In 2009, the Greek economy crashed, sending the country into a severe recession. The financial failure also churned up political unrest: the government was forced to call for an early election and lost power.
The economic crisis that rattled Greece was not isolated—it affected several countries in the European Union. Norway, for example, also experienced a dip in its gross domestic product. Yet there was little political turmoil in the country.
Why the difference?
“There was a lot of news about people being angry with the Greek government and feeling lied to by their politicians,” says Nancy Qian, professor of managerial economics and decision sciences at the Kellogg School. “It was striking that in other European countries going through a recession, we weren't hearing about political unrest or riots.”
Throughout history, economic crises have sparked differing political responses. But it can be difficult to pinpoint the underlying reason for any differences because countries like, for instance, Greece and Norway differ in many ways, including their overall levels of prosperity.
So Qian and coauthors decided to focus on one specific metric: how much the people of a country typically trust other people.
“It’s about how likely I am to attribute the economic problems to circumstance or luck versus to the political leadership.”
They found that in countries where general trust levels are high, an economic crisis is less likely to trigger political unrest. But in countries where people feel less trusting of others, signs of a flailing economy are more likely to lead to the ruling party being voted out of power.
“If I’m a less trusting person, I might say something like, ‘I don’t understand the details of what our leader is doing, but most politicians are bad and they're lazy, so it is probably his fault,’” Qian explains. Alternately, a trusting person might blame factors beyond politicians’ control. “It’s about how likely I am to attribute the economic problems to circumstance or luck versus to the political leadership.”
Measuring Political Unrest and the Economy
To test the role of trust in political fallout, Qian and her colleagues Nathan Nunn of Harvard University and Jaya Wen of Yale University merged publicly available data sets that included events across most of the world’s countries.
They gathered global data from 1945 to 2014 from the Archigos database on whether the head of government was replaced in any given year. (Rather than looking at the designated leader—in some instances, a ceremonial monarch—the researchers considered effective rulers of state, such as the prime minister in parliamentary systems or the chairman of the party for communist countries.)
Separately, they calculated the average level of trust in a country using sources such as the World Values Survey, which samples a representative population in each country. Questions include ones such as: “Generally speaking, would you say that (A) most people can be trusted or (B) that you need to be very careful in dealing with people?”
The researchers then defined a metric of average trust as the fraction of respondents from a country who said most people can be trusted. Norway ranked highest, with a value of 0.79, and Cape Verde scored the lowest at 0.03, meaning only 3 percent of people said they generally trusted others.
Precisely why trust varies in countries is unclear, Qian says, but it is a measure that remains relatively unvarying with time.
The researchers also used the data to measure how political turnover correlated with economic downturns—defined as periods of negative GDP growth—in countries with differing levels of trust. “We compared the probability of political turnover for any given country between when they had a recession and when they did not,” Qian says.
Then, they studied how those odds varied with differing levels of average trust in a country.
Trust’s Role in Political Turnover
Indeed, trust played a part in determining which countries voted their leaders out of office after a recession.
Economic downturns were less likely to cause political turnover in high-trust countries than in low-trust ones. For example, a recession was 12 percent more likely to trigger a change in political leadership in Italy, where only 29 percent of people say they are generally trusting, than in Sweden, where 63 percent of the population has high levels of trust.
This relationship was only seen in democracies, where people had the power to vote officials out of office.
“We didn’t see this pattern in autocracies, which makes sense,” Qian says. “You can change your leadership in an autocracy by having a revolution or a coup, but that is more difficult to pull off, so there’s not much people can do,” even if they are generally slow to trust.
So does electing a new government help an economy recover after a recession?
The researchers found that high-trust countries that did not vote governments out of office after a recession tended to bounce back faster than nations that voted for a change in the ruling party. The correlation is not a causal link, Qian explains. High-trust countries also have other factors, such as more media freedom, higher incomes, and stronger democracies, which may play a part in this recovery.
Lastly, the researchers examined the international impacts of economic slumps. After all, recessions can ripple across borders. But the team found that a downturn in one country did not trigger political events in a neighboring nation or a trade partner.
“We didn’t expect to find such a clear result that it’s only recessions in one’s own country that matter,” Qian says. “It implies voters are able to separate what's happening in their home country, where they might want to blame their politicians, from what's beyond their leaders’ control.”
Parsing Economic Policies
The results have direct implications for how nations approach economic interactions.
“For example, we are now entering a trade war with several countries,” Qian says. “If we think our trade decisions are going to have economic effects in those nations, we need to also consider the potential political consequences.”