RiLACS: Risk-Limiting Audits via Confidence Sequences

TL;DR

This paper introduces RiLACS, a new method for conducting risk-limiting audits in elections using confidence sequences, which provide statistically reliable assurance of election outcomes efficiently across various election types.

Contribution

The paper develops a suite of tools based on confidence sequences and martingales for more efficient and versatile risk-limiting audits within the SHANGRLA framework.

Findings

Provides computationally efficient confidence sequences for election audits

Ensures high-probability correctness of election outcome verification

Applicable to diverse election types with improved statistical guarantees

Abstract

Accurately determining the outcome of an election is a complex task with many potential sources of error, ranging from software glitches in voting machines to procedural lapses to outright fraud. Risk-limiting audits (RLA) are statistically principled "incremental" hand counts that provide statistical assurance that reported outcomes accurately reflect the validly cast votes. We present a suite of tools for conducting RLAs using confidence sequences -- sequences of confidence sets which uniformly capture an electoral parameter of interest from the start of an audit to the point of an exhaustive recount with high probability. Adopting the SHANGRLA framework, we design nonnegative martingales which yield computationally and statistically efficient confidence sequences and RLAs for a wide variety of election types.