3. AQUACULTURE PRICE POLICY AND CONSUMPTION ANALYSIS

3.1 Credit for development of aquaculture and marketing of aquatic products

A sizeable part of the credit flow in the aquaculture sector comes from the private non-institutional sector, which is linked to marketing and production of aquatic products. Merchants provide financing for fishing operations in inland capture areas. The informal sector is highly flexible and also provides loans to fulfil consumption needs. Apart from marketing agents, professional moneylenders advance credit against securities of gold and agricultural properties. Whenever a fish farmer borrows money from a merchant, the farmer receives Rs 1-2 less per kg, compared to open market prices, even though the merchant does not charge interest on the loan.

During the late 1980s, institutional agencies shifted their emphasis from marine and inland capture to aquaculture. The National Bank for Agriculture and Rural Development (NABARD), as an apex organization, provides refinance options through commercial banks, cooperative banks and regional rural banks. A number of other agencies provide credit for projects in aquaculture, such as the Industrial Finance Corporation, Industrial Development Bank of India, State Finance Corporations and the National Cooperative Development Corporation. NABARD has been extending refinancing loans, and Andhra Pradesh has been in the forefront of receiving credit for refinancing purposes. One of many serious problems institutional lending agencies are facing is the fact that aquaculture has not officially been recognized as a land-based activity. Furthermore, seed-hatcheries are relatively few in number and hence, farmers are not assured supply of seed at appropriate times. Multiple controls over water tanks and reservoirs are causing confusion as to their legal ownership. Absence of long-term leases also makes it difficult for lending agencies to advance credit for capital investments. So far, NABARD has funded Rs 294 million for refinancing projects in fisheries, including aquaculture. However, no scientific estimate has been conducted to determine the actual amount of credit from informal sources. Credit extended by institutions for marketing aquatic products needs to be emphasized more. Fish is highly perishable, and the lack of cold storage and other processing facilities make farmers in aquaculture vulnerable to market intermediaries. Although fish farmers can withhold harvesting stock, timely credit could help them realize better prices.

Usually, the system of contract-farming with buy-back arrangements is only found in the corporate sector and among farmers with larger holdings. The National Cooperative Development Corporation (NCDC), a government of India Corporation, has supported an integrated scheme to develop inland fish markets through primary and state-level cooperative organizations. The NCDC has funded Rs 230 million to establish marketing networks, cold storage, retail markets, wholesale markets etc. in Karnataka state.

3.2 Price performance of products from aquaculture

This section presents an examination of price changes of aquatic products over the years. Many economists emphasize that right price is essential for providing incentives to expand output. In developing countries like India, prices may, at times, be volatile and may not be representing actual conditions of supply and demand. Poor farmers cannot afford to invest and specialize when faced with prices below the cost of production. The traditional marketing system does not regulate production to match future demand.

Analysis of price changes of some of the important species in aquaculture shows a relatively high increase in fish prices over a period of ten years. Prices increased the most for rohu from 1988 to 1999 in Haryana state. During 1988-94, the price for rohu increased by 126% and by 30% in 1999. Similarly, there has been a substantial rise in the price of catla, mrigal and other carps during the first five years of the period from 1988-89.

However, the rate of increase in prices slowed during the second half of the decade (i.e. during 1993-94 to 1998-99). We can see from the data that, on average, during the period 1993-94 to 1998-99, prices increased only around 10-15%. The cooling of prices from early 1990's levels can be attributed to a variety of reasons. First, volume of production has gone up significantly and, consequently, market prices have slumped. Second, due to recession and other changes in the economy, the rate of inflation has come down significantly. Inflation climbed to around 10-12% during the early 1990s and started to fall to 3-4% from 1995-96 onwards. During the same period, production-costs of fish per kg increased enormously and eroded the profitability of fish farming. Increased intensity of production resulted in higher production costs.

Table 10. Wholesale price changes of rohu (Rs/kg) during 1988-1999

States/Year	1988-89	1993-94	1998-99	% increase between 1988-89 and 1993-94	% increase between 1993-94 and 1998-99
Bihar	17.85	34.5	40	93.28	15.94
Orissa	16.57	34	41	105.19	20.59
West Bengal	21.31	38.5	42	80.67	9.09
Haryana	11.07	25	32.5	125.84	30.00
Himachal Pradesh	10.7	40.5	44	278.50	8.64
Jammu and Kashmir	15.5	37	41.5	138.71	12.16
Uttar Pradesh	15.6	39	44	150.00	12.82
Assam	21.9	45	50	105.48	11.11
Manipur	30.62	47.5	54	55.13	13.68
Andhra Pradesh	7	27.5	30.5	292.86	10.91
Karnataka	10	24	26.5	140.00	10.42
Gujarat	11.79	36	39.5	205.34	9.72
Madhya Pradesh	12.85	35.5	40.5	176.26	14.08
Rajasthan	14	39	43.5	178.57	11.54
All India	15.48	35.93	40.68	132.05	13.22

SOURCES: 1. Government of India (1998) for 1988-89 and 1993-94. 2. Field survey data for 1998-99.

State/Year	1988-89	1993-94	1998-99	% increase between 1988-89 and 1993-94	% increase between 1993-94 and 1998-99
Bihar	16.92	33	39	95.04	18.18
Orissa	16.14	34.5	38.5	113.75	11.59
West Bengal	20.97	38	41.5	81.21	9.21
Haryana	10.98	24.5	30	123.13	22.45
Himachal Pradesh	10.48	37	40.5	253.05	9.46
Jammu and Kashmir	15.53	36.5	40	135.03	9.59
Uttar Pradesh	15.34	34.5	41.5	124.90	20.29
Assam	21.07	34	49	61.37	44.12
Manipur	30.19	46.5	52.5	54.02	12.90
Andhra Pradesh	7.55	24.5	27	224.50	10.20
Karnataka	10	22	25	120.00	13.64
Gujarat	11.79	33.5	38	184.14	13.43
Madhya Pradesh	12.85	34.5	40	168.48	15.94
Rajasthan	12.62	36.5	41.5	189.22	13.70
All India	15.17	33.54	38.86	121.01	15.87

SOURCES: 1. Government of India (1998) for 1988-89 and 1993-94. 2. Field survey data for 1998-99.

State/Year	1988-89	1993-94	1998-99	% increase between 1988-89 and 1993-94	% increase between 1993-94 and 1998-99
Bihar	16.41	30	34.5	82.82	15.00
Orissa	14.83	32	37	115.78	15.63
West Bengal	19.35	33.5	38	73.13	13.43
Haryana	9.96	24	29.5	140.96	22.92
Himachal Pradesh	10	34	38	240.00	11.76
Jammu and Kashmir	14.38	33.5	38.5	132.96	14.93
Uttar Pradesh	15.29	37.5	40.5	145.26	8.00
Assam	19.85	44	47.5	121.66	7.95
Manipur	30.24	50	53	65.34	6.00
Andhra Pradesh	5.83	22	24	277.36	9.09
Karnataka	10	24	26	140.00	8.33
Gujarat	10	34	37.5	240.00	10.29
Madhya Pradesh	12.5	38	40	204.00	5.26
Rajasthan	13.24	36.5	40	175.68	9.59
All India	14.42	33.79	37.43	134.30	10.78

SOURCES: 1. Government of India (1998) for 1988-89 and 1993-94. 2. Field survey data for 1998-99.

State/Year	1988-89	1993-94	1998-99	% increase between 1988-89 and 1993-94	% increase between 1993-94 and 1998-99
Bihar	15.79	34.5	37	54.23	7.25
Orissa	6.84	30	34.5	77.20	15.00
West Bengal	16.17	33	38.5	51.00	16.67
Haryana	8.1	22	28.4	63.18	29.09
Himachal Pradesh	10	34.5	38	71.01	10.14
Jammu and Kashmir	13.42	30.5	34	56.00	11.48
Uttar Pradesh	14.62	32	38	54.31	18.75
Assam	19.46	38	40	48.79	5.26
Manipur	25.35	40	45	36.63	12.50
Andhra Pradesh	4.86	22	24	77.91	9.09
Karnataka	10	21	25	52.38	19.05
Gujarat	10	32	37	68.75	15.63
Madhya Pradesh	8.4	37	39.5	77.30	6.76
Rajasthan	10	36	38	72.22	5.56
All India	12.36	31.61	35.49	60.90	12.29

SOURCE: 1. Government of India for 1988-89 and 1993-94. 2. Field survey data for 1998-99.

Changes in consumer prices in major cities such as Calcutta and Delhi, are presented in the Tables 14 and 15. Variations in monthly retail prices for the selected cities show that the average monthly price for rohu and catla was higher during the winter season (November - January) and lower during the summer season. Consumption preferences also reveal that people eat more during the winter season. However, fluctuations and increases in retail prices are much less compared to those of producer prices. Average retail price in Calcutta increased from Rs 35/kg in 1988 to Rs 60 in 1996, indicating a 100% increase. Price analyses of markets in Delhi show that market prices were more constant between months, even though they rose substantially over the ten-year period. The producer's share in the consumer's price is around 50-60% for fresh fish. The data indicate that, although fish has a place in the market, the spread of around Rs 25-30 reflects marketing costs that are very high. The average retail price of rohu in markets in Calcutta was Rs 60 in 1996, and the producer/wholesale price in supplying Andhra Pradesh was Rs 30, which indicates that farmers received only 50% of the consumer price. If we consider the farm-gate price, a farmer's profit will be reduced even further. Inefficiency in the marketing of products from aquaculture is evident.

Table 14. Monthly retail price of rohu in Delhi during 1988-96 (Rs)

Month	88	89	90	91	92	93	94	95	96	Average	Price Spread	Producer's share in consumer price (percentage)
April	20	36	35	35	35	38	45	45	50	37.67	16.08	57
May	22	35	35	35	25	36	45	45	50	36.44	12.36	66
June	20	35	36	35	26	38	45	45	50	36.67	13.64	63
July	25	35	36	35	24	38	45	45	50	37.00	14.22	62
August	28	35	36	36	24	38	45	45	50	37.44	12.51	67
September	30	36	36	36	25	38	45	45	55	38.44	14.63	62
October	28	36	36	36	25	40	45	45	55	38.44	12.97	66
November	30	36	40	38	28	40	45	50	60	40.78	14.89	63
December	32	36	40	37	30	44	50	50	60	42.11	17.72	58
January	32	36	35	37	28	39	50	50	60	40.78	15.83	61
February	34	38	35	37.5	40	42.5	45	50	60	42.44	19.19	55
March	34	35	35	35	35	40	45	50	60	41.00	16.28	60

SOURCE: Government of India (1998).

Month/year	88	89	90	91	92	93	94	95	96	Average	Price spread	Producer's share in consumer price (percentage)
April	38	35	40	45	50	50	50	45	60	45.89	20.67	55
May	40	40	40	45	50	50	50	45	60	46.67	21.67	54
June	40	40	40	50	50	50	50	50	60	47.78	22.50	53
July	40	40	40	50	50	50	50	50	60	47.78	21.72	55
August	40	40	40	45	50	50	50	60	60	48.33	22.17	54
September	38	40	45	45	50	50	50	50	60	47.56	21.39	55
October	35	35	40	50	50	50	50	50	60	46.67	19.72	58
November	35	40	45	50	40	45	50	50	60	46.11	18.67	60
December	35	40	40	50	40	45	50	60	60	46.67	21.17	55
January	35	40	40	50	45	45	45	55	60	46.11	20.00	57
February	35	40	40	50	45	50	50	50	60	46.67	19.06	59

SOURCE: Source: Government of India (1998).

Observations of consumer prices and price-spread over the years suggest that the adverse effect of an increase in input prices has not been compensated for, as marketing margins remained high throughout the years.

Input prices advanced during 1990-98 to the extent of 175%. However, the price increase during the corresponding period amounted to only 10-15%, which has marginalized the profitability of commercial fish farming.

3.4 Household expenditure on food, fish and elasticities

According to a study conducted in 1976 by the National Council for Applied Economic Research (NCAER, 1980), 59% of households in Bangalore, 88% in Calcutta and 45% in Delhi consume fish. The proportion of household expenditure on the consumption of fish, out of the total expenditure on food, was 7.6% in Bangalore, 14.6% for Calcutta and 6.3% in Delhi. Average monthly consumption of fish per household in 1976 amounted to 3.86 kg for Bangalore, 5.56 kg for Calcutta and 3.02 kg for Delhi. Per caput monthly consumption of fish was 0.76 kg, 1.01 kg and 0.56 kg for Bangalore, Calcutta and Delhi respectively (NCAER, 1980). The NCAER study is a benchmark exhaustive study of fish consumption patterns in some of the metropolitan cities of India. However, newer and similar studies for these mega-cities are not available. A recent study on fish consumption patterns in Coimbatore, another major city in southern India, revealed that average fish consumption increases with rising income. On average, urban consumers buy around 4.5 kg of fish per month and spend around 7.5% of their total food expenditure on fish (Sekar et. al., 1996).

Table 16. Estimated per caput monthly expenditure as a percentage of total food expenditure by states

State/Region	Food Expenditure		Total non food expenditure	Total expenditure
	Food grains	Total food
North:
Haryana	16.9	60.2	39.8	100
Himachal Pradesh	25.3	60.2	39.7	100
Punjab	15.8	57.8	42.2	100
Upper Central
Bihar	43.3	69.5	30.5	100
Uttar Pradesh	29.5	62.9	37.1	100
Lower Central
Madhya Pradesh	34.6	69.5	38.1	100
Orissa	42.5	68.9	31.1	100
Rajasthan	19.7	59.4	40.6	100
East:
North Eastern states	39.9	71.7	28.3	100
West Bengal	38.9	70.4	29.6	100
West:
Gujarat	23.7	65.9	34.1	100
Maharastra	38.9	70.4	29.6	100
South:
Andhra Pradesh	29.2	59.2	40.8	100
Karnataka	32.5	65.6	34.4	100
Kerala	22.5	62.5	37.8	100
Tamil Nadu	32.2	64.9	35.1	100
All India	30.5	63.9	37.41	100

Table 16 presents estimated total food expenditure per caput per month at current prices. Mean per caput monthly household expenditure (MPCE) for rural India amounted to Rs 287 in 1998. This expenditure ranged from as high as Rs 554 for Punjab to a low of Rs 208 in Madhya Pradesh and Rs 210 in Orissa. Expenditure on food alone is about 31%. This proportion is low in Haryana, Punjab and Rajasthan. On the other hand, expenditure on food is as high as 43.3% and 42.7% of total expenditure in Bihar and Orissa respectively, states which are relatively poor.

The per caput consumption of fish varies marginally with income, as shown in Table 17. Lower consumption levels could also be due to geographic and cultural factors.

Table 17. Household income, consumption and prices

Region/State	Household income (Rs/year)	Per caput income (Rs/year)	Per caput fish consumption (kg/month)	Average fish price (Rs/kg)
North:
Haryana	39 956	6 368	0.02	32
Himachal Pradesh	23 973	4 168	0.02	45
Punjab	37 418	6 380	0.02	50
Upper Central:
Bihar	22 459	3 691	0.13	38
Uttar Pradesh	26 733	4 185	0.02	45
Lower Central
Madly Pradesh	25 319	4 166	0.05	40
Orissa	17 208	3 028	0.32	40
Rajasthan	27 184	4 229	0.01	45
East:
NE states	28 160	5 070	-	-
West Bengal	18 113	3 157	0.66	40
West:
Gujarat	29 356	5 288	0.02	40
Maharastra	29 929	5 525	0.14	40
South:
Andhra Pradesh	24 776	5 046	0.08	30
Karnataka	27 372	4 769	0.07	26
Kerala	29 101	5 778	1.59	35
Tamil Nadu	23 271	5 122	0.11	35
All India	25 653	4 485	0.17	38.73

Table 17 shows that states such as Bihar, with lower per caput income, consume more fish per caput per month compared to Punjab, which enjoys a higher per caput income. On the other hand, price does not seem to play a great role in per caput fish consumption. At the macro-level, fish consumption does not seem to be linked primarily to income, but rather to consumers' cultural and traditional food habits.

For example, fish consumption in Punjab and Haryana is relatively low in spite of higher production and income levels and the fact that migrant labourers of Uttar Pradesh and Bihar, as well as lower income groups, mostly consume fish. On the other hand, some of the micro-level studies tend to support Engel's law of consumption.

Table 18. Fish consumption pattern in Coimbatore city during 1997-98

Income group (Rs)	Avg house hold size	Annual expenditure on food (Rs)	Expenditure on non-food items (Rs)	Total expenditure (Rs)	Expenditure on fish (Rs)	Percentage expenditure on fish to total food expenditure	Percentage expenditure on fish to total expenditure
0-26 000	4.46	1 1013.67	7 725.0	18 768.7
		(58.84)	(41.16)	(100)	813.27	7.36	4.33
26 001-39 000	4.29	12 690.55	15 493.7	28 184.2
		(45.03)	(54.97)	(100)	967.99	7.62	3.43
> 39 000	4.96	15 731.50	22 544.9	38 276.59
		(41.10)	(58.9)	(100)	1201.9	7.63	3.14
	4.57	13 155.24	15 254.6	28 409.8
		(46.30)	(53.70)	(100)	992.2	7.54	3.49

SOURCE: C. Sekar et. al., 1996.

Birthal and Singh (1997) estimated the demand for livestock and fish products in Uttar Pradesh using extended linear expenditure system (ELES) based on the primary data collected from 153 rural households of Uttar Pradesh. Annual household incomes of the selected households were as follows: less than Rs 20 000, Rs 20 000-30 000, Rs 30 000-40 000 and above Rs 40 000. The results showed that livestock products such as milk, mutton, eggs and fish together accounted for 18% of total consumption expenditure. The average expenditure share of fish of total food expenditure increased initially with an increase in income, but declined marginally for higher income classes. This phenomenon was more due to socio-cultural factors than to economic ones. One possible explanation is that households in the lowest income group are generally lower in the caste hierarchy and face no social stigma as a result of consuming fish and other meat products. On the other hand, higher income classes generally belong to higher castes and will not consume non-vegetarian foods. At the commodity level, meat, fish, eggs together accounted for 4.83% of total expenditure (Birthal and Singh, 1997). Average expenditure share of fish across three income groups showed increases from 0.29% to 0.33% from the lowest to the highest income class. Income elasticity of the demand for fish was found to be 0.37, irrespective of the income class, whereas it was 2.25 and 0.35 for chicken and eggs respectively. The Indian diet is based on cereals. Historically, animal protein sources such as fish, meat, eggs and milk have played a very small role in the diet of the average Indian. A comparative picture of consumption of cereals, pulses and fish by consumer expenditure groups is presented in Table 19.

Table 19. Per caput quantity of fish consumed per month by mean per caput consumption expenditure (MPCE) (kg)

Monthly per caput expenditure class	Rural			Urban
	Cereals	Pulses	Fish	Cereals	Pulses	Fish
1	9.79	0.42	NA	8.92	0.43	0.11
2	11.32	0.50	NA	10.11	0.58	0.16
3	12.04	0.84	NA	10.61	0.64	0.20
4	12.66	0.67	NA	10.75	0.71	0.25
5	13.22	0.68	NA	10.88	0.78	0.29
6	13.38	0.74	NA	10.99	0.87	0.36
7	13.77	0.79	NA	10.92	0.93	0.43
8	14.14	0.83	NA	10.95	1.00	0.51
9	14.50	0.94	NA	10.74	1.08	0.59
10	14.74	1.01	NA	10.66	1.20	0.70
11	15.11	1.17	NA	10.18	1.28	0.90
12	15.95	1.48	NA	10.28	1.45	1.24
Average	13.45	0.79	NA	10.64	0.92	0.40

NOTE: Per caput consumer expenditure in groups in 1993-94 for rural areas was <120, 120-140, 140-165, 165-190,190-210, 210-235, 235-265, 265-300, 300-355, 355-455, 455-560 and > 560. For urban areas it was < 160, 160-190, 190-230, 230-265, 265-310, 310-355, 355-410, 410-490, 490-605, 605-825, 825-1055 and >1055.

SOURCE: Based on NSS Data.

Proteins normally supply 10-12% of energy in most diets. It is desirable that energy derived from fat should not exceed 30%; the rest may be derived from carbohydrates (60%). The most apparent energy deficiency encountered in India is associated with deficiency in protein intake. Table 20 shows levels of surplus and deficiency from recommended levels of energy intake in rural and urban expenditure classes. The figures show that a substantial proportion of the population, both in rural areas (80%) and urban areas (70%), is not getting the required levels of 2400 and 2100 kcal respectively. Energy intake in 1993-94 is more than 10% below the required level.

Table 20. Per caput intake and surplus/deficiency according to various NSS rounds for all of India.

Period	1972-73	1983	1993-94	1993-94
Rural
Energy( kcal)	2266	2221	2146	2153
Deficiency/surplus from recommended level of 2400 kcal	-134	-179	-254	-247
Protein (in g)	62	62	61	60.2
Surplus deficiency from recommended 60 g	-2	-2	-1	-0.2
Fat (in g)	24	27	31.4	29
Deficiency from recommended 40 g	-16	-13	-8.6	-11
Urban
Energy	2107	2089	2071	1984
Deficiency	-93	-111	-211	-129
Protein (in g)	56	57	57.2	54
Deficiency from recommended 60 g	-4	-3	-2.8	-6
Fat (in g)	36	37	42	37
Deficiency from recommended 40 g	-4	-3	+2	-3

SOURCE: Shariff and Mallic (1999).

	Rural	Urban
Total energy consumption (kcal)	2100	1924
Per caput fish consumption (g)	50	56
Contribution of energy from fish consumption (kcal)	9.7 (0.46)	10.99 (0.57)
Total protein consumption (g)	59	54
Contribution of protein from fish consumption (g)	0.55 (0.93)	0.62 (1.15)

NOTE: Figures in parenthesis indicate percentage to total

Table 22. Estimated per caput expenditure per day (Rs) by states

States/Region	Food Expenditure
	Food grains	Other food	Estimated percentage share of fish in other food expenditure	Total food
North
Haryana	1.96	5.05	0.66	7.01
Himachal Pradesh	3.71	5.12	0.05	8.84
Punjab	2.92	7.75	1.56	0.67
Upper Central
Bihar	3.14	1.90	6.81	5.05
Uttar Pradesh	3.00	3.40	0.80	6.41
Lower Central
Madhya Pradesh	2.39	1.88	2.50	4.28
Orissa	2.98	1.85	10.14	4.83
Rajasthan	2.68	5.42	0.18	8.11
East
NE states	3.40	2.81	24.52	6.22
West Bengal	3.70	3.00	23.94	6.71
West
Gujarat	2.30	4.07	0.84	6.41
Maharastra	2.14	2.74	6.05	4.89
South
Andhra Pradesh	3.04	3.12	1.99	6.17
Karnataka	2.89	2.94	2.31	5.84
Kerala	2.50	4.41	21.16	6.92
Tamil Nadu	3.32	3.3	3.76	6.70
All India	2.91	3.19	5.21	6.10

SOURCE: India Human Development Report, 1999.

Some authors (Bhalla et al., 1999) and the National Sample Survey (NSS) suggest that there has been a long decline in per caput cereal consumption since the early 1970s. The movement away from eating cereals and pulses in favour of meat, eggs and fish is a particular feature of the recent past. Table 23 shows markedly higher levels of per caput consumption in the food group of meat, eggs and fish. This is a composite category and fish constitutes about half of it. NSS data indicate very low levels of consumption of this food group, particularly in the north and west. But levels of consumption are appreciably higher in the south and east, which, of course, partly reflects the consumption of fish in Kerala and West Bengal.

Table 23. Levels, trend and projections of consumption of meat, eggs and fish

Region	Monthly per caput consumption (kg) 1993-94		Average growth rates of per caput consumption (%)		Projected monthly consumption demand (kg) by 2020		Projected annual consumption ('00 t) by 2020
	Rural	Urban	Rural	Urban	Rural	Urban
South	0.58	0.57	3.77	3.73	1.58	1.80	3875
West	0.15	0.33	0.95	1.09	0.19	0.44	697
North	0.14	0.24	1.29	-0.62	0.20	0.24	1082
East	0.44	0.77	1.64	2.09	0.68	1.35	4097
All India	0.33	0.47	2.68	1.20	0.55	0.85	10451

SOURCE: Dyson and Hanchate (2000).

Table 24. Definition of income groups (based on monthly expenditure)

Income groups	Average household size	Rural		Urban		Average expenditure (Rs)
		Sample size	Expenditure (Rs)	Sample size	Expenditure (Rs)
Very poor	6.21	159	< 359.62	335	< 535.43	348.00
Poor	5.42	158	359.621-636.2	335	535.43-762.71	606.00
Medium	4.48	159	636.2-1036.51	335	762.71-1205.33	899.00
Rich	3.90	158	> 1036.50	334	>1205.33	3358.00

Table 25 shows per caput total fish consumption by species per month across different income groups, ranging from very poor to rich. Among the freshwater fishes, rohu is consumed most frequently, followed by catla, mrigal and other marine fishes. Monthly per caput fish consumption is 0.93 kg, which is substantially higher than per caput fish consumption suggested by NSS rounds. Annual per caput consumption is around 15 kg, which is two to three times higher than earlier estimates supporting observations of changing food preference and consumption patterns. The study also strengthens the research findings of Kumar (1998), which foresee a continued shift of demand towards non-cereal foods.

Table 25. Total per caput fish consumption (kg/month) by species for all of India

Species	INDIA (1973)
	Very poor	Poor	Medium	Rich	Total	Annual
Rohu	0.27	0.36	0.55	0.84	0.50	6.00
Catla	0.19	0.27	0.31	0.31	0.27	3.24
Mrigal	0.07	0.07	0.06	0.04	0.06	0.72
Common carp	0.002	0.32	0.61	0.05	0.03	0.36
Other carps	0.07	0.04	0.03	0.02	0.05	0.60
Total carps	0.61	0.77	1.01	1.25	0.91	10.92
Other marine fishes	0.23	0.19	0.13	0.79	0.33	3.96
Total	0.84	0.96	1.14	2.04	1.24	14.88

Per caput fish expenditure by species and across income groups per month in Rs is shown in Table 26. As income increases, so does the money spent for all species, except for mrigal and other minor carps. Annual per caput expenditure on fish consumed is Rs 645.00 (US$ 14.02), which is somewhat higher than all previous NSS figures.

Table 26 shows annual fish consumption by income classes and by individual species. Per caput consumption by the rich income class is 24.48 kg per year, which is 10 kg above the average. Annual fish expenditure (in US$) shows a similar trend. In terms of fish consumption by species, nearly 50% of total fish consumed by the rich class comprised marine fishes and freshwater species, except carps. For carps, 30-40% of total fish expenditure goes to rohu, followed by catla (19.4%).

Table 26. Annual fish expenditure by income classes and by individual species

Item		Rich	Medium	Poor	Very poor	Average
Annual per caput consumption		24.48	13.8	11.52	10.08	15
Annual fish expenditure (US$)		24.48	12	10.2	9.12	14.04
Average price (US$)		1	0.87	0.89	0.9
Species (% of fish expenditure)
	Rohu	34.7	41.4	35.2	35.1	36.3
	Catla	14.3	23.7	24.9	20.7	19.4
	Common carp	0.7	3.8	2.3	0.2	1.6
	Other carps	1.4	2.9	4.3	9.9	3.5
	Other marine fishes	47	24.4	27.1	24.8	34.9

The total food, fish, protein expenditure per caput per month by income groups is presented in Table 27. Total per caput monthly expenditure by the rich group is four to five times higher than that of the lower income group. However, the percentage of fish to total food expenditure decreases as income increases, illustrating Engel's law. The percentage of fish to protein expenditure also decreases as income increases. This indicates that in terms of equity, higher fish production and improved accessibility to fish will benefit the poor more rather than rich. The percentage of fish to total animal protein expenditure reveals that people in poor income groups receive proportionally more protein from fish than do their counterparts in richer groups.

Table 27. Total food, fish protein expenditure and percentage by income groups for all of India

Variables	N = 1973
	Very poor	Poor	Medium	Rich	Overall
Total per caput monthly expenditure (Rs)	338.50	609.33	902.15	1644. 77	873.68
Percentage of fish to total food expenditure	0.15	0.10	0.096	0.12	0.12
Total per caput fish expenditure (Rs)	35.05	40.00	45.95	93.94	53.72
Total per caput food expenditure per month (Rs)	255.00	378.00	460.42	745.97	459.42
Percentage of fish to total protein expenditure	0.60	0.40	0.30	0.34	0.42
Total protein expenditure (Rs)	66.00	110.65	155.95	306.48	160.75
Percentage of fish to total animal protein expenditure	0.72	0.56	0.45	0.57	0.58
Total animal protein expenditure	48.00	68.70	89.98	150.20	88.92

Consumer preferences and product-value depend upon the properties or characteristics of the fish. The question we must ask is this: To what degree does a particular fish product supply preferred characteristics that are desirable to the buyer? The fish is likely to have a number of combinations of these characteristics, such as taste, texture, colour, shape and nutritional quality.

In order to understand such preferred characteristics for various freshwater fishes, the consumer survey contained inquiries into most preferred and least preferred species, preferred shape, size (number of fish per kg), preferred body parts (eggs, back meat, tail meat, belly meat, etc.). Such information will be helpful in developing fish products with those kinds of characteristics in the future in order to realize higher prices in the market. Table 28 shows the preferred shape for carp species among rural and urban consumers. The results indicate that consumers prefer mostly short and thick catla (43%) and common carp (54%), whereas long and thin rohu, mrigal and grass carps are preferred by 71%, 75% and 97% of consumers respectively.

Table 28. Preferred shape of fish

Species	Long and thin	Short and thick	Short and deep	Others
Catla	9.34	43.44	34.92	12.30
Rohu	71.36	7.69	9.82	68
Mrigal	74.95	12.42	8.50	4.14
Common carp	11.04	54.35	27.59	7.02
Grass carp	97.60	1.80	0.60	-
Others	88.89	3.70	7.41

The size of fish, expressed as preferred number per kg, is an important attribute. For most freshwater fishes, consumers prefer one or less per kg. Table 29 shows that 80-90% of consumers preferred catla, rohu and grass carp of this size. Mrigal and common carps were favoured in numbers 1-2 per kg.

Table 29. Preferred size (number of fish/kg)

Species	No. of fish/kg
	1	2	3	4	5	6
Catla	81.22	17.26	1.52	-	-	-
Rohu	68.10	25.20	3.83	2.07	0.48	0.32
Mrigal	44.12	39.29	6.51	3.15	0.42	0.21
Common carp	63.20	30.20	4.62	1.65	0.33	-
Grass carp	93.41	6.59	-	-	-	-
Others	59.09	13.64	18.18	9.09	-	-

Table 30 shows the preferred body parts for different fishes, in terms of weighted scores with 1 indicating high preference and 10 lowest of preferences. Back meat and belly meat in catla and rohu are highly preferred. Consumers show relatively little preference for fish eggs of any species.

Table 30. Weighted average score for preferred body parts.

Species/trait	Eggs	Back meat	Tail meat	Belly meat
Rohu	4.67	2.60	3.53	3.23
Catla	4.54	2.92	2.93	2.35
Mrigal	4.83	4.64	4.84	4.49
Common carp	6.11	5.43	5.95	4.86
Other carps	2.50	4.32	4.18	3.54

3.6 Demand projections

Given the income elastic demands for milk, fish, eggs and meat and the rise in per caput income, demand will continue to grow rapidly in the years ahead. Several studies have indicated that total cereal demand in India will exceed 200 million t by the year 2000 (Sarma and Gandhi, 1990). When we consider the total demand for food grain of 170 million t in 1995 and a slowdown in the growth of the population along with a decline in per caput cereal consumption, it now appears that demand projections are on the high side. A longer-term outlook for 2020 would be of interest to see whether different demand patterns might lead to significant changes from the past. Any projection should consider changes in population growth, growth in per caput income and urbanization. India's population was 846 million in 1991 and reached 1 billion in March 2000. Population growth-rate slowed to 2.1% per annum in the 1980s and to 1.9% in the 1990s. Most experts expect the rate of population-growth to decline further in the next century and to reach 1.3 billion by 2020. India is also becoming more urbanized. In 1991, 26% of the population lived in urban areas. Urbanization is projected to increase to 42% by 2020 (Natarajan, 1993). The rural population is expected to increase by only 19%, whereas urban population is projected to grow by 153% to 550 million people.

Expenditure elasticities are less elastic in urban areas than in rural areas. National estimates for livestock products including meat, fish and eggs, which are mostly elastic, average 1.04 and 0.75 for rural and urban groups respectively (Bhalla and Hazell, 1998). Urban consumers have consistently lower elasticities for all food groups than their rural counterparts. Hence, given similar growth rates in per caput income, increased urbanization is expected to slow the growth in food demand. However, if income rises faster in urban than in rural areas, this may more than offset the effects of lower urban expenditure elasticities for foods. Many studies have been conducted on patterns of consumer expenditure in India. Table 31 presents elasticity expenditures from different studies.

Table 31. Alternative estimates of expenditure elasticities in India

Items	Radhakrishna and Ravi		Kumar and Rosengrant		Bhalla
	Rural	Urban	Rural	Urban	Rural	Urban
Rice	0.40	0.46	0.01	0.06	-	-
Wheat	0.38	0.44	-0.09	-0.07	-	NA
Other cereals	-0.01	0.03	-0.13	-0.18	NA	NA
Milk &milk products	1.09	1.15	0.37	0.46	1.01	1.47
Edible oils	0.70	0.76	0.23	0.39	0.63	0.77
Sugar	0.88	0.78	0.06	0.14	0.55	0.95
Fish, meat and eggs	0.89	0.93	0.63	0.85	0.75	1.04
Pulses	0.64	0.71	0.22	0.31	0.47	0.62
Fruits & vegetables	0.81	0.81	0.25	0.39	0.60	0.60
Other foods	0.95	0.87	0.70	0.94	-	-
Non foods	1.52	1.61	1.87	2.25	0.60	1.81

SOURCE: Bhalla and Hazell (1998)

Table 32. Consumer demand projections, 1990 to 2020 under alternative scenarios, India (million t)

Item	Baseline Trend to Continue			Poverty Removed			Well-Fed India
	1990 actual	2020 with 3% PCY	2020 with 5.5% PCY	1990 actual	2020 with 3% PCY	2020 with 5.5% PCY	1990 actual	2020 with 3% PCY	2020 with 5.5% PCY
Milk	43.99	203.57	496.28	51.60	237.17	581.91	83.69	379.88	942.98
Eggs	1.13	4.23	7.68	1.28	4.73	8.63	2.33	8.02	15.10
Beef	0.59	2.14	3.95	0.63	2.24	4.16	1.18	3.95	7.53
Pig meat	0.27	0.97	1.79	0.29	1.02	1.89	0.54	1.80	3.44
Poultry meat	0.41	1.49	2.75	0.44	1.56	2.90	0.80	2.68	5.11
Sheep meat	0.22	0.81	1.50	0.24	0.85	1.57	0.42	1.41	2.68
Total meat	1.50	5.42	9.99	1.59	5.67	10.52	2.94	9.84	18.77
Fish	1.83	6.15	11.70	2.13	7.05	13.50	3.56	11.78	22.58
Cereals	144.62	267.83	327.32	151.98	280.41	342.94	138.21	259.81	316.50
Pulses	10.40	25.77	38.19	9.88	24.52	36.33	13.11	32.19	47.81
Oils	4.15	12.46	20.23	4.57	13.61	22.13	5.88	16.79	27.50
Sugar	10.15	30.65	53.86	11.54	34.69	61.17	15.17	45.17	80.17
Roots and tubers	17.17	43.31	65.75	13.87	34.64	52.60	17.42	30.17	33.81

NOTE: PCY = per caput income growth

Expenditure elasticities are used to project annual per caput consumption. We observed that under alternative growth scenarios of the per caput GDP, the magnitude of expenditure elasticities is likely to decline in the future. In the long run, and with growing income, income elasticity of demand for food in general will converge to zero; income effect on per caput consumption will be negligible. However, studies have shown that the income effect on consumption of fish, meat and eggs is positive and significant. Further, increasing trends in consumption towards high-value commodities will generate growth in the demand for livestock, fisheries and horticultural products.

Income elasticity of demand for the fish, meat and eggs group is 0.892. Food demand for human consumption is generally estimated by multiplying projected per caput consumption with projected population (Kumar, 1998). Household demand for fish is projected to grow annually three to four times and will reach a maximum level of 14-18 million t (Kumar, 1998). The percentage share of demand for fish of total world demand for food is highest in India (68%), followed by Bangladesh (18.9%) in South Asia. Annual growth rate of demand for fish is estimated to be around 3% with per caput GDP growing annually by 3.5%. Annual per caput food consumption (with low income growth) is projected to increase to 5 kg by 2015 and 7 kg by 2030, and with high growth of income it could increase to as much as 9.2 kg per caput/year. It is estimated that by 2015, India will need around 12.61 million t of food under a high-income scenario, and 8.46 million t, if income grows slowly.

3.7 Conclusion

The analysis in this section shows that not only do people who live below the official poverty line suffer from malnutrition and insufficient energy intake, but that the majority of the population above the poverty line also lacks adequate nutrition. Food expenditure as a proportion of total expenditure seems to decrease with an increase in per caput income across the states. Income elasticity of demand, as reported by a few micro-level studies, is positive and elastic for fish in urban areas. However, further studies are required to establish the relationship. A general lack of protein, fats, etc. leads to nutritional deficiency. Fish contributes only 1.15% and 0.93% of total protein intake in urban and rural areas respectively. However, actual protein intake is likely to be higher than these levels, as some micro-level studies suggest. Demand projections under alternative scenarios of income-growth, based on expenditure elasticity, have shown that the demand for fish is likely to far exceed the projected growth in production.

Examination of the results of some of the recent studies from six major Indian states on consumption expenditure reveals that, with the exception of a few regions, per caput fish consumption is generally much higher than results of earlier NSS rounds indicate. One of the reasons for this positive, higher trend in fish consumption is an overall change in food habits, urbanization and rising income levels. Consumption patterns by species also show that more than 50% of households now include freshwater fishes in their diets. The results clearly link fish consumption with income. As income rises, total expenditure on fish also increases, yet the proportion of total expenditure decreases. In India, rohu is most preferred by consumers, followed by catla and other carps among the freshwater fishes.