Can Partisan News Shift Political Preference and Voting Behavior? An Experimental Evidence from Taiwan’s General Election 2016

How to cite

Lin, Ming-Jen. 2019. "Can Partisan News Shift Political Preference and Voting Behavior? An Experimental Evidence from Taiwan’s General Election 2016." AEA RCT Registry. AEARCTR-0004934. https://doi.org/10.1257/rct.4934-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Ming-Jen Lin*

* Primary investigator

Registry ID
AEARCTR-0004934
DOI
https://doi.org/10.1257/rct.4934-1.0
First registered
Last updated
Status
Completed
Trial dates
2015-11-18 – 2016-02-02
Countries
Taiwan, Province of China (Taipei)
JEL codes
D72, L82, C92, C93
Registry record
http://www.socialscienceregistry.org/trials/4934

Abstract

We conducted a field experiment to investigate the effects of partisan news on the 2016 Taiwan Presidential and Legislative Elections. Subjects are randomly divided into four groups: KMT the ruling party, DPP, the opposition party, the new third-party and control, and provided them with distinct partisan news articles. We find that the pro DPP news articles have significantly increased the support for the DPP candidate Tsai by reinforcing the existing preferences. Pro third-parties articles have changed subjects' voting intentions to support the new party GSD. We also find that the estimated baseline and reinforcement treatment effects are driven by undecided subjects.

Intervention

We randomly assigned partisan news articles to our subjects during one month before the 201 General Election in Taiwan. During the period of intervention, we asked subjects to complete reading comprehension to the news articles to ensure the readership. Also, we asked subjects their political preferences and voting intentions toward the 2016 election.

Experimental design

The subjects are divided into four treatment groups by the article they received: articles favoring KMT (the ruling party), favoring DPP (the opposing party), favoring NPP and GSD (new emerging party) and control group (receiving news articles unrelated to politics, like sport or entertainment).

Design details

Randomization method
Randomization was done in office by computer with STATA, seed id 20160116.
Randomization unit
We assigned treatments individually. By the initial survey, we divided subjects into two blocks: people initially favoring KMT or people initially favoring DPP. Within each block, we randomize the treatment assignments evenly.
Clusters
200 subjects.
Observations
200 subjects.
Arms
50 in KMT treatment group, 50 in DPP, 50 in THIRD PARTIES (NPP/GSD), 50 in Control
Primary outcomes
The voting decisions subjects reported in 2016 General Election
Secondary outcomes
1. The voters' preference to parties or candidates. 2. The voters' media consumption behavior.

Cited by

No citation data available for this registration yet.

Keywords