Dual approach to proving electoral fraud via statistics and forensics (Dvojnoe dokazatel'stvo falsifikazij na vyborah statistikoj i kriminalistikoj)

Andrey Podlazov,Vadim Makarov

2024-12-06

Abstract:Electoral fraud often manifests itself as statistical anomalies in election results, yet its extent can rarely be reliably confirmed by other evidence. Here we report complete results of municipal elections in Vlasikha town near Moscow, where we observe both statistical irregularities in the vote-counting transcripts and forensic evidence of tampering with ballots during their overnight storage. We evaluate two types of statistical signatures in the vote sequence that can prove batches of fraudulent ballots have been injected. We find that pairs of factory-made security bags with identical serial numbers are used in this fraud scheme. At 8 out of our 9 polling stations, the statistical and forensic evidence agrees (identifying 7 as fraudulent and 1 as honest), while at the remaining station the statistical evidence detects the fraud while the forensic one is insufficient. We also illustrate that the use of tamper-indicating seals at elections is inherently unreliable. -- -- Tezis po-russki est' v russkoj versii stat'i (normal'noj kirillicej, ne translitom)

Applications,Physics and Society

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to double - verify the phenomenon of election fraud through statistical and forensic means. Specifically, the author attempts to solve the following key problems:

1. **Detecting election fraud**:
- The author uses statistical methods to analyze abnormal patterns in the voting sequence to identify possible fraudulent behavior. They designed two simple statistical tests to calculate the probabilities of the longest consecutive voting sequences and non - voting sequences (\(\alpha_0\) and \(\alpha_1\)), as well as the probability of the number of consecutive sequences (\(\tilde{\alpha}\)). These probability values can help determine whether there is a situation of mass - forging ballots.

2. **Providing physical evidence**:
- The author not only relies on statistical data but also combines forensic evidence. Through a detailed inspection of the seals on the security bags, signs of tampering were found in the security bags in multiple constituencies. For example, pairs of security bags with the same serial number were found, indicating that these bags were deliberately made in duplicate in the factory and were replaced or swapped when there was no supervision at night.

3. **Verifying the scope of fraud**:
- The author conducted a comprehensive analysis of the local elections in the town of Vlasikha near Moscow, including 9 physical polling stations and one electronic polling station. By comparing the statistical and forensic evidence of each polling station, it was confirmed that 7 out of 8 polling stations had fraudulent behavior, while in the remaining 1 polling station, although the statistical data indicated fraud, the forensic evidence was insufficient.

4. **Evaluating the effectiveness of existing anti - counterfeiting measures**:
- The study also explored the effectiveness of the anti - tampering seals used in the current elections, pointing out that these seals are not reliable in themselves and can be easily bypassed by motivated fraudsters. Through a detailed analysis of the seals, the author revealed the traces of the seals being replaced or removed.

5. **Revealing the impact on election results**:
- Finally, through the combination of statistical and forensic evidence, the author revealed the huge impact of fraud on the election results. Originally, opposition party candidates (such as candidates nominated by the Communist Party) might have won the majority of seats, but due to fraudulent behavior, all the finally elected candidates were those who supported the government.

### Formula summary

- **Calculation of statistical significance**:
- First test: Calculate the probabilities \(\alpha_0\) and \(\alpha_1\) of the longest consecutive voting sequences and non - voting sequences.
- Second test: Calculate the probability \(\tilde{\alpha}\) of the number of consecutive sequences.
- Corrected significance: For a constituency with \(c\) candidates, the minimum significance is \(\alpha' = 1-(1 - \alpha_{\text{min}})^{2c}\approx 2c\alpha_{\text{min}}\) and \(\tilde{\alpha}' = 1-(1 - \tilde{\alpha}_{\text{min}})^c\approx c\tilde{\alpha}_{\text{min}}\).

- **Probability representation**:
- Represent probability using power: \(p_\alpha = -\log_{10}\alpha\), where each increase of 1 unit indicates a decrease of 1 order of magnitude in probability.

Through these methods, the author can identify the existence of election fraud and its scope of influence with a high degree of confidence.