Polls and the postal ballot: The core question is trust

In a video that has gone viral on Facebook, Islami Andolon Bangladesh’s assistant secretary general Principal Maulana Sheikh Fazle Bari Masud claimed on Ekushey Television’s talk show Ekushey Raat that before election survey organisations took to the field, Jamaat leaders allegedly called Islami Andolon leaders and told them to have people ready at every intersection so that when the pollsters arrive, they can be shown people to indicate voting trends. I have not been able to independently verify this claim, but because it comes from a responsible figure, it has drawn significant public attention.

Although this statement and the recent controversy over expatriate postal ballots are two separate issues, in terms of social psychology they converge on the same question: how pre-election surveys and, on polling day, the ballot itself can both build public trust, and how, quietly, they can also undermine it.

Many people assume that surveys are simply scientific exercises involving a few questions and some numerical calculations. But in electoral politics, surveys are often not merely tools for gauging public opinion, they can also become instruments for shaping it. In other words, surveys do not only measure voters’ preferences, they can influence voters’ decisions as well.
Two of the most discussed effects of election surveys are the “bandwagon” effect and the “underdog” effect.

In the Encyclopedia of Survey Research Methods published by the internationally recognised academic publisher SAGE, Michael Traugott shows that after survey results are published, many voters tend to gravitate towards the likely winner. This is known as the bandwagon effect. It reflects a mindset that “it is safer to be on the winning side.”

On the other hand, some voters are motivated by sympathy for the trailing side and become inclined to support them. This is the underdog effect, driven by the reasoning that “they are weaker; they deserve a chance too.”
Traugott’s key argument is that once surveys are published, voters perceive them as social signals that can influence decision-making. This is where the political power of surveys lies. They do not determine the outcome in advance, but they can help establish in society the perception of who is likely to win.

As a result, surveys can influence election outcomes in three ways: by creating a “winning momentum” around the leading side; by portraying the trailing side as “deprived” and evoking sympathy; by shifting voters away from ideology-based voting towards strategic voting.

This is why, during elections, surveys are not just numbers. They become powerful political tools that shape public psychology. The question, then, is this: if a party or group wants to influence a survey, how do they “prepare” people in advance? This becomes possible by intervening at the most sensitive point of a survey -- sampling.

A survey may claim to have questioned 100 people; but if a large proportion of those 100 were already “prepared” or belonged to a particular political circle, the results no longer reflect the average opinion of society. The outcome then becomes what is known as staged responses, manufactured answers in which people are not expressing reality but presenting a “performed reality.”

In many countries, surveys are also used as political strategies. In particular, in “push polling,” the primary objective is not to measure opinions but to change them under the guise of a survey. Researchers and election observers often regard this as a form of negative campaigning.
In the United States and Europe, the combination of survey results, media framing, and last-minute campaigning creates an environment in which voters consider not only their preferences but also perceptions of “who is going to win.” This often increases strategic voting, abandoning one’s first choice to vote for a candidate perceived to have a realistic chance of victory.

In contrast, in India, rumours, misinformation, and digital campaigning have in many cases “hijacked” electoral public opinion. According to Al Jazeera Journalism Review’s report Elections and Misinformation: India Case Study (30 April 2024), online misinformation during the 2019 election not only shaped opinions but, in some cases, incited violence; WhatsApp emerged as a major conduit for spreading political falsehoods.
In this context, people no longer see surveys as neutral information; for many, surveys have become political weapons.

According to information from the Election Commission published in Prothom Alo (6 January 2026), a total of 1,527,000 postal voters were approved for the 13th national parliamentary election, of whom just over 760,000 were expatriate voters. In other words, a large proportion are abroad, while the rest are within the country.

Within Bangladesh, most postal voters are not ordinary voters; rather, they are government officials and employees, election-duty administrators, and members of law enforcement and service agencies who cannot vote at polling centres due to their official responsibilities.

According to official information from the Election Commission’s “Postal Vote BD” app, the app is primarily used for voter registration and ballot tracking. However, voting itself is still conducted through manual postal ballots, that is, the ballot is sent abroad by mail, filled out by the voter, and then returned by post to the relevant election official in Bangladesh.
While digital tracking is a strength, in practice a large part of the system depends on addresses, postal services, timely delivery, and intermediary controls. As a result, this is less a test of technology than a test of trust.

This system includes security measures such as facial verification and unique numbering/scanning on postal envelopes. However, voters see only an “OK/match” result and have no way of knowing where the approval logs are, where the audit reports are, who controls the data, or what avenues exist for appeal or verification in case of error. For this reason, the crisis of trust is not merely technological, it is also ethical and procedural.

Moreover, even with face recognition in place, the risk of spoofing (deception using photos, videos, or masks) cannot be entirely ruled out. Technical reports such as the European Commission Joint Research Centre’s Biometric Spoofing: A JRC Case Study in 3D Face Recognition, as well as reporting by the technology magazine Wired, show that the presence of facial verification alone does not eliminate the risk of manipulation; the real threat of spoofing or presentation attacks remains.

In this postal voting system, one of the least discussed yet major risks is the design of the ballot paper. Recent controversies have raised allegations about the placement of symbols and boxes. An international report by the Brennan Center, Design Deficiencies and Lost Votes, demonstrates that poor design can lead to over-voting (voting for more than one candidate), causing many ballots to be excluded from the count.

Similarly, the union of Concerned Scientists’ report Equitable Ballot Design and Voter Education Materials notes that poorly designed ballots result in a large number of votes not being counted. The harm caused by unclear design falls most heavily on ordinary voters. On a ballot, the eye tends to go first to boxes at the top or on the right. In particular, less educated or elderly voters may inadvertently stamp the wrong box.

Research on residual votes (undervotes/overvotes) by the Brennan Center and by Hanmer and his co-authors shows that votes are often lost not because of voter error, but because of flaws in design and systems. In the United States in 2000, Florida’s “butterfly ballot” design confused voters and ultimately took the election to the courts. The episode shows that design flaws are not trivial, they can change history.

Recently, following objections raised by a political party, the Bangladesh Election Commission decided to change the design of the postal ballot used within the country. In the new postal ballot, the names and symbols of candidates in the relevant constituencies are presented more clearly.
Taken together, the biggest test ahead is not who will win, but whether public trust in the vote will endure. Because once people begin to believe that surveys are used to “manufacture minds” and ballots to “manufacture results,” the damage goes beyond a single election, it undermines the moral foundations of the entire society.

* Mohammad Jalal Uddin Sikdar is a teacher and researcher,
Department of Political Science and Sociology, North South University
* The views expressed here are those of the author.