Bangladesh has two dominant political parties: the Awami League (AL) and the Bangladesh Nationalist Party (BNP). For most of the country's parliamentary history they have alternated in government, with caretaker periods between. They also regularly meet on the streets. The BPO record codes incidents where party workers from each side are present together as primary and secondary actors — "AL × BNP" events in shorthand.
The total volume is meaningful but not overwhelming: AL × BNP direct-confrontation events are a few percent of the 162,315-incident dataset. They concentrate in specific windows, follow a recognisable script, and have predictable triggers.
Three windows that account for most
1996. The anti-Khaleda movement, hartal- driven, including the boycotted February election and the June re-election. Heavy concentration in Dhaka division.
2013. The international crimes tribunal verdicts. AL and BNP-aligned student wings, plus the wider Islamist groupings, confronted each other and the security services. The casualty counts are the highest in the dataset for any non-election year.
2014 and 2018. The pre-election windows. AL × BNP confrontations spike in the weeks before voting and again in the post-poll arrest waves.
Outside those windows, direct AL × BNP confrontations are sparse — usually local nominations, local committee disputes, or single-upazilla flare-ups. The base rate is low.
What the script looks like
Read 30 AL × BNP incidents end-to-end and a script emerges. Setup: a procession (mosse, rally, victory parade) crosses territory the other party considers theirs. Confrontation: words, then stones, then sticks, occasionally weapons. Police intervention. Arrests on both sides. A statement from each party-office. A press cycle of 24-48 hours. Most cases do not escalate beyond this pattern.
The cases that do escalate share features: pre-existing local rivalries between specific named ward-level leaders; contested election outcomes within a year; or a triggering event in the national news cycle that arrived in town hot.
What changed after 2014
After the 2014 election the BNP's organisational footprint in many districts contracted. AL × BNP direct confrontation rates fell — not because the rivalry softened, but because BNP presence on the street thinned. The BPO record picks up clearly: total AL × BNP coding drops sharply post-2014, recovers partially in 2018, and remains below the pre-2014 baseline through 2024.
Read this as a structural change in how the two-party rivalry plays out, not as a reduction in political polarisation. The polarisation is visible in other parts of the record — Hartal, DSA arrests, the Religious-minorities story on this site — just not in direct AL × BNP street confrontations.
The honest caveat
AL × BNP actor coding has been refined over the years. Some earlier incidents the desk would re-code under more specific actor tags now (Chhatra League, Jubo Dal, etc.) than under the parent-party tag. The trend lines are real but the absolute year-over-year comparisons benefit from a methodological footnote: actor refinement adds noise to comparisons before and after roughly 2015. The cross-cutting Election theme, analysed in a separate Story on this site, is the more stable lens for inter-party violence.