Here is why India's declining consumption inequality deserves recognition

Hans India

21-07-2025

Therecent decline in India's consumption-based Gini coefficient—from 28.8 in 2011-12 to 25.5 in 2022-23, as reported by the World Bank—has prompted considerable scrutiny, particularly when juxtaposed with income-based estimates from the World Inequality Database (WID), which peg India's Gini at an ostensibly alarming 62 in 2023. Reconciling this apparent dissonance necessitates a closer interrogation of the underlying metrics, data sources, and conceptual frameworks. What follows is a reasoned engagement with the criticism—one that distinguishes methodological incompatibilities from substantive economic realities and highlights the perils of conflating fundamentally distinct measures of inequality.
At the core of this divergence is a critical conceptual distinction: the difference between consumption inequality and income inequality. In a country like India—characterised by a large informal workforce, extensive in-kind transfers, and a rapidly expanding welfare architecture—income is often volatile, underreported, or difficult to capture comprehensively. Consumption, by contrast, tends to be smoother over time and more reflective of actual living standards.
The World Bank's Poverty and Inequality Platform (PIP) adopts this logic, using either disposable income or consumption expenditure depending on national context.
Firstly, it is worth pointing out that the World Bank in its paper titled 'The World Bank's New Inequality Indicator' gives a way of converting consumption gini to income gini and vice versa. The bank estimated that the average ratio of income-to-consumption Gini coefficients across 84 country-years where data was available for both is 1.13.
Applying this directly to India's consumption-based Gini of25.5 yields an approximate income Gini of 28.8. This still places India at 12th, even under income-equivalent assumptions. This simple approximation gives a way of comparing welfare types within PIP database.
This raises a pertinent question: why has this not been more widely acknowledged? The answer perhaps lies in the tendency to selectively emphasise outlier estimates.
When the simple approximation given is used for comparison across nations, India's inequality even when measured in income terms is significantly lower than the United States and UK. Among the 48 nations where welfare approach is consumption based; India ranks third.
India's consumption-based Gini coefficient of 25.5 in the PIP database is also internationally striking. China's consumption of Gini, for instance, stands at 35.7, according to the same database and using the same welfare concept. This 10-point difference is significant.
Secondly, why is the impact of large-scale social welfare schemes conspicuously spared from criticism? In a country like India, where large-scale social welfare programmes—such as subsidized food, LPG, housing, rural employment, health insurance, and direct cash transfers—have significantly boosted the living standards of the poor, consumption will inevitably be higher and more equitably distributed than income. These forms of public provisioning raise welfare especially in rural and informal segments.
In BE 2025, the Union Government's spending on beneficiary schemes amounts to ₹7.1 lakh crore, and states together add another ₹7.4 lakh crore. This totals to nearly ₹14.5 lakh crore.
According to PLFS data, the average monthly earning by regular salaried worker is approx. Rs 21,000 and approx. Rs 14,000 for self-employed. The average earning per day by a casual labourer is Rs 433. Using these approximations and accounting for dependency assuming a family of four, this translates to an income of Rs 65,000 per capita. Assuming 80 per cent of the total beneficiary schemes reaches bottom 50 per cent, this translates into Rs 15,000 per year/per person accounting for leakages and overlaps through direct and indirect benefits. This uplift of approx. 20% in effective resources translates into consumption. Thus, even under these conservative assumptions, this significantly compresses effective inequality.
These interventions have also led to a dramatic fall in poverty, with the extreme poverty rate dropping from 16.2 per cent in 2011–12 to 2.3 per cent in 2022–23. At the lower-middle-income line of $3.65/day, poverty fell from 61.8 per cent to 28.1 per cent.
Before accepting WID's estimates at face value, shouldn't we ask what exactly they are measuring? Coming onto the WID database, their benchmark income concept is: 'Pre-tax, post-replacement national income', that is, before taxes and transfers, except for social insurance components like pensions and unemployment benefits.
This means that they exclude most non-contributory welfare transfers — like India's Direct Benefit Transfers (DBT), food subsidies, LPG schemes, Ayushman Bharat, rural housing, and more.
India's social protection system relies much more heavily on non-contributory transfers than contributory insurance. These are not counted in the WID's income concept, even though they materially raise real income and purchasing power.
This creates a systematic downward bias when WID measures inequality in India by ignoring the redistributive effect of these targeted schemes and Inflating the apparent concentration of national income at the top. So, under WID's income inequality framework, we are essentially saying that major upliftment schemes in India — have zero impact on measured inequality.
Secondly, WID relies heavily on tax records to compile its database. Now, even if we look at tax records, Gini coefficient estimated using ITR data of taxable income of individuals shows that individual income inequality has decreased from AY15 (FY14) to AY23 (FY22) from 0.472 to 0.402. 43.6 per cent Individual ITR filers belonging to the Income group of less than Rs four lakh in AY15 (FY14) have left the lowest income group and shifted upwards.
A comparison of disparity in income during FY14 and FY23 shows that there is a clear rightward shift in the income distribution curve signifying people in lower income brackets are increasing their income to converge towards their share in population. The bell-shaped curve for AY24 speaks more!!
In FY14, the share of the top one per cent in total income was 1.64 per cent, which fell to 0.77 per cent in FY21. Furthermore, tax buoyancy of 1.1 alongside falling cost of collection shows better compliance and hence must not be misread as rising inequality.
If India's official tax data shows improving progressivity, and large-scale consumption surveys indicate a sustained reduction in inequality, then it is worth asking why WID estimates tell such a different story.
To argue that India remains deeply unequal based solely on selectively elevated income estimates is much like claiming the country lacks water because Rajasthan faces water scarcity.
Inequality, like deprivation, is not monolithic—it varies across dimensions, regions, and measurement tools but that does not invalidate the broader progress being made. Any evaluation that ignores these dynamics in favour of a narrow, partial view risks obscuring the very progress it seeks to critique.
As we move forward, two takeaways are critical.
First, improved reporting is not the same as increased disparity—and we must resist reacting to shadows cast by better data.
Second, and most importantly, welfare economics must always return to its core question: what improves the lived experience of the bottom half?
In that, India's story over the past decade is less about divergence at the top and more about convergence at the base—quiet, broad, and measurable in the resources people use.
(The authors are a member of the 16th Finance Commission and Group Chief Economic Advisor, State Bank of India and the other an Economist at State Bank of India. Views are personal)