On January 4, 2012 an explosion killed a man in an apartment in the Ukrainian port city of Odessa. Police arrested another occupant. One month later, on February 4, a second man was arrested in connection with the explosion. On February 27—six days before the March 4 Russian presidential election—Russian state controlled television station Channel One broke the story that the two detainees had been part of a plot to assassinate Russian Prime Minister, and presidential candidate, Vladimir Putin. “Channel One said it received information about the assassination attempt 10 days [earlier] but did not explain why it did not release the news sooner.”
Two points of this anecdote are noteworthy. First, information about the alleged plot was not released as soon as it was available. Instead, state television dropped the bomb- shell at a later, strategically-chosen time. Second, voters drew inferences from the timing of the release.
In this paper we analyze a Sender-Receiver game which connects the timing of information release with voters’ beliefs prior to elections. Early release of information is more credible, in that it signals that Sender has nothing to hide. On the other hand, such early release exposes the information to scrutiny for a longer period of time—possibly leading to the information being discovered to be false. […]
We show that fabricated scandals are only released sufficiently close to the election. […] Perhaps more importantly, we make predictions about the time pattern of campaign events. We show that for a broad range of parameters the probability of release of scandals (authentic or fabricated) is U-shaped, with scandals concentrated towards the beginning and the end of an electoral campaign.
{ When to Drop a Bombshell, 2016 | PDF }
The concentration of scandals in the last months of the 2016 campaign is far from an exception. Such October surprises are commonplace in US presidential elections. […] Political commentators argue that such bombshells may be strategically dropped close to elections so that voters have not enough time to tell real from fake news. Yet, if all fake news were released just before an election, then voters may rationally discount October surprises as fake. Voters may not do so fully, however, since while some bombshells may be strategically timed, others are simply discovered close to the election.
Therefore, the strategic decision of when to drop a bombshell is driven by a tradeoff between credibility and scrutiny. […]
This credibility-scrutiny tradeoff also drives the timing of announcements about candidacy, running mates, cabinet members, and details of policy platforms. An early announcement exposes the background of the candidate or her team to more scrutiny, but boosts credibility. The same tradeoff is likely to drive the timing of information release in other contexts outside the political sphere. For instance, a firm going public can provide a longer or shorter time for the market to evaluate its prospectus before the firm’s shares are traded.
{ When to Drop a Bombshell, 2017 | PDF }
