Census data baffles seekers of the middle class

HOW large is the middle class in India? The first answer was provided by some observers in 1991, shortly after the then Finance Minister Manmohan Singh had presented a new budget, and a statement of the new economic policy.

By From Kumaresh Chakravarty In New Delhi

  • Follow us on
  • google-news
  • whatsapp
  • telegram

Published: Mon 13 Oct 2003, 12:05 PM

Last updated: Wed 1 Apr 2015, 9:34 PM

Almost everybody concerned about the economy was keen to assess whether or to what extent foreign capital will be interested in investing in India, and how far foreign capital will initiate expansion of the domestic market. Some reform enthusiasts said India had a 250-millon strong middle class while some others went one step further and put the size of the middle class at 300 million.

The issue is yet to be settled. Experience has proved that the initial estimates, 250 or 300 million, were nothing more than wild guesses. More important, nobody, not even the makers of various estimates, ever explained what was meant by the term 'middle class,' and by what method the estimates were arrived at. No wonder, an organisation of British industrialists even issued a statement expressing their dissatisfaction over what they were told about the size of the middle class. They had expected a much larger market for their products than what it turned out to be. Several foreign companies which came to India for selling their products, closed shop while some others cut down the number of distribution outlets.

The most important reason for all that confusion and controversy lay in the fact that there was no dependable data base for arriving at an estimate. One had to depend on incomplete data. Here is an example. Suppose, five hundred thousand passenger cars were sold last year, as many in the year before, and another five hundred thousand have been sold this year. You may assume that the yearly sale, which indicates new demand for cars, is ten per cent of the potential aggregate demand which will materialise over a period of ten years. Which means that, the section of middle class which buy or will buy cars, is at least half-a-million at the going car prices. You can then think that if prices come down and/or the real income of a section of household goes up to a level where they would feel induced to buy a car, then the aggregate potential demand will go up at a certain rate.

You can go on doing the same thing about scooters, air-conditioners, refrigerators, television sets, and a good number of household gadgets. Depending upon your own concept of 'middle class,' you can classify the total number of households buying these things as different segments of the class, and project a growth rate in the relative numbers. If you are doing so, then you are actually defining 'middle class' by consumption baskets. Since consumption is a function of income (at given prices) and taste, you are actually taking income as the most important determinant. But, that is where the problem starts. And the problem has been concretely posed by the new census data for 2001.

Census is not a sample survey, it is a case-by-case enumeration. Therefore, logically speaking, census data are more reliable than sample survey data. Besides, whatever its limitations, Indian census is internationally accepted as substantially reliable.

Now take a cursory look at some of the highlights if this data. In early 2001, about 4.75 million households in India owned a car, a zeep or a van. Another 22.5 million owned a scooter or a motorcycle. Assuming that a car owner was not also a scooter-owner, the sum of these two groups of owner of personal vehicles was 27.25 million households. Taking them as individuals the number will be about 141 million. In percentage terms however, they were only 14.2 per cent of all households. Thirty one per cent of all households, or about 61 million, had a television each. Now, we can very justifiably think that a household owning a car or a scooter, surely had a television set also. So, if we want to find out those owning a television but not owning a personal vehicle, then we deduct the latter from the former. The number comes to about 34 million, or 16.8 per cent of all households.

Can we say that every television-owning household is a middle-class household? That will no be tenable. In Delhi, for example, 74 per cent households owned a television, though we know that those who live in slums or slum-like residential areas, are at least two-thirds of all households. Nobody will count slum-dwellers in such cases, as a part of the middle class. Also, in Delhi again, more than half of all households which had a television, did not have a telephone. And it is the same story, if not worse, in all states. Telephone owners in many states were only one-fifth to one-third of television owners. Be it fashion, or necessity, it is almost impossible to assume that, ordinarily a household not owning a telephone-when it is easily available-belongs to a social group which has the income and taste of a middle class in a country like India.

How then do we define the criterion for qualifying a household as middle class? If personal vehicle is made the criterion, then the size of the class will be as small as 27 million households or 135 million persons. If all television owners are also included, then the size goes up to 61 million households or 305 million persons.

But even the most successful seller of these goods will not accept it. They know a high proportion of television buyers come from low-income groups.

The latest census data has, thus put the seekers of middle class in a great quandary. What they do not see is that, the term 'middle class;' has proved to be completely redundant for assessing the actual or potential size of a class of consumers.

More news from