Understanding the poll

All over the world there are people working to help their communities understand and respond to risks – from radio presenters in farming communities affected by floods to restaurant owners protecting city economies from COVID-19. Sense about Science and the Lloyd’s Register Foundation are working with many of them to understand their needs. We sought their help to create this guidance for understanding the World Risk Poll.

I. About the World Risk Poll

Who was asked? The sample

In most of the 142 countries and territories where the World Risk Poll was conducted, researchers spoke to a sample of around 1000 people aged 15 or above through telephone or face-to-face interviews. They spoke to larger samples in China, Russia and India, which have large populations.

The samples were nationally representative; they closely match certain characteristics of the entire nation’s adults, such as the number of people in each age group. This means the sample’s responses to questions about risks and perceptions are likely to be similar to the way everyone would answer them.

Representativeness makes it possible to compare countries and groups. A sample of 1000 per country includes enough people of middle-income level, for example, to be confident their views are representative and can be compared with the same group in another country. But only some comparisons are possible. For example, the World Risk Poll found that big categories such as Western Europeans, older people, and people who’ve received more education were more likely to worry about internet fraud. But it cannot compare the worry levels of Filipino and Mauritian women aged 15-29 about internet fraud, because that category is too small in the poll sample to be confident that it represents their views.

In some cases, the World Risk Poll targeted questions to the relevant populations. Questions related to the internet were asked to people who had been online in the previous 30 days, and safety at work questions only asked to employed people. This is to avoid misleading results from people answering questions about experiences they have not had.


We can group people in many ways—for example, by their age, their sex, how much education they’ve received or how much money they make. These different factors are used to understand how certain populations act, which helps create targeted action; the finding that men, especially young men, suffer more than women from physical injuries at work can lead to more efforts to ensure young men enroll in safety training.

Demographic analysis is a central component of the World Risk Poll. You can select age and sex as demographic variables in the interactive graphs on each subject page. By clicking the check boxes on the right of the screen, you can select age and gender as demographic variables in the interactive graphs on each subject page. Using the menu to select what is displayed, you can see how the results change when you look just at women, or people aged 50-64.

In a climate change example, the colours depict how serious a threat it is seen as in each country, showing how that varies with age and sex.

II. Using the World Risk Poll results to think about risk in different communities

Risk is complicated. As the World Risk Poll shows, risk can be viewed both positively and negatively, depending on culture and language. The poll explores perceptions. People’s perceptions of risk are subjective and influenced by lots of factors, but that doesn’t mean they are not useful. In fact, perceptions can tell us a lot: by comparing risk perception with self-reported experience of risks and with recorded incidents (the incidents that make it into official statistics), we can understand the gaps between these three measures, and investigate the reasons behind them.

These measures – of risk incidents, risk experience and risk perception – and gaps between them can also be mapped over time, which means that even if we don’t understand all the factors behind perceptions, we can see how they’re changing. The World Risk Poll is a ten-year project, so it will show some of these trends. The next set of interviews will be conducted in 2021 and could show how, for example, COVID-19 has influenced perceptions and experiences of risk.

If you engage with people in your community about risk, you will be very aware that values and culture and context affect how risks are perceived and communicated. The global commentary on risk can feel distant. The World Risk Poll’s focus on perceptions and experiences makes international discussions about risk easier to relate to.

Examples highlighted by our community collaborators included:

  • The World Risk Poll shows that perceptions of climate change risk are highly affected by whether people have experienced a severe weather event. But there are big differences within groups who have similar experiences. In Japan, older people were more worried than younger people. Neighbouring countries Lesotho and South Africa had a difference of 20%.
  • The Safety at Work finding that farming and fishing are experienced as the most dangerous jobs gives international affirmation to very local experiences and might help people make a common cause.
  • The World Risk Poll findings give a picture of the world before the COVID-19 pandemic. People everywhere are thinking about how their lives are changing and how other people are coping. There will be a lot of interest in this aspect of the next wave of polling.  

From rural Kisumu, Kenya to urban Boston, USA, our collaborators on the uses of the World Risk Poll wanted to think about the findings on people’s risk experiences alongside the possible solutions that are manageable within the context of their community.

In both perception and action, people are not confronted by an issue in isolation but by risk trade-offs – including the costs of tackling the risk – as well as long and short term effects and the particular impact of the consequences for them.  

Risk trade-offs

We all weigh risks up against each other. Eliminating one risk usually means being exposed to a different one. Going to work exposes people to threats to safety in the workplace, but people trade these against the risks and consequences of unemployment.

The fishermen in our community in Bangladesh will go out sea fishing even with a cyclone or a storm approaching, because the risk of not catching fish and not being able to eat feels more present.

Dr Sazedul Hoque, Bangladesh

People are capable of complex and nuanced risk calculations on many topics. But we often also have blind spots where we overestimate or underestimate the risks or the benefits. As a chef, I see this with people choosing to cut out food groups, overestimating the benefit this might have and underestimating the risks it could pose to their health.

Anthony Warner, UK

Short- vs long-term risks

The subjects covered in the World Risk Poll range from immediate to long-term threats. People’s perception of them is likely to reflect things like whether they live day to day or plan for the future and not just whether they receive information or have experience of them.

We have found that it can be hard even for our local policy-makers to understand the risks and communicate why issues, such as addressing the risks that rising sea levels pose, matter now rather than leaving it for the future. Especially when there are so many immediate needs.

Dr Sazedul Hoque, Associate Professor at Patuakhali Science and Technology University, Bangladesh


Perception of risk may similarly be influenced by how often people think something will happen. While statistics might show that a bad storm will happen on average once every 20 years, this might occur as two close together and nothing for 39 years, making it outside of lived experience for many people. This is important for understanding that people might feel safe even when they are not safe, and conversely that a relatively rare hazard might still mean some people’s actual experience is of great losses or harms coming close together.

Weighting risks

The World Risk Poll’s Worry Index uses the poll’s findings to summarize a person’s overall level of worry over seven everyday hazards: food, water, violent crime, severe weather, electrical power lines, household appliances and mental health. The Experience Index summarizes the combined harm they, or someone they know, experienced in the two years before polling in these areas. The hazards and experiences are given different weight in the indices to take account of the degrees of seriousness.

This weighting of risks makes it possible to come up with a combined figure using the same formula for each country and group, which then means they can be compared. This is not the same as saying that risks are the same in different places. For example, drinking water may be a low-level risk for people in developed countries, who are used to clean drinking water. For people in other parts of the world, however, drinking water may be life-threatening.

Even in our region of southeast USA, different farmers experience risks differently. It matters whether they own the land or rent it, what crops they are planting, the geography of where they are and even their values. For some what matters is to make the most yield out of that year’s crop, for others the biggest fear is losing their land.

Pam Knox, agricultural climatologist, USA

People’s worries about risk reflect not just their understanding and experience, but also their priorities and which behaviours they’re willing or able to change – such as what crops to plant or how to build a house. Those who work to engage communities in conversations about risk therefore often focus on the benefits of specific actions and pay a lot of attention to what makes people feel safer and the experiences that they are drawing on. The World Risk Poll provides a systematic picture of these perceptions and experiences and in doing so makes it possible, for the first time, to consider how other communities manage similar conversations and act to make themselves safer.

In our community the rainfall patterns have changed so we need to plant different crops that grow quicker and adapt to these new conditions.

Bernard Okebe, Community Empowerment and Media Initiative in Kisumu, Kenya

When we arrived in Fukushima, six months after the nuclear accident the people there told us they had had plenty of scientific lectures on radiation and associated risks but they still didn’t know whether they could drink the tap water or eat the fish from the market.

Dr Mariko Nishizawa, Associate Member of the Science Council of Japan

When we shame people as a way to try to get them to avoid risky behaviors, it doesn’t generally make the behaviors stop — it just makes people want to hide the behaviors. Rather than shaming, what we can do is try to meet people where they are and engage with empathy.

Dr Julia Marcus, infectious disease epidemiologist at Harvard Medical School and Harvard Pilgrim Health Care Institute, USA

Working in Nepal after the 2015 earthquakes, the communities saw the concrete buildings that had survived the earthquakes as the buildings they would feel safe living in, but as an engineer my perception of the risk was very different. In a different earthquake many of those would not have been safe either. The buildings we designed had to be structurally sound and also make the community feel safe in them, as they were the ones that had to live in the buildings. Dr Alejandra Albuerne, UK
World Risk Poll Report