Skip to content

Design Documentation: Fraud Detection

Christopher H. Gian edited this page Mar 19, 2019 · 2 revisions

CIVIC Software Foundation Proposal for the Open and Accountable Elections (OAE) Program Application

Purpose

Mechanisms are intended to assist OAE staff in detecting fraud and do not need to be 100% automated. The idea is to provide a flexible interface for the campaigns to easily report their transactions—ideally with as little manual data entry as possible on their part—and to provide OAE staff as many ways as possible to sort and compare the data for “gut checks” in addition to some automated detection of unusual patterns.

Internal Note The data should be as linked as possible to allow for easy and fast comparisons.

Types of Fraud

  • Forged signatures on attestation forms
  • Person is not of age or does not have City of Portland Residency
  • Rich donor divides up donations among real (or fake) small donors to help candidate reach threshold (e.g. with money orders attributed to different givers)

Verification Mechanisms

  • Is the donor of age?
  • Does the donor have residency in City of Portland?
  • Is the donor eligible to donate?
  • Has the donor reached the matching threshold on eligible donations?

Data Sources

  • Oregon voter file
  • Arts Tax?
  • Tax Lots
  • Zip Code Data
  • Additional age/address verification: from API if need be, but minimize as second-line due to cost concerns