E

This paper provides empirical evidence on the nature and extent of long-term changes in consumption patterns and nutritional status of various socio-economic groups at the household level in rural and urban India as well as in some of the poorer states. It also takes stock of the changing consumption patterns of the poorest strata of population in rural and urban India. It is hypothesised that the dietary changes during the last few years have improved the food and nutritional status at the household level in both rural and urban areas for all the socio-economic classes and states of India.

The unit data on dietary patterns and consumer expenditures collected by the National Sample Survey Organisation (NSSO) at the national level was used for this study. The sample covered 71,385 rural and 48,924 urban households, making a total of 1,20,309 households. The analysis was based on the 38th and 55th rounds of the National Sample Survey (NSS) and pertained to the years 1983 (January to December) and 1999-2000 (July to June), respectively.

The sample households were grouped into poor (bottom) and non-poor (upper) classes. The non-poor class comprised households which were above 150 per cent of the poverty line, whereas, the poor class consisted of households below the poverty line. The poverty line for rural and urban areas in different states corresponding to various NSS rounds, as defined and adopted by the Planning Commission, was used in the study (Appendix I). The per capita expenditure of the sample households was used as a proxy for per capita income. Prices of food items were computed by dividing the expenditure by the quantity consumed. The changes in consumption pattern were assessed using the indicators of quantity of food items, budgetary share and calorie intake. The food items included rice, wheat, coarse cereals, pulses, edible oils, vegetables, fruits, milk, meat, fish, eggs and sugar. Budgetary share referred to the expenditure on individual food items in the total food expenditures. The calorie intakes referred to the calories from different food items and were calculated using conversion factors provided by NSSO (Report No 405). The minimum (threshold) food-energy requirement was taken as 1,800 kcal/person/day for rural households and 1,575 kcal for the urban households. It represented 75 per cent of the recommended values, 2,400 kcal/person/day for rural and 2,100 kcal/person/day for urban [for more information, see Dandekar 1996]. An intake below this threshold was considered not sufficient for maintaining health and body weight and carrying out light physical activity. The threshold level of food-protein intake was taken as 48 g/person/day for an average Indian. The households consuming below this level were treated as malnourished. The threshold level for fat was taken as 16 g/person/day.

The per capita annual consumption of different food items, in 1983 and 1999-2000 by different income groups and changes

Economic and Political Weekly September 1, 2007 therein are presented in Table 1, a perusal of which reveals two types of effects. (i) Changes in consumption pattern of a group (poor or non-poor) over time, which have been termed as “structural shift” on account of “consumption diversification effect”. This change is the result of easier access to supply, changes in tastes and preferences, and changing relative prices,

State Rural Urban Bottom Group Upper Group Bottom Group Upper Group

Gujarat 61.8 38.4 19.8 7.1 Karnataka 58.9 33.5 37.8 15.4 Maharashtra 46.8 32.8 21.4 5.7 Rajasthan 39.2 23.5 6.2 3.3 Andhra Pradesh 15.6 4.3 3.9 0.6 Himachal Pradesh 18.8 15.4 4.7 0.7 Madhya Pradesh 15.7 3.8 2.4 0.9 Tamil Nadu 7.9 1.7 1.8 0.3 Bihar 4.5 1.7 0.1 Negligible Orissa 4.0 0.9 1.3 0.5 Uttar Pradesh 2.6 1.6 0.6 0.4 Haryana 4.7 3.1 0.7 0.5 India 10.0 8.8 5.4 3.7

Note: Bottom group – Below poverty line; Upper group – Above 150 per cent of poverty line. States with negligible share are not included.

Source: Chand and Kumar (2002).

Year Bottom Group Upper Group Rural Urban All India Rural Urban All India

Calories 1983 56 51 54 9 8 9 1999-2000 64 46 58 12 5 9 Protein 1983 51 64 55 9 20 13 1999-2000 65 65 65 14 14 14 Fats 1983 61 40 55 10 4 8 1999-2000 48 16 36 4 2 3

Notes: Bottom group : Below poverty line. Upper group : Above 150 per cent of poverty line.

Source: Authors’ calculations using data from National Sample Survey Organisation.

State 1983 1999-2000 Rural Urban Rural Urban

Andhra Pradesh 72.66 106.43 262.94 457.40 Assam 98.32 97.51 365.43 343.99 Bihar 97.48 111.80 333.07 379.78 Gujarat 83.29 123.22 318.94 474.41 Haryana 88.57 103.48 362.81 420.20 Himachal Pradesh 88.57 102.26 367.45 420.20 Jammu and Kashmir 91.75 99.62 367.45 420.20 Karnataka 83.31 120.19 309.59 511.44 Kerala 99.35 122.64 374.79 477.06 Madhya Pradesh 83.59 122.82 311.34 481.65 Maharashtra 88.24 126.47 318.63 539.71 Orissa 106.28 124.81 323.92 473.12 Punjab 88.57 101.35 362.68 388.15 Rajasthan 80.24 113.35 344.03 465.92 Tamil Nadu 96.15 120.30 307.64 475.60 Uttar Pradesh 83.85 110.23 336.88 416.29 West Bengal 105.55 105.91 350.17 409.22 Delhi 88.57 123.29 362.68 505.45

Source: Poverty line of Assam was used for Sikkim, Arunachal Pradesh, Meghalaya, Mizoram, Manipur, Nagaland and Tripura. Poverty line of Maharashtra was used for Goa and Daman and Diu. Poverty line of Tamil Nadu was used for Pondicherry and Andaman and Nicobar Island. Urban poverty line of Punjab was used for both rural and urban poverty of Chandigarh. Poverty line of Maharashtra was used for Dadra and

India for the period 1983-2000 (Table 2) indicates that the dietary pattern is converging and becoming similar in nature. The per capita annual consumption of cereals during this period declined from 182 to 157 kg in rural areas and from 142 to 131 kg in urban areas while that of fruits and vegetables increased from 49 to 84 kg in rural areas and from 77 to 115 kg in urban areas. There have been considerable changes in consumption of different food items of animal origin. For example, per capita annual milk consumption increased from 37 to 63 kg in rural areas and from 56 to 91 kg in urban areas. The consumption of meat, fish and eggs has also shown an increasing trend, though the per capita annual consumption was still on the lower side

(6.7 kg in rural and 9.5 kg in urban population); it could be due to non-consumption of these items by a considerable proportion (41 per cent) of inhabitants. The annual per capita consumption of livestock products in the year 1999-2000 was estimated to be 5.3 kg for meat, 5.9 kg for fish and 30 (number) eggs [Birthal and Kumar 2002].

The rural consumers spent 61.5 per cent of their income on food items compared to 50.7 per cent by urban consumers in

Note: Bottom group – Below poverty line, Upper group - Above 150 per cent of poverty line.

Source: Authors’ calculations using data from National Sample Survey Organisation.

Economic and Political Weekly September 1, 2007 the year 1999-2000 (Table 3). The corresponding figures in 1983 were 69.1 per cent and 59.1 per cent. It indicates a decline of about 7 to 8 per cent on food budget during the period 198399 for rural as well as urban households. This shows the improvement in the social welfare of the consumers. The increase in per capita calorie intake during this period was substantial in both groups, though much higher in urban areas (468 kcal/ person/day) as compared to rural areas (127 kcal/person/day). This higher calorie intake was being met through consumption of non-cereals and livestock commodities.

cent) and Uttar Pradesh (31.2 per cent). The changes in the food basket between 1983 and 1999-2000 for the bottom and upper income groups in these states of India are presented in Appendix IV. It was found that the poorer Indian states spent a larger portion of their budget on food as compared to other states. The bottom group spent a higher proportion of its budget on lowcalorie foods, while upper group did so on high-calorie food items. The results do corroborate that in the poorer states also, diets are diversifying towards horticultural and livestock food items, particularly in the upper income households. A similar pattern was observed for the agricultural labourers and small farmers (Appendix Table 5).

A pertinent question in the context of dietary transition was how these changes were affecting the energy-nutrition balance, particularly of the poor. The calorie intakes from different food commodities by both expenditure groups in rural and urban areas in the years 1983 and 1999-2000 are recorded in Appendix Table 6. The cereals were the major suppliers of calories and non-cereals like pulses, edible oils, horticultural products, animal and fishery products were major providers of proteins, fats, vitamins, minerals, etc. Closely following the trends described earlier, the cereals accounted for 76.7 per cent of the total calorie intake in the rural poor and 68.3 per cent in the urban poor in the year 1999-2000. More significantly, the bottom income group derived more than three-fourths of its total food intake from cereals. The upper income group consumed more than half

(57.7 per cent) of the total calories intake from cereals. The decline in calorie intake from cereals on account of decrease in consumption was compensated by a marked increase in intake of calories from milk, vegetables, fruits, meat, sugar, etc. It has been concluded that as income rises, households generally diversify their food consumption pattern by shifting towards highvalue and high-quality food items.

It may, however, be necessary to examine the distribution aspects of calorie consumption, as it may be misleading at the aggregate level. For this purpose, the population below the threshold levels of calories, proteins and fat was estimated (Table 5). Table 5 reveals a general worsening of calorie and protein consumption, particularly in the bottom income group in rural areas, where the population below threshold level had increased from 56 per cent in 1983 to 64 per cent in 1999 in terms of calories intake, and from 51 per cent to 65 per cent in terms of protein intake. Ideally, the source of protein should be pulses and meat. But the data showed that cereals contributed 67 per cent of the protein source in rural India. With lesser consumption of cereals and more of non-cereals, the situation was expected to improve. This, however, needs a detailed empirical analysis. But in the case of fat, the situation had improved in all cases. For example, in the bottom income group in rural areas, the percentage of population below threshold level consuming fat decreased from 61 in 1983 to 48 in 1999, indicating some imbalances in food habits. Thus, at the disaggregate level one can see the worsening of nutritional situation with the dietary transition, particularly in the rural areas. It can perhaps be explained in terms of lack of purchasing power for procuring adequate quantity of high-value non-cereal commodities to compensate for loss in nutrition owing to replacement of cereals.

Increase in income, urbanisation, and consumer perceptions regarding food quality and safety are effecting changes in the food-consumption pattern and these are pervasive changes. Food consumption pattern is diversifying towards vegetable oils,

Item 1983 1999-2000 Annual Growth (Per Cent)

Agricultural labourers Cereals 164.1 148.5 -0.7 Pulses 8.1 8.5 -0.0 Edible oils 2.7 5.1 3.8 Vegetables and fruits 39.0 72.6 3.9 Milk equivalent 16.7 28.9 3.1 Meat, fish and eggs 3.3 4.9 2.6 Sugar 8.0 7.9 -0.3 Marginal farmers (< 1 ha) Cereals 176.4 161.0 -0.7 Pulses 9.0 11.8 1.2 Edible oils 3.1 9.8 6.7 Vegetables and fruits 50.6 86.9 3.3 Milk equivalent 25.3 59.2 4.8 Meat, fish and eggs 4.8 7.3 2.3 Sugar 7.9 14.5 3.3 Small farmers (1-2 ha) Cereals 187.9 165.2 -0.9 Pulses 11.4 12.3 -0.1 Edible oil 3.4 6.8 3.9 Vegetables and fruits 51.8 83.1 2.9 Milk equivalent 37.6 103.4 5.8 Meat, fish and eggs 3.7 4.5 1.1 Sugar 10.0 15.2 2.1

Source: Authors’ calculations using data from National Sample Survey Organisation.

Commodity Bottom income Group Upper Income Group 1983 1999-Change, 1983 1999-Change 2000 Per Cent 2000 Per Cent

Rural Cereals 83.6 76.7 -8 70.9 57.7 -19 Pulses 3.9 3.6 -8 5.8 5.1 -12 Edible oils 3.0 5.9 +97 5.0 9.3 +86 Vegetables and fruits 3.6 5.3 +47 3.6 5.8 +61 Milk equivalent 1.9 2.5 +32 7.2 10.4 +44 Meat, fish and eggs 0.47 0.62 +32 0.7 1.1 +57 Sugar 3.4 3.6 +6 7.2 6.9 -4

Urban Cereals 75.8 68.3 -10 59.7 44.0 -26 Pulses 5.1 4.4 -14 6.7 6.1 -9 Edible oils 5.8 7.9 +37 9.3 15.3 +64 Vegetables and fruits 3.4 5.2 +53 4.8 6.4 +33 Milk equivalent 3.9 4.8 +23 10.8 12.4 +15 Meat, fish and eggs 0.15 0.88 +487 0.33 1.38 +318 Sugar 5.5 5.3 -2 7.4 7.6 +3

All India Cereals 81.4 73.7 -9 66.4 51.7 -22 Pulses 4.3 3.9 -9 6.1 5.6 -8 Edible oils 3.8 6.6 +74 6.5 11.9 +83 Vegetables and fruits 3.5 5.3 +51 4.0 6.0 +50 Milk equivalent 2.4 3.3 +37 8.5 11.3 +33 Meat, fish and eggs 0.38 0.71 +87 0.60 1.21 +102 Sugar 4.0 4.2 +5 7.3 7.2 -1

Note: Bottom group – Below poverty line; Upper group – Above 150 per cent of poverty line.

Source: Authors’ calculations using data from National Sample Survey Organisation.

Economic and Political Weekly September 1, 2007 fruits, vegetables, milk and eggs. With increasing urbanisation, reducing poverty and rising disposable income, the demand for these items is expected to increase. Singh et al (2002) have estimated that the demand by the end of the Tenth Five-Year Plan of India (i e, 2007) for vegetables would be 103-118 million tonnes (mt), fruits 58-66 mt, milk 91-101 mt, meat 6.5-7.7 mt, fish 7.3

8.5 mt, and eggs 57.5 -67.5 billions. In spite of a decline in the real prices and increase in income, the per capita consumption of cereals has declined during the last two decades. In particular, coarse cereals have been severely affected, though these still dominate the food basket in states like Gujarat, Rajasthan and Maharashtra in India. However, the impact of these changes in terms of nutrients, calories, vitamins, etc, is not clear. Some of the estimates in this paper have indicated that the nutritional status of rural people is becoming worse. Deeper insights and analytical investigations are needed to decipher this puzzle. These changes have significant bearing on demand/trade for food, agricultural diversification, market access, research priorities and resource allocations to achieve food and nutritional security. Public and private sectors would have to respond to these signals. These changes should be guided by suitable reforms in technologies, institutions and policies to accelerate the diversification process, and inter alia improve the food and nutritional security of the poorest of the poor.

EPW

Email: pkumariari@rediffmail.com

[The authors are grateful to P K Joshi and Ashok Gulati for their valuable comments and suggestions. Views expressed in the paper are those of the authors and not of the organisations they are working.]