Discovering and Defining Electoral Violence

Discovering Electoral Violence

Most of our violent events were collected from digitised Victorian newspaper reports.  As such, this map is not an entirely comprehensive or precise record of election violence.  Newspapers did not always report smaller-scale or individual events, and were unable to report on those events so small that they entirely escaped journalistic notice.  Moreover, Victorian newspaper reports are often biased and selective, particularly with regard to the partisan, religious, and national identities of perpetrators and targets.

Defining ‘Electoral Violence’

The events contained in our map conform to the following working definition of electoral violence:

Physical damage, or the explicit and immediate threat of physical damage to persons and/or property that directly results from the local and national electoral cycle.

Separating election violence from election NON-violence

Physical damage to persons or property is included as electoral violence. Moreover, there are cases in which the threat of violence was so explicit and immediate that they merit inclusion in our map. For instance, a member of a crowd throwing a rock at a candidate falls under our broader definition of election violence, even if the stone did not actually hit its intended target, because the candidate managed to dodge the missile.

As such, events are be included if the threat of damage was:

EXPLICIT (an unambiguous and recorded threat from potential perpetrators e.g. perpetrators waving bludgeons at potential targets, or making credible verbal threats of physical harm)


IMMEDIATE (likely to occur imminently during a discrete occasion, such as an individual confrontation or public meeting, potentially within minutes of the threat’s presence).

Separating ‘ORDINARY’ violence during elections from ELECTORAL violence

Violence may occur for reasons unrelated to the election, which merely happens during election time. We do not count this as electoral violence. All instances connected to the local and national electoral cycle are included in the map, even if motivation is not discussed, but an explicit electoral connection is discussed in the source material.

Clear accidents arising from elections which result in harm, are excluded from the map (e.g. a carriage carrying voters to poll breaking apart due to wear and tear, causing injury). Also excluded are cases of electoral excitement, in which the excitement of an election is the suggested cause of injuries such as heart attacks.

Separating electoral violence from ‘POLITICAL’ violence

Given the central place of elections in Victorian political life, we class all explicitly political violence taking place during election periods as electoral violence. Elections were thought to be important (if not pivotal) political events, and close attention was paid to their processes and outcomes – thus, political activities such as demonstrations or protests which occurred during election periods were generally connected to the occurrence of contests.

Where this is made clear in source material, we exclude political violence taking place during General Election periods which relate to non-national contests.  This includes by-elections, municipal elections, elections of churchwardens, poor law commissioners etc.

Defining an ‘Event’

This is based largely on (1) The location of events first, then (2) the time of their occurrence. In cases where events explicitly move from one location to another (e.g. a violent procession), they are combined; in most other cases, they are separated into different events. Generally, the content of newspapers was used as a guide to separate or combine events taking place in the same place, but at different times. Thus, a newspaper report on ‘rioting over two days’ might be combined, whereas a report (or separate reports) on seemingly-discrete disturbances on a Tuesday, then a Wednesday, might be separated.