Poverty and inequality in India: measures and policy

Poverty and inequality questions look easy and are marked hard. Everyone can write "poverty has fallen but inequality remains". Examiners reward the candidate who knows how each number is produced, why different methods give different answers, and what that means for policy. This guide gives you the measurement toolkit and the Indian history in one place. It pairs with the growth and development guide, which covers the HDI.

Absolute and relative poverty

• Absolute poverty: inability to afford a minimum basket of goods and services, judged against a fixed poverty line.
• Relative poverty: being poor compared with the rest of society, for example below half the median income. Common in rich countries.

India's official approach has always been absolute and consumption-based, because consumption is measured more reliably than income in household surveys and smooths out seasonal swings in farm income.

How India drew the poverty line

For 2011–12, the Tendulkar method put poverty at about 21.9% of the population and the Rangarajan method at about 29.5%. The same survey gave different answers because the lines were different. That is the single best example to use when a question asks why poverty numbers are contested.

Poverty measures beyond the headcount

The headcount ratio (HCR) counts how many are poor but not how poor they are. Two further measures, from the Foster–Greer–Thorbecke family, fix that:

• Poverty gap index: the average shortfall below the line, as a share of the line, across the whole population (non-poor count as zero).
• Squared poverty gap: squares each shortfall, so it gives more weight to the poorest. It captures severity.

Worked example: poverty line = ₹100; five people with incomes ₹60, ₹80, ₹90, ₹120 and ₹150.

1. Poor: three people, so HCR = 3/5 = 60%.
2. Shortfalls as a share of the line: 0.40, 0.20, 0.10 (the other two count as 0).
3. Poverty gap = (0.40 + 0.20 + 0.10)/5 = 0.14, or 14%.
4. Squared gap = (0.16 + 0.04 + 0.01)/5 = 0.042.

Now suppose a transfer moves the ₹90 person to ₹105. The HCR falls to 40%, but the poorest person is untouched. A transfer to the ₹60 person would lower the squared gap far more. Policy that chases the headcount can neglect the poorest; that is why severity measures matter.

Multidimensional poverty

Income lines miss deprivations that money alone does not fix, such as a lack of clean cooking fuel or sanitation. The Multidimensional Poverty Index (MPI), developed by the Oxford Poverty and Human Development Initiative with the UNDP using the Alkire–Foster method, measures deprivations in three dimensions:

NITI Aayog's national MPI uses 12 indicators, adding maternal health and bank accounts.

A person is multidimensionally poor if their weighted deprivation score is at least one-third. Then:

• MPI = H × A, where H is the headcount of multidimensionally poor and A is their average intensity of deprivation.

Example: in a district, 20% of people are poor by this test (H = 0.20), and the poor are on average deprived in 45% of weighted indicators (A = 0.45). MPI = 0.20 × 0.45 = 0.09. If a sanitation drive lowers A to 0.40 without changing H, the MPI still falls to 0.08. That is the MPI's advantage: it rewards reducing the depth of deprivation, not only pulling people across a line.

Measuring inequality

• Lorenz curve: plots the cumulative share of income (or consumption) against the cumulative share of population, from poorest to richest. The 45-degree line is perfect equality.
• Gini coefficient: the area between the line of equality and the Lorenz curve divided by the whole area under the line of equality. It runs from 0 (perfect equality) to 1 (one person has everything).
• Palma ratio: the income share of the top 10% divided by the share of the bottom 40%. It focuses on the tails, where most change happens.
• Top income and wealth shares: from tax and wealth data.

Growth and inequality

The Kuznets hypothesis suggests inequality first rises, then falls as an economy develops (an inverted U). Evidence is mixed: many countries have seen inequality rise again at high incomes. For India, answers usually argue that the pattern of growth matters: growth in labour-intensive sectors reduces poverty faster than capital-intensive growth.

Causes of persistent poverty and inequality

• Low-productivity work in agriculture and the informal sector.
• Unequal access to quality schooling, health care and credit.
• Regional gaps between states, and within states.
• Social exclusion by caste, tribe and gender.
• Shocks: illness, crop failure and price spikes push the near-poor back into poverty. Health spending is a major cause.

Policy responses

A universal basic income was discussed in the Economic Survey 2016–17 as an alternative to many scattered schemes. It remains a debate, not policy.

Model answer outline (15 marks)

Question: "Poverty numbers in India are contested because of method, not data." Examine, and suggest how poverty should be measured.

1. Opening: absolute versus multidimensional measures; no official line since 2014.
2. Method matters: Tendulkar versus Rangarajan on the same 2011–12 survey (21.9% versus 29.5%).
3. Data also matters: gaps between consumption surveys; changed survey design; national accounts versus survey consumption.
4. Beyond headcount: poverty gap, severity, MPI (H × A).
5. Inequality: consumption versus income and wealth measures.
6. Way forward: an official line updated regularly; a dashboard combining consumption and MPI; faster, comparable surveys.
7. Conclusion: measures guide targeting, so method choices have real policy costs.

Short model answer (10 marks)

Question: Why is a multidimensional approach to poverty useful for policy?

Income or consumption poverty lines show whether a household can afford a basket of goods. They miss deprivations that income does not automatically remove, such as unsafe cooking fuel, poor sanitation, children out of school or malnutrition.

The Multidimensional Poverty Index measures these directly across health, education and living standards. A person is poor if deprived in at least one-third of the weighted indicators. The index equals the headcount of the poor multiplied by the average intensity of their deprivation, so it falls both when fewer people are poor and when the poor are less deprived.

This makes it a practical tool. Indicator-level results show which deprivation to attack in which district, whether that is cooking fuel, housing or school attendance, and they allow programmes to be tracked. Its limits are that weights are a judgment call and it does not measure income. The best approach uses both: a consumption line for material poverty, and the MPI for the non-income side.

Practice questions

1. Which committee's method gave a poverty ratio of about 21.9% for 2011–12?
   • Tendulkar.
2. What are the calorie norms set by the 1979 Task Force?
   • 2,400 kcal rural and 2,100 kcal urban per person per day.
3. If H = 0.25 and A = 0.40, what is the MPI?
   • 0.10.
4. Which poverty measure is most sensitive to the condition of the poorest?
   • The squared poverty gap.
5. What does a Gini coefficient of 0 mean?
   • Perfect equality.
6. What does the Palma ratio compare?
   • The income share of the top 10% with that of the bottom 40%.

What to do next

• Redo the poverty gap example with your own five incomes.
• Draw a Lorenz curve and mark the Gini areas.
• Note the latest official poverty and MPI estimates with their sources.
• Read the employment guide next; poverty and job quality are one story.