MacroScan (Printable Version) - The Calorie Consumption Puzzle

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 55^th 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.



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 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.



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.





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.

So is it the case, then, that the decline in per capita calorie consumption is simply a positive reflection of development in various forms? Unfortunately, such an optimistic conclusion cannot be arrived at so easily, and may probably be misleading. While very rigid norms may not be successful at allowing for differential requirements of individuals, there are other ways to estimate a possible nutritional deficit. One way is in terms of defining a reference "consumer unit" and then applying conversion factors to all the population according to age and gender.

This is what the NSS has been doing with its definition of a consumer unit at 2700 calories, which it has taken as the daily calorific requirement of a normal male person doing sedentary work and belonging to the age group 20-39 years. Other persons are assigned conversion factors based on age and gender, so that each household can then be defined in terms of the number of consumer units, which would be less than the number of household members.

This provides some indication of aggregate nutritional deficiency. Calorie consumption at less than 90 per cent would indicate nutritional deficiency. There are also suggestions of the prevalence of far greater nutritional deficiency at less than 70 per cent of the consumer unit norm. Nevertheless, here we consider only the broader estimate.

Chart 6 provides estimates of how such nutritional deficiency has moved over time. The results are quite stark. In the rural areas, the percentage of population with less than 90 per cent of the norm of calorie intake increased from 40 per cent in 1983 to 45 per cent in 1999-2000. And if the sense that the latest period survey overestimated food consumption is correct, the ratio is likely to be even higher. In the urban areas, the proportion of nutritionally deficient population declined to around 48 per cent by the early 1990s, but appears to have remained at that level thereafter.



Of course, this all-India figure hides very substantial regional variation. Charts 7 and 8 indicate just how marked such variation is across states, in terms of average per capita calorie consumption across the rural and urban areas of the major states. The lowest rural average nutrient intake is to be found in the states of Assam, Gujarat, Kerala, Maharashtra and Tamil Nadu. Note that these are not (except Assam) states with above-average incidence of rural poverty, and indeed have higher than average per capita income. Furthermore, in several of these states (except Kerala and Maharashtra) the average intake worsened over the 1990s.





Another state with relatively poor nutrient intake is Andhra Pradesh, in which also the situation appears to have worsened over time. This is especially surprising in view of the fairly extensive and highly subsidised Public Distribution System that was set up by the state government, although the reduction in food subsidy over the 1990s may have played some role in the worsening intake in rural areas in that state. Karnataka also indicates low average calorie intake in the rural areas.

It is of course possible that such state-wise variations are related to differing physiological requirements and cultural habits. However, another factor may also be at work: the deficiency of cereals production in these states. The Report of the High-Level Committee on Long-Term Grain Policy (July 2002) makes the following point: "Along with Bihar, these states have the lowest per capita cereals production among all the major states. This not only emphasises the continued importance of cereals even in relatively rich states, but also the limitations of trade.

A production deficit by itself does not imply food insecurity but involved additional costs and effort of getting supplies from elsewhere. Deficit production and relatively long distances from surplus regions lead to relatively high prices, e.g. NSS implicit prices of cereals purchased were 30-70 per cent higher in the South Indian states of Karnataka, Kerala and Tamil Nadu than in Uttar Pradesh. This restrains cereal consumption and keeps total nutrient intake even below that in the poorest states.” (page 120).

Two important points are made here: the relevance of local production in states, and the role of price movements of food, and cereals in particular. This latter factor may be a key element in explaining the calorie consumption puzzle, especially in the 1990s, since the decline in calorie intake is closely related to the fall in cereals consumption.

This was a decade marked by rising relative prices of cereals, all over India. While the NSS data show that cereals consumption has declined and that of non-cereals has increased, it is important to note that the share of cereals in food expenditure has not fallen. The fact that cereals prices increased faster than other food prices may have prompted a shift towards other kinds of food to the extent possible, and caused the decline in actual cereal intake (and therefore calorie intake). It should be remembered that cereals still account for half the food expenditure of the poor, and its share in household food budgets remains largely unchanged.

Tables 1 and 2, which present data on the changes in per capita calorie consumption in the rural and urban areas of the major states, bring to light further puzzles. The general decline in rural per capita calorie consumption is not universal; some states like Orissa (one of the poorest states) and Maharashtra actually show improvement after 1983. The decline is very sharp in rural Haryana and Punjab, which are richer states, but also in Madhya Pradesh and Rajasthan, which are poorer states. All these issues deserve much more careful investigation and analysis.





In urban areas, by contrast, average calorie consumption appears to have increased after 1983, in most states. The states where it has declined (such as Haryana, Karnataka and Kerala) are not distinguished by any common feature that could explain the decline.

However, in terms of extent of nutrition deficiency, the trend appears to be much more definitive. As Table 3 shows, across the major states of India, the proportion of people with nutritional intake less than 90 per cent of the norm, increased, especially after 1993. In some cases, as in West Bengal, this reflected an increase after a decade of decline, such that the 1999-2000 figure was still below that of 1983. But in most other states, the ratio of nutritionally deficient population kept increasing. In some states this increase was quite sharp, including in Andhra Pradesh.



In urban areas, the picture is more mixed. For the decade 1983-93, there appears to have been a decline, but substantially the proportion of nutritionally deficient population has stagnated at fairly high levels.

Another way to estimate the extent of adequacy of nutrition is to estimate chronic food deficiency based on body-mass indicators. Such an estimate for the major states, based on work done by the M.S. Swaminathan Foundation, is represented in Chart 9. This indicates very high levels of chronic energy deficiency of 50 per cent or more in many of the major and most populous states. The outlier here is Assam, for unknown reasons.



All in all, the picture that emerges throws up more questions than can be answered here, especially in terms of the trend and regional pattern of calorie consumption. But the basic conclusion that must be drawn is that the current state of nutritional intake in the country is quite appalling, and needs immediate policy attention to ensure adequate access to food to people across the country.



Themes > Features	11.02.2003


The Calorie Consumption Puzzle

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 55^th 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. 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 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. 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. 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. So is it the case, then, that the decline in per capita calorie consumption is simply a positive reflection of development in various forms? Unfortunately, such an optimistic conclusion cannot be arrived at so easily, and may probably be misleading. While very rigid norms may not be successful at allowing for differential requirements of individuals, there are other ways to estimate a possible nutritional deficit. One way is in terms of defining a reference "consumer unit" and then applying conversion factors to all the population according to age and gender. This is what the NSS has been doing with its definition of a consumer unit at 2700 calories, which it has taken as the daily calorific requirement of a normal male person doing sedentary work and belonging to the age group 20-39 years. Other persons are assigned conversion factors based on age and gender, so that each household can then be defined in terms of the number of consumer units, which would be less than the number of household members. This provides some indication of aggregate nutritional deficiency. Calorie consumption at less than 90 per cent would indicate nutritional deficiency. There are also suggestions of the prevalence of far greater nutritional deficiency at less than 70 per cent of the consumer unit norm. Nevertheless, here we consider only the broader estimate. Chart 6 provides estimates of how such nutritional deficiency has moved over time. The results are quite stark. In the rural areas, the percentage of population with less than 90 per cent of the norm of calorie intake increased from 40 per cent in 1983 to 45 per cent in 1999-2000. And if the sense that the latest period survey overestimated food consumption is correct, the ratio is likely to be even higher. In the urban areas, the proportion of nutritionally deficient population declined to around 48 per cent by the early 1990s, but appears to have remained at that level thereafter. Of course, this all-India figure hides very substantial regional variation. Charts 7 and 8 indicate just how marked such variation is across states, in terms of average per capita calorie consumption across the rural and urban areas of the major states. The lowest rural average nutrient intake is to be found in the states of Assam, Gujarat, Kerala, Maharashtra and Tamil Nadu. Note that these are not (except Assam) states with above-average incidence of rural poverty, and indeed have higher than average per capita income. Furthermore, in several of these states (except Kerala and Maharashtra) the average intake worsened over the 1990s. Another state with relatively poor nutrient intake is Andhra Pradesh, in which also the situation appears to have worsened over time. This is especially surprising in view of the fairly extensive and highly subsidised Public Distribution System that was set up by the state government, although the reduction in food subsidy over the 1990s may have played some role in the worsening intake in rural areas in that state. Karnataka also indicates low average calorie intake in the rural areas. It is of course possible that such state-wise variations are related to differing physiological requirements and cultural habits. However, another factor may also be at work: the deficiency of cereals production in these states. The Report of the High-Level Committee on Long-Term Grain Policy (July 2002) makes the following point: "Along with Bihar, these states have the lowest per capita cereals production among all the major states. This not only emphasises the continued importance of cereals even in relatively rich states, but also the limitations of trade. A production deficit by itself does not imply food insecurity but involved additional costs and effort of getting supplies from elsewhere. Deficit production and relatively long distances from surplus regions lead to relatively high prices, e.g. NSS implicit prices of cereals purchased were 30-70 per cent higher in the South Indian states of Karnataka, Kerala and Tamil Nadu than in Uttar Pradesh. This restrains cereal consumption and keeps total nutrient intake even below that in the poorest states.” (page 120). Two important points are made here: the relevance of local production in states, and the role of price movements of food, and cereals in particular. This latter factor may be a key element in explaining the calorie consumption puzzle, especially in the 1990s, since the decline in calorie intake is closely related to the fall in cereals consumption. This was a decade marked by rising relative prices of cereals, all over India. While the NSS data show that cereals consumption has declined and that of non-cereals has increased, it is important to note that the share of cereals in food expenditure has not fallen. The fact that cereals prices increased faster than other food prices may have prompted a shift towards other kinds of food to the extent possible, and caused the decline in actual cereal intake (and therefore calorie intake). It should be remembered that cereals still account for half the food expenditure of the poor, and its share in household food budgets remains largely unchanged. Tables 1 and 2, which present data on the changes in per capita calorie consumption in the rural and urban areas of the major states, bring to light further puzzles. The general decline in rural per capita calorie consumption is not universal; some states like Orissa (one of the poorest states) and Maharashtra actually show improvement after 1983. The decline is very sharp in rural Haryana and Punjab, which are richer states, but also in Madhya Pradesh and Rajasthan, which are poorer states. All these issues deserve much more careful investigation and analysis. In urban areas, by contrast, average calorie consumption appears to have increased after 1983, in most states. The states where it has declined (such as Haryana, Karnataka and Kerala) are not distinguished by any common feature that could explain the decline. However, in terms of extent of nutrition deficiency, the trend appears to be much more definitive. As Table 3 shows, across the major states of India, the proportion of people with nutritional intake less than 90 per cent of the norm, increased, especially after 1993. In some cases, as in West Bengal, this reflected an increase after a decade of decline, such that the 1999-2000 figure was still below that of 1983. But in most other states, the ratio of nutritionally deficient population kept increasing. In some states this increase was quite sharp, including in Andhra Pradesh. In urban areas, the picture is more mixed. For the decade 1983-93, there appears to have been a decline, but substantially the proportion of nutritionally deficient population has stagnated at fairly high levels. Another way to estimate the extent of adequacy of nutrition is to estimate chronic food deficiency based on body-mass indicators. Such an estimate for the major states, based on work done by the M.S. Swaminathan Foundation, is represented in Chart 9. This indicates very high levels of chronic energy deficiency of 50 per cent or more in many of the major and most populous states. The outlier here is Assam, for unknown reasons. All in all, the picture that emerges throws up more questions than can be answered here, especially in terms of the trend and regional pattern of calorie consumption. But the basic conclusion that must be drawn is that the current state of nutritional intake in the country is quite appalling, and needs immediate policy attention to ensure adequate access to food to people across the country.



© MACROSCAN 2003