TCA Anant: A household income survey will be valuable if clarity beats confusion

In this context, let us review some of the challenges underlying income measurement in a developing country like India. To grasp the nuances, we need to understand that the economist’s conception of terms like ‘income’ and ‘profit’ differs markedly from what is in common use.

Also Read: All-India income survey: Let’s not miss this opportunity to get it right

When I asked ChatGPT to provide an economist’s understanding of these two terms, it gave me this:

“Income is defined as the total flow of resources or command over goods and services received by an individual, household, or entity over a specified period of time. It includes all forms of compensation, returns on assets, and unearned transfers that enhance economic well-being and can be used for consumption or saving." And: “Profit is the residual income that remains after all explicit and implicit costs of production have been deducted from total revenue. It represents the reward for entrepreneurship and risk-taking in the production process." In contrast, when asked to define these terms in common usage, the definitions it offered were:

“Income is money they receive or earn regularly which they can use to pay for things they need or want." And: “Profit is the money left over after paying all the costs."

Common perceptions of both often overlook certain sources of income or types of expenses. For instance, employer-provided medical insurance or casual earnings from a temporary food stall set up for a festive occasion may not be seen as income. Similarly, expenses on travel to sell goods, work-related clothing or meals consumed away from home may not be considered costs. These distinctions are complicated by tax provisions, which allow some and disallow others.

Also Read: India’s informal sector isn’t off the map: It’s being tracked better than ever

The challenge of measuring income is particularly acute in developing countries, where a large portion of the workforce defines itself as self-employed. In such cases, household income becomes a composite of both wages and profits. The problems are manifold. Earnings can be seasonal, with different activities in different seasons. But a sample survey, even if it covers these households, may do so only during a particular season. 

A version of this problem occurs when we seek to measure the income of agricultural households. The solution adopted for the Situation Assessment Survey of agriculture conducted as part of the National Sample Survey’s (NSS) 70th and 77th rounds was to visit each household twice—once in each six-month period—on the assumption that there are two crop seasons. This approach was criticized as many parts of the country now have more than two crops a year. 

An appropriate sampling design would need to accommodate all occupational categories beyond farming.

Another difficulty lies in defining and measuring costs, particularly of non-purchased inputs. Many home enterprises use dual-use durables. A sewing machine, for example, may be used to make clothes for sale as well as personal use. 

Also Read: The mystery of India’s missing middle class: Is it really part of our growth story?

After encountering these problems, the NSS partly addressed them by designing questions and training investigators carefully. An additional challenge here is to account for different patterns of production-tool usage over the year and possibly across different seasonal activities.

One of the stated objectives of the survey is to provide information on income distribution. This requires that income from different forms of employment be measured in a strictly comparable manner, which is not easy. Consider a worker earning a regular salary who spends on transport, clothing or eating out as part of the cost of retaining that job. 

Should these be treated as costs and deducted from income? 

In the case of a self-employed person, some of these (like travel and clothing expenses) may be allowed as legitimate costs and deducted from gross earnings. But for a salaried worker, our usual approach—based on tax laws—is to treat gross earnings as income. To make income measures more comparable, we would need to account for such costs and properly value job-related perquisites.

These challenges are in addition to the usual difficulties in measuring the earnings of hard-to-track groups like traders, lawyers, chartered accountants, doctors, etc, and in measuring capital gains. The first issue is addressed by the department of income tax through presumptive income provisions; the second was discussed in an earlier column (shorturl.at/e0jdh).

Also Read: Himanshu: India’s economy is too complex to afford less-than-robust statistics

This is not to argue against a household income survey, which will be very valuable, but to urge caution in interpreting its results, especially when comparing data across countries (or even states) where the composition of work varies significantly. Care must be taken to ensure that such differences are accounted for. The reporting challenge will be particularly acute, with hard-to-survey areas like gated colonies and high-income neighbourhoods likely to worsen a ‘truncation bias,’ with the upper end of India’s income distribution at risk of being left out.

Getting India’s official income survey right will require utmost clarity in its design as well as instruction manual for investigators to explain concepts to respondents. So, bon voyage and caveat emptor.

The author is a visiting professor at the Institute for Studies of Industrial Development and former chief statistician of India.