Measuring household income is tricky and expensive. Collection of reliable data on income has got obvious limitations due to several conceptual as well as data collection problems. This is even more so in a country like India where the majority of people are self-employed, and have multiple sources of sustenance, coming in lumpy and unpredictable flows, the very idea of income varies from person to person.
Report of the Canberra City Group of UN Statistical Commission, called ‘Expert Group on Household Income Statistics’ states that while collecting income data in the field, problems may arise due to ambiguities in defining income concept, in choice of unit of sampling, sampling frame and reference period of data collection. In addition to these, seasonality effect, lack of availability of accounts for employer households, significant amount of purchases through credit, hidden income generated through wages paid in kind, etc., are other factors affecting reliable income data collection. In India, National Survey Organisation (NSO), after making some early attempts in collecting income data, has subsequently refrained from collecting data on household income and concentrated only on household consumer expenditure surveys.
Collecting data on income is an arduous and expensive task, complicated by the high propensity for intentional or unintentional respondents’ biases. While several surveys report income as claimed by respondents, PRICE has chosen a methodology which is more rigorous though more difficult to implement, using the Canberra City Group guidelines for income estimation.
Canberra City Group Report has suggested a conceptual framework for income distribution analysis based on reconciliation of micro and macro approaches. It has identified a set of 106 components of income to obtain reliable estimates for total income, of which 36 are considered essential. ICE 360 surveys try and measure about 56 items of income that are suggested in the recommendations of Canberra Group. For example, in the ICE 360 surveys, a hierarchy of components of income is built up which provides definitions of total disposable household income. The major components of income covered in the survey are income from regular salary/wages, income from self-employment in non-agriculture, income from wages (agricultural labour and casual labour), income from self-employment in agriculture (crop production, forestry, livestock, fisheries, etc), income from other sources such as rent (from leased out land and from providing accommodation and capital formation), interest dividends received, employer-based pensions. In addition, when paid in kind (example in grain), the value of that is also considered as income.
Also, the accounting period used for collection of income data and income distribution analysis is one year as per Canberra City Group recommendation, as well as the a household as a basic statistical unit, defines as a group of two or more persons living together in the same house and sharing common food or other arrangements for essential living).
Ask people what their household income is, and they will answer. But the errors either in distribution of income or in the absolute number are not known. Even with a single answer of respondent claimed income, the categorization of households usually remains the same as in lower income households are unlikely to report income that classifies them as upper income households. Hence customer sentiment or employment tracking type studies that only need to categorize respondents by income do not need more rigorous income measurement (however, they do need more rigorous sampling both in terms of geographic spread and occupational representation given how economic growth and development and political environment factors are).
However when data is needed on, say, absolute income , income growth, inequality in income distribution, or market potential sizing, it requires more rigorous methodology. Using income distributions from analogous countries begs the question of what an analogous country is to India where the heterogeneity between states on politics and policy are ever increasing and complicating the dynamics of income generation and distribution.
A common problem faced by such surveys is the under-statement of economic data (income, expenditure and savings) by the respondents. Based on the adopted concept of income in ICE 360 surveys (which includes wages, salaries, bonus, business, profession, farm income and other forms of labour income, pensions, rent, interest, and dividend), the ICE360 survey 2021 estimates of the aggregate income of Indian households is about 60.3 per cent of the total personal disposable income, as provided by the National Accounts Statistics (NAS) for entire country.
These differences in estimates can be attributed to the following factors. One, this survey did not cover some of the smaller states and union territories which account for about 4 per cent of the population. Two, according to the National Statistical Organisation (NSO), the household sector by definition comprises of individuals, non-government non-corporate enterprises of farm business and non-farm business like sole proprietorships and partnerships, and non-profit institutions. This survey, on the other hand, covers only households. Three, certain components of income are not perceived as income by the respondents and hence they get excluded from incomes reported in income surveys. Items like reimbursements for travel, medical and other such expenses are not reported correctly in this survey.
There are different and equally compelling points of view on this. One school of thought is that household income as measured by the survey is what should be reported. Just as NSS reports expenditure data exactly as captured from the survey.
The other school of thought is that the distribution of income can be taken from the survey and applied to the macro all India income number that is provided by the National Accounts Statistics. We refer to this as “adjusted income” in our data, when income is reported this way ICE 360 survey “adjusted income” data takes the distribution of income in each decile (or 10% slabs based on income) and applies this percentage distribution to the Personal Disposal Income as reported by NAS. This effectively pushes up the income number for everyone, and it is obvious that we are making assumptions that we have not captured total income across rich and poor households to the same tune.
Where absolute income numbers are mentioned in all ICE 360 survey outputs and reports, we ensure that a footnote states “adjusted” or “observed” survey estimates.
Popularly used classifications of Upper / Middle / Lower Middle etc income groups, defined in terms of absolute income, use income cut off to define groups with no stated logic for using these cut offs. Also popular are labels of strivers / aspirers / rich and so on, which have the same issue. For all ICE 360 analysis and reporting, we have decided not to use any labels but to use income percentiles so that income groups with different characteristics emerge through the data emerge through of population, quintiles usually, deciles sometimes, of households.
We have found that income quintiles (20% slabs) offer very good ‘natural’ thresholds for analysing consumption and we mostly report data based on that. For certain analyses, we do this “quintiling” separately for urban India and separately for rural India.