, a political and social challenge — not a logistical one. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.