Results from a fresh round of the NCAER's Market Information Survey of Households (MISH) are now becoming available. They prompt reflection both on where we have come from and where we might be headed.
The MISH is a household survey of consumer spending on selected commodities. The survey has been undertaken annually (with a few gaps) by the National Council of Applied Economic Research since 1985-86.
The sample and questionnaire design have remained substantially unchanged over this period. As such, the sequence of surveys provides a consistent perspective on the Indian consumer market over almost twenty years.
The MISH survey complements the consumption surveys of the National Sample Survey Organisation (NSSO). MISH asks questions about fewer commodities than the NSS, but more frequently. The commodities covered by the MISH are a specified group of consumer durables and fast moving consumer goods (FMCG).
More importantly, MISH asks the head of household to estimate household income over the reference period (the prior fiscal year), while the NSS does not attempt to measure income.
The MISH concept of perceived household income differs from (and is typically about two-thirds of) the estimate of personal disposable income estimated in the National Accounts. Despite their limitations, the MISH data are one of the few sources of data on income distribution for India.
Given the size of the data collection and processing effort (about 300,000 urban and rural households are canvassed), tabulations can only become available well after the reference year, and data for fiscal 2001-02 are now becoming available.
Key statistics are given in the table; these data extend a series previously published in NCAER's publication Indian Market Demographics Report 2002, to which the reader is referred for definitions and details of sample design.
The opportunity provided by the new data has also been used to update (and extend) NCAER's projections of distribution of income by household class to 2009-10.
The income bands used are designed to be consistent with bands used in the original MISH surveys. For convenience, five bands have been collapsed into three, which we now call Low Income (L), Lower Middle (LM) and Middle High (MH).
These have been updated to 2001-02 using national rural and urban price indices. There is no clear link between these numbers and the official poverty estimates.
The poverty numbers refer to consumption, not income, and they are based on per capita household consumption.
Many stories emerge from this table. In the eleven years following 1989-90, the total number of households increased by 32 per cent. New household formation is a potent source of demand the world over, and India is no exception.
Strikingly, the rural share of total households has remained virtually unaltered, at around 71 per cent. While India is already the country with the largest urban population in the world, the big urbanisation spurt still lies ahead, and will continue for a long time to come.
As one would expect in a period of steady, sustained growth, there has been a migration of households from lower to higher income bands (at constant prices).
The increase in the aggregate number of households, when combined with this process of migration, has meant a leap of 260 per cent in the number of households in the Middle High category over the last eleven years.
With average household size just above five, a household monthly income of more than Rs 7,500 at 2001-02 prices (approximately $155 at the exchange rates then prevailing) may not feel like any kind of affluence. It is worth reminding ourselves that only 14 per cent of households enjoyed such real income at the beginning of the decade.
The growth in incomes in the top band has been experienced by both urban and rural households, both of which have roughly doubled as a share of the total population over the 1990s.
Given that the number of rural households remains almost two-and-a-half times that of urban households, the absolute numbers of households in this band are not too far apart: 29 million urban households versus 24 million rural households.
The implied aggregate purchasing power of the two sectors is even closer, given that the real purchasing power of a given income level is higher in a rural setting than an urban. The rural market is thus a significant source of demand for a number of consumer goods.
There have been palpable reductions in the absolute numbers of households in the lowest income band. This is particularly clear in the case of urban households, where the number of low-income households has been approximately halved, from 14.9 million households in 1989-90 to 7.6 million in 2001-02.
The number of low-income rural households has also declined, from 69 million in 1989-90 to 58 million in 2001-02, despite the relatively indifferent performance of agriculture over the decade.
What is behind this growth in household incomes? Economic growth is clearly the principal driver. But it is also important to remember that in India the share of GDP going to consumption has remained relatively high.
In its monthly newsletter, Straight Talk (February 2004), the authoritative Conference Board of the United States focuses on China. It notes that the immense scale of fixed investment in the Chinese economy, much envied by many Indian economists, has meant that the consumer sector has played a less important role in China's growth.
As a result the Chinese economy is highly leveraged to investment spending. Since investment is more volatile than consumption, Chinese growth is prone to greater instability. The Indian growth pattern is closer to that of the US, driven by buoyant consumer demand and rising productivity.
Unlike the US though, we are able to meet our investment requirements entirely from our own savings (despite the dissaving by government), and even occasionally run a small surplus on the current account.
Absent a political or financial crisis, there seems to be no inherent reason why this process will not continue. In 1998, the NCAER attempted to project the evolution of households by income class, based on assumed sectoral growth rates.
As (RBI Deputy Governor) Rakesh Mohan noted in his recent Ayaz Peerbhoy lecture, the outcome for 2001-02 is close to the projections, providing some confidence in the methodology. Using 2001-02 as the base year, sectoral GDP has now been forecast till 2009-10 using NCAER's medium-term macro model.
The forecast sectoral growth rates for the period 2001-02 to 2009-10 are 3.21 per cent for agriculture, 7.0 per cent for industry, 8.1 per cent for infrastructure and 8 per cent for services. This implies real GDP growth of 6.75 per cent over the period, taking into account the very low growth of 2002-03.
The impact on household incomes by income band and sector are shown in the third panel of the table. The impact on rural incomes is particularly noteworthy, with further steady decline in the number of families in the lowest income band.
India's growth experience in the last decade appears to have been balanced and sustainable, save for the fiscal dimension. The links between GDP growth and household incomes seem to be working well.
Some might argue that this is because of government transfers and subsidies, particularly in the food economy, but personally I see little evidence of sustained aggregate demand deficiency.
Given the apparent strength of consumer demand, now seems to be an ideal time to take fiscal adjustment seriously. Indian households would be the beneficiaries.
The author is Director-General of NCAER, New Delhi. The views expressed here are personal.