The Calorie Consumption Puzzle
Feb 11th 2003

One of the more startling, but less discussed, features of Indian development over the past several decades, is the decline in per capita calories consumption which is revealed by the official National Sample Surveys. Average calorie consumption in India was already low by international standards, and that it has actually declined despite apparently high aggregate economic growth rates is clearly something that merits much more attention.
Of course, the most recent data that we have on calories consumption, from the 55th Round of the NSS, is unfortunately not comparable with data emanating from the earlier rounds. This has been discussed at length in this column as well as by Abhijit Sen (2001, 2002) so we will not dwell on it further here. However, it does mean that, along with estimates of consumption expenditure, estimates of food consumption are likely to be overestimates when compared to the earlier rounds.
Nevertheless, even these (relatively) inflated data indicates a decline in per capita calorie consumption for rural India in 1999-2000, as shown in Chart 1. The more significant trend, of course, is the long-term decline since the early 1970s. In addition, there is what appears to be a convergence between rural and urban patterns of calorie consumption by the most recent period.

Chart 1 >> Click to Enlarge
Indeed, in 1999-2000, the estimates of per capita calorie consumption in urban India were higher than for rural India, at 2156 Kcalories per day compared to 2149. This is surprising given the perception that the rural population tends to consume more calories because of the greater intensity of work in rural areas.
The pattern in overall calories consumption is mirrored in per capita protein consumption, which is shown in Chart 2. Here also, rural areas have experienced a substantial decline from the early 1980s in particular. This decline has extended up to 1999-2000 even though the estimate for that period may be higher relative to the earlier rounds because of the change in pattern of questioning. Meanwhile, urban protein consumption appears to have increased, to the point of convergence.

Chart 2 >> Click to Enlarge
Chart 3, which describes trends in per capita fat consumption, actually indicate an increase over time for both rural and urban areas. This is part of a diversification in food consumption that does certainly appear to have occurred on average over the past decades. Overall, while cereal consumption seems to have fallen as a share of total calories, within the category there has been a shift from the so-called "inferior grains" to rice and wheat, especially among the poorer categories.

Chart 3 >> Click to Enlarge
The range of non-cereal foods has diversified, with the greater significance of milk products (which would contain more fat) as well as edible oils, along with fruits and vegetables. The current patterns of food consumption for rural and urban India are indicated in Charts 4 and 5. It should be remembered that these are the aggregate tendencies and that there are significant variations across expenditure classes.

Chart 4 >> Click to Enlarge

Chart 5 >> Click to Enlarge
There has been some discussion on how to interpret the long-term tendency towards declining per capita calorie consumption, especially in rural India. One argument that is frequently made is that the early NSS Rounds, including those of the 1970s and early 1980s, tended to overestimate calories (and especially food grain) consumption. This is then extended to argue that the subsequent estimates are mere corrections that give a closer approximation to reality.
Even if this were the case, there are still other questions to ponder. In rural India, it seems to be fairly clear that per capita calorie consumption fell over the period after the early 1980s, including in the period when all the estimates suggest (without controversy) that the incidence of absolute poverty was on the decline.
Some explanations of this trend have rested on the idea that this reflects a natural and positive change in dietary patterns, consequent upon the change in rural work patterns and life styles. Thus, Hanumantha Rao has argued that the increasing mechanisation of agricultural operations as well as the greater availability of mechanised transport has reduced the amount of manual labour and physical activity related to transport that is required in most of rural India. He suggests that this has meant a reduction in the biological requirement of energy.
Similarly, it has been argued that the increasing "urbanisation" of rural areas has meant that urban lifestyles have penetrated into rural areas, and have influenced the narrowing down of rural-urban differences in food consumption. This is used to explain the convergence in calorie consumption as well. However, it should be noted here that the convergence that appears at the aggregate level does not appear when the population is disaggregated according to expenditure classes. Indeed, the top 30 per cent in the urban areas appear to be consuming substantially more calories at the end of the period, even while the bottom 30 per cent do not show such an increase.
These arguments all relate to the notion of a "norm" of calorie consumption, which has itself come under question. Of course, a minimum calorie consumption line has been used in India since the 1950s to determine the food consumption necessary for survival, and this has then been used to derive "poverty lines" based on consumer expenditure patterns. But several decades ago, P. V. Sukhatme had argued that such rigid norms were not valid, since the human body has adaptation mechanisms with different metabolic properties for those with lower body weights. In other words, those who already weigh less could also require less calories per day, even to do similar kinds of work
In India the norm that has been conventionally used is that of 2400 calories per day (for a rural male engaged in moderate activity). The norm specified by the Food and Agriculture Organisation (FAO) for South Asia as a whole is much lower, at 2110 calories per day. The FAO has an even lower cut-off for its "lowest range of food requirement" of only 1810 calories per day.
The FAO has in fact brought down its calorie norms over time, reflecting factors such as those mentioned by Hanumantha Rao, as well as the reduced incidence of certain types of disease that also has reduced the need for the body to build up resistance to it. There are also social and cultural factors involved. This is even apparent in regional variations in India, where it is clear that the median requirement of calories in states like Gujurat or Tamil Nadu seems to be much lower then in Punjab or Haryana.

 | 1 | 2 | Next Page >>

Print this Page


Site optimised for 800 x 600 and above for Internet Explorer 5 and above