Kaustubh, Satadru Das, Pawan Gopalakrishnan, and Debojyoti Mazumder^{^}

This article assesses consumption inequality in India by analysing convergence or divergence in household consumption expenditure among various groups. Further, the article estimates the incidence of poverty across states by updating the Rangarajan Committee poverty line. The study finds that there is a broad trend towards convergence in consumption expenditure among households but there are a few notable exceptions. A significant decline in poverty incidence across states in India is also found.

Introduction

In recent years, India has experienced a deceleration in consumption growth, a trend with significant implications for economic development, poverty reduction, and social equity. Consumption, a critical driver of economic growth, reflects not only aggregate demand but also the living standards and well-being of households. Analysing consumption inequality and regional convergence is vital, as disparities in consumption can perpetuate economic and social inequities, while convergence can indicate progress towards more inclusive growth. While trends in real per capita gross state domestic product (GSDP) show growing disparity among states over the past decade (Kaustubh and Ghosh, 2023; 2025), it becomes essential to examine how is it materialising in the context of consumption. This prompts a crucial question: Is the growing spatial income inequality leading to a corresponding rise in consumption inequality? Several important factors over the last decade like the implementation of the goods and services tax (GST), and occurrence of unexpected shocks, like the COVID-19 pandemic, are postulated to have substantial effect on inequality. While some analysts have pointed to anecdotal evidence and data from other sources to argue that these events contributed to rising inequality (Himanshu, 2017; Jha and Lahoti, 2022; Ghosh, 2024), other studies suggest that the impacts of these events were either temporary or led to a reduction in inequality (Chanda and Cook, 2022; Gupta et al., 2021). Enhanced labour mobility across states and well-targeted policy measures could help mitigate the impact of rising income inequality on consumption.

Considering the contrasting findings and hypothesis, the Household Consumption Expenditure Survey of 2022-23 by National Sample Survey Organization (NSSO) provides an unparalleled opportunity to examine these issues in detail. The survey was released by the Ministry of Statistics and Programme Implementation (MoSPI) in July 2024, after a gap of eleven years. Covering 2.6 lakh households, this dataset offers granular insights into consumption behaviours across income groups, states, and rural-urban divides, enabling a comprehensive analysis of the drivers of inequality and the extent of regional convergence. Understanding these trends is particularly relevant for India, where addressing consumption disparities is essential to achieving sustainable development goals and fostering equitable economic progress. By investigating these dynamics, this study aims to inform policy interventions that can promote balanced growth and ensure that the benefits of development are more evenly distributed across the population.

Most of the discussions in media and academic research regarding the HCES 2022-23 have focussed on the aggregate numbers from the survey. These articles have analysed the major trends emerging in the consumption pattern of the Indian economy by comparing the surveys of 2011-12 and 2022-23. The findings of these articles can be broadly summarised as follows: there is an increase in the average monthly per capita expenditure (MPCE) between 2011-12 and 2022-23 in both rural and urban areas; a decrease in the rural and urban gap in consumption; a decline in the share of food in total consumption expenditure; and finally, a decrease in consumption inequality across income deciles (Rampal, 2024; Iyer, 2024; Nageswaran et al., 2024).

Previous studies have found a greater spatial inequality in consumption of non-food items compared to food items and spatial inequality in rural MPCE is more than urban MPCE between states in terms of MPCE based on HCES 2022-23 (Mitra and Shrivastav, 2024). In similar line, (Bonu, 2024) analysed the consumption inequality among social groups. It is found that Scheduled Tribes in Odisha, Jharkhand and Chhattisgarh are at the bottom of the consumption pyramid. Broadly, they provided evidence that the scheduled tribes are at bottom of the consumer pyramid followed by scheduled castes in almost all states. These studies focused only on HCES 2022-23 to analyse various dimensions of consumption inequalities. However, the importance of comparative analysis between different HCES rounds were broadly overlooked.

The present article contributes to this discussion by analysing the trends in spatial and temporal variation in consumption expenditure by analysing micro-data from the HCES 2022-23 and comparing them with previous rounds of NSSO consumption surveys.

Through comparative analysis of NSSO survey rounds of 2011-12 and 2022-23, the study finds a reduction in expenditure inequality across several dimensions, including between urban and rural households, between high consumption and low consumption households, and across states. This decrease in consumption inequality could also have implications for poverty levels. To explore this, the poverty incidence is estimated by appropriately adjusting the consumption expenditure based poverty line recommended by the Rangarajan Committee in 2014 to current prices. Results from these estimations reveal a significant decline in poverty incidence across states.

The rest of the article is organised as follows: Section II analyses variation in consumption expenditure across expenditure classes, Section III presents variation between urban and rural households, while Section IV examines variation across states. Section V updates the Rangarajan Committee poverty line for the large states and calculates the incidence of poverty. Finally, Section VI concludes by offering some policy insights.

II. Convergence across Expenditure Classes

To analyse the disparity across expenditure classes, the ratio between the mean expenditure of each expenditure class and the median expenditure of the entire sample is calculated. The exercise is done separately for urban and rural households¹. For convergence, the ratios should increase for the fractiles below median and decrease for fractiles above median between 2011-12 and 2022-23.

It is observed that in urban areas, the ratio increased for most of the fractiles below median and decreased for most of the fractiles above median (Chart 1). In rural areas, the ratio has remained unchanged for most of the fractiles below median but has decreased for the top four fractiles (Chart 2). It can be concluded that inequality across expenditure classes has decreased for both urban and rural households compared to 2011-12, although the decline has been more in urban areas.

III. Convergence between urban and rural households

The urban-rural gap is measured using the urban premium which is defined as the ratio of real urban MPCE and real rural MPCE. The real values are calculated by discounting the nominal MPCE for urban and rural households with their respective consumer price indices (CPIs) so that all values can be expressed in terms of 2011-12 prices. This ratio has seen a decline in trend since reaching its peak in 2009-10, indicating that the urban-rural gap in terms of consumption has continued to decline since 2009-10 (Chart 3).

However, further analysis shows that the fall in urban premium is not uniform across expenditure classes. Specifically, the urban premium has increased among the bottom 30 per cent and has decreased for top 70 per cent of consumption fractiles (Chart 4). This implies that the urban-rural gap has fallen for better-off households but has increased for poorer households.

Next, the estimated urban premium is compared across states to assess if it is lower in richer states (states with higher average MPCE) as compared to poorer states and if that relationship has remained same over time. State-wise average MPCE are calculated by weighting the average urban and rural MPCE in a state by the respective urban and rural population from the 2010-11 Census (the same used by NSSO for the survey). The relationship between the state-wise urban premium and MPCE is examined separately for 2011-12 and 2022-23.

The urban premium has a negative relationship with states’ average MPCE. This implies that states with higher MPCE have lower urban-rural gap (Chart 5). The relationship, however, has changed between 2011-12 and 2022-23. Comparing the slopes of the fitted lines in the scatter plots, it is observed that this relationship has become weaker. The fall in slope is primarily due to the urban premium falling in states which have lower MPCE than the national average. This indicates the urban-rural gap has substantially decreased in the lower MPCE states between 2011-12 and 2022-23.

IV. Convergence across States

The literature on inter-state disparity in economic outcomes and its dynamics in the Indian context has mostly focussed on the growing divergence in per capita incomes² among Indian states, leading to increasing spatial economic inequality. Broadly, these studies have shown that there exists a beta-divergence in per capita GSDP across Indian states during recent decades (Ghosh and Kaustubh, 2023; 2025). As regarding to the dynamics of MPCE between Indian states, (Acharya et al., 2021) had done a detailed study based on earlier rounds of NSSO consumption expenditure surveys and found evidence of overall divergence in MPCE between the states over the period between 1993-94 to 2011-12, and some evidence of conditional convergence when states were grouped based on their standards of living. Some opinion pieces in news media like (Mahambare and Jyoti, 2024) in the context of the household consumption survey round of 2022-23 has pointed to anecdotal evidence of dichotomy between divergent per-capita GSDP/NSDP and convergent MPCE in some states. For instance, while Gujarat and Maharashtra have higher per capita NSDP than Rajasthan, the MPCE in rural Rajasthan is higher than rural Gujarat and Maharashtra. They speculate that such a dichotomy may be attributed to several factors such as high labour mobility which generates higher incomes in labour-exporting states compared to their GSDP levels, and government policies and schemes designed to mitigate the impact of growing spatial economic inequality on the living standards of people.

Given this context, the present study investigates the inter-state dynamics of MPCE between the period 2011-12 to 2022-23. A state-level beta-convergence analysis is performed for the period 2011-12 to 2022-23, separately for urban and rural areas. Specifically, beta-convergence occurs when regions with lower initial values experience faster growth than their wealthier counterparts. This relationship is examined by regressing growth in urban and rural real MPCE between 2011-12 and 2022-23 (Δ Real MPCEi) of the ith state on the logarithm of its MPCE in 2011-12 (Log(MPCE)_i,2011-12) for 32 states and UTs as shown in equation 1. The standard errors (ei) are made robust to heteroskedasticity. A statistically significant negative coefficient of log of MPCE in 2011-12 (β) will denote unconditional beta-convergence and the higher magnitude of the coefficient will denote higher speed of convergence (Barro and Sala-i-Martin,1992; 1995).

Testing the above regression obtains a significant negative relationship between growth in MPCE and the log of MPCE in 2011-12 for both urban and rural areas, indicating beta-convergence in per capita consumption across states (Table 1). This implies that states with lower MPCE in 2011-12 experienced faster consumption growth, resulting in a reduction of inter-state inequality in consumption between 2011-12 and 2022-23. However, the speed of convergence in rural areas is higher than urban areas as reflected by the higher magnitude of β for rural areas compared to urban areas.

To better visualise the beta convergence in MPCE across states, a scatter diagram between state-wise nominal MPCE of 2011-12 (in log) and the corresponding growth in real MPCE from 2011-12 to 2022-23 is plotted for both rural and urban sectors (Chart 6). For beta convergence, the fitted line of the scatter plot must have a negative slope, as is observed. The vertical dashed line in black is the national average MPCE in 2011-12. States on the left of the vertical line had lower MPCE than the national average in 2011-12.

States like Odisha, Madhya Pradesh, Uttar Pradesh, Chhattisgarh, Jharkhand, and West Bengal had lower MPCE than the national average and lie below the fitted curve (red line), implying that these states will take comparatively longer time to reach the all-India average MPCE as their real MPCE growth was below the level suggested by the convergence estimation in relation to their MPCE in 2011-12. On the other hand, most north-eastern states as well as Tamil Nadu and Andhra Pradesh (including Telangana) had higher MPCE than the national average in 2011-12 and lie above the fitted line, indicating that their real MPCE growth was above the level suggested by the convergence estimation.

Thus, the data shows evidence of unconditional beta-convergence in per-capita consumption expenditure across states for both rural and urban areas for the period of 2011-12 to 2023-24. This is in contrast to the results found by (Acharya et al. 2021) for the period between 1993-94 to 2011-12-. However, similar to the findings of the earlier study, there are heterogeneities in the speed of convergence between rural and urban areas, and across states. It should be noted that while the per capita GSDP of Indian states has been diverging (Ghosh and Kaustubh, 2023; 2025), the MPCE in both rural and urban areas is showing signs of convergence. This suggests that living standards across these regions are becoming more similar, even as their per capita economic production continues to diverge.

V. Estimating the Poverty Line

This article follows the methodology proposed by the Rangarajan Expert Committee on poverty, constituted by the Planning Commission in 2011 to estimate incidence of poverty. Consumption-based poverty lines determine the amount of money required to purchase a basket of essential items for a household. Micro-level data on household consumption expenditure can reveal the percentage of households spending more than the amount specified by the poverty line. A decline in inequality in a growing economy will imply expenditure of poorer households is increasing faster than richer households. To the extent that there exists spatial variation in convergence of consumption expenditure, there may be difference in poverty reduction across states.

A few recent papers have calculated the state-wise incidence of poverty in India based on consumption data of HCES 2022-23 (Bhasin and Bhalla, 2024; Sethu et al., 2024). They have either used the World Bank purchasing power parity (PPP) poverty lines of $1.9 and $3.2 or domestic poverty lines (suggested by Tendulkar Committee or the Rangarajan Committee) after inflating them with CPI headline inflation to account for change in prices. However, there are issues with these approaches. There is a lack of consensus regarding the PPP conversion factor required to convert World Bank’s poverty lines, which are defined in dollars, into rupees. For inflating the Rangarajan or Tendulkar poverty line with headline CPI inflation, the weights of consumption items in the CPI basket can be very different from the weights of those items in the poverty line basket (PLB). For example, the weight of food in the Rangarajan PLB for rural areas is 57 per cent compared to 54 per cent weightage in the rural CPI consumption basket. Similarly, the weight for food in Rangarajan PLB for urban areas is 47 per cent compared to 36 per cent weightage in urban CPI consumption basket. Thus, the headline CPI inflation may not be the appropriate metric to gauge the inflation in the PLB.

To tackle this, a price index based on the weight of each item in the Rangarajan PLB, rather than using the item weights from the CPI basket, is developed. The price changes for each item between 2011- 12 and 2022-23 are calculated, adjusting for these changes by the respective item’s weight in the PLB to determine inflation for the PLB during this period. This analysis was carried out separately for urban and rural households in each state. Following this, the Rangarajan Committee poverty lines in nominal terms were updated by adjusting the 2012 poverty line with the inflation rates calculated as discussed in the previous step. Finally, the percentage of households below the updated poverty line in urban and rural areas of each state using the HCES 2022-23 micro-data are obtained.

The method suggested by the Rangarajan Committee, which submitted its report to the Planning Commission in 2014, had some distinct advantages over the previous official poverty line constructed by the Tendulkar committee. First, the Rangarajan Committee’s methodology defines different PLBs for urban and rural populations based on the difference in the consumption pattern between the two groups. Second, the methodology re-links poverty line with minimum calorie needs after the Tendulkar Committee’s methodology delinked them. Third, the Rangarajan Committee report gives the weightage of each item in the PLB for urban and rural areas, which makes it possible to calculate the nominal value of these baskets in 2022-23 prices for each state.³

Table 2 contains the state-wise poverty line in 2011-12 as reported in the Rangarajan Committee Report and the poverty line for 2022-23, which is estimated using the methodology described above for large Indian states.

Using micro-data from HCES 2022-23, the percentage of population below the poverty line in each of the states in 2022-23 were calculated. These are reported in Table 3 along with the poverty percentages of the states in 2009-10 and 2011-12, respectively, from the Rangarajan report of 2014. A substantial decrease in the incidence of poverty is observed in both rural and urban areas across each state in HCES 2022-23 compared to HCES 2011-12. Other studies which have used the HCES 2022-23 have also found substantial decline in poverty during the period although the poverty rates calculated by them is different from the estimates in this study, this is due to differences in methodology. The paper by (Bhasin and Bhalla, 2024) estimated total poverty headcount ratio of 2.3 per cent overall in 2022-23 which was 12.2 in 2011-12. The poverty line they estimated is based on the World Bank’s measure of consumption expenditure of 1.99 USD per day which they convert into Rupee terms using purchasing power parity measure. (Sethu at al., 2024) also estimate poverty incidence using the HCES 2022-23 and found poverty rates of 23 per cent and 27.4 per cent for rural and urban respectively, for all India. Their poverty rate is based on a poverty line constructed by them which has updated calorific needs based on the occupation and age of the family members of the household. It is to be noted that the present study does not construct any poverty line but updates the Rangarajan poverty line using item-wise price indices.

Conclusion

This study compares the consumption expenditure numbers from the Household Consumption Expenditure Survey of 2022-23 by the NSSO with the survey done in 2011-12 focussing on convergence across various dimensions such as expenditure classes, urban-rural areas, and states. The results show that there is a broad trend of convergence in MPCE. A closer analysis reveals that there are some exceptions to the broad trend. The decrease in inequality in consumption expenditure across expenditure deciles and regions may be explained by more effective delivery of various government schemes. Also, the study re-estimated the poverty line for each state with this data keeping the Rangarajan Committee definition. The estimated incidence of poverty shows an overall reduction of poverty across states in India.

References

Acharya, D., Rath, B.N., Parida, T.K. (2021). Convergence in Monthly Per Capita Expenditure and the Rural–Urban Dichotomy: Evidence from Major Indian States. In: Batabyal, A.A., Higano, Y., Nijkamp, P. (eds) Rural–Urban Dichotomies and Spatial Development in Asia. New Frontiers in Regional Science: Asian Perspectives, vol 48. Springer, Singapore. https://doi.org/10.1007/978-981-16-1232-9_10

Barro, R. J. and Sala-i-Martin, X. (1992). Convergence. Journal of Political Economy, 100(2), 223-251.

Barro, R. J. and Sala-i-Martin, X. (1995). Economic Growth. London. McGraw-Hill.

Bhasin, Karan and Bhalla, Surjit (2024). Poverty in India over the last decade. Economic and Political Weekly.

Bonu, S. (2024, August 22). Consumption Expenditure by Social Groups: Findings from India’s Household Consumption Expenditure Survey, 2022-23. Isaac Centre for Public Policy, Ashoka University.

Chanda, A., and Cook, J. C. (2022). Was India’s demonetization redistributive? Insights from satellites and surveys. Journal of Macroeconomics.

Dev, M. and Rangarajan, C. (2024, March 12). With new consumption survey, the need for new indices, Indian Express.

Ghosh, S. (2024, January 15). India’s K-shaped recovery debate has economists divided. Mint.

Ghosh, T. and Kaustubh. (2023). Growth decomposition of the Indian states using panel data techniques. Applied Economics, 56(39), 4664–4684.

Ghosh, T., and Kaustubh. (2025). Growth Divergence between Indian States. Economic and Political Weekly, 60(6)

Gupta, A., Malani, A., and Woda, B. (2021). Inequality in India declined during covid. NBER Working Paper.

Himanshu. (2017, November 6). Demonetisation, inequality and informal sector. Mint.

Iyer, Vaidyanathan P. (2024, February 26). Five questions the Household Consumption Expenditure Survey 2022-23 answers. The Indian Express.

Jha, M., and Lahoti, R. (2022). Who was impacted and how? COVID-19 pandemic and the long uneven recovery in India. WIDER Working Paper.

Mahambare, Vidya and Jyoti, Amar. (2024, March 19) What tells us more about prosperity: Consumption or GDP per head? Mint.

Mitra, A., & Shrivastav, P. K. (2024). Interstate inequality in consumption: Emerging evidence from HCES 2022–23. Economic & Political Weekly, 59(28), 17–20.

Nageswaran, V. A., Guru, Anuradha and Bisht, D. S. (2024, March 6). India’s consumption numbers fit into our big picture of progress. Mint.

Rampal, Nikhil (2024, February 27). Here are 5 key takeaways from Household Consumption Expenditure Survey 2022-2023. The Print.

Sethu, C. A., Abhinav Surya, L. T. and Ruthu, C. A. (2024), Poverty in India: The Rangarajan Method and the 2022–23 Household Consumption and Expenditure Survey. Review of Agrarian Studies, 14(2).

^{^} The authors are from the Department of Economic and Policy Research. The views expressed in this article are those of the authors and do not represent the views of the Reserve Bank of India.

¹ Inequality is defined as the difference between the average MPCE of the top fractiles and the bottom fractiles. In the charts, the ratio between the average MPCE of each quantiles and the median MPCE are plotted as bars. This ratio will be lesser than 1 for quantiles below 50 and greater than 1 for quantiles above 50. An increase in the ratio for the fractiles below 50 will imply a movement towards the median and a fall in the ratio for the fractiles above 50 will also imply a movement towards median, therefore, implying a decrease in inequality.

² State level per capita incomes are measured using state level gross state domestic product (GSDP).

³ It may be noted that the poverty line estimated in the article is based on the methodology proposed in the 2014 Rangarajan Committee report. Ten years have passed since the release of the report, and consumption patterns may have shifted with new items becoming more important in the consumption basket. This may necessitate creation of a new poverty line which may capture the current reality better (Bhasin and Bhalla, 2024; Dev et al., 2024).