Download - Did the online election campaign fail?
Did the Online Election Campaign Fail?:
a Survey Analysis of the Effects of the Online Campaign in the 2013 Japan Upper House Election
Morihiro OGASAHARA
2014 ANPOR Annual Conference, Niigata, Japan
29, November, 2014
Background
• 2013 Japan Upper House Election – The interest was low level
– The first online election campaign in Japan
• Expectations for online election campaign – increase the political interest of voters
– enhance the importance of political issues
– promote dialogues between candidates and voters
Voter turnout
• 52.1% (third-lowest since WW2)
Failure of the online election campaign?
Research Questions
• How much was the exposure?
– To the online election campaign
• How much was the effect?
– To political interest
– To voting behavior
– To party support change
Research Methods
• 2-wave panel survey
– 1st wave: before the election (6/29 ~ 6/30)
– 2nd wave: after the election (7/21 ~ 7/22)
• Sample
– Online survey panel of the research company
– 1st : 2,691 respondents
– 2nd: 1,523 respondents (56.6% of 1st )
Variables
• Dependent variables
– Political interest at 2nd wave
– Voting behavior
– Party support change
• Independent variables
– Political interest at 1st wave
– Exposure to electoral information
– TV, newspaper, conversation, election campaign
Exposure rate to traditional sources
• Television was the highest
Offline election campaign 59.0%
(%)
Exposure rate to online election campaign
• Rather low as a whole
Online election campaign 18.3%
(%)
To political interest
• Positive effect
– Regression model explaining political interest(2nd) β t
(Intercept) 9.713 ***
Age 0.026 1.426
Sex -0.079 -4.238 ***
Political interest (1st) 0.619 31.219 ***
Television 0.041 1.999 *
Newspaper 0.055 2.559 *
Conversation 0.100 4.530 ***
Offline election campaign 0.029 1.311
Online election campaign 0.041 2.132 *
Adjusted R2 0.537
†:p<0.1, *:p<0.05, **:p<0.01, ***:p<0.001
To voting behavior
• Positive effect
– Logistic regression model
†:p<0.1, *:p<0.05, **:p<0.01, ***:p<0.001
B SE
(Intercept) -1.851 0.371 ***
Age 0.016 0.006 **
Sex -0.237 0.133 †
Political interest (1st) 0.538 0.062 ***
Television 0.033 0.170
Newspaper 0.320 0.152 *
Conversation 0.662 0.153 ***
Offline election campaign 0.200 0.154
Online election campaign 0.420 0.207 *
Nagelkelke R2 0.229
B SE
(Intercept) -2.143 0.342 ***
Age 0.007 0.005
Sex 0.015 0.116
Political interest (1st) 0.180 0.056 **
Television 0.621 0.174 ***
Newspaper 0.204 0.129
Conversation 0.190 0.134
Offline election campaign 0.188 0.136
Online election campaign 0.322 0.149 *
Nagelkelke R2 0.093
To party support change
• Positive effect
– Logistic regression model
†:p<0.1, *:p<0.05, **:p<0.01, ***:p<0.001
Discussion
• Push or pull – Push media: passive
– Pull media: active
• One-way or two-way – One-way: audience
– Two-way: participant
• Short term or long term – Short term: election
– Long term: engagement
Conclusion
• The online election campaign succeeded in fact
– Statistically significant positive effects
– More effective than traditional sources
• Online election campaign should be conducted in suitable ways
– Integrate online and offline
– Build relationships
– Connect with voter’s social network, COI
Thank you for your attention!
E-mai: [email protected]