Voting in a Pandemic: Lessons From the 2020 South Korean Legislative Election

Abstract

In this paper, we examine the extent to which a government's response to a pandemic affects election outcomes. Using detailed data on South Korea's 21st legislative election, held in April 2020 during the COVID-19 pandemic, we find that a candidate of the ruling Democratic Party was less preferred in electoral districts where the COVID-19 infection rate was higher. We also find that the South Korean government's successful control of the disease contributed significantly to the overwhelming victory of the ruling party against the leading opposition party. Specifically, our counterfactual analysis predicts that each party would have taken 129 and 102 of the 231 constituency seats considered in the analysis, respectively, had the COVID infection rate been the same as the OECD average during the election period. Given that the observed result was 147 to 84, this implies that 18 electees of the ruling party would have lost to the candidates of the leading opposition party, which in turn would have granted the opposition party the ability to block any attempt by the ruling party to fast-track debatable bills.

Keywords:

• Elections
• politics
• voter utility
• COVID-19

JEL Classification Numbers:

• D72
• H75
• P16

Disclosure Statement

No potential conflict of interest was reported by the authors.

Data and Code Availability

The data and code used in the analysis are available at GitHub (https://github.com/kimik0121/election2020).

Notes

1 https://covid19.who.int/

2 Here are a few examples of news articles and reports covering the success of South Korea:- https://www.wsj.com/articles/lessons-from-south-korea-on-how-to-manage-covid-11601044329- https://ourworldindata.org/covid-exemplar-south-korea- https://www.bloomberg.com/features/2020-south-korea-covid-strategy/

3 COVID-19 might have also affected consumer choices over products. For instance, Kim (2021) showed that the worsened image of China in South Korea after the COVID-19 outbreak caused consumers to drink Chinese beers less than before.

4 The cap on expenditures for a district is equal to 100 million KRW + (population of the district × 200 KRW) + (number of counties in the district × 2 million KRW).

5 This report on campaign expenditures was available on NEC's website only for the first three months after the election.

6 http://www.realmeter.net/category/politics/

7 https://thediplomat.com/2020/04/elections-in-the-time-of-covid-19/

8 Gyeongsang consists of two provinces and three metropolitan cities including Busan, the second largest city in South Korea after Seoul, while Jeolla is composed of two provinces and one metropolitan city.

9 For instance, the Gangnam district in Seoul is composed of three electoral districts.

10 The minimum voting age is 18 in South Korea.

11 The radical Justice Party won one seat, and another four seats were taken by independent candidates who joined the United Future Party after the election.

12 Military service is not imposed on women in South Korea.

13 Wealth is censored below zero. There are five candidates with negative wealth.

14 After excluding 275 cases detected under the quarantine after entry at the airports and seaports, we have 10,175 confirmed cases nationwide by that date.

15 The data analyzed in this paper come from the following sources: (i) election data from NEC (http://info.nec.go.kr/), (ii) COVID-19 data from KDCA (http://ncov.mohw.go.kr/index.jsp), and (iii) local tax data from WeTax (https://www.wetax.go.kr/main/?cmd=LPTIIA7R3).

16 The average and median of the voting share of minor party candidates and independent candidates are 2.3 percent and 0.7 percent, respectively.

17 Democratic Party candidates may have more criminal incidents on record due to their involvements in the democratization movement in 1980s–1990s.

18 1 is added before taking log values of wealth, criminal, and covid. Also, $\ln covid$ and $\ln pci$ are centered about their mean values.

19 As a robustness check, we interact indicator Dem with the proportion of 40s among voters in the district. This age group is known to be the most loyal supporters of the president Moon Jae-in and the ruling Democratic Party. Our empirical findings (available from the authors upon request) are robust to this inclusion.

20 Gerber (1998) showed that caps on campaign expenditures lower the advantage of incumbency.

21 1 is added before taking the log.

22 The generalized error ${\hat{r}}_{jt}$ is derived as $incumbent\cdot \frac{\varphi (\mathbf{z}\hat{\gamma })}{\mathrm{\Phi }(\mathbf{z}\hat{\gamma })}-(1-incumbent)\cdot \frac{\varphi (-\mathbf{z}\hat{\gamma })}{\mathrm{\Phi }(-\mathbf{z}\hat{\gamma })}$ where $\mathrm{\Phi }(\cdot )$ and $\varphi (\cdot )$ denote the standard normal cdf and pdf.

23 See Section 21.4 of Wooldridge (2010) and the related discussion on why this nonlinear instrument, based on the conditional mean, can be optimal in this setting.

24 For an electoral district composed of multiple second-tier administrative districts, we compute the voting share of a candidate by dividing the total simulated number of votes for the candidate by the total population size in the electoral district.

25 The nationwide percentage of ballots cast for United Future Party candidates among all valid ballots would have increased by one percentage point from the observed 41.5 percent to 42.5 percent, potentially resulting in the election of one additional proportional representation candidate from the party.

26 https://www.hrw.org/news/2020/04/03/how-authoritarians-are-exploiting-covid-19-crisis-grab-power

Additional information

Notes on contributors

In Kyung Kim

In Kyung Kim is an assistant professor of economics at Kyungpook National University. He completed a B.A. (2003) in economics at Seoul National University and a Ph.D. (2014) in economics at Indiana University. His research focuses on industrial organization and quantitative marketing.

Kyoo il Kim

Kyoo il Kim is a professor of economics at Michigan State University. He completed a B.A. (1997) and an M.A. (2001) in economics at Seoul National University and a Ph.D. (2006) in economics at UCLA. His research has been focused on diverse topics in econometrics and industrial organization such as differentiated products demand, consumer choice and firm behavior, random coefficient model, hedonic pricing, semiparametric and nonparametric identification/estimation, control function method, and production function and firm productivity.