Quick Blog · Data Notebook

Data tidbit shows Jamaat’s showing among women voters

In about 100 centres where men and women voted separately, Jamaat fared better among women, especially when they carried a constituency.

Nazmul Ahasan·February 27, 2026

There’s a common assumption I’ve heard repeatedly since the election: Jamaat-e-Islami’s failure to nominate a single woman candidate, combined with several controversies around its problematic views on women, must have hurt it badly among female voters.

That sounds logical.

But when we look at the data -- even cautiously -- the picture becomes more complicated.

For this analysis, I zoomed in on 49 polling complexes where men and women voted in separate centres (footnote 1). That gave us 98 centres in total: 49 male-only and 49 female-only. Importantly, these were places where BNP and Jamaat directly faced each other. I excluded constituencies where alliance partners were in the race.

Why does that matter? Because this lets us compare almost the same number of men and women within the same locality, inside the same polling complex, under the same political conditions.

Across these 98 centres, BNP won 61 and Jamaat won 37. So BNP clearly led overall. It also won more female-only centres than Jamaat did, taking 29 to Jamaat’s 20.

However, that also means, Jamaat won more women-only vote centres than men-only ones, whereas BNP had won more men-only ones.

Since the number of women and men registered voters is almost equal in all but a very few centres, it's less problematic to compare the vote totals. So let's do that.

In these 98 centres combined, BNP received 49.2% of the valid votes cast, while Jamaat got 46%. That’s already pretty modest. But when you isolate female-only centres, the gap narrows even further. BNP received 48.2% of female votes. Jamaat received 47.6%.

Another way to see it: women made up 47.8% of BNP’s total votes in this sample. For Jamaat, women made up 50.5% of its total support. So women formed a somewhat larger share of Jamaat’s own voter base.

Now, before anyone runs too far with this, a big caveat.

This subset favours Jamaat more than the national picture does. Nationally, in constituencies where BNP and Jamaat competed directly, Jamaat received 41.3% of the valid votes cast compared with BNP’s 50.4%. That is a 9.1pp advantage for BNP. In this sample, however, the gap is only 3.2pp. So the sample is not nationally representative. It likely overrepresents areas where Jamaat was relatively competitive or strong.

Still, even within that context, the gender breakdown is revealing.

In constituencies that BNP won overall, it secured 52.5% of female votes and 53.4% of male votes cast -- a clear edge. But even there, women made up 49.7% of Jamaat’s total support, compared with 48.2% of BNP’s. So Jamaat’s female base was far from disappearing in BNP strongholds.

In constituencies that Jamaat won, the pattern becomes sharper. Women accounted for 51.7% of Jamaat’s total votes, while they made up 46.7% of BNP’s. Where Jamaat succeeded, women were often central to that success.

The total turnout in this sample was 60.9%, not very far from the national turnout of 59.4%. But turnout differed by gender. Male-only centres saw a turnout of 63.5%, while female-only centres recorded 58.2%. In constituencies Jamaat won, overall turnout was slightly higher at 62%. In both BNP- and Jamaat-winning constituencies, the gap between male and female turnout hovered at around 5pp, with men consistently turning out at higher rates.

If anything, that makes the pattern more striking. With women turning out at lower rates than men, you might expect their weight in the final outcome to be slightly reduced. Yet Jamaat’s relatively strong showing in female-only centres -- and the larger share of women within its vote base in seats it both won and lost -- stands out even more under those conditions.

Dataset

What makes this dataset analytically useful is that it’s geographically controlled. We’re not mixing different constituencies of wildly different sizes. These are matched male and female centres within the same polling complex. The gender balance in registration is also tight. The median share of female registered voters was 50.8%, and the average was 50.9%

Geographically, the 98 centres were spread across 27 constituencies. Khulna-2 contributed the largest share, followed by Rajshahi-2, Satkhira-2, Khulna-3, Sylhet-4, Naogaon-6 and Dinajpur-2. Seventy of the centres were located in constituencies won by BNP and 34 in constituencies won by Jamaat. Again, not nationally balanced.

Behind this small slice sits a much larger data effort. At Netra News, we’ve gathered (footnote 3) more than 17,000 polling station result sheets outside Dhaka and Mymensingh divisions, which we estimate is about 60% coverage of those regions. From that universe, we analysed roughly 8,000 centres. Of those, around 1,600 to 1,700 were single-sex centres. A smaller subset involved direct BNP versus Jamaat contests. The rest were mixed centres, where you simply cannot isolate male and female voting behaviour.

In earlier analysis of all available single-sex centres combined, I found a similar pattern: in centres Jamaat carried, female-only stations outnumbered male-only ones, while BNP’s ratio skewed more male. This smaller, tightly matched sample reinforces that broader finding.

So what does this mean?

It doesn’t mean Jamaat dominates among women nationwide. It doesn’t overturn the national results.

But it does challenge a comforting narrative.

In parts of rural Bangladesh, especially outside Dhaka and Mymensingh, Jamaat’s appeal among women appears stronger than many assumed. In constituencies where it performed well, women were often a decisive share of its base. Even in areas where it lost, its female vote share was still respectable.

Please note, however, that this dataset is not statistically random. It may suffer from clustering effects. It likely overrepresents certain regions. So it should be read with caution.

But if the assumption was that women voters would automatically reject Jamaat because of its rhetoric or candidate choices, the numbers -- at least in these localised comparisons -- suggest something more complex.

And politically, that complexity matters.

Download the full 98-centre sample dataset below.

Could not load data/same-premise-centers.csv. Keep this file at quick-blog/jamaat-women-votes/data/same-premise-centers.csv.

Loading data…·Source: same-premise-centers.csv