The clearance pipeline: PARIVESH EC data as the IEM A→B leading indicator
Between filing intent (IEM Part A) and commencing production (Part B) sits the gate this repo could not previously see: land acquired, Environmental Clearance obtained. This bulletin fills it — gate 3 of the lifecycle schema — with the complete national EC register: which states' implementation surge has pipeline behind it, which sectors are loading next, and which flagship projects have (and haven't) entered the clearance system at all.
SOURCE: PARIVESH 2.0 open JSON API, full extraction 19-Jul-2026 — 113,804 EC proposals, submissions 2022→present, SEIAA state-level clearances included (104,583 SEIAA + 9,221 MoEFCC). Industrial = IND-1/2/3 sectors; mining and delisted records excluded. Aggregate CSV: data/parivesh_ec_state_sector_status_2026-07-19.csv.
The methodological find: PARIVESH 2.0 exposes a fully open JSON API — no key, no pagination, no CAPTCHA — returning the complete national EC register in one call: parivesh.nic.in/parivesh_api/trackYourProposal/advanceSearchData?majorClearanceType=1&…. One GET replaces what took nine research agents elsewhere in this repo. And the feared coverage gap is dead: state-level (SEIAA, Category B) clearances ARE in the central portal, since PARIVESH became the mandatory single window in Feb 2024. Now on the quarterly checklist.
State-wise industrial EC pipeline
Green = EC granted (ready to build) · amber = ToR granted (EIA underway, ~12-18 months before EC)
State
EC granted
ToR granted
Granted | ToR
In process
Dead
IEM impl. rank
Read
Grant flow is accelerating: by last-update year, industrial grants ran 198 (2023) → 734 (2024) → 1,102 (2025) → 579 (2026 YTD) — consistent with the IEM implementation ratio breaking upward (82% → 109.9%). Gujarat holds ~31% of all industrial EC grants and the deepest ToR pipeline; Maharashtra is the most forward-loaded state (ToR 519 > granted 390).
Does the register predict implementation? The correlation, quantified
16 states · ECs granted (2022+) vs IEM Part B (2024–25) · Pearson / Spearman / log-log
Pair
Pearson
Spearman (rank)
Read
EC granted vs Part B implemented value (₹cr, 24+25)
0.961
0.794
The headline: clearance stock explains ~92% of the variance in recent implementation value — and the rank correlation confirms it is not just the Gujarat outlier.
log(EC) vs log(Part B value)
0.864
0.794
Elasticity ~1.7 — implementation value scales super-linearly with clearances: more-cleared states also host bigger projects.
EC granted vs Part B filing count (24+25)
0.857
0.482
Weaker in rank terms — many Part B filings need no fresh EC (expansions, estate piggybacking).
ToR granted (forward pipeline) vs Part B value
0.849
0.676
The EIA-stage pipeline already tracks current implementation.
EC granted vs Part B value growth (24+25 ÷ 21+22)
0.612
0.385
EC stock predicts implementation levels far better than growth rates — ratios are small-base noise.
The residuals are the actionable layer — from the log-log fit (log₁₀ Part B value = 0.67 + 1.71·log₁₀ EC granted), who converts clearances into implemented value above and below trend:
Over-performers — Andhra Pradesh (+0.49: few clearances, giant projects), Madhya Pradesh (+0.47), Odisha (+0.40: steel value-per-clearance), UP (+0.25), Rajasthan (+0.22). Under-performers — Jharkhand (−0.44), Assam (−0.41: post-Tata-surge lull), Karnataka (−0.35: many clearances, modest value — its pipeline-building read quantified; watch this gap close if the Foxconn-era projects land), West Bengal (−0.32), and even Gujarat (−0.19: its ₹4.55 lakh cr sits slightly below what 1,049 clearances predict — the EC book skews to mid-size chemical units).
Caveats: n=16, cross-sectional, not causal; the EC window (2022+) only partially leads the Part B window (2024-25); counts vs rupee values are different units — hence the rank and log checks. The supported claim: the register is a valid leading indicator of implementation levels, with the residuals flagging which states' pipelines are converting.
What's loading next — granted ECs by EIA activity
The clearance register corroborates the implementation sector leaders
Flagship projects in the clearance register
Direct pipeline evidence, with proposal numbers — including two negative findings
Project
Proposal no.
EC status
Read
The two absences matter as much as the presences
POSCO-JSW's $7.73bn steel JV has no EC application in the portal at all — the Apr-2026 cash movement this repo flagged will show up in FY27 FDI data before any clearance-visible construction exists. And RRPCL (~$44bn) is entirely absent from the 2022+ register — the "stalled at MoU" verdict now has clearance-register proof: a project of that scale with zero EC activity is not under implementation.
Using EC as the pipeline indicator — four stated filters
EC categories ≠ IEM scheduled industries. Clean overlap only for metallurgy, cement, chemicals, pharma. Electronics, semiconductor and battery plants have no EC category of their own — they appear under INFRA-2 building-construction (if >20,000 m² built-up) or not at all; category aggregates miss them, so only named-project search covers them.
No grant date in the bulk record — submission date plus a last-update proxy only; exact grant dates need per-proposal calls.
One EC ≠ one IEM-scale project — a 30 TPA API unit and an 18 MTPA steel plant each count once. Read counts alongside investment values, never instead.
2022+ submissions only; legacy ECs live on environmentclearance.nic.in (needs pinned DNS), and estate-based projects can piggyback on an industrial estate's EC without appearing individually. DELISTED_BY_SYSTEM records (25,032) are auto-cleanup and excluded from all conversion math.