In the last year, there have been several peasant protests in different parts of the country, a reflection of low incomes and distress among the peasantry. We now have a new source of data on the extent of savings and investment of rural households. The recently published NABARD National Financial Inclusion Survey or NAFIS 2016-17, based on a sample of 40,327 households, has data on income, expenditure, saving, investment and asset ownership.

According to NAFIS, on average, a rural household had a monthly income of Rs 8,059 in the survey year (the figure was Rs 8,931 for an agricultural household). The monthly consumption expenditure was Rs 6,646 and Rs 7,152 for all and agricultural households respectively. In short, excess of income over expenditure was positive and Rs 1,413 a month for all households and Rs 1,779 a month for agriculture households. If true, these numbers are remarkable, as they suggest positive savings for rural households including agricultural households on average. The main point of this Note is to argue that this picture is flawed — it is very unlikely that rural households in distress, particularly those in the lower income deciles, have positive savings.

There are three possible reasons why these numbers are not right. First, as I argued in my earlier blog, income, particularly crop income is likely to be overestimated in NAFIS as it does not take account of all costs of production. Further, expenditure is likely to be underestimated as the questions on consumption are much fewer (less detail) than in the NSS Consumer Expenditure Survey. Together, this suggests that the excess of income over expenditure, that is, savings, is likely to be overestimated.

The second argument is based on data on decile-wise income and expenditure. Among poor households, we expect expenditure to exceed income, that is, for a situation of negative savings or dissaving. In other words, a poor household is likely to borrow money to meet its expenditure or consumption needs. This is the pattern observed internationally. Surprisingly, the NAFIS data show that income is higher than consumption/expenditure for all deciles except the third and fourth decile. We definitely expect dissaving in the lowest two deciles (the poorest of the poor). So, again, this pattern suggests problems with the data.

The third point is the finding on investment. As per NAFIS, only 10 per cent of rural households undertook any investment during the reference period (one year). And, of those who made an investment, only 40 per cent of the funds for that investment came from their own funds. This result does not square up with the earlier findings on positive savings and is probably a truer reflection of the extent of household savings.

Additionally, the section on savings has a major definitional problem. The Report begins by stating that savings is income minus expenditure. However, it later defines “any money deposited with bank, post office, SHGs, chit funds or money kept aside for emergencies even at home” as savings. This is erroneous. A household deposits money in the bank, say after the harvest, and withdraws it two months later to say, pay the school fees. This money in the bank if only adequate or less than adequate to meet the school fees is not savings (it is not excess of income over required expenditure).The Report states that 47 to 49 per cent of households in the bottom four deciles reported savings during the last year. This is not so. These numbers only reflect the fact that many households now have bank accounts and do put cash in the bank in a “savings account” during the course of a year. This does not imply that these 47 per cent of the poorest households in rural India had savings in the economic sense during the year. The results on “savings” as already mentioned do not square up with the findings on investment.

Major changes need to be made to the questionnaire and approach of NAFIS if the next round is to help us with concrete evidence on savings and investment of rural households.