Cost of living: Should India's tracker of retail inflation account for free food handouts under PMGKAY?

The question gets its urgency from the Pradhan Mantri Garib Kalyan Anna Yojana (PMGKAY). Under this integrated food security scheme, the government provides free foodgrains to about 75% of India’s rural population and 50% of urban households. When a programme of this magnitude touches so many lives, how we account for it in our economic statistics matters profoundly.

The conventional economic logic seems straightforward. The CPI measures what consumers pay for goods and services. If something is free, or if there’s no monetary transaction, then expenditure by the consumer is zero and its weight in the index, typically captured by its share in total consumption expenditure, should be zero.

This isn’t the same as subsidized electricity or regulated fuel prices, where consumers still open their wallets. Free grains involve no payment whatsoever.

But here’s where it gets complicated. Consider, for illustration, a household that consumes 100kg of rice annually at ₹1 per kg. This implies a bill of ₹100. Now suppose the government provides 25kg free. The household then pays ₹75 for only 75kg.

This situation can be interpreted two ways. First, we could say the average price remains ₹1 per kg, calculated by dividing the money spent ( ₹75) on the quantity bought (75kg). Alternatively, we could say the household still consumes 100kg, but the average price has fallen to ₹0.75 per kg, calculated as household spending ( ₹75) as a ratio of the total quantity consumed (100kg). As long as the magnitude and proportion of free grains stays constant, these approaches are equivalent, it turns out.

A divergence becomes evident if the government expands the scheme. If the quantity of free rice increases from 25kg to 30kg, the first interpretation shows no price change. The average price remains ₹1 per kg. The share of the consumer’s outgo on rice simply decreases. The second interpretation, however, shows deflation: the average price drops from ₹0.75 per kg ( ₹75 for 100kg) to ₹0.70 ( ₹70 for 100kg).

Which approach is correct? The answer depends on what we’re trying to measure. If our goal is tracking the growth rate of prices, the first method may be appropriate. But if we want to understand the actual cost of living, particularly for rural beneficiaries of the PMGKAY, perhaps the second approach better captures reality.

This leads to deeper questions. If increased transfers of free rice reduce the cost of living, why would cash transfers not do the same? And why stop at rice? What about free medicines, healthcare and education? One cannot argue that we should include only free foodgrains because surveys provide imputed expenditures for it. That’s a technical convenience, not a conceptual foundation.

Some suggest treating free grains like production for own consumption, which the CPI already includes. But this comparison fails on closer inspection. Homegrown rice, even if fully consumed, has an ‘opportunity cost,’ as farmers could sell it in the market. Similarly, we use imputed rent for owner-occupied housing because owners forgo rental income. Free government grains involve no such opportunity cost ‘for the consumer.’

India isn’t alone in grappling with these issues. The measurement of the digital economy faces similar issues globally. Free digital services like Facebook, Google and weather apps create consumer value but have no charge and therefore zero weight in GDP and CPI calculations.

Researchers like Erik Brynjolfsson at MIT and his co-authors have introduced novel metrics like GDP-B to capture benefits from free goods, using experimental methods to determine what consumers would pay to retain access. They suggest that Facebook alone might add 0.5 percentage points to annual US real GDP growth, with consumers valuing it at more than $50 per month.

Notably, no major economy includes free healthcare, for example, in its CPI. This perhaps reflects a choice to maintain conceptual clarity in inflation measurement rather than mixing it with broader welfare considerations.

The decision India makes now will set a precedent. We face a choice between two distinct objectives. One path preserves the CPI as a clean measure of market price inflation, technically rigorous, internationally comparable and useful for policy. The alternative path transforms it into a comprehensive cost-of-living index that incorporates government transfers and free services.

The better approach might be to keep these measurements separate. Maintain the CPI as a focused inflation gauge, but develop parallel indicators for broader living standards. For indexation purposes or poverty measurement, we could create a new cost-of-living metric that includes all free goods and services while subtracting cash transfers to measure how much money households need to achieve a given consumption basket.

Further, if we truly want to measure standards of living comprehensively, we could think beyond consumption, incorporating health outcomes, educational attainment and environmental quality. These are all active areas of research, from willingness-to-accept surveys for free goods to multidimensional poverty indices. The new initiative of conducting an all-India income distribution survey is a first step in that direction.

The temptation to meet several goals with a single statistic is understandable. But mixing price inflation measurement with welfare policy evaluation risks compromising both objectives.

As we try to build world-class statistical systems, statistical agencies may need to strike a fine balance between clarity and convenience. Mospi’s discussion paper offers an opportunity for India to lead in thinking through such methodological challenges. Let’s articulate clearly what each measure aims to capture and create appropriate tools for each.

Laurence Ball of Johns Hopkins University and Gourab Saha of Ashoka University contributed to this piece.

These are the author’s personal views.

The author is professor of economics at Ashoka University and director and head of Ashoka Isaac Centre for Public Policy.