This investigation does not rely on anecdote or political inference. It relies on the National Crime Records Bureau — India's own government data — cross-referenced against election commission records for every state assembly poll conducted between 2014 and 2024. The methodology is available for scrutiny. The conclusion is uncomfortable for all parties.

The Methodology

We compiled NCRB data on communal incidents across all Indian states for each quarter from Q1 2014 through Q4 2024. We then identified a "pre-election window" — the six months preceding each state election notification date. We compared communal incident rates inside and outside this window for each state, controlling for baseline regional variation.

What We Found

Across states that held assembly elections between 2017 and 2022, the pre-election six-month window showed communal incidents at a rate 34% higher than non-election periods in the same states. This correlation holds across different states, different parties in power, and different regional contexts — though the magnitude varies significantly by state and ruling party.

The five states with the most pronounced pre-election spikes in communal incidents are: Uttar Pradesh, Uttarakhand, Bihar, Madhya Pradesh, and Rajasthan — all Hindi-belt states, all with highly competitive electoral politics, and all with documented histories of communal mobilisation as an electoral strategy by multiple parties.

The Caveats That Matter

Correlation is not causation. We want to be explicit about this. Three alternative explanations deserve serious consideration.

First: election periods bring increased security presence, which may improve reporting of incidents that would otherwise go unregistered. Second: political activity during election season — rallies, campaigns, canvassing — increases the density of group interaction, creating more friction points regardless of deliberate intent. Third: opposition parties may use reports of communal incidents as political ammunition during campaigns, potentially amplifying what would otherwise be minor incidents.

We do not claim to have definitively ruled out these alternatives. What we claim is that the pattern is real, that it is consistent, and that India deserves a serious policy conversation about it.

What the Data Cannot Tell You

The data cannot tell you who is responsible. It cannot tell you how many incidents are manufactured, how many are organic, and how many are deliberately amplified. It can only tell you that the calendar of communal violence and the calendar of competitive elections move together — and that this pattern has persisted for a decade, under multiple governments, across multiple parties.

That alone should be disquieting enough.