Benford's Law and its Applicability in the Forensic Analysis of Electoral Results

Authored by Gonzalo Castaneda

Date Published: 2011

Sponsors: No sponsors listed

Platforms: NetLogo

Model Documentation: Other Narrative Mathematical description

Model Code URLs: Model code not found

Abstract

This article analyses the viability of Benford's Law in the forensic study of the detection of electoral frauds. According to this law, the initial digits of a set of numbers follow a logarithmic distribution as long as the data has not been disturbed. Implementation of this law in socioeconomic data-or in other types of data-depends on the presence of a two-fold randomized process: events of a distribution, and probability distributions selected from a reduced group. Through agent-based models, it is demonstrated that this law does not offer a robust test to distinguish between clean elections and elections that have been manipulated. In the model-in-question, partisan preferences are modified by means of social transmission and the parameters are calibrated with data from the Mexican 2006 elections.

Tags

Agent-based models Benford's Law Mexican 2006 elections analysis of electoral fraud political preferences