The Two Biggest Parties in Parliament Ask About Everything. Everyone Else Doesn't.

Rank India's 51 parliamentary parties by how many questions their MPs asked in the 18th Lok Sabha and current Rajya Sabha sessions, and BJP wins, then INC — by margins that just track how many MPs each party has. That ranking answers a boring question. A different cut of the same 59,381 questions — each party's focus on a specific ministry, divided by what its overall size would predict — answers a more interesting one: which parties are actually specialists. The two biggest are not. Nearly everyone smaller is, sharply, on one or two ministries tied to exactly what you'd guess from their base.

The ranking everyone already expects

Every Lok Sabha question since the 18th Lok Sabha convened (June 2024) and every Rajya Sabha question across the matching sessions (265 to 271, the same calendar window) were matched to the asking MP's party via both chambers' own member rosters. Rank the 51 resulting parties by raw question count and the order is exactly what a seat count would predict.

Source: sansad.in and rsdoc.nic.in question APIs, matched to each chamber's own current/former member roster for party and state. See Sources for the full method and its limits.

That is nearly the whole story if you stop here — and stopping here is exactly what a raw ranking invites you to do. BJP holds roughly 240 of 543 Lok Sabha seats and a large Rajya Sabha bloc; asking twice as many questions as the next party is not a finding, it is arithmetic.

A better question: not how much, but how disproportionate

Divide each party's share of questions on a given ministry by that party's overall share of all questions, and you get lift: a value of 1.0 means the party asks about that ministry exactly as often as its overall size would predict; a value of 3.0 means three times more than predicted. Lift strips out size and isolates focus. To keep the numbers meaningful, the cut below is restricted to parties with at least 100 total questions and at least 10 questions on the specific ministry — a party with two total questions can post a mathematically enormous but meaningless lift score on whichever ministry those two happen to fall under.

Filtered to parties with ≥100 total questions and ≥10 questions on the ministry shown. Ministry names as recorded in each chamber's own data; near-duplicate labels across chambers (e.g. "Road Transport & Highways" vs "Road Transport and Highways") were reconciled to one canonical name before this table was built.

None of this needs a political-science theory to explain. A party built around India's Scheduled Tribes asks disproportionately about Tribal Affairs. A Muslim League asks disproportionately about Minority Affairs. A coastal Andhra Pradesh party with a long shipping and ports policy history asks disproportionately about Ports, Shipping and Waterways. The pattern is legible because it is exactly what each party's own base would predict — which is precisely what makes it a real signal and not noise.

The confirming contrast: BJP and INC specialise in nothing

That is the actual finding, and it is structural rather than partisan: a party large enough to need MPs covering every portfolio ends up asking about everything roughly in proportion to its size, because covering the floor is the job. A smaller or regional party, with fewer MPs and a narrower base, concentrates its limited number of questions on the two or three ministries that matter most to the people it represents. Size doesn't just win the raw count — it flattens the profile. Specialisation is what's left over for everyone who isn't trying to cover everything.

Grouping by what parties emphasise, not by size

Clustering all 51 parties by their relative ministry emphasis (not raw volume, which is dominated by every party's shared interest in Health, Railways and Education) sorts them into four groups: a 24-party mainstream bloc spanning both the ruling party and the principal national opposition, whose ministry profiles are broad and unremarkable by construction; a 13-party bloc with a shared lean toward minority-affairs- and left-coded ministries; an 8-party bloc of smaller regional and first-time parties with less-defined profiles; and BRS standing alone, its lone standout being a 200× lift on Prime Minister's Office questions — a single-question artefact of the size filter above, included here only to show why that filter exists.

Explore the full data

The tables above show the highlights; every party and every qualifying party–ministry pair is here, sortable. Click a column heading to sort; click again to reverse.

Party membership, for context

Question volume tracks party size closely because it's mostly a function of headcount — a party with 356 sitting MPs across both Houses simply has more people entitled to file questions than one with three. Here is that headcount itself: every sitting member of the 18th Lok Sabha and the current Rajya Sabha, mapped to party.

What this dataset does and doesn't cover

• One term, not history. This covers the 18th Lok Sabha (since June 2024) and Rajya Sabha sessions 265–271 over the same calendar window, through August 2026. It says nothing about whether these specialisation patterns hold in earlier Lok Sabhas.
• 138 of 34,720 Lok Sabha questions (0.4%) are excluded from every party breakdown above — they couldn't be matched to a roster member by name (two MPs' name-spellings didn't reconcile against the roster used).
• A jointly-tabled question is credited to its first-listed MP's party only, to avoid one question inflating several parties' counts at once. Real cross-party co-tabling happens more than this framing shows: the underlying data logs 7,428 distinct pairs of MPs who have jointly asked a question together across party lines.
• Ministry, not full question text, is the primary unit here. The Lok Sabha's own listing API exposes each question's subject line and ministry but not the full question text (that sits in a per-question PDF); the Rajya Sabha's does give full question text, but not answer text either way — both chambers publish the actual answer only as a PDF.
• Lift on a small base is noise, not signal — which is why every party/ministry pair shown above passes a ≥100-question, ≥10-question floor, and why party-level counts for very small parties (a handful of nominated MPs, independents) are shown for completeness but not treated as specialisation claims.

What a Parliament Question actually is

Per Rules 32–54 of the Lok Sabha's Rules of Procedure (Rules 47–50 in the Rajya Sabha), an MP files a notice — capped at 150 words, barring policy-level or sub-judice matters, at most five notices a day — at least 15 days ahead of the sitting (10 for a short-notice question). A starred question gets an oral answer and floor follow-up, capped at one per MP per sitting day; an unstarred question gets only a written reply on the record. Across this dataset, unstarred questions dominate heavily — 31,944 of 34,720 in the Lok Sabha (92%), 22,552 of 24,661 in the Rajya Sabha (91%) — which matches the procedural cap on oral questions, not a lack of MP interest.

Sources. Lok Sabha question metadata (subject, ministry, asking member(s), type, date, session — 34,720 questions, 22 July 2024 to 12 August 2026) from `sansad.in`'s undocumented `api_ls` JSON API. Rajya Sabha question data, including full question text (24,661 questions, 22 July 2024 to 13 August 2026, sessions 265–271) from `rsdoc.nic.in`'s `Question/Search_Questions` endpoint. Both chambers' current and former member rosters (party, state, constituency), used to match each question to its asker's party, from the same two undocumented APIs' member endpoints. Party and ministry naming was reconciled across the two chambers' independently-maintained conventions (e.g. Lok Sabha's "SHSUBT" and Rajya Sabha's "SS-UBT" both resolve to "SHS(UBT)") before any cross-chamber total was computed. Procedural definitions (starred vs. unstarred, notice rules) cross-checked against Drishti IAS's "Questioning in Parliament," a standard UPSC-preparation civics reference, used here only to confirm procedure against a reliable secondary source, not as a data source. Lift and party-clustering figures computed from this data using a graph-based method: full methodology, code and the underlying DuckDB database are in this account's own lok-rajya-sabha-qa research repository. An interactive version of this dataset — sortable party and ministry tables, the same signature-ministry lift graph, and the party-community clustering — is available here. None of this is an endorsement or criticism of any party; the finding is about how legislative attention structurally concentrates as a function of a party's size, not about any party's merit.

About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.