Automatic Margin Computation for Risk-Limiting Audits
A risk-limiting audit is a statistical method to create confidence in the correctness of an election result by checking samples of paper ballots. In order to perform an audit, one usually needs to know what the election margin is, i.e., the number of votes that would need to be changed in order to change the election outcome.
In this paper, we present a fully automatic method for computing election margins. It is based on the program analysis technique of bounded model checking to analyse the implementation of the election function. The method can be applied to arbitrary election functions without understanding the actual computation of the election result or without even intuitively knowing how the election function works.
We have implemented our method based on the model checker CBMC; and we present a case study demonstrating that it can be applied to real-world elections.
KeywordsRisk-limiting audit Margin computation Software bounded model checking Static analysis
This work has been partly supported by COST Action IC1205 on Computational Social Choice. This publication was made possible in part by the DemTech grant 10-092309 from the Danish Council for Strategic Research, Program Commission on Strategic Growth Technologies and in part by NPRP Grant #7-988-1-178 from the Qatar National Research Fund (a member of Qatar Foundation). The statements made herein are solely the responsibility of the authors.
- 1.Andoni, A., Daniliuc, D., Khurshid, S.: Evaluating the “small scope hypothesis”. Technical report, MIT Laboratory for Computer Science, Cambridge, MA (2003)Google Scholar
- 3.Beckert, B., Goré, R., Schürmann, C., Bormer, T., Wang, J.: Verifying voting schemes. J. Inf. Secur. Appl. 19(2), 115–129 (2014)Google Scholar
- 5.Blom, M.L., Stuckey, P.J., Teague, V., Tidhar, R.: Efficient computation of exact IRV margins. Computing Research Repository (CoRR) abs/1508.04885 (2015)Google Scholar
- 6.Cary, D.: Estimating the margin of victory for instant-runoff voting. In: Conference on Electronic Voting Technology/Workshop on Trustworthy Elections (EVT/WOTE). USENIX Association (2011)Google Scholar
- 8.Statistik, D.: Befolkning og valg (2015). http://www.dst.dk/valg/Valg1487635/other/2015-Folketingsvalg.pdf. Accessed 23 August 2016
- 9.Eén, N., Sörensson, N.: An extensible SAT-solver. In: International Conference on Theory and Applications of Satisfiability Testing (SAT), Selected Revised Papers, pp. 502–518 (2003)Google Scholar
- 10.Elklit, J., Pade, A.B., Nyholm Miller, N.: The parliamentary electoral system in Denmark (2011). http://www.ft.dk/Dokumenter/Publikationer/Engelsk/The_Parliamentary_Electorial_System_Denmark.aspx. Accessed 23 August 2016
- 13.Jackson, D.: Software Abstractions: Logic, Language, and Analysis. MIT Press, Cambridge (2006)Google Scholar
- 15.Magrino, T.R., Rivest, R.L., Shen, E., Wagner, D.: Computing the margin of victory in IRV elections. In: Conference on Electronic Voting Technology/Workshop on Trustworthy Elections (EVT/WOTE). USENIX Association (2011)Google Scholar
- 16.Sarwate, A., Checkoway, S., Shacham, H.: Risk-limiting audits and the margin of victory in nonplurality elections. Stat. Polit. Policy 4(1), 29–64 (2013)Google Scholar
- 17.Smith, A.M., Butler, E., Popovic, Z.: Quantifying over play: constraining undesirable solutions in puzzle design. In: International Conference on the Foundations of Digital Games (FDG), pp. 221–228 (2013)Google Scholar
- 18.Stark, P.B.: Super-simple simultaneous single-ballot risk-limiting audits. In: Conference on Electronic Voting Technology/Workshop on Trustworthy Elections (EVT/WOTE), pp. 1–16 (2010)Google Scholar
- 19.Stark, P.B., Teague, V.: Verifiable european elections: risk-limiting audits for D’Hondt and its relatives. USENIX J. Elect. Technol. Syst. (JETS) 1, 18–39 (2014)Google Scholar
- 20.Vorobyov, K., Krishnan, P.: Combining static analysis and constraint solving for automatic test case generation. In: Fifth IEEE International Conference on Software Testing, Verification and Validation (ICST), pp. 915–920 (2012)Google Scholar