Ahead of the much-anticipated 13th parliamentary election held on February 12, analysts were busy debating possible outcomes, predictions, and political scenarios. Social media surveys, in particular, played a major role in shaping public discussion. From the capital to remote parts of the country, online polls increasingly appeared to function as an alternative indicator of political popularity.
For example, thousands of pages including those run by mainstream media organised comment-based voting on Facebook. Many YouTube channels also conducted polling, attracting large numbers of participants. In nearly 95 percent of these surveys, Bangladesh Jamaat‑e‑Islami was far ahead by a wide margin, with Bangladesh Nationalist Party (BNP) typically appearing as the closest competitor. Given historical trends and current political realities, such results seem surprising even improbable. So why does this happen? What are the real factors behind it?
Most netizens in Bangladesh are now familiar with the concept of ‘bot’ accounts and often attribute unusual online activity to them. But focusing only on bots overlooks deeper social and psychological dynamics. In many developing countries, including Bangladesh, responsible social media use remains a longstanding challenge. Sharing news without fact-checking, reacting emotionally to fake or AI-generated content, and spreading religious messages in the belief that doing so earns spiritual reward these behaviours have become part of everyday online culture.
Research shows that content built around religious emotion spreads extremely fast. Even a simple message like “Click like button if you love Islam” can reach millions despite containing no discussion of religious philosophy. When social media campaigns by Islamic political parties reach such audiences, they often receive overwhelmingly positive engagement simply because of the parties’ religious associations. This pattern is not unique to Bangladesh. Similar trends have been observed with Shiv Sena in India, a country where religious identity also strongly influences public response. As a result, religiously affiliated parties often outperform others in social media polls.
However, this is where the divergence begins. Many individuals who express support for religious parties online do so from a purely emotional or symbolic standpoint. A closer look at their profiles often reveals that they hold diverse and sometimes conflicting social views. This is where online support and actual voting behaviour split apart.
Even if someone taps a button in support of an Islamic party on their phone, their final decision at the ballot box is shaped by practical realities. Voters tend to consider which candidate is more likely to solve their personal problems, who has stronger social ties within their community, and which relationships or local dynamics might affect their daily lives. Personal and social considerations often outweigh symbolic or ideological preferences. Because the BNP maintains broader social representation across different segments of society, it has been able to secure major victories practically even after suffering heavy defeats in online polls.
To understand this phenomenon more deeply, we can turn to the concept of social desirability bias, developed by psychologist Allen Louis Edwards. Bangladesh has a unique socioeconomic structure in which, aside from dedicated party activists, most voters prefer not to publicly reveal their political preferences. When candidates campaign door-to-door, many voters politely assure nearly every one of their supports.
The same pattern appears online. Some voters intentionally avoid voting for their preferred candidate in public social media polls so others cannot identify their true political affiliation. Edwards’ theory explains precisely this kind of behaviour how perceptions can diverge from reality and how misleading impressions can spread through indirect channels. The 13th parliamentary election offered yet another clear demonstration of this dynamic in action.
The writer is a columnist
The views expressed in this article are solely those of the author
