The Survey Bangladesh's Media Cannot Afford to Ignore

Bangladesh has no shortage of political commentary. Television talk shows dissect every speech, every controversy and every maneuver by the country’s political parties.

Newspapers and online portals routinely turn statements by political leaders into banner headlines. Social media amplifies rumor, anecdote and partisan claims within minutes.

Yet one of the most substantial recent measurements of what Bangladeshis themselves think has received remarkably little attention. That should concern us.

The International Republican Institute’s Center for Insights in Survey Research recently completed its National Survey of Bangladesh, with fieldwork conducted from May 16 to June 24 this year.

Whatever one thinks of IRI or of individual findings, this is not a casual internet poll.

The survey interviewed 5,000 voting-age adults across all 64 districts and all eight divisions, using computer-assisted, in-person interviews in respondents’ homes.

It employed a multistage probability sample, stratification by division and urbanicity, household selection procedures and Kish Grid respondent selection.

The results were weighed using the 2022 census for division, age, gender and education.

The reported margin of error for the full sample is ±1.7% points at the 95% confidence level.

In a country where political claims are routinely built on rallies, Facebook engagement, elite conversations and anecdotal impressions, a nationally representative survey of 5,000 people deserves serious attention.

The more interesting question is why it has received so little.

The Bangladesh in the Survey is More Complicated Than the Bangladesh in Our Headlines

Perhaps part of the explanation is that the findings do not fit comfortably into anyone’s preferred political story. Consider the country's general direction.

The survey finds 46% saying Bangladesh is heading in the right direction and 44% saying it is heading in the wrong direction, with 9% unsure or declining to answer.

That is essentially a divided country, rather than the picture of overwhelming national despair or overwhelming satisfaction that partisan narratives sometimes suggest.

The reasons respondents give are equally instructive.

Among those who believe the country is moving in the right direction, the leading spontaneous explanation is improved economic conditions, at 23%, followed by improvement in law and order at 17%.

Among those who believe the country is going in the wrong direction, commodity-price increases lead at 18%, followed closely by deterioration in law and order and the judicial situation at 17%, and corruption at 10%.

In other words, ordinary Bangladeshis appear less obsessed with the daily political theater than much of the political class is.

Their concerns are strikingly practical: prices, jobs, corruption, security and governance.

When respondents were asked, without being given a list, to identify the single most important problem facing Bangladesh, corruption ranked first at 22%, followed by drugs at 15%, crime and security at 12%, inflation at 10% and unemployment at 9%. Political instability came later, at 7%.

That ought to generate weeks of serious journalism. Instead, much political coverage continues to revolve around what politicians say about one another.

Some Findings Should Force Political Elites to Reconsider Their Assumptions

The survey's findings on democracy are particularly significant. Despite all the anxiety surrounding Bangladesh's political transition, 62% describe the current state of democracy as either very good or somewhat good, compared with 32% describing it as somewhat or very bad.

That does not mean Bangladeshis are satisfied with their political establishment. Far from it.

An extraordinary 83% agree that there is a large gap between political elites and ordinary people, with 48% strongly agreeing and another 35% somewhat agreeing.

That may be among the most consequential findings in the entire report.

Citizens can believe democracy is functioning reasonably well while simultaneously believing that the people who dominate politics have become detached from them.

Those positions are not contradictory. They suggest that Bangladeshis may value democratic institutions more than they value the political class operating within them.

There are similarly important findings on religious pluralism.

74% say elected office should be open to people of all religions, while 20% say it should be restricted to Muslims. Among Muslim respondents themselves, 73% favor opening elected office to all religions.

That finding deserves particular attention because Bangladesh's political discourse is frequently framed through sweeping claims about religious polarization.

A serious press should test those narratives against evidence.

Then There Are The Political Parties

Here the survey becomes especially interesting. IRI asked respondents whether they liked or disliked various political parties, rather than simply asking which party they would vote for.

The results therefore provide an unusually revealing picture of each party's broader political standing.

BNP has the largest positive rating: 27% strongly like it and 30% somewhat like it, for 57% combined. Jamaat-e-Islami follows with 16% strongly and 32% somewhat liking it, or 48% combined.

The NCP records 37% combined positive sentiment, while Islami Andolan Bangladesh receives 33 percent.

But the trend deserves as much attention as the snapshot. Between October 2025 and June 2026, the BNP's combined positive rating increased from 51 to 57%.

Jamaat's moved from 53 to 48%. The detailed trend chart also shows the changing balance of strong and softer support across the other parties.

These are precisely the kinds of numbers that political journalists normally spend enormous energy debating. So where is the sustained discussion?

Where are the front-page analyses examining what the findings mean? Where are the television roundtables comparing demographic groups and regional differences?

Where are the interviews with pollsters challenging their methodology? Where are the serious examinations of why particular parties are gaining or losing favor?

The Election Findings are Even Harder to Dismiss

IRI asked respondents about the February 2026 parliamentary election. The answer is striking.

Seventy-three percent regarded the election as either mostly free and fair or somewhat free and fair: 38% said it was mostly free and fair and 35% somewhat free and fair.

Only 22% described it as somewhat or mostly rigged.

Moreover, 77% are optimistic that future elections will be free and fair, compared with 19% who are pessimistic.

These findings do not settle arguments about Bangladesh's democracy. No opinion poll can do that.

But they certainly complicate narratives suggesting that ordinary Bangladeshis overwhelmingly regard the country's electoral system as illegitimate.

And complicating narratives is precisely what good journalism is supposed to do.

This is Not an Argument that the Survey is Beyond Criticism

IRI should not be treated as an oracle. Its methodology should be examined. Its questionnaire wording should be scrutinized. Researchers should look carefully at sampling, weighting, response effects and question order.

IRI itself notes an important methodological change: Education was incorporated into the weighting scheme using the completed 2022 census data, making direct comparisons with earlier surveys more complicated.

That is how intellectually serious people should approach polling. Challenge it. Replicate it. Compare it with other surveys. Interview the researchers. Invite statisticians to criticize it. But ignoring inconvenient evidence is not methodological criticism. It is an editorial selection.

And That Brings Us to the Media 

The irony becomes almost impossible to miss when one reaches the survey's own section on media. Television remains overwhelmingly the principal news source for Bangladeshis.

59% say television is the medium from which they are most likely to obtain news, compared with 19% for social media, 12% for non-social-media internet sources, and only 4% for newspapers.

The survey also asked which specific outlet respondents considered most trustworthy for political news.

No outlet dominates. Somoy TV leads at 15%, followed by Jamuna TV at 12% and NTV at 8 %, while 9% say none.

Even more revealingly, when respondents were asked to evaluate major institutions, 59% approved of the media's performance while 37% disapproved, producing a net approval of +22.

That is not a catastrophic verdict. But neither is it a blank cheque.

The media should therefore ask itself an uncomfortable question: Are we reporting public opinion, or are we selecting which public opinion deserves to become visible?

Media bias does not always take the crude form of publishing false information. Often it operates through selection, prominence and omission.

A newspaper need not falsify a survey to distort the public conversation. It can simply ignore it. A television channel needs not manufacture numbers. It can devote hours to politically convenient anecdotes while giving almost no airtime to statistically rigorous evidence that challenges the assumptions of its preferred guests.

That form of bias is harder to detect precisely because every individual story may be factually correct. The distortion emerges from what repeatedly gets covered and what repeatedly does not.

Selective Silence Deserves Scrutiny

It would be irresponsible to declare that limited coverage of this survey automatically proves a coordinated political agenda. Newsrooms have deadlines, competing priorities and limited resources.

But selective enthusiasm for polling is itself a legitimate subject for scrutiny. If surveys become prominent when their findings reinforce prevailing narratives, but disappear when equally rigorous findings complicate those narratives, readers have every right to ask why.

The standard cannot be that opinion polls are authoritative when their findings are politically convenient and irrelevant when they are not.

A credible media culture must apply the same standard of scrutiny to evidence regardless of who benefits from it.

Bangladesh Needs An Evidence Culture

There is a larger issue here than BNP, Jamaat, NCP or any other party. Bangladesh badly needs a political culture in which empirical evidence matters more than elite assertion.

We cannot complain about misinformation on Facebook while mainstream journalism itself gives insufficient attention to high-quality data.

We cannot demand informed voters while feeding citizens an endless diet of speeches, accusations and talk-show confrontation.

And we cannot claim to speak for "the people" while showing limited curiosity about what a carefully constructed national sample of those people says.

The IRI survey should not be celebrated because its findings please one political camp. Nor should it be dismissed because some findings irritate another.

It should be reported. Its methodology should be examined. Its surprising findings should be debated. Its weaknesses should be identified.

Political parties should be asked to respond to it. Independent researchers should compare it with other evidence.

That is what journalism does in a democracy. Because the most troubling media bias is sometimes not contained in what appears on the front page.

It lies in what never makes the front page at all.

Dr Mohammad Nakibur Rahman is a professor of Finance at the University of North Carolina and the US spokesperson for Jamaat-e-Islami Bangladesh.