Popular in Political Science

American Politics

Keeping the Republic

Boston College Marc Landy

British Politics

Entrepreneurial thinking: the politics and practice of policy impact

University of Sheffield Matthew Flinders

Canadian Politics

Parties and Elections: An End to Canadian Exceptionalism?

St. Francis Xavier University James Bickerton

Asian Politics

The Politics of Knowledge in Development

University of Delhi Madhulika Banerjee

American Politics

Fifty years of declining confidence and increasing polarization in trust

University of California Berkeley Henry Brady

American Politics

The Opening Dilemma

Cornell University Richard Bensel

Political Economy

Wait and See?

University of Liverpool Mariaelisa Epifanio

Stealing an Election: Violence or Fraud?

View more insights in Social Sciences - Political Science - International Relations - War and Conflict Studies

View more figures from Brown University

View the page of Dawn Brancati

Read the Study

Political actors often resort to electoral violence to gain an edge over their competitors even though violence is harder to hide than fraud and more likely to delegitimize elections as a result. Existing explanations tend to analyze violence in terms of the same factors as fraud, or to treat violence as a means of last resorts given its overtness. Dawn Brancati argues that political actors often use violence for the very reason that it is hard to hide.

Image courtesy of interviewee. June 12, 2024

Generating Workspace

Export as PDF

In a discussion about electoral violence, the speaker provides an overview of their research aimed at understanding how and why candidates resort to violence instead of electoral fraud to influence election outcomes. They define electoral violence as actions that either prevent elections or alter their results, including intimidation, coercion, and violent attacks. The speaker argues that the observability of violence makes it a more reliable method for candidates because it reduces the risk of "shirking" by those contracted to implement such strategies, unlike the more difficult-to-monitor nature of electoral fraud.

The study they reference examines past elections, particularly looking at the 2005 Egyptian elections, where increased electoral monitoring by judges made fraud more difficult, leading to an uptick in violence used by candidates. The findings suggest that when fraud becomes harder to commit due to monitoring, candidates may lean towards using violence instead.

The speaker concludes by indicating future research directions, including a book project focused on observing electoral violence globally and exploring the broader implications for international relations, foreign aid, and investment based on the types of electoral manipulation used.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.000 --> 00:01.725
Electoral violence is a broad term

00:01.825 --> 00:04.085
and it includes many different forms of violence.

00:04.745 --> 00:07.525
It can include violence intended to delegitimize

00:07.525 --> 00:09.325
or stop an election from occurring.

00:09.905 --> 00:11.085
The Taliban when it was

00:11.085 --> 00:12.965
outside government used violence in this way.

00:13.705 --> 00:15.645
It also can include violence that occurs

00:15.645 --> 00:18.245
after an election in order to object to the outcome

00:18.245 --> 00:21.005
of the election, as we saw in the United States

00:21.005 --> 00:22.765
with the January 6th insurrection.

00:23.705 --> 00:26.205
But the type of electoral violence that my co-author

00:26.205 --> 00:29.445
and I analyze in this research is violence that is used

00:29.445 --> 00:33.045
to change the outcome of an election, and it can occur

00:33.105 --> 00:35.445
before, during, or after an election.

00:36.805 --> 00:38.045
Violence can change the outcome

00:38.045 --> 00:39.565
of election in a number of ways.

00:40.505 --> 00:43.005
It can force candidates to drop out of an election.

00:43.975 --> 00:46.355
It can prevent candidates from running an election

00:46.535 --> 00:49.115
by either killing them or severely injuring them.

00:50.215 --> 00:53.075
It can deter voters from not going to the polls to vote

00:53.075 --> 00:54.155
for certain candidates.

00:55.135 --> 00:58.155
It can even compel voters to vote for certain candidates

00:58.155 --> 01:00.675
that they wouldn't otherwise do in.

01:00.695 --> 01:03.785
In another context, it can even be used

01:03.835 --> 01:05.465
after votes have been counted to,

01:05.965 --> 01:07.625
to destroy the actual ballots.

01:07.685 --> 01:09.025
In this study, my co-author

01:09.025 --> 01:11.705
and I try to understand why candidates use electoral

01:11.825 --> 01:14.905
violence to steal elections instead of electoral fraud.

01:15.835 --> 01:18.785
There are many different reasons why candidates should not

01:18.965 --> 01:20.545
use violence to steal an election.

01:21.405 --> 01:24.025
So violence is potentially more repugnant

01:24.025 --> 01:25.105
to voters than fraud

01:25.105 --> 01:27.945
because of the physical harm that it causes and

01:27.945 --> 01:29.825
therefore could potentially alienate a

01:29.825 --> 01:31.025
candidate's own supporters.

01:32.465 --> 01:35.625
Violence is also easier to observe so that anyone

01:35.645 --> 01:37.745
who uses it is more likely to get caught.

01:39.275 --> 01:41.495
So our argument in this paper relates

01:41.515 --> 01:44.375
to the last point about the observability of violence.

01:45.675 --> 01:48.015
Our argument is distinct from the existing literature

01:48.015 --> 01:50.455
and that we argue that the observability of violence

01:51.375 --> 01:52.375
actually provides an incentive

01:52.395 --> 01:53.935
to candidates to use violence.

01:54.275 --> 01:57.135
So the observability we argue may co sit less likely

01:57.165 --> 02:00.255
that people that candidates hire to steal elections

02:00.255 --> 02:02.175
for them shirk their responsibilities

02:02.395 --> 02:04.495
by not committing either fraud or violence.

02:05.855 --> 02:08.105
With fraud, the risk of shirking is very high.

02:08.665 --> 02:11.225
A candidate, for example, can offer voters $50

02:11.245 --> 02:12.705
in exchange for their vote.

02:13.245 --> 02:14.745
But once inside the polling booth

02:15.285 --> 02:16.665
and outside anyone's view,

02:17.005 --> 02:18.465
the voter can do whatever they want.

02:18.775 --> 02:20.425
They can vote for another candidate.

02:20.425 --> 02:22.785
They could submit a blank ballot or an invalid ballot,

02:23.045 --> 02:24.385
or they may not even vote at all.

02:25.015 --> 02:27.425
With violence, the risk of CI is much lower

02:27.425 --> 02:29.065
because violence is more observable

02:29.065 --> 02:31.945
and a candidate can better monitor whether this person

02:31.975 --> 02:33.425
commits violence for them or not.

02:34.735 --> 02:37.135
A candidate, for example, can offer a person $50

02:37.195 --> 02:38.495
to attack another candidate.

02:38.955 --> 02:41.135
And unlike in the case of electoral fraud,

02:41.435 --> 02:44.015
the candidate can monitor well whether this person attacks

02:44.015 --> 02:45.015
the other candidate or not,

02:45.015 --> 02:46.375
because violence is more observable.

02:47.095 --> 02:49.195
The candidate may not personally observe the person

02:49.275 --> 02:50.395
attacking the other candidate,

02:50.415 --> 02:52.155
but they will hear about it in the news

02:52.215 --> 02:53.555
or by word of mouth from others

02:53.695 --> 02:55.955
who observe the attack or its aftermath.

02:57.125 --> 02:58.905
As a result, the person who's

02:58.975 --> 03:01.375
Paid to commit violence is less likely to shirk.

03:01.795 --> 03:03.055
If they do shirk, they're likely

03:03.055 --> 03:04.375
to get punished in some way.

03:04.405 --> 03:07.175
They may not get paid or hired for another job,

03:07.315 --> 03:08.815
or they themselves might get attacked.

03:08.815 --> 03:10.775
Testing this argument is really difficult

03:10.775 --> 03:13.335
because you need to measure something you can't actually

03:13.355 --> 03:14.775
see, and that's electoral fraud.

03:15.395 --> 03:18.055
So what we did was the best, the next best thing,

03:18.055 --> 03:20.495
and we looked at something we think affects the likelihood

03:20.515 --> 03:23.095
of persons to commit fraud, to shirk.

03:23.395 --> 03:25.135
And when the incentive to shirk is high,

03:25.315 --> 03:27.535
we expect electoral violence to be higher

03:27.535 --> 03:30.295
because the candidate can't rely on those.

03:30.325 --> 03:33.015
They they contract to commit vi fraud for them

03:33.035 --> 03:34.415
to actually do so and

03:34.415 --> 03:36.095
therefore switch to electoral violence.

03:37.315 --> 03:39.015
The thing that we look at in this study

03:39.015 --> 03:41.415
that affects the risk of shirking when it comes

03:41.415 --> 03:43.375
to fraud is electoral monitors.

03:44.075 --> 03:46.215
So electoral monitors can increase the incentive

03:46.215 --> 03:47.255
to shirk for two reasons.

03:47.755 --> 03:49.935
One, they make it harder to commit fraud,

03:50.195 --> 03:52.615
and two, they make it harder candidates

03:52.615 --> 03:54.255
to monitor whether the persons they hire

03:54.275 --> 03:56.735
to commit fraud do the do it for them or not.

03:57.525 --> 03:59.615
Electoral monitors do not affect the risk

03:59.615 --> 04:00.935
of shirking when it comes to violence

04:00.935 --> 04:02.855
because violence is observable whether

04:02.855 --> 04:04.175
monitors are present or not.

04:04.875 --> 04:05.975
So in this project,

04:06.115 --> 04:08.615
we look at the electoral monitors in Egyptian elections

04:08.615 --> 04:11.015
and specifically the 2005 elections under

04:11.045 --> 04:12.615
President Jose Mubarak.

04:12.985 --> 04:15.575
There were reputable monitors in these elections

04:15.575 --> 04:16.615
in the form of judges.

04:17.325 --> 04:19.015
They made it harder to commit fraud.

04:19.165 --> 04:20.855
They made it harder for candidates

04:20.855 --> 04:23.415
to observe whether fraud was committed on their behalf

04:24.035 --> 04:25.695
and consistent with our argument.

04:26.155 --> 04:27.615
As a result, we saw less fraud

04:27.635 --> 04:29.655
and more violence committed in these elections

04:29.655 --> 04:30.815
than in past elections.

04:30.875 --> 04:34.015
In this project, we test our argument at the micro level in

04:34.035 --> 04:36.495
one case, which was excellent at getting at really

04:36.855 --> 04:39.695
pinpointing the detailed mechanisms by which fraud

04:39.695 --> 04:41.015
and violence operate.

04:42.115 --> 04:44.615
I'm currently now working on a book project called Observing

04:44.615 --> 04:46.815
Electoral Violence, in which I look at this question

04:47.095 --> 04:49.495
globally based on interviews and quantitative methods.

04:50.355 --> 04:52.615
And in this book project, I look at two questions.

04:52.875 --> 04:55.695
So I look not only at why political actors use violence

04:55.695 --> 04:57.295
instead of fraud to steal elections,

04:57.295 --> 04:59.295
as I did in this project with my co-author,

04:59.755 --> 05:01.735
but I also look at the consequences

05:01.795 --> 05:03.215
of using violence versus fraud

05:03.215 --> 05:05.415
to steal elections both domestically

05:05.475 --> 05:07.815
and internationally in terms of foreign aid

05:07.815 --> 05:10.295
and investment for birth questions,

05:10.455 --> 05:12.415
I focus on the observability of violence

05:12.515 --> 05:15.215
and how it affects the decision either to use violence

05:15.515 --> 05:18.255
or how the public, the international,

05:18.255 --> 05:20.255
the international community responds to it.

00:00 - Electoral violence is a broad term and it includes many different forms of violence. It can include violence intended to delegitimize or stop an election from occurring. The Taliban when it was outside government used violence in this way.

00:13 - It also can include violence that occurs after an election in order to object to the outcome of the election, as we saw in the United States with the January 6th insurrection. But the type of electoral violence that my co-author and I analyze in this research is violence that is used to change the outcome of an election, and it can occur before, during, or after an election. Violence can change the outcome of election in a number of ways.

00:40 - It can force candidates to drop out of an election. It can prevent candidates from running an election by either killing them or severely injuring them. It can deter voters from not going to the polls to vote for certain candidates. It can even compel voters to vote for certain candidates that they wouldn't otherwise do in. In another context, it can even be used after votes have been counted to, to destroy the actual ballots.

01:07 - In this study, my co-author and I try to understand why candidates use electoral violence to steal elections instead of electoral fraud. There are many different reasons why candidates should not use violence to steal an election. So violence is potentially more repugnant to voters than fraud because of the physical harm that it causes and therefore could potentially alienate a candidate's own supporters.

01:32 - Violence is also easier to observe so that anyone who uses it is more likely to get caught. So our argument in this paper relates to the last point about the observability of violence. Our argument is distinct from the existing literature and that we argue that the observability of violence actually provides an incentive to candidates to use violence. So the observability we argue may co sit less likely that people that candidates hire to steal elections

02:00 - for them shirk their responsibilities by not committing either fraud or violence. With fraud, the risk of shirking is very high. A candidate, for example, can offer voters $50 in exchange for their vote. But once inside the polling booth and outside anyone's view, the voter can do whatever they want. They can vote for another candidate. They could submit a blank ballot or an invalid ballot,

02:23 - or they may not even vote at all. With violence, the risk of CI is much lower because violence is more observable and a candidate can better monitor whether this person commits violence for them or not. A candidate, for example, can offer a person $50 to attack another candidate. And unlike in the case of electoral fraud, the candidate can monitor well whether this person attacks the other candidate or not,

02:45 - because violence is more observable. The candidate may not personally observe the person attacking the other candidate, but they will hear about it in the news or by word of mouth from others who observe the attack or its aftermath. As a result, the person who's Paid to commit violence is less likely to shirk. If they do shirk, they're likely to get punished in some way.

03:04 - They may not get paid or hired for another job, or they themselves might get attacked. Testing this argument is really difficult because you need to measure something you can't actually see, and that's electoral fraud. So what we did was the best, the next best thing, and we looked at something we think affects the likelihood of persons to commit fraud, to shirk. And when the incentive to shirk is high, we expect electoral violence to be higher

03:27 - because the candidate can't rely on those. They they contract to commit vi fraud for them to actually do so and therefore switch to electoral violence. The thing that we look at in this study that affects the risk of shirking when it comes to fraud is electoral monitors. So electoral monitors can increase the incentive to shirk for two reasons. One, they make it harder to commit fraud,

03:50 - and two, they make it harder candidates to monitor whether the persons they hire to commit fraud do the do it for them or not. Electoral monitors do not affect the risk of shirking when it comes to violence because violence is observable whether monitors are present or not. So in this project, we look at the electoral monitors in Egyptian elections and specifically the 2005 elections under

04:11 - President Jose Mubarak. There were reputable monitors in these elections in the form of judges. They made it harder to commit fraud. They made it harder for candidates to observe whether fraud was committed on their behalf and consistent with our argument. As a result, we saw less fraud and more violence committed in these elections than in past elections.

04:30 - In this project, we test our argument at the micro level in one case, which was excellent at getting at really pinpointing the detailed mechanisms by which fraud and violence operate. I'm currently now working on a book project called Observing Electoral Violence, in which I look at this question globally based on interviews and quantitative methods. And in this book project, I look at two questions. So I look not only at why political actors use violence instead of fraud to steal elections,

04:57 - as I did in this project with my co-author, but I also look at the consequences of using violence versus fraud to steal elections both domestically and internationally in terms of foreign aid and investment for birth questions, I focus on the observability of violence and how it affects the decision either to use violence or how the public, the international, the international community responds to it.

Faculti AI can make mistakes