Diets, Nutrition, and Poverty - The Oxford Topic of Food, Politics, and Society - page 333

conducted jointly by the University of Maryland and the National Council of Applied
Economic Research permits analysis and answers. Eating out refers to meals or snacks
served in restaurants, roadside eating places, tea and snack shops, and street vendors.
Our focus is on the socioeconomic status of households eating out, and their spatial
distribution.19
Eating out was more common among the metro residents, as one would expect; they
also spent larger amounts. Yet eating out is pervasive; about 30% of the households did
so. A large proportion of those eating out (about 42%) spent under rupees (Rs) 99 per
month, and about a quarter spent over Rs 200 per month (at 2004-2005 prices). Eating
out is a feature not just of the metros or urban areas, but also of urban slums and rural
areas, though it is less pervasive in the last two areas. In the six largest metros (Mumbai,
Delhi, Kolkata, Chennai, Bangalore, and Hyderabad), about 34% of the households ate
out, as compared to about 27% elsewhere. Over 47% of the former spent Rs 200 or more
per month on eating out, and less than one-quarter of the latter did so. There are also dif-
ferences by caste. About 25% of the Scheduled Castes (SCs), about 27% of the Scheduled
Tribes (STs), and about 31% of the Other Backward Castes (OBCs), and Others ate out.
Even some of the most deprived and socially excluded groups—especially the SCs and
STs—have switched from traditional staples to fast foods and opted for greater variety
in food consumption. When the sample is split into the poor and nonpoor households
using the official poverty line, we find that a much larger proportion of the nonpoor
households (about 32%) ate out, but about 12.5% of the poor ate out as well.20 A more dis-
aggregated classification of the households into four Monthly Per Capita Expenditure
(MPCE) classes (less than Rs 300, between Rs 300-500, between Rs 500-1000, and
greater than Rs 1000) further qualifies the common perception that eating out as a man-
ifestation of dietary transition is mostly a middle-class phenomenon.
Using an econometric model, we obtain additional insights into the marginal contri-
bution of household traits, their (relative) affluence, and locational characteristics.21 The
results show that location of households, their demographic and caste characteristics,
and, above all, their relative affluence determine both the decision to eat out, and, con-
ditional on it, the amount spent. Further details are omitted to focus better on the latter.
Amounts spent on eating out vary with location. Households located in both met-
ros and nonmetro urban locations are likely to spend larger amounts on eating out,
relative to rural areas. Between the metros and nonmetros, households in the former
are likely to spend much larger amounts. SCs, STs, and OBCs are likely to spend lower
amounts relative to Others. The higher the number of adult males in paid employment
in the age-group 25-45, and of females in the older age-group, >45 years, the greater is
the amount spent. The effect of higher per capita expenditure relative to the poverty
line is large and significant, confirming that the more affluent are not just more likely to
eat out but also likely to spend larger amounts. The higher the share of salary in house-
hold income, the lower is the amount spent. By contrast, the higher the share of business
income, the larger is the amount spent.
Our analysis thus broadly confirms the important role of urbanization, demographic
changes, expansion of middle class and its growing affluence in eating out, or, more