Turnout became “the main focus” of these contests. The Egyptian government devoted extensive resources to “turnout buying,” using a variety of inducements to get citizens to the polls. An Egyptian newspaper disclosed tactics ranging from distributions of cash and other goods, to cash prizes for locales exceeding 40 percent turnout, to businesses and state enterprises forcefully “encouraging” their employees to vote. The government also publicly threatened abstainers with a fine and reportedly sent police officers door to door to encourage participation.
As a result of forced turnout and a lack of true opposition candidates (Musa openly proclaims support for the incumbent), it appears that the most viable option for expressing dissent during this contest was to intentionally ruin one’s ballot. As one Egyptian explained, “if I voted, I’d spoil my ballot by ticking both boxes. I don’t want either of them.” Judging by the official results, spoilage was indeed a popular option this year. Sisi won 97 percent of the vote and Musa placed second with under 3 percent. But the 1,762,231 ballots spoiled in this election more than doubled the 656,534 total ballots cast for Musa, and far exceeded the number of ballots spoiled during the 2014 election.
Ballot spoilage can serve as an important metric to understand dissent and to read between the lines for what this election tells us about the state of opposition in Egypt. Research shows that ballot spoilage is increasing in many countries, including where voting is compulsory and in established democracies. For example, invalid votes surged as frustration with Brazil’s military regime (1964-1985) rose. In Morocco’s elections before the Arab Spring, observers claimed that disaffected citizens often spoiled their ballots with anti-regime slogans.
I explore vote spoilage in the 2018 presidential elections by combining demographic information from Egypt’s 2006 census with official election returns from the Egyptian electoral commission. The data include Egypt’s second-round presidential vote in June 2012, featuring Mohamed Morsi of the Muslim Brotherhood and former Mubarak regime official Ahmed Shafiq; results from the May 2014 which pitted the current president against Nasserist candidate Hamdeen Sabahi; and official results from last month’s election. This analysis draws on ongoing joint work with Dr. Steven Brooke at the University of Louisville about voting behavior in Egypt before and after the 2013 coup.
In Egypt, official data comes with several caveats. First, it is aggregated to the district level, creating an ecological inference problem when making claims about individuals’ behavior. Second, I openly register skepticism: These tallies may have been manipulated by the regime. Recent statistical advances have allowed for forensic tests to identify irregularities in the reporting of vote totals. However, even these techniques are unable to distinguish between deliberate fraud and human error, in the absence of other indications.
My preliminary analysis and firsthand accounts from election monitors suggest whatever fraud occurred was related to the regime’s vote-buying strategy and coercion of opposition candidate before the vote itself, rather than in reporting.
In my analysis, I control for a district’s population, percent female, percent illiterate (under Mubarak, illiterate voters were found to be more likely to spoil their ballots due to problems understanding how to vote correctly), and whether the district is part of the Sinai frontier. I also account for the district’s turnout in 2018. Finally, I control for spoilage in the second round of the 2012 president election — the most free and fair election in Egyptian history — as a baseline expectation for “true” levels of spoilage. Whatever variation remains driven by independent variables of interest suggests a protest vote.
Political and economic protest
In the current context, ballot spoilage could indicate at least two different kinds of protest: political and economic. First, spoilage may indicate a political protest from those whose first electoral choices were eliminated during this tightly controlled contest. This includes both 2012 Morsi supporters and 2014 Sabahi supporters. And indeed, there is a significant and positive relationship between a district’s Morsi vote share in 2012 and spoilage in 2018. While the expected positive relationship exists between Sabahi vote share in 2014 and spoilage in 2018, it is not a significant predictor.
Second, ballot spoilage could indicate economic protest. In classic pocketbook voting, voters punish the incumbent for a declining economy. I find that a district’s reliance on the tourism industry, which made up over 11 percent of the Egyptian economy in 2011 but has since been decimated by political instability and terrorism, is not significantly related to spoilage in 2018. However, districts with a higher percentage of public-sector employees do spoil ballots at higher rates. While public employees can be induced to show up, they spoil their ballots in the privacy of the voting booth.
One former Egyptian government official optimistically characterized spoiled ballots as a sign of credibility and a healthy democratic process. These preliminary results confirm that, but perhaps not in the way this official intended. Because the Sisi regime has closed off many avenues of opposition and coerced citizens into participation, ballot spoilage remains one of the ways that citizens can voice their dissent. Ballot spoilage may be one of the few remaining metrics through which to understand sources of dissent in contemporary Egypt.
Elizabeth R. Nugent is a postdoctoral research fellow in the Middle East Initiative at the Harvard Kennedy School of Government’s Belfer Center for Science and International Affairs.