In January 2020, SARS-CoV-19 reached the United States. With that came a virus that was spreading even faster – xenophobic rhetoric referring to the epicenter of the pandemic in Wuhan, China. Politicians flooded news outlets and social media with distrust of the Chinese government, labeling COVID-19 “Chinese flu”, “Wuhan flu”, “Kung flu” and more.
News attributing COVID-19 to its Asian country of origin seemed to spill onto the streets. Americans who are Asian, Asian American and Pacific Islander (AAPI) began to experience racial harassment, discrimination and physical assault at higher rates than before the pandemic. According to the Stop the AAPI Hate Initiative, an online tool for self-reported hate incidents against AAPI individuals, there were 9,000 hate incidents in the first year of COVID-19. In our online age, understanding whether racial trends on social media are causing harm in the real world is crucial. If so, how can we best predict where racist attacks are most likely?
A new University of Utah under direction to learn has taken a first step. Researchers mapped anti-Asian hatred using Twitter’s dataset of geolocated tweets containing keywords reflecting COVID-19 and anti-Asian hatred in November 2019 and May 2020.
They found that anti-Asian hate speech spiked between January and March 2020, with clusters of hateful tweets spread across the contiguous US that varied in size, distribution of strength, and location. The authors identified two peaks in hateful tweets: the first in late January 2020, when COVID-19 first made its way to the US; the second in mid-March, after President Donald Trump tweeted about the “Wuhan flu” and the “Chinese virus.” Finally, the volume of anti-Asian tweets decreased but remained higher than before the pandemic.
“We can interpret that as hatred lasting once hatred is high. At least as long as our student days,” says Alexander Hohl, assistant professor at the University of Utah and first author of the study. “During the outbreak of COVID there have been reports of hate crimes being committed against Asians and Asian Americans in the United States. We think this warrants a research perspective.”
This study is the first to put anti-Asian hate tweets on one map. The authors hope this method will one day help officials allocate resources to respond to pandemic-related racism as a public health threat.
“In American culture, we see Asians as the exemplar minority, right? They tend to do well academically and make money, or at least that’s our perception,” said Richard Medina, associate professor at U and co-author of the study. “Often in America, Asians are overlooked as targets of hate speech or hate crimes. But COVID-19 brought that out.”
The authors bought 4,234,694 geolocated tweets from Twitter. From this data set, they searched for tweets that met the following criteria: written in English, sent within the study period, localized in the contiguous US, and contained COVID-19 (coronavirus, SARSCoV2, etc.) language. They then classified the remaining 3,274,614 tweets as hateful or non-hateful based on the presence of additional anti-Asian hate-related keywords (Kungflu, Wuhan virus, etc.) and put them on the map.
They identified 15 clusters as the geographic regions where the number of anti-Asian hate tweets was statistically higher than expected based on underlying population. The authors also calculated the so-called relative risk for each US state, which describes the ratio of hateful tweets sent within a county compared to outside. The darker the color of the counties, the higher this relative risk.
The authors found low levels of hate from November 2019 to January 2020 (0% to 0.1% of daily tweets were hateful). Then the rise began, culminating in the first of two peaks in the study period.
The strongest cluster was in Ross County, Ohio, where the proportion of hateful tweets was about 300 times higher than the rest of the country. There were no identifiable patterns of anti-Asian hatred along urban-rural or geographic gradients; They were scattered across the country and varied greatly in size, population, location, and number of tweets. The authors plan further analysis that may reveal demographic and socioeconomic factors to explain cluster locations.
Racism is a public health crisis
The study is in the early stages of what the authors hope to become a public health tool to protect marginalized communities from racist hate crimes. Her next steps are to improve her method of classifying hateful tweets and explore other clustering algorithms.
They also want to understand if hate tweets lead to violence breaking out against the group. They plan to track the time and place that AAPI hate crimes were committed, although hate crime data is notoriously unreliable due to inconsistent reporting by states to federal databases.
“It’s difficult to track hate because the FBI’s hate crime database contains such a small sample of incidents compared to what actually happens, especially because some forms of verbal harassment are not illegal. But that’s increasing, and it’s not just because we’re paying more attention,” Medina said. “We’re seeing more Asian immigrants across the country, and we’re seeing more hatred of them than the immigrant population in general. It will just keep going and we have to pay attention to that.”
The authors also want to develop a description of a hateful place. They plan to study the socioeconomic and demographic variables in counties or cities with elevated levels of anti-Asian hatred.
“The hope is that we can use hate on social media as an indicator of hate on the streets, to alert communities and public health officials to get the help and protection they need.”
Other authors of the study are Moongi Choi, Neng Wan and Ming Wen from the University of Utah and Aggie Yellowhorse from Arizona State University.