"No one has enough food to eat, and the food prices are soaring. Most are quarantined in their homes. Until now, no one has asked about the disease that is ravaging my country and killing my friends and family" — Liberian responses to Gallup interviews during the 2014 Ebola outbreak
COVID-19 has infected over 431 million people and has killed 6 million people as of this writing. But the real death toll might be even higher. According to advanced modeling by The Economist using “excess deaths,” the actual figure could be as high as 23 million deaths.
The pandemic also wrecked the global economy, which shrank by 3.5% in 2020 — the largest global contraction in 60 years, according to the IMF. That same year, nearly a third (32%) of workers globally told Gallup they lost their job, and 50% said they earned less money.
So how were people’s lives going during the pandemic? Most of what we learned will not surprise you, but one thing will.
The COVID-19 pandemic impacted almost everyone: 80% of people worldwide told Gallup in 2020 that the pandemic affected their lives, including 45% who said it affected their lives a lot (the most extreme response). The poor, women with children, and the middle-aged were affected most.
And the impact on women with children was particularly acute. Nearly half of women with children younger than 15 (49%) said their lives were affected a lot by the pandemic. They also experienced more stress, sadness, worry, physical pain, and anger compared with all other adults.
These findings confirm conventional wisdom. But here is the surprise. Although people experienced their lives worse in 2020, they rated their lives better. The number of people who rated their lives high enough to be considered thriving increased.
Did people’s views of their lives really improve during the pandemic? If so, it would be a massive finding. But it’s hard to believe that people rated their lives better during one of the worst years in recent history. There must be another explanation.
Here are two possible reasons for such a shift: a change in Gallup’s methodology and the timing of our surveys.
A World Poll During a Pandemic
Gallup conducts the majority of our global surveys face to face. But in-person interviewing during a pandemic is just not possible. However, we still need to ask people how their lives are going, especially at such a critical time.
Gallup has faced something like this before.
In December 2013, an 18-month-old boy came down with a mysterious illness in a Guinean village. Scientists believe bats infected him. Soon after, five more people died from similar symptoms. The government issued a health warning on January 24, 2014.
Seven weeks later, the illness reached Guinea’s capital, Conakry. By mid-March, there were 49 infections and 29 deaths. That same month, the Pasteur Institute in France confirmed what the illness was—the Ebola virus.
Over the next three years, the virus spread to nine more countries—infecting 28,652 people and killing 11,325. One of the most affected countries was Liberia. To better inform its decision-making as the crisis was unfolding, the Liberian government wanted near real-time economic data. But how do you conduct surveys when a virus is raging?
The preferred methodology—face-to-face interviewing—was off the table. To reach Liberians, we had to do something different. So, the World Bank, Gallup, and the Liberia Institute of Statistics and Geo-Information Services (LISGIS) came up with a solution.
LISGIS had access to Liberians’ phone numbers, and the national language in Liberia is English. So English-speaking Gallup interviewers could call Liberians from the U.S.
The project worked. We conducted five studies that helped gauge the economic environment throughout the outbreak. While the economic insights informed policy, people’s stories from the field, such as the one at the beginning of this chapter, had a profound impact on those conducting the research. Sadly, comments like these were common:
Please just stay on the phone with me; I need someone to keep me company.
My family and I are just waiting to die.
I would rather experience war.
Conducting a global survey in 2020 was a lot like our experience in Liberia in 2014. We had to change the way we did everything. 2020 was harder because instead of pivoting our methodology in one country, we had to do it for the entire world.
At a moment’s notice, we changed our global methodology from 30% phone and 70% in-person to nearly 100% phone. This shift in methodology may have contributed to the changes in how people see their lives.
For example, in Egypt and Ethiopia, we moved interviewing from face to face to phone in 2020. And thriving unexpectedly increased in both countries in 2020. What changed in Egypt from 2019 to 2020 that would have caused people to see their lives better? Ethiopia is even more perplexing. Thriving reached its highest point in our nine years of tracking despite the country descending into a civil war with the Tigray region.
Egypt and Ethiopia were not the only developing countries that saw an increase. Thriving rates rose an average of five points in developing countries between 2019 and 2020.
In wealthier countries, it was different. There, Gallup saw little change in life ratings, and the European Commission (EC) found similar results. Using the Eurobarometer, the EC looked at life ratings in 30 European countries and found that Europeans rated their lives the same from 2019 to 2020—exactly what Gallup found in Europe.
So, from 2019 to 2020, life ratings remained the same in many rich countries but improved in many developing countries. Why?
The only places where we switched from face-to-face to phone interviewing were developing countries; in wealthy countries, we continued to conduct interviews over the phone.
Further, when we switched back to face-to-face interviewing in some countries in 2021, we saw thriving decline. In Egypt, for example, thriving went back down to 8% (from 18% in 2020). Unfortunately, changing how you ask people questions can change their answers.
And the science behind this is not new.
One of the largest migrations of survey research from in-person to phone interviewing took place in the 1950s. In wealthy countries, survey researchers stopped showing up at people’s homes and began calling them instead. Some of the most groundbreaking research focused on these differences known as “mode effects.” Georgetown professor and former head of the U.S. Census Bureau, Robert Groves, even co-wrote one of the initial texts about the differences between the two methodologies in the 1979 book Surveys by Telephone: A National Comparison with Personal Interviews.
Switching from in-person to phone interviews can change the outcome of your surveys. According to academics Allyson Holbrook, Melanie Green, and Jon Krosnick, “[Phone] respondents were more likely to satisfice, to be less cooperative and engaged in the interview, and were more likely to express dissatisfaction with the length of the interview. … Telephone respondents were also more suspicious about the interview process and more likely to present themselves in socially desirable ways than were face-to-face respondents.”
The Pew Research Center found similar mode effects for phone surveys compared with web surveys. They found that people rate their lives higher on the phone (compared with web) but report experiencing life (for example, positive emotions) the same. Mode adjustments might disproportionately affect the way people report how they see life compared with how they live life.
Other studies suggest that mode switches do not dramatically affect life ratings. The U.K. government tested this idea by simultaneously conducting in-person and phone surveys using the same questions. If mode effects were present, the impact on how people rate their lives was negligible. People rated their lives 0.04 points higher on the phone compared with in-person surveys.
Just because mode changes did not affect results in the U.K. does not mean they will not influence results somewhere else in the world. For example, following COVID-19 protocols, Gallup conducted some interviews in India using face-to-face methodology. Although we saw an increase in life ratings from 2019 to 2020 in the phone and face-to-face interviews, Indians tended to rate their lives higher on the phone.
But if the methodology change is what caused people to rate their lives differently, then why was everything else in the results of our survey so accurate? Not only did the rest of our data highly corroborate what other surveys were finding, our resulting data also perfectly aligned with conventional wisdom: Almost everyone was affected by the pandemic, especially the poor, the middle-aged, and women with children.
While the change in methodology might have affected our results in developing countries, there may have been another reason people rated their lives better—the timing of our surveys.
The Timing of a Survey
In 2006, only 12% of people in China were thriving. In 2019, before COVID-19 started to spread, that figure was 19%. In our 2020 survey, we found a massive increase. Almost one-third (31%) of people in China were thriving—the most in the history of our tracking there.
Like in most countries, we changed our methodology in 2020 from face to face to phone. The survey began in mid-September and lasted for about six weeks—a few months after Wuhan emerged from its lockdown in early April. That fall, it looked like most of China was back to life as usual. Citizens of other countries, who were still in lockdown, saw images of Chinese citizens participating in large events such as concerts. In August of 2020, the BBC ran an article titled “Wuhan coronavirus: From silent streets to packed pools.”
While livelihoods appeared to be improving, so was China’s economy, which grew 2.3% in 2020 despite the pandemic. So, people may truly have been rating their lives higher at that time.
In terms of how people lived life in China, there were also stark differences from 2019 to 2020. Negative emotions increased 10 points—a highly unusual increase for any country. Positive emotions were also affected, dropping nine points. In fact, the decline in laughing and smiling in China was one of the steepest worldwide.
Negative and positive emotions moved predictably in China, but why didn’t life ratings?
To further explore how perplexing these shifts are, let’s look at another country that was in the news frequently in 2020—Sweden.
Sweden gained international attention for being one of the few countries in the world that did not immediately go into a lockdown. The country’s strategy at the onset of the pandemic was to adopt herd immunity. The government eventually implemented a lockdown of its own, but there was no statistical change in how people rated their lives in 2020 compared with 2019.
Again, the timing of the survey matters. The 2020 survey took place during April, which was one of the months when COVID reached a peak for deaths in Sweden.
The results in Sweden are among the most difficult to explain. The survey modes were the same from 2019 to 2020, and many Europeans rated their lives the same year over year (as was true in the EC’s data). One of the reasons the numbers did not decline might be because we did not conduct surveys frequently enough.
Looking at how people rate their lives once per year is like capturing GDP during only one quarter of the year. Imagine measuring GDP during the second quarter of 2020 and considering that the measurement for the entire year. The drop in the official statistics would have been spectacular. But that drop would not have accurately described the economic reality for the whole year. Yet, in most countries, that is what we did to measure how people’s lives were going.
Ideally, we would track wellbeing in every country at least quarterly, especially during a time of monumental change like that brought on by a pandemic. We have tested a greater frequency of tracking in the U.S. since 2008 and have seen meaningful shifts at key moments. Here is the complete trend:
Notice the volatility when you increase the frequency of measurement. More than half (52%) of people in the U.S. were thriving in early 2008, immediately before the global economic crisis. When America’s economy collapsed, so did thriving—dropping almost six points. When the U.S. emerged from the financial crisis, thriving not only rebounded, it exceeded the 52% recorded before the 2008 economic crisis, reaching 57% in late 2017.
When COVID-19 spread to the U.S., thriving collapsed, just like it did during the 2008 financial crisis, hitting the identical low point of 46%. As the economy rebounded, so did thriving, climbing to 54%. As COVID persisted, thriving dropped again to 48%. But when people started getting vaccinated and some lockdowns were lifted, U.S. thriving reached a high never before seen in our tracking (59%).
Would more data points globally have better explained why people rated their lives better? Did an abrupt methodological switch because of the pandemic cause a change in how people see their lives? Or did the pandemic cause people, especially in developing countries, to reflect better on their lives than they had previously?
We do not know for sure. But when we do know more definitively, this knowledge will help advance wellbeing research tremendously.
During the 2014 Ebola crisis, a Gallup interviewer said to us:
We have received multiple affirmations from Liberians that they are so grateful we have called. Until now, no one has asked them how they feel about the disease that is ravaging their country. So, the importance and significance of this work has been confirmed by the very people we are calling … Even though it is hard … we will continue this work to give a voice to the Liberian people. My hope is that voice will be heard loud and clear around the world.
We approached 2020 with this same aspiration.
a global affairs media network
How the Pandemic Shaped Happiness
Photo by Midjourney.
September 25, 2022
While 80% of people worldwide told Gallup in 2020 that the pandemic affected their lives, the number of people who rated their lives high enough to be considered thriving increased. This is a shift that could be caused by two possible reasons, writes Gallup CEO Jon Clifton.
"No one has enough food to eat, and the food prices are soaring. Most are quarantined in their homes. Until now, no one has asked about the disease that is ravaging my country and killing my friends and family" — Liberian responses to Gallup interviews during the 2014 Ebola outbreak
COVID-19 has infected over 431 million people and has killed 6 million people as of this writing. But the real death toll might be even higher. According to advanced modeling by The Economist using “excess deaths,” the actual figure could be as high as 23 million deaths.
The pandemic also wrecked the global economy, which shrank by 3.5% in 2020 — the largest global contraction in 60 years, according to the IMF. That same year, nearly a third (32%) of workers globally told Gallup they lost their job, and 50% said they earned less money.
So how were people’s lives going during the pandemic? Most of what we learned will not surprise you, but one thing will.
The COVID-19 pandemic impacted almost everyone: 80% of people worldwide told Gallup in 2020 that the pandemic affected their lives, including 45% who said it affected their lives a lot (the most extreme response). The poor, women with children, and the middle-aged were affected most.
And the impact on women with children was particularly acute. Nearly half of women with children younger than 15 (49%) said their lives were affected a lot by the pandemic. They also experienced more stress, sadness, worry, physical pain, and anger compared with all other adults.
These findings confirm conventional wisdom. But here is the surprise. Although people experienced their lives worse in 2020, they rated their lives better. The number of people who rated their lives high enough to be considered thriving increased.
Did people’s views of their lives really improve during the pandemic? If so, it would be a massive finding. But it’s hard to believe that people rated their lives better during one of the worst years in recent history. There must be another explanation.
Here are two possible reasons for such a shift: a change in Gallup’s methodology and the timing of our surveys.
A World Poll During a Pandemic
Gallup conducts the majority of our global surveys face to face. But in-person interviewing during a pandemic is just not possible. However, we still need to ask people how their lives are going, especially at such a critical time.
Gallup has faced something like this before.
In December 2013, an 18-month-old boy came down with a mysterious illness in a Guinean village. Scientists believe bats infected him. Soon after, five more people died from similar symptoms. The government issued a health warning on January 24, 2014.
Seven weeks later, the illness reached Guinea’s capital, Conakry. By mid-March, there were 49 infections and 29 deaths. That same month, the Pasteur Institute in France confirmed what the illness was—the Ebola virus.
Over the next three years, the virus spread to nine more countries—infecting 28,652 people and killing 11,325. One of the most affected countries was Liberia. To better inform its decision-making as the crisis was unfolding, the Liberian government wanted near real-time economic data. But how do you conduct surveys when a virus is raging?
The preferred methodology—face-to-face interviewing—was off the table. To reach Liberians, we had to do something different. So, the World Bank, Gallup, and the Liberia Institute of Statistics and Geo-Information Services (LISGIS) came up with a solution.
LISGIS had access to Liberians’ phone numbers, and the national language in Liberia is English. So English-speaking Gallup interviewers could call Liberians from the U.S.
The project worked. We conducted five studies that helped gauge the economic environment throughout the outbreak. While the economic insights informed policy, people’s stories from the field, such as the one at the beginning of this chapter, had a profound impact on those conducting the research. Sadly, comments like these were common:
Please just stay on the phone with me; I need someone to keep me company.
My family and I are just waiting to die.
I would rather experience war.
Conducting a global survey in 2020 was a lot like our experience in Liberia in 2014. We had to change the way we did everything. 2020 was harder because instead of pivoting our methodology in one country, we had to do it for the entire world.
At a moment’s notice, we changed our global methodology from 30% phone and 70% in-person to nearly 100% phone. This shift in methodology may have contributed to the changes in how people see their lives.
For example, in Egypt and Ethiopia, we moved interviewing from face to face to phone in 2020. And thriving unexpectedly increased in both countries in 2020. What changed in Egypt from 2019 to 2020 that would have caused people to see their lives better? Ethiopia is even more perplexing. Thriving reached its highest point in our nine years of tracking despite the country descending into a civil war with the Tigray region.
Egypt and Ethiopia were not the only developing countries that saw an increase. Thriving rates rose an average of five points in developing countries between 2019 and 2020.
In wealthier countries, it was different. There, Gallup saw little change in life ratings, and the European Commission (EC) found similar results. Using the Eurobarometer, the EC looked at life ratings in 30 European countries and found that Europeans rated their lives the same from 2019 to 2020—exactly what Gallup found in Europe.
So, from 2019 to 2020, life ratings remained the same in many rich countries but improved in many developing countries. Why?
The only places where we switched from face-to-face to phone interviewing were developing countries; in wealthy countries, we continued to conduct interviews over the phone.
Further, when we switched back to face-to-face interviewing in some countries in 2021, we saw thriving decline. In Egypt, for example, thriving went back down to 8% (from 18% in 2020). Unfortunately, changing how you ask people questions can change their answers.
And the science behind this is not new.
One of the largest migrations of survey research from in-person to phone interviewing took place in the 1950s. In wealthy countries, survey researchers stopped showing up at people’s homes and began calling them instead. Some of the most groundbreaking research focused on these differences known as “mode effects.” Georgetown professor and former head of the U.S. Census Bureau, Robert Groves, even co-wrote one of the initial texts about the differences between the two methodologies in the 1979 book Surveys by Telephone: A National Comparison with Personal Interviews.
Switching from in-person to phone interviews can change the outcome of your surveys. According to academics Allyson Holbrook, Melanie Green, and Jon Krosnick, “[Phone] respondents were more likely to satisfice, to be less cooperative and engaged in the interview, and were more likely to express dissatisfaction with the length of the interview. … Telephone respondents were also more suspicious about the interview process and more likely to present themselves in socially desirable ways than were face-to-face respondents.”
The Pew Research Center found similar mode effects for phone surveys compared with web surveys. They found that people rate their lives higher on the phone (compared with web) but report experiencing life (for example, positive emotions) the same. Mode adjustments might disproportionately affect the way people report how they see life compared with how they live life.
Other studies suggest that mode switches do not dramatically affect life ratings. The U.K. government tested this idea by simultaneously conducting in-person and phone surveys using the same questions. If mode effects were present, the impact on how people rate their lives was negligible. People rated their lives 0.04 points higher on the phone compared with in-person surveys.
Just because mode changes did not affect results in the U.K. does not mean they will not influence results somewhere else in the world. For example, following COVID-19 protocols, Gallup conducted some interviews in India using face-to-face methodology. Although we saw an increase in life ratings from 2019 to 2020 in the phone and face-to-face interviews, Indians tended to rate their lives higher on the phone.
But if the methodology change is what caused people to rate their lives differently, then why was everything else in the results of our survey so accurate? Not only did the rest of our data highly corroborate what other surveys were finding, our resulting data also perfectly aligned with conventional wisdom: Almost everyone was affected by the pandemic, especially the poor, the middle-aged, and women with children.
While the change in methodology might have affected our results in developing countries, there may have been another reason people rated their lives better—the timing of our surveys.
The Timing of a Survey
In 2006, only 12% of people in China were thriving. In 2019, before COVID-19 started to spread, that figure was 19%. In our 2020 survey, we found a massive increase. Almost one-third (31%) of people in China were thriving—the most in the history of our tracking there.
Like in most countries, we changed our methodology in 2020 from face to face to phone. The survey began in mid-September and lasted for about six weeks—a few months after Wuhan emerged from its lockdown in early April. That fall, it looked like most of China was back to life as usual. Citizens of other countries, who were still in lockdown, saw images of Chinese citizens participating in large events such as concerts. In August of 2020, the BBC ran an article titled “Wuhan coronavirus: From silent streets to packed pools.”
While livelihoods appeared to be improving, so was China’s economy, which grew 2.3% in 2020 despite the pandemic. So, people may truly have been rating their lives higher at that time.
In terms of how people lived life in China, there were also stark differences from 2019 to 2020. Negative emotions increased 10 points—a highly unusual increase for any country. Positive emotions were also affected, dropping nine points. In fact, the decline in laughing and smiling in China was one of the steepest worldwide.
Negative and positive emotions moved predictably in China, but why didn’t life ratings?
To further explore how perplexing these shifts are, let’s look at another country that was in the news frequently in 2020—Sweden.
Sweden gained international attention for being one of the few countries in the world that did not immediately go into a lockdown. The country’s strategy at the onset of the pandemic was to adopt herd immunity. The government eventually implemented a lockdown of its own, but there was no statistical change in how people rated their lives in 2020 compared with 2019.
Again, the timing of the survey matters. The 2020 survey took place during April, which was one of the months when COVID reached a peak for deaths in Sweden.
The results in Sweden are among the most difficult to explain. The survey modes were the same from 2019 to 2020, and many Europeans rated their lives the same year over year (as was true in the EC’s data). One of the reasons the numbers did not decline might be because we did not conduct surveys frequently enough.
Looking at how people rate their lives once per year is like capturing GDP during only one quarter of the year. Imagine measuring GDP during the second quarter of 2020 and considering that the measurement for the entire year. The drop in the official statistics would have been spectacular. But that drop would not have accurately described the economic reality for the whole year. Yet, in most countries, that is what we did to measure how people’s lives were going.
Ideally, we would track wellbeing in every country at least quarterly, especially during a time of monumental change like that brought on by a pandemic. We have tested a greater frequency of tracking in the U.S. since 2008 and have seen meaningful shifts at key moments. Here is the complete trend:
Notice the volatility when you increase the frequency of measurement. More than half (52%) of people in the U.S. were thriving in early 2008, immediately before the global economic crisis. When America’s economy collapsed, so did thriving—dropping almost six points. When the U.S. emerged from the financial crisis, thriving not only rebounded, it exceeded the 52% recorded before the 2008 economic crisis, reaching 57% in late 2017.
When COVID-19 spread to the U.S., thriving collapsed, just like it did during the 2008 financial crisis, hitting the identical low point of 46%. As the economy rebounded, so did thriving, climbing to 54%. As COVID persisted, thriving dropped again to 48%. But when people started getting vaccinated and some lockdowns were lifted, U.S. thriving reached a high never before seen in our tracking (59%).
Would more data points globally have better explained why people rated their lives better? Did an abrupt methodological switch because of the pandemic cause a change in how people see their lives? Or did the pandemic cause people, especially in developing countries, to reflect better on their lives than they had previously?
We do not know for sure. But when we do know more definitively, this knowledge will help advance wellbeing research tremendously.
During the 2014 Ebola crisis, a Gallup interviewer said to us:
We have received multiple affirmations from Liberians that they are so grateful we have called. Until now, no one has asked them how they feel about the disease that is ravaging their country. So, the importance and significance of this work has been confirmed by the very people we are calling … Even though it is hard … we will continue this work to give a voice to the Liberian people. My hope is that voice will be heard loud and clear around the world.
We approached 2020 with this same aspiration.
