Econvitals IndiaMoney & banking / MPC Speak
MPC Speak
How hawkish or dovish is each member of the Monetary Policy Committee? Every member writes a statement in the minutes, published about two weeks after each meeting. We count phrases that lean hawkish (worry about inflation, a case for tighter policy) or dovish (worry about growth, a case for easier policy) in each statement. At the Oct 2026 meeting the committee's average score was +0.43, which leans hawkish.
Committee tone at each meeting
Average of the six members' scores; above zero leans hawkish, below zero dovish
Members at the latest meeting (Oct 2026)
| Member | Score | Lean | Phrases counted | Rate vote |
|---|---|---|---|---|
| Indranil Bhattacharyya | +0.43 | hawkish | 4 hawkish, 1 dovish phrases | for |
| Nagesh Kumar | +0.33 | hawkish | 1 hawkish, 0 dovish phrases | for |
| Poonam Gupta | +0.71 | hawkish | 5 hawkish, 0 dovish phrases | for |
| Ram Singh | +0.54 | hawkish | 9 hawkish, 2 dovish phrases | for |
| Sanjay Malhotra | +0.46 | hawkish | 7 hawkish, 2 dovish phrases | for |
| Saugata Bhattacharya | +0.09 | neutral | 5 hawkish, 4 dovish phrases | for |
Recent meetings
| Member | Aug 2025 | Oct 2025 | Dec 2025 | Feb 2026 | Apr 2026 | Jun 2026 | Aug 2026 | Oct 2026 |
|---|---|---|---|---|---|---|---|---|
| Indranil Bhattacharyya | −0.85 | −0.82 | −0.71 | +0.20 | 0.00 | +0.20 | +0.43 | |
| Nagesh Kumar | −0.50 | −0.75 | −0.67 | −0.50 | −0.50 | −0.50 | 0.00 | +0.33 |
| Poonam Gupta | −0.82 | −0.80 | −0.71 | −0.67 | −0.50 | +0.33 | −0.33 | +0.71 |
| Ram Singh | −0.80 | −0.93 | −0.84 | −0.86 | 0.00 | −0.17 | +0.60 | +0.54 |
| Sanjay Malhotra | −0.71 | −0.82 | −0.78 | −0.71 | −0.43 | 0.00 | +0.43 | +0.46 |
| Saugata Bhattacharya | −0.60 | −0.67 | −0.78 | +0.33 | +0.33 | +0.71 | +0.20 | +0.09 |
| Rajiv Ranjan | −0.87 |
† voted against the rate resolution. A blank cell means the person was not a member at that meeting. The vote table in the minutes does not say which way a dissenting member wanted the rate to move; read their statement for that.
How the score works
Score = (hawkish phrases − dovish phrases) / (hawkish phrases + dovish phrases + 2). The +2 keeps a statement with only one or two matching phrases close to zero, so a single phrase scores ±0.33 rather than ±1. The phrase lists are fixed in advance and the same for every member and meeting. No language model is involved, so anyone can reproduce a score from the minutes and the lists below.
What it is not: a model of what members believe, or a forecast. Phrase counts miss negation and context ("we no longer see upside risks to inflation" counts as hawkish), and statements differ in length and style from member to member. The score is most useful for changes over time in the same member and for the committee as a whole. Checks against the committee's own decisions are on the Methods page.
Phrase lists (version 1)
Hawkish
upside risks? to (?:the )?(?:headline |core |food )?inflationdownside risks? to (?:the )?growth (?:have|has) (?:receded|eased|diminished)inflationary pressures?second[- ]round effects?(?:de|un)-?anchor(?:ing|ed)? (?:of )?(?:inflation )?expectationsremain(?:s|ing)? vigilantvigilancewithdrawal of accommodationtighten(?:ing)?rate hikes?(?:increase|raise|raising|hike) (?:in )?the (?:policy )?(?:repo )?ratepremature(?:ly)?durable (?:alignment|disinflation)persistent inflationelevated inflationinflation (?:remains |is )?(?:well )?above (?:the )?targetstickydisinflation (?:is|has been) (?:slow|incomplete)
Dovish
downside risks? to (?:the )?(?:growth|inflation)upside risks? to (?:the )?growth(?:support|supporting|supports) (?:to )?growthgrowth[- ]supportiveaccommodativerate cuts?(?:reduce|reducing|reduction in|cut in|cut) the (?:policy )?(?:repo )?rate(?:monetary |policy )?easingpolicy spacenegative output gapslackbenign(?:softening|moderation) (?:of|in) (?:headline )?inflationinflation (?:remains |is )?(?:well )?below (?:the )?targetundershoot(?:ing)?real (?:policy )?rates? (?:is|are|remains?) (?:too )?high
Regular expressions, matched case-insensitively on whole words. Changes are logged in registry/mpc/lexicon.yaml and re-score the full history.
Source: Reserve Bank of India, minutes of the Monetary Policy Committee meetings since October 2016; site calculations. Downloads: Data (mpc_member_tone, mpc_member_dissent, mpc_committee_tone).