Korea had the highest number of cumulative COVID-19 cases outside China between mid-February and mid-March, outpacing even China in terms of new cases confirmed at a point during this period. Then since the cumulative cases reached over 10,000 in early April, new cases have been kept under 100 per day, with a clear declining trend. And as early as on April 30, the country recorded zero local infection, the first such day since February 18 when the outbreak took off. Experts and the public in and outside Korea consider this successful containment largely as a result of the government’s responses, and growing attention is being paid to the Korean case. One aspect of the Korean case that draws particular attention is that the country never imposed a major lockdown.

While the Korean government’s containment policies and certain Korean products such as test kits, masks and hand sanitizers are drawing attention from around the world, the Korean legislative election held amid the crisis also offers worthwhile observations.

The election for the 21st National Assembly of Korea, the country’s unicameral legislature, was scheduled on April 15. In late March, the National Election Commission (NEC hereinafter), the government body governing electoral administration, announced its plan for preventing COVID-19 infection during elections. Voters were obligated to wear a mask when coming to vote and, while waiting to enter a polling station, waited at pointed spots separated at least 1 meter from each other. When entering the polling station, voters had their temperature checked by non-contact thermometers, sterilized their hands and put on sanitary gloves. Special procedures, such as distance voting, and separate polling booths were prepared for those already confirmed to be infected or showing symptoms.

The official campaign period began on April 2, during the period of the government-initiated social distancing. The number of new cases per day was still around 100. While some governments postponed their electoral procedures, the Korean government decided to go ahead.

One obvious concern was that gathering people at the polling stations could exacerbate the spread of the virus. According to the government, however, there have been no confirmed cases resulting from the election, thanks to the thorough preemptive measures.

Some experts warned that the fear of infection could suppress turnout, as was observed in the French mayoral election in mid-March and the 1918 U.S. mid-term elections during the Spanish flu outbreak. Surprisingly, the national turnout was the highest since the 14th National Assembly election in 1992 and 8.2%p higher than the previous 20th National Assembly election. At the largest local administrative division level, each of 17 divisions had a higher turnout than in the previous National Assembly election. Still, the aforementioned hypothesis might hold if the magnitude of the increase in turnout from the previous National Assembly election decreased with the prevalence of the virus. However, a first look at the data appears to suggest the opposite: local divisions with more cumulative cases on the day of election tended to have larger increases in turnout. For instance, the percentage point increase from the previous National Assembly election, held in 2016, is positively and significantly correlated with the log of cumulative cases on election day. And this relation remains when using different measures of prevalence (different dates or per capita measures, for instance), with or without log. Of course, more rigorous analyses are needed to properly test the hypothesis. But this simple observation suggests that some other important factors might have been at play.

One possibility underlying this counter-intuitive observation is that the election actually became an evaluation of the government’s responses to the crisis. In that case, while the voters in regions with higher prevalence had higher voting costs due to the risk of infection, they might have also had more information about the situation and about the government’s responses, and hence been more certain about their evaluations, which might have mobilized more of them to the polling stations.

In fact, the presidential job approval rate was showing a declining trend since at least the end of last year. Then when the number of new cases started declining, very rapidly, the approval rate also made a turnaround and started increasing, very rapidly. In a Gallup survey conducted right before the 21st National Assembly election, more than half of the respondents who approved the president’s job named the government’s responses to the crisis as the reason for their approval. And the results of the election, held in this political climate, are seen as a landslide for the ruling party. All these observations are in support of the aforementioned possibility.

World-wide attention is being paid to the Korean case of COVID-19. Korea’s legislative election held amid this crisis also offers important lessons to other countries. Even though the situation and circumstances differ from one country to another, the Korean experience shows at least that holding elections during this crisis time, without sacrificing much turnout or exacerbating the outbreak, if any, is not impossible when proper precautions are taken. It also suggests the possibility that the election becomes largely about evaluating the government’s responses to the crisis.