Bibtex
Endnote
RIS
The 2022-23 Household Consumption Expenditure Survey (HCES23) released last fortnight is important for several reasons. First, it provides estimates of private consumption for rural and urban segments, not available since 2011-12 (there are other surveys which provide estimates, but this is the most comprehensive). GDP data on private consumption provides only the total composite. Second, we get estimates of consumption spends across “affluence” bands, which is important for analysing inequality. Third, HCES23 estimates the number of households (HHs) in India on a large sample; we have 195 million rural and 86 million urban households and an average HH size of 4.9 persons. This is the closest to insights on the structure of India’s demographics in the absence of the 2021 census.
Changing consumption patterns
The most notable feature is the sharp drop in the share of food in the total consumption basket relative to the shares in the National Sample Survey 2011-12, which is the closest possible comparison available. In rural areas, the share of food in monthly per capita expenditures (MPCE) fell to 46.4%, a 6.4 percentage point (ppt) drop from 2012 levels. Almost all of this came from lower spends on cereals (-5.8 ppt). Spending on pulses, vegetables and edible oils have also fallen (by much smaller magnitudes). On the other hand, spends on beverages and processed foods had increased significantly (by 1.7 ppt), indicating the spread of retail outlets (and probably changing tastes). Spending on “fuel and light” was also down by 1.3 ppt.
The Household Consumption Expenditure Survey 23 estimates the number of households (HHs) in India on a large sample; we have 19.5 rural and 8.6 urban households and an average HH size of 4.9 persons. This is the closest to insights on the structure of India’s demographics in the absence of the 2021 census.
The reduced spend on food has been widely attributed to the free cereals provided under the PMGKAY to more than 81 crore persons. But is this correct? The data also provides “imputed” monthly per capita consumption expenditure (MPCE), which adjusts for subsidies and freebies. Versus the nominal MPCE in rural areas of Rs 3,773, the imputed spend was Rs 3,860; i.e., a difference of just Rs 87 per person per month due to the free provision of rice and wheat. The difference in these amounts for urban areas is even lower (Rs 62). These just sound too low to account for the effects of subsidised food and services.
The reduced spend on food has been widely attributed to the free cereals provided under the PMGKAY to more than 81 crore persons. But is this correct?
So, should the data and calculations thereof be revisited, or are the consumer preferences of Indians changing as malls and brands grow and take root in smaller towns? The inconsistencies with anecdotal references point to the need for a cautious reading of the upsides of the data, and the need for critical inputs to correct what clearly are structural weaknesses in the system, notably when it comes to health and education.
Spends on education have fallen 0.2 ppt; even the per capita monthly spends (Rs 125 in rural and Rs 374 in urban areas) seems low in the context of anecdotal evidence. This might largely reflect a shift from private schools to government ones, supported by the Annual Status of Education Report (ASER) data. Similarly, the share of spends on medical care rose only 0.5 ppt, somewhat inconsistent with reports on high and increasing cost of medical care. It is likely that various interventions and subsidies by the Centre (like the Ayushman Bharat scheme) and State governments have helped keep spends in check. Rising health insurance coverage might also have played a role.
The surge in consumer loans
One of the drivers of these higher spend categories is “conveyance” (+3.4 ppt), followed by “consumer durables (+2 ppt) and “consumer services” (+1.1 ppt), on the back of a sharp rise in retail (consumer) credit over the past three years. A logical corollary is rising household debt, and the consequence of dipping into financial savings. Recall the concerns raised in early January when FY23 savings and investment data showed that household financial savings had dropped in FY23 (to 10.9% from 11.1% in FY22 and a pre-COVID FY20 11.6%).
HCES23 points to the need for caution in interpreting the data and hence the continued imperative for multiple policy interventions on socio-economic weaknesses which help India achieve a sustained high growth trajectory.
Distribution of consumption by amount of spends (fractiles)
More inconsistencies show up in spends across consumption levels. First, the average urban MPCE is just 1.7 times the rural MPCE, including rent. Urban cost of living is significantly higher than the rural cost of living. This does not square with anecdotal narratives of relative costs; although rural food inflation has indeed been higher over the past couple of years, this has been due to higher food inflation, which has a lower share in the consumption basket. House rent, in particular, has increased sharply in the post-pandemic years. Moreover, consumption of the top 5% is just 3.2 times the average for urban and 2.7 times in rural areas.
This does not fit in with data on the recent number of income tax filers (or the somewhat older data on income tax paid), which tells a very different story. For instance, data indicates that the average annual growth rate (over FY15 to FY23) of income tax filers with incomes over Rs 10 lakhs increased 15% vs 10% in the Rs 5-10 lakh bracket and 5% in less than Rs 5 lakhs, respectively. Do treat these numbers with some caution, but the orders of magnitude are likely correct.
More inconsistencies show up in spends across consumption levels. First, the average urban MPCE is just 1.7 times the rural MPCE, including rent. Urban cost of living is significantly higher than the rural cost of living. This does not square with anecdotal narratives of relative costs
Are the GDP private consumption and HCES numbers consistent?
HH consumption survey estimates are typically lower than GDP, given that methodologies vary and consumption surveys do not capture activity of trusts, endowments, etc. However, this divergence is accelerating over time. The gap between the two estimates was 45% in 2009-10 and 2011-12 (using the earlier 2004-05 GDP base), and has now increased to 53%. Note that GDP private consumption had increased 6.8% yoy in FY23, but is estimated to have slowed to 3.0% in FY24.
In conclusion, HCES23 is a critical report, opening a window on encouraging and improving patterns of private consumption across “income” stratas, States and geographies but also point to the need for caution in interpreting the data and hence the continued imperative for multiple policy interventions on socio-economic weaknesses which help India achieve a sustained high growth trajectory.