Just how bad was this year? These professors found answers on Twitter.
But exactly how unhappy was the year?
On Twitter, it was the most miserable in at least the last 12 years, according to Peter Dodds and Chris Danforth, who run the Computational Story Lab at the University of Vermont and use data from the site to measure our collective happiness.
Since 2008, the duo has taken a random 10 percent of everything tweeted each day, seeking truths hidden in plain sight. (While acknowledging, as Danforth put it, that “Twitter is a nonuniform subsample of utterances made by a nonuniform subsample of humans who are on the Internet.”)
They’ve used it, for example, to explore fame, finding that Donald Trump and K-pop band BTS are mentioned as commonly as some regular words (think: “after,” “would.”). As Dodds put it, “The word ‘Trump’ has been in the top 300 words all year this year, which he’s never done before. That’s more common than the word ‘God.’ ”
But the Hedonometer, their flagship tool, seeks to quantify the world’s collective happiness — or lack thereof. They selected several languages, then chose the 10,000 most frequently used words in each and asked native speakers to rate them on a 1 (most negative) to 9 (most positive) scale. (This piece refers to the English-language version throughout.)
Each day, an algorithm counts how many times each word is used, discarding any words that scored between four and six — a slice that includes neutral words and words that people use in different contexts, such as slang and curse words. (They continue to refine their list as meanings change. They had to mute “thirsty,” Dodds said. “It just took on a different meaning.”)
The Hedonometer then creates a weighted average of the day’s collective Twitter happiness or unhappiness, based on those words.
“We don’t go in and say this sentence is a happy sentence or this tweet is a happy tweet or a sad tweet or an angry tweet,” Dodds said. “We take a whole collection of them and put them in a big box to get the collective voice.”
In that way, he points out, someone with no knowledge of current events could look at a particular day and probably figure out generally what happened by piecing together the trending words — almost an experimental form of storytelling. Throughout the pandemic, for example, words like “virus,” “outbreak” and “dying” were repeatedly among the most popular.
Some trends have emerged through the years. All else being equal, Saturday is the week’s happiest day on Twitter, Tuesday the saddest. National holidays cause huge spikes in happiness, with Christmas being the most cheerful. Major sporting events and birthdays of pop stars, particularly K-pop stars, tend to make for gleeful days. On the flip side, natural disasters and mass shootings tend to spark more unhappy days.
Until this year, some of saddest days as recorded by the Hedonometer were those of the Boston Marathon bombing in 2013 (5.89), the Las Vegas massacre (5.75) and the school shooting at Marjory Stoneman Douglas High School (5.86), all of which led to spikes of words such as “terrorist” and “murder.”
Two days in 2020 surpassed them. As the grave seriousness of covid-19 began emerging, the Hedonometer recorded happiness ratings roughly equivalent to those lows for weeks, reaching a historic low of 5.71 on March 12.
“The pandemic was an unprecedented event in our lifetime, and there wasn’t the background of sports and music and leisure that there typically is to talk about, which normally balances out the negative things,” Danforth said.
Just as happiness levels began to stabilize, George Floyd died under the knee of a police officer, helping to spark a summer of protests concerning racial inequality. On May 31, at the height of the protests, the Hedonometer recorded a happiness level of 5.63, even lower than March 12, with “terrorist,” “protest,” “violence” and “racist” trending.
“For seven or eight years, there have been a lot of police killings of Black people that end up showing up on the instrument briefly,” Danforth said. “The thing that stands out the most from this year is the sustained staying power of that story, in the midst of the pandemic. … People were really sad during that time, but there was also a lot of activation. People who hadn’t previously engaged in that conversation, people who had not previously protested, became a part of it. I think that’s an inspiring story coming out of a terrible year.”
While those two particular days were deeply unhappy, what truly set 2020 apart from other years is the slowness of the recovery.
“Negative events are tragedies. They’re shocks to the system, but the system usually recovers pretty well,” Dodds said. But after the dual momentous events of 2020, the world’s happiness “took a long time to get back to roughly where it was before. That’s a signal of collective trauma.”
While this year has clearly been an outlier, the decline is in keeping with the trends — the Hedonometer’s happiness level generally has been declining since 2015. The usage of words associated with sorrow and/or anger has skyrocketed.
One reason, Danforth suggested, is the Trump presidency: “The 2016 election really changed how people interacted with Twitter because of the way the president used it, and the stories he was able to tell his supporters through it really engaged people.”
Suddenly, people were using Twitter in a new way, echoing or ranting against the president. Meanwhile, much as the news cycle was upended, the usual trend on Twitter of people discussing more news and politics during the week and movies and weddings and sports on weekends fell apart.
“In the last five years, we’ve seen the usual weekly cycle just get busted,” Dodds added. “It’s sort of all over the place now. Events are happening any day of the week. It’s much more what I would call emotional turbulence.”
Dodds said there was one positive note from 2020. Twitter’s happiness on Christmas peaked in 2014 and had declined every year since. “But this year, it popped up again a little bit,” ahead of 2019′s, he said. “The end of 2020, at least for one day, it went up again.”
Read more:
