Imported Fertiliser & Fuel Prices → Diesel → CPI/WPI — India

MoSPI Dataset Analysis — Statistical Bulletin

Diesel passes straight through to WPI. Fertiliser mostly doesn't — because subsidy absorbs it.

Correlates the WPI sub-indices for diesel (HSD) and fertilisers (urea, DAP) — the domestic price points where this repo's own import-value growth for mineral fuels (HS27) and fertilisers (HS31) actually lands — against WPI and CPI headline and food inflation, 2012–2026. Cross-checked against five published studies on import-price pass-through in India.

SOURCE: MoSPI Wholesale Price Index (base 2011-12) and Consumer Price Index (base 2012), via the MoSPI MCP connector — item-level series for HSD (diesel), Urea, and Di-ammonium Phosphate (DAP); retrieved 18 Jul 2026

The input-cost side: diesel, urea, DAP, and the WPI they feed

April-anchor annual snapshot, WPI 2011-12 = 100, FY2012 through FY2026.

HSD (diesel) Urea DAP WPI headline

Since April 2012: diesel (HSD) is up 79.0%, moving almost exactly with global crude and tracking WPI headline itself (same-year growth correlation r = 0.92) — diesel isn't just correlated with WPI, it's one of the components that defines WPI's fuel weight, so a nearly 1:1 relationship is closer to accounting than causation. Urea is up only 13.8% over the same 14 years, despite global fertiliser-price shocks in 2008, 2021–22 and 2023 — the subsidy regime holds the retail/wholesale price down almost regardless of what happens at the border. DAP, with a smaller and less consistent subsidy share, is up 38.4% — roughly midway between urea's near-total insulation and diesel's full exposure.

Where it's supposed to land: WPI and CPI, headline and food

Same April-anchor method, CPI 2012 = 100, 2013 through 2025.

CPI headline CPI food & beverages CPI oils & fats

CPI oils & fats is the one series here that moves in genuinely dramatic swings — not shown by the April-anchor line alone: the actual monthly year-on-year print went from -18.17% in June 2023 to +21.24% in August 2025, the sharpest reversal of any major CPI sub-group in this window. That matches Mishra and Roy's (2011-12) finding that edible-oil inflation is the one food category with near one-to-one pass-through from global prices — unlike cereals or sugar, which India insulates with trade and support-price policy — and lines up with this repo's own trade-balance bulletin showing edible-oil imports (HS15) up 97.9% over FY2018-19 to FY2025-26.

So how strong is the correlation, actually?

Pearson r on year-on-year growth rates, same-year unless marked.

The one strong number here — diesel vs. WPI headline, r = 0.92 — is closer to a mechanical fact than a discovery, since HSD is itself inside the WPI fuel basket. Every other pair — fertiliser or fuel input costs against CPI headline, CPI food, or CPI oils & fats — comes out weak or inconsistent (r between -0.52 and +0.29) at this annual, year-on-year-growth resolution. That is not evidence the pass-through described in the literature doesn't exist — it's evidence that annual data is the wrong resolution to see it in.

Why the annual correlation looks weak when the literature says pass-through is real
Because the transmission happens in months, not years — and other shocks dominate the annual average.
Bhattacharya and Gupta's SVAR of Indian food inflation finds a 10% rise in fuel inflation raises food inflation by about 1% within the same month, and a 10% rise in agricultural-wage inflation raises it by 1.6% immediately, rising to 2.4–5.5% within four months — real, fast, but also transitory: the effect fades within the year it hits. Collapsing twelve months of that dynamic into a single year-on-year growth figure, then correlating it against another year-on-year figure, averages the fast pass-through together with whatever else moved food prices that year — a monsoon shock, an MSP hike, a demand-side shift — which is exactly why James Hamilton's framing paper on this subject warns that "pass-through" is not a single number: it depends on what caused the price to move in the first place, and a raw historical correlation answers a different question than "what happens if diesel goes up."

Reading this against five studies

Hamilton (International Journal of Central Banking, 2012) supplies the framing caveat this whole bulletin leans on: import-price "pass-through" isn't a fixed constant, because a price rise caused by a supply shock behaves differently from the same price rise caused by currency depreciation or a demand boom. A raw correlation conflates these.

Bhattacharya and Gupta (NIPFP Working Paper 2015-151) is the direct quantitative anchor: their Structural VAR puts numbers on fuel-to-food and wage-to-food transmission (above), and separately finds edible-oil and sugar inflation carry significant global pass-through, while cereals largely don't — protected by India's own trade and procurement policy.

Mishra and Roy (India Policy Forum, 2011-12) reach the same edible-oil conclusion independently, adding that the degree of pass-through tracks how tradable a commodity actually is — oils, which India imports heavily, move almost one-for-one with world prices; rice and wheat, protected by export bans and procurement, barely move with them at all.

Elahi and Nath (JETIR, 2025) give the plain-language version of the mechanism this bulletin quantifies: "a classic example of imported inflation in India is the impact of global crude oil prices on domestic fuel costs" — and note the RBI's own finding that inflation above 6% starts dragging on GDP growth, the threshold India's CPI food print crossed repeatedly during the 2022 and 2024 spikes shown above.

The Yogeshwaran thesis (Indian Maritime University, 2025) on India's import composition puts the two inputs analysed here in proportion: crude petroleum alone is 23.4% of India's entire import bill by value — the single largest line item — while the various fertiliser categories (mixed mineral/chemical, nitrogenous) add roughly another 2%. Diesel's outsized pass-through to WPI isn't a coincidence of one sub-index; it's the largest import category in the whole basket.

Finally, a prior analysis in this workspace on India's Minimum Support Price vs. Cost-of-Production data (2020–26, sourced from PIB/CACP releases) found the same urea-vs-DAP divergence quantified here from the supply side: fertiliser subsidy has kept urea's farmer-facing cost nearly flat even through the 2021–23 global fertiliser-price shock, while CACP's own cost-of-production data is a lagged projection that understated the real in-year cost spike — meaning the squeeze from that shock likely hit farmers before it ever showed up in an official series, on either the input-cost or the MSP side.

View annual index values, WPI inputs and CPI/WPI outputs
Errata & methodology caveats
  • WPI series use the April value of each year as an annual anchor (FY2012 = April 2012, etc.) rather than a full monthly series or FY average — chosen to keep this bulletin's methodology consistent with the annual/FY convention used elsewhere in this repo (GDP, trade-balance, HSN-trends charts). A monthly-frequency version would very likely show materially different, probably stronger, short-lag correlations — see the verdict box above for why.
  • CPI (base 2012) only covers 2011–2025; WPI (base 2011-12) extends to April 2026. The two series are therefore compared over their 2013–2025 overlap, not the full WPI window.
  • HSD is one of the ~10 items inside the WPI "Mineral oils" group, which is itself one of three groups making up the "Fuel & Power" major group (weight 13.15% of WPI). Its high correlation with WPI headline growth (r=0.92) partly reflects that HSD is mechanically inside the index being correlated against, not a fully independent external check.
  • Urea and DAP are wholesale (ex-factory/import-parity) prices, not the subsidized retail price actually paid by farmers — India's fertiliser subsidy operates as a per-unit payment to manufacturers/importers, so the WPI series shown here already reflects most of that subsidy's dampening effect on the price that reaches the market, not a pre-subsidy world price.
  • Correlation is not causation, and small-n annual correlations (12–14 points) are noisy — treat every r value on this page as indicative, not a rigorous econometric estimate. The published SVAR/VAR studies cited above, run on decades of monthly data with proper lag structure, are the actual rigorous evidence; this bulletin's own correlation table is a sanity check against that literature, not a replacement for it.
  • Underlying data: data/fertiliser_fuel_price_transmission_2012_to_2026.json.
herrrickshaw/mospi-dataset-analysis — derived from MoSPI WPI/CPI series, not an official MoSPI/RBI publication

Comments

Popular posts from this blog

📚 Article Index — Start Here

Chemical Import Substitution — Full HSN-8 Chapter-Wise Registry (827 Codes)

CPI Inflation Heatmap — India, June 2026