Inflation and the Business Cycle: CPI versus WPI, and What Actually Causes Prices to Rise

This chapter has produced 5 real Prelims questions so far: on the precise scope difference between CPI and WPI (2020), on what drives demand-pull inflation (2021), on the correct fiscal response to a recession (2021), on which method of financing a deficit is most inflationary (2021), and on which institution is mandated to control inflation in India (2022). That last question, and the institutional machinery behind it, is deliberately not this note's job: this site's separate note on the RBI and monetary policy already covers the 4% CPI inflation target, the 2-6% tolerance band, the MPC's exact composition, and the accountability mechanism triggered when inflation strays outside the band for three consecutive quarters, in real depth. Go there for the institutional response. This note covers everything upstream of that: how inflation is actually measured in India, the two classic theories of what causes it, the broader business-cycle concept inflation sits inside, and how fiscal policy and deficit financing interact with both.

CPI versus WPI: two indices measuring different things

UPSC's single favourite trap in this chapter is treating CPI and WPI as two measurements of "the same inflation" that simply disagree. They are not. They price different baskets at different points in the supply chain, for different purposes, compiled by two entirely different government bodies.

The Consumer Price Index (CPI) measures the change in retail prices paid by a fixed representative consumer basket, the actual prices households pay at the point of final purchase. Its basket spans far more than goods: NSO's own CPI release groups spending under headings like food and beverages, clothing, housing, health, transport, communication, education, and restaurant and accommodation services, a mix of goods and services that mirrors how a household actually spends. CPI is compiled and released by the National Statistical Office (NSO), under the Ministry of Statistics and Programme Implementation (MoSPI), drawing on retail prices collected by NSO's own field staff across a large sample of rural and urban markets every week.

The Wholesale Price Index (WPI), by contrast, measures the change in prices at an early stage of transaction, the wholesale or producer level, before goods reach a retail shelf. Its basket of 697 items is built entirely from three groups: Primary Articles, Fuel and Power, and Manufactured Products. There is no services group; WPI excludes services entirely, a scope restriction that is exact, not approximate. WPI is compiled and released by the Office of the Economic Adviser (OEA), inside the Department for Promotion of Industry and Internal Trade (DPIIT), Ministry of Commerce and Industry, on the 14th of every month.

That NSO/MoSPI-versus-OEA/DPIIT split is exactly the kind of institutional attribution fact secondary sources garble, and exactly what UPSC likes to test: a statement that swaps which body compiles which index is a classic wrong-but-plausible option. Keep the anchor simple: CPI is a Ministry of Statistics job (NSO/MoSPI); WPI is a Commerce Ministry job (OEA/DPIIT), which is itself a small tell for why WPI was historically built around goods moving through industry and trade, not household consumption.

Why the two diverge in the same period

Because CPI and WPI price different baskets at different points in the supply chain, they routinely show different inflation rates in the very same month, and this is normal, not an error in either series. Three mechanical reasons drive the gap: food carries a materially higher weight in the CPI basket than in the WPI basket; WPI captures prices before taxes, retail margins, and transport costs are added, while CPI captures the price after all of that is baked in; and WPI has no services component at all, so a services-driven episode, a rise in transport fares or education fees, shows up in CPI and is structurally invisible to WPI. A real 2020 Prelims question tested exactly this cluster together, including the (false) claim that the RBI targets WPI rather than CPI for monetary policy, a trap that folds this chapter's measurement content directly into the institutional content covered in the RBI note.

The other CPI sub-indices, and who compiles each

"CPI" is not one index; India publishes several, for different audiences and different purposes, and they are not all compiled by the same body. This is the second institutional-attribution trap in the chapter.

• CPI-Combined: the headline all-India number, a weighted combination of rural and urban prices, the figure the RBI's inflation target actually tracks.
• CPI-Rural and CPI-Urban: the same underlying methodology, split by area, useful for seeing whether inflation is hitting rural or urban households harder in a given month.

All three of the above are compiled by NSO, MoSPI, exactly like the headline CPI described above; they are simply three cuts of the same survey.

• CPI-IW (Industrial Workers): a separate index, weighted toward what industrial working-class households consume, compiled monthly by the Labour Bureau, under the Ministry of Labour and Employment, not NSO/MoSPI. Its primary use is wage indexation: the Dearness Allowance (DA) paid to central and state government employees, and to workers in scheduled industrial employment, is revised twice a year using a 12-month average of the all-India CPI-IW.
• CPI-AL (Agricultural Labourers) and CPI-RL (Rural Labourers): also compiled by the Labour Bureau, using retail prices collected separately across 20 selected states and weighted by each state's share of agricultural or rural labour households. CPI-AL feeds directly into statutory minimum wage revisions for agricultural labour, exactly why it exists as its own series rather than being folded into the general CPI.

So the split to hold in your head: CPI-Combined/Rural/Urban is NSO, MoSPI; CPI-IW and CPI-AL/RL are both Labour Bureau, Ministry of Labour and Employment. A question that assigns CPI-IW to NSO, or the headline CPI to the Labour Bureau, is testing exactly this line.

A recent, genuinely exam-relevant development worth flagging: the CPI series most students learn on base year 2012=100 has itself just been revised. NSO's own February 2026 press release reports the headline CPI on a new base year of 2024=100, alongside the GDP base-year revision to 2022-23 covered in this site's National Income & Basic Economic Concepts note. Both revisions land in the same short window; know that the base year has moved, even if most current textbook material still describes the older series.

Demand-pull versus cost-push inflation: the two classic theories

Once the measurement is settled, the theory question UPSC asks most often is what actually causes the price level to rise, and the standard framework splits inflation into two causal mechanisms.

Demand-pull inflation happens when aggregate demand in the economy outpaces aggregate supply: "too much money chasing too few goods." Anything that pushes total spending power upward without a matching rise in output can trigger it: expansionary monetary policy, a fiscal stimulus that raises government expenditure, or a general rise in consumers' purchasing power. A concrete Indian example: a large rural income-transfer or wage scheme rolled out quickly, putting more money in households' hands faster than farms and factories can expand output to match it, pulls prices up from the demand side.

Cost-push inflation happens on the supply side instead: the cost of producing goods rises, and firms pass that higher cost through to prices, even if demand has not changed at all. A concrete Indian example: a sharp rise in global crude oil prices raises input costs across transport, fertiliser, and manufacturing simultaneously, since diesel and petroleum derivatives sit inside so many Indian supply chains, and firms facing higher costs raise output prices to protect margins, pushing the price level up even though nobody is spending any more than before.

The trap UPSC has actually tested: wage indexation (automatic wage adjustments that track inflation) is not itself a demand-pull trigger. It is better understood as a mechanism that propagates an inflation episode once it has already started, part of a wage-price spiral, rather than a root cause of aggregate demand rising in the first place. A 2021 Prelims question built its wrong option around exactly this confusion.

The GDP deflator: the third measure, briefly

CPI and WPI are not the only price index India uses. The GDP deflator (Nominal GDP divided by Real GDP, multiplied by 100) is the broadest price measure of the three, since it implicitly covers every final good and service produced domestically in a year, not a fixed household basket like CPI or a goods-only wholesale basket like WPI. This site's National Income & Basic Economic Concepts note derives the deflator in full, including MoSPI's double-deflation methodology for manufacturing and agriculture; the exam-relevant point to carry into this chapter is simply that CPI, WPI, and the GDP deflator can each report a different inflation rate for the identical quarter, and none of the three is "wrong": they are pricing different, only partially overlapping baskets.

Core inflation versus headline inflation

Headline inflation is the full CPI number, including every item in the basket. Core inflation strips out the most volatile components, food and fuel, and reports the rate for everything else. Policymakers watch it separately because food and fuel prices swing sharply for reasons unrelated to the underlying state of the economy, a poor monsoon, a global oil-price shock, a temporary vegetable price spike, and those swings often reverse within a month or two on their own. A central bank that reacted to every headline spike driven by a transient food-price shock would be tightening credit across the whole economy to fight a price rise that was never a symptom of excess demand. Core inflation is the better signal of underlying, demand-driven price pressure precisely because it removes the noise headline inflation cannot.

The business cycle: four classical phases

Inflation does not move in isolation; it sits inside the broader rhythm of the business cycle, the recurring, roughly wave-like pattern of expansion and contraction that market economies exhibit over time. The classical framework identifies four phases:

• Expansion (boom): output, employment, and incomes all rise; business and consumer confidence is high. As the expansion runs on, demand begins to outpace productive capacity, and inflationary pressure typically builds toward the end of this phase.
• Peak: the high point, where growth stops accelerating. The economy is often at or near full capacity here, which is why inflation tends to be highest around the peak: there is little spare capacity left to absorb further demand without prices rising.
• Contraction (recession): output, employment, and incomes fall; investment slows or reverses; unemployment rises. Inflationary pressure typically eases, since weaker demand takes the pressure off prices.
• Trough: the low point, where the contraction bottoms out before the next expansion begins.

The exam-relevant instinct: expansion and peak pair with rising output/employment and typically rising prices; contraction and trough pair with falling output/employment and typically easing prices. Stagflation, covered below, is exactly the case where this normal pairing breaks down.

Fiscal policy's countercyclical role

Because the business cycle is not something government policy can switch off, fiscal policy is conventionally used to lean against it, smoothing the extremes rather than eliminating the cycle itself. This is the countercyclical principle, and it is precisely what a real 2021 Prelims question tested: during a recession, the textbook response is expansionary fiscal policy, raising public expenditure (or cutting taxes) to boost aggregate demand, even though this widens the fiscal deficit in the short run. During a boom, when demand risks outrunning capacity and stoking inflation, the textbook response flips to contractionary fiscal policy, reducing spending or raising taxes to cool demand down. Cutting spending during a recession, or raising interest rates as a fiscal tool, are wrong-answer patterns built around confusing the direction of the required response, or confusing fiscal policy with monetary policy.

Automatic stabilisers are the part of this mechanism that needs no fresh legislation at all: certain elements of the tax-and-transfer system counter the cycle automatically, simply because of how they are already structured. Progressive income tax collects less revenue automatically when incomes fall in a downturn, cushioning household spending power without any new policy decision; unemployment benefits pay out more automatically as joblessness rises, putting money back into the hands of exactly the households whose spending has just been hit. Both push in the countercyclical direction without anyone legislating a fresh stimulus package in real time, which is what makes them "automatic": the stabilising effect is built into the structure of the system itself, not into a discretionary decision made during the downturn.

Stagflation: the combination classical theory could not explain

Stagflation is the simultaneous occurrence of high inflation with high unemployment and stagnant (or negative) growth, a combination classical Keynesian demand-management theory struggled to explain, since that framework generally expected inflation and unemployment to move in opposite directions: strong demand brings rising prices and falling unemployment together, weak demand brings falling prices and rising unemployment together. Stagflation defies that pairing, and its defining historical association is the 1970s oil shocks: the OPEC oil embargo of 1973, and the second shock following the 1979 Iranian Revolution, both delivered a severe cost-push supply shock to Western economies while growth was already weak, driving inflation and unemployment up together and forcing a rethink of the assumed inflation-unemployment trade-off (the Phillips curve) that had shaped policy until then. The exam-relevant takeaway is the mechanism: stagflation happens when a supply-side shock, not a demand-side one, drives inflation, because a supply shock can push prices up and push output and employment down at the same time, exactly the combination a purely demand-side model cannot produce.

Deficit financing and its inflationary channel

A government running a fiscal deficit has several ways to finance the gap between spending and revenue: borrowing from the public through government-bond sales, borrowing from commercial banks, or financing it through the creation of new money, effectively having the central bank monetise the deficit. Of these, creating new money is the most directly inflationary route, because it expands the money supply without any matching increase in the output of goods and services, the textbook demand-pull mechanism operating at the level of government finance itself. Borrowing from the public or from banks largely recycles money already in circulation rather than creating new purchasing power out of nothing, which is exactly why a real 2021 Prelims question identified deficit monetisation, not bond or bank borrowing, as the most inflationary financing method.

India's own history with this channel is precise and worth knowing exactly. Until the mid-1990s, the Government of India routinely financed shortfalls by issuing ad hoc Treasury Bills to the RBI, an arrangement in place since the mid-1950s that converted government deficits into fresh central-bank money in a way that was close to automatic. In March 1997, the Government of India and the RBI signed a formal agreement to end this: ad hoc Treasury Bills were discontinued with effect from 1 April 1997, replaced by a Ways and Means Advances (WMA) scheme, under which the RBI gives the Central Government short-term accommodation for temporary cash mismatches, repayable within three months, capped by mutually agreed limits rather than issued without limit against the deficit itself. The Ministry of Finance's own contemporaneous account is explicit that this was a deliberate structural reform: the shift to WMA "means elimination of automatic monetisation of fiscal deficit," since, unlike the old ad hoc bills, a WMA drawdown is not itself a source of financing the deficit, only a short-term liquidity bridge that must be repaid. India built an institutional mechanism specifically to stop deficit monetisation from operating automatically, rather than merely warning against it in theory.

Quick revision points

• CPI: retail consumer basket, goods and services, compiled by NSO, MoSPI. WPI: wholesale/producer-level prices, goods only, no services, compiled by the Office of the Economic Adviser, DPIIT.
• CPI-Combined/Rural/Urban: NSO, MoSPI. CPI-IW and CPI-AL/RL: both compiled separately by the Labour Bureau, Ministry of Labour and Employment, and used for DA revision and minimum-wage revision respectively, not for the RBI's inflation target.
• CPI and WPI diverge because of different basket weights, different points in the supply chain, and WPI's total exclusion of services.
• Demand-pull: too much money chasing too few goods (monetary/fiscal stimulus, rising purchasing power). Cost-push: rising input costs passed through to prices (an oil-price shock is the classic trigger). Wage indexation propagates inflation, it is not itself a demand-pull cause.
• GDP deflator (Nominal GDP / Real GDP x 100) is the broadest price measure of the three; see the National Income note for the full derivation.
• Core inflation strips out volatile food and fuel; policymakers watch it as the better signal of underlying, demand-driven price pressure.
• Business cycle: expansion, peak, contraction (recession), trough; output/employment/prices generally rise through expansion and peak, and fall through contraction and trough.
• Fiscal policy is countercyclical: expansionary during a contraction, contractionary during a boom. Automatic stabilisers (progressive taxes, unemployment benefits) counter the cycle without new legislation.
• Stagflation: high inflation with high unemployment/stagnant growth, driven by a supply-side shock; historically tied to the 1973 and 1979 oil shocks.
• Creating new money to finance a deficit is the most inflationary financing method. India ended automatic monetisation via ad hoc Treasury Bills from 1 April 1997, replacing them with the Ways and Means Advances scheme.

Every fact in this chapter pairs a mechanism with an exact institutional or historical detail, which body compiles which index, which year a practice ended, which shock defines stagflation, and UPSC's questions are almost always built by testing whether you know the detail precisely, not just the general idea.