Plausible deniability
| DOI | 10.1177/0951629819875518 |
| Date | 01 October 2019 |
| Published date | 01 October 2019 |
| Author | Joshua A Strayhorn |
| Subject Matter | Articles |
Article
Journal of Theoretical Politics
2019, Vol.31(4) 600–625
ÓThe Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0951629819875518
journals.sagepub.com/home/jtp
Plausible deniability
Joshua A Strayhorn
University of Colorado,USA
Abstract
One explanation for why voters sometimes fail to hold elected officials accountable for corruption
is failure to correctly attribute blame. Yet existing theories of how voters attribute responsibility
do not consider how voters assessments may be shaped by the possibilitythat politicians can stra-
tegically delegate corrupt activity. This paper develops a formal model of an electoral accountabil-
ity environment where politicians can pursue malfeasance directly or indirectly, but where ‘rogue
agents’ occasionally pursuemalfeasance independently. Corruptioncan arise via multiple pathways,
and politicians sometimes possessplausible deniability. In one equilibrium,voters rationally reelect
after plausibly deniable corruptiondue to a non-obvious and novel mechanism. Politicians are also
more likely to delegate malfeasance to agents when they anticipate lenience. Voter lenience is
non-monotonically related to many parameters, including politician competence, the agent’s mal-
feasance preferences, and transparency.
Keywords
Accountability; blame attribution; corruption; delegation
Why do some corruption scandals result in punishment at the ballot box, while
others do not? Electorates do not seem to punish politicians for corruption nearly
as harshly or consistently as we might expect (see De Vries and Solaz, 2017, for a
recent overview). While many explanations have been offered for voters’ inconsis-
tent responses to corruption, recent attention has focused on responsibility attribu-
tion problems. This growing literature has accurately noted that, because
government is complex, voters may face difficulty assigning blame for corruption
to specific politicians. Factors such as voters’ own biases (Anduiza et al., 2013), the
structural complexity of government institutions (Schwindt-Bayer and Tavits,
Corresponding author:
Joshua A. Strayhorn,University of Colorado, 333 UCB, Boulder, CO 80309, USA.
Email: joshua.strayhorn@colorado.edu
2016), or the inherent ambiguity of certain scandals (Winters and Weitz-Shapiro,
2016) may intervene between the voter’s receipt of information about some corrup-
tion scandal and their decision to vote.
What this literature has yet to grapple with, however, is that ambiguity over
responsibility for corruption is subject to strategic manipulation by politicians.
Incumbents control the production side of corruption: a corrupt incumbent might
intentionally delegate corrupt acts precisely to muddy the informational waters if dis-
covered by the public (Hood, 2010). A savvy politician might look the other way as
subordinates engage in corruption, keeping their own hands clean despite a tacit
understanding that corrupt behavior is encouraged. If such strategies can sway voter
perceptions, politicians may be able to insulate themselves from accountability. We
know relatively little about how voters’ perceptions might be shaped by such
strategies.
We might think (or hope) that voters who understand that politicians may
attempt to manipulate their perceptions would be especially skeptical of attempts
to shift blame, and respond particularly strongly to evidence of corruption.
However, I argue in this paper that this is not necessarily true. Instead, even when
politicians are strategic and voters are sophisticated, voters may prefer to reelect
incumbents due purely to the informational structure created when responsibility
for corruption is unclear. The ability to delegate malfeasance to subordinates gives
politicians a strategic advantage, which in turn can help us understand voters’
apparent indifference to corruption.
To support this argument, I develop a formal model that considers voters’
responsibility attribution problem in an environment where corruption is produced
endogenously by politicians. The model’s setting is characterized by plausible denia-
bility: politicians can pursue malfeasance either directly or indirectly, with the indi-
rect route leaving no firm evidence of responsibility. The model additionally allows
for the possibility that innocent politicians are occasionally the victims of ‘rogue
staffers’ acting independently. In one of the game’s two equilibria, voters reelect
politicians despite hard evidence of corruption, as long as there is no smoking gun
linking the politician directly to the scandal. This equilibrium demonstrates that an
‘incompetence defense’ by the politician—a claim to have failed to restrain their
agent rather than to have participated in malfeasance—can be successful even
when voters are sophisticated.
This argument has three major implications for the study of electoral account-
ability. First, the model describes a novel causal mechanism that may account for
voter behavior in a broad class of corruption events. The model closely considers
an oft-overlooked distinction between hard evidence of malfeasance and hard evi-
dence of the politician’s involvement in malfeasance, introducing a parameter reflect-
ing a ‘smoking gun probability’—a chance that the politician’s guilt can be firmly
established. This parameter influences voters’ interpretation of corruption in coun-
terintuitive ways. Simply treating these two forms of information as theoretically
distinct, is, on its own, sufficient to explain voters’ decision to retain an incumbent.
This mechanism can account for reelection of incumbents despite the deliberate
Strayhorn 601
To continue reading
Request your trial