The social media network, Twitter, has been at the heart of many a public debate not least the national and international response to the COVID-19 pandemic. New research from the USA published in the International Journal of Business and Systems Research, has examined public opinion on “lockdowns” and “reopening for the economy” during the first summer of the pandemic as revealed by more than a million unique Twitter updates about COVID-19.
Sina Shokoohyar and Julianne Dang of Saint Joseph’s University in Philadelphia, Pennsylvania, and Hossein Rikhtehgar Berenji of Pacific University in Forest Grove, Oregon, classified 1.3 million Twitter updates, whimsically known to users of the microblogging website and application as “tweets”. Their classification divided updates into three camps. The first, were in favour of removing lockdown restrictions to allow the US economy to “reopen”. The second category had those updates aligned with continuing lockdown restrictions for the sake of public health. The third category were neutral tweets offering facts rather than opinion.
Rather than using logistic regression, decision tree, random forest, neutral network or multinomial naïve Bayes, the team turned to a gradient boosting classifier algorithm, which they demonstrate had an accuracy of 88% and so outperformed those other classifiers in their research.
The fundamental conclusion from the analysis is that there were significantly more tweets in favour of reopening the economy rather than persisting with lockdown measures, such as ongoing educational and business closures and stay-at-home orders and that this opinion became increasingly prominent as time passed during the early stages of the pandemic lockdowns. The team suggests that the perceived and real socioeconomic impact of lockdowns on stock markets, gross domestic product (GDP), unemployment rates, and rates of household consumption were drivers for the offered opinions of many Twitter users.
Of course, lockdowns led to an increase in social media activity and so this in itself partly underpins the increase in tweets offering an opinion on lockdown, public health, and the socioeconomic impact of COVID-19.
“Perhaps one of the most surprising side effects of the outbreak is the increase in US residents’ engagement in expressing their opinion on social media,” the team writes. “People from all walks of life are suddenly reading statistical analyses and epidemiology charts and sharing them as if they were popular music videos or comedy memes,” the researchers add.
There are implications for policymakers of this study, the team suggests. Twitter and other social media can be used to extract public opinion quite widely and so reveal how public attitudes to any given policy or regulation might change in an emergency situation such as a global pandemic.
The team adds that “Analysing these tweets can shorten the time to observe the consequences of the pandemic, and can facilitate faster response by policymakers.” Whether or not policymakers should be chasing public opinion in a crisis of this sort is perhaps a different matter when there are direct implications for public health to be weighed against long-term implications for the economy and ultimately its effects on public wellbeing and public health.
Shokoohyar, S., Rikhtehgar Berenji, H. and Dang, J. (2021) ‘Exploring the heated debate over reopening for economy or continuing lockdown for public health safety concerns about COVID-19 in Twitter’, Int. J. Business and Systems Research, Vol. 15, No. 5, pp.650–672.