ABSTRACT
This research note scrutinizes official vote counts from the 2018 Egyptian presidential election. Drawing on data for 13,807 polling stations published by the Egyptian Electoral Authority, it uses the Election Forensics Toolkit to estimate a series of digit and distribution tests. The results point to statistical anomalies in voter turnout for polling stations located in Egypt’s major population centres. The findings align with journalistic accounts of very low turnout in those areas and demonstrate how tools from electoral forensics can be applied to the study of elections in the Middle East and North Africa.
Acknowledgment
Walter Mebane provided guidance on using the Electoral Forensics Toolkit and generously shared the code to estimate the finite mixture models. The brilliant team at Daftar al-Ahwal provided access to an unreleased version of the WikiThawra dataset. Steven Brooke, Thoraya El-Rayyes, Marc Lynch, and two anonymous reviewers gave useful feedback and suggestions on the text.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. See Hicken and Mebane (Citation2017, pp. 37–38) for a detailed explanation of how typical values are derived.
2. On elections held under the Mubarak regime, see Kassem (Citation2004) and Masoud (Citation2014).
3. The exception to this trend is the 2014 parliamentary election. This may reflect the regime’s priorities: an enfeebled parliament with no popular mandate is less able to hold El-Sisi to account.
4. At the time of writing, this information was recently removed from the website only to subsequently reappear. Regardless, the original page and corresponding Excel file can be recovered through the WayBack Machine (see the bibliography).
5. This information could potentially be recovered by triangulating polling station names with the known location of polling stations used in previous elections.