Working Paper: NBER ID: w26321
Authors: Rafael Di Tella; Sebastian Galiani; Ernesto Schargrodsky
Abstract: We study a propaganda campaign sponsored by the government against the main political challenger in the days preceding the 2015 Argentine runoff presidential election. Subjects in the treatment group watched an “ad” initially aired during soccer transmissions that was part of this campaign and were then asked about their political views. Relative to subjects in the control group, their declared preference for the challenger drops by 6.5 percentage points. We find no effects of the three types of defenses employed by the challenger (a positive message unrelated to the “ad”, an answer to the accusations in the “ad”, and a counter-attack). The propaganda effect is driven by women.
Keywords: propaganda; voter preferences; Argentina; political communication
JEL Codes: D72; P48
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Exposure to the government's propaganda ad (M38) | Reduced intention to vote for the challenger, Mauricio Macri (D79) |
| Exposure to the government's propaganda ad (M38) | Increased intention to vote for the incumbent candidate, Daniel Scioli (D79) |
| Exposure to the government's propaganda ad (M38) | Increased number of undecided voters (D79) |
| Exposure to the government's propaganda ad (M38) | Reduced support for Macri among women (J16) |
| Antidotes employed by Macri's campaign (D79) | No significant alteration of voter intentions (D79) |