Skip to main content

Smiles, turnout, candidates, and the winning of district seats: Evidence from the 2015 local elections in Japan

  • Masahiko Asano (a1) and Dennis P. Patterson (a2)

Research has shown that a candidate’s appearance affects the support he or she receives in elections. We extend this research in this article in three ways. First, we examine this relationship further in a non-Western context using 2015 local elections in Japan. Next, we show that this positive relationship is more complicated depending on the characteristics of the election under consideration. Specifically, we distinguished election contests by levels of turnout and found that despite a positive relationship between turnout and the extent to which smiling increases a candidate’s support levels, the marginal increase in support declined as turnout increased and, in fact, became negative when some high-turnout threshold was crossed. Finally, we show that the number of candidates competing in an election is negatively related to the impact of a candidate smiling, confirming research conducted by the Dartmouth Group.

Corresponding author
Correspondence: Dennis P. Patterson, Department of Political Science, Texas Tech University, 112 Holden Hall, Lubbock, Texas, 79409. Email:
Hide All
1The Columbia School has emphasized the sociological characteristics of individual voters, while the Michigan School or social-psychological school has emphasized partisan attitudes and attachments. On the former, see, e.g., Berelson, B., Lazerfeld, P., and McPhee, W., Voting (Chicago: University of Chicago Press, 1954), On the later, see A. Campbell, P. Converse, W. Miller, and D. Stokes, The American Voter (New York: Wiley, 1960).
2The classic text on the rational voter model is, Riker, W. and Ordeshook, P., An Introduction to Positive Political Theory (Englewood Cliffs, NJ: Prentice Hall, 1973).
3 Stewart, P. A., Salter, F. K., and Mehu, M., “Taking leaders at face value: Ethology and the analysis of televised leader displays,” Politics and the Life Sciences , 2009, 28(1): 4874.
4This literature has its roots in the work completed by the Dartmouth Group. See, e.g., Sullivan, D. G. and Masters, R. D., “Happy warriors: Leaders’ facial displays, viewers’ emotions, and political support,” American Journal of Political Science , 1988, 32(2): 345368, and R. D. Masters, D. G. Sullivan, J. Lanzetta, G. J. McHugo, and B. G. Englis, “The facial display of leaders: Toward an ethology of human politics,” Journal of and Biological Structures, 1986, 9(4): 319–343. For more recent work that builds on these contributions, see Alexander Todorov, A. V. Mandisodzo, A. Goran, and C. C. Hall, “Inferences of competence from faces predict election outcomes,” Science, 2005, 308: 1623–1626; and P.A. Stewart, E. P. Bucy, and M. Mehu, “Strengthening bonds and connecting with followers: A biobehavioral inventory of political smiles,” 2015, Politics and Life Sciences, 34(1): 73–92. For an evolutionary perspective, see G. R. Murray, “Evolutionary preferences for physical formidability in leaders,” Politics and the Life Sciences, 2014, 33(1): 33–53.
5Stewart, Salter, and Mehu, pp. 48–74. The phrase “facial displays” is employed throughout this article, reflecting the behavioral ecology view of Alan Fridlund as more applicable to candidates seeking support in election contests as opposed to the basic emotions view of facial expressions. On the former, see A. J. Fridlund, Human Facial Expressions: An Evolutionary View (San Diego, CA: Academic Press, 1994). On the latter, see P. Ekman, “Universals and cultural differences in facial expressions of emotion,” in Nebraska Symposium on Motivation, J. Cole, ed. (Lincoln: University of Nebraska Press, 1971), pp. 208–288.
6See, e.g., Masters, R. D. and Sullivan, D. G., “Nonverbal displays and political leadership in France and the United States,” Political Behavior , 1989, 11(2): 123156.
7Sullivan and Masters, p. 361, explain that as the number of competing candidates was reduced, voters were “quicker to convert their own emotional responses to gestures into more generalized attitudes of submission or support”.
8Todorov et al., pp. 1623–1626.
9Stewart, Salter, and Mehu, pp. 48–74.
10 Berggren, N., Jodhal, H., and Poutvaara, P., “The looks of a winner: Beauty and electoral success,” Journal of Public Economics , 2010, 94(1–2): 815.
11 King, A. and Leigh, A., “Beautiful politicians,” Kyklos , 2009, 62(4): 579593.
12 Atkinson, M. D., Enos, R. D., and Hill, S. J., “Candidate faces and election outcomes: Is the face-vote correlation caused by candidate selection? Quarterly Journal of Political Science , 2009, 4(3): 229249.
13 Ballew, C. C. II and Todorov, A., “Predicting political elections from rapid and unreflective face judgements,” Proceedings of the National Academy of Sciences , 2007, 104: 1794817953.
14 Olivia, C. Y. and Todorov, A., “Elected in 100 milliseconds: Appearance-based trait inferences and voting,” Journal of Nonverbal Behavior , 2010, 34(2): 83110.
15 Murray, G. R., “Evolutionary preferences for physical formidability in leaders,” Politics and the Life Sciences , 2014, 33(1): 3353, This study showed that it was physical prowess as opposed to beauty that defined preferences for leaders.
16 Lawson, C., Lenz, G. S., Baker, A., and Myers, M., “Looking like a winner: Candidate appearance and electoral success in new democracies,” World Politics , 2010, 62(4): 561593.
17 Rosar, U., Klein, M., and Beckers, T., “The frog pond beauty contest: Physical attractiveness and electoral success of the constituency candidates at the North Rhine-Westphalia state election of 2005,” European Journal of Political Research , 2008, 47(1): 6479.
18 Horiuchi, Y., Komatsu, T., and Nakaya, F., “Should candidates smile to win elections? An application of automated face recognition technology,” Political Psychology , 2012, 33(6): 925933.
19 Matsumoto, D., “American-Japanese cultural differences in the recognition of universal facial expressions,” Journal of Cross-Cultural Psychology , 1992, 23(1): 7284.
20 Biehl, M., Matsumoto, D., Ekman, P., Hearn, V., Heider, K., Kudoh, T., and Ton, V., “Matsumoto and Ekman’s Japanese and Caucasian Facial Expressions of Emotion (JACFEE): Reliability data and cross-national difference,” Journal of Nonverbal Behavior , 1997, 21(1): 321.
21Sullivan and Masters, p. 361.
22 Franklin, M. N., “The dynamics of electoral participation,” in Comparing Democracies 2: New Challenges in the Study of Elections and Voting, LeDuc, L., Niemi, R. G., and Norris, P., eds. (Thousand Oaks, CA: Sage, 2002), pp. 148168.
23 Campbell, J. E., “The revised theory of surge and decline,” American Journal of Political Science , 1987, 31(4): 965979.
24 DeNardo, J., “Turnout and the vote: The joke’s on the Democrats,” American Political Science Review , 1980, 74(2): 406420.
26 Budge, I. and Farlie, D., Explaining and Predicting Elections: Issue Effects and Party Strategies in Twenty-Three Democracies (London: Allen and Unwin, 1983).
27 Petrocik, J. R., “Issue ownership in presidential elections with a 1980 case study,” American Journal of Political Science , 1996, 40(3): 825850.
29Almost all of the designated cities completed uploading their official gazette of faces for the elections by April 8, five days after the official announcement of the voting date for the election. These uploads were ceased on election day, April 12, 2015.
30This should not undermine the results we report here because of the broad availability of campaign posters to all potential voters and the fact that we are calibrating their relationship to support levels at the aggregate level.
31Horiuchi, Komatsu, and Nakaya.
32For a general treatment of the political importance of smiling in a hierarchical situations, see, Mehu, M., Grammer, K., and Dunbar, R. I. M., “Smiles when sharing,” Evolution and Human Behavior , 2007, 28(6): 415422.
33 Mehu, M. and Dunbar, R. I. M., “Naturalistic observations of smiling and laughter in human group interactions,” Behavior , 2008, 145(12): 17471780.
34Since one candidate, Tsuji Yoshitaka, out of 1,380 in Osaka City did not provide his photo for the official gazette for elections (senkyo koho), we did not evaluate his facial expression.
35Almost all of the designated cities completed uploading their official gazette of faces for the elections by April 8, five days after the official announcement of the voting date for the election. These uploads were ceased on election day, April 12, 2015.
36Campbell et al.
37In Japan, these voters include both weak partisans and unaffiliated electors, depending on how they are measured in public opinion surveys. NHK Yoron Chosa, Gendai Nihonjin noh Ishiki Kozo [The structure of modern Japanese consciousness] (Tokyo: NHK Books, 1991).
38Jijitsushinsha, Sengo no Seito toh Naikaku Shijiritsu [Postwar party and cabinet support rates] (Tokyo: Jijitsushinsha, 1986).
39In our analysis of city council elections, some involved districts in which population densities were lower than others, thus providing the appropriate conditions for this effect to obtain.
40See the Appendix for the descriptive statistics on the variables we use in our analysis.
41 Bookstein, F. L., Morphometric Tools for Landmark Data: Geometry and Biology (Cambridge: Cambridge University Press, 1991).
42 Lestral, P. E., Fourier Descriptors and their Applications in Biology (Cambridge: Cambridge University Press, 1997).
43 Viola, P. and Jones, M. J., “Robust real-time face detection,” International Journal of Computer Vision , 2004, 57(2): 137154.
44This process revealed no inconsistencies in the way the software calibrated smiling in candidates’ campaign posters.
45We thank Mr. Uetsuji of Omron Co. for his help on appropriately interpreting the smile index for the latest version of the OKAO Vision.
46We measured the smile index for 1,379 candidates at the following website that Omron set up:
47These designated cites are Chiba, Fukuoka, Hamamatsu, Hiroshima, Kawasaki, Kita-kyushu, Kobe, Kumamoto, Kyoto, Nagoya, Niigata, Okayama, Osaka, Sagamihara, Saitama, Sakai, Sapporo, Sendai, Shizuoka, and Yokohama.
48These local elections take place where the number of available seats varies from district to district. District magnitudes then refer to the number of seats available in each district.
49The reason we only use 1,379 candidates out of 1,477 candidates (total candidate number) is that because we only use the data on 16 cities (except Sendai, Shizuoka, Kita-kyushu, and Hiroshima). Three cities (Sendai, Shizuoka, and Kita-kyushu) are excluded because they did not have elections in 2015. Hiroshima is excluded because Hiroshima city office did not officially release its candidates’ photo information on the website.
50Descriptive statistics on our variables are presented in the Appendix.
51We included this interaction term to capture the expected declining marginal impact of smiles on candidates’ obtained vote shares.
52To be clear, this means that a candidate is a challenger in the 2015 election but a winner in at least one election before the 2011 election.
53We did not detect any extensive heteroskedasticity in our variables, which allows us to conclude that OLS is the correct estimation method. Moreover, as discussed in more detail later, we captured the potential nonlinear relationships we discussed earlier through the inclusion of several interaction terms in the models for estimation.
54The substantive and statistical insignificance of the gender variable can most likely be explained by the fact that we are examining local (city council) elections, which generally involve more female candidates, rendering them much more gender friendly.
55 SD means standard deviation. Since the standard deviation on “turnout” is 6.369 and its mean is 44.437, we get the high value (mean_turnout 1 SD  51 and the low value (mean_turnout 1 SD  38.
56We calculated the slope of vote share on smile at the mean (mean 11), a high value of the number of candidates (mean 1 SD   17), and at a low value (mean 1 SD   6). We then estimated the three equations and found the regression models are statistically significant at the three levels of the number of candidates.
57Sullivan and Masters.
58We appreciate the comment that one reviewer made, suggesting that we explore the impact of smiling not as an index per se but rather in terms of its impact at its highest and lowest values. The discussion in this section is based on our attempt to adhere to this suggestion.
59This is the predicted vote gain a candidate will receive when his or her smile goes from a score of 0 to a score of 100 is 1.2 percentage points in a district with lower turnout rates (38% or less).
60There were three districts where three candidates ran: Higashiyama-ku (Kyoto city), Fukushima-ku (Osaka city), and Konohama-ku (Osaka city). There was one district in which as many as 29 candidates ran: Minami-ku (Sagamihara city).
61These are happiness, neutral, surprise, anger, and sadness.
62We tried to access to the same older version of OKAO Vision that Horiuchi, Komatsu, and Nakaya used, but it turned out that no older versions were on sale or available from Omron.
63We realize that since our analysis is at the aggregate level, it is characterized by inherent limits and that an analysis of these relationships at the individual level would add more insight into these issues.
64We are grateful to a reviewer for comments that suggested these issues for possible future research.
Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

Politics and the Life Sciences
  • ISSN: 0730-9384
  • EISSN: 1471-5457
  • URL: /core/journals/politics-and-the-life-sciences
Please enter your name
Please enter a valid email address
Who would you like to send this to? *



Altmetric attention score

Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed