The Analytical Angle: Vaccine hesitancy in Pakistan is growing. Here's how it can be tackled

Pakistanis have become less likely to say they will accept a Covid vaccine.

Since the start of the pandemic, Pakistanis have become less likely to say they will accept a Covid vaccine. Saher Asad, Javaeria Qureshi, Mariam Raheem, Taimur Shah, and Basit Zafar look at the findings of a new survey into vaccine hesitancy and suggest how the government could overcome it.

Even rich countries have often struggled to roll out vaccines but in the Global South, things are much worse. In Pakistan, access to jabs has so far been limited to healthcare workers and people over 50. As of March 2021, an average of 0.2 doses had been administered for every 100 people in Pakistan, strikingly lower than the regional average for Asia (4.5 doses for every 100 people).

While it is imperative to sort out the supply issues, there is another side to the picture. How many Pakistanis are willing to take up the vaccines when they become available? A recent survey conducted by Gallup Pakistan and the Pakistan Islamic Medical Association (PIMA) showed that 81% of healthcare workers were willing to get a vaccine, while 19% were not yet convinced about their efficacy. Yet we know little about how willing the general population is to get a jab.

To shed light on people’s intentions, as well as factors that may change their minds, the Economic Vulnerability Assessment (EVA), a survey that has been run by the Centre for Economic Research in Pakistan (CERP) is useful. The EVA is a telephone-based survey that examines the economic impact of Covid-19 on households in Punjab. The first, second and third rounds were conducted in June/July 2020, September/October 2020 and December 2020/January 2021, respectively. The last two waves of the survey asked respondents whether they would get a Covid vaccine if it became available.

Respondents who were earlier unsure are now saying they would not get the vaccine

In the most recent wave, fielded in December 2020/January 2021, 65.7% of respondents said they would get vaccinated (that is, they answered “Yes” or “Absolutely Yes” to the question: “If a vaccine against the coronavirus becomes available, do you plan to get vaccinated?”). Thirty per cent said they would not. The remaining 4.6% were unsure. This wave was conducted when the vaccination campaign for citizens was imminent: Pakistan started jabbing healthcare workers just a few weeks later on February 2, and expanded eligibility to older individuals on March 10. A similar question was asked in wave 2 of the survey, in September/October 2020, when vaccinations were not on the horizon (and so the question was largely hypothetical). As the blue bars in the figure below show, the share of respondents who said they were planning on getting vaccinated was about 67%, not very different from the level in the most recent wave. However, a fifth of the respondents back then were unsure about their plans, showing that respondents who were previously on the fence have almost all decided not to take the vaccine.

*Figure 1: If a vaccine against the coronavirus becomes available, do you plan on getting vaccinated?*

Notably, attitudes towards getting vaccinated do not differ across rural/urban areas, respondents’ education and income levels. Roughly two-thirds of urban and rural respondents say they plan on getting vaccinated. Likewise, 67% of individuals with more than matriculation schooling said they will get vaccinated, which is only slightly higher than the 63% of individuals with lower education levels who say they would do the same.

Nonetheless, people’s willingness to take the vaccine is unsurprisingly higher among those who report being worried about Covid (69% of respondents who said they were worried about the virus said they would take it). The overall trend in intentions is somewhat puzzling since the fraction of households reporting that they were at least moderately worried about the spread of Covid increased from 68% to 73% over the same period. This change in attitudes is seen in sub-populations across the board.

Respondents cite safety as their main concern

So why are nearly a third of respondents unwilling to take the vaccine? We presented respondents with a list of reasons for not getting the vaccine, and asked them to choose all possible reasons. The leading concern in rural and urban areas alike, reported by 54% of respondents, is the safety of the vaccine. Safety worries aside, perceptions about the public health threat Covid poses are worrisome. When asked why they would not get the vaccine, 38% cite their healthy immune system, which they believe will help them fight off the virus without difficulty. Religious concerns seem to be a minor factor. These reasons match up surprisingly well with the two most common reasons cited by US citizens who said they would not or might not get the shot in mid-March, including a feeling that the shot was “still too untested” and that they were “just not concerned about coronavirus”.

*Figure 2: What would be the reasons for not getting vaccinated?*

People who report not wanting the vaccine seem to have made up their minds

So what can policymakers do to increase vaccine acceptance? Studies from various contexts have shown that social influencers and religious leaders can shift attitudes towards public health measures. For example, a large-scale messaging campaign in India featuring a video of Indian-origin Nobel laureate Abhijit Banerjee was effective in increasing compliance with preventive behaviour for Covid-19. Recent findings from Pakistan show that one-on-one engagement with local community leaders, imams in mosques, is likely to have positive spillover effects on swaying the community’s behaviour and curtailing the spread of disease. These results are not limited to South Asia — a randomised experiment conducted in the US also showed that culturally-tailored messaging is likely to reduce knowledge gaps. In our survey, we asked respondents if they would get the vaccine if their favourite political leader or well-known religious leaders did so.

The responses are quite discouraging. The vast majority of the respondents who would refuse the vaccine were unlikely to change their decision. Only 5% of respondents say they would get vaccinated if their favourite political leader did so. The involvement of religious leaders is marginally more effective, with 9% reporting they would follow suit. Interestingly, we find that respondents with higher education have slightly more malleable attitudes.

*Figure 3: If a vaccine against the coronavirus becomes available, do you plan on getting vaccinated if the following leaders get it?*

Respondents who do not plan on getting the jab severely underestimate the share of households around them who will

Respondents who are planning to be vaccinated think that about 57% of the people around them will do the same; those who aren’t think that only 17% of the people around them will. Note that, in our survey, two-thirds of the respondents said they were planning on getting a jab. This means they routinely underestimated the proportion of other people who will do so, and the underestimation is significantly more pronounced among those who are reluctant to have the vaccine themselves.

*Figure 4: Once a vaccine against coronavirus becomes available, what proportion of households in your neighbourhood do you think would be willing to get the vaccine?*


It is concerning that, over time, we have seen the fraction of the population who were undecided or unsure about the vaccine — a fifth of the population — shift largely towards deciding against it. The government of Pakistan should take these findings into account as they build on their Rs1bn vaccination awareness campaign.

  1. Our results suggest that focusing on public announcements or televised clips of government leaders getting vaccinated will likely not be enough. Involving local religious leaders might be more promising.

  2. There is a systematic bias in respondents’ perceptions of the proportion of other people who will get vaccinated. Rather than taking cues from national leaders, people may be waiting to see what people they know and others in their own communities decide. A recent poll conducted in the US found that believing that those around them want to get vaccinated strongly predicts people's vaccine uptake or intended uptake. It is therefore imperative that the government communicates coverage rates in sub-populations as the campaign moves ahead. As well as enlisting local leaders who can act as agents of change, they should experiment with various grassroots campaigns. The goal should be to encourage people to talk about their plans to get vaccinated and share their experiences with their friends and family. Using hashtags on social media or giving out “I got the shot” or “Pakistan vaccinates” stickers or bracelets can create a nudge similar to that induced by “I voted” stickers or inked thumbprints. Such social signals have been quite effective in increasing vaccination rates in other contexts.

  3. Given that safety concerns are the main reason for not wanting to get a vaccine, efforts are needed to reassure the population about safety and to educate them about normal side effects. Here, too, having people share these messages with their friends and networks will be more effective than relying on national leaders to give reassurance.

  4. As a growing fraction of the population gets vaccinated, without any notable incidents of severe side effects linked to the vaccine, hesitancy may diminish. However, it is important that the government does not rely on this alone, since it is not guaranteed and will likely result in large disparities in vaccine hesitancy. Fortunately, vaccine hesitancy is not correlated with socioeconomic background so far, but we find that attitudes seem to be more malleable among more educated individuals. These people will be more likely to know more people who have had the vaccine, and so may update their priors on concerns about vaccine safety. People with lower education and in rural areas, where the rollout has not yet begun, may have fewer opportunities to update their priors and have their worries assuaged. Targeted efforts should be made to reach out to groups with more stubborn attitudes or fewer opportunities to learn.

The Analytical Angle is a monthly column where top researchers bring rigorous evidence to policy debates in Pakistan. The series is a collaboration between the Centre for Economic Research in Pakistan and The views expressed are the authors’ alone.